Drainage, pore water pressures, and; slope stability
in London clay at Stag Hill, Guildford.
by
Jamshid Sadrekarimi
M.Sc.
A thesis submitted to the University of Surrey for the Degree of Doctor of Philosophy in the Department of Civil Engineering, Geotechnical Division.
September 1988. ProQuest Number: 10804448
All rights reserved
INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a com plete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion. uest
ProQuest 10804448
Published by ProQuest LLC(2018). Copyright of the Dissertation is held by the Author.
All rights reserved. This work is protected against unauthorized copying under Title 17, United States C ode Microform Edition © ProQuest LLC.
ProQuest LLC. 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml 48106- 1346 Sum m ary
Stag Hill is an example of a slope failure in London Clay. In 1964 it was proposed to develop the area, so the slope was stabilized using a drainage system and heavy buildings were constructed on piled foundations. Piezometers were also installed to establish the pore pressure pattern.
In 1985 a general evaluation of the history of the failure, stabilization and development of the area was undertaken. This gave rise to the view that this area would be a very suitable place to establish a system for studying the efficiency of drainage systems. The presence of the drainage system, which initiates rapid fluctuation of ground water pressure in response to infiltration, brought about the necessity of attention to the errors associated with time lag which would be involved in monitoring of pore water pressure.
Later on, when the ground water pressure was being monitored, it was observed that changes of atmospheric pressure significantly affects the records of vibrating wire piezometers. In this way the necessity of the evaluation of difficulties and errors associated with atmospheric pressure in monitoring of ground water pressure automatically merged into the research.
Accordingly, to fulfill the above mentioned aspects, all available piezometer records since 1964 were collected and plotted. All piezometers which survived were tested to see if they were still reliable. Drainage trenches were located at appropriate locations and over 60 new piezometers of different types were installed at different depths, in drained and non-drained areas, and in the trenches. Two types of pore water pressure measurements, daily and weekly, were taken. To record the duration and intensity of daily rainfall, an automatic tilting bucket system was installed in the area. To include the effect of the atmospheric pressure on the evaluation of the pore water pressure pattern, records of atmospheric pressure, taken at Gatwick airport were used.
The main conclusions are that:
Monitoring of ground water pressure, using stand pipe piezometers, in drained areas may lead to serious errors, say 1.0 meter head of water. Application of rigid piezometers is necessary but not enough, unless a reasonable correlation is established between piezometer reading times and rainfall periods. Changes of atmospheric pressure are a significant source of error in monitoring of ground water pressure techniques. A method to correct piezometer readings for changes in atmospheric pressure was established.
At Stag Hill, the ground water pressure pattern, and in turn the stability of slopes is governed by layers of higher permeability in the range of 10-5- 10~6 cm/sec.
The main drainage system having a spacing to depth ratio of S/D = 3.4 - 3.8 has not efficiently improved the stability of the slopes. Surface covering (buildings and paving), however, has a dominant effect on the improvement of safety factor. At Wates House, the drainage system having S/D =1.1, has lowered the ground water pressure with 92% efficiency.
It takes some years before the ultimate efficiency of a drainage system takes effect. The efficiency of drainage trenches below invert level varies between the maximum and minimum efficiency of drains at invert level, in the long term.
The theoretical methods for design of drainage trenches do not agree with practice. A practical design curve was suggested. Acknowledgements
I am very grateful to my Supervisor, Professor N. E. Simons, for many helpful discussions and suggestions during my period of research. I also wish to express my sincere thanks for his kindness and pastoral care during my illness.
I should like to thank Mr. M.A. Huxley for various discussions and linguistic advice.
I am grateful to Professor A.D.M. Penman, Dr. C.R.I. Clayton, and Dr. E.N. Bromhead for the technical and academic assistance they gave.
I acknowledge the help of many other people and organisations who contributed to the work described in this thesis. These include:
Mrs M. Wicks for her help with organising and providing instrumentation facilities.
Mr. I.D.Rose and Mr. T.J. Webb, from Thames Water, for the provision of an automatic rain-gauge.
Gatwick Airport Authority for the provision of atmospheric pressure records.
Members of the technical staff of the Soil Mechanics and Heavy Structures Laboratories of the University of Surrey.
Mrs M A Simms of the Works Department, University of Surrey.
Mrs M. Janes for typing this thesis.
This research was carried out with financial support provided by the University of Tabriz - Iran.
Finally, my thanks are due to my wife and children for their patient suffering over three years, virtually without husband and father. Contents Page No.
Sum m ary. I Acknowledgements HI Contents. IV Symbols. XI Introduction. XV
CHAPTER ONE - LANDSLIDES AND CORRECTIVE MEASURES. 1
1-1 Introduction. 1 1-2 Field Investigation. 3 1-2-1 Topography. 3 1-2-2 Geology. 4 1-2-3 Monitoring of the ground water pressure. 5 1-2-4 Weather. 5 1-2-5 Shape of sliding surface. 6 1-3 Preliminary planning. 7 1-4 Corrective and remedial methods, Treatment of a slope shape and drainage. 8 1-4-1 Introduction. 8 1-4-2 Treatment of slope shape (removal of weight). 8 1-4-3 Drainage. 11 1-4-3-1 Surface drainage. 11 1-4-3-2 Subsurface drainage. 12 a - Trench and counterfort drains; definition and Literature review. 13 b-Blankets. 18 c-Galleries. 19 d - Horizontal drains. 19 e - Well systems (vertical drains). 20 1-5 Safety Factor, Literature review. 21
CHAPTER TWO - ERRORS IN FIELD MEASUREMENTS OF GROUND WATER PRESSURE. 33
2-1 General definitions and principles of pore water pressure measuring systems. 33 Page No.
2-1-1 Types and principles of operation. 33 - open stand pipe piezometers. 33 - closed system hydraulic piezometers. 33 - diaphragm type piezometers. 34 2-1-2 Porous element in piezometers. 36 2-2 Sources of errors in ground water pressure measurements. 37 2-2-1 Introduction. 37 2-2-2 Hydrostatic time lag. 38 2-2-3 Stress adjustment time lag. 39 2-2-4 Theoretical and experimental studies of time lag, Literature review. 41 2-2-5 Effects of changes of atmospheric pressure; Literature review. 48 2-2-6 Other sources of error. 50 - General instrumental errors. 50 - Seepage along conduits. 52 - Effects of gas bubbles in the system and/or soil. 52
CHAPTER THREE - STAG HILL. 60
3-1 Introduction. 60 3-2 Topography. 60 3-3 Geology of the area. 61 3-4 History of the site investigations and remedial works. 61 3-5 History of piezometer installations and pore water pressure records. 66 3-6 Engineering properties of the soil. 68 3-6-1 London clay, general definitions. 68 3-6-2 Mechanism of failure. 69 3-6-3 Strength parameters. 70 3-6-3-1 Peak strength parameters. 70 3-6-3-2 Strength parameters relevant to first-time slide (fully softened). 71 3-6-3-3 Residual strength parameters. 71 Page No.
3-6-4 Index properties. 73
CHAPTER FOUR - DETAILS OF THE MEASURING SYSTEMS, AND LABORATORY WORK. 90
4-1 Standpipe piezometers. 90 4-1-1 Details of piezometers. 90 4-1-2 Preparation for installation. 91 4-1-3 The Readout Unit. 91 4-2 Pneumatic piezometers. 91 4-2-1 Details of piezometer. 91 4-2-2 Preparation for calibration and installation. 92 4-2-3 The Readout Unit. 94 4-3 Vibrating wire piezometers. 94 4-3-1 Details of the piezometer. 94 4-3-2 Preparation for calibration and installation. 95 4-3-3 The Readout Unit. 96 4-4 Automatic tilting bucket rain gauge. 97 4-5 The sealing efficiency of bentonite pellets. 97
CHAPTER FIVE - SITE EXPERIMENTAL WORK. 110
5-1 Introduction. 110 5-2 Inspection and service of previously installed piezometers. 112 5-2-1 Inspection against penetration of surface water into the piezometers - Improvement of commercial caps. 112 5-2-2 Inspection against blocking, cleaning of piezometers. 112 5-2-3 Equalization tests, as a criterion for reliability of piezometers. 112 5-3 Locating of drainage trenches. 113 5-4 Installation of new piezometers. 114 5-4-1 Reasons and justifications for the new installations. 114 5-4-1-1 Introduction. 114 5-4-1-2 Installation along profile 5-5. 116 5-4-1-3 Installations along profile A-A. 117 Page No.
5-4-1-4 Installations along profile 6-6. 119 5-4-1-5 other piezometer installations. 122 - zone 2. 122 - zone 8 . 122 - zone 9. 123 -piezometer J50. 124 5-4-2 General procedure of piezometer installation. 124 5-4-2-1 Drilling. 124 5-4-2-2 Installation and sealing in clay. 125 5-4-2-3 Installation in the trenches. 127 5-4-3 Borehole logs and special cases encountered during borehole drilling and piezometer installations. 128
5-5 In-situ evaluation of the response time of standpipe and pneumatic installations. 143 5-6 Installation of the Automatic tiling bucket rain gauge. 144 5-7 In situ tests. 145 5-7-1 Equalization tests. 145 5-7-1-1 Equalization tests Series ”A”. 145 5-7-1-2 Equalization tests series ”B”. 146 5-7-2 Daily measurements of pore water pressure. 148 5-7-3 Daily measurements of the amount and duration of rainfall. 150 5-7-4 Weekly measurements of pore water pressure. 150
CHAPTER SIX - RESULTS. 171
6-1 Summary of characteristics of the piezometer installations. 171 6-2 Results of equalization tests and permeability analysis. 177 6-2-1 Results of equalization tests series ”A”. 177 6-2-2 Results of equalization tests series ”B”. 178 6-2-3 Results of permeability measurements and calculations. 178 6-2-3-1 Records of permeability tests carried out in the past (1966). 178 Page No.
6-2-3-2 Results of permeability measurements carried out during the research. 178 6-3 Records of daily measurements of ground water and atmospheric pressures. 181 6-4 Records of daily measurements of the amount and duration o f rainfall. 182 6-5 Piezometer records, since 1965. 182
CHAPTER SEVEN - STABILITY ANALYSES OF SLOPES. 241
7-1 Introduction. 241 7-2 Stability of slope along profile 6-6 . 242 7-2-1 Stability at the failure time. 242 7-2-2 Stability after failure and before drainage and development of the area. 243 7-2-3 Stability after failure and change of the shape of slope, 1988. 244 7-3 Stability of slope along profile A-A. 245 7-3-1 Stability at the failure time. 245 7-3-2 Stability after failure and before drainage, 1965. 246 7-3-3 Stability after drainage and development area, 1988. 246
CHAPTER EIGHT - DISCUSSIONS AND CONCLUSIONS. 251
8-1 Assessment of the equalization tests. 251 8-1-1 Evaluation of the first time equalizations. 251 8-1-2 Evaluation of the subsequent time equalizations of piezometer J37, (tests Series "A”). 253 8 -1-3 Evaluation of the equalization curves Series ”B”. 255 8-1-4 Conclusions. 259 8-2 Assessment of the difficulties and errors in measurement of ground water pressure. 260 8-2-1 Instrumental errors. 260 8-2-2 Errors and difficulties associated with atmospheric pressure. 262 8-2-2-1 Introduction. 262 Page No.
8-2-2-2 The efficiency with which atmospheric pressure affects ground water pressure. 262
8-2-2-3 Errors in measurement of the ground water pressure associated with atmospheric pressure. 267 I - vibrating wire piezometers. 267 II- pneumatic piezometers. 272 HI - standpipe piezometers. 274 8-2-3 Errors associated with time lag. 274 8-2-4 Conclusions. 280 8-2-4-1 Errors associated with atmospheric pressure. 280 8-2-4-2 Errors associated with time lag. 281 8-3 Assessment of the ground water pressure pattern. 283 8-3-1 Summary of borehole logs, and ground water pressure profiles. 283 8-3-2 Discussion. 285 8-3-3 Conclusions. 290 8-4 Evaluation of the stability of slope. 291 8-4-1 Discussion. 291 8-4-2 Conclusions. 293 8-5 Evaluation of the mechanism and efficiency of the drainage system in lowering of the ground water pressure. 295 8-5-1 Introduction. 295 8-5-2 Drainage system phase I, part 1. 295 8-5-3 Drainage system phase I, part 2. 298 8-5-4 Drainage system phase V, (Wates House). 300 8-5-5 General demonstration of the mechanism of the process, through which the efficiency of the drainage system takes effect. 301 8-5-6 A comparison of theoretical solutions for trench and counterfort drains with the field data. 304 - Comparison with Hutchinson’s Method. 306 - Comparison with Bromhead’s solution. 308 - Comparison with Stanic’s solution. 308 8-5-7 Conclusions. 308 Page No.
References 327
A ppendix 334 Records of piezometers A, B, C & D. 335 List of Symbols
The symbols used in this text have been mostly chosen in accordance with current conventional usage; and are listed as below:
A cross directional area, pore pressure coefficient, trenches spacing, a effective radius of a piezometer, equivalent intake radius of a piezometer. B pore pressure coefficient, width of a trench, b width of a slice, c coefficient of consolidation, c' cohesion in terms of effective stress. c's fully softened cohesion. c'r residual cohesion. D trench depth (Hutchinson’s method), d depth of impermeable layer. e error in measurement of ground water pressure. Natural log. base, ei & e2 errors corresponding to minimum and maximum values of atmospheric pressure, assuming Si = S]. e' i & e'2 errors corresponding to minimum and maximum values of atmospheric pressure, assuming Si = 100%. F shape factor of a piezometer intake. f frequency fs shape factor which correlates curved phreatic surface to an average plane surface. g acceleration due to gravity. G Ground level. H trench depth (Bromhead & Stanic’s methods), h Pore water head on invert level at any point between trenches, ho trench invert depth from ground water table. hav average piezometric level on drains invert level considering piezometric levels in trenches. h average piezometric level between two adjacent trenches on drain invert level. hd mean piezometric level on drainage invert. hm maximum piezometric level on drainage invert level. nw average ground water pressure head on trenches invert level. i hydraulic gradient. J prefix for new piezometers. JT prefix for a new piezometer installed in a trench. K coefficient of permeability. k vibrating wire piezometer constant. kh horizontal coefficient of permeability. kv vertical coefficient of permeability. L length II Liquidity index. LL Liquid limit. 1 vib. wire length. In length of base of slice N o.n. m or mv Coefficient of volume compressibility. mb millibar(s). m b.g.l. meter(s) below ground level. No pre-installation base reading of a vib .wire piezometer. Ni after installation reading of a vib.wire piezometer. n slice no. P pressure
P a o atmospheric pressure when the pre-installation base reading (No) is recorded. Pal atmospheric pressure when the after-installation reading (Ni)is recorded, Pa min the average minimum atmospheric pressure of the area under investigation. Pal max & Pai min maximum and minimum amounts of atmospheric pressure during a certain month, PI plasticity index. PL plastic lim it.
Pm maximum piezometric level between two trenches. Pp average piezometric level. Rk permeability ratio, kh/kv . r radius, average rate of change of piezometric level.
tm normalized maximal piezometric level at the slip surface. rp normalized average piezometric level at the slip surface . ru pore pressure ratio. S trenches spacing (apparent). S' trench spacing (real). Sr degree of saturation. T basic time lag. time factor. Tw period of ground water pressure fluctuation. t time t5o,90 time lags for 50 & 90% equalizations, ts phase shift. u pore water pressure in saturated soil. ua pore air pressure. uw pore water pressure. Uq pore water pressure at t = o Uoo pore water pressure at t = oo ut pore water pressure at t = t W weight X a parameter depending on the degree of saturation. Xa amplitude of fluctuation of water pressure as measured with a piezometer, in incompressible soils. yt the unbalanced pressure at the time t = t after beginning of an equalization test. yQ the unbalanced pressure at the beginning of an equalization test. Za amplitude of the ground water pressure fluctuation, in incompressible soils. a angle. rate of change of ground water pressure. P angle Yw unit weight of water. A finite difference. APa Pal max - Pai min AP change of a piezometer reading due to a change of atmospheric pressure. 50 efficiency of transmission of major constant part of atmospheric pressure on the ground water pressure at any depth below ground level. 51 efficiency of transmission of a change of atmospheric pressure on the ground water pressure at any depth below ground level (si = AP/APa). t\ efficiency of trench drains on a given cross section with respect to the intervening mass of soil only, riav overall efficiency of trench drains, on any given cross section considering piezometric levels in trenches also. jj stiffness of a pore water pressure measuring system. V volume factor of a piezometer p density of a vibrating wire I sum a total normal stress, tension in a vibrating wire a ' effective normal stress principal total stresses
O'hO'3 principal effective stresses T shear stress 0 angle of shearing resistance in terms of total stress o' angle of shearing resistance in terms of effective, stress 0 r residual angle of shearing resistance
0 S first-time slide (fully softened) angle of shearing resistance Introduction.
Stag Hill is an example of a slope failure in London clay. The fact that such a suitable site so close to a tourist centre like Guildford remained undeveloped until 1964 is due entirely to a geotechnical reason, namely, that a landslip had taken place here on the slopes of Stag Hill (Simons, 1977). In connection with the age of this landslip, estimations are available in literature. Certainly it must have occurred more than 130 years ago, since a tree 110 years old was cut down on the back scarp in 1966. The scarp, however, is still a clearly expressed feature which is unlikely to be more than a few centuries old, and there are signs of fresh movement at the toe (Skempton & Petley, 1967). Probably a landslip took place here some time during the 19th century (Simons, 1977).
After the proposal to construct a new University at Guildford, Surrey, the first subsoil investigation, consisting of seventeen borings and associated soil tests was carried out between 9th May and 7th July 1964.
During this period, a preliminary inspection of the site in June 1964 revealed that a major slip had occurred there in the past. Evidence of a slip was particularly clear in the field to the North-East of Guildford Cathedral, where a number of hawthorn trees had become established along the line of the slip scarp and along the line of overthrusting at the toe.
In November 1964, Professor R E. Gibson evaluated the stability of the area with respect to the results of the above mentioned investigation and proposed remedial proposals. Professor Gibson suggested a drainage system with trenches 5m deep, 0.75m wide, and spaced at about 12m centres for the area where slipping is extensive. To the West where the slopes are slightly flatter, trenches of 3.6m deep, 0.75m wide and spaced at about 18m centres were suggested. In order that the present slope should remain stable it was proposed that the weight of the heavy buildings should be carried by bored reinforced concrete piling, isolated from the top 3m of soil by the provision of a sleeve, down to an adequate depth (Gibson, 1964).
In 1965 a further site investigation was carried out on the site of the proposed University. The objective of this investigation was to provide further information on the ground conditions at the site that would assist the engineer in his design of the foundations. Further investigations were carried out in 1965-66 (Skempton & Petley, 1967) and the slip surface was located through a section. The main slip surface, which was highly polished, could be seen in three exploration pits.
The final conclusions at that time, September 1965, were that drains 5m in depth spaced at 15m centres should be installed over the slipped area.
At the present time the drainage system of this area has been constructed in five phases.
Since 1964, a short period before construction of a drainage system and development of the area, piezometers have been installed in Stag Hill.
This research was intended to evaluate the stability conditions of Stag Hill and the efficiency of the drainage trenches in lowering of the ground water pressure. The strong possibility of the involvement of errors in monitoring of the ground water pressure in this area led to the evaluation of the range of errors in monitoring of the ground water pressure. Following the appearance of very special and complicated pore pressure records, which is not usual in London clay, it was decided to study the ground water pressure pattern as another main branch of the research.
The present research consists of eight chapters.
In Chapter one, the importance of the correction of landslides and sequence of investigation is briefly explained. Then corrective and remedial methods with emphasis on the trench and counterfort drains, are reviewed. A summary of proposed permissible safety factors, suggested by different sources, is included.
In Chapter two, the general principles of pore pressure measuring systems and sources of errors in measurements of ground water pressure are studied. Concepts of the hydrostatic and stress adjustment time lags are explained and then previous theoretical and experimental studies are reviewed. Previous work in connection with the effects of changes of atmospheric pressure on the ground water pressure is summarised. Other sources of errors, relevant to this research, are then briefly explained. Chapter three consists of the history of the site investigations, remedial works, piezometer installation and development of the area since 1964. London clay is defined and the engineering properties of soil are given.
In Chapter four, the measuring systems, which were used during the research, are described. The sealing efficiency of the bentonite pellets was also investigated in the laboratory. The results are discussed and conclusions made.
Chapter five discusses the site experimental work. Sequences of inspection and cleaning of previously installed piezometers, location of drainage trenches, borehole drilling, piezometer installation and sealing, installation of drive-in piezometers, and the installation of the Automatic Tilting Bucket rain gauge are extensively explained. Borehole logs and special cases encountered during the borehole drilling and piezometer installations, which would help to visualize the stratography and the ground water pressure pattern conditions, are also given.
In situ tests, consisting of equalization tests, measurements of pore water pressure, and rainfall, are described.
Two series of equalization tests were carried out:
Tests Series ”A”, to evaluate the response characteristics of pneumatic piezometers to changes of ground water pressure, and effects of installation time lag on the first time equalization of piezometers
Tests Series ”B”, to evaluate in situ permeability of soil.
In Chapter six, the characteristics of all piezometer installations, results of equalization tests, in situ permeability measurements, records of piezometer installations from 1965 to 1988, are given. Records of daily measurements of appropriate piezometers and the rainfall from May to December, 1987 are also included in this chapter.
In Chapter seven, the stability of slope along two longitudinal profiles is analysed at different stages, and the results are summarized in a table.
In Chapter eight, the results and data, given in chapters six and seven are discussed in five sections and conclusions are given at the end of each section. CHAPTER ONE - LANDSLIDES AND CORRECTIVE MEASURES
1.1 INTRODUCTION
Movements of sliding earth masses, separated from the underlying stationary part of a slope by a definite failure are called ’’slope movements” and in the case of natural slopes are designated as ’landslides’. The sudden transfer of potential eqergy into kinetic energy when a slope fails can be very catastrophic and destructive, in other words, the uncontrolled releases of energy during the failure of a slope can cause tremendous physical catastrophies that take large numbers of lives and cause heavy property damage. According to Cedergren(1977) a single large landslide of modem times consumed energy at a rate of at least 20 billion horsepower. Normally, however, slow moving slope failures usually do not take lives but add substantially to the total economic loss by the gradual destruction of railroads, highways and other structures.
Stable earth slopes, both natural and man-made, are of great importance to mankind, so the permanent control and maintenance of the slopes in many cases is necessary, although it may be expensive. In a discussion of economic losses due to landslides, Smith (1958) declared that the direct cost of landslides to United States railroads alone was more than $5,000,000 each year. He estimates that the average annual cost of landslides in the United States runs into hundreds of millions of dollars. As another example, the initial cost of the drainage system for stabilization of only 6.0 hectares of the northern side of Stag Hill to establish the University of Surrey was estimated at about £153,000 in 1965, (Simons 1977). In Hong Kong, since 1976, landslip preventive works have been carried out to 180 slopes at a cost of $430 million (Bowler and Philipson, 1984).
Landslides and slope movements may threaten: (a) single houses or entire settlements; (b) agricultural and forest lands (c) the operation of quarries and exploitation of mineral deposits; (d) communication under construction or in use; (e) tunnel constructions (0 water, sewage and gas conduits, telephone and electrical wires; (g) the functioning of submarine cables which maybe interrupted due to subaqueous slides; (h) hydrotechnical works, particularly dam constructions; (0 diversion canals, penstocks; Slope movement phenomena are usually studied from two different points of view: - geological studies; - engineering studies. Geologically they are considered as a natural process in sculpturing the land surface and the cause of their origin, and the resulting surface forms are studied as the significant exogenic denudation processes.
The approach of engineers and engineering geologists is quite different. Engineers investigate the slopes and the safety of the constructions to be erected on them, in the case of natural slopes, and study the degree of safety and reliability of man- made earth structures against failure during their service life. Furthermore, engineers develop methods for a reliable assessment of the stability of slopes as well as the controlling and stabilization works needed. Engineering studies of slopes are based on the principles of soil mechanics. The best results of stability studies can be achieved only by the combination of both these approaches. The quantitative determination of the stability of slopes by the method of soil mechanics must be based on a knowledge of the geological structure of the area, the detailed composition and orientation of strata and the geomorphological history of the land surface. On the other hand, geologists may obtain a clearer picture of the origin and character of sliding processes by checking their considerations against the results of static analyses and the research done by means of soil and rock mechanics. For example the degree of stability of an old reactivated landslide and a first time landslide, having all other features the same, can be quite different. From what has been mentioned above, it follows that the study of sliding phenomena is of theoretical and practical importance both for the engineer and the geologist. It is the recognition of the causes, character and development of landslides which makes it possible to appreciate the extent of the danger and to find an adequate solution for the control and correction of sliding area. Finally landslides and earth movements that are not anticipated or not well understood may endanger the results of human work and human lives. Before design and construction of any remedial measure for correcting a slide it is necessary that appropriate field investigations and preliminary planning are carried out. 1-2 FIELD INVESTIGATION
A reasonable project for the corrective and preventive treatment of a landslide should be based on a detailed field investigation. The scope of field investigation should include topography, geology, ground water regime, weather, and the shape of sliding surfaces.
The presence or absence of pre-existing slip surfaces must be ascertained and hence continuous sampling or closely spaced alternating samples in adjacent boreholes should be carried out, or visual inspection in deep test pits may provide crucial information (Simons and Menzies, 1980).
1-2-1 Topography.
The topography or geometry of the ground surface is an overt clue to post land slide activity and potential instability. As the topography of a landslide is continually changing, so the area must be mapped at different times, from several years before construction to several years after remedial measures are undertaken. Ultimately, the effectiveness of corrective measures is expressed by whether the topography changes.
Air photographs are most useful for the investigation of landslides, because they offer a perfect three-dimensional view of the area. From an air photograph, one can determine precisely the boundaries of a landslide, as the slope surface below the scarp is irregularly undulated with ponded depressions. A very good example showing how air photographs are useful in slope stability investigations is given in figures (3-5) and (3-6). Back scarps and the toe of the slip are easily recognisable.
In photographs showing a clear picture of vegetation, the disturbed vegetation cover can reveal recent landslides and the age of movement could be inferred from the curvature of trunks and new growth.
Aerial photographs may be of great help in preparing a programme of research work, especially the surveying plan, the layout of cross-sections and a reasonable location of boreholes and trial pits. 1-2-2 Geology
Geologic structure is frequently a major factor in landslides. Although this topic includes major large scale structural features such as folds and faults, minor structural details including joints, small faults, and local shear zones may be even more important. The landslide and the surrounding area should be mapped geologically in detail. On the map, the shape of the head scarp and the area of accumulation, outcrops of beds, offsets in strata, and changes in joint orientation, dips, and strikes should be identified. In particular, the depth and the dip of sliding surfaces must be measured, when exposed. Of special importance for the recognition of the character of the movement is the examination of cracks.
The principal parts of a landslide and the characteristic cracks may be defined as follows (Zaruba and Mencl, 1969):
On the root area can be seen open cracks perpendicular to the direction of movement, Fig (1-la). The head scarp is represented by the outcrop of the main slip surface along which the rock mass moved down, fig (1- lb). If the landslide is not so old that the initial form of the scarp would be obliterated by the disruption of the upper edge, the inclination of the sliding plane can be inferred from the shape of the scarp. On the landslide body itself, a series of transverse cracks is observable, which in the upper part are generally open and are of the tension type, Fig. (1- lc). In the lower part the cracks are closed, sometimes even deformed by pressure. The bulged part of a slide is occasionally cut by cracks arranged radially in relation to its arcuate outline, Fig (1-ld). At both flanks of a landslide, longitudinal shear cracks develop,along which lateral ridges maybe squeezed out Fig(l-le).
root orec
—lonque
toe. slidinq plane
Fig.(1-1) The principal parts of a landslide and characteristic cracks (Zaruba and Mencl, 1969) a - lunar cracks, b - head scarp, c - transverse cracks, d - radial cracks, e - lateral ridges. 1-2-3 Monitoring of the groundwater pressure.
Water is a major factor in most landslides. A plan for corrective measures requires a good knowledge of the hydrological conditions of the slide itself and of its surroundings. The first task is to determine the depth of the ground water table and its fluctuation and to map all streams, springs, seeps, wet grounds, undrained depressions and permeable strata.
Groundwater flow exerts pressure on soil particles, which impairs the stability of slopes and can also wash out soluble cement and thus weaken the inter granular bonds; consequently, cohesion decreases and the coefficient of internal friction drops. The moving ground water washes out fine sand and silt particles from the slope and the underground cavities thus formed weaken the stability. Huang, (1983) mentions that the confined groundwater acts on the overlying impervious beds as uplift, and, the impervious bed can be a slip surface.
A thorough knowledge of the distribution of pore water pressure in a slope is therefore, essential for a proper analysis of a long-term stability problem and, in addition to seasonal changes, any variation of pore water pressure across the site or with depth must be measured. The distribution of pore water pressure with depth will have a controlling influence on the depth of a potential slip surface (Simons and Menzies, 1980).
It is to be mentioned, however, that the monitoring of the groundwater pressure is a complicated technical task itself. There are many factors which may initiate considerable errors in monitoring of the groundwater pressure. These factors are explained in Chapter 2 and evaluated later in section 8-2.
1-2-4 Weather
The climate of the area, including rainfall, temperature, evaporation, wind, snowfall, relative humidity, and barometric pressure, is the ultimate major factor influencing most landslides. The effects of these factors can seldom be evaluated analytically because the relations are too complex. Empirical correlations of one or more of these factors, particularly rainfall, snow and melting temperatures, with episodes of movement or movement rates can point out influences that must be controlled to minimize movements. There are indications that in some landslides, chemical changes like hydration, ion-exchange in clay, induced by percolating water, are another deleterious factor. Thus for instance, areas built up of clays and glauconitic sandstones show a susceptibility of sliding.
Accordingly all of the available meterological data should be collected and analysed, and attempt should be made to establish a correlation between the fluctuations of the meteorological records and the response of the sliding slope.
The mechanism of rain-induced slope failure has received little attention. The earliest attempt to provide a rational link between infiltration and soil suction has probably been made by Lumb (1962). Lamb (1975) also provided a regional case study of rain-induced failures in residual soil slopes. The mechanism of rain-induced failures has been discussed by Morgenstem & de Matos (1975), Blight (1977) and Morgenstem (1980) in very general terms. In Brand’s opinion(1981) the great majority of the world’s slope failures, particularly those large enough to be classified as ’’landslides”, are rain-induced failures which occur in residual soils. He has elaborated the mechanism of rain-induced failure and mentioned that the stress path which is followed in the field during a rain-induced failure is very different from the stress path, followed in the usual forms of triaxial tests which are conducted to determine the shear properties for slope design and slope analysis. Because, in the field, in contrast to the triaxial test, ai and (J 3 are sensibly constant and the pore pressure increases as it rains. Brand (1984) has then studied and summarized the landslide situation in eight countries of Southeast Asia and has again concluded that the rain-induced landslides in the residual and colluvial terrain are common in all the countries.
1-2-5 Shape of sliding surface.
A knowledge of the slip surface profile beneath a landslip is essential for stability analysis and for the design of remedial works.
The subsurface geometry of a landslip is defined by the shape, orientation, and depth of a slip surface or zone. These dimensions are important, as they may highlight horizons where groundwater pressures or shear strength parameters are critical and they are essential for stability analyses and for the design of remedial works.
Site investigations are designed on the basis of available geomorphological, geological, geotechnical, and historical information. If such desk studies are augmented by a prior knowledge of the geometry of the failure surface, the efficiency of site investigations, feasibility studies, and costing exercises could be greatly enhanced. The usual method of locating the slip surface is by means of trial pits and boreholes. Trial pits are dug and inspected until the slip surface has been exposed. This requires a costly site investigation and, even then, the position of the slip surface may be difficult to identify; consequently, some form of instrumentation is usually required. This may include slip indicators, columns of blocks which can be examined by subsequent excavation, piezometer tubes which are plumbed with slip rods, and the electromechanical inclinometers which are guided down aluminium tubes (Carter, 1983) The inclinometer measures the change in inclination or tilt of a casing in a borehole and thus allows the distribution of lateral movements to be determined as a function of depth below the ground surface and as a function of time. Geophysical methods are sometimes used to help locate the slip profile but interpretation of the results is often difficult (Bogoslovsky and Ogilvy, 1977). A useful review of the various methods used for locating the slip surface beneath landslides is given by Hutchinson (1983).
Graphical methods for estimating the depth of slip surfaces using elementary cross sections have been described by Vamess (1958, 1978) and methods using surface measurements are given by Jakobson (1952) and Philobrick and Cleaves (1958). A new method of location of a slip surface is due to Carter and Bently (1985). In this method the geometry of slip surfaces beneath landslide is predicted from surface measurements. This technique requires a number of survey stations to be established on the land slip and to be monitored over a period sufficiently long for movements to occur that are larger than errors inherent in the survey procedures employed.
1-3 PRELIMINARY PLANNING
When a slide takes place, it is necessary to determine the causes of the slide, so that proper remedial measures can be taken. The detection of the causes may require continuous observations, and a final decision cannot be made in the short term. To prevent further disasters or loss of property which may be caused by the sliding mass during this investigation period, some well known factors which add to the instability should be quickly eliminated. Even in the case of sliding slopes where the main causes are known, it would be wrong to plan an extensive remedial measure prior to preliminary investigation. For example if extensive earthworks for slope adjustment is planned prior to the drainage of surface water, the redeposited soil could possibly be soaked by water. Accordingly, initially all streams and temporary water courses should be prevented from entering the sliding area. In addition, all springs issuing within the slide area,especially at its head, must be diverted outside the slide area, and ground water should be pumped out from all wells in the slide area. At the same time, the ground surface should be levelled out and drainless depressions and all cracks filled so as to ensure a continuous run-off of surface water.
1-4 CORRECTIVE AND REMEDIAL METHODS, TREATMENT OF A SLOPE SHAPE AND DRAINAGE.
1-4-1 Introduction.
Corrective methods can be used either to decrease the driving forces or to increase the resisting forces. Factors affecting the driving and resisting forces are interrelated and sometimes H is difficult to determine whether the measure results in a reduction of driving forces or an increase of resisting forces.
There are, of course, many remedial methods which are installed appropriately with respect to the characteristics of the failure and cost estimations. In the Stag Hill area, however, the remedial measures include drainage and some treatment of the slope shape. In spite of the fact that the treatment was not proposed in the beginning, i.e. in 1965, recent surveying of the slope during this research in 1986 revealed that the slope shape has been considerably treated along some profiles. Accordingly, in this section, the two above mentioned methods, treatment of the slope shape and drainage, will be explained and appropriately reviewed with the emphasis on the trench and counterfort drains.
1-4-2 Treatment of slope shape (removal of weight)
The concept of the corrective methods can be well understood from the interpretation of the stability equation of Standard Method of Slices. in which: ES. = factor of safety c; = cohesion (effective) o' = angle of friction (effective) Wn = weight of slice (total) u = pore pressure b = width of slice (bn = ln cos a n) a = slope angle (inclination of the radius of the slip surface with respect to vertical at any points), Fig.(l-2). Fd and Fr = driving and resisting forces respectively.
with reference to equation (1-1), to minimize the driving force £ Wn.Sinan, the weight and the slope of the sliding mass should be corrected so that the summation of the £ Wn Sinan is a minimum. It must be noted that any change in the weight of the sliding mass will affect the normal effective stress and in turn the shear strength, however.
In general, the magnitude of a n changes from slice to slice and increases from the toe of the slip to its head, it can even be negative for one or a few of the slices near the toe, Fig(l-2b), so if the weight of the soil near the toe of the slip is increased, the total driving force £ Wn Sinan will decrease. Conversely, if the weight of the sliding mass is concentrated near the head of the slope the safety factor will be decreased.
In fig.(1-2b) if the surcharge load AW is applied to a slice having a negative value of an, the driving force decreases to (Fa - AW Sin c^)
The change of pore pressure in the short term will be:
Au = B[Ao3 + A(Aai-Ao 3)]
in which Aai = A W cos 0Ln/ ] n assuming B = 1, A = 1 then Au = Aai = A Wcos a n A It can be seen that the change of pore pressure is the same as the change of total normal stress in the short term. In other words the effective stress does not change in the short term, therefore the safety factor is increased.
In the long term while the driving force remains unchanged i.e:
(Fd - AW Sin an), the shear strength increases to
[Fr + (AW cosan)tan 0'] which in turn increases the factor of safety furthermore.
If AW is applied on a slice having = 0, in the short-term neither driving force nor shear strength changes and the load is actually inert. The point of application of the inert load is called ’’neutral point in the short term”. In the long term, however, while the driving force Fd does not change, the shear strength increases to [Fr + (AW cos On) tan 0'], which means that the neutral point in the short term becomes beneficial in the long term.
As this small load moves from the slice having a = 0 to slices having a > 0, toward the head of the slope, the driving force is being increased while the shear strength does not change in short term. So the effect of the surcharge load on the stability decreases from neutral to detrimental in the short term, for all values of a > 0.
In the long term, there is a slice having On = Ob up to which the ratio of [Fr + (AW cos a n) tan 0'] / (Fd + AW Sin an), is still greater than Fr/Fd- So up to 0Cn When a n = cxb, then: [Fr + (AW cos ab) tan 0'] / (Fd + AW Sin ocb) = Fr/Fd- which means that the surcharge load will have no effect on the stability of slope in the long term. The point of application of this load is called ’’neutral points in the long term”. For all a n>(Xb values the surcharge load has a destructive effect on slope in short and long term. With respect to the discussion above the stability of the slopes may be substantially increased either by reducing the load of soil at the head or direct reduction of the slope and its enlargement at the toe. In the case of the toe or shallow slips, as the dumping of extra soil at the toe may not be effective, the stabilization is affected by direct reduction of slope, Fig (1 -3a), or flattening by cutting Berms, Fig (1 -3b). Deep and base failures can be stabilized by dumping extra material at the toe without any flattening or cutting of slope, or both flattening and damping at the toe, fig.(l-3c). 1-4-3 Drainage An unfavourable groundwater condition is a frequent cause of slope sliding. Water lowers stability and contributes to slope failure in the following ways: By reducing or eliminating cohesive strength By increasing pore water pressures which reduce effective stresses, thereby lowering shear strength. By producing horizontally inclined seepage forces which increase the overturning moments and the possibility of failure. By lubricating failure planes after small initial movements have occurred. By supplying an excess of fluid that becomes trapped in soil pores, and then during earthquakes or other severe shocks, leading to liquefaction failures. To prevent water lowering the stability, both surface and ground water must be led away from the sliding areas as much and as soon as possible. This is achieved by drainage systems. Drainage systems can be classified as surface or subsurface drainage. 1-4-3-1 Surface Drainage The surface of the area affected by sliding is generally uneven hummocky and traversed by deep fissures and cracks. In depressions and fissures water accumulates and wet grounds develop. The source of water can be from rainfall or an initial small movement in an unstable area which may cause the rupture of water mains, sewer pipes, and so on. Although re-shaping the surface of the sliding mass can be extremely beneficial in that cracks and fissures are sealed and water-collecting surface depressions are eliminated, any seal will be broken by the slightest further movement due to reactivating the slide or even changes of moisture content. It is better therefore to make arrangements to drain cracks. Streams and temporary water courses should be prevented from entering the sliding area and in addition all springs issuing within the slide area, especially at its head, must be controlled and led outside the slide. All of this can be done with a surface drainage system. Although surface drainage in itself is seldom sufficient for the stabilization of a slope in motion, it can contribute substantially to the surface drying and thus also, controlling of the landslide. 1-4-3-2 Subsurface drainage. The purpose of subsurface drainage is to modify the shape of the seepage regime rather than to control its hydraulic boundary conditions, or in other words the lowering of the groundwater table is not of primary importance, rather the change of the flow net pattern is the crucial point in stabilizing by drainage, (Nonveiller, 1981). Cedergren (1977) mentions that, many natural slopes become saturated during periods of heavy precipitation, in which the water table rises to nearly ground surface, and water flows essentially parallel to the direction of the slope and the seepage force destabilizes the slope. If the slope is underlaid with a highly pervious gravel layer, the water will flow vertically downward to the gravel and the energy of free water in the soil will be consumed harmlessly. This view is, of course, reasonable only if the performances and properties of the soil and the drainage layer are in such a way that the downward flow is established. In clay, and probably in most other soils, however, the flownet is a function of the stress dependent permeability characteristics of the soil. Accordingly, even if a very efficient under drainage system is installed the ground water pressure profile above the drainage system may never tend toward zero. In other words, a vertically downward flow is never established. These cases are encountered in Stag Hill and discussed in section 8-3. Subsurface drainage as a direct method of slope stabilization can be very effective but as a long term solution it (and surface drainage ditches filled with granular materials) suffers greatly because the drains must be maintained if they are to continue to function. Subsurface drains can be classified as: a - Trench and Counterfort drains b - Blankets c- Galleries d - Horizontal drains e- Well systems (vertical drains) a - Trench and counterfort drains; definition and literature review. The counterfort drain seems to have been the first of these types to be used as a principal stabilization measure in France and England during the first half of the the 19 th Century. The trench and counterfort drains are generally narrow and aligned directly downslope thus largely avoiding the risk of reactivating the landslide being treated. There are sometimes supplemented by shallower drains laid in a chevron or herringbone pattern. An earlier version of the trench drain is the”counterfort drain”. The term ’’counterfort drain” is used to describe trench drains which penetrate into solid ground below the soil which is being drained, and may therefore provide some mechanical buttressing effect as well as their effect on the pore water pressure and hence shear strength in the drained soil, while the term ’’trench drains” is reserved for those which do not thus penetrate, and so contribute to stability only through their drainage actiori.(Nonveiller, 1981). The major use of these types of drains is to stabilize shallow slides and slides of a principally translational character. The trench drains are generally limited by construction considerations. Usually trenches are excavated to depths of 3 to 5 meters with power equipment. The open filled drain trenches running cross-slope are sometimes built above the crest of the slip or slope, when they are usually termed interceptor or cut-off drains. Shallow ones merely intercept surface run-off; deeper ones are intended to intercept groundwater flowing towards the slope. Care must be taken to avoid siting cut-off drains so that they could act as a tension crack in any future land slide. Schweizer and Wright (1974) mention that the bottom of an interceptor drain should be founded on unyielding material, and if the drain is located within the slide mass the base should be below the elevation of the slip surface and if movement destroys the base of the drains, its effect is lost and stability will not be increased. This is, of course, an old belief and the basis for such a view has been removed since Trezaghi’s enunciation of the principle of effective stress in 1925. In connection with the ultimate efficiency of a drainage system, Hutchinson (1977), with reference to the field evidence suggests that in clay soils it may take some years for a trench or counterfort drain system to take effect fully This view seems to he reasonable, because the stress release occasioned by the excavation of the drains generally produces a significant short-term modification of the original pore water pressure, which is followed by an intermediate stage of consolidation and possibly swelling before the long-term condition has been reached. The long-term condition, of course, is the major interest in the context of slope stabilization, however. Up to the last decade the design of trench drains for slope stabilization in the UK had been carried out on a semi-empirical basis (Hutchinson, 1977). In the fields of soil physics and agricultural engineering, an extensive literature on groundwater flow to drains is available (e.g. van Schilfgaarde J., 1970). This is largely concerned, however, with the prevention of water-logging of crop roots and thus with the determination of the phreatic surface between horizontal drains of relatively shallow depth, usually 1.0 to 1.5m. Drainage for slope stabilization on the other hand is much deeper and thus pore water pressure monitoring at deeper depths, say 5 to 12 m., is required. However, there are some solutions of the flow of groundwater to drains which have been recently devoted for slope stabilization purposes. These solutions are briefly reviewed as follows: I - Hutchinson’s solution Hutchinson (1977) approached the problem of the ground water flow to drains through the use of flow nets, as was suggested by Henkel (1957). The solution was conducted using a finite element technique with the following assumptions: 1 - The ground surface and original piezometric surface are horizontal. 2 - Ground permeability is homogenous and isotropic, Rjc= = 1 Js.v 3 - There is an impermeable layer which drains may penetrate to that depth. 4 - The drains are of great length L upslope compared with their spacing S, so that a two dimensional approach will have some validity. 5 - A horizontal ponded surface exists at some depth at p = D - ho below ground level. 6 - The trench drains are rectangular cross section and parallel to each other. The definition of the various symbols used are shown in the key cross section of Fig (1-4). Then assuming that the ponded surface exists at ground surface (p = o & D = ho) and that the piezometric head at the drain invert is zero (ha = o) figures (l-5),(l-6) and (1-7), showing the flow net and variation of piezometric level, were obtained. With reference to Fig (1-5) and (1-6) it can be observed that the efficiency of a drainage system at invert level increases with increasing depth of impervious surface. However, it has been mentioned that the increase of n above 4.5 appears to have little effect on the flow net within the drain depth. Nevertheless, the interesting point is that if it can be assumed that the slip surface is comparatively impervious then it can be concluded that for a defined drainage system the deeper the slip surface the higher the efficiency. This is of course in contrast to the past general belief that drains which did not penetrate beneath the seat of sliding were of little or no use. Although the past belief was recently confirmed by Schweiger and Wright (1974), the basis for such a view has been removed by Terzaghi’s enunciation of the principle of effective stress in 1925. With reference to Figs (1-6) and (1-7) it is seen that: For a constant depth of drains (D = cte), the less the drain spacing the higher the differential rate of increase of ultimate efficiency of the drainage system in lowering of pore water pressure at trench invert level. II - Bromhead’s solution. Later on, the problem of seepage into counterfort drains was solved analytically and numerically by Bromhead (1984) on the basis of the analytical solution that defeated Richardson in 1908. The solution is carried out with the following assumptions: 1 - The effects of the sloping surface of the ground are negligible. 2 - The sides of the drains are wetted, and are at atmospheric pressure and the drains operate with a negligible head of water in them. 3 - The trench drains penetrate all the way through a soil layer of finite thickness H and uniform (but possible anisotropic) permeability, to an impermeable layer. 4 - Flow is two-dimensional. 