Geological and Engineering Analysis of Residual Soil for Forewarning Landslide from Highland Area in Northern Thailand

Open Geosci. 2015; 1:637–645

Research Article Open Access

Thanakrit Thongkhao, Sumet Phantuwongraj, Montri Choowong*, Thanop Thitimakorn, and Punya Charusiri Geological and engineering analysis of residual soil for forewarning landslide from highland area in northern Thailand

DOI 10.1515/geo-2015-0059 Received Jan 14, 2015; accepted Aug 27, 2015 1 Introduction

Abstract: One devastating landslide event in northern Landslide is a worldwide natural hazard, which in most Thailand occurred in 2006 at Ban Nong Pla village, Chiang cases, occurs as a debris flow and soil creep. In a tropi- Klang highland of Nan province after, a massive amount cal climate region, landslide is commonly caused by in- of residual soil moved from upstream to downstream, via tense and continuous heavy rainfall [1]. Therefore, a better creek tributaries, into a main stream after five days of un- understanding of landslide processes requires the precise usual heavy rainfall. In this paper, the geological and en- characterisation of the triggering factors and relationship gineering properties of residual soil derived from sedimen- among geological and engineering parameters. The trig- tary rocks were analyzed and integrated. Geological map- gering factors are often time dependent and for this reason ping, electrical resistivity survey and test pits were car- it is complicate to have a quantitative approach without a ried out along three transect lines together with system- permanent in-situ investigation measurement [2]. In most atic collection of undisturbed and disturbed residual soil of the landslide processes, loss of soil equilibrium caused samples. As a result, the average moisture content in soil by heavy rainfall is considered as one of the most impor- is 24.83% with average specific gravity of 2.68, whereas the tant triggering factors. liquid limit is 44.93%, plastic limit is 29.35% and plastic in- Thailand is located in a warm and tropical climate dex is 15.58%. The cohesion of soil ranges between 0.096– region. The tropical monsoons and typhoons from both 1.196 ksc and the angle of internal friction is between 11.51 the Andaman Sea and the South China Sea contribute and 35.78 degrees. This suggests that the toughness prop- to the heavy rain inducing landslide to occur around the erties of soil change when moisture content increases. Re- country. Rainy season starts in June and commonly pro- sults from electrical resistivity survey reveal that soil thick- duces over discharge in the basins locating in the north- nesses above the bedrock along three transects range from ern part of the country. Beside the effect of heavy rain in 2 to 9 m. The soil shear strength reach the rate of high de- the granitic terranes from the north that triggered land- creases in the range of 72 to 95.6% for residual soil from slides, but long and continue heavy rain in southern Thai- shale, siltstone and sandstone, respectively. Strength of land also causes the overbank full discharge leading to soil soil decreases when the moisture content in soil increases. creep in many places [3–5]. The average annual rainfall Shear strength also decreases when the moisture content ranges from 1,000 to 1,500 mm for northern, northeastern changes. Therefore, the natural soil slope in the study area and central parts of the country, whereas at the eastern will be stable when the moisture content in soil level is and southern part of the country, rainfall is between 2,000 equal to one, but when the moisture content between soil to 3,000 mm. Rainfall inducing landslides normally occurs particle increases, strength of soil will decrease resulting in mountainous area due to single intent or long period in soil strength decreasing. of rain. In case of prolonged rainfall, the flood and debris flow commonly follow landslide and cause more damages Keywords: landslide; multi-stage direct shear test; shear to villages along flow passages and on the alluvial plain. strength; Nan; northern Thailand

Thanakrit Thongkhao, Sumet Phantuwongraj, Thanop Thiti- *Corresponding Author: Montri Choowong: Department of Geol- makorn, Punya Charusiri: Department of Geology, Faculty of Sci- ogy, Faculty of Science, Chulalongkorn University, Bangkok 10330, ence, Chulalongkorn University, Bangkok 10330, Thailand Thailand; Email: [email protected]; [email protected]

© 2015 Thanakrit Thongkhao et al., published by De Gruyter Open. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivs 3.0 License. failure of slope from the highland or steep terrain in northern Thailand is often reported to occur mostly in conjunction with flash flooding from heavy rainfall and in those areas where human activities such as exploitation of land by changing soil slope for agriculture purpose are present. The Ban Nong Pla area is considered to be situated in one of the landslide risk zones. 638 Ë Thanakrit Thongkhao et al.