5 - In addition to the wetted drain sides, the upper surface of the soil is also assumed wetted and at atmospheric pressure. The achieved solution is graphically shown in Fig (1-8), in which: To include soil anisotropy the real drain spacing S' can be scaled to an apparent spacing S by means of the transformation: S = S/VRk, in which Rk = III - Stanic’s solution. Stanic (1984) examined the determination of the necessary spacing of trenches for a required increase in factor of safety for long-term (ultimate efficiency), steady state condition. Similar to both previous solutions, this method is also independent of the coefficient of permeability. The solution is based on the Laplace differential equation AH(Xjy)Z) = 0,with the given boundary condition. Stanic mentions that in the materials with high permeability the drains would lower the groundwater table in certain period, but at other times the water level would be at the same level as in the case without drains. The latter case is more critical and therefore the analysis of the drain influence with the assumption of no change in the groundwater table is appropriate. This and some other thoughts,which were mentioned previously, are most probably due to lack of sufficient and precise experimental work and case histories. However, if this view was generally true, then the only stabilizing effect of a drainage system would be the deformation of phreatic surface between the adjacent trenches, from a straight to a curved shape. As the shape factor in these cases is about 0.8 (Hutchinson, 1977), so the efficiency of a drainage system would always be only about 20%. In section 8-5, however, it is shown that very considerable efficiencies can be achieved even in London clay, having a permeability value as low as 10'8 to 10-9 cm/sec, with a properly designed system. Stanic schematised the problem with the following assumptions: 1 - The slope is infinite in all directions with the inclination B to the horizontal (tanB = 1/n). 2 - The soil is homogenous and isotropic. 3 - The slip surface is impermeable and parallel to the slope. The drainage trenches of depth h from water table are dug to the slip surface and at spacing S. 4 - The trenches are long in comparison with the spacing S. 5 - The ground water table is parallel to the slope and does not change with time. 6 - The drain trenches are in the direction of maximal slope inclination. The characteristic profile and cross section of slope and the diagrams for determining the average and maximum piezometric levels on the slip surface constructed by parametric study, are shown in Figures (1-9) and (1-10). Analysing models of variable length, it was shown that the influence of the boundary condition on the potential in the middle of the model is negligible when the model length is greater than four times the height of the model. Although the solutions reviewed are appreciable attempts to formulate the function of drainage trenches, there are some points and questions which spring to mind,however. In spite of the fact that no rigorous research or clear evidence to verify the impermeability of slip surface is seen in the literature, all of the reviewed solutions are derived assuming an impermeable slip surface. Huang (1983) mentions that slip surfaces are generally impervious. The evidence which he has given is the appearance of new springs and wet ground when a slip surface approaches the ground surface. This view, of course, does not seem to be true in the case of clay slopes, because, the permeability of such a soil is so low that water can never flow easily through the clay to make a spring. Even in the case of fissured clays it is, of course, true that the polished slip surface joins up all discontinuities along a thin surface, but it does not mean that the slip surface is considerably more impermeable than the adjacent clay. Although the fissures may provide a path to moving groundwater of greater permeability than the intact lumps of clay, underneath ground level the high lateral pressure must tend to keep the fissures closed (Skempton, 1961). The other point is that, in all methods it is assumed that a ponded surface exists at some depth below ground level or that the groundwater table does not change in time. This means that the ground surface is subjected to a continuous infiltration and rainfall. Although the efficiency of a drainage system in the worst meterological condition is evaluated in this way but it may lead to a very conservative result in the design of a drainage system. In the nature, however, the meterological condition of a defined area is probably the most important and cmcial factor affecting the groundwater pressure especially at comparatively shallow depths, say 8 meters and less, at which most landslides occur and drainage trenches are installed. As the rainfall and in turn the infiltration are time dependent, the efficiency of a drainage system will also be a function of time. In other words, a continuous consolidation and swelling process dominates the efficiency of a drainage system in compressible soils. This process is completely ignored and, of course, cannot be easily involved in analysis and design, because the meterological regimes are very complicated. Accordingly, the analysis, design and study of drainage systems must be based on a theoretical-experimental basis. The compatibility of the reviewed methods with practice and the extent to which they can be adequately used to design a drainage system will be discussed in Section 8-5, however. b - Blankets Blankets, or thin horizontal sheets of more pervious material, such as sand, are sometimes incorporated in the structure of embankments in nearly saturated clays. The reason is that when wet fill is placed in rolled earthfill embankments, the pore-pressure that develops may become high enough to endanger the stability. The vertical spacing of these drainage layers is usually determined by assuming that they act with full efficiency and the pore water through them remains at atmospheric pressure. Gibson and Shefford (1968) have theoretically evaluated the efficiency of drainage blankets in terms of excess pore-pressure set up in them. They have shown that in order to be fully effective a drain should have, typically, a permeability at least 106 times that of the surrounding clay fill, although an acceptable efficiency would be achieved with a ratio of about 3 * 104. A drainage blanket only 103 times more pervious than the fill is likely to be useless. Sills (1974) has assessed the above mentioned theoretical concept by comparison with the pore pressure distribution observed in three specific drained embankments. She has mentioned that the effect of the drainage system is directly related to the thickness and spacing of the blankets. c - Galleries (Drainage Tunnels) Where the depth of unstable, saturated soil is too great for construction of drainage trenches, drainage galleries are sometimes used to discharge and lower the groundwater pressure. Owing to their large perimeter, a large amount of water may be discharged. Their effectiveness may be increased by drainage borings in the walls, floor or roof of the gallery. Construction of galleries are costly and a large amount of engineering work is required. d - Horizontal drains (Drainage borings) Horizontal drains are at least five times cheaper than drainage galleries (Zaruba and Mencl, 1969). A horizontal drainage system usually consists of 2 inch to 4 inch diameter steel pipes installed in the face of the slope (Schweizer and Wright, 1974). Although they are described as horizontal drains, the pipes usually vary in inclination from 2 to 20 degrees above the horizontal. Horizontal drains have been successfully used on a wide variety of slope profiles and in soils of markedly different engineering characteristics. Kenney, Pazin and Choi (1977) have established charts to determined drain length and spacing in order to achieve a certain increase of stability of a particular slope. Nonveiller (1981), has numerically studied the efficiency of horizontal drains on slope stability and concluded that the time needed to achieve an adequate safety gain increases with decreasing coefficient of consolidation. Thus longer drains and a closer spacing is needed in clay slides of low permeability and high compressibility, in order to reduce the time to stabilize the slide. Lau and Kenney(1984), studying five horizontal drains in a natural clay slope concluded that the effectiveness of horizontal drains depends on the spacing and diameter of the drains and on their location relative to the critical slip zone. e - Well Systems (Vertical drains) As a landslide prevention or correction measure, well systems are most commonly used in conjunction with horizontal drains to provide relief of hydrostatic pressure and gravity discharge of the subsurface water, respectively. When employed for slope stabilization, vertical drains have been used for three basic purposes (Schweizer& Stephen, 1974): (1) to provide a drainage path between lenses or strata of water-bearing material which are separated by impervious layers (Palmer, 1950; Parrott, 1955) (2) to relieve artesian conditions which may develop at or below the surface of rupture (Smith, 1964; Holm, 1969; Smith, 1969); and (3) to relieve excess hydrostatic pressures in slopes of saturated clay and therefore expedite consolidation and increase the shear strength of the soil (Fellenius, 1955; Holm, 1969). In addition to the above mentioned ’’drainage methods”, interconnected vertical wells have often been used in place of cut-off trenches where pervious water bearing strata lie beyond the reach of conventional trenching equipment. Installations of well systems ranging 0.6 to 0.9 metres in diameter and 7.5 to 25 meters in depth are seen in the literature. For successful installation the pervious backfill material must satisfy the standard filter rules. An advantage of well systems is their flexibility, because additional wells can be installed at intermediate points if initial spacing proves inadequate to control seepage and groundwater pressure. Cedergren (1977) has developed design charts which may be used either to estimate the,rate of seepage into pipes through filter materials of known permeabilites, or to determine the required permeabilities of filter materials to permit the given rate of seepage to enter the vertical pipes surrounded with such material. 1-5 SAFETY FACTOR, LITERATURE REVIEW In the stability analysis of slopes, as a part of the geotechnical analyses, many design factors cannot be determined with certainty. Therefore, a degree of risk should be assessed in an adopted design. The Factor of Safety fulfills this requirement. The Factor of Safety should take into account not only uncertainties in design parameters, but also the consequence of failure. Where the consequences of failure are slight, a greater risk of failure or a lower safety factor may be acceptable. Two types of the safety factor are used usually in geotechnical slope stability analyses. ’’Total factor of safety” which is used in the conventional working stress design method (WSD) and the alternate method which uses ’’partial factor of safety” and commonly referred to as limit state design (LSD). This method has been used for a number of years in geotechnical engineering. Table (1-1) shows the total factors of safety suggested by various sources for mining operations. All of these stipulations are based on the assumptions that the most critical failure surface is used on the analysis, that strength parameters are reasonably representative of the actual case, and that sufficient construction control is ensured. Meyerhof (1970,1984), examined the conventional total factors of safety in relation to the probability of failure for earthworks, retaining structures and foundations. The results of this study are given in table (1-2). The upper values of these safety factors apply to normal loads and service conditions, while the lower values can be used for maximum loads and worst environmental conditions. The total and partial safety factors are interrelated, because the partial factors suggested for use in limit state design have been obtained by comparison with conventional geotechnical analyses (Baikie, (1985). However, in order that the magnitude of the partial load and resistance factors remains constant for each type of design, it is necessary to introduce load and resistance modification factors (Meyerhof, 1984). These factors are given in the tables (1-3) and (1-4). It is customary in conventional practice to select a total factor of safety that is dependent upon the loading conditions and the consequence of slope failure. Highway and railway cut and embankment slopes tend to be designed using a lower total factor of safety than would be employed for a slope where a structure is to constructed near the top or base. In the latter case, for an earth slope composed of intact homogenous soil, when the strength parameters have been chosen on the basis of good laboratory tests and a careful estimate of pore pressure has been made, a total safety factor of at least 1.5 is commonly employed (Lambe and Whitman, 1969). With fissured clays and for non-homogenous soils, however, larger uncertainties will generally exist and more caution is necessary. Lee (1974) has given some typical safety values, as recommended by codes of practice or regulations [Lamson (1967), EAU (1971)]. These values which are chosen by consensus of opinion on ’’good” practice are given as in table (1-5). Whitlow (1983) mentions that from the point of view of economics, each problem should be treated on its own merits and a minimum value of F.S. decided on the basis of all of the factors involved. Then he suggested the safety factors, given in table (1-6), as a general guide. Mitchell (1983), has recommended safety factors as shown in table (1-7) for permanent slope design. He mentions that the recommended safety factors should be increased by 0.25 where there is limited material testing and by an additional 0.25 where there is limited construction control. He added that it is commonly found that slope creep movements develop in many earth slopes where the calculated safety factor is less than about 1.25. The safety factor recommendations on table (1-7) are based on the assumption that site investigation and testing procedures have identified the most probable mechanism of failure and the applicable strength envelope. It is emphasised that when failure can develop in weak clay layers, samples of this clay should be subjected to multiple shear reversals in a shear box to investigate the possibility of residual strength developing in these layers. Matsuo and Suzuki (1985) examined safety factors in specifications from the view point of reliability-based design. They concluded that the optimum central safety factor decided by the reliability-based design of embankments on saturated clay layers becomes F.S. = 1.2 - 1.35 which are the values used in the present general safety factor method. With respect to the literature it can be concluded that authors and specifications all agree that the appropriate value of a safety factor for any earth structure is partially dependent on the confidence of engineer in his design and the reliability of the available data. This view is, of course, reasonable because the soil is a product of nature and the products of nature are so complicated (Terzaghi, 1936). Nevertheless, as given in table (1-1) to (1-7), there are some numerical values given as a general guide and basic criterion for design. These values can be modified with respect to the special characteristics of any earth structure to achieve the appropriate safety factor. Primarily it should be mentioned that the F.S. values, given by the above sources are for ’’good” practice, in which good site investigation, soil tests and construction control is carried out. In experience of the author, however, the site investigation, including soil properties and pore pressure pattern, which have been collected on this site are much more detailed than the data usually obtained from a normal site investigation. Probably the term ’’excellent practice” is appropriate for this site. For example, in practice, the pore pressure pattern is never evaluated as accurately as evaluated here. Accordingly, the F.S. values given by the above sources must be considered somewhat conservative for this area. In connection with the appropriate F.S. value, for the time before development of the area, in other words, if the slope was left as it was and neither remedial measure nor any building was constructed, at first glance it would appear that a factor of safety of about F.S. = 1.1 - 1.2 could be acceptable [British National Coal Board (1970) and Whitlow (1983)]. On the other hand, however, according to Whitlow’s experience creep movements develop in many earth slopes where the calculated safety factor is less than about 1.25. The interesting point is that at Stag Hill, fresh movements at the toe were reported in 1965 (Skempton and Petley, 1967a) when the corresponding F.S. values calculated in this research are about 1.22, table (8-5), confirming Whitlow’s experience. Accordingly, for undeveloped conditions a safety factor of F.S. = 1.25 seems to be reasonable in this area. At the present time, however, the area is mostly developed and multi-storey buildings are constructed on the slope. Therefore, a higher safety factor must be chosen, because the buildings cannot tolerate any significant movement and there is a very high risk of life loss, as the buildings are all residential. For this case, a safety factor of F.S. = 1.5 is the most common. Although F.S. = 2 is alternatively recommended by Whitlow (1983), it appears to be very conservative, however. With reference to the scope of a ’’good” practice, explained above, and with respect to the fact of that the effect of the sides actual three dimensional slide, which may improve the F.S. value by 5% (Skempton, 1985), are ignored in stability analyses, therefore, an F.S. value somewhat less than 1.5, say F.S. = 1.4 - 1.45 is probably the most appropriate value for this site. UNITED STATES (FEDERAL REGISTER, 1977) Minimum Safety Factor I End of construction 1.3 II Partial pool with steady seepage saturation 1.5 TIT Steady seepage from spillway or decant crest 1.5 IV Earthquake (cases II and III with seismic loading) 1.0 Suggested minimum factor of UNITED STATES safety with hazard potential (D’Appolonia Consulting Engineers Inc. 1975) High Moderate Low Designs based on shear strength parameters measured in the laboratory 1.5 1.4 1.3 Designs that consider maximum seismic acceleration expected at the site 1.2 1.1 1.0 FACTOR OF SAFETY BRITAIN (NATIONAL COAL BOARD, 1970) I* jp * (1) For slip surfaces along which the peak shear stress is used. 1.5 1.25 (2) For slip surfaces passing through a foundation stratum which is at its residual shear strength (slip circles wholly within the hank should satisfy (1)). 1.35 1.15 (3) For slip surfaces passing along a deep vertical subsidence crack where no shear strength is mobilized and which is filled with water (slip surfaces wholly within intact zones of bank and foundations should satisfy(l)). 1.35 1.15 (4) For slip surfaces where both (2) and (3) apply. 1.2 1.1 FACTOR OF SAFETY CANADA (MINES BRANCH, CANADA, 1972) I* ip * Design is based on peak shear strength parameters 1.5 1.3 Design is based on residual shear strength parameters 1.3 1.2 Analyses that include the predicted 100-year return period accelerations applied to the potential failure mass 1.2 1.1 For horizontal sliding on base of dike in seismic areas assuming shear strength of fine refuse in impoundment reduced to zero 1.3 1.3 *where there is a risk of danger to persons or property ** where no risk of danger to persons or property is anticipated Table (1-2) Total factors of safety and probability of failure (after Meyerhof 1970,1984) Total factor Probability of safety of failure Earthworks 1.3 - 1.5 lO-2 Earth-retaining structures and excavations 1 .5 -2 10-3 Foundations 2-3 10-4 Table (1-3) Load and load modification factors (after Meyerhof 1984). Load Load Modification Item Symbol Factor Factor Dead loads f d 1.25(0.85)* Live loads, wind or earthquake h 1.50 Water pressures f \l 1.25(0.85)* Resultant force f q 1.25 (overturning, bearing capacity) Apparent earth pressure fq 1.25 (excavation supports) * Factors in parenthesis apply where load is beneficial. Table (1-4) Resistance and resistance modification factors (after Meyerhof 1984) Resistance Resistance Modification Item Symbol Factor Factor Shear strength Cohesion(c) (stability, earth pressures) f c 0.65 Cohesion(c) (foundations) f c 0.50 Friction(tan 0) 0.80 Passive earth pressure f r 0.80 Basal stability of excavations Sand f T 0.6 Clay f r 0.8 Shaft resistance (effective stress) / r 0.6 Table (1-5) Typical Safety Factors (Lee, 1974) Strength Cohesion 1.3-1.6 tano 1.1-1.2 Load Dead. Soil weight, static water pressure 1.0 Live. Dynamic water pressure 1.2-1.5 Design Method Bearing capacity 2 .0 -3 0 Slope stability 1.0-1.5 Earth pressure 1.0-1.5 Anchor piles 1.5-2.0 Table (1-6) suggested safety factors (Whitlow, 1983). End of construction (embankments and cuttings): 1.3 Steady seepage condition 1.25 After sudden drawdown 1.20 Natural seepage of long standing 1.10 - 1.20 Soil tip 1.50 Problems involving buildings 2.0 Table (1-7) Typical safety factors (Mitchell, 1983) Slope description and High risk Low risk (no loss of design conditions (loss of life or severe life and moderate damage damage from failure) from failure) permanent slope in geologically stable material, all conditions 1.3 1.15 permanent slope in geologically metastable materials, all conditions 1.5 1.25 non-impoundable dikes and embankment, all conditions 1.3 1.3 impoundment dams (a) end of construction 1.3 1.3 (b) normal operation 1.5 1.3 (c) rapid drawdown 1.3 1.1 (d) earthquake loadings 1.2 1.1 (e) earthquake in combination with (a), (b) or (c) 1.1 1.0 A geologically metastable material is intended to refer to a material susceptible to earthflow or where low safety factors may lead to creep movements and progressive softening. AWr Toe and shallow slip AWt Deep and Base Slip Fig (1-2) Concept of slope shape treatment in toe and base slips. A to B - Fills always beneficial atB - Fills neutral in short-term and beneficial in long term B to C - Fills detrimental in short term and beneficial in long term at C - Fills detrimental in short-term and neutral in long term CtoD - Fills always detrimental /— r (a) Direct reduction of slope. (b) Flattening by cutting berms Toeslip Toeslip (c) Dumping at toe. Baseslip. Fig (1-3) Treatment of Slope Shape in toe and base slips r < - nho G|H 'A/2 Fig(l-4). Key diagram: cross-section of typical trench drains. (1) Ground surface. (2) Original piezometric level on plane Fig(l-5) EH. (3) Piezometric levels on plane FG after drainage. Typical flow nets for Rfc = 1, drawn from finite element analyses: (4) Mean piezometric level on plane FG after drainage. a) for fully penetrating drains (n = 1) (5) Mean piezometric level on drainage inverts after drainage, b) for partially penetrating drain with n = 4.5. (6) Trench on counterfort drain. (7) Clay seal. (1) Trench drain (2) Impermeable lower boundary. (8) Impermeable boundary at depth. (3) Ponded upper boundary. VO 1.0 S /D « 6. o.e h_ \ w 0.i 0 0.2 Fig(l-7) Curves showing the'relationship between hm/ho and S/ho, and between h/hQ and S/hQ, for fully penetrating drains (with n = 1.0) and partially penetrating drains, with n ** 4 5 Rfc is taken as 1. 0-2 C-6 N.B. The note appended to the caption for Fig.(l-6) also applies here Fig(l-6) Curves showing the variation of piezometric level (at drain invert level and assuming hd “ 0) between trench drains, for = 1 and various values of S/p. a) For fully penetrating drains (n = 1) b) For partially penetrating drains (n = 4.5) Note. The ratios S/j) and h/o are also taken to be approximately equivalent to S/h/0 and h/h0 respectively. The effect of Rjj 1.0 can also be obtained from these curves by the appropriate scale transformation. Hutchinson'© solution BROMHEAD 1.0 /V / / ------:---- S ------/ / / • a • . 1 / #.v # • / / •*# #• / ■ / ✓ * # / * $ - } ✓ / / / b / ' / .. la! d=H Ibl d;3/H fc! d=%'H / / A 0 .5 / H ✓ s Id] d=%H «• s / s' s' s s’* i s' /'■ ^ , 0 * S’ ^ s ' ^ 0 S ------___ ------—" Id] 0.5 1.0 0.5 0 R< 1 /R C Fig (1-8) Variations of average pore water pressure with drain spacing and depth. CROSS SECTION Gr.L s m r ©I— r G-w .L i I H H \ / i T ' \ ' I j\l/ X ' Fig(l-9) . Characteristic profile and cross section of slope with drainage trenches. (1) Ground w. L. • (2) Maximum piezometric level between two drains. (3) Average piezometric level. (4) Slip surface. (5) Trench-soil interface. 100 1XO 3.0 0.80 080 0.70 0.70 2.0 080 050 3.0 050 0.50 2D 020 0.10 0D0 00__ 0.00) 0.0 1:2 1:3 1U 1:5 1*6 1:7 1 -8 He n'l l*-2 1:3 UA 1*5 T-6 1:7 VeQgTj Fig(l-lO) (a) Diagram of normalized average piezometric level at slip surface Op). (b) Diagram of normalized maximum piezometric level at slip surface (fm) Pp Pm rp - h >rm- h S/h = normalized drain distance S t a n i c 7 © eolation CHAPTER TWO - ERRORS IN FIELD MEASUREMENTS OF GROUND WATER PRESSURE. 2-1 GENERAL DEFINITIONS AND PRINCIPLES OF PORE WATER PRESSURE MEASURING SYSTEMS. 2-1-1 Types and principles of operation Ground water pressure is measured using piezometers. Piezometer systems may be classified into three major categories. These categories and operating principles of piezometers of each category are as follows: Open standpipe piezometers: The Casagrande (Casagrande, 1949) and Geonor (Bjerrum and Johannsen, 1961) piezometers are examples of this type of piezometer. These piezometers comprise a standpipe tube and a porous piezometer tip which is connected to the lower end of the tube. After installation, ground water can enter the tube only via the tip and the water level in the tube virtually shows the ground water pressure. The water level is usually measured with a dip meter. These piezometers are simple and cheap and can be used for permeability tests, but of course, their initial equalization takes a comparatively long time and in certain cases they fail to measure the ground water pressure with an acceptable error. These cases are discussed in Section 8-2-3. Closed system hydraulic piezometers; The U.S.B.R., hydraulic piezometer (Armstrong 1956, Earth manual 1960), the B.R.S. hydraulic piezometer (Penman, 1956), and the Imperial College type piezometer (Bishop, Kennard and Penman, 1961) are examples of this type of system. These piezometers usually comprise a porous piezometer tip which is connected to the readout location by twin tubes. De-aired water is circulated through the tubes until both the tubes and the tip are completely filled. The pore pressure can then be measured at the remote end of either water tube, making a correction between the tip elevation and the measuring gauge elevation. These piezometers, because of having de-airing facilities, can also be used in partially saturated soils. Permeability tests can be carried out with these piezometers. An important limitation of these piezometers is that neither the readout gauge nor the highest level of tubing should be higher than a specified level (usually 5 meters) above the piezometric level. For this reason, installation of these piezometers in areas having complicated ground water pressure patterns, like Stag Hill, (Refer to 8-3) might lead to waste of money. There is another type of closed system hydraulic piezometer (BAT system) which is different from the above mentioned systems. In this piezometer, the filter tip and pressure sensing unit are separate. At the time of measurement, the pressure sensing unit is connected to the filter tip by a unic hypodermic needle which passes through a rubber disc into the tip (Torstensson, 1975 & 1978). Diaphragm type Piezometers: These piezometers consist of a diaphragm which operates as a valve (pneumatic piezometers), or controls strain and in turn period of vibration of a wire (vibrating wire piezometers). A pneumatic piezometer tip comprises a porous element integral with a diaphragm transducer. Twin nylon tubes in a polythene sheath connect the transducer to the readout unit. To take readings, air, nitrogen or hydraulic pressure is supplied from the read-out unit to one side of the flexible diaphragm valve incorporated in the transducer. When this applied pressure is sufficient to balance the ground water pressure acting on the reverse side of the diaphragm, the valve opens allowing flow along a return line to a detector in the readout unit. The balance pressure is displayed on a burden tube gauge or digital display. These piezometers cannot show negative pressure. The elements of a pneumatic piezometer are shown in Fig. (4-2). Vibrating wire piezometers consist of a porous element integral with a diaphragm vibrating wire transducer. A cable connects the transducer to a terminal unit or direct to the readout unit. Readings are made measuring the time for 100 cycles and converting the time into frequency. In vibrating wire piezometers, a wire under tension is fixed between a fixed point and diaphragm. The pore water pressure acts on the reverse face of the diaphragm, so the change of the pore water pressure changes the tension of the wire. This change of tension is converted to the change of pressure in the following manner: When the wave is sent into oscillation, in its simplest mode its movement is that of a half wave between the fixed ends, oscillating at a frequency of f Hertz, where erg f= — (2- 1) 2p P in which 1 = vibrating wire length a = tension in the wire g = acceleration due to gravity p = density of the wire Equation (2-1) can be re-written a s : a = - pi2 f2 (2-2) g or a = k f2, k = | p | 2 where k is the gauge constant. In terms of period: (2-3) for rise of the pore pressure from P0 to Pi the change of tension of the wire will be equal to: 1 1 (2-4) in which: N0 = Pressure reading displayed by readout for a certain pressure of P0, before the piezometer has been installed. The N0 is called pre-installation base reading. Ni = Pressure reading displayed by readout after piezometer has been installed. So if the amount of N0 is determined for a certain value of P0, before the piezometer has been installed, then for any value of Ni after installation, the change of pressure can be calculated using the above formula. The value of k is determined by the manufacturer, or can be determined by calibration. 2-1-2 Porous element in Piezometers. Probably the most important application of the measurement of pore water pressure is for calculation of the effective stress. For fully saturated soils the expression for effective stress proposed by Terzaghi, (1924) is applicable. o' = a - uw (2-5) where a' denotes effective stress, a denotes total stress uw denotes pressure in the pore water measured with respect to the same datum. For partly saturated soils having both air and water under differing pressures in the voids, the effective stress equation is of the form (Bishop, 1959): o' = cr - ua + X (ua - uw) (2-6) in which ua denotes the pore air pressure and X is a parameter depends on the degree of saturation (Sj). In a partly saturated soil, the pore water pressure is lower than the pore air pressure due to the effects of surface tension or capillarity and in compacted clay soils this pressure difference may have a magnitude of several atmospheres even at relatively high degrees of saturation (Bishop et al, 1964). With respect to equation (2-6) it can be concluded that the estimation of effective stress on the basis of the measurement of pore water pressure alone or pore air pressure alone will be over estimated or under estimated respectively. The above brief explanation reveals the importance of pore pressure (air or water) measurement in practice and the use of appropriate instrumentation to measure the pore water pressure or pore air pressure. To measure pore water pressure, air bubbles must be prevented from entering the piezometer and this can be done by the mounting of an appropriate filter (ceramic) on to the piezometer tip. Ceramics can be classified in terms of pore size and hence air entry value, which is defined as the difference between air pressure on one side of the saturated filter and water pressure on the other, at which air bubbles pass through the ceramic. Typical air entry values for low and high air entry commercial ceramics may be 5 and 100 kN/m^ respectively. (Site investigation, 1982, Clayton, Simons & Mathews). In the case of saturated soil the properties of the filter are not critical, provided it retains the soil particles and is reasonably permeable. In partly saturated soil the pore size must be sufficiently fine for the air entry value to be greater than the amount by which the air pressure in the soil exceeds the water pressure (i.e. ua - uw). In both cases, reliable measurements cease when the pressure is so low as to cause cavitation in the water between the filter and measuring device (Bishop et al, 1964). Vaughan, (1974) has mentioned that even with a high air entry value filter, air will still enter the enclosed system cavity. He also mentioned that a thick high air entry value filter with a sealed cavity electric piezometer may operate without breakdown (breaking of the continuity of the water in the system) for many years if pore water pressures above atmospheric are measured. The other function of a high air entry value ceramic is maintaining rigidity of a closed system piezometer. The influence of trapped air is significant if rapid response is required (Vaughan, 1974). In connection with the effect of ceramic on the equalization process and permeability tests, Penman, (1961) concluded that the fine ceramic being much more permeable than the clay did not impair the rate of response. Gibson, (1966) has also shown that the permeability of the piezometer installation can be assumed to be infinite if the piezometer ceramic is at least ten times more permeable than the surrounding soil. 2-2 SOURCES OF ERRORS IN GROUND WATER PRESSURE MEASUREMENTS. 2-2-1 Introduction This research largely deals with pore water pressure pattern. So it was clear that the errors in measurements of the ground water pressure in this area had also to be appreciated. The presence of a large drainage system which initiates quick and considerable fluctuations in ground water pressure gave more emphasis to the importance of the evaluation of the errors in measurements of the ground water pressure. Sources of these errors and history of the theoretical and experimental studies have been reviewed in this section, and special attention has been paid to the errors associated with time lag and atmospheric pressure. Time lag is probably the major source of error in measurement of ground water pressure. After a piezometer has been installed it takes some time until the piezometer shows (virtually) the ground water pressure. The same phenomenon happens when ground water pressure changes. The time which a pore pressure measuring device requires to equalize a certain percentage of the out of balance pressure, between the intake point and ground, is called time lag. Time lag is a function of the characteristics of the soil, pressure measuring device and specially equation of the fluctuations of the ground water pressure. Time lag in general consists of two parts. Each part is defined in Sections 2-2-2 and 2-2-3, and then a review of the history of the theoretical and experimental studies is given in Section 2-2-4. Previous work in connection with the effects of changes of atmospheric pressure on the ground water pressure is reviewed in Section 2-2-5. The other sources of errors are discussed briefly in Section 2-2-6. 2-2-2 Hydrostatic time lag. When the water content of the soil in the vicinity of intake for a pressure measuring device remains constant, in other words, if the soil is assumed to be incompressible, and when other sources of error are negligible, the time required for water to flow to or from the device until a desired degree of pressure equalization is attained is called the hydrostatic time lag. (Hvorslev, 1951). For a given pressure change the volume of water which must flow into or out of the instrument is the representative of the energy required to operate a piezometer. Piezometers which operate with a little volume of water (rigid piezometers) can show the ground water pressure change much quicker than those operating with relatively greater volume of water. Hydrostatic time lag is a function of the permeability of the soil, shape and dimensions of the intake end of the piezometer, volume factor of the piezometer and the equation of the fluctuation of the ground water pressure. The volume factor (v) is defined as the volume of water flowing into or out of the system due to unit pressure increase or decrease respectively. The volume factor is a representative of the volumetric flexibility of the pressure measuring device. In a laboratory, evaluation of time lag is rather simpler, because an artificial ground water pressure is applied to the system, and theoretically, the solutions corresponding to simple equations of ground water pressure are also developed. In nature, however, the problem is very complicated, because the equation of the fluctuations of ground water pressure at any point is a function of soil properties, position and characteristics of natural or man-made drainage systems (in general, pattern of distribution of permeability); meterological characteristics of the area (frequency, duration and amount of rainfall); and other factors. 2-2-3 Stress adjustment time lag. Stress adjustment time lag is the time required for changes in water content of the soil in the vicinity of the intake point as a result of changes in stress conditions. Stress adjustment time lag, indeed, is governed by the consolidation and swelling characteristics of soil, so, for an incompressible soil, stress adjustment time lag can be negligible. A distinction must be made between ’’initial stress adjustment” which occurs only during and immediately after installation of a pressure measuring device, and the ’’transient but repetitive stress adjustment” which occurs each time water flows to or from the intake point during subsequent pressure observations (Hvorslev, 1951). Callstenius and Wallgren, (1956) have called the initial stress adjustment time lag as’’installation time lag”. 9 The initial stress adjustment time lag is initiated by disturbance of the stress condition of the soil in the vicinity of the intake during the piezometer installation. In the case of piezometers which are installed in boreholes, the influence of the installation upon the nature pore pressure is dependent on the method of hole boring and pore pressure can either be increased or decreased. If the hole boring is carried out with drilling, and if the piezometer responds more rapidly than the zone of perturbed pore pressure can equalize, and if the cavity pressure is initially higher than the pore pressure around the bore hole, the piezometer may appear to equalize to a pressure lower than true pore pressure. (Vaughan, 1974). In practice, the effect of the initial stress adjustment is augmented by effects of setting of sealing grout and swelling of the bentonite pellets. All these effects, however, are eliminated gradually. The first time equalization process is considerably affected by these factors . In the case of drive-in piezometers, which are pushed into the ground, after the piezometer has been installed a consolidation process begins. This causes pressure gradients, which generally decrease with time. The consolidation time varies within wide limits, with many factors. Installation time lags varying from a few minutes to two weeks have been observed in practice (Kallstenius & Wallgren, 1956). These people have compared the first time equalization of two standpipe piezometers of 42 and 60 mm diameter. It seemed to indicate that the influence upon the (installation) time lag, is roughly proportional to the square of the pipe diameters. Repetitive stress adjustment time lag is a consequence of swelling or consolidation of the soil in the vicinity of the intake point. Due to any unbalanced pore water pressure between the intake point and ground, water flows to or from a pressure measuring device and the effective stresses in the soil in the vicinity of the intake point will be subject to changes. If the ground water level is assumed to be constant and the pore pressure at intake point of a piezometer is altered manually the above mentioned changes will be more or less transient. With decreasing difference between the piezometer and ground water pressures, the stress conditions and void ratios will approach those corresponding to the pore water pressures in the soil mass as a whole (Hvorslev, 1951). This process always takes some time and is called ’’repetitive stress adjustment time lag”. In nature, as the ground water pressure mostly fluctuates, so the stress adjustment is always in process, not only in the soil in the vicinity of the intake point but also in whole mass of the soil. It is to be noticed that in equalization tests of piezometers in which the water pressure in the piezometer is lowered or increased manually and ground water pressure is virtually assumed to be constant, only the soil in the radius of influence of the piezometer goes under stress adjustment. The stress adjustment process has retarding characteristics and prevents changes of ground water pressure or equalization of a piezometer to proceed quicker. So, in pore pressure measurement practice, the shape factor and volume change of piezometers remain the dominant factor causing time lag, but this is not a general law. The equation of fluctuations of the ground water pressure is not only a function of the permeability and compressibility of the soil but also the geometry of the ground, boundary and meterological conditions all are crucial factors affecting this equation. The stress adjustment time lag is a function of the properties of the soil and also the volume factor of a piezometer installation. It is theoretically shown that in a same soil when the volume factor of a piezometer is increased, the radius of influence of the piezometer in the equalization process is also increased and more volume of soil goes under the stress adjustment process which in turn increases the corresponding stress adjustment time lag. The radius of influence of a piezometer is defined as the distance from piezometer centre to a region in the soil where the maximum change in pore pressures throughout the entire equalization process does not exceed 5% of the total pressure change in the piezometer system (Premchitt and Brand, 1981). 2-2-4 Theoretical & experimental Studies of time lag, Literature review. Hvorslev, (1951) formulated the equalization process for piezometers in incompressible soils. He derived the theoretical time-lag for a piezometer in an incompressible isotropic, fully saturated and infinite in extent soil from Darcy’s Law and the basic differential equation that governs the saturated flow through a falling head permeameter. If the piezometer and soil are initially at a pressure Uq and the pore pressure in the soil increases instantaneously to u©o the piezometer will theoretically record u©« after an infinite time. The time lag ”t” before the piezometer records a pore pressure (Ut), Fig. (2-1), is expressed as: R = ?°p~ Pt = 1-e^ (2-7) * o t_ Pt T or 8 = p^-= e (2-8) in which: T _ v Yw 1 FK Pq = Uoo - Uq Pt = Uoo - Ut 8 = equalization ratio (error at any time) R = Response of the piezometer system K = soil permeability (cm/sec) Yw = unit weight of water (gi/cm3) F = shape factor of the piezometer, or the shape factor of the piezometer installation in the field (cm). V = volume factor of the piezometer system, which shows the volume of water required to balance a pressure of unit head of water. It is expressed in cm5/gr and numerically is equal to the cross sectional area of the standpipe in cm2. For linearly changing ground water pressure, when the ground water pressure is rising or falling at a uniform rate of + a or - a respectively, the ground water pressure equation will be as follows, Fig. (2-2). Z = P0 + at (2-9) in which: P0 = groundwater pressure at time t = o with ut = o for t = o, the Hvorslev’s solution leads to: U t' a t - l - e T (2-1 P 0 - aT Theoretically a, T, and PQ may be determined by observing three successive changes in piezometer level at equal time intervals, t, and expressing the results by three equations similar to equation ( 2-10). Hvorslev has also extended his solution for the case in which the ground water fluctuates sinusoidally Fig. (2-3), and for steady state he has concluded: ts _ Za Xa=Za Cos27r ^ - = (2-11) Tw V i+(2rfi/rw)2 and arctan27T (2-12) Aww in which: Za = amplitude of the ground water pressure fluctuation. Xa = amplitude of fluctuation of the water pressure as measured with a piezometer. Tw = period of the ground water pressure fluctuation ts = phase shift T = ^^"7 is called ’’Basic time lag”. Josselin de Jong, (1953) appreciated the problem of equalization considering consolidation of the soil around the pore pressure meters. He considered a pore pressure meter having a spherical intake with radius ao and then solved the problems for the Rigid type and also the Cavernous type intake points. The solution was given only for constant pore water pressure and was based on the theory of the three- dimensional consolidation (Biot, 1941) and Darcy’s Law. Kallstenius and Wallgren, (1956) have also formulated the time lag in incompressible soils. Their method is based on the same basis as Hvorslev’s method. They concluded the general expression as: (2-13) In which, with respect to Fig ( 2-1) Au t_ = P0, the unbalanced water pressure between the ground and device at the beginning of the equalization test. Au tmt = Pt, the unbalanced water pressure between the ground and device at time t . f(t) = general equation of ground water pressure. k = permeability of soil v = volume factor of the device A = filter area In this method, as with Hvorslev’s method, the effect of volume change of piezometer during the equalization process on the water pressure in the soil around the piezometer has been ignored. For constant ground water pressure, f(t) = cte, the equation (2-13) coincides with the equation (2-7) deduced by Hvorslev, (1951). The validity of Hvorslev’s method was experimentally evaluated by Penman, (1961). The tests were carried out in a laboratory. Hydraulic, electrical resistance strain gauges, and vibrating wire piezometers were used. The testing apparatus was like a triaxial cell, a piezometer tip could be placed at the centre of the sample of 11 inches diameter and 17 inches long. A reconstituted London clay was used in the cell. A total pressure which was applied through a rubber envelope was kept steadily constant during each test. This is equivalent to an equalization test with a field piezometer which is installed in ground having constant pore water pressure. The importance of the effects of redistribution of pore pressure within the soil (stress adjustment) on the equalization of a piezometer was proved and it was shown that Hvorslev’s theory could be used to obtain 199.99 for a hydraulic piezometer and it cannot be used to estimate the pore pressure in a clay soil from piezometer readings taken long before the equilibrium has been established. Although the effect of compressibility of soil on the time lag had already been pointed out theoretically by Josselin de Jong, (1953), and experimentally by Penman, (1961)>Gibson (1963) investigated further the relationship between time lag, compressibility of the soil, and system flexibility, on the basis of Terzaghi’s consolidation theory (Terzaghi, 1943). Gibson indeed, introduced a very realistic and useful method for studying the equalization process. He considers an isotropic and homogenous soil having a non-fluctuating pore water pressure. A standpipe which is connected to a suitable porous element whose permeability is large compared with that of clay is considered, and it is assumed that this element can be replaced by an equivalent spherical element whose radius ”a” is very small compared with its depth ”D” below the ground surface. The associated boundary conditions are shown in Fig. (2-4) and briefly explained as follows: I - At a very appreciable distance from the element the ground water pressure will be unaffected by the test, u -*> Uq at r -*■ 00 Uq is the ground water pressure which is assumed to be constant. II - The initial ground water pressure, immediately after the pressure in the piezometer is lowered, at any distance greater than ’a’ is assumed to be Uq . u = u0, at t = o for r > a This type of boundary condition is only acceptable for equalization tests in which the pressure in the piezometer is lowered or increased manually. In practice, especially in compressible soils, sudden changes of ground water pressure due to meterological conditions is impossible unless an external load is applied. HI - The water pressure (during the equalization process) is continuous between the porous element and soil. u (a,t) = Ywht for t > 0 This is an important assumption which excludes Gibson’s method from the Hvorslev’s and Kallstenius & Wallgren’s methods. Indeed, the effect of the equalization process on the ground water pressure around the piezometer is considered. The stiffness of the measuring system was characterized by the dimensionless quantity ji and time factor T. 47ra3m p = ------v T = ct/a2 in which: m = coefficient of volume compressibility c = coefficient of consolidation v =A/Y w> volume factor of piezometer a = equivalent radius of intake point A = cross sectional area of standpipe. Considering the above mentioned boundary conditions, Gibson employed Laplace transforms to obtain the solution to the equation of consolidation in spherically symmetrical form, namely: , 02U 2 0U v Su c(e?r-7 W)=dT’r>a (2-14) and then introducing the ’’equalization ratio” 6 at any time as: E - iU O r O ^ " UA r - = f(M’T) ( 2 - 1 5 ) the curves correlating errors £ to pT and pT2 values were provided, Fig. (2-5). Gibson suggested the use of a curve fitting method in which the measured response time relationship is superimposed on the theoretical relationship to obtain the best fit and so determine the values of ji and T/t. Since its publication, Gibson’s important solution for pore pressure equalization had received little attention. It had not been verified by experimental data or numerical solutions until 1981. Neither had it been used for in situ measurements of ’k’: and ’c’ values, because it entails falling head tests, which are more difficult to interpret than constant head tests (Premchitt & Brand, 1981). Premchitt & Brand, (1981) presented verification and further extension of Gibson’s theoretical solution. They established response characteristics by direct numerical evaluation of equations (2-15) and also by means of a simple explicit finite difference scheme for different values of p, Fig. (2-6). It was shown that the results from the finite difference method were almost identical to those obtained directly from Gibson’s solution. Premchitt and Brand characterized the shape of the response-time curve by the ratio T 9 (/T5o, Fig. (2-7), (which is equivalent to a real time ratio t 5(/t9 o), where T 90 and T 50 are the time factors at 90% and 50% responses respectively. They also established an equalization chart, Fig. (2-8), which enables the soil properties to be determined directly and conveniently from the response-time measurements (for constant ground water pressure) without resort to curve fitting which was suggested by Gibson. They also showed that the laboratory equalization test results provided experimental confirmation of Gibson’s theoretical solutions for spherical piezometers and that falling head tests in piezometers provide a quick and accurate means of measuring in situ soil permeability. However, the coefficient of consolidation ”c” is overestimated, which can probably be more representative of the coefficient of swelling. The solutions which have already been reviewed have principally concentrated on the equalization of a piezometer in a ground having constant water pressure. The unbalanced pressure, between piezometer and ground, is caused manually by lowering or increasing the pressure in the piezometer and, of course, an abrupt discontinuity of pore pressure between piezometer and ground is assumed either during (e.g. Hvorslev, 1951) or at the beginning of the equalization process (e.g. Gibson, 1963). These solutions, specially Gibson’s method extended by Premchitt & Brand, (1981), are extremely important and appreciable in the sense of predictions of the ground water pressure (non-fluctuating) before the piezometer has equalized and also in situ estimation of the properties of the soil surrounding the piezometer. In practice, however, piezometers are basically installed to establish the ground water pressure pattern, and long term records with errors within a permissible range are expected. After a piezometer has virtually equalized for the first time after installation, the discontinuity of pore pressure between ground and piezometer virtually vanishes. Then the piezometer and surrounding ground operates as a whole body and indeed the service life of a piezometer starts. The pore pressure measurement practice may suffer from the errors which are caused due to fluctuations of the ground water pressure. Hvorslev, (1951) has simplified the problem and given a solution for sinusoidal fluctuating ground water pressures. In nature, however, equations of fluctuation of ground water pressure is not sinusoidal but a complex function of meterological conditions, topography and seepage patterns (e.g. the presence of natural or man made drains), especially the swelling and consolidation characteristics of the soil. The latter factor causes the ground water pressure to show even different equations at different depths. In Hvorslev’s method, however, ignoring the compressibility of soil, the same equation (same amplitude and period of fluctuation) is necessarily assigned to the ground water pressure at any depth. Besides, the time lag is also affected by compressibility of the soil (stress adjustment) and in Hvorslev’s method it has not been considered. Theoretical evaluation of errors of this type is veiy difficult, and probably impossible, because of the complexity and uncertainty of the equation of the ground water pressure at a proposed point. Some general viewpoints in connection with the errors in monitoring of the ground water pressure, caused by the changes of groundwater pressure are seen in Literature (e.g. Vaughan, 1974). A detailed intensive investigation which can give at least some typical values of errors are not available, however. Vaughan, (1974) has mentioned that adequate measurements can be made even with a standpipe piezometer of slow response, irrespective of the soil permeability, provided it is not close to a drainage boundary where rapid changes in pore pressure can occur. This is, of course, a general viewpoint and no experimental work or detailed elaboration is given. 2-2-5 Effects of changes of atmospheric pressure, Literature Review Although the effects of atmospheric pressure changes on the ground water level in aquifers is appreciated by many researchers, the difficulties and errors in measurement of ground water pressure, associated with changes of atmospheric pressure, have been left in uncertainty. However, parts of previous works, especially the feature of transmission of changes in atmospheric pressure through the layer lying between the water table and ground surface which could be of interest to the current research are reviewed and given here. Tuinzaad, (1954), has mentioned that in fine grained surface layers in which the soil above the phreatic surface is completely saturated up to the ground surface with capillary water, atmospheric pressure changes could affect the elevation of phreatic water. He did not give examples of the magnitude of fluctuation. If it can be ascertained that the atmospheric pressure changes are applied as a sudden external load to the ground surface then this view will be reasonable, because in a fully saturated clay soil any external load is borne by the pore water, A a = Au in short term. Turk, (1975), has attributed the mechanism of transmission of the effects of changes of atmospheric pressure on water table aquifers to the infusion or expulsion of air through the material lying between the water table and ground surface, and mentioned that a layer of low-permeability material lying between the water table and the ground surface and restricting the flow of air, may give rise to ’’blowing” or ’’sucking” wells. This means that the low-permeability material absorbs the loading changes and the water table, surrounding a well does not undergo pressure change, and the water table in the well itself is directly affected by the changes of atmospheric pressure. This unbalance pressure, between well and surrounding water may give rise to blowing or sucking wells. This view is, of course, against statics and principle of effective stress. He has also studied the effects of changes of atmospheric pressure on the water level in a shallow aquifer (depth to water table below land surface less than 2m) with a high permeability intervening layer, and concluded that the effect of atmospheric pressure changes, in shallow aquifers overlain by highly permeable materials, is reflected in the water level almost instantaneously. Weeks, (1979) has explained the feature of transmission of changes of atmospheric pressure through the soil overlying confined and unconfined aquifers. He has mentioned that the water levels in wells screened only below the water table in deep unconfined aquifers fluctuate in response to atmospheric pressure changes. He has attributed the reason to the point that the materials composing the unsaturated zone overlying an unconfined aquifer resist air movement and have capacity to store air with a change in pressure. Consequently, the transmission of any pressure change at land surface is slowed as it moves through the unsaturated zone to the water table, but it reaches the water surface in the well instantaneously. As time passes, however, the entire pressure change, as shown in Fig. (2-9-b) is transmitted through the unsaturated zone and the water level recovers to its initial position. It should be mentioned that, although the atmospheric pressure changes may be attenuated by air bubbles in soil, it should reach the water surface surrounding the well, instantaneously, otherwise the equilibrium condition is not fulfilled. In the case of confined aquifers, Weeks has given a substantially different explanation of the transmission of changes in atmospheric pressure through the overlying soil. As shown in Fig. (2-9-d), the pressure change is transmitted instantaneously without attenuation through the confining bed to the interface between the confining material and the aquifer. It is of interest to note that he mentions that at the interface, part of load change is borne by the confined water over the area of its contact with the interface (bAhw), and the other portion (ASk) is borne by the aquifer skeleton, Fig. (2-9-c). As the pressure change (AH0) is borne entirely by the water in the well an imbalance pressure equal to (AH0 - bAhw), results in an equivalent water level change in the well. This type of justification is, of course, against the principle of effective stress, because in the interface between the confining material and the aquifer the pressure is borne entirely by pore water and the aquifer skeleton does not bear any stress. Further observations, although very limited, in connection with the response of water tables, in standpipe piezometers and abstraction boreholes tapping aquifers, to the changes of atmospheric pressure are given by Walthall and Ingram (1984). The geology of the site is some seven metres of sandy clay overlying the Chester Pebble Bed Formation consisting of red-brown medium to coarse-grained well-cemented sandstones with occasional pebbles. The approximate natural ground water level is 32 m b.g.l. which is much below the overlying sandy clay layer. Eggboro and Walthall, (1985) have shown the response of two mutually remote boreholes in completely separate aquifers. They have not given details of the stratography and groundwater pressure pattern of the sites under investigation. From the similarity between the changes of atmospheric pressure and water table, they have concluded that the dominant effect on the ground water level in aquifers is barometric. They have mentioned that while the full effect of barometric changes is felt on the water in the open bore hole tapping an aquifer, the water in the aquifer itself will only feel a partial effect, as the structure of the aquifer will absorb part of the loading change. This thought is, of course, the same as the justification given previously by Weeks (1979) for confined aquifers. They have also mentioned that where a confined aquifer is overlain by a thick sequence of clay the ratio of the change in water level in the borehole to the change in atmospheric pressure may approach 100%. because the clay absorbs all the loading changes. This view is highly disputable in ground having a continuous pore pressure distribution with depth, because the loading change must necessarily be borne by the water in the aquifer, otherwise neither equilibrium condition nor principle of effective stress will be fulfilled. 2-2-6 Other sources of error. There are many other sources of error in measurement of ground water pressure. In this section, however, only the errors which are sensibly relevant to this research are briefly explained as follows: General instrumental errors. Several sources of error may be found in the design, construction and method of operation of the piezometer installations. In the case of open standpipe piezometer installations, the main sources of instrumental errors maybe inaccurate determination of the depth to the water surface in the piezometer, evaporation of water or condensation of water vapours and penetration of the surface water into the piezometer through the protective cap. The depth to water surface in standpipe piezometers is usually measured by an electrical dipmeter. When the probe enters water an audio signal is emitted. When the sensitivity of the probe is reduced due to dirt or oxidation, especially when the water in the piezometer contains less soluble, false measurements are recorded. Evaporation of the water in the piezometer always tends to decrease water level in the piezometer. If the rate of evaporation is slow its effect might be compensated by equalization of the piezometer. Part of the evaporated water escapes through the ventilation hole of the cap and the rest which is condensed, either accumulates in the cap or drops into the standpipe, which in this case partially recovers the effects of evaporation. Penetration of surface water or rainfall into the piezometer through the protective cap is another source of error. Probably most of the commercial caps suffer this difficulty. In spite of the fact that it is sometimes thought the air pressure entrapped in the standpipe prevents the water passing through the cap into the piezometer, during this research it was clearly observed that the water can penetrate through the relatively loose threads and the tiny ventilation holes, into the piezometer. In the case of the pneumatic installations the main source of error is the sensitivity of the transducer, Fig. (4-2), against the minimum differential pressure. The minimum differential pressure is the minimum pressure difference between two sides of the membrane which makes it operate. This error is augmented with the errors of readout unit and makes a total instrumental error. For a vibrating wire installation, errors in the measurement of the frequency of the vibrating wire and sensitivity of the piezometer transducer against the minimum pore water pressure changes are the main sources of instrumental errors. These piezometers are very sensitive to temperature changes and adaptation take a considerable time. In clay soils changes of temperature are very slow, but in soils of high permeability, like free drainage materials which are filled into the drainage systems, these piezometers may show false records, because the water temperature may considerably change. Basically, installation of such piezometers in drains is not recommended, however. - Seepage along conduits. Seepage along the casing piezometer tubing, or other conduits may take place, especially when irregular ground water conditions are encountered (Hvorslev, 1951). When a piezometer is installed, actually three types of materials of different properties (soil, bentonite-cement grout, piezometer tubing) are intimately interfaced. The soil is always deforming due to changes of ground water pressure while the sealing grout becomes crumbled and comparitively solid after it has set. If the water content of soil decreases due to hot weather and decreasing of the ground water level, cracks will appear between the seal and surrounding soil and this might be a good passage for surface water to penetrate into the intake point and/or disturb the pore water pressure pattern in the soil adjacent to the piezometer. The depth and width of the cracks depends on the properties of the soil and meterological conditions. This difficulty may be minimised if piezometers are installed in Summer, but it should be noticed that subsequent swelling of the ground when the wet season starts can cause tension in the seal and piezometer tubing. As a case record, standpipe piezometers (J4, J5, J6, &7) were installed in Spring 1986; in the following Summer the protective caps stayed about one inch above the ground level because of shrinkage and consolidation of the ground. To protect and strengthen the caps against grass cutting machines the gap between the ground and the concrete holding the caps were filled with sand-cement mortar. In the following Winter when the soil had swelled again about a 10 mm slide had occurred between the standpipe and the seal. - Effects of gas bubbles in the system and/or soil. Air or gas bubbles in an open standpipe piezometer affect the time lag and cause the stabilized water level to rise above actual piezometer level. So, it is recommended that the pipe should be smooth, and downward protruding edges or joints should be avoided (Hvorslev, 1951). The diameter of the standpipe should be large enough to cause bubbles filling the cross section to rise to the surface, and for this reason it is not possible to use very narrow pipes to minimize the volume change of standpipe piezometers. Air or gas bubbles in the soil are accumulated around the ceramic (porous stone) of piezometers. In the standpipe piezometers having low air entry ceramics, the bubbles pass through and rise to the surface. If they have high air entry ceramic the bubbles increase the response time and influence the equalization curves. In the case of closed diaphragm systems the air or gas bubbles penetrate gradually through the ceramic into the piezometer. The piezometer gradually tends to show the air or gas bubbles pressure, which is higher than the water pressure. The use of thick high air entry value filters, for lengthening the diffusion path, may help the system to operate without breakdown for many years if pore water pressures above atmospheric are measured (Vaughan, 1974). However, in partly saturated soils or soils having gas bubbles, systems having de-airing facilities are usually used. G.L. t=t Uo — 1=0 Fig (2-1) Piezometer equalization, constant groundwater pressure (Hvorslev, 1951) G.L. S-G .W .L. t=t «t S-G.W.L. t = 0 Ut _ - t = t Uo —t = 0 u Fig (2-2) Piezometer equalization, linearly changing ground water pressure (Hvorslev, 1951) Tw = WAVE PERIOD t 5 = PHASE SHIFT X PIEZOMETER-STEADY STATE'-' CROUND-WATER PRESSURE Fig (2-3) Piezometer equalization, sinusoidal fluctuating ground water pressure (Hvorslev, 1951) G.L. G.W.L. t=0&oo U a .t U a ,o t = 0 Fig (2-4) Piezometer equalization, pore water pressure distribution around an ideal spherical piezometer tip in a ground with a constant water pressure (Gibson, 1963) 1*0 0 -9 0-6 e 0 -7 0-6 •jj.lO 0-6 0 -3 0-2 0 - 0 01 OOI 0-1 IO PT 1-0 0 -9 0-6 06 - 0 4 02 IO 0-1 OOI 0-1 IO ICO Fig (2-5) Variation of degree of piezometer equalization with pT and pT2 (Gibson, 1963) 100 r 80 60 40 20 0 10'6 1 10 7 *= cf/a2 Fig (2-6) Theoretical response-time relationships obtained by Premchitt & Brand (1981) from Gibson’s (1963) solution. 100 s o»o O £ 10 Fig (2-7) Relationship between p and the ratio T 90 /T50 (Premchitt & Brand, 1981). 4jrak/Vy( Fig (2-8) Equalization chart derived from Gibson’s theory Gibson’sfrom theory derived chart Equalization (2-8) Fig (Premchitt & Brand, Brand, 1981). & (Premchitt 10*3 F 8 0 000 0 8 F 100000 50 000 50 * O A 20 000 8000 10000 s 5000 a h . Land surface Ahf AH Water table A A I A__A___ Well screen a Confining material B arom etric pressure Time AWL Water level surface 0 = Potentiometric surface Confining material 1 * “ •' / See ’• 1 2 Z - ' y Inset.' . V . . . : : ..••.••..•;Aquifer;.-/.‘- •: .".v.-.-: I • • ** .*.* I : r Confining material Barometric pressure 1 AH0 Inset pressure Increasing ! Time ------— c» d .£ o «/> t .AHn-1 b A h w £ • § 3 Water level Fig (2-9) Effect of a change in barometric pressure on the water level in a well tapping unconfined and confined aquifers.(Weeks, 1979). (a) Idealized section of an unconfined aquifer. (b) Idealized barograph and hydrograph showing water level response with time. (c) Idealized cross section of a confined aquifer. (d) Idealized barograph and hydrograph showing water level response with time. CHAPTER THREE - STAG HILL 3-1 INTRODUCTION. Stag Hill is an example of a slope failure in London clay. The fact that such a suitable site so close to a tourist centre like Guildford remained undeveloped until 1964 is due entirely to a geotechnical reason, namely, that a landslip had taken place here on the slopes of Stag Hill (Simons, 1977). In connection with the age of this landslip, estimations are available in literature. Certainly it must have occurred more than 130 years ago, since a tree 110 years old was cut down on the back scarp in 1966. The scarp, however, is still a clearly expressed feature which is unlikely to be more than a few centuries old, and there are signs of fresh movement at the toe (Skempton & Petley, 1967). Probably a landslip took place here some time during the 19th century (Simons, 1977). In this chapter a brief and clear picture of Stag Hill is given from different aspects. A general view of the site with all developments since 1965 until 1987 is shown in Fig. (3-1). The topography and geology of the area is firstly explained and then the history and objective of site investigations and remedial works reviewed. Folowing this, the history of piezometer installations and the ground water pressure records since 1965 will be given. Finally, the London clay will briefly be defined and reviewed and its engineering properties explained. 3-2 TOPOGRAPHY. The topography of the site is shown in Fig. (3-2). The Northern side of the Hill on which the main part of the University is built is remarkably uniform varying from about 8° to 10°, except where a scarp has been exposed by slipping. The positions of the toe and back scarp of the slip are also shown in this figure. In Fig. (3-3) the limits of areas with slope angles equal to or greater than the figures shown on the border lines are shown (Clayton & Hodder, 1981). In both Figs. (3-2) and (3-3) it is seen that the slopes become gentler as they move further to either the East or West. 3-3 GEOLOGY OF THE AREA. The University site lies at the Southern edge of the London basin at the junction of the Eocene and Upper Cretaceous beds. With respect to the geological map of the area (Sheet No.285) it is observed that the University area and also the Southern side of the Hill are underlain by London clay and then in turn towards the Southern toe of the slope Reading beds and upper chalk outcrop. With respect to the direction of inclination of these layers in this area, which are tilted to the North, it can be concluded that the whole site is underlain by London clay and this is in turn underlain by the Reading beds, and then chalk. An appreciation of the borehole logs of the area (Northern side of the slope) shows that the depth of weathering (oxidation), shown by brown clay, is not the same all over the area and varies from nearly 15m in the upper part of the slope to about 9m at the bottom and the flat area. The material between the blue London clay and the surface, weathered layer, is not uniformly firm brown clay. The presence of pockets of sand, patches of silt, and even deposits of gravel are mentioned in Borehole log reports. Toward the top of the Hill undisturbed brown London clay, overlying the blue clay, extends to ground surface. Close to the top of the slope, and beyond, the undisturbed clay is overlain by about 6m of a quite distinct brown, mottled, sandy and silty clay. Confirmation that this upper zone was redeposited was obtained from a trial pit more than half-way up the slope where fragments of chalk were found at a depth of 1.5m. It is unlikely to be more than about 2m thick at this elevation (Gibson, 1964). A study of borehole logs, for site investigation, piezometer installation and piling works, shows that many^ boreholes with depths varying from 2 to 45m have been drilled in this area since 1964, and none of them have proved the chalk, although the Woolwich-Reading beds were encountered at 39m below ground level (communication from M.C. Wicks and also S.A. Bloss, 1986) when the 45m deep borehole was drilled to install piezometer No.701. This indicates that the distance of chalk from the ground surface increases steeply from the South toward the Northern side of Stag Hill. 3-4 HISTORY OF THE SITE INVESTIGATIONS AND REMEDIAL WORKS After the proposal to construct a new University at Guildford, Surrey, the first subsoil investigation, consisting of seventeen borings and soil tests was carried out between 9th May and 7th July 1964 (by Foundation Engineering Ltd). The boreholes were sunk to between 4.5m and 20m below ground level, using standard shell and auger equipment. The positions of the boreholes are shown in Fig. (3-4) and the appropriate parts of borehole logs are summarized in section 8-3-1. During the boring operations the depth at which water entered the boreholes was recorded. These operations gave no reliable guide to the elevation of the water tables or to the water pressures to be expected at any point in the clay. They merely indicated the depth at which a major fissure was encountered or a locally more pervious horizon penetrated. The level to which water rises in a borehole left open for a number of days provides a better estimate of the elevation of the water table, but this could still be seriously in error (Gibson, 1964). Water level indicators were therefore installed at two levels in both boreholes 7 and 8, and the water levels, recorded some weeks after installation when equilibrium conditions should have been attained, were as follows: B.H. No. Depth of Indicator Water Level Below G.L.(m) Below G.L. (m) 7 3.35 2.56 7 12.80 2.56 8 3.35 0.91 8 12.80 0.24 The very low water level indicated by the deeper indicator in borehole 8 was considered to be unreliable and ignored. The other readings were typical for the water level conditions in London clay slopes during the summer months (Gibson, 1964). In a trial pit at the toe of the slope, a thin film of water was seen to cover the exposed shear face which was found at a depth of 0.7m below ground level. The water was therefore assumed to be above this level and probably very close to ground level. Borehole logs and details of the laboratory test results were presented in Report No.F.69/882 August 1964, Foundation Engineering Limited. During this period, a preliminary inspection of the site in June 1964, suggested that a major slip had occurred there in the past. Evidence of a slip was particularly clear in the field to the North-East of Guildford Cathedral, where a number of hawthorn trees had become established along the line of the slip scarp and along the line of overthrusting at the toe. Aerial photographs taken subsequently showed that these scarp and toe lines extended to the West into the field immediately to the North of the Cathedral. Two of these photographs are shown in Figs. (3-5) & (3-6). In Fig. (3-5) the back scarp and in Fig. (3-6) the toe and back scarp are clearly shown. In November 1964, Professor Gibson evaluated the stability of the area with respect to the results of the above mentioned investigation and introduced remedial proposals. For stability analysis, the slope was treated as an infinite slope, and stability calculations were carried out in terms of effective stress. The soil parameters were considered to be as follows: - Residual strength parameters, c r = 0 and o'r = 14° - Soil unit weight (saturated) y = 20 kN/m3. No field evidence was available to show at what depth the slip surface was actually located. So, to estimate the average depth of slip surface, the value of slope inclination (6) before the slip occurred was inferred to be 10°. The depth of water table at the time of failure was assumed to be 0.9 m b.g.l and the average depth of the slip surface (assuming F.S. =1) was calculated to be 2.25 m b.g.l. At that time (1964) the average slope was about 9° and the factor of safety was calculated to be F.S. = 1.1. Therefore it was concluded that any additional loading on the surface of the slope, either during the construction operation or on a long term basis, was very likely to lead to further slipping. The only economically viable method of founding buildings on the slipped area was to carry their load by means of piling to an appreciable depth below present ground surface. If this measure was adopted the stability of the slope itself would have been in no way improved, and if slipping did occur these piles would experience very considerable lateral loading. It was very difficult to be certain that any piling, however heavily reinforced against bending and shear, would be able to withstand that loading. It was considered essential, therefore, to increase the stability of the slopes before any buildings had commenced. This could be achieved in two ways: I. by regrading the slopes to a flatter profile or II by lowering the ground water table permanently by providing a system of deep drainage trenches backfilled with gravel. The first possibility had to be rejected because of the proximity of the Cathedral and because of the very heavy cost likely to be involved, and so the drainage system was adopted. It was assumed that the drains would reduce the pore water pressure to zero on the slip surface. Using this assumption (hw = 0)and the parameters mentioned above, it was computed that a 10° slope on the site would have a safety factor of F.S. =1.42. Professor Gibson suggested a drainage system with trenches 5m deep 0.75m wide, and spaced at about 12m centres for the area where slipping is extensive. To the West where the slopes are slightly flatter, trenches of 3.6m deep, 0.75m wide and spaced at about 18m centres were suggested. In order that the present slope should remain stable it was proposed that the weight of the heavy buildings should be carried by bored reinforced piling isolated from the top 3m of soil by provision of a sleeve, down to an adequate depth (Gibson, 1964). In 1965 a further site investigation was carried out on the site of the proposed University. The object of this investigation was to provide further information on the ground conditions at the site that would assist the engineer in his design of the foundations. The site work consisted of 6 borings and 10 trial pits. The borings were carried out between the 15th and 26th June, 1965. The 6 inch (150mm) diameter borings were put down to depths of 18. 7m using standard shell and auger equipment. The results are described in Report No. F6a/371, August 1965 by Foundation Engineering Limited. The positions of the boreholes and trial pits are shown in Fig. (3-7) and the borehole logs are appropriately summarized in section 8-3-1. Further investigations were carried out in 1965-66 (Skempton & Petley, 1967) and the slip surface was located through a section. This section, the Section A-A, is positioned on Fig. (3-2) and Fig. (3-3) and the longitudinal profile is shown in Fig. (3- 9). In Fig. (3-8) the positions of some boreholes are shown on the topographic plan. The main slip surface, which was highly polished, could be seen in each of the three exploration pits. A thin section examined under the microscope (Morgenstem & Tchalemko, 1967) showed that the slip surface consists of a band roughly 50 microns wide in which the particles are strongly oriented. It is located at or near the lower boundary of a shear zone about 6mm in width containing numerous minor shears, Fig. (3-10). Outside the shear zone the degree of particle orientation was slight (Skempton andPetley, 1967). During these investigations it was found that the disturbed zone associated with the landslide movements had a maximum depth of about 8m. The final conclusions at that time, September 1965, were that drains 5m in depth spaced at 15m centres should be installed over the slipped area. The material used for filling the drainage trenches was proposed to be from natural sources and to be clean, hard, stable and have a grading within the limits specified on the grading chart shown in Fig. (3-11). At the present time the drainage system of this area has been constructed in five phases. Phase I itself consists of two parts. The drainage trenches of each phase or part, lead to a ’’collector drain” and in turn to the sewer. Five phases of the drainage system are shown in Figs. (3-1) & (5-2). The drainage system construction work was commenced in December 1965. During this month a trial slope drain, Fig. (3-1), and then between March and June 1966 the remainder of the phase I (parts 1 & 2) were constructed. For further elucidation about the drainage system, details of the trial trench are shown in Fig. (3-12). Additional slope drainage systems, phases II and III, were constructed during the following years, up to 1969, and the slope drainage system phase IV was constructed in 1970. In 1972, the drainage system phase V was constructed to lower the pore pressure in the ground on which Wates House has now been constructed. This phase is different from the others in that the trenches were constructed at about 6m centre line spacing. The reason was that, the field measurements of piezometric heads in the previously drained area, suggested that for drains 5m deep spaced at 15m centres to the factors of safety in the steeper areas would be unacceptably low, e.g. for a=9°, F.S. =1.11, (Report No. 3771/DJH/TRS, OVE ARUP & PARTNERS with contributions by Professor N.E. Simons). Professor Simons has carried out a long term study on the pore water pressure pattern and stability of the area (e.g. Inaugural Lecture 1977, and Report No. 3771/DJH/JRS,OAP and also Simons et al, 1987). The later study of characteristics of the area was carried out by J.P. Hodder, in March 1981, under the supervision of Dr. Clayton (MSc. thesis, University of Surrey, 1981). In this study, limits of areas with different slope angles and probable order of land slip events were determined using aerial photographs. These are shown in Fig. (3- 3) and (3-13) respectively. The line A-A on Fig. (3-3) shows the alignment of the slip surface which was located in 1965 by Skempton. In May 1986, a site investigation was undertaken for the proposed Centre for the Performing Arts building. Five trial pits were dug and it is interesting to note that in trial pit No.2, Fig. (3-14), over 20m to the North of the toe of the slip, where the ground becomes more or less flat, a slip surface was observed at about lm below ground surface. The position of trial pits and the view of slip surface are shown in Figs. (3-14) and (3-15) respectively. 3-5 HISTORY OF PIEZOMETER INSTALLATIONS AND PORE WATER PRESSURE RECORDS All of the piezometers which have been installed in Stag Hill over the past twenty two years are standpipe piezometers. Positions of the piezometers are shown in Fig. (5-2), and characterized in table (6-1). In October and November 1965,19 piezometers were installed. The 101 to 112 and 201 to 207 series, and records were taken, but most of these piezometers were soon lost during development of the area and construction of the drainage system. The following list shows when piezometers were lost: InMarch 1966 piezometer Nos. 103,104 & 108. In June 1966 piezometer Nos. 101,102,109,110 & 111. In August 1966 piezometer No.205 In October 1966 piezometer Nos. 105, 106, 107, 112 & 201. In April 1966 piezometer No. 204 In April 1967 piezometer No. 203 In March 1968 piezometer No. 204 In September 1968 piezometer No.206 In December 1968 piezometer Nos.202 & 207. In January 1966, piezometer No.208 was installed but was lost in June 1966. In December 1966, 6 additional piezometers were installed. Piezometers A,B,C and D, on a line down the slope West of Wates House and through what is now the Leggett Building, and 201 and 204 were reinstalled. Piezometers A,B,C and D were in ground not affected by drainage measures at that time, so that the readings there could be compared with those affected by the drainage system. There is no knowledge about the depths of these piezometers however. The following list shows when the above mentioned piezometers were lost: In September 1967 piezometer 201 In March 1968 piezometer 204 In December 1968 piezometers A,B,C and D . In March 1967, 17 new piezometers were installed, the 401 to 417 series. The following piezometers were lost or blocked as shown: In September 1969 piezometer 408,411 & 414 In February 1970 piezometer 409 In January 1971 piezometer 405 In December 1972 piezometer 410 In April 1976 piezometer 407 In September 1976 piezometer 413 In December 1976 piezometer 412 In April 1986 piezometer 403 In April 1979, 14 piezometers, the 501 to 514 series, were installed to measure the pore pressure pattern in the drained ground, where Wates House was constructed. No data are available on piezometers 512, 513 and 514 so far. In February and November 1982 two other piezometers, 601 and 602 respectively were installed, followed by six piezometers, the 701 to 703 and KP1 to KP3 series in December 1985. These series are along a straight line a few metres to the North of the BB Building, all except 602. Piezometers KP1 and KP3 differ from the others in that while all other piezometers have been backfilled with standard materials (i.e. sand, bentonite plug, bentonite-cement slurry), these two have been fully backfilled with clay and cement slurry respectively. Of the 63 abovementioned piezometers, only 26 survived to 1986 (Simons et al, 1987). 3-6 ENGINEERING PROPERTIES OF THE SOIL 3-6-1 London clay, general definitions. London clay is an ’overconsolidated’ clay, having been subjected at one time to considerable overburden pressures from other sedimentary deposits. The removal of these pressures, by geological processes which ease the internal stress pattern, produces a reduction in strength. Slopes in this material tend therefore to fail until a residual strength condition has been reached. The presence of discontinuities in London clay is well known, and to the Civil Engineer the importance of discontinuities and their effects on the strength parameters of the clay cannot be questioned. London clay, in common with other sedimentary rocks, contains significant discontinuities in the form of bedding surfaces and joints (Skempton et al, 1969). If it has been sheared by land sliding of tectonic forces, additional discontinuities will be formed in the shear zone. It has been shown that the shear zone consists of minor shears of limited extent and, usually, one or more principal slip surfaces (Skempton, 1966). The Riedel, thrust and displacement shears which appear during the shearing process can collectively be grouped under the heading of minor shears (Skempton & Petley, 1967a). The small non-systematic joints are known as fissures and their significance and effects on the strength of clays has been recognised since Terzaghi (1936). Skempton et al (1969) have mentioned that the fissures in London clay within the top 9-12 meters below ground level are planar or conchoidal fractures, rarely more than 15cm in size with a matt surface texture. Such fissures are probably a consequence of the high brittleness index and swelling characteristics of clays and clay shales of high plasticity index (Bishop, 1971). In connection with the effects of fissures on the mass permeability and seepage characteristics of the clay, it is, of course, true that the fissures can provide a path to moving ground water and in turn provide greater permeability than the intact clay, but beneath ground surface the high lateral pressures (Skempton, 1961) tend to keep the fissures closed. That the fissures do not affect the insitu permeability of the soil is supported by the available field evidence which indicates that the rate of pore pressure equilibration in cuttings, despite the fissured structure of the clay, is very slow. Pore pressure equilibration time lags in the order of 50 years are seen in the literature (Skempton, 1977). 3-6-2 Mechanism of failure When an over-consolidated clay is subjected to simple shear five successive stages can be recognised as the deformation increases (Skempton, 1966). These stages are briefly defined as follows and shown in Figs. (3-16) and (3-17). I In the first stage, before the peak is reached, continuous non-homogenous strain occurs. II In the second stage, at or just before the peak, ’’Riedel Shears” at an inclination usually between 10° and 30° to the direction of general movement, and conjugate shears are sometimes formed. HI In the third stage, at which slip along the Riedel shears is no longer kinematically possible, new slip surfaces parallel or sub parallel to the direction of general movement are developed. These are the ’’displacement shears”. IV In the fourth stage, which is reached following greater movements, some of the displacement shears link up to form a ’’principal displacement shear” or ’’slip surface”. ’’Thrust shears” typically inclined at about 160° to the direction of movement, also tend to develop in the third and fourth stages. V In the fifth stage, the slip surface is appreciably flattened due to further movements. The development of shears in clays is accompanied by particle orientation more or less in the direction of movement (Morgenstem & Tchalenko, 1967). In a flattened principal slip surface, in which the maximum particle orientation has attained, the resistance to shear tends to be the minimum possible value, and this is the ’residual strength’ of the clay. If a test is carried out on a flat slip surface with full particle orientation, (with the same characteristics of movement in the test and nature) a stress-strain curve of type 1 in Fig. (3-17) is expected. There will be no peak and with a small displacement the residual strength (Sr) will be achieved. In practice, however, the tests often show a small peak, although with further displacement the strength soon falls to the residual, curve 2 in Fig. (3-17). The difference between the peak and the residual strength is denoted by AS. The reasons for AS being greater than zero in tests on principal slip surfaces are explained by Skempton and Petley (1967a). 3-6-3 Strength Parameters 3-6-3-1 Peak strength parameters The standard shear box (6cm) tests on specimens of clay, taken several centimetres away from the shear zone (Stag Hill 1965-66), showed peak strength parameters as follows (Skempton and Petley, 1967a), see Fig. (3-18). c = 16 kN/m2 a =20' Further data in connection with the effects of specimen size on the peak strength parameters of brown London clay are given by Skempton (1977). These parameters as measured in 6 cm shear box tests or 38mm diameter triaxial tests, are: c = 14 kN/m2 0=20’ Tests on 250mm diameter triaxial samples, which could include a representative assemblage of fissures, gave the following results: c = 7 kN/m2 0 = 20* Skempton, (1977) has concluded that even with results obtaining from the.large sample (250mm diameter) the strength is around 30 to 70 per cent in excess of the field values obtained from back analysis, and so some progressive failure mechanism appears to be involved. 3-6-3-2 Strength parameters relevant to first-time slides (fully softened). The fully softened strength parameters of an over-consolidated clay c s and 0 S are equal numerically to the peak strength parameters of the normally consolidated clay and present a practical approximation to the critical state (Skempton, 1970). In the critical state, in a drained test, any further increment in shear distortion of a saturated clay does not result in any change in water content. It appears that the displacement which reduces the strength towards the fully softened or the lower limit of fissure strength is a complex resultant of minor shears such as the Riedel, thrust and displacement shears, but no principal shear surface has yet been formed (Skempton 1966, 1970, 1977). The shear strength parameters of the brown London clay relevant to first-time slides have been given as follows (Skempton, 1977): c's = lk N /m 2 0's = 20" 3-6-3-3 - Residual strength parameters: Fourteen tests were carried out on the slip surface specimens of Stag Hill in 1965. The best linear fit to the fourteen residual points Fig. (3-18), gave the following parameters (Skempton & Petley, 1967a): cr=2.5kN/m2 o'r= 12° and approximately the same results were achieved with reversal shear box tests on the specimens away from the shear zone. The residual strengths as measured on the Guildford specimens were plotted together with points obtained from tests on samples taken from other sites. Then, considering the non-linearity of the residual strength failure envelope, for normal effective stress range 30-60 kN/m 2 the reasonable linear approximations were: c r =1.4kN/m2 , 0r = 13° or c r=0 o'r= 15° and for pressures between 60 and 240 kN/m 2 the approximate values were: c'r = 3.5 kN/m2, 0r = 11.5° Of the above mentioned sets of residual strength parameters the parameters c r =1.4 kN/m2 and 0 r =13° have received more appreciation, because they agreed well with the analysis of post-slip movements at another area in brown London clay (Skempton, 1977). Skempton compared all available residual strength parameters. A summary of this comparison is given in Fig. (3-19) and Table (3-1). It was concluded that the back analysis of reactivated landslides and slip surface tests (shear box), at the relevant effective pressure, both give the field residual strength (Skempton, 1985). He also pointed out that for most clays the relationship between residual strength and normal effective stress is non-linear, but for design purposes it is often useful to take a ”best fit” linear envelope over the range of pressures involved, in the form S = c + a tan o' Table (3-1) Comparison of 14 case records between back analysis of reactivated landslides and slip surface test results (Skempton, 1985). ______ Parameter Angle of A 0/ 0 :% shearing resistance: deg Mean 0 from analysis 12.8 Mean 0 from tests 13.4 Mean A0 ±0.6 +4.5 Standard deviation in A 0 ±1.2 ±9 Maximum A 0 +2.5 + 17.5 Minimum A0 -2.2 -17 With respect to Table (3-1) it can be concluded that the most reasonable o r for design purposes is 13°. It can be observed in Fig. (3-19) that points corresponding to the Guildford area appear to be a little above the best fit line. Accordingly it may be reasonable to accept a c r value greater than zero for this area. The c r value corresponding to o r = 13°, as given previously, is c r =1.4 kN/m2. 3-6-4 Index properties The average values of the index properties of the clay at Stag Hill are as follows: - Liquid Limit (LL) 83% - Plastic Limit (PL) 32% - Water content (samples away from the slip surface) 33% - Liquidity index (LI), corresponding to 33% water content +0.02 - Water content (in the shear zone) 34% - Clay fraction (samples away from the shear zone) 55% - Clay fraction (in the shear zone) 57% - Activity 0.9 - Unit weight 18.9 kN/nP liversity of Surrey site plan showing the drainage system and positions of the piezometers Fig (3-1) Fig o o ro 00331 0 10 8 o I. in ro o C4 to Q.O) ARY/ ■RES'TAURANI BAGHjSeARg = J-1 2 5 0 Fig.(3-2 CD CD i_ CD CD X) XJ LO CO *© Cl A Q) CD C o O) O o \ CD CL 0 (/) o JZ cn * } ■"* *o a CD 1__ o l d o co M—o CO l------CDa £ CD i— Lf) 1— co. \LH CO / o 00 CO o o Fig (3-3) Limits of areas having slope angles equal to or greater than figures shown on the border lines (Clayton and Hodder, 1981). mm mmrv'.&j&teto-; :*->&•$ J&K • \ m c-V; S «*!{••■» f e • • mm S ta s : • •t.t? - .. • ■ * 1 : I/; -V&54 H > , : \ : : i : «?H 8 ft Fig (3-4) Locations of boreholes, 1964 Fig (3-5) Air Photograph of the Slip Fig (3-6) Air Photograph of the Slip (back scarp) i 37 Lctoso oeoe n ra is 16. = 1/3520 2 5 1 3 =»/ Locations S of boreholes and trial 1965.pits, Fig (3-7) a# g. Fig (3-8) Locations ofboreholes on Topography plan, 1965 S «* 1 / 1 7 6 0 C t C O \ O O CD I Q O O CD O C7 > CL (D CL JO o E inE ~o c D O t— CD o Fig (3-10) Details of shear zone, - 4 — > CD 04 a O Pit 11 (Skempton & Petley, 1967a) CD o CO <4— CD L_ L_ D CL C/1 O O £ O O O CD UD CO < Fig (3-9) Longitudinal profile through landslide (Skempton & Petley, 1967a) OOQOOOOO a o a CD cd c d i n ^ cd c\ j . -—* m U !N 33d3d Cv<°«? a a c k 'ri'.u Qgfluuuag Sgctgn A-A pg. ?Q-!a^gfe=ga-^.a- s? mm?. m s s s m i SirI1 • , r ~ l:5i i L^srsaji&ge Section A -A Through Manhole *j£.Ta£C v->ov->.M s^.ST-t-gg-.SJi’C gfc^- Pton E*E &c*,e c-c Fig (3-12) The trial trench sections and details. "D CD - a CL CO C I— D D c O JZ CD > c r c o CD CL t- i_ h CL O t- o [- n To CO - "U CD H c CL O CD a X > if) CD D > o o JZ C o D CO o o CD "O -X a CL < < < < w CD i_ > < O c_ < < < ^ r v ^ ) $ a t x H r - * ^ Q) CQ x : < < < < CO > t JD < < o o Fig (3-13) Probable order of landslip events (Clayton & Hodder, 1981) Pit. 1 1. 5 me t e r deep Pit 2 At 1.0 m b. g. 1. an el ip eurfaee was expaxed Pit 3 1. 5 meter deep Pit 4 4.5 metere deep 7he-oniginal ground ourfaoe wae exposed at 2. 0 m b. g. 1 Pit 5 1. 5 motor deep Pit 3 Pit 2 LJ □ ED'' Sports Pit 4 Pit 5 TOE <§? S L ip a Teaching Block Court Phase IV piezom eter centreline of drainage trench limits of slip surface Fig (3-14) Locations of trial pits (1986) Fig (3 -1 5 ) View of Slip Surface in Pit 2 1386 Stage (ii) Stage (iii) Stage (iv) Thrust(P) Riedet (R) — Displacement (D) Fig (3-16) Successive stages in the development of a shear zone in clay, from laboratory tests (Skempton, 1966) peek strength of ‘intoct* cloy s s Shear stress test on ‘inloct* cloy residucl strength = s t tests on discontinuities 0 Displacement Residual factor on discontinuity R • 1 - .^ s- ■ sf - sr Fig (3-17) Stress - displacement in tests on a discontinuity and on intact clay (Skempton and Petley, 1967a) 20 Brown London Cloy LL = 83 PL- 32 ‘intoct1 w = 33 CF - 55 15 slip w = 34 CF - 57 10 psi 5 0 10 20 30 A s psi psi Fig (3-18) Guildford; test results. Block 2 and k (Skempton and Petley, 1967a) London Clay LL = 80 PL = 29 CF = 55 o Guildford Tests a Dedham k r» Sudbury Hiii 100 on -v Walthamstow analysis { ♦ Hadleigh surface ° Wsro"en Point I 9 Herne Bay ( &c Wraysbury 50 0 ^ = 12-3* 100 150 200 250 300 Normal effective stress a': kPa Fig (3-19) Field residual strength for London clay (Skempton, 1985) CHAPTER FOUR - DETAILS OF THE MEASURING SYSTEMS , AND LABORATORY WORK. 4-1 STANDPIPE PIEZOMETERS. 4-1-1 Details of piezometers The piezometer tip used for installation is a Casagrande porous plastic piezometer tip comprising a cylindrical element protected by perforated rigid plastic tube with plastic end fittings to fit p.v.c. standpipe. For non drive-in piezometers, heavy gauge rigid p.v.c. tubes with both ends threaded, were used. The characteristics of the porous element and standpipe are summarized in the following table: Porous Stand element pipe Nominal internal diameter (mm) 19 19 External diameter (mm) 32 - Length (mm) 300 300 Overall weight (gr) 150 - Pore diameter (micron) 60 - Permeability (cm/sec) 3 x 10-2 - It should be mentioned that the internal diameter of p.v.c. pipes was not found to be 19mm and up to 3mm tolerance was observed. So, for permeability tests the measured diameters were used. For drive-in piezometers, a Casagrande drive-in tip comprising a porous element the same as mentioned above, protected by a perforated galvanised mild steel pipe with cadmium plated mild steel end cone and with coupling to fit galvanised mild steel standpipe tubes with 19mm internal diameter were used. Details of p.v.c.(non drive-in) and drive in piezometers are shown in Fig.(4-1). 4-1-2 Preparation for installation Special attention was paid to the fact that the standpipe piezometers are susceptible to leakage through the connections. Accordingly, all threaded couplings were sealed with P.T. F.E. tape and to investigate the tightness of the connections against leakage, a simple test was carried out as follows: Three p.v.c. pipes were coupled to each other, using the threaded connectors and P.T.F.E. tape with the same standard which is usual in practice. Then one end of the coupled pipes was plugged tightly and the other end connected to a public water supply tap using a suitable connector and a hose pipe. The system was de-aired by loosening the plug at the other end and then left under water pressure. After one hour the connections were inspected and no leakage was observed. The tap water pressure was measured to be 45 p.s.i. by the Thames Water Authority. The piezometer tip was soaked and saturated in de-aired water for at least 24 hours. At the time of installation it was coupled to the standpipes, installed in the hole or driven into the ground. 4-1-3 The Readout Unit For measurement of the depth of the water level a dip meter comprising a graduated flexible co-axial cable mounted on a winding drum, an audio signal indicator and sensitivity control fitted with an 11mm diameter brass probe, was used. When the probe enters water an audible signal is emitted from the drum. The dip meter is shown in Fig. (4-1). 4-2 PNEUMATIC PIEZOMETERS. 4-2-1 Details of piezometer. Details of the piezometer tip are clearly shown in Fig. (4-2). A commercial pneumatic piezometer tip for low and medium pressure (up to 300m head of water) comprising a brass transducer body, Viton diaphragm with two brass connectors to receive twin tubing was used. The transducer was fitted with a ceramic clamped by a brass nose cone threaded onto a central spindle. Viton is a special mineral oil resistant rubber stable in high temperatures, and with memory (100% elastic). The maximum volume change of the tip, due to the deflection of the diaphragm is less than 0.03cm3 according to the manufacturer. The characteristics of the piezometer tip and the relevant ceramic are summarized in the following table. Piezometer Ceramic Tip External diameter (mm) 38 38 Overall Length (mm) 270 127 Weight (kg) 1.2 Nominal accuracy (meter head of water) ± 0.2 Nominal maximum volume change (cm3) <0.03 Permeability (cm/sec) 2x10-6 Pore diameter (micron) 1 A twin Nylone 11 tubing comprising twin 1.9 mm internal diameter, 3.2mm overall diameter, sheathed with 1mm thick polythene was used. According to the manufacturer the burst pressure of the twin tube was 13.8 MPa and its weight was 3.6 kgs per 100 metres. For more information about Nylone 11 refer to Dupont Ltd. To connect the piezometer to the readout unit, self sealing and quick release brass couplings for pressure up to 6.5 x 103 kpa were used. 4-2-2 Preparation for calibration and installation. In preparing the pneumatic piezometers, special attention was paid to ensure the tip was as de-aired as possible and that the connections were airtight. The sequence of preparation for calibration and installation of a pneumatic piezometer was as follows: ______Firstly, to saturate the ceramic it was kept in de-aired water for five to seven days. Then for more safety the ceramic was put in a container filled with de-aired water and boiled for one hour and subsequently left under about 50 KN/m2 pressure flow of de-aired water for 2 hours so that any remaining air bubbles could be pushed out of the ceramic. The device which was made to de-air the ceramic using water pressure is shown schematically in Fig (4-3). To apply the de-aired water pressure into the ceramic it was tightened on to the central spindle using the bearing nut and two seals at both ends of the ceramic. To de-air the cavity between the ceramic and the spindle, assembly of the ceramic on to the central spindle was carried out immersed in a big bucket of de-aired water, keeping the spindle in a vertical position under the water level. After assembling the ceramic onto the spindle the de-aired water pressure was imposed into the already de-aired cavity between the ceramic and spindle through the hole of the spindle. The de-aired ceramic was kept in a container of de-aired water for a few hours before assembly to the piezometer tip. Secondly, a twin tube of appropriate length, usually equal to the depth of bore hole plus 10 cm, was connected to the transducer using brass connectors and olive nuts. It was found necessary to do this before touching the transducer with de-aired water, as the penetration of water into the transducer via the twin tube fitting holes would be prevented in this way. The other end of the twin tube was connected to the self sealing spring connectors in the same manner as above. To make sure the connections were quite air tight the transducer was connected to the read out unit via the self sealing connectors and then the transducer and self sealing connectors were submerged in a bucket of de-aired water and a nitrogen pressure of 1.6 Bar (whole range of the Borden gauge) was applied to the system to see if any gas bubbles leaked from the system. After the system was proved to be airtight the transducer was retrieved from the water and the top cap was firmly screwed to the top of the transducer, Fig (4-2). Finally, to fill the cavity of the transducer with water and to assemble the ceramic, the transducer was submerged again in the de-aired water and the tubing side kept down. Although the cavity was immediately filled with water, in order to ensure a margin of safety, more water was injected into the cavity using a syringe. The subsequent appearance of air bubbles proved the entrapment of air bubbles in the cavity. After the cavity and ceramic were fully de-aired they were put together (under Jhe water) and tightened, by screwing the spindle on to the transducer and the brass cone onto the spindle. Now the piezometer was ready for calibration and installation, and it had to be kept in de-aired water until it was installed. A pneumatic piezometer, ready for installation, is shown in Fig (4-4). To calibrate piezometers an eight metres deep hole was drilled on the site and then a 2 inch diameter p.v.c. tube, one end plugged and eight metres long, was placed in the hole and the cavity between the hole and pipe was Filled with sand-cement grout. Meanwhile the pipe was filled with water and a protective cap was mounted. This hole was subsequently used to calibrate piezometers. For each of the piezometers a profile of piezometric pressure with depth was established by plumbing the piezometer tip into the pipe and reading the pressures against the different positions of the tip under the water level. The average of the results of calibrating each of the piezometers (which were not significantly different) three times,was accepted as the calibration criterion. It was observed that as a general rule the accuracy of piezometers decreases remarkably when the pore pressure decreases below 0.05 bar (0.5 water head). 4-2-3 The Readout Unit An old readout unit was modified and used to monitor the pneumatic piezometers. The readout unit had initially comprised a 150mm diameter Borden gauge calibrated from zero to 60 metres head of water, and a gas reservoir bottle re-chargeable from an external nitrogen supply up to 70 bars pressure. The Borden gauge was not found accurate enough for the range of pressures which would be anticipated. So another 250mm diameter Borden gauge, calibrated from zero to 1.6 bar was mounted on the unit. The readout unit, after modification, is shown in Fig (4-5). 4-3 VIBRATING WIRE PIEZOMETERS. 4-3-1 Details of the Piezometer. The vibrating wire piezometer tips with 5 atmospheres pressure range fitted with high air entry ceramics were used. Characteristics of the piezometer tip and the relevant ceramic are summarized in the following table: Piezometer Ceramic Tip External diameter (mm) 38 38 Overall length (mm) 240 62 Weight (kg) 1.3 Nominal accuracy (per cent of full scale) 0.10 Volume change (against 5 atm) (cm3) 0.005 Permeability (cm/sec) 2 x 10-6 Pore diameter (micron) 1 For cabling, three core copper conductors with heavy duty insulation, armoured, and with p.v.c. outer cover, with 12 ohms resistance per 1000 metres were used. The third core copper line was for when piezometers are fitted with thermistors for temperature measurement. The piezometer together with the readout unit are shown in Fig. (4-6). 4-3-2 Preparation for calibration and installation. The process of saturation and de-airing of the ceramic and the transducer and mounting of the ceramic on to the tip was exactly the same as the process for the pneumatic piezometers. Cables of appropriate lengths were connected to the piezometer tips and sealed by the manufacturer (Soil Instruments Ltd). To initialize a vibrating wire piezometer (refer to 2-1-1) a container was filled with de-aired water in the laboratory and was kept in a place where the temperature was not significantly changing. The piezometer was plumbed into the water down to the level of the transducer. After 24 hours, the pre-installation base reading, (refer to 2-1-1) was recorded. Calibration of these piezometers was carried out in the same way as the pneumatic ones. The constants of piezometers k were determined by the manufacturer. Vibrating wire piezometers are completely closed systems. They are sensitive not only to the changes of pore water pressure but also to changes of atmospheric pressure, which affect their records. Therefore to take into account the effects of changes of atmospheric pressure over the reading of piezometers, when the No value (pre-installation base reading) was measured the time was also recorded. To determine the atmospheric pressures corresponding to the recorded time, records of atmospheric pressure at Gatwick Airport were used. The above mentioned values are given in table (4-1). Table (4-1) Summary of results of initialization of the vibrating wire piezometers (No) and piezometer constants (k). Piezometer Date of Time of Atmospheric No k No. Calibration Calibration pressure mb J870 11.12.87 13 995 3973 27.13 J871 5.5.87 17 1010 3874 23.103 J872 3.2.87 16 1020 3760 18.84 J874 20.8.87 10 1019 4001 23.9 J875 26.2.87 15 1015 3771 22.74 J877 7.5.87 14 1030 3685 23.6 J878 8.6.87 10 1005 3940 20.86 J879 26.2.87 15 1015 3833 21.9 4-3-3 The Readout Unit. A miniature portable digital period counter utilising a 100 kHz crystal frequency reference was used to initialize, calibrate and monitor the vibrating wire piezometers, Fig(4-6). The internal circuits include a digital counter by means of which the period of one hundred cycles of transducer output signal (N Value) is presented on a 5-digit liquid crystal display. An accuracy better than 0.05% of reading in temperature range - 10 to 40oc is mentioned by the manufacturer. The High-Low switch provides two levels of transducer energisation voltage. The piezometers were monitored for both of the High and Low modes and the results obtained were found to be equally consistent on both modes. The most consistently displayed figure (Ni, refer to 2-1 -1) is accepted as the representative of ground water pressure. 4-4 AUTOMATIC TILTING BUCKET RAIN GAUGE. This system has been developed to provide accurate logging and analysis of rainfall over a 30 day period. The system consists of a bevelled ring which is mounted on a cylindrical cover fitted with a copper funnel which leads rain to a specially shaped and plated twin bucket. Each tip of the bucket is monitored by a switch connected to a module which has a part to accept a data cartridge and is battery powered, Fig (4-7-2). The rain gauge is placed on a flat surface and levelled. A cartridge is inserted in the module. In operation, when one side of the bucket is filled it tips and discharges its contents while the other side fills. Each tip of the bucket operates a switch and the event is recorded by the cartridge which also logs the time. In fact the number of times the bucket is discharged at any period of time can be calculated after the recorded data has been processed. As the discharge volume of each side of the bucket is known, so the recorded data can be used as the representative of the amount (mm/day or mm/hr) and duration (hrs/day) of rainfall. In the case of the system used here, the bucket would discharge against every 0.2mm rainfall and the clock accuracy was 1 min/month. The system is shown in Figs (4-7-1) & (4-7-2). 4-5 THE SEALING EFFICIENCY OF BENTONITE PELLETS. Application of bentonite pellets for piezometer installations is a common practice. They are used to form a low permeability plug which prevents the grout from penetrating into the sand filter. They are also sometimes used to backfill the borehole. In this research they have only been used to make a plug. Previous tests have shown that they can initiate a final permeability in the order of 9.2 * 10-9 cm/sec (Rocha Filho, 1976), but there was uncertainty about the minimum standing time in which a certain amount of pellets in a certain installation, can form a suitable plug against penetration of grout into the sand filter. In practice it is not always possible to leave the borehole unsealed for a long time until the pellets have completely swollen. To evaluate the sealing efficiency of pellets a standpipe piezometer was installed in a 75mm diameter transparent pipe. The pipe was 2 metres high and plugged at the base. The installation procedure was the same as proposed for installation of site piezometers. The procedure of installation of site piezometers is explained in section 5-3-2. Here only the relevant parts are briefly explained. A quarter litre of saturated sand (grain size 0.6 - 1.2mm) together with one litre de-aired water was poured into the pipe. The piezometer, which was kept under de-aired water for one day, was installed and 1.75 litres of saturated sand together with two litres of de-aired water was poured into the pipe and tamped with a one end plugged 19mm diameter pipe. Forty bentonite pellets were dropped into the pipe and then deliberately tamped immediately afterwards. It was observed that when the pellets were dropped into the hole they were not evenly distributed around the standpipe unless they were tamped afterwards. The bentonite pellets are shown in Fig (4-4). The grout consisting of 3 volumes of bentonite powder, one volume ordinary Portland cement and 5.7 volumes water was prepared in an electrical mixer. This process took 20 minutes and to take into account the elapse of time for carrying of grout from mixer to the site the grout was left as it was for 10 minutes. Then as the grout had partially set, 0.3 volume more water was added and stirred manually, and then poured into the pipe while being tamped. It was observed that a large part of the water which was previously poured into the pipe was spilled over the pipe due to tamping and higher density of the grout. The sealing process took 40 minutes and afterwards the standpipe was filled with de-aired water. After a few minutes, the surface of the pipe was cleaned and inspected. The interfaces of the bentonite pellet plug with the sand filter and the grout was very clear and distinguishable. The grout, in spite of strong tamping, had not penetrated into the bentonite plug. The bentonite plug had penetrated about 1cm into the sand filter. At the same time, some tiny cracks were seen in the bentonite pellets zone. One day later, large air filled cracks and cavities had formed, Fig (4-8). In Fig (4-9) the longitudinal section of the installation is shown. It is seen that the cores of pellets are still dry. The cracks and cavities are ruined due to cutting and so are not seen here. The concluding points are that the bentonite pellets, in less than one hour, produced an efficient plug against penetration of grout into the sand filter. The cavities could be due to the release of air voids in the pellets. These cavities could have probably been removed if the pipe had been left unsealed for a longer time, until the pellets had swelled to the core and tamped before sealing. In the case of standpipe piezometers these cavities do not appear to cause subsequent difficulties, but for rigid piezometers they might affect the response time. If these cavities are in direct contact with the pore water in sand filters the diffusion of air bubbles into the cavity of the piezometer will also be possible. ■ • i r i > • ’ - Fig ( 4 - 1) Details of P.V.C. and drive-in standpipe piezometers and the dipmeter. a Top cap u >? cOlive nuts Transducer M(Diaphragm o Spindle U Fig (4-2) Details of the pneumatic piezometer tip. 38 m n u Seal Fig (4-3) Fig Device to deair a ceramic using deaired water pressure water deaired using ceramic a deair to Device 125mm Ceramic Seal Bearing Nut eard Water De_aired 1 ? Fig (4-4): 1 - A pneumatic piezometer connected to the guide rod, ready for installation. 2 - The hand pump designed to lower the water level in the standpipe piezometers. 3 Bentonite pellets. Fig (4-5) The pneumatic piezometer readout unit connected to a piezometer. Fig (4-6) A vibrating wire piezometer and the read out unit. Fig (4-7-1) The Automatic tilting bucket rain gauge. Fig (4-7-2) Details of the Automatic tilting bucket tain gauge. Appearance of the plug of bentonite pellets after swelling for 24 hours. Fig (4-8) FFa r j CM i cnI Fig (4-9) A profile of the bentonite pellets plug after 24 hours. CHAPTER FIVE - SITE EXPERIMENTAL WORKS 5-1 INTRODUCTION The proposed research programme in the beginning was as follows: 1. Evaluation of the efficiency of the drainage system in the lowering of the ground water pressure and in turn on the improvement of the stability of the area. 2. Evaluation of the efficiency of the drainage trenches to discharge the drained water to some twenty-two years after installation, and assessment of the effective life span of the drainage trenches To achieve these aims it was necessary to provide a clear picture of the pore water pressure pattern of the area at the time before and after installation of the drainage system, especially at the present time. For the time before installation of the drainage system there were some records of piezometers, which had been installed a short time before construction of the drainage trenches and all lost during the installation of the trenches and development of the area. These records were not consistent enough to get a clear picture of the pore pressure pattern at the time before drainage and, of course, there was no way to supplement these records, because, now some twenty-two years had passed since the time of the installation of the first drainage phase. For the time after construction of the drainage system there were also some records of standpipe piezometers which had been installed between the drainage trenches and also in the non-drained area at different dates. Some of these piezometers were still surviving and needed some attention when the research started. Although the records of these piezometers were more consistent than the records of the previous group, further piezometers were necessary to supplement the records of these piezometers, however. As discussed in Section (2-2), equation of fluctuation of ground water pressure is a complicated function of engineering properties of soil, atmospheric pressure, meteorological conditions and especially topography and geometry of the natural or man-made drainage systems. In general the higher the amplitude and the shorter the period of fluctuation of ground water pressure, the more significant and critical are the errors (corresponding to time lag) in measurement of the ground water pressure. So, in a non-drained area in which the ground water pressure does not fluctuate quickly, adequate measurements of the ground water pressure could be made even with a standpipe piezometer of slow response, irrespective of the soil permeability (Vaughan, 1974). But here, there is a large drained area initiating quick changes of ground water pressure. Accordingly there was the possibility of considerable divergences from the real values in measurement of the ground water pressure. As there was no appreciable research to quantify the feature and range of this divergence in literature, the reliability of the records of the piezometers in the drained area went under question and the necessity of evaluation of the errors in measurement of the ground water pressure arose. To accomplish this important evaluation it was decided to install different types of piezometer (standpipe, pneumatic, vibrating wire) at the same positions with respect to the drainage trenches and then make a correlation between the records of these piezometers to discover the most critical case and conditions (in terms of the type of piezometer, position in depth and distance with respect to the trenches, atmospheric pressure, amount and duration of rainfall) in which the maximum relative errors appear in the measurement of the ground water pressure. To measure the amount and duration of rainfall an automatic tilting bucket rain gauge was installed. To consider the effects of the atmospheric pressure, records from Gatwick Airport were processed and used. To understand how quickly a pneumatic installation responds to a change of ground water pressure and how closely such a system follows the changes of the ground water pressure a system was made and installed, refer to section ( 5-5). To monitor water pressures in the trenches their locations were recovered in appropriate zones and piezometers were installed in them. For better interpretation of the pore water patterns some in situ permeability tests were also carried out. The following chapter elaborates the details of the site works carried out to fulfil the above mentioned aspects. 5-2 INSPECTION AND SERVICE OF THE PREVIOUSLY INSTALLED PIEZOMETERS All of the previously installed standpipe piezometers which survived, were inspected from three points of view, as follows: 5-2-1 Inspection against penetration of surface water into the piezometers. Improvements of commercial caps. The protective caps of the piezometers were inspected to see if they were tight enough against penetration of rainfall and surface water into the piezometers. The commercial threaded p.v.c. caps were not found to provide a reliable shelter for the piezometers, because, the rainfall penetrated through the ventilation hole into the piezometer. These caps were improved and used for most of the previously installed piezometers and also for the new installation. A commercial cap before and after improvement is schematically shown in Fig. (5-1). Any piezometer cap found to have sunk into the ground was reinstalled 1-2 inches above the ground surface. 5-2-2 Inspection against blocking. Cleaning of piezometers. Piezometers were probed to see if they were filled with debris and/or blocked. Those filled with soft materials were cleaned, whilst those which were blocked with stiff materials, like cement mortar etc, were assumed lost. To clean up the piezometers, a hollow spindle was mounted on top of the probes and pushed into the sedimented debris and then pulled out, Fig. (5-3). The rest of the sedimented part and floating materials were flushed out by flushing clean water into the piezometer tip using a copper tube and a hosepipe. 5-2-3 Equalization tests, as a criterion for reliability of piezometers. Piezometers with tight caps and showing reasonable and consistent yearly pore water pressure fluctuations were assumed to be reliable. Those piezometers having apparently strange and or scattered records were checked against leakage through the standpipes and connections. This was accomplished by equalization tests. The procedure of the test and the subsequent results, which were also used for permeability measurements, are explained and shown in Sections (5-7-1) and (6-2),respectively. The relevant discussion is given in Section (8-1-3). The positions of all the piezometers are shown on Fig. (5-2). The characteristics of all piezometer installations, since 1964, are summarized in table (6-1). 5.3 LOCATING OF DRAINAGE TRENCHES Determining the exact location of the drainage trenches, their depth and shape of lateral cross section at appropriate points was of great importance for the subsequent piezometer installations and other research work. Although there were some available drawings showing the position of the drainage trenches, field work showed that these drawings and what had actually been laid out and constructed were not in exact agreement. This difficulty, plus regrading of the slope during the construction period and collapsing of the edges of the trenches during excavation made the accurate locating process more complicated. However, the first possible step was to determine the approximate position of the centre lines of the trenches in appropriate zones with respect to the available drawing, using surveying techniques. Then, a Mackintosh probe was used to find out the response of the soil beneath the ground surface against penetration of the probe at different points around the approximate centre line (laid out by surveying). Sounding of granular materials proved the presence of a trench under a probed point. The Mackintosh probe and its accessories are shown in Fig. (5^3). For determination of the shape of the lateral cross section of a trench, the probe was driven into the ground until it had passed through the superficial deposits into the granular material. Driving of the probe through the granular material was continued carefully until the clay was felt by the operator. This moment was very important and the depth of penetration of probe (from ground level) showed the depth of one of the interface points between the granular material and clay. This process was repeated for other points 10 - 20cm apart from each other along a straight line. Four to six times repetition of the driving of the probe into the ground was found to be reasonable enough to locate the lateral cross section of a trench. Then the results were plotted and the cross section of a trench was determined. The determination of the shape of cross section using a probe depends greatly on the skill of the operator. The following experiences were found to be very useful: As long as the probe is in the granular material it feels very loose and can be simply pulled up or shaken (at least a few centimetres) without winding back; while in the clay it is tight and impossible to pull up (even to shake it) and it comes up only with winding back. Driving of the probe through the granular material becomes easier if the turning of the top lever is accompanied by vibrating or a continuous hammering; while in the clay, winding is the most effective. The resistance of the clay against the winding of the probe is much higher in comparison with the resistance of the granular material. The granular materials make a noise when they are probed, whilst the clay does not make any noise. Six trenches were located in four zones and their lateral cross sections determined along four profiles kl-k2 (Zone 7), k3-k4 (Zone 4), k5-k6 (Zone 2) and k7 - k8 (Zone 6). These profiles and their positions are shown in Figs. (5-4), (5-5), (5-6), (5-7) and Fig. (5-2)/respectively. 5-4 INSTALLATION OF NEW PIEZOMETERS 5-4-1 Reasons and justifications for the new installations. 5-4-1-1 Introduction Fifty nine new piezometers of different types have been installed during the research. Before elaborating on the reasons and justifications for the new installations it is necessary to mention the following points: To simplify the positioning of the piezometers and also to shorten the text and avoid repetition, zones of concentration of piezometers have been enumerated. Nine zones have been distinguished and shown in Fig. (5-2). Installations have been carried out with respect to the availability of the appropriate resources, so a systematic installation has not been possible. In installation of adjacent piezometers attention has been paid to keeping them out of the radius of influence of each other. The radius of influence of a piezometer (r) is the distance from centre of intake point to a region in the soil where the maximum change in pore pressure throughout the entire equalisation process does not exceed 5% of the total pressure change (Premchitt and Brand, 1981). This radius is a function of volume change of a piezometer. The greater the volume change the larger the radius of influence. So, for example, if a rigid piezometer is installed inside the radius of influence of a flexible standpipe piezometer, any equalization test on the standpipe piezometer will disturb records of the adjacent rigid piezometer. The same phenomenon happens if the ground water level changes. The maximum radius of influence of a piezometer, assuming ji = 0.01, is about 5a (Premchitt and Brand, 1981) in which ”a” is the equivalent radius of intake point of a piezometer. Details of the new and previous piezometer installations are summarized in Table (6-1). The idea of the new installations was, in general, to establish a system which would fulfil the following aspects: 1 - To supplement the records of the previously installed piezometers to evaluate the pore water pressure pattern. 2 - To compare the pore pressure pattern in the past (before drainage) and at the present time and also in drained and undrained areas to study the efficiency of the drainage system in the lowering of the pore water pressure. 3 - To evaluate the stability of the slope before and after drainage 4 - An experimental correlation between the factors affecting the ground water pressure equation and records of the different sorts of piezometers, (refer to 2-2) to evaluate the difficulties and errors in measurements of the pore water pressure associated with time lag and flexibility of the piezometers. 5 - A study of the discharge potential of the drainage trenches some twenty-two years after construction. To establish such a system, it was initially decided to install piezometers along three profiles, and in some other places because of the appropriate reasons. Description of these profiles and places and the appropriate reasons and justification for the new piezometer installations are as follows: 5-4-1-2 Installations along profile 5-5 Profile 5-5, Figs. (5-2) and (5-8) was found suitable because,it would allow some piezometers to be installed in a non-drained part of slope. A study of the effect of construction of the proposed new building (The Centre for the Performing Arts) over the pore water pressure pattern (if any) would also be possible in this way. Profile 5-5 was pegged and levelled and the achieved profile was compared with the topography of the area at the time before construction of the University. It was observed that the slope was regraded by dumping soil at the toe of the slope. The profile is shown in Fig. (5-8). Installation of the new piezometers was started along this profile. Characteristics of the installations are summarized in the following table: Date of Piezo Type Depth Relevant installation No. (m) Figs. 6.5.86 J1 standpipe 7.5 (5.2),(5.8 ) 6.5.86. J2 99 5.5 (5.2),(5.8 ) 6.5.86 J3 r> 2.5 99 99 8.5.86 J4 » 9.0 99 99 8.5.86 J5 99 5.0 99 99 9.5.86 J 6 99 7.0 99 99 12.5.86 J7 99 3.0 99 99 20.5.86 J l l 99 4.0 99 99 21.5.86 J10 99 3.0 99 99 21.5.86 J12 99 2.5 99 99 22.5.86 J8 99 8.0 99 99 22.5.86 J9 99 6.0 99 99 19.3.87 J38 99 5.0 99 99 19.3.87 J39 99 3.0 99 99 The reason for the installation of piezometers J38 and J39, in Zone 1, at the same depths as piezometers J5 and J7 was that in November 1986 the piezometric level in piezometer J5 suddenly jumped from its usual pattern (3 to 3.40 m b.g.l.) to a piezometric level more or less the same as the piezometric level in piezometer J7, refer to Fig.(6-71), so, these two piezometers were installed to prove the real pattern. Piezometers J10, J11 and J 12 should have been installed along profile 5-5, but as the original position fell under the proposed location of the new building, their position was inevitably shifted 30m to the west of the profile. 5-4-1-3 Installations along profile A-A. This is the profile along which the slip surface had been located in the past [Refer to Fig. (3-9), Skempton, 1965], so, installations along this profile would not only fulfil part of the aspects mentioned in the beginning, but the achieved pore water pattern could also be correlated with the pre-defined slip surface to check and control the stability of the slope at the present time. This profile is shown in Figs. (5-2) and (5-9). The only suitable and also most fruitful place to install the piezometers along this profile was Zone 4, Fig. (5-2). The other places along this profile are now covered by buildings or pavements. Following the location of trenches in this zone, the appropriate piezometers were installed. The characteristics of these installations are summarized chronologically in the following table: Date of Piezo Type Depth Relevant installation No. (m) Figs. 2.7.87 J13 Standpipe 5.0 (5-2),(5-16),(5-ll) 11.8.87 J27 99 4.8 99 99 99 11.8.87 J29 n 5.0 99 99 99 27.11.87 J20 Pneumatic 5.0 (5-11),(5-12) 3.12.87 JT21 Drive-in 4.2 » (5-10) 30.3.87 J40 Pneumatic 10.0 » (5-12) 15.4.87 J41 Pneumatic 8.0 99 99 99 29.4.87 J42 Pneumatic 9.0 99 99 99 6.5.87 J44 Standpipe 8.0 99 99 99 6.5.87 J871 Vib.wire 8.0 » (5-12) 7.5.87 J877 5.0 », (5-17),(5-13) 20.8.87 J874 5.0 », (5-13), (5-14) 26.11.87 JT52 Drive-in 4.2 **»(5-13) The purpose of the installation of piezometers J13, J27, J29 and JT21 was basically to establish the distribution of pore water pressure between two adjacent trenches at a depth of about the same as trench invert level, and also to monitor the pore pressure in a trench. To study the range of errors associated with different types of piezometers and possibly the effects of the fluctuations of the atmospheric pressure on the ground water pressure, the pneumatic piezometer J20 and vibrating wire piezometer J877 were installed at the same position with respect to the trenches as J13. Subsequently, for quick establishment of the pore pressure distribution with depth, the pneumatic piezometer J40, 10m deep, was installed at the same position with respect to the trenches as piezometer J20, see Fig. (5-12). As the achieved pore pressure pattern diverged considerably from the hydrostatic condition, two other supplementary pneumatic piezometers, J41 and J42 were installed. The reason for the choice of pneumatic piezometers was that: Fluctuations of the pore water pressure at different depths in the drained area would quickly be assessed using a rigid system. The present pore pressure pattern would be precisely compared with the pore water pressure pattern at the time before installation of the drainage system, which was available from piezometers 103 and 104. From the above, the mechanism of the drainage system on the stabilization of a slope would also be assessed. Now, the system for evaluation of the errors in measurement of the pore water pressure at about trench invert level had been established. To compare the records of the different types of piezometers at a depth well below the trench invert level the standpipe piezometer J44 and also vibrating wire piezometer J871 were then installed, Fig. (5-12). A comparison between the records of piezometers J27 and J874, which had been installed at the same distance from the drainage trench, Fig. (5-14), showed a considerable difference. This was not due to the quicker response of the piezometer J874, because it was steady. To justify the reason it was necessary to prove the shape of flow surface in the trench. So, the piezometer JT52 was installed in the trench, zone 4 , 1.85m to the North of piezometer JT21. 5-4-1-4 Installations along profile 6-6 This profile passes through the centreline between the trial trench and the consecutive trench to the East, Figs. (5-2) and (5-15). The advantage of this profile was that there is no appreciable obstacle against sub-surface explorations to locate trenches and installation of the new piezometers all the way along this profile, and also it was probably the steepest profile of the slope which would help to assess the most critical factor of safety in the area. The new installations are concentrated at three places along this profile. These places are Zone 7, Zone 6 and a place near the top of the hill just outside the back scarp, Figs. (5-2), and (5-15). Zone 7: Some characteristics of the piezometers which are installed in this zone are summarized chronologically in the following table: Depth Relevant Date of Piezo Type m b.g.l. Figs. installation No. 9.7.86 JT14 standpipe 4.80 (5.2) 14.8.86 J16 pneumatic 4.15 (5.2),(5.16) 14.8.86 J17 99 4.00 (5.2),(5.15),(5.16) 14.8.86 J18 99 4.10 99 99 26.9.86 JT15 drive-in 4.10 99 99 26.9.86 JT19 99 4.00 99 99 23.1.87 J22 standpipe 4.20 (5.2),(5.15) 5.2.87 J872 vib.wire 4.0 (5.2),(5.15)(5.16) 5.2.87 J870 99 8.00 99 99 12.3.87 J24 standpipe 2.50 99 99 Following the location of the trial trench the standpipe piezometer JT14 was initially installed as a trial piezometer installation in the trench. The purpose of installation of piezometers JT15, J16, J17, J18 and JT19 was an immediate establishment of the pore pressure between and in the trenches at a depth equal to the depth of the trench invert level, to evaluate the efficiency of the drainage system and stability of the slope along profile 6-6 and also to compare the achieved pattern with the corresponding patterns in the other places. Following the unexpected records obtained from the pneumatic piezometers J16, J17 and J18, the standpipe piezometer J22 was installed to prove the records of the above mentioned piezometers. According to the pneumatic piezometers the piezometric level was at the same depth as the trench invert level all the way between the trenches, which was surprising indeed, Figs. (6-65) & (5-15). The reason for installation of piezometers J870, J872 and J24 was to establish the distribution of the pore water pressure with depth and also to compare the results of the different types of piezometers (J17, J22, J872) and again to prove the apparently odd pattern which had been observed with piezometers J17 and J22. - Zone 6 : The further necessity to establish the pore pressure pattern at another place toward the top of the slope along profile 6-6, was revealed when the above mentioned pattern was observed in Zone 7. It was basically necessary to understand to what extent the observed pattern would dominate. Otherwise the appropriate evaluation of the stability of the slope would not have been possible. For this reason, the trenches were located again at 39m toward the top of the hill from section K1-K2, along the cross-section K7-K8, Figs. (5-2), (5-17) and (5-18), and the appropriate piezometer installations were carried out. Some characteristics of the installations are summarised chronologically in the following table: Date of Piezo. Type Depth Relevant installation No. m b.g.l. Figs. 27.2.87 J875 vib.wire 8.00 (5-2),(5-17),(5-18) 27.2.87 J879 99 4.20 99 99 99 5.3.87 JT23 drive-in 4.30 99 99 99 12.3.87 J25 standpipe 2.50 99 99 99 29.4.87 J43 99 6.00 99 99 99 9.6.87 J878 vib.wire 11.80 99 99 99 7.5.87 J45 standpipe 4.20 99 99 99 17.6.87 J46 99 4.20 99 99 99 10.7.87 J51 99 11.80 99 99 99 Five piezometers J875, J879, J25, J43, and J878 were installed primarily to establish the pore pressure distribution pattern with depth and to study the feature of fluctuations of the pore water pressure at different depths. It is to be noted that in order to get a consistent profile of the pore water pressure profile it was proposed in the beginning to install a series of vibrating wire piezometers all the way from 4 m b.g.l. to 12 m b.g.l., but when installing a vibrating wire piezometer in borehole J43, the piezometer fell on the concrete pavement, so a standpipe piezometer was installed in the borehole. The drive-in piezometer JT23 was installed in the trench to the east of the trial trench for further monitoring of the discharge ability of the trenches, although the pore water pressure in trenches was monitored previously by piezometers JT15 and JT19. As part of the system for analysis of the errors and difficulties in monitoring pore water pressure using different types of piezometers, the standpipe piezometer J45 was subsequently installed. To supplement the establishment of the pattern of distribution of the pore pressure between the two trenches at trench invert level (i.e. approx. 4m) and also 12m b.g.l., two supplementary piezometers J46 and J51 were installed adjacent to the interface of the trench and clay. - Installations in the vicinity of the back scarp To complete the establishment of the pore water pressure pattern along profile 6-6, three piezometers were installed along this profile just in the vicinity of the back scarp. Some characteristics of these piezometers are summarized chronologically in the following table: Date of Piezo. Type Depth Relevant installation no. m b.g.l. Figs. 12.3.87 J26 standpipe 2.50 (5-2), (5-15) 12.3.87 J28 99 4.50 99 99 17.3.87 J47 99 10.00 99 99 5-4-1-5 - Other piezometer installations - Zone 2 Three standpipe piezometers, P503, P504 and P505 were previously installed in this zone between the trenches, Figs. (5-2) and (5-19). After the two adjacent trenches were located the vibrating wire piezometer J873, 5m deep, was installed 2.5m to the north of piezometer P505, at the same position with respect to the trenches as piezometer P505, Fig. (5-20). The purpose of this installation was basically to make a comparison between the records of a standpipe and a vibrating wire piezometer at an area with closer drains and also to compare the results with the results of the appropriate piezometers in Zones 4,6 and 7. - Zone 8 There are eight standpipe piezometers previously installed in this zone. These piezometers and their depths are as follows: Piezo Depth No. m b.g.l. 601 6.0 602 12.0 701 45.0 702 15.0 703 30.0 KP1 9.0 (backfilled with clay) KP2 9.0 KP 3 9.0 (backfilled with cement slurry) This Zone with piezometers from six to forty-five metres deep could give a clear picture of the distribution of pore water pressure with depth. In order to supplement the results of these piezometers, two standpipe piezometers J48 and J49, depths of 5m and 2.5 m b.g.l, were installed on 1.7.1987. - Zone 9 As will be explained in Section (5-5), Zone 9, where the water level stays above the ground level for a considerable period of the year, could be a suitable place, not only to establish a profile of distribution of the pore water pressure with depth, but also to study the time dependent characteristics of the different types of the piezometer installations (explained in section 5-7-1). Some characteristics of the piezometers which were initially installed in this zone are summarised in the following table: Date of Piezo. Type Depth installation no. m b.sr.l, 1. 10.86 J31 standpipe 4.0 1. 10.86 J33 99 1.5 12.10.86 J34 pneumatic 4.0 12.10.86 J35 99 1.5 27.11.86 J36 99 4.0 Piezometer J31, of course, would form part of the system explained in Section (5-5). Subsequently on 27.1.87, the system J37 was installed to evaluate the response time of pneumatic piezometer installations. Further details are given in section (5-5). - Piezom eter J50 At a place near the Eastern boundary of the area some subsurface exploration was carried out during the summer of 1987, for further construction works. The standpipe piezometer J50 was installed in a 15m deep borehole which had been drilled for sampling. In the beginning it had not been proposed to install any piezometers in this place. However, when the borehole was drilled it was found worthwhile to install a piezometer in the borehole for further elucidation of the pore water pattern. The position of this piezometer is shown in Fig. (5-2). 5.4.2 General procedure of piezometer installations In order to avoid repetition, piezometer installation procedure, common to all piezometers is explained in this section. Subsequently in Section 5-4-3 the borehole logs, and any other distinctive and special features of each of the borehole installations will be described. 5-4-2-1 Drilling A mobile drilling system mounted on a single axle carriage, with solid stem continuous flight augers of 75mm and 100mm diameter, fishtail and mining bit drills were used for drilling. The drilling machine and augers with drills are shown in Figs. (5-21) and (5-3) respectively. The 75mm diameter augers were used for drilling unless otherwise mentioned in section (5-4-3). For drilling in clay a fishtail bit was used and whenever the drill was faced with a hard obstacle, the fishtail bit was replaced with a mining bit. It was observed that when a borehole was completed with a mining bit some fragments of soil were left in the borehole after the mining bit drill was extracted from the hole. This made the piezometer installation very difficult and sometimes impossible. Therefore, whenever a borehole was completed using the mining bit drill, the fishtail bit was also driven all the way into the borehole to clear up the clay fragments. These cases, were, of course, very time consuming. As the soil is disturbed due to drilling, it is not possible to establish a borehole log describing the soil type. Also it is impossible to realize the depth from which a particular soil specimen is coming up. In this research, to overcome the second difficulty, a step by step drilling method was employed. For 5m depth and less the drill was driven into the ground in steps of 0.5m. After any 0.5m driving of the drill into the ground the drills was fixed in its position, while rotating, until all the drilled soil was cleared out of the borehole and then the drill was driven again one more step into the base of the borehole. By this method it was possible to get an idea of the stratiography of the ground and thickness of the different layers. Beyond 5m b.g.l. the drilling steps were usually controlled by the power of the engine and resistance of the soil. At depths below 8-9m steps even less than 10cm were sometimes experienced. 5-4-2-2 Installation and sealing in clay Immediately after a borehole was drilled, the appropriate piezometer was installed using the following procedure unless otherwise mentioned in (5-4-3): (1) Twenty four hours before installation, an appropriate amount of fine grained veiy clean sand (grain size 0.6 - 1.2 mm) was saturated in de-aired water. (2) A quarter litre of saturated sand together with one litre of de-aired water was poured into the hole to make a suitable bed for the piezometer tip and to keep it submerged. (3) The appropriate piezometer which was resting in a container of de-aired water was quickly embedded in the hole. Standpipe and vibrating wire piezometers were directly plumbed into the hole. For pneumatic piezometers a guide rod, which was screwed onto the cap of the piezometer tip, was used to place the tip correctly at the base of the hole. Rigidity of the twin tube usually makes difficulties if the pneumatic tip is installed without employment of a guide rod. The guide rod is shown in Fig. (4-4). All precautions, for closed systems, were made to shorten the time of installation (the elapse of time from the moment at which the tip was removed from the de-aired water until it was submerged again in the de-aired water of the base of the hole). In the most unfavourable conditions this time did not exceed about twenty seconds. (4) 1.75 litres of saturated sand were poured again into the hole and the sides of the hole were washed with two more litres of de-aired water to clear grains of sand which could have stuck onto the sides of the hole. (5) The sand filter was tamped using a one-end plugged 19mm diameter pipe. Then the height of sand filter was measured and the tamping pipe was pulled out of the hole. (6) A certain number of bentonite pellets, Fig. (4-4), (40 in most cases) were dropped into the hole and then deliberately tamped. The hole was left as it was for as long as possible, to let the pellets swell and make a suitable plug. The standing time longer than say 5 hours was not practically possible. (7) During the standing time of the hole, an appropriate volume of bentonite-cement grout was prepared in the laboratory. To make the grout, three measures of bentonite powder were mixed with one measure of ordinary Portland cement in an electrical mixer and then about 5*7 measures of water was added gradually to the mixture while the mixing was continued until a homogenous grout was produced. The grout was taken to the site and mixed manually with 0.3 volume more water, in order to eliminate setting of grout on the way from the laboratory to the site. Then the grout was poured directly into the hole and tamped meter by meter up to ground level. Then a suitable cap was mounted and appropriate measurements were started. It should be mentioned that in order to minimise the shrinkage, the grout was produced somewhat thicker (i.e. less water) than that which is usual in common practice and as it was not possible to place the thicker grout into the hole through a conductive pipe, it was directly poured while being strongly tamped. To understand the efficiency of tamping in making of a sound and compact seal, pressures were measured with some pneumatic and vibrating wire piezometers immediately after the installation had been completed. These measurements are summarised in table (5-1). Table (5-1) amounts of pressure generated immediately after sealing: Piezometer Type Piezo Po Pav. No. Depth m. head m. head m b.g.l. of water of water. J37 pneumatic 4 4.2 3.80 J87 vib.wire 5 5.5 2 J877 99 99 5 4.4 3 J40 pneumatic 10 9.20 3.20 In this table: P0 = the pressure measured immeditely after installation was completed. Pav = the average ground water pressure at the time of installation. In the following hours these pressures decreased and within two or three days tended towards Pav value. With respect to the Table (5-1) it can be observed that in spite of the fact that tamping and use of a thicker grout takes much effort and a longer time, a compact seal is built up, however. It is seen that even pressures above the ground level are sometimes generated. The other advantage of the application of a concentrated grout was revealed when it was used to seal boreholes which the ground water had entered and accumulated in. In these cases, the greater part of the water was squeezed out of the borehole due to tamping and the considerable density of the grout. 5 .4-2-3 Installation in the trenches In the beginning, the collapse of granular materials immediately after drilling made it impossible to install any piezometers in the trenches. To overcome this difficulty, two alternative methods were used, as follows: In the first method a casing comprising 1.1/4 inch diameter steel pipes of 1.20m long and both ends threaded, connectors, and a cone shaped steel nose was made and used. The steel nose could be fitted onto the top of the casing pipe, not by screwing, but by friction. Detrails of casing pipes are shown in Fig. (5-22). To drive the casing into the ground a borehole was firstly drilled to the appropriate depth. This was necessary, because hard obstacles would be crushed due to drilling. After drilling the casing was advanced down to the appropriate depth by hammering. Then a standpipe piezometer, 1.5m longer than the proposed depth of borehole was embedded into the casing. In this way, a 1.5m length of standpipe stayed above the ground level. The casing pipes were then gradually withdrawn and unscrwewd section by section, whilst the standpipe was being pushed firmly over the steel nose to release the nose from the casing pipes. The nose, of course, was left permanantly in the ground. In the second method, drive-in piezometers were used. The Mackintosh probe, Fig. (5-3), was initially driven into the ground to ease the subsequent driving of the piezometer into the trench and to find out the depth and thickness of the granular material in the trench. Then the piezometer was driven into the trench. Although the second method was easier than the first, there was a risk of bending the standpipe and/or flattening and breaking the connectors. Further details in connection with this,are given in Section (5-4-3). 5-4-3 Borehole logs and special cases encountered during borehole drilling and piezometer installations. The following points are noticed again: (1) All boreholes were drilled using 75mm diameter augers unless otherwise stated. (2) All drilling installation and sealing processes were carried out according to Section (5-4-2) unless otherwise mentioned. (3) Piezometer installations were carried out immediately after drilling unless otherwise stated. Borehole No. G.L. Depth Borehole log and mOD m b.g.l. description J1 60.10 7.6 0.0 - 0.4m topsoil 0.4 - 2.0m backfill,(clay, concrete,brick) 2.0 - 7.6m brown London Clay The borehole was dry. J2 60.10 5.6 0.0 - 0.4m topsoil 0.4-2.0m backfill,(clay,concrete,brick) 2.0 - 5.6m brown London Clay. J3 60.10 2.6 0.4-2.0m backfill (clay,concrete,brick, etc) 2.0-2.6m brown London Clay The borehole was dry. It was proposed to experience the possibility of installation of more than one piezometer in the same borehole, so, this borehole was drilled using the 100mm diameter augers. Piezometer J 2 was installed in the borehole according to Section (5-4-2) except that 3.5 litres of sand was used in steps 2 and 4. The borehole was sealed up to 2.5m b.g.l. and then a few hours later, when the grout had become somewhat crumbled, piezometer J2 was installed in the borehole in the usual way. Because of the difficulties with drilling with larger augers and the elapse of a long time, installation of more than one piezometer was not found worthwhile and abandoned. Borehole No. G.L. Depth Borehole log and mOD m b.g.l. description J4 62.30 9.2 0-0.3m topsoil 0.3 - 1.0m backfill 1.0-9.2m brown London Clay. The borehole was dry. J5 62.30 5.1 0-0.3m topsoil 0.3 - 1.0m backfill 1.0-5. lm brown London Clay The hole was drilled down to 7.0m b.g.l. by mistake. It was backfilled to 5. lm with a mixture of bentonite grout and clay fragments. The borehole was quite dry. J 6 62.30 7.3 0-0.3m topsoil 0.3 - 1.0m backfill 1.0 - 7.3m brown London Clay The borehole was dry. J7 62.30 3.2 0-0.3m topsoil 0.3 - 1.0m backfill 1.0 - 3.2m brown London Clay The borehole was dry. J8 69.00 8.1 0 - 0.3m topsoil 0.3-8.10m brown London Clay. Water was gradually entering into the hole. Piezometer was installed ------immediately after drilling and the hole sealed up. Borehole No. G.L. Depth Borehole log and mOD m b.g.l. description J9 69.00 6.2 This borehole was initially drilled lm to the East of borehole J 8 . At 5.5m b.g.l. the drill faced with a claystone of about 0.3m thick. Whilst the claystone was being drilled, the borehole was dry. Immediately after the drill passed through the claystone, water flushed into the hole and rose to 3m g.b.l. Too much mud was generated in the hole and a subsequent attempt to wash out the hole using a water supply and a hose pipe was not found to be successful. This hole was backfilled with cement bentonite grout and clay fragments and another hole was drilled lm to the west of J 8 . The borehole log was as follows: 0 - 0.3m topsoil 0.3 - 6.2m brown London Clay. This time (only 2m to the VVest of the previous hole), there was not any sign of claystone and/or water. The borehole was dry just after drilling. J10 57.20 3.0 0.0 - 0.3m topsoil 0.3 - 3.0m brown London Clay. At 3m b.g.l. a hard layer appeared, but it was not drilled, and the augers removed from the borehole. Water was penetrating into the hole from the base. The hole was left open overnight. Next Borehole No. G.L. Depth Borehole log and mOD m b.g.l. description day the water level was measured to be 1.5m b.g.l. The water level just after drilling was 2.5m b.g.l. A piezometer was installed according to Section (5-4-2) except that no de-aired water was poured into the hole. Jll 57.15 4.2 0.0 - 0.3m topsoil 0.3 - 3.0m brown London Clay 3.0 - 3.2m Claystone 3.2 - 4.2m brown London Clay. The water level in the hole just after drilling was measured to be 3.3m b.g.l. Piezometer was installed according to Section (5-4-2) except that no de-aired water was poured into the hole. J12 57.15 2.25 0.0 - 0.3m topsoil 0.3 - 2.25m brown London Clay. The borehole was dry just after drilling and was left open overnight. Next day, the water level was 0.3m above the base of the borehole. A standpipe piezometer was installed and the hole was sealed up. When the top cap was being hammered into the borehole, the piezometer was also found to have been driven about 0.25m ------into the base of the hole. ------Borehole No. G.L. Depth Borehole log and mOD m b.g.l. description J13 63.26 5.10 0.0 - 0.3m topsoil 0.3 - 5.0m brown London Clay. The borehole was dry. JT14 59.75 4.85 The Mackintosh probe was initially driven into the trench to prove the depth of the trench invert level, which was previously determined, subsequently a standpipe piezometer was installed using casing according to Section (5-4-2-3) Borehole log according to the Mackintosh probe was as follows: 0.0 - 0.6m topsoil 0.6 - 4.2m granular material 4.2 - 4.85m London Clay. Water level was at 4. lm b.g.l. JT15 59.90 4.10 TheMackintosh probe and then a drive-in piezometer were advanced into the trench. The borehole log according to the Mackintosh probe was as follows: 0.0- 0.6m topsoil and fill 0.6 - 4.0m granular material 4.0 - 4.10m brown London Clay. The water level was measured to be 3.95m b.g.l. Borehole No. G.L. Depth Borehole log and mOD m b .g.l. description J16 60.00 4.50 0.0 - 0.5m topsoil 0.5 - 3.5m backfill material (clay, stone, brick) 3.5 - 4.5m brown London Clay. It appears that the trench must have collapsed before it was backfilled and the collapsed zone must have been subsequently been backfilled with miscellaneous material. The borehole was dry. J17 59.80 4.40 0.0 - 0.4m topsoil 0.4 - 4.4m brown London Clay. The borehole was dry. J18 60.00 4.40 0.0 - 0.4m topsoil 0.4 - 4.4m brown London Clay. The borehole was dry. JT19 60.00 4.10 0.0 - 0.5m topsoil 0.5 - 3.95m granular material 3.95 - 4.0 Clay Water level was measured to be 3.77m b.g.l. Borehole No. G.L. Depth Borehole log and mOD m b.g.l. description When a drive-in piezometer was being driven into the trench, one of the connectors which was already driven down to 2.2m b.g.l. suddenly became disconnected and so another piezometer was driven in again to the trench a few inches away from the broken one. J20 63.78 5.10 SameasJ13 JT21 63.05 4.20 0.0 - 0.5m topsoil 0.5 - 4.2m granular material After the Mackintosh probe was advanced into the trench to find out the depth of the trench invert level, a drive-in standpipe piezometer was installed. J22 60.20 4.40 0.0 - 0.4m topsoil 0.4 - 4.4m brown London Clay. The hole was quite dry and the piezometer was installed immediately after drilling. JT23 4.40 0.0 - 0.4m topsoil and backfill 0.4 - 4.3m granular material 4.3 - 4.4m brown London Clay. A drive-in piezometer was driven into the trench. The water level was measured to be 4.17m b.g.l. immediately after the piezometer was installed. Borehole No. G.L. Depth Borehole log and mOD m b.g.l. description J24 61.20 2.60 0.0 - 0.3m topsoil 0.3 - 2.6m brown London Clay The borehole was dry. J25 65.18 2.60 0.0 - 0.4m topsoil 0.4 - 2.6m brown London Clay. The borehole was dry. J26 71.20 2.60 0.0 - 0.3m topsoil 0.3 - 2.6m brown London Clay The borehole was dry. J27 63.06 4.90 0.0 - 0.5m topsoil 0.5 - 4.9m brown London Clay The borehole was dry. J28 71.60 4.60 0.0 - 0.4m topsoil 0.4 - 4.6m brown London Clay The borehole was dry. J29 62.46 5.10 0.0 - 0.3m topsoil 0.3 - 5.10m brown London Clay The borehole was dry. J31 4.30 0.0 - 0.3m topsoil 0.3 - 4.3m brown London Clay The borehole was dry. J33 1.80 0.0 - 0.3m topsoil 0.3 - 1.8m brown London Clay The borehole was dry. J34 4.30 SameasJ31 Borehole No. G.L. Depth Borehole log and mOD m b.g.l. description J35 1.80 Same as J33 J36 4.30 SameasJ31 J37 4.30 Refer to Section (5-5) J38 62.30 5.10 0.0 - 1.0m topsoil and backfill 1.0 - 5.10m brown London Clay The borehole was dry. J39 62.30 3.10 0.0 - 1.0m topsoil and backfill 1.0 - 3.10m brown London Clay The borehole was dry. J40 63.60 10.20 0.0 - 0.3m topsoil 0.3 0- 8.0m brown London Clay 8.0 - 10.20m a very stiff mixture of brown and grey London Clay. At 8m b.g.l. augers were stuck in the hole. With further attempts they were finally released and the rest of the hole was drilled very deliberately in steps of a few centimetres. The borehole was dry. J41 63.98 8.20 0.0 - 0.3m topsoil 0.3 - 8.20m brown London Clay The borehole ws dry. J42 63.90 9.20 0.0 - 0.3m topsoil Borehole No. G.L. Depth Borehole log and mOD m b.g.l. description 0.3 - 9.20m brown London Clay with some trace of grey Clay becoming very stiff after 8m b.g.l. The borehole was dry. J43 65.22 6.20 0.0 - 0.4m topsoil 0.4 - 6.2m brown London Clay. The borehole was dry. J44 63.96 8.10 Same as J41 J45 65.66 4.50 0.0 - 0.4m topsoil 0.4 - 4.5m brown London Clay. The borehole was dry. J46 65.66 4.50 0.0 - 0.6m topsoil and backfill 0.6 - 4.5m brown London Clay The borehole was dry. J47 71.56 10.00 0.0 - 0.3m topsoil 0.3 - 5.0m brown London Clay 5.0 - 10.0m brown London Clay with scattered fragments (or layers) of claystone. Water entered the hole from 5m b.g.l. and lm thick mud was generated due to drilling and penetration of water into the hole and accumulated at the bottom of the hole. The piezometer was inevitably installed in the mud, no sand filter was Borehole No. G.L. Depth Borehole log and mOD m b .g.l. description used, but 40 pellets were dropped into the hole as usual. J48 50.50 5.10 0.0 - 0.5m topsoil and backfill 0.5 - 5.10m brown London Clay The borehole was dry. J49 50.50 2.60 0.0 - 0.5m topsoil and backfill 0.5 - 2.6m brown London Clay The borehole was dry. J50 15.0 0.0 - 0.3m topsoil 0.3 - 8.5m brown London Clay 8.5 - 15.0, grey highly fissured Clay with lenses of fine sand and/or silt. The borehole was advanced using a percussion system of 15cm diameter. At 4.5m b.g.l. water entered the hole. Leaking of water into the hole was more or less impounded with subsequent driving of casing into the hole down to 5m b.g.l. The water level immediately after the borehole was completed, was 0.3m above the base of the borehole. To install a piezometer in this hole the same procedure as in Section (5-4-2) was followed,except that 10 litres of sand plus 5 litres of de-aired water were used in steps 2 and 4 and also fragments of clay Borehole No. G.L. Depth Borehole log and mOD m b.g.l. description were dropped into the hole when it was being filled with bentonite-cement grout. The borehole was sealed up to 4.5m b.g.l. Next day the casing was removed and the rest of the hole was sealed and completed. J51 65.74 12.0 0 .0 -0.3m topsoil 0.3 - 10.0m brown London Clay becoming very stiff after 8m b.g.l. 10.0 - 12.0m very stiff grey Clay. No water entered the hole, during or immediately after drilling. JT52 62.61 4.2 SameasJT21 J870 60.50 8.2 0.0 - 0.4m topsoil 0.4 - 6.5m brown London Clay 6.5 - 7.4m claystone and brown London Clay 7.4 - 8.2m brown London Clay. Down to 6.5m b.g.l. the hole was dry and while the claystone was being drilled water flushed into the hole and rose to 4.5m b.g.l., 0.7m of loose mud was generated due to drilling and subsequently accumulated at the base of the borehole. The piezometer tip (vibrating wire) was ------put in a small linen bag of sand and installed in the liquidy mud immediately after drilling. Borehole No. G.L. Depth Borehole log and mOD m b.g.l. description J871 63.92 8.10 SameasJ41. J872 60.65 4.50 0.0 - 0.5m topsoil 0.5 4.5m brown London Clay The borehole was dry. J873 70.00 5.40 0.0 - 1.2m topsoil and backfill consisting of clay, brick and concrete 1.2 - 5.4m brown London Clay. The borehole was dry. J874 62.63 5.10 0.0 - 0.3m topsoil 0.3 - 5.10m brown London Clay. The borehole was dry. J875 65.40 8.20 0.0 - 0.3m topsoil 0.3 - 4.0m brown London Clay 4.0 - 5.0m a layer consisting of hard claystone and Clay. 5.0 - 8.2m brown London Clay Down to 4m b.g.l. the borehole was dry. At 4.0 - 5m b.g.l. water entered into the hole, with the advancing of drilling, penetration of water into the hole was stopped. Probably the seepage passage was sealed with the clay which was coming up due to drilling. The borehole was dry when completed. J 877 62.83 5.20 SameasJ13 Borehole No. G.L. Depth Borehole log and mOD m b.g.l. description J878 65.00 12.00 SameasJ51 J879 65.50 4.50 SameasJ45 5-5 IN-SITU EVALUATION OF THE RESPONSE TIME OF STANDPIPE AND PNEUMATIC INSTALLATIONS If the ground water pressure is assumed constant, measurement of the response time (time lag) of an open standpipe piezometer corresponding to any percentage of equalization is not complicated. The water level in the piezometer is simply lowered or increased to a certain level and then the rate of change of the water level in the piezometer is measured and plotted against time until the water level has stabilized again. In the case of the closed systems, such as a vibrating wire or pneumatic piezometer, since it is not possible to change the water pressure in the intake point, assessment of the response time is not as simple as an open standpipe. In this research, to assess the errors in the measurements of the pore water pressure associated with the different types of piezometers it was initially important to compare the response times of a flexible open system with a rigid system in the site for the same soil and ground water conditions. There is a concave zone in the flat area, Zone 9, Fig. (5-2), where the rainfall water accumulates during the wet season and the water level remains steadily a few centimetres above the ground level for a few months of the year. A system comprising a standpipe piezometer and a pneumatic piezometer coupled with a tiny valve was installed in the above mentioned place as follows: (1) A standpipe piezometer (J31) was installed in a 4.3m deep borehole in the same way as explained in Sections (5-4-2). (2) A very delicate and tiny valve was fitted into the connector between the porous tip and the pipe of a standpipe piezometer. The valve was a spring valve which could be opened by pushing the conical nose on the top of the valve. The valve would automatically be closed after removal of the load from the nose. The maximum displacement of the conical nose was about 1mm. Details of the valve and the connector are shown in Fig. (5-23). To make sure that the valve was quite airtight it was initially connected to a nitrogen pressure supply unit and immersed in a bucket of water and then a 1.6 bar nitrogen pressure was applied to the valve. No gas bubbles were observed and after 15 minutes the Bordon gauge did not show any drop in pressure. To make sure that the valve was properly sealed in the connector, the piezometer was placed in the in situ calibration tube (refer to 4-2-2) under 7.5m head of water. After five days the piezometer was removed from the tube and inspected against any leakage from the tip through the connector and/or valve into the pipe. The system was found to be quite airtight and no leakage was observed. Then a pneumatic piezometer was prepared and calibrated in the laboratory (according to 4-2-2). Finally the pneumatic piezometer and the above mentioned standpipe were assembled to each other side by side, Fig. (5-23), so as the membrane of the pneumatic transducer and the nose of the valve stayed at the same level. The assembled system was then installed deliberately in a borehole of 7.5cm in diameter and 4.3m in depth (the borehole J37). Installation was carried out according to Section (5-4-2), except that special attention was paid to keep the nose of the valve and the interface of the sand filter and bentonite plug at the same level. The reason for this was that in this way any possibility of the replacement of the pore water of the sand with air bubbles during the opening of the valve would be eliminated. The position of this system in the borehole is schematically shown in Fig. (5-24). Now the duty of the standpipe was not measurement of the pore water pressure but to act as a supplementary element to release the pore pressure from the pneumatic transducer. To release the pressure a steel bar could be lowered into the standpipe to push the nose of the spring valve. The appropriate tests are explained in Section (5-7-1-1) and the results are shown in Fig. (6-1). 5-6 INSTALLATION OF THE AUTOMATIC TILTING BUCKET RA IN GAUGE Details of the automatic tilting bucket rain gauge have been given in Section (4-4). To install the device two conditions had to be met: (1) Preparation of a horizontal bed (platform) so that the concrete slab can be horizontally settled. It is a crucial factor to mount the device on an horizontal platform otherwise the bucket will not operate properly. (2) The device should be installed away from the rain shadow of any building, the recommended distance being at least two times the height of the nearest buildings. The device was installed at the far Northern end of Senate House car park on a gravel bed. The gravel was flattened and levelled and then the concrete slab and the device were installed respectively. The measurements are explained in Section (5-7-3). 5-7 IN SITU TESTS 5-7-1 Equalization tests In general, to carry out equalization tests the water pressure in the piezometer is decreased or increased to a certain level and then the rate of the compensation or dissipation of pore pressure in the piezometer is measured and plotted against time. In this research the ’’first time equalization test” is attributed to an equalization test carried out immediately after a piezometer has been installed. The characteristic of a first time equalization test is that the effect of the disturbance of the soil due to drilling and swelling of bentonite pellets and also setting of grout on the equalization of a piezometer can be studied if the first time and subsequent equalization curves of a piezometer are compared. More discussion on equalization can be seen in Section (2- 2-4). Two major series of equalization tests (Series A and B ) were carried out as follows: 5-7-1-1 Equalization test series "A” These tests relate to the system explained in Section (5-5) i.e. piezometers J31 and J37. The aims and procedure of these tests were as follows: To evaluate the rate of equalization of pneumatic and standpipe piezometer installations due to a sudden change of the ground water pressure, to see how much faster a pneumatic installation responds to a ground water pressure change (in a certain soil). To study the trend of effect of the disturbance of the soil due to drilling, swelling of the bentonite pellets and setting of the grout on the response time of a piezometer installation. The procedures for the tests were as follows: (1) The standpipe piezometer J31 was installed on 10/10/86 and subjected to the first time in flow (rising head) equalization test on the same date. On 24/2/87 the water level in this piezometer was lowered using the pump shown in Fig. (4-4), to enable the second time equalization to be carried out. (2) The pressure in the pneumatic piezometer J37 was released (lowered to zero) immediately after the borehole was sealed. As mentioned in Section (5-5), the release of pressure was carried out by lowering a steel rod into the standpipe and pushing the nose of the spring valve. Then at the same moment as the steel rod was removed from the nose of the spring valve, the time was recorded. Subsequently, the pressure build up Pt in the pneumatic piezometer was measured in progressive lag of time, until Pt had tended toward a constant value of Poo. Pt Variations of p—were plotted against time lag on a semi-logarithmic scale. This test was carried out again on subsequent dates. The dates and some characteristics of the equalization tests Series ”A” are summarized in Table (6-2), and the equalization curves are also plotted in Fig. (6-1) in Chapter 6. The first time equalization of piezometers installed in Zone 9,including piezometers J31 and J37,have also been shown in Fig. (6-1). 5-7-1-2 Equalization tests Series These tests were carried out with the following aims: As a criteria to assess the reliability of the standpipe installations and to check the appropriate piezometers against leakage of water into or from the standpipe. So, as mentioned in Section (5-2) any standpipe piezometers showing a doubtful pore pressure pattern was subjected to an equalization test. To estimate the engineering properties of the soil; k, m, c, in order to aid the correct interpretation of the records of piezometers and also to elucidate the visualisation of the pore pressure pattern and stratiography of the area. To evaluate the nature of the stress adjustment time lag (after drilling) and the effect of swelling of bentonite pellets and also setting of grout on the equalization. The equalization tests were accomplished by the following steps: (1) Depth of the water level from the ground level was measured initially, ho . (2) The standpipe was filled with de-aired water to the top of the plastic pipe and the time was recorded. (3) The depth of water level in the standpipe was measured in progressive time lags since the time which was recorded in step 2, ht.* The time was measured in minutes. Measurement of ht was continued until it tended toward a constant value. (4) The ultimate virtually equalized depth of water level in the standpipe was measured and recorded, (hoo). In most cases hoo was not appreciably different from h0. / r \ xr . . r. ht x 100 ■ (5) Variations of g- were plotted against time lag on a semi-logarithmic scalel The plottings are shown in Figs. (6-2) to (6-7). Measurement of ho is important in that one can get an idea about h» and so at any time during the equalization test the percentage of equalization can be estimated by h t x 100 calculating— . In this research, in equalization tests it was tried to concentrate the time lag measurements around the 50% and 90% equalizations. Since the subsequent estimations of the in situ permeability of soil; k, would be based on the above mentioned percentages of equalizations. Equalization of piezometers is more elaborately explained in Chapter (2). To minimize the effect of the fluctuations of the ground water pressure on the equalization tests, especially in the drained area, the equalisation tests were predominantly carried out in August and September during which piezometers were tending to show minimum pressure changes. For the piezometers installed in Zone 9, Fig. (5-2), the best time for equalization testing was the wet season, when the piezometers had initially been equalized and the water level was steadily a few centimetres above the ground level. In the place of piezometers J31 and J33, the water level is very close to the ground surface Therefore, the equalization was carried out through the inflow (rising head) test. The equalization curves are shown in Figs. (6-2) to (6-10) and also the piezometers which are subjected to the equalization tests, and the results of estimations of soil permeabilities are summarized in Section (6-2-3), table (6-4). 5-7-2 Daily measurements of pore water pressure Of the piezometers which have been installed in this area, 14 piezometers were subjected to daily measurements of pore water pressure for a minimum period of 7 months. These piezometers are: Zone Piezo. Piezo. Depth no. type m b.g.l. 2 505 standpipe 5.0 J873 vib.wire 5.0 4 J13 standpipe 5.0 J20 pneumatic 5.0 J877 vib.wire 5.0 J44 standpipe 8.0 J41------pneumatic 8.0 J871 vib.wire 8.0 J27 standpipe 4.8 Zone Piezo. Piezo, Depth no. type m b.g.l. J874 vib.wire 5.0 6 J45 standpipe 4.2 J879 vib.wire 4.2 J875 8.0 J878 11.80 Details of the position of these piezometers are shown in Figs. (5-11), (5-14), (5-12), (5-19) and (5-20). The objectives of these measurements are as follows: A comparison between the records of the three types of piezometers having same depth and position with respect to the drainage trenches, to evaluate the range of the co-relative errors in measurement of pore pressure using different types of piezometers. To evaluate the effect of the variations of the atmospheric pressure on the ground water pressure and relevant difficulties and errors. Precise evaluation of the mechanism of the efficiency of the drainage system on the stability of the area. Evaluation of the effect of the amount and duration of rainfall on the fluctuations of the ground water pressure in drained area and in turn on the mechanism of the efficiency of the drainage system on the lowering of the ground water pressure. It was tried to concentrate the daily measurements between 10am to 2pm. All of the records have been appropriately processed and plotted in terms of the depth of the piezometric level from the ground surface versus date, and are shown in Figs. (6- 11) to (6-16), 5-7-3 Daily measurements of the amount and duration of rainfall. The amount and duration of rainfall was automatically recorded with the automatic tilting bucket rain gauge installed in the Senate House car park. The device was adjusted to start its measurements at 9am GMT each day. So the records show the amount (mm/day) and the duration (hours/day) of the rainfall from 9am each day until 9am next day. The purpose of these measurements, as mentioned earlier, is first to correlate different patterns of rainfall (as part of the factors affecting fluctuations of the ground water pressure) and the fluctuations of the pore water pressure to find the most critical pattern of rainfall which can cause the most critical pore water pressure condition. The rainfall records have been processed and plotted for each month in Chapter 6, Figs. (6- 17). 5-7-4 Weekly measurements of pore water pressure All of the piezometers are subjected to weekly measurements of pore pressure, whether they were installed to establish pore pressure or for the error analysis. The piezometric levels have been measured either every week or every other week depending on the trend of rainfall and position of a piezometer. Any significant change of pore pressure which was encountered in daily measurements of pore water pressure has also been included here. The objective ° f these measurements was basically as follows: Establishment of the pore pressure pattern of the area. Evaluation of the efficiency and mechanism of the drainage system on the lowering of the piezometric level and in turn on the stability of the area. Evaluation of the pore water pressure pattern in the trenches. The records of the measurements have been processed and plotted in terms of the depth of the piezometric level from the ground surface versus date and shown in Figs. (6-20) to (6-77). 60 mm 6 0 mm 45 34m m 2.5 2.5 '— Verrfcilation hoi© New ventilation befon© improvement hoi© S=1/1 S=1/1 Fig (5-1) A commercial cap before and after improvement Fig.(5-2) The General site plan showing: - Drainage systems - Locations of piezometers - Trial trench (refer to the back of hard cover) J 1 I« I Fig (5-3): 1 - The probes for clearing standpipe piezometers. 2 - The Macintosh probes. 3 - The drilling Auger and bits. 0.00 .0 .0 0 -.1- -v■: Granular Hc*eri;’s - 4.10 I ■---- 1______1 0 HS.V Scale 4m Fig (5-4) Locations of trenches, section kl- k2, Zone 7 J------16-90 —------*J 00 0.50 Granular Material, 0 H&V Scale 4r Fig (5-5) Locations of trenches, section k3 - k4, Zone 4 -6-30- ! East u 10-0 m f--0-80 i 1 i it? %-5-0 I.-5.0 i------1______I______u I 0 H& V Scale 4 m Fig (5-6) Locations of trenches, section k5 - k6, Zone 2 ooo Fill Materials Granular— M aterials —S :•!:> o H £ V Scale 4m Fig (5-7) Locations of trenches, section k7-k8, Zone 6 Fig. (5—8) Longitudinal Profile i 58 Longitudinal profile 5-5,in undrained area Fig (5-8) cn P -H C •P ■ 0_Q 0 0-0 0-0 3: h 0 .H {.0.-H 0. f -H f j 0 « > n a 0 o £ IN (0 N (V (0 CD (0 10 to rJn? CD mOD 70 o “ | tandPipe r n e u t n o t i c 7 Vil=. Wir*e 0Q* 56 VO 52 10 '—' H & V S e a l 58 60. 93.I c«—► Cu 54 £3* fil 13 23.8 w « | - S. 2 a 50 O 50. > 45 42 1 3 .5 38 4 0 70 A l—t 0 L 3 4> U 0 nc 0 0) 0 CM V r - i H © e > > 8 H r-i • • 0 0 N H 0 e n •rt £L Q. • ft X X c 0 a n 2 s S n B H ca O a CD r - 0-M0 0 l0 •* -P M 0.0 5: "0 = 0C 30 _Q . c CO 0 . > n n n CO > o3 CM X Fig (5-10) Locations of piezometers J13, J27, J29 & JT21 along cross section k3-k4, and piezometric levels, Zone 4. lOD 65- 1.40 ■0-97 0. 9 0 - • 1.15 T 1.47 64. 63- 6 2 ■ 61- 60' 59 58 57 56 J.42 55- 0 H&V Scale 2 m 5 4 •~ St andpip ® = Pnsumati J 40 6-10.0 V = Vab T T T T T T T 112 ' 110 ' 108 ' : 106 Fig (5-11) Details of locations of piezometers J877, J13, J40, J20 and J41 along profile A-A Fig (5-12) Locations of piezometers J20, J40, J41, J42, J44 J44 J42, J41, J40, J20, piezometers of Locations (5-12) Fig j*1.0 *^-1.0 ^ 7.90 CM CM & J871 along cross section k3-k4, section cross along J871 & ■H 5: 0 1 {.0.-H -rl v D CD 2 a: l Lii LlJ CO to M 10 3s -i O O Fig (5-13) Locations of piezometers J877 & J874 along cross section cross along & J874 J877 of piezometers Locations (5-13) Fig 16-90 CM k3-k4, and piezometric levels, Zone 4. Zone levels, piezometric and k3-k4, ts- o - 0)0. > 0 E "0 rl N N > > M M Ui UJ ! -J _! . 0.Q. O O LU Ld Lii I - Ui I - “I'M" •h P C ■P I I II II II I C 3 . . 3 C O JO 0 0 0.0 5: 0.0 C CL-h 0 0 0 £ X 0 > LT) 3 u 4 -H) ft! i n - 1.85 0.45 GL 1 - MAX. PIEZO. LEVEL 2 - MIN. PIEZO. LEVEL J874 '-5-0 ° H&V Scale 2m ^ = Standpip© ° “ Pneumatic V 53 Vib. Win© Fig (5-14) Longitudinal profile showing locations of piezometers J27 & J874 and piezometric levels, Zone 4. L l O Q ig. (5-15) Longitudinal Profilo E WH> •ri Q-05: 0.-H L o 4 *r| ) CD i 51) Longitudinal profile 6-6 Fig (5-15) * u gs 4 c - - ft-! _ (O to CP v > + " + 14.90 Trial Trench H-0 6.45 6-45 1-0-1 Ease K2 0-05 .0-0 0 0-08 0.00 Fill Materials GranuUr^MgfgHsls" JT15 Tl9 -4.10 I872 -3 .9 5 k20 j l l ? MAX. PIEZO. LEVEL , J870 4m MIN. PIEZO. LEVEL . J870 H&V Scale Standpi po Pneumatic J870 Vi b. Win© — 8.20 1 « MAX. PIEZO. LEVEL INFERED FROM THE RECORDS OF PIEZOMETER J24 2 « MIN. PIEZO. LEVEL IN J24 3 « MAX. PIEZO. LEVEL AT TRENCH INVERT LEVEL ACCORDING TO PIEZOMETERS JT15 . J16 . J17 , J18 & JT19 4 *= MIN. PIEZO. LEVEL AT TRENCH INVERT LEVEL ACCORDIG TO PIEZOMETERS JT15 , J16 . J17 . J18 & JT19 THE GROUND-WATER PRESSURE PROFILE IS SHOWN IN Fig. <8-3> P R O F I L E 9 Fig (5-16) Locations of piezometers JT15, J16, J17, J18, JT19, J22, J24, J870 & J872 along cross section kl-k2, and piezometric levels, Zone 7. o C\J inE - I CM i— C3 CO ->o _! d cd CM o COM N \ o a co n . CD 10 '3‘ CD io 10 LO LO CO CO Fig (5-16-a) Longitudinal profile showing locations of the trial trench invert level and piezometers JT14 & JT15. 65 64 63 ~J25 1-2.5 62 J45 -4.2 J87Q 61 60 J43 '-6 .0 59 58 J875 0 H&V Scale 2m J -8.0 57 — Standpip© O — Pneumatic V = Vi b. Wi o© 56- MAX. PIEZO. LEVEL MIN. PIEZO. LEVEL 55' THE GROUND-WATER PRESSURE PROFILE IS SHOWN IN Fig. (8-3> PROFILE 10 54- J878I 70 69 68 67 66 65 v_11.8 64 63 53- -17) Details of locations of piezometers J25, J43, J45, J875, J878 & J879 along profile 6-6, and piezometric levels, Zone 6. 14.80 ~ 7.40 6.40 0.00 Q.QO 0.00 Fill Materials Granular Material J25 -2.50 J45 JT23 -4-0 .J879 J46 -4.20 -4.30—' 4m J43 0 H&V Scale -6-0 '— Standpip© O — Pneumatic J875 -8-0 l j8 7 8 u—11.80 1-11.80 1 « MAX. PIEZO. LEVEL AT TRENCH INVERT < CONJECTURAL) 2 « MAX. PIEZO. LEVEL AT TRENCH INVERT 3 ~ MIN. PIEZO. LEVEL AT TRENCH INVERT THE GROUND-WATER PRESSURE PROFILE IS SHOWN IN ON Fig. <8-3) PROFILE 10 Fig (5-18) Locations of piezometers J25, JT23, J43, J45, J46, J51, J875, J878 & J879 along cross section k7-k8, and piezometric levels, Zone 6. •6.30— 2.50 cL—1.35->-j—1-30- -1.25- ■ 2.40 r East 1 0.0 0.0 K6 -0-80 5 o 4 ^ £ so 3 P505 J873 -5 .0 -5 .0 -5 .0 - 5 .0 * = Standpip© MAX. PIEZO. LEVEL 4m ° — P n e u m a t i c 0 H&V Scale MIN. PIEZO. LEVEL 7 - Vib. Win© Fig (5-19) Locations of piezometers 503, 504, 505 & J873 along cross section k5 - k6, and the profile showing locations of piezometers 505 & J873, and piezometric levels, Zone 2. 1.85 0.45 GL Average piezo- metric level in tnench. R e f e r to Fig. C - 3 3 5 JT21 .= 4.4 J T 5 2 -4.4 S t a n d p i p 7 — Vib. W i n © !______>______I 0 2m H&V Scale Fig (5-20) Longitudinal profile showing locations of piezometers JT21 & JT52 in a trench and piezometric levels in Zone 4. Fig (5-21) The drilling rig (Minuteman) I tm rsim m m ,ium l.*, »m ?. / r j Fig (5-22) The casing pipes and the conical tips. Fig (5-23) Details of the system J37. Spooial rod to open the spring valve to release pone water preseure from eand filton and the diophrogm. The pneumatio piezometer t w i n t u b e GL 0 . 0 0 Bentonite~Cement g n o u t 3 100 H&V Scale Bentonite pellets The epning valve p l u g nipple level_____ Pneumatio piezometer diaphnagm level St. pi p e piezometer- p o r o u s tip •rt Sand fi1ten Diameten “ 75 mm ~4. 50 m Fig (5-24) Schematical details of installation of system J37 in the bore hole. CHAPTER SIX - RESULTS 6-1 SUMMARY OF CHARACTERISTICS OF THE PIEZOMETER INSTALLATIONS. To provide a quick reference to the details of the piezometer installations, the characteristics of all piezometers, installed in Stag Hill since 1965, are summarized in table (6-1). Piezometers series J and JT including forty-one standpipe, eleven pneumatic and nine vibrating wire piezometers which have all been installed during this research since 1986. Series JT are piezometers installed in the drainage trenches. The methods of calculation of the shape factors F, and radius of the equivalent spherical intake point ae of piezometer installations, given in table (6-1), are explained in section 6-2-3. The F and ae values for piezometers KP1 and KP3 are not given, because they are entirely backfilled with clay and cement slurry respectively. Table (6-1) Summary of characteristics of piezometer installations (New piezometers) Piezo Piezo Depth of Ground Bore Hole St.Pipe Height of Shape Equiv. No. Type Element level diameter Diameter sand filter factor radius (F) (a*) m b.g.l. mOD an mm cm cm cm J1 St.pipe 7.5 60.10 7.5 20.2 60 153 12 J2 99 5.5 60.10 10 99 99 169 13 J3 99 2.5 60.10 10 99 99 169 13 J4 99 9.0 62.3 7.5 99 99 153 12 J5 99 5.0 62.3 99 99 99 153 12 J6 99 7.0 62.3 99 99 90 202 16 J7 99 3.0 62.3 99 99 60 153 12 J8 99 8.0 69.0 99 99 99 99 12 J9 99 6.0 69.0 99 99 99 99 12 J10 99 3.0 57.2 99 99 ?? ? J l l 99 4.0 57.15 99 99 60 153 12 J12 99 2.5 57.15 99 99 ? ? ? J13 99 5.0 63.26 99 99 60 153 12 99 99 JT14 4.8 59.75 - --- 99 99 JT15 4.1 59.90 - --- J16 pnum 4.15 60.0 7.5 60 153 12 J17 99 4.0 59.80 99 - 99 99 99 J18 99 4.1 60.00 99 - 99 99 99 JT19 St.pipe 4.0 60.00 - 20.2 - - - J20 pnum 5.0 63.78 7.5 - 60 153 12 JT21 St.pipe 4.2 63.06 20.2 -- - J22 99 4.2 60.20 7.5 99 60 153 12 99 99 JT23 4.3 65.70 - --- J24 99 2.5 61.20 7.5 99 50 136 11 J25 99 2.5 65.18 99 99 99 99 99 J26 99 2.5 71.20 99 99 99 99 J27 99 4.8 63.06 99 99 60 153 12 J28 99 4.5 71.60 99 99 99 99 99 J29 99 5.0 62.46 99 99 99 99 99 Table (6-1) Summary of characteristics of piezometer installations(Cont.) (New piezometers) Piezo Piezo Depth of Ground Bore Hole St.Pipe Height of Shape Equiv. No. Type Element level diameter Diameter sand filter factor radius (F) (ae) m b.g.l. mOD an mm an an cm J31 yy 4.0 7.5 » yy yy 12.0 J33 yy 1.5 10.0 » yy 169 13.0 yy J34 pnum. 4.0 7.5 - 153 12.0 » yy yy yy yy J35 1.5 - yy yy yy yy J36 4.0 - yy yy yy yy J37 4.0 - J38 St.pipe 5.0 62.30 yy 20.2 yy yy yy J39 yy 3.0 62.30 yy » 40 118 9.5 yy J40 Pnum. 10.0 63.60 - 60 153 12.0 yy yy yy yy J41 8.0 63.98 - 12.0 yy yy yy yy yy J42 9.0 63.90 - J43 St.pipe 6.0 65.22 yy 20.2 yy yy yy J44 yy 8.0 63.96 yy yy yy yy J45 yy 4.2 65.66 yy » yy yy yy J46 yy 4.4 65.70 yy » yy yy yy J47 yy 10.0 71.56 yy » 53 4 J48 yy 5.0 50.50 yy 20.8 60 153 12 J49 yy 2.5 50.50 yy yy yy yy 12 J50 yy 15.0 15 yy yy 199 16 J51 yy 11.8 65.75 7.5 yy yy 153 12 JT52 yy 4.2 62.61 _ 20.2 . - J870 Vib.Wire 8.0 60.50 7.5 - 60 153 12 yy yy yy yy J871 8.0 63.92 7.5 - yy yy yy yy yy J872 4.0 60.65 - yy yy yy yy yy J873 5.0 70.00 - yy yy yy yy yy J874 5.0 62.63 - yy yy yy yy yy J875 8.0 65.40 - J877 _____ yy ------5.0 62.83 - y y — yy — yy yy yy yy yy yy yy J878 11.8 65.0 - yy yy yy yy yy J879 4.2 65.5 - Table (6-1) Summary of characteristics of piezometer installations(Cont.) (Previous piezometers) Piezo Piezo Depth of Ground Bore Hole St.Pipe Height of Shape Equiv. No. Type Element level diameter Diameter sand filter factor radius (F) (ae) m b.g.l. mOD an mm cm on cm 101 St.Pipe 4.2 71.6 15 19 60 199 16 102 99 18.0 71.75 99 99 . 99 99 99 103 99 4.4 64.3 99 99 99 99 99 104 99 7.7 64.37 99 99 99 99 99 105 99 3.2 60.0 99 99 99 99 99 106 99 6.0 60.0 99 99 99 99 99 107 99 13.5 60.2 99 99 99 99 99 108 99 2.8 56.78 99 99 99 99 99 109 99 1.5 53.60 99 99 99 99 99 110 99 14.0 53.66 99 99 99 99 99 111 99 18.0 71.72 99 99 99 99 99 112 99 13.3 59.90 99 99 99 99 99 201 St.pipe 4.7 68.03 99 99 99 99 99 202 99 6. 63.62 99 99 99 99 99 203 99 12.4 63.40 99 99 99 99 99 204 99 3.6 60.08 99 99 99 99 99 205 99 3.1 53.94 99 99 99 99 99 206 99 6.1 63.40 99 99 99 99 99 207 99 5.7 63.40 99 99 99 99 99 208 99 9.0 62.80 99 99 99 99 99 401 99 4.8 69.20 99 99 99 99 99 402 99 5.7 63.60 99 99 99 99 99 403 99 5.5 64.00 99 99 99 99 99 404 99 5.5 64.00 99 99 99 99 99 405 99 5.4 61.40 99 99 99 99 99 _ __ 406 99 — 6.5 58.30 99 99 99 99 407 99 5.3 66.40 99 99 99 99 99 Table (6-1) Summary of characteristics of piezometer installations(Cont.) (Previous piezometers) Piezo Piezo Depth of Ground Bore Hole St.Pipe Height of Shape Equiv. No. Type Element level diameter Diameter sand filter factor radius (F) (ae) m b.g.l. mOD an mm an a n cm 408 6.0 64.50 409 6.0 59.03 15 19 60 199 16 410 6.0 58.98 411 6.0 61.37 412 6.20 59.24 413 6.0 58.93 414 3.6 62.80 415 3.3 57.70 416 3.3 57.70 417 4.8 57.60 501 5.0 69.2 7.5 15 50 136 11 502 5.0 72.2 503 4.8 67.1 504 5.0 67.10 505 5.0 67.10 506 5.0 65.50 507 5.0 61.00 508 4.8 55.50 509 4.8 71.30 510 5.0 67.30 511 5.0 63.50 512 5.0 71.60 513 5.0 ? 514 5.0 ? 601 6.0 50.50 15 15 60 199 16 602 12.0 50.10 Table (6-1) Summary of characteristics of piezometer installations(Cont.) (Previous piezometers) Piezo Piezo Depth of Ground Bore Hole St.Pipe Height of Shape Equiv. No. Type Element level diameter diameter sand fitter factor radius (F) (a j m b.g.l. mOD cm mm an cm cm 701 45.0 50.50 15 2 2 60 199 16 702 15.0 51.50 yy yy . yy yy yy 703 30.0 51.50 yy 15 yy yy yy KP1 9.0 51.80 15 22 . KP2 9.0 51.80 yy yy 60 199 16 yy yy KP3 9.0 51.80 -- - A ” ? 70.10 15 19 60 199 16 B » 9 67.06 yy yy yy yy yy C ” ? 64.01 yy yy yy yy yy D ” ? 60.96 yy yy yy yy yy 6.2 RESULTS OF EQUALIZATION TESTS AND PERMEABILITY ANALYSIS. 6.2.1 Results of equalisation tests series ”A” Results of the equalization tests series ”A” (Refer to 5-7-1-1) are processed and shown in Fig (6.1) Some characteristics and results of tests are summarized in Table (6-2). In column one (from left), the tests are tabulated in the order they were conducted. For example, J37/1 is attributed to the first time equalization test on piezometer J37. The first time equalization test is an equalization test which is carried out immediately after a piezometer has been installed. To carry out equalization tests with piezometer J31, the water level in the standpipe was decreased to 3.5m below ground level. In test J31/1 to shorten the time of duration, the record of pneumatic piezometer J34 was accepted as the ultimate pore pressure in piezometer J31. For this reason the corresponding equalization curve, Fig (6-1), is not complete. Table (6-2) Some characteristics and results of equalization tests series "A”. Test start Poo Po App. *50 *90 *90450 order date bar bar pressure bar min. min. J31/1 10/10/86 25x103 11.66x104 4.70 J31/2 24/2/87 - - - llx lO 3 39x103 3.50 J37/1 27/1/87 0.38 0.00 0.90 87x101 28x 102 3.20 J37/2 3/2/87 0.38 0.00 0.90 18 17x101 9.4 J37/3 10/7/87 0.35 0.00 0.90 18 12.5x101 6.9 J37/4 13/7/87 0.35 0.00 0.90 30 18x101 6 J37/5 22/7/87 0.36 0.01 1.60 18 11x101 6.1 J37/6 23/7/87 0.36 0.01 0.90 16 95 5.9 Poo = Pressure corresponding to 99.99% equalization Po = pressure corresponding to 0.0% equalization App.pressure = nitrogen pressure applied to transducer during the test, tso & t 90 = times for 50 and 90 per cent equalization. Note: For more details refer to Fig (5-24) and section 5-7-1-1. 6.2.2 Results of the equalization tests series ”B” Results of the first and subsequent time equalization tests series ”B” (Refer to 5- 7-1-2) are processed and plotted in Figs (6.2) to (6.3) and Figs (6.4) to (6.8) respectively. It is again to be noticed that the first time equalization test is attributed to a test which is carried out immediately after a piezometer has been installed and the borehole sealed up. In figs (6-9) and (6-10) the first and second time equalizations of standpipe piezometers are compared. 6.2.3 Results of permeability measurements and calculations. 6-2-3-1 Records of permeability tests carried out in the past (1966). The available permeability values yielded by constant and/or falling head tests carried out prior to the digging of any drainage trenches and also after completion of the drainage system, are given in Table (6.3). 6-2-3-2 Results of permeability measurements carried out during the research. To estimate the permeability of the ground at appropriate places using piezometer equalization curves, Brand and Premchitt’s method (1981) based on Gibson’s solution (1963) was used (Refer to 2.2.4). The shape factor values (F) for permeability analysis were calculated using the following equation (Premchitt and Brand, 1980): F ______2.4tt! Ln[1.2 l/d + A/H7ITl/d)2 ] I = height of sand filter d = diameter of borehole and then a^, the radius of the equivalent spherical intake point, was calculated from: F Values of F and ae are tabulated in Table (6-1). Permeability calculations are based on the ratio of times corresponding to 50% and 90% equalizations (t9(/tso) and the results are summarized in Table (6-4). Table (6-3) Results of permeability measurements in 1966. Piezo. Depth Measured prior Measured after No. m b.g.l. to drain trenches completion of drain trenches Const. Falling Const. Falling Head Head Head Head cm/sec cm/sec cm/sec cm/sec 101 4.2 8x10-8 102 18.0 1.5x10-5 104 7.7 2.3x10-7 105 3.2 1.1x10-5* 2.3x10-5 106 6.0 1.4xl0-8 107 13.5 1.7x10-5 108 2.80 8x10-6 109 1.5 4.4x10-6 1.1x10-6 110 14.0 1.3x10-5 111 18.0 2.2x10-8 201 4.7 1.7x10-7 6.6x10-8 202 6.0 2.3x10-8 4x10-8 2.4x10-8 203 12.4 5.6x10-8 5x10-8 204 3.6 205 3.1 2.7x10-6 1.7x10-6 206 6.1 6.8x10-8 207 5.7 5x10-8 * limit of apparatus Table (6-4) Permeability measurements using pore pressure equalization of piezometers (Premchitt & Brand’s Method, 1981). Piezo. Depth k value from first k value from No. m b.g.l. time equalization subsequent time test equalization test cm/sec cm/sec J1 7.5 - 4.8*10-8 J4 9.0 - 3.1x10-8 J6 7.0 - 2.9X10‘8 J8 8.0 - 8.4X 10'8 J9 6.0 - ■ 3.4X10-6 J22 4.4 - 2.8X10-8 J24 2.5 l.lxlO'8 - J25 2.5 - 3.1 x 10-8 J31 4.5 5.1xl0-9 1.5 x10-8 J33 1.5 - 5.7X10-8 J38 5.0 - 4.4X10-8 J39 3.0 - 2.4X10-7 J43 6.0 6xl0-9 - J45 4.2 - 2.5X10'8 J48 5.0 4.7xl0-8 - J49 2.5 - J50 15.0 1.23xl0-8 - 701 45.0 - 3.3X10-6 702 15.0 - l.OxlO-5 703 30.0 - 2.1x10“8 601 6.0 - 1.8X10-8 602 12.0 - 2.3X10-5 416 3.3 - 2.9X10-6 417 4.8 _ 1.8X10-5 Note: The first time equalization is a test which is carried out immediately after a piezometer has been installed. 6.3 RECORDS OF DAILY MEASUREMENTS OF GROUNDWATER AND ATMOSPHERIC PRESSURES Daily records of five groups piezometers of different types, in different zones, are shown in Figs.(6-11) to (6-16) and also Figs.(6-18) and (6-19). The piezometers subjected to daily readings, figures showing their locations in plan and profile and figures showing corresponding records (bar charts) are summarized in table (6-5). Table (6-5) List of figures corresponding to the piezometers subjected to daily measurements Piezometers Location in plan Location Records No. Fig(5-2) in profile (Barcharts) J45,J879,J875,J878 Zone 6 Fig(5-17) Figs(6-ll),(6-12) &(6-18) J13,J20,J877 Zone 4 Fig(5-ll) Figs(6-13)& (6-19) J44,J41,J871 Zone 4 Fig(5-ll) Fig(6-14) J27,J874 Zone 4 Fig(5-14) Fig(6-15) P505,J873 Zone 2 Fig (5-19) Fig(6-16) Records of each piezometer at any point are shown in terms of depth to piezometric level from the ground surface at that point. To facilitate the comparison between the records of piezometers and changes of atmospheric pressures, the latter records are plotted on the same graphs as the records of piezometers. Records of two sets of piezometers are corrected against the atmospheric pressure and then shown in Figs (6-18) and (6-19). The method of correction is discussed in Section 8-2-2-3. 6-4 RECORDS OF DAILY MEASUREMENTS OF THE AMOUNT AND DURATION OF RAINFALL. Records of daily measurements of the amount (mm/day) and duration of rainfall (hours/day) are shown in fig (6-17). The automatic device was not in working order from June 1st until June 6th, so, the amounts of rainfall were measured using a simple basic rain gauge and obviously the measurements of duration of rainfall was not possible in that period. 6-5 PIEZOMETER RECORDS, SINCE 1965. The available records of piezometers, installed in Stag Hill since 1965, are collected and plotted together with the records of 59 new piezometers installed during the current research. The plottings are shown in figures (6-20) to (6-77). To plot the records they are classified into six groups as follows: 1 Records of piezometers which were lost before or during construction of trenches. 2. Records belonging to the time before and a certain period after installation of drainage system. 3. Records of piezometers after drainage. These records belong to the piezometers which were installed in the drained area, or in the areas affected by drainage trenches. 4. Records of piezometers installed in the trenches to monitor the pore water pressure in the fill material. 5. Records of piezometers in undrained areas of slope. These records belong to the piezometers which were installed on the hill side, away from the drainage system. 6. Records of piezometers in a low lying, undrained area. This area is located on the flat playing fields to the north of Stag Hill, Figs(3-2) and (5-2). To simplify the visualization of the groundwater pressure pattern, piezometers with different depths and close to each other and in more or less the same positions with respect to drainage trenches, are assumed to have been installed on the same vertical profiles. Accordingly their records are mainly plotted on the same figure. Records of piezometers which are installed in the same depth and along the same cross section between two adjacent trenches are also plotted on the same figure. At the end, the records of piezometers in three places before and some years after installation of trenches are compared in Figs (6-78) to (6-80). It should be mentioned that as the depths of piezometers A, B, C and D remained unknown, their records are not included here. Records of these piezometers are given in the appendix, however. Finally, Stag Hill is an important case of a stabilized slope in London clay, having the ground water pressure and other engineering data since the time before installation of trenches. Accordingly, the available records of piezometers, regardless of the importance of them in this research, are all included in this thesis for the benefit of the people who will carry out further work. The collection of the records of piezometers installed years ago and putting them in a correct and appropriate order was a very difficult and time consuming task requiring much patience. RATE Of EQUALIZATION ( P. I ) RATE OF EQUALIZATION ( P. 100 100 20 30 50 60 70 90 80 40 10 20 50 90 30 60 70 80 40 10 0 0 1 i. 6 2 RESULTS TIME FIRST EQUALIZATION OF ) 2 PIEZOMETERS OF - 6 ( Fig. i.( - ) EUT O EULZTO TSS EIS "A" SERIES TESTS EQUALIZATION OF RESULTS ) 1 - ( 6 Fig. 10 , 7 3 J , 7 3 J 100 100 TIME ( minute ) (minute TIME IE (minute) TIME 1000 0 010.000 1000 , 1 / 7 3 J . 7 3 J 10000 / 9 4 J 1 / 0 5 J 100000 100000 / 4 4 J 1 / 3 3 J /3 3 4 < J 1000000 icooooo RATE OF EQUALIZATION ( R 1 ) RATE OF EQUALIZATION ( R I ) 100 100 20 30 30 50 60 70 90 50 70 20 80 90 60 80 40 40 10 10 0 0 . 1 10 00 00 100 1000000 100000 10000 1000 100 10 1 i. 6 4 RESULTS EQUALIZATION OF ) 4 ) PIEZOMETERS OF "B" - TESTS SERIES 6 ( ( Fig. i.( - ) EUT O FRT IE QAIAIN F IZOMETERS PIEZ OF EQUALIZATION TIME FIRST OF RESULTS ) 3 - ( 6 Fig. 1 1 /I 2 P K 10 2 / 2 0 6 /2 6 1 4 TIME ( ninutc ) ( ninutc TIME 1E ( 1 min. T1HE ' * 100 1 / 4 2 J 1000 1 / 1 5 J RATE OF EQUALIZATION ( R Z ) RATE OF EQUALIZATION ( R Z ) 100 100 100 80 90 20 30 50 70 60 40 10 20 30 50 90 70 60 80 40 10 0 0 1 10 00 00 100.000 10000 1000 100 10 1 i. 6 6 RESULTS EQUALIZATION OF ) 6 ) OF "B" PIEZOMETERS - SERIES TESTS 6 ( ( Fig. 1 i. 6 5 RSLS F QALZTO O PE MTR ( ET SRE "" ) "B" SERIES TESTS ( OMETERS PIEZ OF LIZATION EQUA OF RESULTS ) 5 - ( 6 Fig. 10 0 1000 100 X'— 2 / 3 3 J /2 1 0 7 TJHE ( minute ) ( minute TJHE IE () minute TIKE ✓ 2 / 4 J 2 / 3 0 7 2 / 8 J 10,000 / 1 3 J l M 100.000 i g . ( 6 7 ) RESULTS OF EQUALIZATION OF PIEZOMETERS ( TESTS SERIES "B" ) 100 90 80 70 60 50 40 30 20 10 0 10 100 1000 10000 100000 TJHE ( minute ) RATE OF EQUALIZATION ( R Z ) RATE OF EQUALIZATION ( R Z ) 100 100 100 20 30 50 20 90 70 90 30 50 70 60 80 40 80 10 40 60 10 0 0 0 0 10 100 000 1000000 100000 10000 1000 100 10 1 10 00 00 100000 10000 1000 100 10 1 i. 6 9 A OPRSN EWE TE IS & EOD IE QAIAINS EQUALIZATION TIME SECOND & FIRST THE BETWEEN COMPARISON A ) 9 - ( 6 Fig. i. 6 0 A COMPARISON A ) 10 BETWEEN - 6 THE 8, SECOND ( FIRST TIME EQUALIZATIONS Fig. ------— - - ______------IE ( minute TIHE TIHE ( minute ) ( minute TIHE ✓ ✓ ✓ ✓ — ✓ ✓ - ) t f t / / t 1 / 0 5 J / / / / 0 5 J / / / / — / / / / : / / r I / / / / / i k /■ t I f , t t / / / . »*- ✓ / f / s 4 / // / / // /; t t / f - ~ i f ;’ j; i i i f 2 // /•' / / ;/ // 1 / , j; /; / / V 2 : DEPTH TO PIEZOMETRIC LEVEL (cm) P DEPTH TO PIEZOMETRIC LEVEL (cm) P DEPTH TO PIEZOMETRIC LEVEL (cm) 150 5 -1 -350 -300 -200 -500 -250 -300 -100 -350 -250 -200 -J50 -250 -500 -500 -350 -300 -350 -100 -50 50 -5 50 -5 ■200 i . AL EOD FPEOEESN. 4 J7 s J875 s J879 . J45 No. PIEZOMETERS OFRECORDS DAILY Fig. AL EOD FPEOEESN. 4 J7 J7 DPH . 4.2 ; 4.2 = DEPTHS J875 ; J879 ; J45 No. PIEZOMETERS OF RECORDS DAILY __ i_ oebr 1987 November uut 1987 August “—!~ T. PRESSURE ATM. VJRE VJB. JB75 STANDPJPE J<5 T. PRESSURE ATK. : VJRE VJB.J875 . STANDPJPE J<5 PRESSURE ATM. VJRE VJB. J785 STANDPJPE J*5 89VB VJRE VJB. J879 89VB VJRE VJB. J879 B9VB VJRE VJB. JB79 j _ : 0 2 3 300 § 1300 1000 500 500 500 § -250 § 3 s —i> m in m 5 n 73 m X 13 o t in tr □ ' UJ L > -350 j * - UJ NJ UJ O o -250 g H u Q . Q s K -350 x -350 x -250 y u e l I -300 -200 -500 -250 -150 -500 -300 -200 -300 -200 -300 -50 -50 •500 -50 <00 350 > 2 - 5 ( . s g i F O T R E F E R DAILY RECORDSOF J45 PIEZOMETERS ;; No. J879 DAILY eebr 1987 December t ) 8 1 - 5 ( a ) 7 1 - 5 < U i.(-2 DIY EOD O IZMTR No PIEZOMETER OF RECORDS Fig. DAILY (6-125 DEPTH TO PIEZOMETRIC LEVEL (cm) DEPTH TO PIEZOMETRIC LEVEL (cm) -1100 0 5 0 -1 3150 5 1 -3 0 0 0 -3 -1000 1350 5 3 -1 950 5 -9 0 0 -9 0 0 -8 0 0 -7 950 5 -9 0 0 -9 800 0 -8 850 5 -8 0 5 -7 650 5 -6 700 0 -7 2 4 6 8 1 1 1 3 1 3 1 1 1 3 20 2 22 2 0 3 9 2 8 2 7 2 6 2 5 2 4 2 23 2 2 21 0 2 39 18 17 16 35 14 33 12 11 10 9 8 7 6 5 4 3 2 1 R E F E R T O F i g s . <5-2> <5-2> . s g i F O T R E F E R AL EOD O IZMTRN. 88 ET = I8 m II.8 = DEPTH J878 No. PIEZOMETER OF RECORDS DAILY oebr 1987 November s 51> <5-18> & C5-17> T . PRESSUREATM. WIRE VJB. J878 T . PRESSUREATM. WIRE VJB. JB78 J878 1000 0 0 31 0 0 0 3 300 0 13 0 0 9 0 0 9 DEPTH TO PIEZOMETRIC LEVEL (cm) P DEPTH TO PIEZOMETRIC LEVEL (cm) rr DEPTH TO PIEZOMETRIC LEVEL (cm) -500 -450 -350 -250 -200 -300 -300 -350 -300 -350 -300 -200 -250 -500 -50 20 u -200 -300 -50 -250 -500 -350 -50 Fig. J20 ; J13 No.PIEZOMETERS <6— OF RECORDS DAILY 135 AL EOD O IZMTR o J3; 2 J7 DPH 5; 5 ; 5 ; 5 = DEPTHS JB77 ; J20 ; J13 No. PIEZOMETERS OF RECORDS DAILY J' , oebr 1987 November uut 97 etme 1987 September 1987 August T. PRESSURE ATM. PNEUHAT J20 J C STANDPJPE JJ3 67VB VJRE VJB. J677 ATM. C Vie. U .A JB77 W E U P 0 2 J J20 PNEUHATJC PNEUHATJC J20 STANDPIPE JJ3 T. PRESSURE ATK. VJRE VJB. JB77 L.. "L PRESSURE VJR£ VJR£ 3000 3000 900 30 § 1300 3000 900 900 x m n i m P3 -a n m x x —i> in -a m o -350 x -250 y uj -350 x j -250 uj -350 x -250 -500 -300 -200 -350 -500 -300 -300 -200 -300 -350 -300 -200 -500 -50 -50 -50 REFER TO F ig e . (5 -2 ) j (5-1B ) ) (5-1B j ) -2 (5 . e ig F TOREFER DAILY RECORDSDAILY OF J13; J20 PIEZOMETERSNo. JB77 s ,r~ ( 13) 3 -1 (5 S eebr 1987 December , 5- ) 2 -1 (5 GL -350 x -350 -300 -250 DEPTH TO PIEZOMETRIC LEVEL (cm) -500 -50 -300 •200 -200 -250 •300 -300 •350 -500 -50 -200 -<00 -50 ■350 •300 •250 ■500 •350 Fig. (6-14) DAILYRECORDS OF PIEZOMETERS No. J44 ; J41 ; J871 DAILY RECORDS OF PIEZOMETERS No. DEPTHS J44= ;8 J41 ; ;B JB71 B; n 9 30 33 12 33 1< 35 35 37 18 39 20 21 22 23 2< 25 25 27 28 29 30 29 28 27 25 25 2<23 22 21 20 39 18 37 35 1< 35 33 12 33 30 9 oebr 1987 November uut 1987 August J O PNEUMATIC PNEUMATIC O J STANDPJPE « J T. PRESSURE ATM. VJRE VIS. JB7J J « STANDPJPE STANDPJPE « J T. PRESSURE ATM. VJRE VJB.JS71 PNPJKATJC J<1 JB7J VJB. VJRE VJRE VJB. JB7J STANDPJPE J<< T. PRESSURE ATM. PNEUMATIC J4J 3 300 3000 500 3000 § 3 300 900 900 3 -250 y -250 y -350 x -250 y -350 x -350 -300 -300 -300 -500 -50 -50 ■200 •500 •200 •350 •300 •200 ■500 •350 ■300 <50 EE T Fio <-> <5-12> S <5-2> iSo. F TOREFER EE T Fie. 52 » <5-12> » <5-2> . iSe F TOREFER DAILY RECORDS OF PIEZOMETERS No. J44 ; J41RECORDS J44DAILY PIEZOMETERS OF No. 3031 12 33 3<9 15 36 383739 20 21 etme 1987 September r e b m e c e D 1987 u August 1987 September 1987 9 30 11 32 13 14 15 16 17 IB 39 2D 21 22 23 24 25 25 27 28 29 30 31 J27 STANDPJPE -50 JB74 VJB. VJRE -50 ATM. PRESSURE -100 -300 -350 -150 -200 -200 w -250 u» -250 -300 -300 x -350 x -350 -500 900 -500 November 1987 December 1987 - J27 STANDPIPE -50 REFER TO Fig©. <5-2) , <5-10) , <5-13) - JB7< VJB. VJRE -50 - ATM. PRESSURE & < 5 - 1 4 ) -300 -200 -200 y -250 y -250 -300 -300 x -350 x -350 -400 -400 -450 -500 “ — '— '— '— ■— 1— 1— '— '— 1— >— 1— 1— *— 1— 1— i— i— i— i— i— i— I I t i g 3 f) -500 DAILY RECORDS OF P3E20KETERS No. J27 ; JB74 DEPTHS - 5 ; 5 n DAILY RECORDS OF PIEZOMETERS No. J27 Fig. (6-15) DAILY RECORDS QF PIEZOMETERS No. J27 J874 CL (cm) (cm) P (c m ) DEPTH TO PIEZOMETRIC LEVEL P DEPTH TO PIEZOMETRIC LEVEL DEPTH TO PIEZOMETRIC LEVEL -100 -50 -500 .-250 -350 -100 -150 -200 -300 -350 -500 -300 -250 -200 -300 -350 0 0 4 - -500 -50 •50 0 DAILY RECORDS OF PIEZOMETERS Ho. P505 ; JB73 DEPTHS = 5 DEPTHS : 5a Ho. P505; OF JB73 PIEZOMETERS RECORDS DAILY coe 18 Nvme 1987 November 1987 October uy 97 uut 1987 August 1987 July J673 VJB. V1RE V1RE VJB.J673 STANDPIPE P50S T. PRESSURE ATK. P505 STANDPIPE STANDPIPE P505 T. PRESSURE ATK. VJRE Y1B.JB73 P505 STANDPIPE STANDPIPE P505 J673 J673 T. PRESSURE ATM. Yia. Yia. V1RE V1RE 10 i 1100 500 30 33 1300 30 a 3300 3000 900 x t o -250 L K 3 n 70 m -o m m tn in cr n n cr CL -350 x uj -350 x j -250 uj 30 - -350 x -250 -150 -200 -300 -50 -500 -300 -350 -200 -300 -50 -500900 -500 -300 -450 -200 -350 -300 -50 Fig. (6-16> DAILYRECORDS OF 2) j 19) 9 —1 5 ( j ) —2 5 ( . ® g i F O T R E F E R DAILY RECORDS OF PIEZOMETERS No. P505 RAINFALL (mm/doy) RAINFALL (mm/day) RAINFALL (mm/day) 25 20 15 10 0 3 0 3 20 5 2 20 5 2 5 1 10 35 5 0 2 3 4 9 10 3 32 13 1 2 30 31 0 3 9 2 8 2 7 2 6 2 5 2 4 2 3 2 2 2 21 0 2 9 1 8 3 7 1 5 1 5 1 14 3 1 2 3 31 0 1 9 8 7 6 5 4 3 2 1 2 3 4 7 8 9 10 3 1 4 15 35 17 18 39 20 2 22 23 24 25 26 27 28 29 0 3 9 2 8 2 7 2 6 2 5 2 4 2 3 2 2 2 23 0 2 9 3 8 1 7 1 5 3 5 1 14 3 1 12 33 0 1 9 8 7 6 5 4 3 2 1 31 0 2 9 2 8 2 7 2 6 2 5 2 4 2 3 2 2 2 23 0 2 39 8 3 7 3 36 5 3 14 3 3 32 31 30 9 8 7 5 5 4 3 2 1 i i l __ I I I I I I I I I _! [_ l i i I _i - 32.6 r e b m e v o N t s u g u A y a K nu DURATION RAINFALL 20 ^ 4 1 ^ 6 1 18 12 £ g 0 3 2 6 0 “ a 8 4 6 < O < £ x o u u —I 5 2 20 5 1 10 20 0 3 5 2 0 3 5 2 20 10 35 5 10 15 0 2 3 4 0 1 12 3 3 35 15 3 1 2 2 21 0 2 9 1 8 1 37 5 1 5 3 34 33 2 1 13 30 9 8 7 6 5 4 3 2 3 2 3 4 1 2 3 4 7 38 19 20 2 22 2 21 0 2 9 1 8 3 37 6 3 5 3 34 13 32 11 0 3 9 8 7 6 5 4 3 2 3 2 3 4 9 1 3 32 1 3 35 1 7 B 192021 22 2 1 2 0 2 9 1 IB 37 16 5 3 34 13 2 3 33 10 9 8 7 5 5 4 3 2 1 Fig. (6-17) DAILYRECORDS OF u h 11 r. r e b . r o t p e S r e b m e c e D a n u J n h d j U J45 STANDPIPE DEPTH ■ J879 VIB. WIRE DEPTH -50 ■50 J875 VIB. WIRE DEPTH -100 100 ~ ~ i i~ - i i-“i -150 150 UJ -200 •200 ,— ' -250 -300 x -350 •350 _ _ — i— 1_ UJ a -400 -400 -450 450 -500 -500 November 1987 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 REFER TO Figs. C5-2> s C5-17> J45 STANDPIPE DEPTH ■ J879 VIB. WIRE DEPTH -50 & <5-18> -5 0 J875 VIB. WIRE DEPTH -100 -100 -150 -150 UJ > UJ -200 -200 a: -250 5 -250 a M UJ £ -300 x -350 i— i -350 UJ a -400 -450 -5 0 0 Fig. (6-185 DAILY RECORDS OF PIEZOMETERS CORRECTED AGAINST ATMOSPHERIC PRESSURE DEPTH TO PIEZOMETRIC LEVEL Ccm) £ DEPTH TO PIEZOMETRIC LEVEL (cm) GL -450 -500 -400 - -350 -300 -200 -150 -100 -500 -450 -400 -350 20-200 -150 -200 -150 -250 -50 -50 Fig.CORRECTED PIEZOMETERS <6-19)OF RECORDS DAILY AGAINST ATMOSPHERIC PRESSURE ATMOSPHERIC AGAINST REFER TO Figs. <5-2> <5-10> j —i i— _J “ r t C5-12) <5-13> & _r-1 oebr 1987 November 87VB WR DEPTH J877VIB. WIRE ■ DEPTH J20 PNEUMATIC— ■ DEPTH J13 STANDPIPE 87VB WR DEPTH J877VIB. WIRE ■ DEPTH J20— PNEUMATIC DEPTH■ JJ3 STANDPIPE -i 1 — . -500 -450 -400 -350 -100 -300 -250 -200 -150 -500 -450 -400 -350 -300 -250 -50 -50 O M CL til N o z LU i- z a i—i (j DEPTH TO PIEZOMETRIC LEVEL . Fig. (6-21) RECORDS OF PIEZOMETERS BEFORE DRAINAGE BEFORE . PIEZOMETERS Fig. OF (6-21) RECORDS Fig.(6-20) RECORDS. 8.BEFOREPIEZOMETERSOF DRAINAGE AFTER A A K101 | £ A 1 A" A A A A A A A I® O a A %>Q A A A A @ A A ^ A 04“ 2. 3 . 2 “ 4 i0 K ^ O A <=> N O Kl 9 5 1966 1965 ® G i ® A O® fi O R E C O R D S O F P I E Z O M E T E R S N o .10 3 ;10DE 4 P T H S = 4 . 4 s7 . 7 m A A 02 EOD O PEOEES o 11 12 ETS 1 m 18 ; 2 . 4 = DEPTHS 102 ; 101 No. PIEZOMETERS OF RECORDS N O O « A = = a O 8 x 1 0 " “ CM. “ " 0 1 x 8 5 . 1 x 95 1966 1965 10"' ^a" " a "^ ^ a A A a x Q§> D 05C 10"5 O O o l & G G A S ft G CM/SEC & ^ 'SEC P CP0 3 0 1 * /SEC M A ^ 4 0 1 STARTED INSTALLATI c o 0 o 0 0 o D a o © e D J 2 0 1 * 01 10 ® N F DRAI OFDN EE TOREFER SAME VERT RECORDS C v ^AA AA & AA AA & AA ^AA v EE T Fi <5- . ig F TOREFER DATE DRAINS PIEZOMETERS WERE C ® ° o O 0 o o o ° P c H SAME VERTIC/ THE PIEZCRECORDS OF F DATE A * , PIEZOMETEF CL PROFIL ICAL i <5-23 . rig O 3 'IS J M 0 O ° ° A a a 23 POIE N ZOf IN PROFILE L METERS INSTALLE! A & A A A * o A o A c A kAA o ° ° o° © )0o S INSTALLERS I ZN 3 ZONE IN I OT UIG INE DURINGLOST |— INSTALL|— A A © COMPLE' A ^ O F AA A A A A A o O VIRTUALLYD TO O DR OF ATION 'ED A O O E 4 IE IN VIRTUALLY1 O O A O A 0 ALTO OF TALLATION M O O AINS M N THE IN A A O J © 1965 1966 N D JF M A M J J A S 0 I 0&D i 1 FIR 0 i^© & ---- INSTALLATI 3N OF D *AINS CC MPLETED a “ a a A A A. 1 0 6 aqa0 A 0 LJ 0 O 0 O A O A A° 0 O ° /&A A&ru, a a® lOtPtPl ,o O A A — INST/ LLATIOI1 OF DRAINS STAR! ED K l05= 2 . 3 x 1 0 "5c m / se C 0 0 K106P 1. 4 x 1 0 ~8c m / si :c • "0 0 K l07“ 1. 7 x 1 0 "S CM/S EC 0 ■ 0 ' 0 tI 0 ■a nfln 0 Qn 0 0 - gj0 -0 EE 0 CTy i-P 0 TJ3 0 0 0 g 00 □ j 0 0 0 0 0 0 0 Q □ 0 0 0 RECORD' ; OF PIE ZOMETEFS INSTA -LED VII?TUALLY IN THE SAME VEFtTICAL PROFILE IN ZONE 5 REFER 10 Ficj. <5-2> RECORDS OF PIEZOMETERS No. 105 ; 106 ; 107 DEPTHS = 3. 2 ; 6 ; 13.5 m (6-22) RECORDS OF PIEZOMETERS BEFORE 8, AFTER DRAINAGE DATE 1965 1966 N D J F M PIEZOMETER WA S LOST DURING INS TALLATION OF----- DRAINS o REFER TO F iq . <5-2> o G G O G G G 0 G O G §b geP o P > CP ©GO < G 0 © G O Q ©O© o o ° GO G O © G (> ©o K108*3 8x10 3c m /s e c RECORDS OF PIEZOMETER No. 108 DEPTH = 2. 8 m (6-23) RECORDS OF PIEZOMETER BEFORE DRAINAGE DATE 1965 1966 N U Q J F M M J K 10[3= 2.7x10 A - ’CM/SEC o - INSTALLATI ON OF DRAI vis STARTED ' t A K m 1“ 1.3x10 ’CM/SEC rA <9 1 1 0 $ <3 A A A £ ^*1 A A A A aAAA& ^&AS&£aAaA£^ a aaAA&a^ k A A A A » ^ A**' A A -- INSTAI.LATION OF L RAINS COMPL ETED RECORDS OF PIEZOMETERS INSTALLED VIRTUALLY :NTHE SAME \ERTICAL PRO -ILE REFER TO F ig. <5-2> RECORDS OF PIEZOMETERS No. 109 ; 110 DEPTHS = 1.5 ; 14 m Fig. (6-24) RECORDS OF PIEZOMETERS BEFORE 8, AFTER DRAINAGE DATE 1965 1966 JF M AM J J ° K = 2.2x1 f 8CM/SEC 1 •> 2 o 3 LU > LU 4 U ►—4 QL 5 H LU 3: o 6 M Q> RECORDS OF PIEZOMETER IU INSTALLED IN ZONE 3 •—i o REFER TO F La. C5-2> Q. 7 O o 8 o G © 9 % u © TION OF DRAI NS COMPLETED O — INSTALL/ OGK-\ 10 W G^Q8GI0£O~ ri0 11 r ? ° °fO'* w © O © © G© © O 12 RECORDS OF PIEZOMETER No. Ill DEPTH =18 ™ Fig. (6-25) RECORDS OF PIEZOMETER BEFORE 8, AFTER DRAINAGE DEPTH TO PIEZOMETRIC LEVEL < m > DEPTH TO PIEZOMETRIC LEVEL Fig.(6-26) PIEZOMETEROFBEFORE RECORDS i. (-7 RCRS F IZMTR EOE ATR DRAINAGE AFTER & BEFORE PIEZOMETER Fig. OF (6-27) RECORDS 5 3 4 2 1 0 1965 1965 E> 9. D — » 1:2x10 » K RECORI REFER F J )S O Fi< TO EOD O PEOEE N. 1 DPH 1. m 13.3 » DEPTH 112 No. PIEZOMETER OF RECORDS F 1 OF EOD O PEOEE N. 0 DPH 47 m 4.7 = DEPTH 201 No. PIEZOMETER OF RECORDS l— 7 - . <5-2) j. •IEZOME "OOqjg NTLAIN OF INSTALLATION CM/SEC M 1966 E IETLE N ZON IN TALLED INE fER ©O© ° IN — A o STALLA1 o M O RIS COMPLETEDRAINS DATE O DATE O OF ION <*>©< J 1966 5 E © DRAINS © J o O ©0, 0 © 0 COMPLE EOD ATR PIEZOMETERRECORDS A-TER REINSTALLED & A ATR DRAINAGE AFTER TED o © ° S 1967 o ° N 0 WAS D O >—iLU N CL o 2E LU i— w J _ a: J ( UJ > LU DEPTH TO PIEZOMETRIC LEVEL •Fig. (6-28) DRAINAGE &AFTERBEFOREPIEZOMETERS OF RECORDS Fig. (6-29) RECORDS OF PIEZOMETER BEFORE .& BEFORE DRAINAGE PIEZOMETER OF AFTER Fig. (6-29) RECORDS 10 9 7 3 6 5 8 4 2 1 1 0 1965 1965 EOD O PEOEES o 22 28 23 ETS 6 9 12.4 ; 9 ; 6 = DEPTHS 203 ; 208 ; 202 No. PIEZOMETERS OF RECORDS 5x10 INST STAR DRY © 0 '3 203 202 .-6 TED klla NTLAIN pF INSTALLATION RECORDS OF PIEZOMETER No. 204 DEPTH = 3.6 m 3.6 = DEPTH 204 No. PIEZOMETER OF RECORDS CM/SEC I 0 on i 1966 O, O f s n i a r d NTLAI RI3 CONPLETE DRAIN3 INSTALLATIO VERT RECO EE T Fi ( 2) -2 (5 . ig F TOREFER □ □ RIS•1PLETE DRAINS 1966 ‘ °?°©003000 DATE 3ROFI DATE K202 K203 1 K203 ZOMETlI E IN LE IDRY| a BEFOR^ RECORDS OF REFER 3x10 R 11 ERS 3x10 ooo ZONE a q TO .-8 -8 SAlE VRULY N THF IN VIRTUALLYNSTAllLED PIE PiS oo° CM/SEC CM/SEC PI 1967 af EZOME' C5r2) | er RIN: ER INSTALLA1IOf ISTAlLEE 1967 1968 dr ZONE DRY ! a IfS o© DEPTH TO PIEZOMETRIC LEVEL < m ) DEPTH TO PIEZOMETRIC LEVEL Fig. (6—30)8,BEFORE PIEZOMETEROF RECORDS DRAINAGE AFTER i . (-1 RCRS F IZMTR EOE ATR DRAINAGE AFTER & BEFORE Fig. PIEZOMETER OF (6-31) RECORDS 0 • &■ 1965 ° o ° O - K N 1965 EOD O PIEZOMETERRECORDS OF EE T Fik <5-2> k i F TOREFER § <*<•> 5 <§> L I 1 L i 2 . 2 x D ® ( 6 “ 0 1 QD o° INSTALL 1966 : m EOD O PEOEE N. 0 DPH 3. m = 1 DEPTH 205 No. PIEZOMETER OF RECORDS EOD O PEOEE N. 0 DEPTH 206 No. PIEZOMETER OF RECORDS / c e s O A M F J qe P CP1 NTLE N ZONE IN INSTALLED > o® COMPLETED DATE DATE o ---- o 1967 INST BETWEEN o 1966 *iLLATION © O M . O AJCN TRENCHES ADJACENT 0 o F DRAINS IF ©o J < O O > COMPLETE 3 1968 o J o o 30® ° © A . Fig. (6—32) 8,BEFOREPIEZOMETER OF RECORDS DRAINAGE AFTER DEPTH TO PIEZOMETRIC LEVEL 95 96 97 1968 1967 1966 1965 > i 6 § s< i 9 i i i ! i t i i i i t i i ! i 1 I i D Sp- M | > L * f I REFI REC = K | 1 i 1 1 1 1 \ i I ! I j ! 1 1 1 i I 1 3RC : , U M o r 5x1 q 1 i 1 i TC 5 O G of 2© •IN.ST O 0“iSC FfS- 0 ! | t j p j EOD O PEOEE N. 0 DPH 5. m = 7 DEPTH 207 No. PIEZOMETER OF RECORDS : SE /S A \L/^T I D EZOM <5r2: 1 ! 1 ! ! I i 1 i | i i ! R E H f DN D s i ! i ! ! 1 ! ! ! I | IO FD OF ©I H IlNST; A L ■ i o ! 1 1 ! ! PA ( > : ,U ns l I i i > ° 0 COM3LE: DIF>JED i M DATE 1 1 0 i l I i 1 ON „ I | J ! D I > EJETWE 1> S : n TW0 0 CNr • ACEN r M G O O 00 RE 0 ■J NC J G ( >o AEi 3— > O ( 0 D s > O DEPTH TO PIEZOMETRIC LEVEL C m ) DEPTH TO PIEZOMETRIC LEVEL i. (-3 RCRS FPEOEE ATR DRAINAGE AFTER Fig.PIEZOMETER OF (6-33) RECORDS 4 3 2 1 0 1979 4 99 90 1971 1970 1969 II ,M 6 8 1980 e J c A rp a a U o or 1981 EOD O PEOEE N. 0 DPH 4. m 8 = DEPTH 401 No. PIEZOMETER OF RECORDS EOD FPEOEE o 41 ET 4. m = 8 DEPTH 401 No. PIEZOMETER OF RECORDS e : r 1982 ?T 1972 E D 931984- 1983 DATE DATE 1973 1974 10^00% 1985 30 © 95 1976 1975 1986 ©O ©O GDI 1987 0©©0 © © P0 P o DEPTH TO PIEZOMETRIC LEVEL C m ) DEPTH TO PIEZOMETRIC LEVEL Fig. 3 4 2 1 0 A 1979 (6-34) 99 90 91 92 93 94 95 1976 1975 1974 1973 1972 1971 1970 1969 !©l!1 0,G 0 EOD O PEOEE ATR DRAINAGE AFTER PIEZOMETER OF RECORDS RECORDS 0,- ESR r 0 7 REISER ;oo 1980 lO l EOD O PEOEE N. 0 DPH 7 . 5 = DEPTH 402 No. PIEZOMETER OF RECORDS 1981 0 0 1979 (6-36) 99 90 91 92 93 94 95 1976 1975 1974 1973 1972 1971 1970 1969 !i!j ' llll ! ii Mil! i I i! ! 1 1 III ! ! i m i i ii O O "(T’Oo i| in Ill L U i ; i o. EOD O PEOEE ATR DRAINAGE AFTER PIEZOMETER OF RECORDS RECORDS RECORDS EE T REFER j i ; j ' Mi 1980 M ! I M ! I | I;N STALLED _ 1 i 6 9 6 1 ’ DATE DATE n 3ET t I I I ! i I 1983 WEEN i I jfwtii I ! I i ! i ADJA ! 1984 o o <•> '1° 95 96 1987 1986 1985 0 O kN ZtONB oo o%! % DATE 1969 1970 1971 EZOMETER WAS LOST CD RECORDS OF PIEZOMETER INSTALLED MI □WAY BETWEEN TWD ADJACENT TRENC HES REF^ER TO F ig. C5-2) RECORDS OF PIEZOMETER No. 405 DEPTH = 5. 4 m (6-37) RECORDS OF PIEZOMETER AFTER DRAINAGE UJ N o 2 lllh- ►H o _ > o; ui hi j : Fig. DEPTH TO PIEZOMETRIC LEVEL 5 4 3 2 1 0 H ! ! l l l i J 1979 i ! Ii miilm 1|yHw ! i h !!|jj!l ! <6-38) 1969 i! I Is !6 ; i ! II! 1 1 i 1 EE jTlREfER RECORDS ! I i ! '! i 'ill EOD O PEOEE ATR DRAINAGE AFTER PIEZOMETER OF RECORDS i'i! ! i ! ! j j I 1970 oo ! © i q l l i EOD O PEOEE N. 0 DPH 65 m 6.5 = DEPTH 406 No. PIEZOMETER OF RECORDS i ! II II EOD O PEOEE N. 0 DPH m 5 . 6 = DEPTH 406 No. PIEZOMETER OF RECORDS ii IPIEZOMEJ 19811980 i I I i I ! i I 1971 1982 II 1972 DATE 1983 DATE 1973 IOW I* 1984 EEN 1974 1985 DJACE sIT 10! 1975 TRENCHES fc 1986 oo AS W oU 1976 l l i 1987 O! J m DATE 1969 1970 1971 A M J J A S 0 N D RECORD 3 OFf PI tzok ETEfc INSTAI-LEC BETWEEN T it 0 / DJA: eni triincf ES [N 2 ONE r f i g . > REF ER TO <5- O Oo RECORDS OF PIEZOMETER No. 407 DEPTH = 5. 3 m (6-39) RECORDS OF PIEZOMETER AFTER DRAINAGE DATE 1969 MA M J j A S 5 - i PIEZOME TER WAS- LOST o RECOIJDS OF PIEZON ETER INSTALLI :d in zone 6 F?FFF? TO Fig. (5 -J •> RECORDS OF PIEZOMETER No. 408 DEPTH - 6. 0 m (6-40) RECORDS OF PIEZOMETER AFTER DRAINAGE DEPTH TO PIEZOMETRIC LEVEL < m ) DEPTH TO PIEZOMETRIC LEVEL Fig.(6—41) DRAINAGEAFTERPIEZOMETEROF RECORDS Fig. (6-42) 4 M J 0N D N 0 S A J J M A 99 1970 1969 6 8 EOD O PEOEE ATR DRAINAGE AFTER PIEZOMETER OF RECORDS 1969 BLQCKIi PIEZOM FER COR RE RE,C F 0 RWA ER II OF EOD O PEOEE N. 0 DPH 6 m 6 = DEPTH 409 No. PIEZOMETER OF RECORDS RECORDS OF PIEZOMETER No. 410 DEPTH = 6 m 6 = DEPTH 410 No. PIEZOMETER OF RECORDS i9 PJEZjOMSTE <5 o ' f 2> EZOM E C5-2> 1971 INSR TEF: -lE LE MIDWAL_EC 1970 EF DATE DATE A ll Ej ll A 1972 BL AY OC KIEID BE TWEEN n IT TC TWD A 1973 i DJACENT A l £N(t 1971 TR: 94 1975 1974 nc O W T HES I -J Z0NE7 zone DEPTH TO PIEZOMETRIC LEVEL < m ) DEPTH TO PIEZOMETRIC LEVEL Fig. Fig. (6-43) (6-44) 99 90 1971 1970 1969 o o M RECORD REFER o° RECORDS OF PIEZOMETER AFTER DRAINAGEAFTERPIEZOMETEROF RECORDS EOD O PEOEE ATR DRAINAGE AFTER PIEZOMETER OF RECORDS RECORD R^ER o g. <5-25 . ig F ro PIEZOME' OF 3 TO RECORDS OF PIEZOMETER No. 411 DEPTH = 6 . 0 m 0 . 6 = DEPTH 411 No. PIEZOMETER OF RECORDS A RECORDS OF PIEZOMETER No. 412 DEPTH = 6.= DEPTH 2 m 412 No. PIEZOMETER OF RECORDS R INSTALLED ER OWE O ER M 92 1973 1972 ST BETWEEN ADJACENT TREI TWO DATE DATE 1969 ED J BETTI rrV O J 94 95 1976 1975 1974 D. A w dent CE I ZONE IN JCHES PIEZOMETER V A DM ES TE O © S LOST- AS 7 ID VAS \s> S O I. 3 A v DEPTH TO PIEZOMETRIC LEVEL E Fig. 5 3 4 2 1 <6-45) ■ " '■ " © o o A M 3 ^ 0 RECORDS OF PIEZOMETER AFTERDRAINAGE PIEZOMETER OF RECORDS g EOD F IZMTR o 44 ET = 3.m = 6 DEPTH 414 No. PIEZOMETER OF RECORD M O DATE 1969 J J o PIEZOMETER A \ A LOST 'AS S ----- O DEPTH TO PIEZOMETRIC LEVEL Cm) DEPTH TO PIEZOMETRIC LEVEL Fig. sr 3 (6-46) 99 90 1971 - 1970 1969 4 68 EOD O PEOEES o 45 46 47 ETS 3 ; .3 48 m 4.8 3. 3.; ; = 3 3 DEPTHS 417 ; 416 ; 415 No. PIEZOMETERS OF RECORDS EOD FPEOEESN. 1 46j 1 DPH . 33; 4.; 3.3 m ; 8 3.3 = DEPTHS 417 j 416 ; 415 No. PIEZOMETERS OF RECORDS K 4 l*c / TWO AD K RECORD REFER i s L6“ > S 1 PS 17*= M EOD O PEOEES FE DRAINAGE AFTER PIEZOMETERS OF RECORDS OFt PH 3 9 . 2 fERE HD f}lg. JACENT ‘ e . i 1 0 l x 0 ! x * 2 IMF C5-2) : ci 5 ' a ~6 ROVi 96 1987 1986 RENpHES zom o © A A J | etef I/SE Sl /S 1 : d u L ° 2 : II S c I - G !2 VII DATE a ITUA VED 2 2 DATE D _LY © o NOW 415 : i 4 IN c 1 S 11 1 FHE 3EJaai P 1 AP M U) fAS 3EPEt < 1976 1975 1974 IMPF s a □ % > ixl ?TIC OVEI A 2 2 2 e a J L F AL ) □ ‘ 2 2 a © II ROF □1 □ ^ LE S 3ETV EEN 5>e© m a T 0-0- L ^ A , © t i B J D DATE 1979 1980 1981 1982 1983 1984 1985 1986 1987 RECORDS EZ E" INS A*' El!' VEL” 3 HE USE III M l iT DRAINAGE ES Ml ll| It REFER TO ElS 00 >'*&> RECORDS OF PIEZOMETER No. 501 DEPTH = 5. 0 m (6-47) RECORDS OF PIEZOMETER AFTER DRINAGE DATE 1979 1980 1981 1982 1983 1984 1985 1986 1987 m ' m 01- ED El. HOUSE !f £r RECORDS OF PIEZOMETER No. 502 DEPTH = 5. 0 m (6-48) RECORDS OF PIEZOMETER AFTER DRAINAGE DEPTH TO PIEZOMETRIC LEVEL C m > DEPTH TO PIEZOMETRIC LEVEL Fig. Fig. 3 2 5 4 1 5 4 3 2 1 0 0 1979 (6-49) 1979 (6-50) R RECORDS OF PIEZOMETER AFTER DRAINAGEAFTERPIEZOMETER OF RECORDS 1980 EOD O PEOEE ATR DRINAGE AFTER PIEZOMETER OF RECORDS ecords 1980 i LR WftTES HOUSE n 11 t n r o EOD O PEOEE N. 0 DPH 4. m = 8 DEPTH 503 No. PIEZOMETER OF RECORDS EOD O PEOEE N. 0 DPH 5. m = 0 DEPTH 504 No. PIEZOMETER OF RECORDS DF 1981 r 1981 I WA e 1 Z om e m E t ] t : r 1982 E S b l l.L 1983 DATE 1983 DATE At A” 1984 19841982 . ’ l.E 1985 1985 1979 1980 1982 1983 19841981 1985 1986 1987 RECORDS OF PIEZOMETER No. 507 DEPTH = 5.0 m (6-53) RECORDS OF PIEZOMETER AFTER DRAINAGE DATE 1979 1980 1981 1982 1983 1984 1985 1986 1987 • RECORDS OF PIEZOMETER No. 508 DEPTH = 4.8 m (6-54) RECORDS OF PIEZOMETER AFTER DRAINAGE DEPTH TO PIEZOMETRIC LEVEL < m > DEPTH TO PIEZOMETRIC LEVEL Fig. Fig. 5 3 5 4 2 1 0 * 99 99 90 91 92 93 94 95 96 1987 1986 1985 1984 1983 1982 1981 1980 1979 (6-55) 1979 (6-56) REG RECORDS OF PIEZOMETER AFTER DRAINAGEAFTERPIEZOMETER OF RECORDS EOD O PEOEE ATR DRAINAGE AFTER PIEZOMETER OF RECORDS 1980 REM EFI !EN E5 EOD O PEOEE N. 1 DPH 5. m = 0 DEPTH 510 No. PIEZOMETER OF RECORDS DF EOD O PEOEE N. 0 DPH 4. m = 8 DEPTH 509 No. PIEZOMETER OF RECORDS 1981 WA wAr ■f' ■$- : " E HU 1982 ]MS ISlfe « ? # lilr Li. : d DATE DATE &S> m 19841983 IE OF 3w 1985 vi : Ml e , m 1986 WU m t 1987 DEPTH TO PIEZOMETRIC LEVEL < m > DEPTH TO PIEZOMETRIC LEVEL Fig. Fig. 5 3 4 2 1 0 1979 (6-57) (6-58) 1979 TRENCH r ■6^1- :c RECORDS OF PIEZOMETER AFTER DRAINAGEAFTERPIEZOMETEROF RECORDS 1980 Eft E EOD O PEOEE ATR DRAINAGE AFTER PIEZOMETER OF RECORDS > % or : * OF EOD O PEOEE N. 1 DPH 50 n, 5.0 = DEPTH 511 No. PIEZOMETER OF RECORDS RECORDS OF PIEZOMETER No. 512 DEPTH = 5.0 m 5.0 = DEPTH 512 No. PIEZOMETER OF RECORDS 1981 1980 I e Z 01* b - ' s HO ER U3E 1982 I f-F *J£T IE2 M OME LED 1981 TER 1983 A DATE DATE t 99 IN AS /E LOS 1984 Lt:v El. •II 1982 M D 1985 E"W ZEN 1986 AF> WO ADJ 1983 ACE. 1987 •H $b .Fig. (6—59) AFTERDRAINAGEPIEZOMETERS OF RECORDS DEPTH TO PIEZOMETRIC LEVEL A K R E F E R C E P S E R t R O C E R ' f E R T F O H T P E D F o 50 ; 5 5 m 5 ; 5 = S H T P E D 3 7 8 J ; 05 5 No. S R E T E M O Z E I P OF S D R O C E R ° A o TO T I P F O S A ^ p O T M . e g i F O W T i C A J D A 3 I E T E M O Z I E V N I H C 3 2 - 5 < (A &© - 5. 0 0 . 5 - . ! E V E L T R / T S N I S * 1 f E R T T N : <5-19 1 - 5 < & » A A » A o o © A A A A * > T A D E L L N I S E H C © O J DATE 1987 I I V k A A 5k . E R I W i. b. g. 1 . g . b m J873 © o 2 E N O Z T R E V N I 0 S 505 / E P I P D N A f i © © I L E V E L A A O C O 5 i S E H T si S A ME POSl S O P E \M A © O A < © © A A W S N O I T 3 3 H T I A 0 ° A DEPTH TO PIEZOMETRIC LEVEL < m > DEPTH TO PIEZOMETRIC LEVEL Fig. 5 3 2 4 1 0 (6-60) CRS F o J1 ; 2 ; T1 5 ; 4 m 4. 2 ; 8 . 4 ; 5. 1 = S H T P E D JT21 ; J27 ; 13 J No. S R E T E M O Z E I P OF ECORDS R REI TRI REi F O S D R O C E R i F C T F E F E R -ER inci : ori M EOD O PEOEES FE DRAINAGE AFTER PIEZOMETERS OF RECORDS EES 1 IS TO F f 3F Fi; t i 1986 : 96 97 1988 1987 1986 PIE ; > f l e e i j ; ZOM s - s : : p - 5 < F o J2 DEPTH = m 5 = H T P E D 29 J No. R E T E M O Z E I P OF S D R O C E R O ome % ---- IL c E T E » ooG > TEF &A ZC t ms NE ] S 2 J 1 J S m <5 5-1 T S M bo 7 AA © > G 3 E I - NS1 - 4 As Z ! TWO W T '! E E V T E 3 ZD L L \ TS,—1 > L L A A ,AA o' 0 DBET ED D O A 3 JT G* A ..A DATE WEEN 7 DATE mm &*i 21 » O * * DE J M WO =TH (% o > K A DAET •ECES N N 4 2 ••REHCHE:SONE\DJACENT IN J D A « • A % OF AS 1987 A & ACENT D A ...,A ° 0 ° TR O INC 4 M V A a ^ II E TRE • • • G 0 AA o o IVEI >JCH • • ©o A S a T I IT S I • • ©A A 0 A O \ND .EVE & . 9 GO AA : > O l IN □ O - 4. 4 ■ • • ONE G o : r 4 1988 « • O A G A Tl r b. i • » L-A- M o EE 5 9-3 % A4* l^fr N» *>0 DEPTH TO PIEZOMETRIC LEVEL < m > DEPTH TO PIEZOMETRIC LEVEL Fig. 5 3 4 2 1 0 (6-61) F o J7 j JT52 DEPTHS H T P E D 2 5 T J ; 4 7 8 J j J877 No. S R E T E M O Z E I P OF S D R O C E R REFIIR TO TRENCHES REF REC 3RDS REC F o J4 J7 DEPH = ; m 8 ; 8 = PTHS E D J871 ; J44 No. S R E T E M O Z E I P OF S D R O C E R R TCER ORDS EOD O PEOEESATR DRAINAGE AFTER PIEZOMETERS OF RECORDS M F P OF PIEZOMETERS f o F i 9< € ' . 5 — 1 2 ) lS F f l : e s t i . - 5 < ». IEZOK A u A O AAA O O IN ^ ) » 2) ETER5 7 8 J 4 4 J 1987 s J 1987 ZON a a i A ^ i s in ; . 4 DPH OF DEPTH 4. E <5-1:> 0© 0© < 5 N'LE3 BET 3 INS'ALLE - 1 p i ; rALLE y O - £ A BET□ o ° A A DATE DATE WEEN WEEN TWO O TRENCH IN A A © A a TWO -ee- JT52 * t f D A N 'RNHS : RENCHES*DJAC ' ENT □ £ N "RENCENT EE - ^ - LEVEL * •••*•* NTLE I TE TRENCH THE IN INSTALLED A O •> E AD I ANDHES . 4 m b. g. b. m 4 . M % 1988 1988 & OOG G O O 3 ZO N ------ •*«••• y ? o o | N O "HE OF ONEv| UJ ■'t J DEPTH TO PIEZOMETRIC LEVEL < m > DEPTH TO PIEZQMETRIC LEVEL Fig. Fig. - 3 RECORDS O PIEZOMETERS N. 0 J1 J4 ; 0 5 8 9 1 m 10 ; 9 ; 8 ; 5 = S H T P E D 40 J ; 42 J ; J41 ; 20 J No. S R E T E M O Z E I P OF S D R O C E R 8 7 6 5 3 4 2 1 0 (6-62) 0 (6-63) GO © REFE BETV Cfe)© REC( PTH PER COR F o J2 ; 4 ; m 2 . 4 ; 4. 2 = S H T P E D 2 5 T J ; JT21 No. S R E T E M O Z E I P OF S D R O C E R EOD F IZMTR ISALD N TRENCH INSTALLEDIN A PIEZOMETERS OF RECORDS RECORDS OF PIEZOMETERS AFTERDRAINAGE OFPIEZOMETERS RECORDS □S 30 1RDS "0 "0 TO R EEN OF TC M OF i i TR i F i i t 'WO A PI IF Fig© 1986 O PI ge NC EZ C E I H <5- (5-2 . 3JACE EZOME • • • • OM ©oo® 0 © 0 J40 J20 J41 J42 2 WERT ITERS ) • • • • q &AAA : ) ) T T NT ters o E1 1987 qe J t LEVEL J INSTAL 5-1 3> 5-1 « • » © o d 5- 2 -1 (5 ?ENCH ©O NTALLECINST E > AL IS LED • © © A A □ □ & 4. C • O 5- L i 5L ( IN 5-113 VRrUALL VIR 1 ONG • O O A A □ $ $ 5- ) 0 -2 (5 $ 1$) S DATE DATE . b • A □ TRENCH R F E -A NZONE IN A A- 3ROFI _E • © A A □ . . o 1. • O El 0» • » 0 0 B A A 0 IN Y O © 3 B T21 JT . A . NZONE IN 1987 A D H SAME THE a t D A A © © • • o l Q tl O VERT • • 4 A A IJ □ O ■ 0 • ' A A' Q U J D [CAL M * © n 1988 • RF E M LEPROF] og i 8 JTI21 O 1988 2 5 T J [DWAY G O O ° £ A J A A b a DATE 1986 1987 1988 0 1 2 r e :ords of piezometers jnstalled IN TWC AC JAC ENT TFENCHES IN ZCHE 7 DEPTHS OF TRENCHES IKVEFTS FOF PI EZOMET ER£ J715 , JT19 = 8> 3. 9 5 m b. i. 1 EC'IVELY AN! I FOR FIEZOME TEF J714 REFER TO F ig e . :5-C) t CS-l£-a>& :5-i 3 JT1 3 3 QXDG^> Qaoo 00 OO GjOO © A k A^A Aa A & A ^ SAa A & A / A A A A 4 □ EBB QB BEEEI B I p S -q q e ; B B J T 14 5 RECORDS OF PIEZOMETERS No. JT14 ; JT15 ; JT19 DEPTHS = 4. 8 ; 4.1 ; 4.0 m Fig. (6-64) RECORDS OF PIEZOMETERS INSTALLED IN THE TRENCHES DATE 1986 1987 1988 0 OF PIE ZON ETERS INSTALLED AT INVERT L EVE ETW' :en TWO AOJAOEN" TFENCHEE 7 . THE TRENCHES iflVETT -EVELS ARE 4 IN TH3S 2 ONE. iFEFif Ttll F ig a . <5-2) » 2 3 J17 GX> O c J1 A i ’cefec °° O O — 2-iLaffiftA. AA-A ■ 4 iAA&A-^-A-A B E Luulju j B B 3BE±)lfe[ El B El B B r m 3 0 B Q B E J16 5 RECORDS OF PIEZOMETERS No. J16 ; J17 ; J18 DEPTHS = 4. 15 : 4.0 Fig. (6-65) RECORDS OF PIEZOMETERS AFTER DRAINAGE DEPTH TO PIEZOMETRIC LEVEL < m > DEPTH TO PIEZOMETRIC LEVEL Fig. 2 5 3 4 1 0 ------<6-66) ECORDS O PIEZOMETERS N. 2 ; 2 ; S H T P E D 0 7 8 J ; J22 ; J24 No. S R E T E M O Z E I P OF S D R O C RE A —a— L A BET! $ REFI REC REF OPE REC J n A A * EOD O PEOEES FE DRAINAGE AFTER PIEZOMETERS OF RECORDS AIG ATRATING R O ig F TOER R S F A OFORDS R TO IR VEENT DRDS -- © M \ ^ A ---- of L,—2k*2 E^tlQE VO AC v Fiq« F o J872 DEPTH = .0 m 4. 0 = H T P E D 2 7 8 J No. R E T E M O Z E I P OF S D R O C E R p : . <5- .. JACEt <5-; . EZOM NEAR VIBR ®©o© 0 B B 0 J2^ J2; 8 J ) <5-13> & 2) ATMOSPHERICLY PRESSURE IHSTAL PIEZOME' WIR ERATINGI 2> VT TR ITERS 70 I 3EJ J DOO o© 1987 aa 97 1988 1987 j e h J » "” ENCHE C O D !i C5-1J B B INS* A / e AL( S > ALLEi A O © B 0 A t c A 5-16) DNG P VIR 3 o o □ A s DATE DATE B ROFIL TUALL o o o O A A □ n 0 J B EJ C 6-1 E IN Y l C l A LED B 3 IN 3 ^A' * O ° DO THE D □ SAME ZONE ZN 7 ZONEN O B B K22 K24 O 7 =■ 1. . 2 = VERT o o A"’A H B ICAL 25 42 8m 4.2; 8 =2.5; 2 1 x 8 2 1 x 1 A M o B 1988 «k 3 ocooo o o ooc ) PROF "8 _SCM cm . L M [LE SEC E /S 'SEC A B IDWA'i O © A A B 0 J O G B 0 DATE 1987 1988 M J s D M J RECORDS OF P [EZOMETERJ ; in s TALLE 3 VIF TUAL .Y IN THE SAME VER1 ICAL PROF ILE N IDWA f BET tfEEN TWO tDJACENT 1REND IES A .ONG PROF LE 6 -6 IN ZONE : 6 % RE rER T0 Fi< i». <5-23 * <5- L55 t < 5 -‘ 7> t <5-1 8> <» O o T 25 o ■ J 43 o 0 n ui A* A A £ > O . 0 LU % >G°o B O A k0 4 A ■ A D 2 •> CJ > 0 o A o M O OA 0 A ° e ° cc (U A- H o I— p o _ o ^ A lxl . s?* -*>© “S o u O J 0 0 0 B 0 " “ 0 £ 2: o 0 B 1 H0 i 0 0 » o \ A 2 2 0 © 4 N A * o 0 A LU 3 i A ■ 0 o i—i ■ 0 0 B CL O » • • • • • K 25 = 3 . 1 x 1 2 "8 c M/SE c • • • • • •• • # K 4c x l 0 * 9 CN /SEC •• * • • • • = E •• < J875 REC O R D S OF P I E Z O M E T E R S No. J25 ; J879;J43 J875 DEPTHS = 2. 5 ; 4.2; 6 ; 8 DATE 1987 1988 M J s DM j —O— 0 « ' ° J 2 5 0 J4 3 a . 1 o § A K o 2 ffA 4 L k A ° k £ • © © AA A>> C 5 ° © e o ° * 2 °o © 0 f l H i " 0 B ~ 0 ~ B * 4 3 |A A b - B f n I 0 0 0 0 0 o ° * Al a J87< 3 ^ a a a a k H i * * A 4 ; B B 3 • • • • • LU • • •• » > • •• • • • LU «*' 4 ( -• DC J87E IUI— 5 O2 N LU 6 RECORDS OF P IEZ0N ETER! 5 INS TALLE D VIE tTUAL .Y IN THE SAME VERi ICAL PROF ILE FIIDWA < BE! WEEN TWO *DJAC ENT ■rRENCHES / LONG PROF ELE E- 6 I t ZON E 6 7 REF E R T ( F i g s . < 5 - 2 ) t < 5 - 1 5 ) t < 5 - 1 ' 0 t < 5 - I E > 8 •’ 9 10 a a a a a J87 8 . a a il □ a a □ Dg 0 E a a s n o a a. a a 0 t in 0 3 g a e 11 RECORDS OF PIEZO. No. J25; J879; J43; J875; J878 DEPTHS=2. 5 ; 4.2 ; 6 ; 8 ; 11.8 Fig. (6-67) RECORDS OF PIEZOMETERS AFTER DRAINAGE DATE 1987 1988 M J S D M J REt ORDS OF F IEZOF ETER ; in s TALLE D AT INVE RT LE VEL 3ETWEEN T 10 AD JACEF T TR :nche 3 AN! 1 IN 3NE CF THE: TRE 'JCHES ITSE LF I A ZONE 6 / 1 A REFER T ) F ig 9. (5- 2) t <5-1 O® o o O A OO •) o c J87S ! A A Aa A A (l 'A A A I A A A A A * A A A A J 4 6 JT 2 3 INSTALE ED I >1 THE TREE CH RECORDS OF PIEZOMETERS No. J879 ; J46 ; JT23 DEPTHS = 4. 2 : 4. 4 ; 4.3 m DATE 1987 1988 RECORDS OF PIEZO PETERS IN STALL ED V [RTUAi_LY ] THS SANE PROFILE ADJACENT TO A TRANCH I v| ZON ALONG THE S \ME CROSS SECT ION AS PI SZOME1TERS J25 t . 879 i J87 5 & . '878 <5 - 18>REFER "0 Fig®. <5- 18>REFER J46 J51 RECORDS OF PIEZOMETERS No. J46 ; J51 DEPTHS = 4. 4 ; 11.8 m (6-68) RECORDS OF PIEZOMETERS AFTER DRAINAGE DEPTH TO PIEZOMETRIC LEVEL < « > DEP™ T0 PIEZQMETRIC LEVEL Fig. 5 3 4 2 1 0 <6-69) K F o J6 J8 J4 DEPTHS = .5 4.5 ; 0 m 10 ; 5 . 4 ; 2. 5 = S H T P E D 47 J ; J28 ; J26 No. S R E T E M O Z E I P OF S D R O C E R REFi RED TRE 5x1 x 5 . 2 “ RFL 6-6 6 PROFILE EfR O iga F TOREFfiR RECCIRDS PI OF J2 JCHES : ]RDS EOD O PEOEES FE DRAINAGE AFTER PIEZOMETERS OF RECORDS r M to .J26 o 3F c 8 0 ALOF Figs p F o J45 D 4.2 m 2 . 4 = H T P DE 5 4 J No. R E T E M O Z E I P OF S D R O C E R : SE /S A EZOMi PRIG EZOMETERS © > : : JFILE ter 97 1988 1987 1987 o o J S ■ ------INST/ <5-If 5 1 - 5 < -6 6 ns e l l a st in o 6 [ to LLED O N Zt IN O O : : JNE 6 T I AT o O virtually S DATE DATE E1LEVI WER1 o OO O 0 in : ■> l THE D mi O DWAY BETWEEN 1 o ; ame o vertical O O ©o M 7 4 J 1988 i WO Al ■> i e fil o r p JJACE 'JT © o o . A A ai O . ong J • DATE 1986 1987 1988 OF p i e ;ZOMZTE!RS LOM ROF ILE 5 - 0 -INEI REt*- 1 °otoo a C3aEafiGsqo3ay>o aG^GOC oojoo 2 3 J2 L 4 5 6 RE COR DS OF ZOM RS INSFAL.ED VIRTUmLL' in THE same : 4. 8x1 ,-8 VERTI CAL PR LE 'EG 3RO-IL: 5 -5 BO REFER TO F i CO 5—8 > 7 8 RECORDS OF PIEZOMETERS No. J1 ; J2 ; J3 DEPTHS = 7.5 ; 5.5 .Fig. (6-70) RECORDS OF PIEZOMETERS IN UNDRAINED AREA OF SLOPE DATE 1986 1987 1988 0 J5 1 J7 r 2 3 J5 Q 0 4 F&r J6 4= c m / s e :c ,-E 5= 4x1 5 6= 6 7 RECORDS OF PIE! :OM?TEf?S JNSI,ALL EDI VIR TU/ LL'i AME VERTICAL PR OF I -E ALO'JG PROFILE !3—5 IN ZONE 3 REIFEF TC F i g o . <5- 5 -E ) 8 RECORDS OF PIEZOMETERS No. J4 ; J6 : J5 ; J7 DEPTHS = 9 ; 7 ; 5 : 3 m Fig. (6-71) RECORDS OF PIEZOMETERS IN UNDRAINED AREA OF SLOPE DEPTH TO PIEZOMETRIC LEVEL < m ) DEPTH TO PIEZOMETRIC LEVEL Fig. 5 3 4 2 1 0 <6-725 RE PR 9 3. 4 . 3 K9= KE RE TO REC F o Jl J0 J12 DEPTHS = ; ; m 5 . 2 ; 3 ; 4 = S H T P E D 2 1 J ; J10 ; Jll No. S R E T E M O Z E I P OF S D R O C E R REF OF I COR -ER THE ORD! R ‘ ER 4 . 8 M _E OS EOD O PEOEES NUDAND RA F SLOPE OF AREA INUNDRAINED PIEZOMETERS OF RECORDS TO OF ; 0 F ■0 WES F o J ; 9 8 6 m 6 ; 8 = S H T P E D J9 ; J8 No. S R E T E M O Z E I P OF S D R O C E R 1986 ig®. r ----- PIE of :ZOMETERS INS"ALLED £ ( i <5-; ZOMi D ---- 9 O k A 96 1987 1986 a <5-; § A PRC J o A A ET I r jfil 25) : 3 ter a A/ O1 1 5- 5 : > 0 E > 0 <5-85 IN 5 'a aa <5 C * GA 5 STAl -e> 3 9 J S oo ‘Ai : b E VIR LED 8 J VIRTUALL' o A 0 1 J1 o A DATE DATE TUAL 2 l l J 3° iA D o □ O , A LY E a oG B N H SM VERTICAL SAME THE IN 33 © 33 : n k i 1987 °© 3 i' Bi n E ‘ HE M A / ; o ) A i 1 ame < A nfT?r .J 9 i a > o a a □ VEF G> ^A 8 J TIC/ & © 3 8 3 J >L P .. 0 A © A RFL *iLC PROFILE G 1 *OFI 1 1 e ° A a J A a le -fixJL 2 1 A A □ □ 1 J 1 J S : AG 1988 0 M 10 3 © A □ A a NG A^ □ E iTER i S □ DATE 1986 1987 1988 □ , - a 4 > 10 CM/S ZC 1 ,-•7, J38 A 10 ICCH/S 2 3 4 5 6 REICOR!DS OF PIE ZOMITE INSTALLED VIRTUALL'' IN THE SAME! PFOFI LE AS PIE ZON-ETE RS J5 J7 AllONG PROF I _E 5-5 IN ZONE L TO PROVE THEIR RECORDS. REIFER TC Fi <5- Z> 5-8 > 7 8 RECORDS OF PIEZOMETERS No. J38 ; J39 DEPTHS = 5 ; 3 m DATE 1986 1987 1988 M J s D M J s DM 0 RE:cof DS OF PIE ZOMETE R I NST **LL ID IN [■HE m 3TE RN SI DE OF HIL L. LOST DJE TO : on 5TR JCT[ON o > *K. O 1 REFER TO F i 3- <5 -2 ) 2 to _j 3 >ui LU n 4 o 5 ci 6 ©e __ o 7 °0 o Q © O a . 8 LUCl -£ K £= 1. 2x: 0 :m /SEI 9 10 11 RECORDS OF PIEZOMETER No. J50 DEPTH = 15 m Fig. C6-73) RECORDS OF PIEZOMETER IN UNDRAINED AREA OF SLOPE DEPTH TO PIEZOMETRIC LEVEL < m ) DEPTH TO PIEZOMETRIC LEVEL Fig. 6 74) (6— K3: K3: 3 ‘ S3 M M EOD O IZMTR I FA UDAND AREA UNDRAINED IN FLAT PIEZOMETERS OF RECORDS F o J1 J33 DEPTHS = ; 1. m 5 ; 4 = S H T P E D 3 3 J ; J31 No. S R E T E M O Z E I P OF S D R O C E R ! F o J35 ;J36 DEPTHS = .5 4 m 4 1. ; 5 = S H T P E D ; 6 5 3 3 J J No. S R E T E M O Z E I P OF S D R O C E R 5. 5: . . T C10- :10 3 3 CM 96 1987 1986 1986 cm J J 'SE1 'SEI S S 3 J j A o A : | s / o o DATE 1 DATE < 3 3 J i & a & D D j : -m^ !6 M (3D © Vlf RE RE IN h L m L A -ER JTUA : A A A ©© ZON orc © A G & G £ i f A A i nflP ( A A 9 E LLY Of S TO A C O M A M i *O ■ ri3- PN • VER A REF REC INS IN q R 1 ER ORDS TIC/ TALL )® e t ^)A®j THE EUMA © 1 p £ % <5-2 PI L F 0 ED SAMI BAg 1987 > TIC OF J J JOFI ^IRT : S STA 9 PIE < • o l ve * 3-2> I -E JALL MDP1 RTIC ZOME * o Z( N E 1 PE II Y a 2 a L F AL TERJ 6 6 D S D s TH 1 INE >IEZ >ROF ; © A G> in 2 9 / s : DME1 A A A [LE STAL A ME ERS A 3 LED ----- DEPTH TO PIEZOMETRIC LEVEL C m ) DEPTH TO PIEZOMETRIC LEVEL Fig. 5 3 4 2 1 0 92 1983 1982 (6-75) d m j s d k j s d m j s d 1982 REOC F o J9 J48 ; 0 DEPTHS = .5 5 6 m 6 ; 5 ; 2. 5 = S H T P E D 601 ; 8 4 J ; J49 No. S R E T E M O Z E I P OF S D R O C E R _UF.Dg *ib ogET EOD O PEOEES N LT NRIE AREA UNDRAINED IN FLAT PIEZOMETERS OF RECORDS ro 1 : > 3 ijjj. < 1983 -5 jf4 CM F o 60 DEPTH » 2 m 12 » H T P E D 02 6 No. R E T E M O Z E I P OF S D R O C E R . ? t -E CM/SEC MM K=> 1984 SEC ZON U ne E ST 1984 AllU i UVI ERI ep ES DATE 1985 DATE 30 _i.'d 1985 QOcS?, IN [N KF S> SAME M J S D 1986 1986 3 /ER ICAL EE D PR M J S D 60 L 60 1987 97 1988 1987 <£> AS 34 4£ T 1985 1986 1987 1988 1 RECORDS OF PIEZO4ET IR INSFAL -ED VI RTUALLY ] N THE SAME VERTICAL pr o f : LE AS PIEZOMETERS 702 » 70 SERIES £00 & \ p ] n ::one :8 2 3 3. 3x1 4 5 “ ElQ esj P] EZCiMETRIt: le:vei . dropped DUE "0 EQUALIZATION TEST 6 RECORDS OF PIEZOMETER No. 701 DEPTH = 45 m DATE 1985 1986 1987 1988 1 -5 1x1 ,-E. 702 2. j xl CM. SE 2 3 4a . CAF WA! T a ' 03 4 __----- 702 5 RE CORDS OF PIEZOMETERS IN3TA .LED VIRTJAL LYIN VEFITICAL PROFILE AS PIEZOME' ERS 701 i SERIES 300 a REFEF: TCi F= g. S -2> 6 RECORDS OF PIEZOMETERS No. 702 ; 703 DEPTHS = 15 ; 30 m g. (6-76) RECORDS OF PIEZOMETERS IN FLAT UNDRAINED AREA LU Q. . Q •—i a O LU ^ N O 3E (■I h- L h- DEPTH TO PIEZOMETRIC LEVEL DC o l ) Fig. 11 10 9 8 7 5 6 2 0 1985 % 1 0 C6-77) D [ )0 3 V Q0 K K PIE RE( o e o F o K1 K2 KP DEPTHS = ; ; 9 ; 9 ; 9 = S H T P E D P3 K j KP2 ; KP1 No. S R E T E M O Z E I P OF S D R O C E R 48 E IZOMET^RS I OF :0RDS ZON : * | ,*S R O c. <5 . ic F TOER ore EOD O PEOEES N LTUDAND AREA UNDRAINED IN FLAT PIEZOMETERS OF RECORDS i "KF. ^A^A ETEI 01 S CRS F o J8 J4 DEPTHS = ; m 5 . 2 ; 5 = S H T P E D 43 J ; J48 No. S R E T E M O Z E I P OF ECORDS R 4. p k OSD p k M x 7 : : ? PI - S J IS A “ J * F e o AA 3 1 133 ----- 1986 SERIES IEZOME 1 0 CM 8 t 48 te®6 Q BE lQ EZOI A*, PIE RE -2) 1986 !ETE : [ABL zom o O A El □ J /SE J4 TERS 60 a » 9 s r JxSt* ITEF E $ IN$ S T A L L E DV 3 i EEC i M : pO jjt y SER NST io i m KP 700 oc IES aa □ \LLE I I e S I 1 £I &. & JtU -E f c L L E X J NOT5 a a KP 0 0 6 V D ' a IRT IN IRTL A a DATE □ A DATE aa i ZON JAL 30© AA A i n i 0 0 ALL D 8 E LY f n z :n o A Id N H SAME V THE IN IN ONE TH S/ I 8 1987 M Ei/ERl ME i A A ERTICA O A ET 4 J 1987 O 3 jj-D- A CI PF ICAI. 49 a a 8 J A PROF _ e © A-4 FI E A OF .E II O O A A LE A A S A AS s t O A 1988 A G a a d A o o o - D A DEPTH TO PIEZOMETRIC LEVEL < m ) DEPTH TO PIEZOMETRIC LEVEL Fig. 8 0 ECORDS O PIEZOMETERS N. 2 ; 4 ; 4 ; 0 S H T P E D 40 J ; J42 ; J41 ; J20 No. S R E T E M O Z E I P OF S D R O C RE <6-78> A GO S>G ' O i4 2. 3xl0~y Ki04= O A EE T Fi a ig F TOREFER BETWEEN "WO A EOD I PIEZOMETERS IF RECORDS CMAIO EWE TERCRS F PIEZOMETERS OF RECORDS THE BETWEEN COMPARISON A a °® °® ® a O F o 13 14 4 ; .7 m. 7. 7 ; 4. 4 = S H T P E D 104 ; 103 No. S R E T E M O Z E I P OF S D R O C E R O , , A A N o A 0 EOE ATRDANG I ZN 4 IN ZONE DRAINAGE AFTER S BEFORE A ^ * 951966 1965 A O O a <-> C-2 a Cj5-ll> a C5-12) i <5-2> . DJACENT TRENCHES ALONG''PROF I J40 J42 A O 0 3 0 1 o ! A A ft 0 /SEC M C ^ 4 0 1 97 1988 1987 J D NTLE VRVERTICAL VIR INSTALLED ° A & AV k o o DATE DRAINS PIEZOMETERS WERE H SM VERTIC/ SAMETHE EOD O PIEZCRECORDS OF o ° ° & \ AA aAA &aa a & A A a A A s\a A DATE o00o( o 0 0 Go o J E - I ZONE IN A-A _E H SAMETHE a a , L PROFILE IN IN PROFILE L METERS INSTALLEt ,0)0 ,0)0 OT UIG INE DURINGLOST G ° ° ° ° ° ° gG° * * * F * * * * * * 201 PROF E 4 IE o * * * * ALTO OF TALLATION 8 9 1 m 10 ; 9 ; 8 ; ITAL IN VIRTUALLY E M LE o M A o O o O DEPTH TO PIEZOMETRIC LEVEL C m ) DEPTH TO PIEZOMETRIC LEVEL Fig. • 5 11 10 10 9 8 7 6 3 4 2 9 1 8 7 1 6 5 3 4 2 F IZ. o 2;J7;J3 85 ETS2 : . ! ? : 18 m 11.8 : 8 ? 6 ! 4.2 : 5 DEPTHS=2. 8 7 8 J J875; No. J43; J879; J25; PIEZO. OF S D R O C E R 1 □ <6-79) [ 1,0 951966 1965 F o 22 28 23 6 9 12.4 ; 9 ; 6 = S H T P E D 203 ; 208 ; 202 No. S R E T E M O Z E I P OF S D R O C E R REC BE! REF CMAIO EWE TERCRSO PIEZOMETERS OF RECORDS THE BETWEEN COMPARISON A 4 J 1 « A AA* AA * ' ORDS WEEN •••< > R I Fi . <5- 3. ig F TI)ER J87E BEFORE & AFTER DRAINAGE IN ZONE 6 IN ZONE DRAINAGE AFTER & BEFORE J87J M ' ^ Jj25 ®©o © )® ^ INSTALLATION STARTED 1 OF TWO a P c r»'"fifi ADJAC IEZON AA 4 J ■ - * °U© aa B |-©-P ©OHnn f A± p 3 *** * i* k TR INS ETER£ ) , 2) N TRENCENT J87 8 J ' GEOQ 1987 6F *» . J DRAINS 00 0 3 e 8 * * *j S ’ H3- • • • J • • <5-1 NTLAIN F RIS CONPLETE DRAINS OF INSTALLATION ETCL RFL I ZN 6 ZONE IN PROFILE VERTICAL EOD O PEOEES iNSTAlil PIEZOMETERSRECORDS OF e GO, 3 FALLE E A HES ) t 5) G B B G 0 Vlf D LONG <5-1" □ □ • • ; s S DATE DATE K20E 5. 3x10 — K202= : ) t ’) - * T tual PROF S 0E 0 IS A l ° 11 £ 03 0 3 • • • L £ [LE IN -Y B ■ 5-IE < 9x10 GG G E 0 0E3 B 8 ■B k • • »• > It 6 - THE M D -8 GO© ZON1 CM/SEC SAME . Q 0. • • ■ B * CM/SEC G b LED VIRTUALLY IN THE SAN THE IN VIRTUALLY LED 5 E VER' n 0 ■ B fi • • 2 00 i .0 0 ICAL Bfl • • 1988 ®0£ 5 Hijj 1 3 I PROF * ° o L FIIDWA ILE BB 1967 0 • 0 o . i f fi • J < -fi • B B © G 10 0 Go DEPTH TO PIEZOMETRIC LEVEL .< Fig. 5 4 3 2 1 0 1965 O 3 (6-80) E F o J1 ; J18 DEPTHS = .1 ; .0 4 m 4. 1 ; 4. 0 ; 4. 15 = S H T P E D 8 1 J ; 7 1 J ; 16 J No. S R E T E M O Z E I P OF S D R O REC RECORDS OF EE T Fi TOREFER N OE . 7 ZONE IN CMAIO EWE TE EOD O PIEZOMETERS OF RECORDS THE BETWEEN COMPARISON A 0 1 x 5 EOE FE RIAE N OE 7 IN ZONE DRAINAGE AFTER & BEFORE .-6 1986 I O TR ISALD T NET L INVERT AT INSTALLED ZON PIE ETERS g e . <5-2> ; 222 : e v a 1987 I PIEZQME - & ETWI 1967 AFTER <5*-2> o o J1 I 7 1 J HI 2ONE. IS TH TWO AOJACEN" TRENCHES TRENCHES AOJACEN" TWO STAliLA' INSTAI 0 3 TlOh I LEII 1988 1968 DRAIN ZONE RY o o CHAPTER SEVEN - STABILITY ANALYSES OF SLOPES. 7-1 INTRODUCTION There are obviously several methods of analyses of stability of slopes. These methods can be classified into different categories with respect to the feature and shape of slip surfaces, the characteristics of the soil, the precision which is expected from the analyses, etc. As the failure surface is non-circular at Stag Hill (Skempton & Petley, 1967a), the Morgenstem and Price method (1965) is used to evaluate the stability of the area. This method was applied using the Clayton & Timche (1983) computer program. The computation was repeated for several different values of the constant factor f until a minimum factor of safety was achieved. It was observed, however, that the effect of different values of f on the resultant safety factors is not considerable and probably less than 2 per cent. In connection with the strength parameters of the soil, with respect to the extensive explanations given in section 3-6-3, the following residual and peak shear strength parameters were used: 0'r =13° c'r = 1.4 kN/m2 o'p = 20° c'p =7 kN/m2 The pore water pressure ratio values, ru, at each stage were deduced from the available piezometer records and appropriate interpolations. Further explanations will, however,be given in the following sections. With reference to the records of shallow piezometers Nos. 109, 205 and J12, and Figs. (6-24), (6-30) & (6-72) and piezometer J26, Fig. (6-69) which are installed at the toe and top of the slope, it is inferred that the phreatic surface, in the worst condition, may rise to about 0.5 and 1.0 meter below ground level (m b.g.l.) at the toe and top of the slope respectively. Accordingly, in analyses of the stability of slope at the time of failure, and after failure, but before installation of drains, the phreatic surface will be located using the above mentioned values. 7-2 STABILITY OF SLOPE ALONG PROFILE 6-6. 7-2-1 Stability at the failure time. To determine the factor of safety at the time of failure and to locate the slip surface, the conjectural ground surface before failure is inferred from the topography of the area and is shown in Fig. (7-1). With respect to the fact that there is no pore water pressure data belonging to the failure time, an accurate and realistic stability analysis is not possible. Even so, it can be reasonably accepted that the critical phreatic surface at that time must have reached the same level as it reached after failure (1965). The reason is that, the critical depths of phratic surface, about 0.5 and 1.0 m b.g.l., at toe and top of the slope, which is observed after failure in Stag Hill, is common to most slopes in London clay. The ground water distribution with depth, at the failure time, remains in uncertainty, however. Although chalk is a classic underdraining layer in London clay, and the pore water in this layer and in turn in the overlaying soil has changed since the time failure happened in Stag Hill (last century probably), this effect in shallow depths, say less than 8 meters in which failures occur, can be adequately ignored. The reason is that the stress dependent permability of soil prevents pore pressure in shallow depths to drop due to decreaing of pore pressure in deep chalk. In Fig. (8-3) there are cases showing that, although the pore pressure is nearly atmospheric at some depth (not deeper than 12 meters) below ground level, the ground water pressure profile in the intervening soil is nearly hydrstatic to some depth below ground level, however. The most imoportant change in ground water pressure pattern due to failure is that, the pore pressure profiles have been shifted somewhat towards the toe of the slope and the pro failure pore pressure pattern, observed in 1965, must have been established afterwards. Accordingly, ignoring the effect of shifting, we will necessarily accept the pro-failure porepressure pattern as a conjectural criterion to estimate the stability of the slope at the failure time. The resultant F.S. value, however, will prove how reasonable the above justification is. With respect to the above, to establish the critical pore water pressure pattern to determine the ru values, the following ground water pressure distribution profiles at three places are established and shown in Fig. (7-1). At the top of the slope, the highest piezometer levels in piezometers Nos. J26, J28 and J47, Fig. (8-3), profile 11, are assumed to be representative of the critical ground water pressure profile since the time before failure occurred. This assumption seems to be reasonable, because, this point is outside the failure zone. At the middle of the profile 6-6, the ground water pressure profile is established using the records of piezometers Nos.202, 203, & 208 at the time before installation of drains, Fig. (8-3) profile 6. Near the toe of the slip surface, however, the pore water pressure profile is interesting. With reference to the records of piezometer No.204, Fig. (6-80), it is observed that there must have been a very low piezometric level at a depth of 3.6m b.g.l. or deeper. It is seen that such a pattern would have existed since the time before installation of drains, and not induced by drains. Subsequent installation of piezometer Nos. J22, J24 and J870, Fig. (6-66), revealed that the pore water pressure is about at atmospheric at 4.2 m b.g.l. and then starts to increase again. The ground water pressure between the ground surface and 4 m b.g.l, however, is considered to be greater than atmospheric and to be distributed the same as the broken line shown in Fig. (7-1). This is confirmed with respect to profile 9 in Fig. (8-3) which is established after installation of drains. It is seen that at 4.2 m b.g.l. pressure is nearly atmospheric. From that depth to ground surface pressure is greater than the atmospheric, however. To locate the most critical slip surface, several surfaces were analysed. The slip surface shown in Fig. (7-1), was located to be the most critical, having a factor of safety of F.S. = 1.01. This indicates that the conjectural ground surface and the pore water pressure pattern have been established reasonably well. 7-2-2 Stability after failure and before drainage and development of area, 1965. Stability of the slope after failure but before development of the area is evaluated using a similar ground water pressure pattern as established in section 7-2-1. The only difference is that the critical phreatic surface is established with respect to the ground surface after failure, 1965. The purpose of this analysis is, of course, to determine the safety factor of slope before installation of the remedial measures. From the analyses, it was concluded that the slope must have had a Factor of Safety of F.S. = 1.21 at that time. The increasing of the safety factor from 1.01, at the failure time, to 1.21, in 1965, is because of change of geometry of slope due to failure, and probably some natural denudation. 7-2-3 Stability after drainage and change of the shape of slope, 1988. With reference to Fig. (7-2), it can be observed that the upper part of the slope along profile 6-6 has been covered with a layer of fill of 1.2 meter maximum thickness. As explained in Section 1-4-2, this type of weighting of slopes is always harmful. Nevertheless, the reason for this dangerous disturbance of the slope shape is not clear. To evaluate the stability of slope at the present time, the critical ground water pressure pattern is established using the piezometer records at three places as shown in Fig. (7-2). At the top of the slope the critical ground water pressure profile is the same as profile 11 in Fig. (8-3), according to which the critical phreatic surface is located at about 1.0 m b.g.l. At the middle of the slope, the critical ground water pressure profile at midway between two adjacent trenches is the same as profile 10 in Fig. (8-3). The phreatic surface between two adjacent trenches is shown in Fig. (5-18), using the records of piezometers J879, J46, J.T.23. Assuming that this surface is a parabola (although it is not parabolic indeed), and taking into account the effect of piezometric levels in the trenches themselves, the average critical phreatic surface is found to be 1.95 m b.g.l. As can be observed in Figs. (6-67) and (6-68) a very interesting pore pressure pattern dominates at depths of 8 m b.g.l. and deeper. At 8 m b.g.l., the pressure in vibrating wire piezometer J875 has gradually increased from 4.2 m b.g.l., regardless of the usual yearly fluctuations of the ground water pressure regime, and tended towards the constant value of 3.2 m b.g.l., in May 1988. At 11.80 m b.g.l., the piezometric levels in both piezometers J878 and J51, Figs. (6-67) and (6-68) increased gradually, and in September 1988, tended towards constant values of 10.10 and 8.0 m b.g.l., respectively. This phenomenon is demonstrated in section 8.3. It is of interest in this section, however, to mention that, as the slip surface is not deeper than 8 m b.g.l., the changes of ground water pressure at 11.80 m b.g.l. will not affect analysis of the stability of the slope. Near the toe of the slip, zone 7, the piezometric levels between two adjacent trenches and the ground water pressure profile at midway between trenches are shown in Fig. (5-16) and Fig. (8-3) profile 9,respectively. It is seen that there are two atmospheric surfaces. The first one is surface 3 which is inferred from the maximum piezometric levels in piezometers J24, JT15 and JT19, Figs. (6-64) and (6-66). The second one is surface 5 which is plotted with reference to the records of piezometers J16, J17 and J18, Fig. (6-65). The location of the upper average maximum phreatic surface, taking into account the effect of trenches, is calculated to be 1.95 m b.g.l. using surface 3. With respect to the ground water pressure pattern explained above, the critical safety factor is computed to be F.S. = 1.33. If the slope shape had not been disturbed, and was the same as in 1965, the critical factor of safety would have been F.S. = 1.42. The computation was also carried out for the case if the drains are blocked. In this case the critical ground water pressure pattern will be the same as the critical ground water pressure pattern, before drainage and development of the area (i.e. 1965). Nevertheless, the safety factor will be less than of that time, because the ground surface has now been disturbed. With respect to the above, the critical safety factor is computed to be F.S. = 1.14. The stability values of profile 6-6 at the different stages are summarized in table (7-1). 7-3 STABILITY OF SLOPE ALONG PROFILE A-A. 7-3-1 Stability at the failure time. The slip surface along profile A-A has been previously located by Skempton and Petley (1967a). This profile is shown in Fig. (7-3). In this section the above mentioned slip surface (ABCDE) will firstly be checked to see if it is theoretically the most critical. The ground surface before the slip, is assumed to be the same as the surface inferred by Skempton. As explained in section 7-1, the phreatic surface at the time of failure is about 0.5 and 1.0 meter below ground level (m b.g.l) at the toe and top of the slope,respectively. To infer the critical ground water pressure condition, the maximum ground water pressure profiles in three places are established using the records of piezometers installed after failure occurred. These profiles are shown in detail in Fig. (8-3) profiles 3, 4 and 5. It is assumed that these profiles are not significantly affected by the landslide, despite the fact that the locations of some of them have been shifted towards the toe of slip. This assumption is explained and justified in section 7-2-1. The profile A-A and the critical ground water pressures are shown in Fig. (7-3). Several slip surfaces were analysed and it was concluded that Skempton’s slip surface is the most critical with a factor of safety of F.S. = 0.95, provided it’s shape is changed slightly from ABCDE to LBCDE. 7-3-2 Stability after failure and before drainage, 1965. The most critical profiles of distribution of ground water pressure in four places along profile A-A, at the time before drainage, and the location of the critical phreatic surface are shown in Fig. (7-4). Details of the above mentioned ground water pressure profiles are shown in Fig. (8-3) profiles 1, 3, 4 and 5. The minimum safety factor with respect to the above mentioned ground water pressure pattern is computed to be F.S. = 1.23. 7-3-3 Stability after drainage and development of area, 1988. Levelling of the profile in some places revealed that the ground surface has not been considerably changed since 1965. In the upper part of the slope a small amount of earthworks which improves the margin of safety must, however, have been carried out. Accordingly, for the analysis of the stability of slope, the same ground profile as in 1965 is used for simplicity. An important factor which is anticipated to have increased the stability of slope along profile A-A, is that about 90 per cent of ground surface is covered by asphalt or buildings, which obviously minimise infiltration. This is demonstrated by a comparison between the records of piezometers Nos. 106 and 406, Figs. (6-22) and (6-38), respectively. Piezometer No. 106 was lost years ago, and the area in which piezometer No.406 was installed, is now fully covered by buildings and paving. According to piezometer No. 106 the piezometric level at 6 m b.g.l. was about 0.7 m b.g.l. before drainage. After drainage, however, records of piezometer No. 406, in 1986 and 1987, show that the maximum piezometric level, in a fully surfaced place, at the midway between two adjacent trenches, at 6.5 m b.g.l., is 2 m b.g.l. It is seen that in spite of the fact that the drainage system is not operating efficiently in this place and the piezometric level in the trenches is about 0.9 meters above the invert level, Fig. (6-62), the piezometric level is decreased from 0.7 to 2 m b.g.l. In zone 6 along profile 6-6, Fig. (5-18), however, despite the fact the drains are operating with full efficiency, the phreatic surface rises to about 1.0 m b.g.l.. Accordingly, the concluding point is that the factor of safety which is computed for stability of the profile A-A at the present time, should not be attributed to the efficiency of the drainage system only. The average critical phreatic surface is located using the records of piezometer No. 406, Fig. (6-38), piezometers installed in zone 4, Fig. (5-10), and finally piezometer No. 401, Fig. (6-33). Records after drainage of piezometers Nos. 101 to 110, cannot be used because at that time the area was not paved and developed. Assuming that the phreatic surface between two adjacent trenches is parabolic (although it is not), the average critical (highest) depths of the phreatic surface at the locations of piezometers Nos. 406, J13 and 401 is calculated to be 2.2, 1.8 and 2.2 m b.g.l. respectively. It is seen that the average phreatic surface located in zone 4, Fig. (5-10), is 0.4 meters higher than that of the two other places. This is, of course, attributed to the fact that the zone 4 is not paved, although it is under the rain shadow of the multi storey buildings. The critical phreatic surface, however, is shown in Fig. (7-5). Using the above mentioned ground water pressure pattern, shown in Fig. (7-5) the minimum factor of safety is computed to be F.S. = 1.58. The results of computation are graphically shown in Fig. (7-6) as a specimen. Finally, results of analyses of the stability of slope are summarized in the following table and discussed in section 8-4. Table (7-1) Summary of slope stability analyses. F.S. F.S. profile profile Description 6-6 A-A 1 F.S. at the time of failure 1.01 0.95 2 F.S. after failure, before drainage and before any development of the area, 1965 1.21 1.23 3 F.S. after drainage, developing and paving, 1988 - 1.58 4 F.S. after drainage, if the area was not paved - 1.35 5 F.S. after drainage and changing of slope shape, 1988 1.33 - 6 F.S. after drainage, if the slope shape had not been changed (disturbed) 1.42 - 7 F.S. In the case of blocking of drains 1.14 - Note: F.S. = 1.35 (in row 4) is explained in section 8-4 m O O 78 — Cr'ound 1 • v s 1 bsfo.—s slip ( CcnJ«cturc 1 5 — Cr'ound lsvsl oft.on *1 ip » 1955 — - Existing Cr'oundCr-ound lsvsl , 1965Existing 74 10 70 7 2 55 TOE CL 1985 58 CL 1955 54 2 0 8 S3 50 10 30 50 60 70 mOD 78 Cnound lsvsl bsfor-s £lip ( Conjsetui-al > Cnound 1 1 oftnr- slip, 1965 Existing Cr'ound1 * 1 * 1985 74 10 70 70 | 55 .62 TOE CL 1986 58 V_8.8 CL 1955 A. 2 54 J 8 7 8 j j, 8 7 8 i . 50 50 10 50 70 G round 1 1965 60. -— rrT -urfoc- FENCE t TOE Slip 20 50. 13.5 x x x ,00 10 H&V Scolo 19S5 20 60_L 20 50_L_ P1E7 13.5 X x x 30 50 60 70 60 ,00 10 Ground level . 1965 3 .2 50. 12.5 x x I x x 0. 0 20 70 1-- 1 m i 0EI-1 LU: m 021- 011- / 00 fe i 001- / R Co • 08- uo He! 00- II U- X +> 09- CD i <4- 0 0S- / to 0*- <4- I 0 II 00- L 0 o “0 02 - +> -Q 0 H 01- 0 O Ll ea ca 0- j —-I-—I— i— 1 8 csi '■r (IM) C“W> ^U0WOW CAPO UDeliS BEX- -081 X 021 BIX- 011 i 021' 001 SIXV-A J I I L n f n (M h i' & *0 [DUUeqUJ ?i. Fig- ( 7 - 6 ) Th© graphical representation of stability a n a ly sis as an ©Xampl© < profile A-A after drainage. 1988 > CHAPTER EIGHT - DISCUSSIONS AND CONCLUSIONS. 8-1 ASSESSMENT OF THE EQUALISATION TESTS 8-1-1 Evaluation of first time equalizations. With reference to Fig. (6.1) and Table (6.2), it can be observed that in an inflow equalization test (rising head),the first time equalization process, regardless of the type of piezometer, takes a much longer time than the subsequent equalizations. It can be observed that for standpipe piezometer J31, time lags for 90% equalization in the first and second time tests are 11.66 x 104 (about 81 days) and 39 x 103 minutes (27 days) respectively. In other words, in first time equalization, there must have been 77.6 x 103 minutes (about 54 days) delay for 90% equalization. It is again pointed out that a first time equalization test is a test which is carried out immediately after a piezometer has been installed. In the case of the pneumatic piezometer J37, in the first time equalization test, 90% equalization occurred in 28 x 102 minutes (about 2 days). In the second time equalization test, which was carried out only 7 days after installation, the time lag for the same percentage of equalization decreased to 170 minutes. Therefore, in the case of the pneumatic piezometer in the first time equalization test there must have been 26.3 x 102 minutes (1.8 days) delay for 90% equalization. In the case of the piezometers which were installed elsewhere, the first time equalization process could not be studied because of the effects of the fluctuations of the piezometric level on the equalization tests. Even so, the results of the first and second time outflow (falling head) equalization tests of two piezometers J22 and J50 at depths of 4.2 and 15 m b.g.l., respectively, are given in Figs. (6-9) and (6-10). Records of these piezometers do not show appreciable fluctuations, so the effects of the changes in ground water pressure, on the equalization tests, can be ignored. It can be seen that in the outflow (falling head) tests, in contrast to the above mentioned cases, the rate of the first time equalization is probably somewhat quicker than for subsequent times. It seems that the inflow (rising head) first time equalization proceeds more slowly than the outflow (falling head) first time equalization. This can be interpreted as follows: In Chapter 2 it was explained that the factors affecting the first time equalization are swelling of bentonite pellets, setting of grout, displacement of the sides of borehole and disturbance of the stress condition due to drilling and the formation of a cylindrical layer of remoulded clay all the way round the borehole which can have properties different from the properties of the undisturbed soil. The above mentioned phenomena are probably partially due to the swelling of the pellets which try to absorb water and in this way retard inflow and advance outflow equalization. Another reason could be that as the pore pressure and in turn, the water content of the soil around the hole is reduced due to drilling, part of the water in the first time equalization process, is absorbed by the surrounding soil moving towards its original moisture content. Interestingly, in the case of the pneumatic piezometer J37, the factors retarding the first time inflow equalization have all been eliminated within 7 days of installation. The results of the second time equalization tests carried out just 7 days after equalization, and subsequent results from tests carried out 6 months later, did not show any appreciable differences. In the case of the standpipe, J31, the retarding factors exerted an influence for a considerably longer time, however. For example, in the first time equalization of piezometer J31, the time lag corresponding to 90% equalization was about 54 days longer than the time lag for the same per cent equalization in the second time. This suggests that the initial stress adjustment time lag (installation time lag), is a function of the volume change of the piezometer. Adjustment of the disturbed stress condition of the soil around the intake point of a standpipe piezometer takes a much longer time than in the case of the pneumatic piezometer. Most probably the swelling process of the soil takes a long time because of the slower pressure build up in the standpipe. This process, however, can be considerably shortened if the standpipe is filled with water immediately after installation has been completed. It seems that the initial stress adjustment process tends to advance or retard the first time equalization process whether the test is falling or rising head. It is recommended that in order to get accurate pore pressure measurements within a comparitively shorter time of installation, standpipe piezometers are filled with water immediately after installation. With respect to the above, installation time lag, (i.e. initial stress adjustment) is governed by the volume factor of a piezometer. The less the volume factor the shorter the installation time lag, and that is why a rigid pneumatic piezometer has the great advantage of being able to come into equalization with the surrounding soil within a few days (7 or less) of installation. 8-1-2 Evaluation of the subsequent time equalizations of piezometer J37, (tests series WA”). For a proper interpretation of the results of subsequent time equalization tests of the pneumatic piezometer J37, the mechanisms of operation of different types of piezometers are briefly discussed below. In the case of an open standpipe, water flows into or out of the system due to any out of balance pressure. Obviously, the response time of a piezometer is a function of the amount of flow of water into the piezometer and the transient redistribution of pore pressure in the soil around the piezometer intake during equalization. This is also tiue in the case of a vibrating wire piezometer. When ground water pressure increases, water flows into the ceramic due to further deflection of the membrane of the cylindrical transducer. The amount of this deflection, which is proportional to the amount of water entering the system, is a representative of the volume factor (v) of the system. In the case of a pneumatic piezometer, however, the concept of volume change or flexibility is not the same as above. After a pneumatic piezometer has been installed the diaphragm never deflects due to an increase or decrease in ground water pressure. So, under ideal conditions, when there are no air bubbles at the intake point (sand filter, ceramic and transducer) no water flows in or out of the piezometer cavity due to fluctuations in the ground water pressure. Elements of pneumatic piezometer are clearly shown in Fig. (4-2). It is seen that there is no cavity between the diaphragm and the base of the transducer. The base of the transducer is even infmitesimally dome-shaped and so the diaphragm naturally (without any external pressure) lays on the transducer. The other member which screws the diaphragm onto the transducer tightly has a concave perforated part in the middle. The diaphragm can deflect into this part when the pore water pressure is measured. The amount of deflection of the diaphragm is nominally called the volume change of the piezometer. The maximum nominal volume change of the pneumatic piezometers used in this research is 0.03 cm3. As the diaphragm deflects for a very short period (say 10 to 20 seconds) when the pore water pressure is measured, any flow of water from the system into the soil during application of the nitrogen pressure,and into the system after the measurement has been finished, can be ignored. Probably an instantaneous elastic deformation is caused at intake point due to unbalanced pressures. This was proven with a simple test as follows: Pore water pressure was measured in piezometer J37 six times consecutively, the whole test taking about 5 minutes. In the first measurement 0.9 bar of nitrogen pressure was applied and pore pressure was measured to be 0.38 bar. Immediately after the first measurement, the second and subsequent measurements were carried out with the same nitrogen pressures. In all measurements pore water pressure was measured to be the same as the first time. No drop in measured pore water pressure was observed. This suggests that immediately after measurement, pore pressure at the intake point returns to its initial value and virtually no water flows from the piezometer intake to the ambient soil and vice versa, otherwise the subsequent readings would not be the same as first time. In another test 0.9 bars nitrogen pressure was applied to the system J37 and pore pressure was measured to be 0.38 bars. Then the system was left under 1.6 bars of nitrogen pressure for 5 minutes and then the pore water pressure was immediately measured. This time the measured pressure had dropped to 0.36 bars. This suggests that a flow must have occurred from the system towards the ambient soil. It does not mean that a gap between the diaphragm and base of the transducer had been left after removal of 1.6 bar nitrogen pressure, because in this case the subsequent measurement of pore pressure would be zero. Accordingly, immediately after any application of nitrogen pressure into the transducer for measurement of pore water pressure or any other purpose, the diaphragm is pushed back onto the transducer. No more water can flow into the transducer due to a subsequent increment of the ground water pressure. This suggests that, any time lag which is observed in an equalization test or an in situ measurement of ground water pressure with pneumatic piezometers is predominantly representative of stress adjustment time lag. Now we return again to Fig. (6-1). Here it can also be observed that within a minute of the pressure on both sides of the diaphragm was decreased to zero, pore water pressure build up was again recorded by the piezometer (measurements in less than one minute were not actually possible). It can be assumed therefore, that no flow could have occurred from the soil into the system within a minute of the test commencing, and these curves are indeed predominantly representative of pore pressure build up in soil itself which is governed by consolidation and swelling characteristics of the soil(stress adjustment) around the borehole. With respect to Table (6-2) it is seen that the average time lag for 90% equalization of a pneumatic installation is about 130 minutes, which is about 286 times quicker than a standpipe installation. With respect to the above discussion, it is probably not possible to achieve quicker responses unless the properties of the soil are changed. The above discussion can lead to the following conclusions which are of interest to this research: A pneumatic installation, under ideal conditions, can record ground water pressure changes with virtually no hydrostatic time lag. The very short time lags which are observed in equalization tests with the pneumatic piezometer J37 are time lags corresponding to transient pore water pressure buildup in the soil in the vicinity of the intake point (stress adjustment time lag). Therefore, the records of a pneumatic installation can be virtually accepted as the real ground water pressure at piezometer intake level. 8-1-3 Evaluation of the equalization curves series In Fig. (6-2) first time equalization curves of some piezometers are shown. Of these curves the curve belonging to piezometer J44 is different from the others and from the theoretical solution (Gibson, 1963), in that it does not tend to an asymptote to the 100% equalization line. Indeed the gradient of the upper part of the curve, in contrast with the theoretical solution, shows an increasing trend. To elaborate the factors causing this type of equalization, average rates of change of water level (r value) in some piezometers during the weekly periods are shown in Table (8-1), in terms of mm/day. J44 and J22 are standpipe piezometers of 8 m and 4.2m depth in the drained area. J50 is a standpipe piezometer of 15m depth in the non-drained area. Attention to this table shows that in the case of J44/1 the equalization process would have continued with repetitions of an r value of 2mm/day at later stages of equalization for nine consecutive weeks. It could probably have continued for longer periods if intensive rainfall had not started in October, 1987, Fig. (6-17). Although in two other piezometers, J22 and J50, tests were being carried out at the same period of the year as piezometer J44, the r value shows a consistently decreasing trend in consecutive weeks. Installation of the pneumatic piezometer J41 in the vicinity of piezometer J44 ,Fig. (5-12), makes it possible to study the state of the ground water pressure at a depth of 8 m below ground level. With respect to the daily records of this piezometer, Fig. (6-14), ground water pressure must have been decreasing with an average rate of r = 1.8mm/day^at 8 m b.g.l., at the location of piezometer J44 during the six months from May to October. Records of piezometers J22 and J50, Figs. ( 6-66) and (6-73) do not show appreciable fluctuations of pore pressure, however. With respect to the above, it can be concluded that the abnormality in the shape of the equalization curve corresponding Table (8-1) Average daily changes of water level (r values) in stand pipes during falling head equalization tests. r values (mm/day) Week piezometer, piezometer, piezometer, Week ending and test order and test order and test order No. on: J44/1 J22/2 J50/1 30/4/87 1 7/5/87 2 37 14/5/87 3 23 21/5/87 4 20 28/5/87 5 11 4/6/87 6 6 11/6/87 7 7 18/6/87 8 7 25/6/87 9 6 2/7/87 10 5 9/7/87 11 4 416 16/7/87 12 5 228 23/7/87 13 4 130 30/7/87 14 4 201 88 6/8/87 15 4 129 62 13/8/87 16 3 81 37 20/8/87 17 3 52 26 27/8/87 18 3 41 20 3/9/87 19 3 33 11 10/9/87 20 3 22 7 17/9/87 21 3 10 5 24/9/87 22 3 0 4 1/10/87 23 3 0 3.5 8/10/87 Table (8-1) Average daily changes of water level (r values) in stand pipes during falling head equalization tests (cont’d). r values (mm/day) Week piezometer, piezometer, piezometer, Week ending and test order and test order and test order No. on: J44/1 nvi J50/1 8/10/87 24 3 0 3 15/10/87 25 2 0 0 22W 87 26 2 0 0 29/10/87 > 27 0 0 0 5/11/87 28 -3 0 to piezometer J44 must have been due to comparitively long term decrease in ground water pressure. This is why the equalization curve shows some discrepancy from the theoretical solution (Gibson, 1963) which is based on a constant ground water pressure. For compressible soils, when the ground water pressure changes with a constant rate of r, there is no theoretical solution for equalization. Ignoring the compressibility of soil, a solution is given in chapter 2 (after Hvorslev, 1951). A summary of the solution is as follows: _t Yt - rt _ t T e Y o -r T in which A T = FK F = intake factor, given in table (6-1) A = standpipe cross sectional area, given in table (6-1) yt = the unbalanced pressure at the time t = t after beginning of the test. yQ = the unbalanced pressure at the beginning of the test, can be estimated using piezometer J41 (y= 3.10m). k = the permeability of the ground has been previously, (1966), obtained from piezometer No. 104, (2.3 x 10 -7 cm/sec), prior to digging the drainage system, r = the average rate of change of piezometric level, 0.18 cm/day The modified equalization curve using this formula (instead of the usual method explained in section 5-7-1) leads to the equalization curve shown in Fig. (8-1). It is seen that when the ground water pressure started to increase again in November only 84% equalization had occurred in piezometer J44. Finally, even if the ground water pressure had continued to decrease, the equalization curve would probably have tended towards a certain equalization rate of less than 100%. Of course, if a less flexible system had been installed, like a narrower standpipe and/or bigger intake point, then the water level (yt) would change more rapidly in piezometer, and the rate of change of ground water pressure, r value, would be less significant in comparison with the rate of change of water level in standpipe (Ayt). Therefore, the equalization curve would have tended toward an asymptote to an equalization line much higher than 84%. Apart from the equalization curve discussed above, the equalization curve KP2/2, Fig. (6-4) also shows an odd shape. This curve inflects when 80% equalization has occurred. Although no exact and certain justification can be given, it is likely to be because of the leaking of the standpipe. Probably water leaks into an empty space (like a gap between the seal and pipe) in the borehole but not into the soil; as 80% equalization does not take more than 30 minutes (discussions with Professor Penman). Whatever the reason, piezometer KP2 is not trustworthy. 8-1-4 Conclusions The first time outflow (falling head) equalization of standpipe piezometers, in the same clay, proceeds faster than the first time inflow (rising head) equalization. Accordingly, in order to measure the ground water pressure more accurately within a comparitively short time (say 3 to 4 months faster for 95% equalization) it is recommended that the standpipe be filled with water immediately after installation has been completed. A pneumatic piezometer installation, under ideal conditions, can record ground water pressure changes with virtually no hydrostatic time lag. The very short time lags which are observed in equalization tests with the pneumatic piezometer J37 are the time lags corresponding to the transient pore water pressure build up in the soil surrounding the intake point (stress adjustment time lag), and the records of a pneumatic installation can be virtually accepted as the real ground water pressure at piezometer intake level. Although Premchitt and Brand’s method (1981), based on Gibson’s solution (1963), is a useful way of estimating in situ soil properties, it suffers from the changes of the ground water pressure, however. Correction of the equalization curve is not possible, unless the rate of change of the ground water pressure (r value) can be estimated. Finally, the corrected equalization curve may not lead to a useful result, even if the rate of change of ground water pressure is known. Such a case may be encountered in a drained area in which the ground water pressure continuously and quickly fluctuates. 8-2 ASSESSMENT OF THE DIFFICULTIES AND ERRORS IN MEASUREMENT OF GROUND WATER PRESSURE 8-2-1 Instrumental Errors: Before evaluating the effects of atmospheric pressure on ground water pressure and associated errors and errors associated with time lag, it is important to appreciate the instrumental errors. The magnitude of instrumental errors relating to the three types of piezometers, used in this research, is briefly explained as follows: standpipe piezometers: Errors in the measurement of the depth to water level in the standpipe are sources of instrumental errors. The precision of dipmeter readings is approximately ± 2 cm water head, provided the brass probe is not damaged and is free from dirt and grease. It was observed that when the brass probe is not clean, the precision of reading may decrease by the whole length of the probe, i.e. 8 cm, and in extreme cases no signal is emitted. pneumatic piezometers: Errors arise from the pressure transducer and readout unit. For piezometer transducers an accuracy of ± 20 cm water head is mentioned by the manufacturer. The standard calibration of piezometers, however, showed that the water pressure is always under-estimated with these piezometers. The calibration pattern tolerates ± 5 cm water head if calibration is repeated for pressure ranging from 0.5 to 10 meters water head. For this pressure range, the magnitude of under-estimation was (10 to 20) ± 5 cm water head depending on the piezometer and water pressure. The concluding point is that having a piezometer calibrated, the ground water pressure ranging from 0.5 to 10 meters water head, can be estimated with an error of ± 5 cm water head. For pressures less than 0.5 meters water head, although the pressure was underestimated, calibration was not possible because of the limitation of the readout unit and lack of repeatability in consecutive calibrations. With respect to the observations, an under-estimation in the order of 20 cm water head can be assumed for pressures of 0.5 m head or less. The above mentioned calibrations are true when a high air entry value ceramic and a pressure gauge readout unit, chargeable up to 60 bars, is used. Water pressure, however, was under-estimated up to 70 cm water head when piezometers with high air- entry ceramic were calibrated with micro digital readout units (Micromatic readout unit). Piezometers without ceramic or with low air entry ceramics, however, showed calibration patterns the same as achieved previously. vibrating wire piezometers: Piezometers reading up to 5 kg /cm 2 pressure, with a nominal accuracy of ± 0.1% of full scale (according to the manufacturer) were used. Calibration of these piezometers in the laboratory, under constant temperature, showed that measurement with an accuracy of ± 7 cm water head was possible. The point is that, unlike the pneumatic piezometers, the water pressure is sometimes under-estimated and sometimes over-estimated. The above mentioned accuracy is, of course, the maximum precision which can be achieved under stable conditions of operation, in which the displayed reading varies by only ± 1 digit on successive impulses. 8-2-2 Errors and difficulties associated with atmospheric pressure. 8-2-2-1 -Introduction. Prior to evaluating the effects of atmospheric pressure on ground water pressure and discussing the associated errors, it must be mentioned that the precision of the evaluation of these effects is limited by the range of instrumental errors, as described in section 8-2-1. So, to achieve a clearer result, the discussions and comparisons will concentrate on the greatest changes of atmospheric pressure possible. The other point is that as the effects of changes of atmospheric pressure on the records of piezometers are generally influenced by the effects of rainfall, the difficulties associated with each of the above must be evaluated carefully with respect to each other. There are some special cases, however, in which the effects of rainfall on the ground water pressure, can be ignored in the short term against the dominant effects of atmospheric pressure. These cases, which simplify the evaluation of the effects of atmospheric pressure on the piezometric records,will be discussed wherever encountered. 8-2-2-2 The efficiency with which atmospheric pressure affects ground water pressure. To reduce the complexity, records of two vibrating wire piezometers J871 and J877, zone 4, will firstly be evaluated. The discussion will then be extended to cover the records of the other vibrating wire piezometers. The compatibility of the conclusions achieved with the records of other types of piezometers, will be evaluated in section 8-2-2-3. Records of the above mentioned piezometers together with some others, at two different depths are shown in Figs. (6-13) and (6-14). Locations of piezometers are given in Figs. (5-12) and (5-13). A comparison between the records of piezometers and records of rainfall, Fig. (6-17), reveals that the ground water pressure at 5 m b.g.l. is considerably and quickly affected by rainfall. At 8 m b.g.l., however, the effects of rainfall on the ground water pressure are not considerable over a short period. It is of greater interest to compare the records of piezometers in October, 1987, the month in which there was 226.8 mm rainfall. It is seen that while piezometer J20, 5 meters deep, showed 1.1m change in piezometric level, piezometer J41, 8 meters deep, did not show any appreciable response to such heavy rainfall. The reason and further elaboration of this phenomenon will be given in section 8.5. It is of interest in this section, however, to conclude that at depths 8 m b.g.l., and deeper, the effects of atmospheric pressure on the ground water pressure can be studied ignoring the effects of rainfall. At the shallower depths, however, the separation of the effects of rainfall and atmospheric pressure is not generally simple. An study of Fig. (6-14) shows that there is a good similarity between the graphs showing records of piezometer J871 and changes of atmospheric pressure. The similarity can clearly be seen in November 1987, the month during which the atmospheric pressure fluctuated by as much as 51 m b. This suggests that any change in atmospheric pressure is reflected instantaneously on the records of this piezometer. With respect to Fig. (6-17) it is observed that in August and September the rainfall was the lowest in 1987. As the evaporation is usually considerable during these months, the infiltration, and in turn the effects of rainfall on the ground water pressure, can be ignored. This assumption is reasonable because, with reference to the records of the pneumatic piezometer J20, Fig. (6-13), it is seen that the piezometric level has shown virtually no response to the rainfall. Therefore, the effects of atmospheric pressure on the records of piezometer J877, during August and September may be evaluated regardless of the effects of rainfall. Now, it is clearly seen that all changes of atmospheric pressure during August and September were reflected on the records of piezometer J877. During the other months, in spite of the fact that the ground water pressure pattern is considerably affected by rainfall, the effects of changes in atmospheric pressure on the records of piezometer J877 are still clearly distinguishable. As an example, it is seen that the sudden change of atmospheric pressure on June 19th, Fig. (6-13), is completely and instantaneously recorded by piezometer J877. This again shows that atmospheric pressure acts as a sudden undrained loading at this location. As the rate of changes of pore pressure due to such loading is not controlled by the soil properties, its effect is rapidly reflected on the ground water pressure. Changes of pore pressure due to rainfall, however, are not so rapid because of the interference of the effects of soil properties (stress adjustment time lag). It is interesting to mention that in contrast to previous views (Eggboro and Walthall, 1985), the change of atmospheric pressure in this point is efficiently transmitted through the clay. Now, the records of three other vibrating wire piezometers, J879, J875 and J878, zone 6, which are shown in Fig. (6-11) and Fig. (6-12) respectively, will be considered. A comparison between the records of these piezometers and changes of atmospheric pressure reveals that here also the changes of atmospheric pressure and the records of piezometers are similar. The interesting and important point, is that at this location in spite of the fact that any change in atmospheric pressure is recorded by piezometer J879, only a small part of it is transmitted to the records of piezometers J875 and J878, however. Records of these piezometers in November give a clear picture confirming the above. The position of these piezometers are shown in Fig. (5-18). It is clearly seen that the similarity between the records of the piezometers J875, J878, and records of the atmospheric pressure is not perfect. In November, of the 51 mb change in atmospheric pressure, only about 15 mb (1 mb = 1cm water head) is transmitted to the piezometric levels at 8 and 11.8 m b.g.l. Now the records of piezometers J871 and J875, Figs. (6-14) and (6-11) are compared. Both piezometers are at the same depth but in different locations. It can be observed that the similarity between the atmospheric pressure records and records of piezometer J871 is much better than that of piezometer J 8 75. A similar observation can be made with the vibrating wire piezometers J877, J874 and J873, Figs. (6-13), (6-15) and Fig. (6-16). All piezometers are 5 meters deep, but J877 and J873 are installed in different zones between two trenches and J874 is installed adjacent to a trench, Figs. (5-13) and (5-19). In the case of piezometer J877, as explained previously, any change of atmospheric pressure is completely reflected on the record of the piezometer. In the case of piezometer J 8 73, of the 22 and 24 m b maximum changes of atmospheric pressure in August and September, only about 8 and 15 mb are reflected on the piezometer records respectively. Perhaps only about 50% of any change of atmospheric pressure is transmitted to the records of the piezometer J873. Finally, for vibrating wire piezometer J874, Fig. (6-15), clear observations are available in September and December, periods in which the effect of rainfall on the piezometric level was not considerable, according to the standpipe piezometers J27. Of the 24 and 37 mb changes of atmospheric pressure in September and December, only 13 and 14 mb are transmitted to the piezometric levels respectively. These observations all reveal that, in general, only part of the changes in atmospheric pressure is transmitted on the ground water pressure and the efficiency of transmission varies from depth to depth and from place to place. The transmission is instantaneous, however. The former phenomenon cannot be attributed to the flexibility of the vibrowire piezometers because, as mentioned in section 4.3.1, the overall volume change of piezometers is 0.005 cm 3 against 5 atm (5kg/cm2) pressure. If the permeability of the clay is taken as 10_9 cm/sec, which is even less than the average of the permeability values given in table (6-4), the hydrostatic time lag corresponding to 99% equalisation will be about 30 seconds. This figure is quite insignificant in comparison with the rate of changes of atmospheric pressure which is very slow and usually is not noticeable in periods of less than one hour. The transmission of change of atmospheric pressure is perhaps a function of the ground water pressure profile, depth of water table and degree of saturation of soil, and further research is, of course, necessary in this connection. The scope of the present research, however, is to evaluate the difficulties and errors, initiated by changes of atmospheric pressure in ground water pressure measuring techniques. The effect of atmospheric pressure on the ground water pressure is indeed a consequence of a major pressure and a minor fluctuating pressure of the orders of 1000 mb and 50 mb, respectively. The reason for this separation is that we do not know if the major part is transmitted through the soil with the same characteristics as the fluctuating part. If, for example, it can be accepted that the efficiency of transmission of the changes of atmospheric pressure through the soil is time dependent (Weeks, 1979), then it should be accepted that each part is transmitted with a separate efficiency. Therefore, to cover the uncertainty and for more generalization, it is assumed that the major constant part (Pa min) is transmitted with the efficiency of S0 and the changing minor part (AP)with an efficiency of Si. S0 is nominally defined as the ratio of the effect of the average minimum atmospheric pressure on the ground water pressure at any point below the ground surface to the average minimum atmospheric pressure at the ground surface. In practice, however, as the changing part contributes only about 0.5 percent of total atmospheric pressure, S0 at any point below the ground surface may be defined as the ratio of the total atmospheric pressure at that point to the total atmospheric pressure at the ground surface. Measurement of the total atmospheric pressure in the ground is not simple, however, because the ground water pressure is involved. The efficiency of the changes of atmospheric pressure at any depth below the ground level £1 is defined as follows: AP £l APa in which: APa = change of the atmospheric pressure at the land surface within a certain period. AP = change of ground water pressure due to APa. Values of Si can be estimated comparing the records of atmospheric pressure and records of vibrating wire piezometers during a certain period in which the rainfall has not had a dominant effect. The Si values at some points below the ground surface have been estimated using the records of vibrating wire piezometers during the most suitable months, and the sequence of estimation and results are summarized in table ( 8 - 2) as follows: Table (8-2) the Si values, deduced from Figs.(6-11) to (6-16) Piezo. Depth Month atm.pressure APa AP Si No. Pai min Pai max m b.g.l. mb mb mb (cm) % J871 8 Nov. 987 1038 51 48 94 J875 8 Nov. 987 1038 51 15 29 J878 11.8 Nov. 987 1038 51 15 29 J877 5.2 Aug. 1004 1026 22 21 95 J873 5 Aug. 1004 1026 22 8 36 J873 5 Sep. 1006 1030 24 15 62 J874 5 Sep. 1006 1030 24 13 54 J874 5 Dec. 1000 1036 36 14 38 J879 4.2 - Aug. 1004 1026 22 22 100 1 mb = 1 cm water head APa = Pal max - Pai min. Pai max & Pai min = maximum and minimum amounts of atmospheric pressures during the month noted, at ground surface. APa = Pai max - Pai min. AP = maximum change of piezometer record during the mentioned month. AP Si = = Efficiency of change of atmospheric pressure at any depth. In section 8 -2-2-3 it will be shown that the uneven transmission of the changes of atmospheric pressure to the piezometric levels is a considerable source of error and difficulty in monitoring the ground water pressure. 8-2-2-3 Errors in measurement of the ground water pressure associated with atmospheric pressure. The effect of the changes of atmospheric pressure on the ground water pressure and its interference in the records of vibrating wire piezometers was revealed in the previous section. The range of errors in measurement of the ground water pressure associated with the effects of atmospheric pressure and methods of correction for different types of piezometers are discussed in this section as follows: I - vibrating wire piezometers In section 2-1-1 the mechanism of operation of vibrating wire piezometers was explained. It was shown that the pressure change AP could be calculated using the following formula AP = k [ % 2 . 1^2] The N 0 value (pre installation base reading) is determined before installation, then for any Ni value after installation, the pressure change AP can be calculated. Now, a vibrating wire piezometer, which is prepared and then installed in a ground with atmospheric pressure efficiencies of 80 and 81 , as defined in the previous section, will be considered. It is assumed that the atmospheric pressure at the time of pre-installation base reading (N0) was Pao. After installation, however, the atmospheric pressure at the ground surface changes to Pai, (AP = Pai - Pao) and the piezometer reading changes to Ni due to the application of the ground water pressure and change of atmospheric pressure from Pa0 at the ground surface to [e0 Pa min + 81 (Pal - Pa min)] at the piezometer intake point. The values of atmospheric pressures Pa, total pressure resulting from atmospheric and ground water pressures, and readings of the piezometer before and after installation will be as follows: atm.pressure total pressure piezometer at ground surface at intake point reading. Pa Pt before installation Pao No after installation Pai 80 Pa min +&i (Pai - Pa min) + Pw N 1 with respect to the above: [80 Pa min + 81 (Pai - Pa min) + Pw] - Pao = k - 1/fy2] from which P w k [ VjSf2 - V tf2] + {[Pao - Pa min (80 - 61)] - 81 PaJ} (8- 1) Assuming C {[Pao - Pa min ( 8 © - £>1)] - 6 1 Pai } (8-2) thus (8-3) In which N0 = Pre-installation base reading of piezometer Ni = Reading of Piezometer after installation Pa min = The average minimum atmospheric pressure of the area Pao = The atmospheric pressure when the N 0 is read (at ground surface) Pai = The atmospheric pressure when the Ni is read (at ground surface). 50 = The efficiency with which the Pa min is transmitted through the soil over the ground water pressure 51 = The efficiency with which the changes of atmospheric pressure, (Pai - Pa min), is transmitted through the soil over the ground water pressure. e = error in measurement of ground water pressure associated with atmospheric pressure, or the amount with which the k [ 1/ft2 - 1/N^] 'value must be corrected. Pw = The ground water pressure, independent of atmospheric pressure. k = Piezometer constant The equation (8-3) is the general equation of the measurement of ground water pressure using vibrating wire piezometers, which takes into account the effects of atmospheric pressure on the ground water pressure. The e value has not been introduced in ground water measuring techniques so far. Consequently, the associated errors are either completely ignored, assuming e = o, or simplified inadvertently, assuming S0 = £1 = 100% (Soil Instruments Limited, vibrating wire piezometer users’ manual). A precise evaluation of the effects of atmospheric pressure over the ground water pressure is still not possible because of the uncertainty of the range of S Q. However, if it is assumed that the major part of the atmospheric pressure is completely transmitted all the way through the soil over the ground water pressure then, taking SQ = 100%, the magnitude of errors associated with the changes of atmospheric pressure can be evaluated using the Si values given in table (8-2). As seen in that table, the Si values may change from 100% to as low as 29%. In table (8-3) the extreme values of errors, ei and e2, which are caused by atmospheric pressure, are given. Pai min and Pai max show the minimum and maximum values of atmospheric pressures during the monitoring period of the ground water pressure. Pa min is the average minimum atmospheric pressure during the last two years, which is accepted to be the major constant body of the atmospheric pressure. To show the range of errors which may be involved if the readings of a piezometer are corrected against the effects of atmospheric pressure simply without checking the Si values, the extreme values of e, (e, and e'2) are also calculated assuming Si = 1 and shown in table (8-3). The positive and negative signs of e and e' show the values by which the ground water pressure is under-estimated and over-estimated respectively if the effect of atmospheric pressure is ignored. The e"i and e'2 are indeed the correction values which are simply suggested in the users’ manual of vibrating wire piezometers. It can be observed that the higher the efficiency of the changes of atmospheric pressure the closer the ei and e2 are to e'i and e'2 values, respectively. On the other hand, when the Si diverges considerably from 100%, then if the Si is not estimated, considerable errors may be involved in the correction of the readings of piezometers for the effects of atmospheric pressure. For example, it can be observed that, in the case of piezometer J 8 78, an error of the order of 38 mb (or 38 cm water head) may be involved (e2 - e'2 = 38). For further elucidation of the importance of the evaluation of the effects of atmospheric pressure in ground water pressure measuring techniques, the records of piezometer J875, before and after correction, are shown in Fig.(8-2). Graph (1) shows the piezometric levels as calculated directly from, [1/ft 2 - 1/ft^] • Graphs (2) and (3) show the ground water pressure levels as calculated from equation (8-3) assuming 61 = 0.29 and 61 = 1, respectively. The clear difference between the graphs (1) and (2) reveals the necessity of the correction of piezometer readings for the effects of atmospheric pressure. It can be observed that when the piezometer readings are corrected using the method which was already introduced, equation (8-3), a consistent and linear piezometric level, graph (2), is achieved. If, however, the correction is carried out regardless of the estimated 81 values (assuming Si = 1) a quite distinctive piezometric level will appear, graph (3), and considerable errors will be involved. It is seen that graph (3) is sinusoidal rather than consistent and linear. It must be mentioned that graph (2) is confirmed further, being in agreement with the records of pneumatic piezometer J41, Fig. (6-14), which is installed in Zone 4 at the same depth and position with respect to the drainage system as J875. It will be shown that records of piezometer J41 are representative of ground water pressure (Pw) only and not affected by the changes of atmospheric pressure, because 81 is unity at this particular point. With respect to the above it is seen that the range of errors which is initiated ignoring estimated values of 81 may be so high that there will be no point in using such Table (8-3) Extreme values of errors associated with atmospheric pressure in measurement of ground water pressure with vib.wire piezometers. " rO h p PH PH P P P S d S d d d d d cd ii ii h h h co 1 m a x *o O n O O - s 0 o n s 0 F-H O - s 0 to P ph B B II P 03 ctf h o CO oo wo wo oo t~- N O oo N O oo H O m in = h j* 'O f £ 13 rO .3 P • bO o CO 3 4> a 4) P 3 •w 13 O h p % ) 4 CO 3 a o b a p p 4) l U H p Ul ? H p T 03 P P 1 3 I .3 *G $ -*-> 13 CO CO e'i & e'2 = The ’e’ values when the atmospheric pressure drops to Pai min or increases to Pai max respectively, assuming Si = 100%. Pal = Atmospheric pressure at the time of reading of piezometer (Ni). 271 an expensive device. To overcome this difficulty, however, it is recommended that the atmospheric pressure and the pre-installation base readings, (Pao & N0) are simultaneously recorded. Then after the piezometer has been installed, beside the usual weekly or monthly readings, a simultaneous daily reading of the piezometer and atmospheric pressure must be taken and plotted for a month, or two weeks at least. Then the £1 value is estimated in the same manner as shown in table (8-2). If it is anticipated that the rainfall will have a rapid and considerable effect over the piezometric level, the daily measurements should be carried out during the most dry period of the year. Finally, piezometer records must be corrected using equations (8-2) and (8-3). II- pneumatic piezometers: In the case of pneumatic piezometers, the readout unit itself is not affected by the changes of atmospheric pressure. However, after it is connected to the piezometer to measure the ground water pressure the whole system will be affected by the atmospheric pressure. If the instrumental and other sources of error are ignored then the amount of error associated with the atmospheric pressure will be as follows: After a piezometer has been installed, the total pressure which is applied to the piezometer diaphragm will be equal to Pw + £0 Pa min + £i (Pai - Pa min). As the readout unit is not isolated against the atmospheric pressure, the displayed total pressure by the readout unit will be reduced by Pai. Therefore, the displayed pressure will be equal to: Pd = Pw + So P a min + Si (Pai - Pa min) - Pai From which: Pw — Pd [Pal (1 " Si) - Pa min (£© - £i)] (8-4) assuming: e ” [Pal (1 - Si ) - Pa min (£0 - £i)] (8-5) Then P w - Pd 6 (8-6) All parameters have the same meaning as before. In equation (8-4), Pd is the pressure which is displayed by the readout unit. The other parameters have the same meaning as before. The ’e’ value in equation (8-5) shows the amount by which the measured pressure Pd is to be corrected so that the real ground water pressure Pw is achieved. Now we will attend to the daily records of pneumatic piezometers J20 and J41, which are shown in Figs. (6-13) and (6-14) respectively. The piezometric level corresponding to piezometer J41 gradually decreased from May 1st to November 6th and then it started to increase steadily regardless of the changes of atmospheric pressure. In the case of piezometer J20, however, the records must be considered carefully because of the considerable involvement of rainfall. As it was explained previously, piezometer records during August and September may be evaluated regardless of the effects of rainfall. A study of the records of piezometer J20 during these two months and records of piezometer J41, as explained previously in section 8-2-2-2, reveals that there is no similarity between the records of changes of pressure and the records of these piezometers. This must not be attributed to the flexibility of the piezometers because, as it was discussed in section 8-1-2, any pore water pressure change due to an external load is recorded instantaneously by this type of piezometer. Neither can it be concluded that the changes of atmospheric pressure are not carried by ground water; this has already been shown by vibrating wire piezometers J871 and J877. An investigation of the mechanism of the measurement of ground water pressure using pneumatic piezometers reveals that in the particular case of piezometers J20 and J41, records which are virtually not affected by changes of atmospheric pressure could have been anticipated. The reason is that at the locations of these piezometers the Si value is about 100 per cent. This can be explained by using the equations (8-4) and (8-5). With reference to the Si values corresponding to piezometers J877 and J871 in table (8-2), the Si values at 5 and 8 meters below the ground level in Zone 4 are 0.95 and 0.94 respectively. Assuming S0 = 1, the maximum values with which the records of piezometer J20 and J41 should be corrected against the effects of atmospheric pressure (e values), are as follows: for J20: e = 1037 (1-0.95) - 985 (1-0.95) = 2.6 mb for J41: e = 1037 (1-0.94) - 985 (1- 0.94) = 3.1 mb It can be seen that maximum disturbance which may be caused by atmospheric pressure in the measurement of the ground water pressure using the pneumatic piezometers in Zone 4 at 5 and 8 m b.g.l. is of the order of 3 cm water head, which is not appreciable in comparison with the instrumental errors, and for this reason the effect of atmospheric pressure has not virtually manifested itself on the records of piezometers J20 and J41. It must be mentioned that if a pneumatic piezometer had been installed instead of piezometer J875, in zone 6, the maximum error which could have been caused by the changes of atmospheric pressure, in measurement of the ground water pressure, would be as large as e = 1037 (1-0.29) - 985 (1-0.29) = 37 mb This degree of error cannot be ignored, and reveals again the importance of the evaluation of the Si value, when the ground water pressure is proposed to be measured with high accuracy using expensive rigid piezometers. Ill - standpipe piezometers. In the case of the open standpipe piezometers the atmospheric pressure and its changes are applied instantaneously over the water table in the standpipe with a 100 per cent efficiency. In the surrounding soil, however, the effects of atmospheric pressure will generally be imposed by some decrease on the ground water pressure. This causes errors in the measurement of the ground water pressure. As the volume change of the piezometer is considerable, compressibility of the soil within the radius of influence of the piezometer (Premchitt and Brand, 1981), will be involved in the evaluation of the errors. If the compressibility of the soil is ignored, however, the errors may be estimated using the formulae given by Hvorslev, (1951). 8-2-3 Errors associated with time lag. Time lag and the corresponding errors in measurement of the ground water pressure are reviewed in section 2-2-4. It was shown that when the ground water pressure is constant or the soil is incompressible, the theoretical evaluation of errors associated with time lag is possible. In practice, however, equation of the ground water pressure is a very complicated function of different factors, (refer to 2-2-2 and 2-2-3), and therefore a rigorous solution is likely to be very difficult. For experimental evaluation of the range of errors in measurement of the ground water pressure, associated with time lag, records of different types of piezometers which have been installed in virtually the same vertical profiles will be compared as follows: Records of standpipe piezometer J45 and vibrating wire piezometer J879 which have been installed at the same depth as the trench invert level and midway between two adjacent trenches are shown in Fig. (6-11). The locations of these piezometers are shown in Figs. (5-17) and (5-18). For more elucidation, records of these piezometers, during October and November, corrected for the effects of atmospheric pressure^are also plotted in Fig. (6-18). Considering records of rainfall, Fig. (6-17), it can be seen that in August and September the amount of rainfall was low. Consequently, according to piezometer J45, the rate of change of ground water pressure at 4.2 m b.g.l. was as low as 3mm/day (8cm in one month) in September, Fig. (6-11). Finally, during the first three days of October, during which the atmospheric pressure was almost constant, records of piezometers J45 and J879 showed virtually parallel trends and the change of ground water pressure was not appreciable. With respect to the above, it can be concluded that in September and especially in the first few days of October the effects of time lag on the records of both piezometers would have been quite insignificant. In other words, the effect of time lag which could have disturbed the records of the stand pipe piezometer J45 may be ignored during this period. With reference to Fig. (6-18) it can be observed that during the first three days of October the records of piezometer J879 are about 22cm higher than that of piezometer J45. With respect to the explanations given above, this difference is due to the employment of different datums of measurements and not because of the greater flexibility of the standpipe piezometer. With reference to Figs. (6-17) and (6-18), it can be seen that there is a good correlation between the records of piezometer J879 and rainfall. Following the intensive rainfall from 5th to 10th October, the ground water pressure reached a maximum value of 1.12 m b.g.l. on October 11th, just one day after the intensive rainfall had stopped. The pressure then decreased gradually due to functioning of the drainage system until October 14th. Just one day after the second rainstorm started on October 14th, the second peak point of the ground water pressure occurred on October 15th. This time the pressure reached 1.05 m b.g.l, which is 0.07 m higher than the first time. When the third rain storm occurred on October 20th, the ground water pressure increased to a peak point of 1.0 m b.g.l. on October 21st and then started to decrease gradually because the rainfall was not appreciable. The standpipe piezometer J45, in contrast to the piezometer J879 which was described above, did not show a quick response to rainfall and as seen in Fig. (6-18) the piezometric level corresponding to this piezometer started to increase gradually from October 9th and reached a peak value of 1.55 m b.g.l. on October 24th, and then necessarily started to decrease on October 25th, because the ground water pressure, according to piezometer J879 had dropped to 1.67 m b.g.l (1.45 + 0.22 = 1.67) on that day. It is observed that if the ground water pressure had been measured by the standpipe piezometer J45, errors of the order of 0.9m, 0.83m and 0.40 meter head of water would have been involved on October 11th, 15th and 21st, respectively. With respect to the records of piezometer J875, Fig. (6-18), it can be observed that at 8 m b.g.l., however, the ground water pressure is not so sensitive to rainfall. This phenomenon is attributed to the consolidation and swelling characteristics of the soil and further elaboration will be given in section 8-5-5. Nevertheless, the point which is of interest in this section is that at such a depth the ground water pressure could have been measured satisfactorily even with a very flexible and cheap open stand pipe piezometer. So, it may be stated that in a drained area, in London Clay, with the drains spacing to depth ratio of the order of S/D = 3.4, the measurement of the ground water pressure midway between the trenches at depths equal to trench invert level and less is critical. Installation of standpipe piezometers should be avoided because of involvement of errors of the order of over 0.9 meter head of water. Installation of a rigid piezometer, e.g. vibrating wire, is necessary but not enough, because the fluctuations of the ground water pressure are very rapid. Accordingly, even with rigid piezometers, the weekly or monthly records of piezometers, which is usual in practice, will not show a realistic pattern of the changes of ground water pressure. The periods of measurements must be correlated to the features of rainfall. It is important to measure and include the records of piezometers within one day after a rainstorm has finished, because within one day after a rainstorm has finished the ground water pressure may reach its peak value. Below the trench invert level, however, the deeper the depth the slower is the response of the ground water pressure to precipitation, and in turn the less are the effects of time lag over the records of piezometers. At 8 m b.g.l. (about 4 m below trench invert) and deeper, the ground water pressure can be adequately measured by a simple open standpipe piezometer, however. In zone 4, the daily records of three different types of piezometers, which are installed at the same depths and positions with respect to trenches, are plotted. The trench spacing and depth to water level in a trench in this zone are shown in Fig. (5-10) and Fig. (6-62). Fig. (6-62) shows the records of piezometer JT21 which is installed in a trench, and it can be observed that the water level is about 0.9 m above invert level. The trench spacing to depth ratio at this location is about S/D = 15.9/4.2 = 3.8. Records and locations of piezometers J13, J20, J877, which have been installed midway between two adjacent trenches, are shown in Figs. (6-13), and (5-11), respectively. Records of these piezometers during two critical months, October and November, corrected for the effects of atmospheric pressure and are shown in Fig. (6- 19). With respect to this figure, it can be seen that at the end of the comparatively dry period and before the rainstorms started in October, i.e. the first three days of October, the piezometric levels according to piezometers J877 and J20 were 0.55 meters above and 0.33 meters below the piezometric level in standpipe piezometer J13 respectively. This shows that the phreatic surface is not parallel to the ground surface. This is probably due to the fact that the slope has a very uneven gradient at this place. We will, however, take the above mentioned piezometric levels as a datum for subsequent comparisons between the records of three types of piezometers for analysis of the errors associated with time lag in the measurement of ground water pressure. It can be observed that both piezometers J20 and J877 responded with a similar pattern to rainfall, and on October 22nd both piezometers showed steady piezometric levels for three days. The standpipe piezometer started to respond on October 9th, five days after piezometers J20 and J877 had responded. The maximum discrepancy between the records of piezometers J20, J877 and J13 occurred on October 11th, one day after the first rainstorm had finished. On that day, the piezometric level in standpipe piezometer J13 was 0.33 meters, (246-158-55 = 33), and 0.47 meters (246-232 + 33 = 47) lower than the piezometric levels in piezometers J879 and J20 respectively. Then the discrepancy decreased gradually and the piezometric level in standpipe piezometer J13 reached its peak value on October 24th, about two or three days after piezometers J20 and J877 had shown. The interesting point is that on October 24th the piezometric levels according to piezometers J877 and J20 had been about 0.1 meters below and 0.2 meters above the piezometric level in piezometer J13. This is probably due to a small rotation of the phreatric surface. It is seen that at this zone, in which the trench spacing to depth ratio (S/D) is greater than that of zone 6, and piezometric level in the trench is 0.9 m above invert, the standpipe piezometer J13 showed the ground water pressure more correctly than the standpipe piezometer J45 in zone 6. Now the records of three deeper piezometers J44, J41 and J781 are considered. Positions of these piezometers and their records are shown in Figs. (5-12) and (6-14) respectively. It is seen that there is good agreement between the records of piezometers J871 and J41. The small discrepancies between the records of these piezometers are obviously caused by the effects of changes of atmospheric pressure on the records of the vibrating wire piezometer J8 71. The records of pneumatic piezometer J41 are not disturbed by the changes of atmospheric pressure, because the Si value is about unity at 8 m b.g.l. The standpipe piezometer J44 after a long period of equalization showed the same pressure as the others in December. The results are in agreement with the records of piezometer J875 which is installed in zone 6, Fig. (6-11). As discussed previously it is observed that at both zones and at 8 m b.g.l. the ground water pressure changes very steadily in comparison with the sudden and considerable fluctuations of the ground water pressure at trench invert level. Consequently it may again be stated that at about 4 meters below the trench invert level precise measurement of the ground water pressure is possible with a cheap and simple standpipe piezometer. To evaluate the range of errors in measurement of ground water pressure in the vicinity of drainage trenches, the daily records of piezometers J27 and J874 are considered. The records of these piezometers and their locations are shown in Fig. (6- 15), and Figs. (5-10) & (5-13) respectively. These piezometers have been installed at the vicinity of the interface of the trench and clay and about 1.80 meters apart in longitudinal profile. Although it had been anticipated that both piezometers would show more or less the same piezometric levels, the vibrating wire piezometer recorded higher pressures than the piezometer J27. This, obviously, cannot be attributed to the higher volume change of piezometer J27, because, the retarding effect of volume change of a piezometer manifests itself when the ground water pressure changes suddenly. At this location, however, even in September during which the piezometric level was virtually constant, the piezometric level according to piezometer J874 was about 0.85 meters higher than the piezometric level in piezometer J27. It seems that this difference in piezometric levels is partially due to the fact that the phreatic surface at this place is not parallel to the ground surface. A comparison between the records of piezometers J27 and J874, and records of piezometers J13 and J877, Fig.(6-13), respectively reveals that the fluctuations of the ground water pressure midway between trenches is much higher than its fluctuations near the trenches. For example, while the piezometric level in piezometer J13 increased 1.1 meters in October, piezometer J27 showed only 0.45 meters change in piezometric level in the same month. This suggests that the focus of errors in measurement of the ground water pressure associated with time lag is midway between two adjacent trenches. Now, we will evaluate the effects of time lag in monitoring of the ground water pressure in zone 2. This is the location in which the drainage trenches are as close as about 6 meters, Fig. (5-19), and the trenches spacing to depth ratio is as low as about S/D =1.1. With reference to the records of piezometers P503, P504 and P505, Figs. (6-49), (6-50) and (6-51), it can be seen that the piezometric level between two trenches is very flat. Even the intensive rainfall in October did not change the piezometric level appreciably . This suggests that even in drained areas there are some cases in which the ground water pressure can be monitored adequately with simple standpipe piezometers. For further elaboration, however, the daily records of piezometers J873 and P505, Fig. (6-16), are studied. It is seen that in May, June and July both piezometers, considering the effects of atmospheric pressure on the records of piezometer J873, showed the same trend. During the following months, however, the piezometer J873 started to show an increasing pattern. This cannot be attributed either to atmospheric pressure or time lag, because, the increasing pattern was continuously carried out, regardless of rainfall and atmospheric pressure, until the end of November, when the measurement was terminated This is most probably due to the diffusion of air into the cavity of the piezometer, because the ground water pressure is very low and the piezometer is operating in a pressure which is nearly atmospheric (Vaughan, 1974). The records of piezometers in October are of special importance. It is seen that at this location the rainstorms have not caused the piezometric level to change as quickly and as much as in zones 4 and 6, and the differences between the records of the vibrating wire piezometer J873, considering the effects of atmospheric pressure, and the standpipe piezometer P505, are not considerable. 8-2-4 Conclusions 8-2-4-1 Errors associated with atmospheric pressure. Atmospheric pressure is an important source of error in monitoring of ground water pressure. The magnitude of errors depends on the range of the changes in atmospheric pressure, the efficiency with which these changes are transmitted through the soil (si value) to the ground water pressure, and the characteristics of the pressure measuring device. In the case of vibrating wire piezometers, if the effects of atmospheric pressure over the records of a piezometer are completely ignored, a maximum error of the order of the maximum change of the atmospheric pressure in the area under investigation may be involved. In the Guildford area the maximum change of atmospheric pressure during the last two years (1986-1987), was 51 mb (51 cm head of water). If the readings of a piezometer are corrected simply by the change in atmospheric pressure when the pre-installation base reading (N0) is recorded and when the ground water pressure is measured (Nj) (as recommended in the users’ manuals), considerable errors may still be involved. Accordingly, if either the effects of atmospheric pressure are completely ignored or corrected as described above, the errors will be many times greater than the small instrumental errors which are mentioned by manufacturers. In this way the advantage of such a very expensive and sophisticated device against the very cheap and simple standpipe piezometer is under question. To achieve satisfactory results with a vibrating wire piezometer, however, it is recommended that the efficiency of transmission of the changes of atmospheric pressure through the soil, the £\ value, be estimated. Then, the piezometer records can be corrected with respect to the Si value. Methods of estimation of Si value and correction of records of piezometers are discussed in section 8-2-2-3. In the case of pneumatic piezometers, if the effects of atmospheric pressure on the records of piezometers are completely ignored, particularly when the efficiency of transmission of the changes of atmospheric pressure (si) is 100 per cent, satisfactory pressures, which are representative of the ground water pressure ( P w )j will be measured directly. No error, associated with the changes of atmospheric pressure will be involved. In general, however, when the £1 value is less than unity, the ground water pressure will be under-estimated, the less the 5i value the larger the under estimation. For example, for £1 = 0.29 the error will be as large as e = 37 mb. This much error, of course, cannot be ignored, and reveals again the importance of estimation of the 5i value, when the ground water pressure is proposed to be measured accurately using expensive rigid pneumatic piezometers. With the standpipe piezometers, quantifying the range of errors associated with the effects of atmospheric pressure is more complicated, because of the involvement of a considerable hydrostatic time lag. Therefore further research is necessary. 8-2-4-2 Errors associated with time lag. If all other sources of errors are ignored the pneumatic and vibrating wire piezometers with maximum nominal volume changes of 0.03 and 0.005 cm3, respectively, are able to record any change of ground water pressure virtually without hydrostatic time lag from the practical view point. In spite of the fact that the standpipe piezometers are cheaper and simpler in comparison with the others, in some cases, however, considerable errors are involved in monitoring of the ground water pressure with these piezometers. One of these cases was encountered in monitoring of the ground water pressure in Stag Hill, a slope in London clay which was stabilized with the construction of drainage trenches. It was observed that the ground in the drained area, midway between two adjacent trenches, and at depths equal to, or less than trench invert level, formed the focus of the errors, which are involved in measurement of the ground water pressure with stand pipe piezometers. The magnitude of the error (discrepancy of the measured pressure from the real ground water pressure) at any time is a function of performance of piezometer installation, meterological conditions, especially duration and intensity of rainfall, and in turn infiltration, engineering properties of soil, and finally details and performance of drainage trenches. The following notation is used: S = spacing of trench drain D = trench depth ho = piezometric level at invert before drainage ha = piezometric level in the trench For S/D = 3.40 & S/h0 = 4.2, installation of standpipe piezometers should be avoided because of the involvement of significant errors, say 0.9 meter head of water. Installation of rigid piezometers, e.g. vibrating wire, is necessary but not enough, because the fluctuations of ground water pressure are very rapid, and may rise over 1.2 meters in a couple of days. Accordingly, even with rigid piezometers the weekly or monthly readings of piezometers, which is usual in practice, will not show the realistic pattern of the changes of ground water pressure, unless the reading times are correlated to the meterological conditions. For S/D = 15.9/4.2 = 3.8, S/ho = 5.1 and ha = 0.9m, with an equivalent S /D = 15.9/3.3 = 4.80, an average maximum error in the order of 0.4 meter head of water is involved with a standpipe piezometer. Then, in a short time, say two or three days, depending on the characteristics of the meterological condition (i.e. duration and amount of rainfall), a standpipe piezometer may tend to show the realistic ground water pressure, however. The reason is that the discharge capacity of the drainage system is not efficient enough. Pore water pressure monitoring, using rigid piezometers is recommended, however. When the trench spacing to depth ratio is as low as about S/D = 5.6/5 = 1.1, and S/ho = 1*3, the efficiency of the drains in lowering of the ground water pressure is practically independent of the meterological condition. Accordingly, the ground water pressure can be adequately monitored with a simple and cheap standpipe piezometer at any depth and position with respect to the trenches. For S/D = 1.1, 3.4 and 4.8, and most probably for all values of S/D, the ground water pressure from 4 meters (and probably from 2 meters) below invert level to any depth can be adequately monitored using simple stand pipe piezometers, however. 8-3 ASSESSMENT OF THE GROUND WATER PRESSURE PATTERN. 8-3-1 Summary of borehole logs, and ground water pressure profiles. The aim of this section is to evaluate and interpret the ground water pressure pattern with respect to the borehole logs, ground water pressure profiles, and permeability values. The available borehole logs, from 1964 to 1986, are appropriately summarized in table (8-4) from the reports Nos F.69/882, 1964 and F69/371, 1965,by Foundation Engineering Limited. The locations of boreholes Nos. 1 to 10 and boreholes Nos. 101 to 106 are shown in Figs. (3-4) and (3-7), respectively. The borehole logs for 59 new piezometers installed during the research are given in section 5-4-3. To visualize the features of distributions of the ground water pressure at different locations, profiles of distribution of ground water pressure are plotted with respect to the piezometer records given in Figs. (6-20) to (6-78). These profiles are shown in Fig. (8-3). It should be pointed out that the measured permeability may be quite different from the true permeability because of the influence of a remoulded layer of clay due to drilling surrounding the piezometer tip, also if a high permeability layer is thin then it is possible that the point of measurement of pore water pressure may not coincide with the position of the thin layer and its presence would not then be reflected by the measured value. Table (8-4) Summary of bore hole logs 1964, 1965. Bore Depth Notes and special events - remarks, hole No. m b.g.l. 1 12.0 Water entered bore hole at 7.6 m b.g.l. on 25-6-1964, but was later found to be 5.0 m b.g.l. on 26.8.1964. 2 12.5 Borehole diy 3 15.5 Water entered borehole at 5.0 m b.g.l. Stiff dark brown clay with patches of light blue silt and a 10 cm layer of claystone at 5.0 m b.g.l. Table (8-4) Summary of bore hole logs 1964, 1965 (cont’d). Bore Depth Notes and special events - remarks, hole No. m b.g.l. 4 18.3 Borehole diy 5 15.2 Water entered borehole at 5.2 m b.g.l. firm mottled brown clay with traces of sand from 1.2 to 6.0 m b.g.l. 6 18.3 Water struck at 14.9 m b.g.l. Stiffblue clay from 12.0 to 18.3 m b.g.l. 7 18.3 Indicator installed at 3.3 and 12.8 m b.g.l., and water levels were measured to be 2.5 m b.g.l. in both cases just one day after boring, (Indicator type not mentioned). 8 18.3 Indicator installed at 3.3 and 12.8 m b.g.l., and water levels were measured to be 0.9 and 10.2 m b.g.l. respectively,(Indicator type not mentioned). 9 18.3 Borehole dry. 10 10.4 Water level 6.0 m b.g.l. Compact fine green clayey sand from 6.0 to 9.7 m b.g.l. Stiff to hard putty chalk appeared at 9.7 m b.g.l. (Borehole is close to the railway station, Fig.(3-4).) 101 18.7 Water entered borehole overnight at 4.6 m b.g.l. and rose to 1.8 m b.g.l. Further seepage was observed at 6.0 m b.g.l. 102 18.7 Water entered borehole at 4.6 and 7.6 to 9.1 m b.g.l. Subsequently rose to 1.2 m b.g.l. Table (8-4) Summary of bore hole logs 1964, 1965 (cont’d). Bore Depth Notes and special events - remarks, hole No. m b.g.l. 103 18.7 Water entered borehole at 6.0 m b.g.l., and subsequently rose to 3.0 m b.g.l. 104 18.7 Water entered borehole while standing open overnight at a depth of 9.6 m b.g.l. and rose to 2.4 m b.g.l. Subsequently with borehole open and unlined at depth of 18.7 m b.g.l. water remained overnight at 7.2 m b.g.l. 105 18.7 Water entered borehole overnight at depth of 7.5 m b.g.l. and rose to 4.5 m b.g.l. Borehole left unlined and dry at 16.0 m b.g.l., water entered overnight and rose to 3.0 m b.g.l. 106 18.7 Water entered borehole below 9.3 m b.g.l. and rose to 7.5 m b.g.l. 602 17 Water level found at 3.2 m b.g.l. Large flow of water below modstone at 7 m b.g.l. 8-3-2 Discussion With respect to the borehole logs, it is observed that in 22 out of 76 cases water seeped or entered rapidly into the boreholes while they were being drilled, or immediately after bore holes had been completed. It is very interesting to note that in most cases the water level rose quickly in the boreholes. For example, in the case of borehole 7 the water level rose at least 10.3 meters only in one day after boring. In the case of boreholes J9 and J870, water flushed into the boreholes and rose instantaneously to 3.0 and 4.5 m b.g.l., while the holes were being drilled. This suggests that there are very permeable water bearing layers possessing artesian potential in some cases, because water can never flow from the ambient clay into the borehole so quickly. Although the borehole logs are ample evidence for the above statement, further evidence is available from the in situ permeability tests. Referring to Tables (6- 3) and (6-4) it is seen that permeability values in the order of 10-5 and 10-6 cm/sec, indicating the presence of layers of silty or clayey fine sands, are frequent. To evaluate the ground water pressure and flow regime in these layers and in turn, their effects on the pore pressure pattern of the area, the stability of slope, and the location of slip surfaces, profiles of ground water pressure at 16 locations are plotted and shown in Fig. (8-3). In this figure, the broken lines show the conjectural patterns and the arrows present the direction of flow. These profiles are discussed and interpreted as follows: - Profiles 1 and 2. Profiles 1 and 2 are representative of a hydrostatic condition. In profile 2, the piezometric level at 8 m b.g.l. and above shows some fluctuation, however. The good agreement between the records of piezometers J48 and J49, Fig. (6-75), and records of rainfall, Fig. (6-17), reveals that these fluctuations are rain controlled, and in summer tend towards hydrostatic. The major point in this profile, however, is that the pressure is hydrostatic having layers with permeabilities as high as about 10-5 cm/sec. This may suggest that the water bearing layers at this location (a flat area) are confined. - Profiles 3 to 11. Profiles 3 to 11 generally represent a ’classical’ under-drained pore pressure profile. The magnitude of diversion from hydrostatic is a function of the stress dependent permeability characteristics of the soil (Bromhead and Vaughan, 1984), and the discharge capacity and pore pressure in under draining materials. In profiles 6 and 8 it is seen that ground water pressure profiles have sharply decreased at a certain depth below ground level and reduced from hydrostatic to a smoother loop. The profile 9 is more interesting. The ground water pressure at only 4.2 m b.g.l. is nearly atmospheric and does not fluctuate. This suggests that the layer at that depth is operating as a very efficient under draining blanket. With reference to the records of piezometer No.204 in Fig. (6-80), it is seen that such a profile would have existed for a long time before installation of the drainage system and therefore has not been initiated by the drainage. - Profiles 12 to 16 As the changes in ground water pressure are usually attributed to the rainfall and infiltration, so it is expected that seasonal changes of ground water pressure in compressible soils decrease with depth. This, of course, can be justified with respect to the swelling and consolidation characteristics of clay and there are cases in the literature confirming this point (e.g. Lau and Kenney, 1984). As extensively explained in section 8-5, in general cases the change of ground water pressure is infiltration- dependent and strain-dependent. The pore pressure change that is caused directly by infiltration is a function of the rainfall period and intensity, the geometry of slope, the diffusivity, porosity and degree of saturation of the soil, and finally the presence of natural or man-made drainage systems, (partially from Lumb, 1962). Pore water pressure, in a compressible soil, starts to change from the soil and infiltration zone interface. Therefore, in a clay ground, the shallower the depth the higher the changes of ground water pressure due to rainfall periods. In Stag Hill, however, some special cases are encountered in which the amplitude of fluctuations of the ground water pressure in some depths is considerably greater than that of the others, so that in some cases the direction of flow is changed. In the following part of this section, these special cases are discussed. Profile 12 shows the ground water pressure distribution at the toe of slope along profile 5-5. The pressure at 2.5 m b.g.l. (piezometer J3) is constant and atmospheric throughout the year, and this is the depth of fill soil. At 5.5 m b.g.l., however, fluctuations of ground water pressure are very considerable. In summer, Fig. (6-70), the pressure decreases to such an extent that the flow is directed towards this level (5.5 m b.g.l.) from underneath and above. When rainfall is intensified, the pore pressure at 5.5 m b.g.l., (in piezometer J2) quickly increases and a downward flow is established again. With respect to the explanation given above, this type of profile is a characteristic of soils having a very efficient drainage blanket at 2.5 m b.g.l., and also at 5.5 m b.g.l., but with less efficiency. Profile 13 is another example of an under drained pore pressure profile. The interesting point, however, is that fluctuation of the ground water pressure at 9 m b.g.l. (piezometer J4) is about 6 times that of 3 m b.g.l. (piezometers J39 and J7). This clearly suggests that there is an unconfined permeable layer at about 9 to 10 m b.g.l. which operates as a drainage system. A comparison between the records of rainfall, Fig.(6-17), and records of piezometer J4, Fig. (6-71), reveals that in August and September 1987 during which rainfall was at a minimum, the pore pressure has decreased to a minimum value. In October, however, immediately after the rainstorms started, the piezometric level in piezometer J4 suddenly increased from 6.7 to 5.8 m b.g.l. In the shallower piezometer J6, however, the pressure increased more slowly. This suggests that the above mentioned permeable layer most likely out crops somewhere so that the rainfall directly infiltrates into it. Profile 14 shows the distribution of ground water pressure at a point about 60 meters up the slope from profile 13. Here, with reference to the records of piezometers J8 and J9 in Fig. (6-70), it can be observed that in summer water in the permeable layer ,at 6 m b.g.l.,drains away quicker and pressure decreases, so that the direction of flow turns upwards from 8 m b.g.l. to 6 m b.g.l. When rainfall starts, however, the pore pressure increases and flow is directed downwards all the way from ground surface to 8 m b.g.l. Profile 15 is another interesting figure showing the distribution of the ground water pressure at the toe of the hill about 30 meters to the west of the longitudinal profile 5-5. As noted in the borehole logs, section 5-4-3, water entered into the borehole J10 from the base (3 m b.g.l.). Now, in profile 15 it is seen that the permeable layer at 3 m b.g.l. operates continuously as a drainage blanket,draining the water from the soil above and below. With respect to the discussion given already and ground water pressure profiles Nos. 12,13, 14 and 15, a natural under drainage blanket is approximately located along profile 5-5 and shown in Fig. (8-4). Finally, profile 16 is another example in which the pore pressure profile diverts considerably from the hydrostatic, indicating the presence of an under drainage condition. An important point which should be mentioned here is that chalk is known as the classical under draining layer in London clay. Profiles similar to some of the profiles given in Fig. (8-3) are encountered in the literature, (e.g. Butler, 1972). Nevertheless, the under draining pattern which is discussed in this section should not be attributed to the presence of chalk, because as explained in section 3-3, there are several boreholes which were sunk as deep as 21 and 45 m b.g.l. in the sloped and flat areas, respectively. None of these boreholes, which are much deeper than the under draining layers, penetrated into the chalk, however. To evaluate the effect of the ground water pressure pattern, discussed already, on the failure of the slope, we refer to the longitudinal profile A-A in figures (7-4) and (7-5). The slip surface 'Was located by Skempton some 23 years ago, (Skempton and Petley, 1967a). At that time there was no idea about the features of the ground water pressure profiles along the above mentioned longitudinal profile. Now, profiles Nos. 4, (piezometers 105, 106 , 107) and 8 (piezometers J20, J41, J42 and J40) are established and shown in figures (7-4) and (7-5), respectively. It is clearly seen that these profiles and the slip surface are in very good agreement. The slip surface has passed exactly through the points at which the pore pressure is the maximum. This reveals that the feature of failure of the slope is a function of the natural drainage pattern. Finally, such a complicated pore water pressure pattern provides major difficulties for the site instrumentation for monitoring of the ground water pressure. The installation of hydraulic twin tube piezometers is generally rejected , because their installation in ground with the possibility of piezometric levels lower than 6 m b.g.l. is not recommended, and may lead to a waste of money. In the case of vibrating wire and pneumatic piezometers, there are some locations under the ground surface in which the pore water pressure is nearly atmospheric (about 14 m b.g.l. in profile 6, 4.2 m b.g.l. in profile 9, about 13 m b.g.l. in profile 10, and etc.). If such p iezometers are inadvertently installed at such a depth, there will be the risk of diffusion of air into the transducer, (Vaughan, 1974). Piezometers J872 and J873, Figs. (6-66) and (6-59),respectively, are such cases. On the other hand, although the stand pipe piezometers do not suffer the above mentioned difficulties, their long first equalization time lag and slow response, and especially the considerable errors which may be involved if they are installed adjacent to the higher permeability layers having large and rapid fluctuations of pore water pressure, are the shortcomings of such piezometers in such areas. In profile 10, with reference to Fig. (6-67), it is seen that the piezometric level in vibrating wire piezometer J875, 8 m b.g.l., has continuously and steadily increased since installation and tended towards a constant value of 4.2 m b.g.l., in June 1988. In piezometer J878, 11.8 m b.g.l., the piezometric level, since installation time, has continuously increased steadily and tended towards a constant value in August. Although neither the records of piezometer J875 nor J8 78 are in agreement with the usual seasonal changes of ground water pressure, they do not appear to be due to the instrumental fault or diffusion of air into the piezometers, because the piezometers were deliberately tested and calibrated before installation,and pore pressure in J8 75 is well above atmospheric. In piezometer J8 78, although the pore pressure is not much above atmospheric, records of stand pipe piezometer J51, Fig. (6-68), installed only about 8 meters to the east of J878, shows the presence of an increasing pore pressure pattern at 11.8 m b.g.l., at that place. 8-3-3 Conclusions: The ground water pressure pattern in Stag Hill is governed by layers of higher permeability values of the order of 10'5 to 10-6 cm/sec, compared with 10-8 - 10_1° cm/sec for the London clay. These layers act as the under draining blankets with different discharge capacities. It is possible that the presence of layers of high permeability soils at Stag Hill is due to the site being located at the southern edge of the London clay basin and these layers have been deposited by water flowing intermittently down the adjacent chalk slope. The complicated ground water pressure pattern which is caused by the above mentioned factor, partially controls the stability of the slope and introduces difficulties with the instrumentation of the slope to monitor the ground water pressure. 8-4 EVALUATION OF THE STABILITY OF SLOPE. 8-4-1 Discussion. Results of analyses of the stability of slope, carried out in Chapter Seven, are summarized in the following table and then discussed. The permissible safty factor values, corresponding to this area, is previously discussed in section (1-5). Table (8-5) Summary of slope stability analyses. Description F . S. F . S. profile profile 6-6 A-A 1 F.S. at the time of failure 1.01 0.96 2 F.S. after failure, before drainage and any development of area, 1965 1.21 1.23 3 F.S. after drainage, developing and paving of area, 1988 - 1.58 4. F.S. after drainage, if the area was not paved - 1.35 5 F.S. after drainage and changing of slope shape, 1988 1.33 - 6 F.S. after drainage, if the slope shape had not been changed (disturbed) 1.42 - 7, F.S. in the case of blocking of drains 1.14 1.45 Note: How the F.S. values of 1.35 and 1.45 are achieved is explained later in this section. Although the stability of slope at the time of failure has been evaluated on the basis of some assumptions and inference of the ground surface profile using the present topography of the area, the computed safety factor values confirm that the assumptions and inference of the ground profiles must have been reasonably good. After failure, in 1965, although the computed factors of safety are above unity (i.e. 1.21 & 1.23), they are not satisfactory, because the allowable factor of safety for the case encountered here is suggested to be F.S. = 1.25 according to the discussion given in section (1-5). Accordingly, the improvement of the stability of the slope must have been necessary at that time. At the present time (1988), profile A-A is quite stable having a safety factor of F.S. = 1.58, and as explained in section 7-3-3, such a high stability is due to paving of the area, and drainage. The efficiency of paving will be estimated later in this section. The safety of profile 6-6, F.S. = 1.33, is below the permissible value (F.S. = 1.40) and of course, if the shape of slope had not been disturbed (refer to section 7-2-3) the drainage system would have increased the factor of sdfety to F.S. = 1.42, which would be acceptable. At the present time, however, the piles around profile 6-6 are nominally supporting some lateral load. The word ’’nominal” is used because the piles are covered with plastic sleeves to the depth of the slip surface, and there is no soil-pile friction in the failure zone. Therefore, as long as F.S.^ 1.25 (refer to section 1-5) piles will not be subject to significant lateral loading. If the drains block, the factor of safety of slope along profile 6-6 will decrease to 1.14, which is even less than the stability of slope at the time before development of the area (1965). This is again due to the change of the slope shape along the most critical profile. In this case, the ambient piles will seriously go under lateral stress. Nevertheless, the grassed area itself, will slide with a safety factor of 1.14, because there is no pile. Therefore, the slope will endanger the safety of the University and new corrective measures will have to be installed. In the case of profile A-A, the drains are not operating properly and probably a siltation phenomenon has started, because with reference to Fig. (6-62) it is seen that the piezometric level in a trench is about 0.9 meters above the invert level. Nevertheless, the interesting point is that even if the drains are fully blocked the slope will still be stable due to the efficiency of the surface coverage. To estimate the efficiency value of the surface coverage of the area against infiltration and in turn on the improvement of stability of profile A-A, profiles A-A and 6-6 are compared. With reference to table (7-1) it is seen that both profiles had approximately the same factors of safety in 1965. After drainage, if the area along profile A-A had not been paved the factor of safety would have been somewhat less than 1.42, because the drains along profile A-A (drainage system phase I, Part 2) are not operating as effectively as the drains along profile 6-6 (drainage system phase I, part 1). If, however, the piezometric level in the trenches along profile 6-6 was 0.9 meters above the invert level, the factor of safety would be about F.S. = 1.35. Accepting this figure as the factor of safety of profile A-A if the area had not been paved, then the efficiencies of the drainage systems, paving, and disturbance of the shape of slope along profile (6-6), on the stability of slope can be evaluated as follows: For profile 6-6: Total improvement of safety (1988) = ^ 21 ^ * 100 = 10%. Efficiency of drainage system (1988) = x 100 = 17% Effect of change of the slope shape = 10 - 17 = -7%. For profile A-A: 1 50 _ I 90 Total improvement of safety (1988) = —1 x 100 = 28% Efficiency of drainage system (1988) = ^ ^123 ^ x = Efficiency of paving (1988) = 28 - 10 = 18% Factor of Safety if the drains block = 1.23 x 1.18 = 1.45 As it is seen, the surfacing is so effectively towering the infiltration that even if the drains are blocked the slope will still be quite stable along profile A-A. 8-4-2 Conclusions: The safety of slope along profile 6-6 is below the permissible level, and the piles are nominally under some lateral stress. Therefore, any further disturbance of the ground surface, any rainstorm heavier than of the October 1987 (268m/month) and especially pore pressure build up in the trenches will seriously endanger the safety of the University. Accordingly it is crucial to monitor continuously the discharge efficiency of the trenches, using piezometers Nos. JT15, JT16 and JT23. The slope is quite safe along profile A-A. The safety has been improved not only by the drainage system but by the paving of the area. The interesting point is that the efficiency of paving in improvement of the safety is the dominant factor, so that even if the drains are completely blocked the slope will still be stable. With respect to the above, the surface covering of slopes appears to be an important factor to be taken into account in economic estimation and technical design of remedial measures. 8-5 EVALUATION OF THE MECHANISM AND EFFICIENCY OF THE DRAINAGE SYSTEM IN LOWERING OF THE GROUNDWATER PRESSURE 8-5-1 Introduction Two types of efficiencies may be defined with regard to a drainage system as follows: a - Efficiency of a drainage system on the improvement of the safety of a slope, which is a combination of an improvement of the safety factor due to lowering of the ground water pressure and due to decreasing the seepage force by changing the direction of flow. b - Efficiency of drainage on the lowering of the ground water pressure only. In this section, however, only the mechanism, and efficiencies of the drainage trenches on the lowering of the ground water pressure, will be evaluated. The evaluation of efficiency of the drainage system will be carried out with respect to the key diagram shown in Fig. (1-4). Definitions of the symbols are the same as given in Hutchinson’s method in Section 1-4-3-2. For simplicity, this figure is repeated at the end of this chapter. In the following sections the efficiency of drainage trenches and mechanisms of their operation will be evaluated and discussed. Then the results of efficiency values will be summarized in table (8-6). Finally, the site data, summarized in table (8-6), will be compared with the theoretical solutions reviewed in section 1-4-3-2. 8-5-2 Drainage System phase I, part 1. To study the mechanism of operation and efficiency of the drainage trenches part 1, phase I, Fig. (1-1), two cross sections Ki - K2 and K7 - were located and piezometers were installed to establish the ground water pressure regime between adjacent trenches. Both cross sections, Ki - K2 and K7 - Ks, are across the longitudinal profile 6-6, in Zones 7 and 6, respectively, Fig.(5-2). The evaluation of the efficiency of the drainage trenches in Zone 7, Fig. (5-16), may not lead to a realistic result because of the domination of a special ground water pressure pattern at this location, Fig. (5-16) and Fig. (8-3) profile 9. In Zone 6, however, the evaluation of the efficiency of the drains can be satisfactorily carried out. With respect to the records of piezometers Nos. 202 and 206, Figs. (6-28) and (6-31), it can be seen that the ground water table was 0.6 - 0.7 m b.g.l. at the location of these piezometers, at the time before installation of trenches (1965). The ground water table at the location of the section K7 - K%, however, which is located some distance uphill from piezometers 202 and 206, must have been a little deeper, say 0.8 m b.g.l. at that time. After drainage, with respect to Fig. (5-18) and records of piezometer J879 in Fig. (6-79), it can be observed that the piezometric level at the trench invert level, midway between trenches, reached 1.0 m b.g.l. in October, 1987. This suggests that the drainage system in this zone, detailed in table (8-6), Row No.2, has not had a significant efficiency on the lowering of the ground water pressure at the trench invert level. The efficiency at invert level is: h = 0.8 hm = 0.8 x 3.15 = 2.52 m 0.1 x 0.8 + 2.5 x 14 hav = ------14-g------= 2.39 m 5.*? -2 .5 2 A#ir r| = ------0.25 ; 25% 3 35 - 2 39 %v = 3 35 = 0.29 ; 29% In connection with the mechanism with which the drainage trenches have improved the ground water pressure profile, we refer to the records of piezometers 202, J43 and J879, Fig.(6-79), Zone 6. With reference to the records of piezometer 202 it is observed that the phreatic surface before drainage was about 0.6 m b.g.l, and after installation of drains the piezometric level, at 6 m b.g.l. decreased, and it varied between 1.9 and 2.5m b.g.l. in 1966 and 1967. Piezometer J43 which was installed in 1987, at the same depth as piezometer 202, showed the same pattern as that of piezometer 202, some 22 years later. This suggests that, just at about 2 meters below invert level, the critical piezometric level has been decreased from about 0.6 m b.g.l., at the time before drainage, to about 2 m b.g.l., after drainage. It is important to emphasise that this pattern is seen to be irreversable, and, in contrast to the piezometric level at invert level, it never increases to nearly the same level as before drainage. At invert level, however, piezometric level midway between trenches does increase nearly to the same level as before drainage. For further elucidation, it is interesting to compare the records of rainfall, Fig. (6-17), and records of piezometers J879 and J43, Fig. (6-79). It can be seen that in September 1987, the rainfall was not appreciable, and the piezometric levels in piezometers J879 and J43 decreased to 2.3 and 2.5 m b.g.l., respectively. In October, following a rainstorm, while the ground water pressure at trench invert level (J879) suddenly increased to 1 m b.g.l. (1.3m increase), the piezometric level 2m below invert level (J43) did not show any quick response to the rainstorm. It must be mentioned that this phenomenon is not due to the higher flexibility of the standpipe piezometer J43, because the standpipe piezometer J45, which is installed at the same depth and location as vibrating wire piezometer J879, showed a sudden response to rainfall also, Fig. (6-69). This clearly indicates that soil between ground surface and trench invert level can swell and go under strain much easier and quicker than the soil below the invert level and vice versa. A survey of the piezometer records of the whole area shows that such a rapid change of ground water pressure, e.g. 1.3 meters in a few days, are shown only by piezometers which are installed at/or above trench invert levels. This phenomenon will be interpreted in section 8-5-5 after the two other cases are reviewed in sections 8-5-3 and 8-5-4. Now, we return to the records of piezometer J43 to see how efficient the drainage system is in lowering the ground water pressure at a point just 2 meters below invert level. It is assumed that at that depth the piezometric level between trenches is slightly curved, say fs = 0.95, and if the effects of trenches on the ground water pressure are ignored, then with reference to Fig. (5-18) and table (8-6): ho = 6 - 0.6 = 5.4, measured from piezometer intake level. h = (5.4 - 1.4) x 0.95 = 3.80 _ _ 5.4-3.80 = Q 3Q Qr 3Q% It can be observed that the efficiency of the drains, at 2 meters below invert level, is approximately the same as at invert level, or probably slightly higher, a long time after the trenches have been installed. 8-5-3 Drainage system phase I, part 2. As a criterion to evaluate the effect of the drainage system Phase I, part 2, Fig. (5-2), some piezometer records in Zone 4, along profile A-A are available. It should be mentioned that at this zone the efficiency of drains are affected by the effects of the paving, explained in section (8-4). The ground water pressure profiles, before and after drainage, midway between trenches are shown in Fig. (8-5). Locations of piezometers J20, J40, J41 and J42, installed in Zone 4, are shown in Fig. (5-12). These piezometers have been installed virtually along the same vertical profile as piezometers 103 and 104. Records of these piezometers are all shown in Fig. (6-78). With reference to Fig. (8-5), and records of piezometer 103 in Fig. (6-78), it can be observed that fluctuation of the ground water pressure, before drainage and at about invert level, was only 0.3 meters. After drainage, although the maximum piezometric level decreased only 0.1 meter, the ground water pressure fluctuated about 1.4 meters. In Fig. (6-78), a comparison between the records of piezometers J20 and 103 reveals that not only has amplitude of fluctuations of ground water pressure at invert level increased, but the period of fluctuations also has been considerably shortened due to drainage. This suggests that the drainage system significantly shortens the stress adjustment time lag of the whole mass of soil from ground surface to invert level. For this reason, flexible piezometers (e.g. Standpipes) fail to follow the changes of ground water pressure, and large errors are involved. As mentioned previously, in section 8-5-2, the reason for this behaviour will be discussed in section 8-5-5, after all the case records are reviewed. Now, with respect to profile (3) in Fig. (8-5), it can be inferred that, the piezometric level at trench invert responded quickly to precipitation, but, the duration of infiltrations were not long enough to increase the ground water pressure at 8 m b.g.l. to the same level as at invert level. It is interesting to note that the ground water pressure at 8 m b.g.l. has been decreased more than at invert level midway between trenches. This suggests that, where dry periods are dominant, the efficiency of a drainage system below invert level increases gradually and may overtake the efficiency value at invert level. To evaluate the efficiency of the drainage trenches in lowering of the ground water pressure, with reference to the records of piezometers installed in the trench, Fig. (6-62), it can be seen that the piezometric level in the trench is h^ = 0.9 meters above invert level. At invert level, from Fig.(5-8), table (8-6) and Fig. (1-4): ho = 4.2- 1.1 = 3.1m hm = 3.1 - 0.1 = 3.0 m h = 0.8 x 3.0 = 2.4 m 0.9 x 1 + 2.4 x 15.9 hav= ------= 2.31m - 3.10-2.4 AAft/ r| = j-jq = 0.22 ; 22% 3 1-231 T)av = 3 9 = 0.25; 25% at 8 m b.g.l., assuming that fs « 1.0 h = hav = 5.2 m ho = 8-1.1 = 6.9 m, h0 measured from the piezometer J41 intake level. Tj = 6' 96' 95' 2 ■" 0.25 ; 25% It can be observed that the efficiency of the remedial measures (drainage and paving) at 8 m b.g.l., (on the slip surface), is approximately the same as at invert level. As explained previously, in section 8-5-2, this suggests that the efficiency of the remedial measure (drainage plus paving) has now taken a significant effect at 8 m b.g.l. some 22 years after drainage. As discussed in section (8-4), it should be remembered that the above values of efficiencies are not due only to the drainage, but also to paving and adjacent multi-storey buildings, which decrease the infiltration. Accordingly, the efficiency of the drainage system itself is less than the above values. 8-5-4 Drainage system phase V (Wates House) Locations of the maximum and minimum piezometric levels, between two adjacent trenches are shown in Fig. (5-19). It can be observed that the fluctuations of ground water pressure here are no longer as large as the two previously mentioned cases. The maximum change of ground water table midway between trenches, following the intensive rainstorms which occurred in October 1987, (264 mm) was only 0.35 meters. Assuming that the water table was about 0.8 m b.g.l. before drainage, the efficiency of drainage system at invert level will be as follows: hm = 0-45 m h = 0.8 x 0.45 = 0.36 m 0.7 x 0 + 0.36 x 5.6 hav = ------5 3 ------= 0.32 m ri = 4-2 '4 °2 36 — 0.91; 91% oavk 4 ,2 4 . 2 32 “ °-92 ; 92% Although there is no deeper piezometer in this place to show the efficiency of the drains below invert level with respect to the previous cases, it can be expected that the ground water pressure to a certain depth below invert level, has now decreased in the same order as at invert level, i.e. ho - hav = 4.2- 0.32 = 3.88m. Probably the same efficiency as at invert level, r)av = 92%, has now been achieved. Here it is appropriate to mention that, for S/h0 ^ 1.3, or hav/h 0 < 8%, the ground water pressure at any depth permanently and irreversably decreases from the practical view point. Therefore, the analysis of drain influence with the assumption of no change in the ground water table, which is suggested by Stanic (1984), is not valid, and such an assumption may lead to very conservative and costly proposals. 8-5-5 General demonstration of the mechanism of the process, through which the efficiency of the drainage system takes effect. In this section, with respect to the previous case records given in sections 8-5-2, 8-5-3 and 8-5-4, and further piezometer records, the mechanism of the process, through which the efficiency of the drainage system in this area (and probably all other locations in over consolidated fissured clays) takes effect, will be demonstrated . A comparison is made between the efficiency values measured in practice and as calculated from theoretical methods. Records of some piezometers, which were installed before construction of drainage system and they survived for some period after the drains were installed, are shown in Figs. (6-20), (6-22), (6-24), (6-27) and (6-28). In these figures, with reference to the records of piezometers which are installed at about the same depths as trench invert levels, (say 4-5 m b.g.l.) or less, it can be observed that the piezometric levels, dropped to certain levels shortly after drainage. The important point, however, is that the piezometric levels after drainage did not fluctuate appreciably during certain periods. Records of piezometers 101, 105, 109, 201 are ample examples to show this. Piezometer 109 is an exceptional case, however, because the sudden initial drop in piezometric level, which increased shortly afterwards, Fig.(6-24), is most likely due to the excavation of a trench in the location of this piezometer. Now, it is interesting to mention that, in contrast to the above mentioned piezometers, piezometers J879, J20, J45, J877 and J13 which are installed at invert level, but some 22 years after installation of the drainage system, showed up to 1.1 meters rapid fluctuations in response to rainfall. The above phenomenon is further elaborated and elucidated with reference to the records of piezometers 401, 402 and 404, Figs. (6-33), (6-34) & (6-36). These piezometers were installed just three years after drainage system phase I was constructed. It can be clearly observed that the piezometric levels gradually started to fluctuate and the amplitude of fluctuations increased steadily year after year and tended towards an ultimate pattern. In the case of piezometers 402 and 404 the ultimate pattern occurred, probably in 1978, twelve years after the drains were installed. In the case of piezometer 401, where the trench spacing is larger, the ultimate pattern appeared in 1983, seventeen years after construction of trenches. From the above observations it can be inferred that immediately after trenches are excavated the lateral earth pressure in the soil between two adjacent trenches is released and the pore pressure drops When the trenches are backfilled the lateral pressure increases, but the soil between the trenches can still go under strain much easier than the soil in undrained areas and the soil underneath the invert level. Initially, after installation of drains, the reduced pore pressure tends to keep the soil mass as a whole body. As time passes, however, the negative pressure is gradually released and the fissures are opened. This may be called a transient process. After the transient process has finished, the soil above invert level shows quick response to infiltration. In general, the change in pore water pressure as infiltration occurs is both infiltration-dependent and strain-dependent. The pore pressure change which is caused directly by infiltration is a function of the amount and duration of rainfall, the geometry of slope and type of coverage, the diffusivity, porosity and degree of saturation of soil (partially from Lumb, 1962). The strain dependent pore pressure is dependent on the properties of the soil and the rate of strain it undergoes. As discussed above, the installation of a drainage system in fissured clay, improves both the above- mentioned factors, and in turn causes the pore water pressure, at and above the invert level, to fluctuate more quickly and by a greater amount than in the soil below invert level. The transient process, of course, takes some years to take full effect. As an example, for trench spacing to depth ratios S/D = 4.8 and S/D = 5.4 it may take 12 and 17 years respectively. Below invert level, however, with respect to the case records and discussions given in section 8-5-2, 8-5-3 and 8-5-4 [Fig.(6-79) in Zone 6, Figs. (6-78) and (8-5) in Zone 4], it is inferred that in an area in which the ground surface is subjected to continuous infiltration (and the surface is ponded), after drainage trenches are installed, the average piezometric level in the soil between ground surface and invert level, decreases and then the phreatic surface deforms from the straight line to a curved shape in the long term. At any point below invert level, however, pore water pressure decreases much more slowly, because the rates of strain and infiltration are less than that of the soil above invert level. This process is carried on, until an ultimate stress and pore pressure condition is achieved. The deeper the soil the slower the rate of pore pressure and stress adjustment process. The average efficiency of a drainage system at any point below invert level will finally tend to be equal to the average efficiency (r)av) at invert level. In reality, however, rainfall is periodic and causes the ground water pressure to fluctuate. In an area in which dry periods are dominant, the pore water pressure at invert level during each dry period decreases to a minimum value and the ground water pressure at a certain point below the invert level decreases slowly by APp When rainfall starts, the piezometric level at invert level increases quickly to a maximum value. Then, if the trenches operate efficiently, shortly after rainfall has stopped the ground water pressure at invert decreases to its previous level. During this wet period, the pore pressure at the same point below invert level increases somewhat, say AP2. The important point, however, is that APi is always greater than AP2. Accordingly, the ground water pressure, at a certain point below invert level, decreases by (AP2 - APi) during each dry and wet period. Ultimately, the pore pressure tends toward a pattern which is a function of the meterological condition, soil properties and details of the drainage system. If the infiltration is not large enough to alter the piezometric pressure at invert level, or the details of the drainage system are in such a way that the system operates with very high efficiency and practically independent of infiltration, (like the drainage system phase V in Wates House), then the average improvement of the ground water pressure at invert level, and to a certain depth below invert level, will be the same in the long term. In practice, however, the average efficiency of a drainage system at any point below invert level varies between the average maximum and average minimum efficiencies of the drainage system at invert level (r|av max, r|av min). This in turn suggests that the long term efficiency of a drainage system, having a constant spacing to depth ratio (S/D) throughout a slope varies from section to section, depending on the depth of slip surface. The deeper the slip surface the higher will be the efficiency of the drains (rjav) in the long term. As observed in previous sections: In Zone 6, [section 8-5-2, Fig.(5-18)]: at invert level: fiav = 29% at 2m below invert level: r|av * 30% In Zone 4, [section 8-5-3, Fig.(5-12)]: at invert level: r|av = 25% at about 4m below invert level: r|av = 25% Finally in section 8-5-4, Zone 2, Fig.(5-19): at invert level rjav = 92% 8-5-6 A comparison of Theoretical Solutions for trench and counterfort drains with the field data. The drain details, and performances in three zones as measured in the field, are summarized in table (8-6). The symbols have the same meanings as defined in the key diagram Fig. (1-4). The effect of paving and buildings on the lowering of the piezometric levels was discussed in section 8-4, and it was concluded that, of the total improvement of slope safety factor along profile A-A, only 36% is due to drainage. Nevertheless, Zone 4 itself is only a small part of profile A-A which is not paved, but, of course, is affected by nearby tall buildings, and also limited to paving and buildings in North and South, which are affecting the phreatic surface in this zone. Therefore, it must be noted that the rj = 22% and r|av = 25% values, given in Row No.l, table (8-6) for Zone 4, are indeed a summation of the efficiency of drains and part of the efficiency of surface covering. To estimate the efficiency of the drains alone it is referred to piezometer 406, installed in the fully surface covered area,, midway between trenches. The fj and r|av according to this piezometer and table (8-6) are: Table (8-6) The drains details and performances, 1987. 2 P •<—» rG Q *G S J « ' O vO § EL G t g S o B & 8 O wo 6*0 wo o r s c o V— (5-12) 4.2 15.9 2 3.8 3.0 2.4 2.31 0.97 (N * /• wo O V wo o r o c NCO CN oo © oo o -- 1 s 4.15 14 1.5 3.4 , 3.15 2.52 2.39 0.94 25 29 i—H o s c \ o (5-19) 0.7 5 5.6 ? 4.2 CO 0.45 0.36 0.32 0.11 92 Note (*) The r| and rjav values in row 1, Zone 4, are the total values of the efficiencies caused by drainage and paving. The r| and r|av values corresponding to the drainage only are: rj= 18 and r|av = 19 per cent. p = 0.6 m;from piezometer 106 ho = 4.2 - 0.6 = 3.60 m hm = 4.2 - 2.0 = 2.2 m h = 0.8 * 2 .2 = 1.76 m 1 x 0.9 + 1.76 * 15.9 = 1.71m 16.9 3.6 - 1.71 nav= 3 6 = 0.52; 52% 3.6 - 1.76 = 0.51 ; 51% n 3.6 It can be inferred that if Zone 4 was paved the r|av value would be in the order of 52%, rather than 25%. The f\ and rjav values in Zone 4, cross section (5-12), corresponding to the drains only are as follows: rjav =0.36* 52= 19% 0 = 0.36*51 = 18% Comparison with Hutchinson’s Method (1977). The theoretical and experimental variations of rj against S/ho are compared in Fig. (8- 6). The curve ’A’ is the best fit to the experimental data. The field n values are 2 in Zone 4 and 1.5 in Zone 6, so that curve ’A’ may be compared with the average theoretical curves C and D corresponding to n = 2. It can be observed that for S/ho ^ 4 and S/ho s 1.6 values, the theory over estimates and under-estimates the r\ values respectively. As the in situ permeability ratio Rfc is not known, a precise comparison is not possible. However, with reference to Fig. (8-6) it can be inferred that in Zone 2, S/ho = 1*3, the rj value is underestimated by 38% to 7%, for Rfc = 1 to 5. In Zone 4, S/h0 = 5.1, the rj value is overestimated by 18% to 182% (fi theory = 2.82 f| practice)? f°r = 1 to 5. Finally, in Zone 6, S/h0 = 4.1, the rj may be overestimated by 1% to 124%, for Rk = 1 to 5. For 1.6 < S/ho < 4,the drains efficiency may be over or under estimated and/or coincide to the field data. Hutchinson, (1977), evaluated his solution comparing with field data, for S/h0 > 2, collected from different sites, including previous data from Zone 4, (Simons, 1977). Then he suggested that, until further observations became available, the curves H and G in Fig. (8-6) could be regarded as the tentative upper and lower bounds for the design of trench or counterfort drains in over consolidated clay slopes under temperate climatic conditions, up to an S/hQ ratio of 5 to 6. Now, with reference to Fig. (8-6) it is seen that there is a considerable discrepancy between Curve ’A* and the pattern which was achieved by Hutchinson. He has inadvertently based his evaluation over the records of standpipe piezometers installed at trench invert level. As discussed in section 8-2-3, installation of standpipe piezometers in drained areas, especially at invert level, may lead to considerable errors, sometimes about one metre head of water, in measurement of ground water pressure, and Hutchinson’s evaluation most likely suffered such an error. For this reason he has come up with such a conservative suggestion. It must be mentioned that, of course, before this research was carried out, the significance of errors in the monitoring of ground water pressure in a drained area had not been revealed and quantified. The other error which his evaluation suffered is that, he thought the rj value taken from Zone 4 was caused by the drainage only, while it is now revealed that paving efficiency is the dominant factor. Comparison with Bromhead’s solution. A comparison of Bromhead’s solution (1984) for counterfort drains with the field data is shown in Fig. (8-7). Curve ’A’ represents the field behaviour of the drains. It can be observed that the field curve ’A* intersects theoretical curves [a] to [d]. The discrepancy is neither due to soil anisotropy, because the effects of soil anisotropy in the general solution is included by scaling S' (the real drains spacing) to an apparent spacing S by means of transformation S = S /^ j^ , in which Rk is the ratio between horizontal and vertical permeability coefficients, and the changes of Rk value shifts the locations of the curves to the left or right only, nor is it due to the fact that the field drains are not counterfort, because, it was already shown that the drains improved the ground water pressure a the location of slip surface in the same order as at invert level. Most likely, the disagreement is initiated due to the assumptions of the solution, like ignoring soil compressibility, taking slip surface as being impervious. Comparison with the Stanic’s solution. The comparison of Stanic’s solution for trench drains with the field data is shown in Fig. (8-8). The Stanic’s Solution Curve E, is plotted for a slope inclination of tan P = 1/6. It can be observed that the curve E is significantly different from curve ’A’. However, for S/ho 1.5 the curve E is more or less parallel to curve ’A’. Accordingly, if the rj values, deduced from the Stanic’s solution, are multiplied by a correction factor, then realistic results may be achieved. 8-5-7 Conclusions a - Immediately after trench drains have been installed, the ground water pressure in the soil between ground surface and invert level drops to a certain level. In the first stages after drainage, the ground water pressure, at invert level and above, does not show a significant response to periodic infiltrations. With time, however, it starts gradually to fluctuate with higher amplitudes and shorter periods in response to infiltration periods. Ultimately, say after about 10 years for S/D = 3.5, the above mentioned transient condition terminates, and the fluctuation of piezometric level, at trench invert, tends toward a consistent pattern. Now, the actual efficiency of the drainage system in improvement of pore water pressure from land surface to invert level has started. In this stage the amplitude of change of pore water pressure at invert level may be many times higher than before drainage at that level. Below invert level, however, the pore water pressure does not change as quickly as that at/or above invert level, the greater the depth the slower the rate of change of pore water pressure. In the long term, the average efficiency of a drainage system at any point below invert level varies between the average maximum and average minimum (r|av max & r|av min) efficiency values of the drainage system at invert level corresponding to the average minimum and average maximum piezometric levels (hav min & hav max) at trench invert level. This in turn, suggests that the long term stabilizing efficiency of a drainage system, having a constant space to depth ratio (S/D) throughout a slope, varies from section to section depending on the depth to the slip surface; the deeper the slip surface the higher the drains efficiency (r|av) in the long term. This conclusion is in contrast to the previous views emphasising the necessity of penetration of trenches down to the slip surface (e.g. Schweiger and Wright, 1974). When S/D is less than a certain value, e.g. S/D ^ 1.2, the ground water pressure at invert level drops significantly, hav/h0 < 8%, and its response to any infiltration is insignificant. The system may be called infiltration independent practically. Therefore the analysis of the drain influence with the assumption of no change in the ground water table (hm = h0) which has been suggested previously, e.g. Stanic, (1984), is not valid, and may lead to very conservative and costly proposals. b - The theoretical solutions for drains reviewed in section 1-4-3-2, do not show a good agreement with the field data. It is, therefore, opportune to repeat Hutchinson’s proposals (1977), with some modifications. It is suggested that, until further observations as accurate as those provided in this research become available, the curve ’A’ in Fig. (8-6) can be regarded as a criterion for the design of trench and counterfort drains, in over consolidated clay slopes under rainfall conditions shown in Fig. (6-17). 6__ u X 'A/2 | Fig(1.4). Key diagram: cross-section of typical trench drains. (1) Ground surface. (2) Original piezometric level on plane | EH. (3) Piezometric levels on plane FG after drainage. | (4) Mean piezometric level on plane FG after drainage. | (5) Mean piezometric level on drainage inverts after drainage. | (6) Trench on counterfort drain. (7) Clay seal. (8) Impermeable boundary at depth. h = fs.hm in which fs = a factor reflecting the shape of the relevant piezometric surface. The theoretical values of fs, range from 0.76 to 0.80. It is suggested that in practical cases where only the mid-point piezometric height, hm is known, an estimate of the value of h can be made by taking this as = fs hm, with fs = 0.8, (Hutchinson, 1977). hav = the average piezometric head on the whole plane EH. j j jj r| = —^ — , the efficiency of trench drains on a given cross section, with respect to the intervening mass of soil. rjav = h° ^av» ^ the overall efficiency of drains, on any given cross section. Fig. (8-1) EQUALIZATION OF PIEZOMETER J44 AFTER CORRECTION AGAINST THE CHANGES OF GROUND WATER PRESSURE O O O O O O O O O Q O ) a a \ o t>- co \ \ to to CBD N OUVZITVnOB OUVZITVnOB N CBD . N. ... \ V n -"* in . N 30 V. 31 VH V. *s co ------ s s c \ \ j \ " " •—* * ■ \ V, t — ------\ \ \ \ '1 j 1 1 1 1 1 — 1 1 I 1 1 I — 1 I 1 1 1 1 I — H 1 1 1 i | 1 100 1000 10000 100000 lOOOOOt TIME (minute) ATMOSPHERIC PRESSURE (mb) CD CD CD O CD CD LD CD CD o o 1—5 C d LU CD I— CO LU CD CM r-J GO LU CM CM CO . O' CM co LD o CD CM QC CD uJ u) CD CM LU CD oc CM Ll_ CM CD CM CL CD O CM LU LU 0 0 CD QC o n co ~ CD _ 00 CM ICQ CD co CD JD in CD LU '0 CD Z CM LU CD CD CD ZZ1 CD GO CO QC Q_ CO ZSL CD in CD CM C\J CO I 00 cn CD CD CD CD CD CD CD CD CD CD CD CD in CD in CD in CD in CD in CM CM CO CO in LD CW3) 13A31 0Idl3W0Z3Id 01 Hld30 GL - 2 \ R©f ©n to Fig. C6-21) • 1.03. - 7 104. 2.3x10 CM/SEC _ 8 . 10 I . i . J 0 H S V Seal© 5 AVERAGE PIEZO. LEVEL CONJECTURAL PIEZO. lovol J SEEPAGE DIRECTION Fig. <8-3) Ground "water pressure profiles GL -x49 - - e -8 CM/SEC J43 - 601 - J ^ CM/SEC -8 KP3- 622" 2 . 3 x 10~5 CM/SEC 702- CM/SEC “ IS Rofar to Fig- <8-751 . <6-76> & C5-77> -24 - 23 7 2 3 . CM/SEC -32 10 HSV SCALE MAX. . & MIN. PIEZO. LEVELS 7 21 - CM/SEC -48 Fig. (8-3) Cont. GL 101 - 8 x 10~8 CM/SEC -6 -8 -10 -12 Rofor to Fig. C6-20) -14 - 1 6 - 5 1.5 x 10 CM/SEC Profile ( 3 ) I____ i____ i___ ;_i____ j____ I 0 H & V Seal© 5 AVERAGE PIEZO. LEVEL CONJECTURAL PIEZO, l.vol J SEEPAGE DIRECTION Fig. <8-3) Cont, GL _ 2 105. 2 . 3 x 1 0 CM/SEC - 4 Refers t o Fig. C6~22) .“8 106.-6. 1. 4x10 CM/SEC _ 8 - 10 _ 12 -5 107 C M / S E C7x10 _14 1____ 1____ !------1------1____ I 0 H S V Seal© 5 AVERAGE PIEZO. LEVEL CONJECTURAL PIEZO. 1 BVol } SEEPAGE DIRECTION Fig. C8~3) Cont. ,-S 109- CM/SE.C ReFor" to Fig. C6—24) Profile C 5 ) -5 110 CM/SEC GL H & V Seal© -2 AVERAGE PIEZO. LEVEL CONJECTURAL PIEZO, level { SEEPAGE DIRECTION Rafon to Fig. CS—28) .-8 202 --6 CM/SEC -8 208-- -10 -12 -8 5. 3 x 10 CM/SEC Fig. (8-3) Cont, RoFon to Fig. <6—465 -2 416- 2 . 9 x 10.-6 CM/SEC -4 417- 1.8 x 10”5 CM/SEC GL • Rofer to Fig. C6—63) -4 Trench invert level H & V S e a 13 -6 J41 *“ 8 J40 -- 10 P r o f i l e C > GL MAX. & MIN. PIEZO. L E V E L S CONJECTURAL PIEZO. LEVEL \ SEEPAGE DIRECTION -8 J24- CM/SEC .-6 '-4 invent level J22- CM/SEC Refen to Fig. C6*~66} -6 Fig. (8-3) Cont GL -2 J25- R©for' to Fig. C6—675 - 4 J879- TnortoFi inv ert lovol J4 3 --6 CM/SEC J875--8 -10 J 878 - 12 ProFi1© C 10 ) 1____ i____ i____ i____ i____ I 0 H S V Seal© 5 r— — MAX, S MIN. PIEZO. LEVELS CONJECTURAL PIEZO. LEVEL \ SEEPAGE DIRECTION Fig- <8-3) Cont. GL H & V Seal© 5 MAX. & MIN. PIEZO. LEVELS J26- CONJECTURAL PIEZO. LEVEL \ SEEPAGE DIRECTION - 4 J2 8 - Refer' to Fig. CO—09} - 6 -8 J 4 7 - - 10 Pr^oFi 1 q C 11 ) GL -2 J3 - Refer* to Fig. CO—70} -4 J2 - - 6 J1 - 4. 8 x CM/SEC - 8 Profi1© C 12 ) Fig. (8-3) Cont. i . \ \\ \ \ J7 CM/SEC - 4 J3 8S J5 CM/SEC -6 .“8 J6 CM/SEC -8 J4 3.1 x 10 - 10 Pr^ofi le C 13 y Refer to Figa. (6—715 & (6—73} GL H S V Seal© MAX. & MIN. PIEZO. LEVELS CONJECTURAL PIEZO. LEVEL \ SEEPAGE DIRECTION Refer to Fig. (6—72} -6 J9 --6 3. 4 x 10 CM/SEC -8 J8 --8 x 10 CM/SEC Refer to Fig. (6—72} J12 - J10 -• Voter bearing layer J1 1 - - 4 Fig. (8-3) Cont, GL r 2 Refor to Fig. C6*"68) r 4 Tr^onoh invert 1 ovel r 6 h e ■---* MAX. a MIN. PIEZO. LEVELS | SE E P A G E D I R E C T I O N Fig- (8-3) Cont. r-l in L i © in X 0 r - l c n c •H o c c . *0■rl •H 3 0 +> ■H L 07 " 0 C t 0 1 _1 CL 0 3 : L 0 “U ■f3 Z C U )(L> D II O *»• rH 0 L D +> 0 C 0 <4- 0 C 0 •H •P LT) 0 1 0 i n 0 I—1 © r—l © •H 0 - 0 0 E L •H CL X 0 CD L C CL 0 c l r-H < 0 r \ 1 0 0 v 0 7 •H LI Q O E to CD Q to CD r> (0 to 10 IT) 'T GL -2 - 4 •TRENCH INVERT LEVEL PI 03 - J 2 0 -- -6 J42-- J40 -10 Scale = _J_ 100 Max. & Min. Ground water pressure prof i les before drainage ( ^ ( 4 ) = Max. & Min. Ground w a ter pressure p r o f i les after drainage (1987) ..refer to Figs. (5~12) & (5—2) Fig. (8-5) Ground water pressure profi1©s before and after drainage, in zone 4 Zone 2 n = 4;5 0-4 Zone 6 0.2 Rk=1 ] n= 4-5 Fig. C8-6) Comparison oF approximate theory (Hutchinson 1977) f on t r ench d r a i n s with the -Field data summarised in table ( 8 ~ 6 ) 1.0 A — ■Field data r 5 ■ • 2 9 • . 9- b 1 \im _L % H *9 *.» * V / ' 9 • t* 1 7 foj,:✓ lol d=H lb] d=%H lc] d-fjH L 0.5 / H Idl d=%H / "A lc 1 IdJ 0 0.5 1.0 0.5 Rc 1/R c F i g . (8-7) Comparison oF Bromhead/s solution(1984) for countefort dnains with the Field data summarised in tab1e(8—6) 1 fi el d data Q8 Sfcanic/8 solution 0.6 0.4 0.2 0 0 2 3 5 6 Fig. (8-8) Comparison of* Stanic's solution C1984) for trench drains with the ■Field data summarised in tab1©(8~S) R eferences Baikie, L.D. (1985) Total and partial factors o f safety in geotechnical engineering. Can. Geotech. J. 22, pp 477-482. Beles, A.A. and Stanculescu, I.I. (1958) Therm al treatm ent as a m eans o fim proving the stability o f earth m asses. Geotechnique, Vol.8, No.4, pp 158-165. Bishop, A.W. (1955) The use o f slip circle in the stability analysis o f earth slopes. Geotechnique, Vol.5, No.l, pp 7-17. Bishop, A.W. (1959) The principle o f effective stress Tekisk Ukeblad, Vol. 39, p 859 Bishop, A.W. and Bjerrum, L. (1960) The relevance o f the tn’axial test to the solution o fstability problem s. Proc. Rec. Conf. on the shear strength of cohesive soils, Denver, Colorado (Special publication, American Society of Civil Engineers) pp 439-501 Bishop, A.W., Kennard, M.F. and Penman, A.D.M. (1961) Pore pressure observations a t Selset Dam. Proc. Conf. on pore pressure and suction in soils, Butterworths, London, pp 91-102 Bishop, A.W., Kennard, M.F. and Vaughan, P.R. (1964) D evelopm ents in the m easurem ent and interpretation o f pore pressure in earth d a m s . Trans. 8th Congress Large Dams, 1, pp 47-72 Bishop, A.W. (1971) The influence o fprogressive failure on the choice o f the m ethod o f stability a n a ly s is . Geotechnique 21, pp 168-172 Bjerrum, L. and Johannsen, I. (1961) Pore pressure resulting from driving piles in soft clay. Proc. Conf. on pore pressure and suction in soils, Butterworth, London, pp 108-111 Bjerrum, L., Nash, J.K.T.L., Kennard, R.M. and Gibson, R.E. (1972) H ydraulic fracturing in field perm eability testing Geotechnique 22, No.2, pp 319-332. Bloss, S.A. (1986) Equalization o f sim ple piezom eters B.Sc. dissertation, Kingston Polytechnic. Bogoslovsky, A.V., and Ogilvy, A.A. (1977) G eophysical M ethods for the Investigation o fLandslides. Geophysics, Vol. 42, No.3, pp.562-571. Bowler, R.A. and Phillipsons, H.B. (1984) Landslip preventive measures - a review o fconstruction Hong Kong Engineer, Vol. 10, Pt.10, pp 13-31. Brand, E.W. and Premchitt, J. (1980a) Shape factors o f cylindrical piezom eters Ceotechnique 30, No.4, pp 369 - 384 Brand, E.W. (1981) Som e thoughts on rain-induced slope failures. Proc. 10th Int. Conf. Soil Mech. Found. Eng., (3) Stockholm, pp 373 - 376 Brand, E.W. and Premchitt, J. (1982) Response cham cteristics o fcylindrical piezom eters Geotechniques 32, No.3, pp 203 - 216 Brand, E.W. (1984) Landslides in Southeast Asia: A STATE-OF-THE-ART-REPORT 4th Int. Symp. on Landslides, Toronto, September 1984, pp 3.1-3.43 Bromhead, E.N. and Vaughan, P.R., (1980) Solutions for seepage in soils with an effective stress dependent perm eability. Proc. 1st Int. Conf. on Numerical Methods for non-linear problems, Swansea, Pineridge Press, pp 567-578. Bromhead, E.N. (1984) A n analytical solution to the problem o fseepage into counterfort drains. Can.Geotech. J.21, pp 657-662. Butler, F. (1972) G eology, Soils and Foundations Thames-side Symposium, Ove Arup Partnership, pp 15-23 Carter, M. and Bentley, S.P. (1985) The G eom etry o f slip surfaces beneath landslides: prediction from surface measurements Canad. Geotech. J. Vol. 22, pp 234 - 238 Casagrande, A. (1949) Soil M echanics in the design and construction o f the Logan Airport. Boston Society of Civil Engineers Contributions to Soil Mechanics, pp 1941- 1953 Cedergren, H.R. (1977) Seepage, dminage and flownets John Wiley & Sons, New York. Chandler, R.J. (1984) Recent European experience o flandslides in over-consolidated clays and soft r o c k s . Proc. 4th Int. Symp. on Landslides, Toronto, 1, pp 61 - 81 Chugh, A.K. (1986) Variable factors o fsafety in slope stability analysis Geotechnique 36, pp 57 - 63. Clayton, C.R.I., and Hodder, J.P., (1981) A n investigation into the landslips on natural slopes o f weathered London clay at Stag H ill, G uildford. M.Sc. Thesis Civil Engineering Department, University of Surrey. Clayton, C.R.I. and Timche, W.A., (1983) N on-circular analysis o fslopes M.Sc. Dissertation, Civil Engineering Department, University of Surrey. Daehn, W.W. (1962) D evelopm ent and installation o f piezom eters for the m easurem ent o f the pore fluid pressure in earth dams. American Society for Testing Materials, Vol. 88 Eggboro, M.D. and Walthall, S. (1985) The value and interpretation and ground water level m easurem ents. Ground water in Engineering Geology, 21st Regional Conference of the Engineering group of the Geological Society, pp 423-436 Elmer, W., Brooker and Lindberg, D.A. (1985) F ield m easurem ents o f pore pressure in high plasticity soils Proc. Int. Research and Eng. Conf. on expansive soils. College Station, Texas, pp 57 - 68. Foundation Engineering Limited (1964) Report No. F. 69/882, August (1964) Gibson, R.E., (1963) A n analysis o f system flexibility and its effects on tim e lags in pore pressure measurements. Geotechnique 13, pp 1 - 9 Gibson, R.E. (November 1964) R eport on slope stability o f Stag H ill, Guildford. Building Design Partnership. Gibson, R.E. (1966) A note on the constant head test to m easure soil perm eability in situ. Geotechnique 16, pp 256 - 259. Gibson, R.E. and Shefford, G.C. (1968) The efficiency o fhorizontal drainage layers for accelem ting consolidation o f clay embankments. Geotechnique 18 pp 327 - 335 Gibson, R.E. (1970) A n extension to the theory o f the constant head in situ perm eability test. Geotechnique 20 (2), pp 193 - 197 Henkel, D.J. (1957) Investigation o f tw o long-term failures in London clay slopes at W ood Green and N ortholt. Proc. 4th Int. Conf. Soil Mech. Found. Engg., London, 2, pp 315 - 320 Huang, Y.H. (1983) Stability analysis o fearth slopes Van Nostrand Reinhold Company Inc. Hutchinson, J.N. (1977) A ssessm ent o f the effectiveness o f corrective measures in relation to geological conditions and types o f slope m ovem ent. Bulletin of the International Association of Engineering Geology No. 16, pp 131-155. Hutchinson, J.N. (1983) M ethods o flocating slip surfaces in landslides. Bull. Assoc. Eng. Geol. XX, 3, pp 235 - 252 Hvorslev, M.J. (1951) Tim e Lag and soil perm eability in ground w ater observations. Waterways Experimental Station, U.S. Corps. Engrs. Vicksburg, Mississipi, Bulletin No. 36. Joselin De Jong, G.DE (1953) Consolidation around pore pressure m eters J. Appl. Phys., Vol. 24, No.7, pp 922 - 928 Kallstenius, T. and Wallgren, A. (1956) Pore w ater pressure m easurem ents in field investigations. Royal Swedish Geot. Inst. Proc., No. 13. Kenney, T.G., Pazin, M and Choi, W. (1977) D esign o fH orizontal drains for soil slopes. ASCE, Journal of the Geotechnical Engineering Division, part 2, pp 1311 - 1323 Lambe, T.W. and Whitman, R.V. (1969) Soil M echanics John Wiley & Sons, New York. Lau, K.C. and Kenney, T.C. (1984) H orizontal drains to stabilize clay slopes. Canad.Geot. J. 21, No.2, pp 241 - 249 Lee, I.K. (1974) Soil M echanics - N ew Horizons Butterworth & Co. Ltd., London Lumb, P. (1962) E ffect o frainstorm s on slope stability. Proc. Symp. Hong Kong Soils, Hong Kong, pp 73-87 Matsuo, M., and Suzuki, H., (1985) Exam ination on safety factors in specifications Proc. 11th Conf. Soil Mech. Found. Engg., San Francisco, pp 831-834 Meyerhof, G.G. (1970) Safety factors in soil m echanics Canad. Geot. J., No.4, pp 349-355 Meyerhof, G.G. (1984) Safety factors and lim it states analysis in geotechnical engineering Canad. Geot. J.21, pp 1-7 Mitchell, R.J. (1983) Earth Stm ctures Engineering Allen & Unwin Inc., USA Morgenstem, N. R. and Price V.E. (1965) The analysis o f the stability o f general slip surfaces Geotechniques, Vol. 15, No. 1, pp 79 - 93 Morgenstem, N.R. and Tchalenko, J.S., (1967) M icrostructum l observations on shear zone from slips in natural clays. Proc. Geotech. Conf. (Oslo), Vol.l, pp 147 - 152 Morgenstem, N.R. (1980) Factors affecting the selection o fshear strength param eters in slope stability a n a ly s is Proc. Int. Symp. Landslides (2), New Delhi,.pp 83 - 93 Nonveiller, E. (1981) E fficiency o fhorizontal dm ins on slope stability. Tenth Int. Conf. Soil Mech. Found. Engg., Stockholm, Vol. 3, pp 495 - 500. Ove Amp & Partners (1970) Report No. 3771/DJH/JRS Papadopoulos, B.P. and Anagnostopoulos, A.G. (1981) Factors affecting the stability o fslopes Proc. 10th Int. Conf. Soil Mech. Found. Engg. Stockholm, pp 501 - 504 Penman, A.D.M. (1956) A field piezom eter apparatus Geotechnique, 6, No.2, pp 57-65 Penman, A.D.M., (1961) A study o f response tim es o f various types o f piezom eter. Prof. Conf. Pore Pressure, Butterworth, London, pp 53-58 Premchitt, J. and Brand, E.W. (1981) Pore pressure equalization o f piezom eters in com pressible soils Geotechnique 31, N o.l, pp 105 - 123. Rocha Filho, P. (1976) Laboratory tests on a new borehole seal for piezom eters Ground Engineering, Vol.9, No.l pp 16-18. Schweizer, R.J. and Wright, S.G. (1974) A survey and evaluation o frem edial m easures for earth slope stabilization. Research report 161 - 2 F CHR, University of Texas, Austin, Texas, USA. Sills, G.C. (1974) A n assessm ent, using three field studies, o f the theoretical concept o f the efficiency o f drainage layers in an em bankm ent. Geotechnique 24, No.4, pp 467 - 474 Simons, N.E. (1977) Slips, settlem ents and sink holes University of Surrey. Simons, N.E., Clayton, C.R.I., Sadrekarimi, J. and Matthews, M.C., (1987) The use of drains to improve the stability of a land slip site: A case study. 9th European Conf. Soil Mech. Found.Engg, Dublin. Skempton, A.W., (1954) The pore pressure coefficients A and B in saturated soils. Geotechnique 4, pp 143 - 147. Skempton, A.W., (1961) H orizontal stresses in an over-consolidated clay. Proc. 5th Int. Conf. Soil Mech. Found. Engg., Paris. 1, pp 351 - 357 Skempton, A.W., (1966) Som e observations on tectonic shear zones. Proc. 1st Int. Conf. Rock Mech., Lisbon, Vol. 1, pp 329 - 35 Skempton, A.W. and Petley, D.J., (1967a) The strength along structural discontinuities in stiff clays. Proc. Geot. Conf., Oslo, 2, pp 29 - 46. Skempton, A.W. Schuster, R.L. and Petley, D.J., (1969) Joints and fissures in London clay at W raysbury andEdgware. Geotechnique 19, No.2, pp 205 - 217 Skempton, A.W., (1970) First-tim e slides in over-consolidated clays. Geotechnique 20, pp 320 - 324 Skempton, A.W. (1977) Slope stability o fcuttings in brown London clay. Special lecture. 9th Int. Conf. Soil Mech. Found. Engg., Tokyo. Skempton, A.W. (1985) Residual strength o f clays in landslides, folded strata and the Laboratory Geotechnique 35, pp 3 - 18 Smith, Rockwell, (1958) Landslides and Engineering practice. Highway Research Board Special Report 29, Washington, D.C., Chapter 2, p 6. Soil Instruments Limited (1985) Vibrating wire piezom eter users* manual. Stanic, B., (1984) Influence of drainage trenches on slope stability ASCE 110, pp 1624- 1635 Terzaghi, K., (1936) Relation between soil m echanics and Foundation Engineering Presidential address, Proc. 1st Int. Conf. Soil Mech. and Foundation Eng., Harvard, Vol. 3, pp 13-18 Torstensson, B-A. (1975) Pore pressure sounding instrum ent Disc., Session 1, Proc. ASCE Spec. Conf. on In-situ measurement of soil properties, Vol. II, Raleigh, North Carolina,pp 48-54 Torstensson, B-A., (1978) The pore pressure probe. Proc. Geoteknikkdagen, Tapirforlag, Oslo, Norway, pp 34.1 - 34.15, Tuinzaad, H., (1954) Influence o f the atm ospheric pressure on the head ofartesian w ater and phreatic w a te r . Assoc. Int. Hydrol. Sci., Assem. Gen. Rome, 2; pp 32 - 37 Turk, L.J., (1975) D um al fluctuations o f water tables induced b y atm ospheric pressure changes. J. Hydrol., Vol. 26, 1/2, pp 1 - 16. Van Schilfgaarde, J. (1970) Theory o f flow to drains Advances in Hydro Science, 6, pp 43 - 106 Vaughan, P.R., (1969) A note on sealing piezom eters in boreholes Geotechnique 19, No.3, pp 405 - 413 Vaughan, P.R., (1974) The m easurem ent o f pore pressure with piezom eters Proc. of conf. on field instrumentation in Geotechnical Engineering, London, pp 411 -422 Walthall, S. and Ingram, J.A., (1984) The Investigation o faquifer param eters using m ultiple piezom eters Groundwater, Vol. 22, pp 25 - 30 Weeks, A.G., (1969) E ffects o f Counterfort drains on the Sevenoaks by-pass Civil Engineering and Public Works Review. Weeks, E.P., (1979) Barom etric Fluctuations in W ells Tapping D eep Unconfined A quifers Water Resources Research, Vol. 15, pp 1167 - 1176 Whitlow, R., (1983) Basic Soil M echanics Construction Press, Longman House, Essex, England. Yamaguchi, Sh. (1977) D eterm ination o f The Location o f drains and Assessm ent o f the effect o f the work b y repeated electric resistivity survey. Bulletin of the Int. Association of Engg. Geology, No. 16, pp 183 - 184. Zaruba, Q. and Mencl, V. (1969) Landslides and their control Elsevier, Amsterdam and Academia, Prague A ppendix DEPTH TO PIEZOMETRIC LEVEL < m > DEPTH TO PIEZOMETRIC LEVEL 5 4 3 2 1 1966 1966 > ©% © © D 1967 1967 ||Ii!¥EESiT¥ CF SORREf i H i i U EOD O PEOEES & B & A PIEZOMETERS OF RECORDS EOD O PEOEES & D & C PIEZOMETERS OF RECORDS DATE DATE 1968 1968 ©o c 6 © 1983 " O N 94 1975 1974 &QU- 1984 5. m 5 OO 95 1986 1985 1976 0 © v A DEPTH TO PIEZOMETRIC LEVEL < m > DEPTH T0 PIEZOMETRIC LEVEL E Fig. 3 4 2 1 4 3 2 1