Figure 3. Location of a village and present condition of highland forest and agriculture. Figure 3: Location of a village and present condition of highland forest and agriculture. Figure 1. Digital Elevation Model (DEM) showing the study area (red square) covering the THE METHODOLOGIES Figure 1:northernDigital provinces Elevation of Thailand Model and (DEM)Thai-Laos showing PDR border the (left) study and index area map (red of square)Thailand covering indicating the northernwhere the study provinces area is located of Thailand (black square) and (right). Thai-Laos Satellite image and interpretation

PDR border (left) and index map of Thailand indicating where the 2001The at THEOS Wang satellite Chin image district, from Geo-Informatics Phrae and province. Space Technology This Development mega- study area is located (black square) (right). Agency is firstly applied in this study for interpreting and classifying patterns of landslide landslidescars, land-use/land was cover, one type of of the common country’s and unusual worst agricultures, records. and boundary Land- of existing forest. Geographical reference data were digitized with the coordinates either in slide subsequently extended to Phee Pun Num mountain terms of latitude and longitude or columns and rows. Attribute data associate a numerical

rangecode to eac coveringh cell or set of Phrae, coordinates Lampang and for each variable, and Sukhothai or to represent actual provinces. values (900 m elevation, 25 degrees slope gradient) or to connote categorical data types (land-uses,

Thevegetation number type, landof cover, casualties rock type) [12] has. reached more than 30 peo- After geomorphological map has been created, test pits and shallow excavation along

plesthe prospected and caused transects the were economic conducted. These losses combined of more methods than provide 300 an academic mil-

mean of acquiring a very detailed record of the complex soil conditions which often exist

lionnear to bahtthe ground [6]. surface. Soon It is after, worth remembering, at Nam Kor-Nam however, that Chun test pits and area, other Lom- outcrops

can also be used for in situ testing and to obtain high quality samples. Disturbed and sak district, Phetchabun province, a landslide occurred undisturbed soil samples from these methods were analyzed for both physical and engineering properties. in August 2001, killing 150 people and damaging about

600Two-dimension housesal electrical with total resistivity amount survey of 645 million baht of eco-

nomicTwo-dimension loss. Duringal electrical that resistivity time, measurement the amounts has been carried of rainfalls out aiming to distinguish a layer of residual soil from rock basement. Basic modes of operation include the wereprofilingreported (mapping), vertical up to electrical 150 mm sounding [7]. (VES In) May,, combined 2004 sounding at Mae and profiling Ra- mat(two-dimensional district, resistivity Tak province, imaging), and one electrical mega-landslide resistivity tomography determined (ERT). The images which are obtained (apparent resistivity pseudo sections) are processed by inversion the loss of 400 casualties, 2,500 damaged houses, and

500 million baht of economic loss. There were more than

250 mm of rainfall prior to the landslide movement [8].

After these events, interested areas were included in the “landslide hazard map” produced by the Land Develop-

ment Department (1:250,000 scale) [10] and associated dis-

asters from nine districts of Nan province [11]. This is the

main reason why scientists seem to consider the amount of rainfall as the main trigger mechanism for mass move-

ment. Notably, all landslide events occur in rainy season.

2 The study area FigureFigure 2. 2:LandLand use a usend land and cover land from cover the study from area, the Nan study highland area,. Ban Nan Nong high- Pla land. Bancommunity Nong consists Pla community of two small consistsvillages. One of is two located small on upper villages. part and One the other is in the lower part of the mountain. The study area lies within a small highland village named is located on upper part and the other is in the lower part of the mountain. Ban Nong Pla that is the center of the study area and is

located in a valley with a steep slope (Figure 1). The location of the village is reported as one of high potential Apart from 2006 Nanlandslide, other well-known landslide risk areas in Nan province [9]. The area is sur- landslide events in northern Thailand, in the past decades, rounded by mountains with elevation ranging from 700 has been recorded. One of these events occurred on 4 May software. The multi-electrode resistivity technique consist a multi-core cable with as many conductors (24, 48, 72, 96 and so on) as electrodes plugged into the ground at a fixed spacing, every 5 m. Three resistivity lines were carried out in this study including BNP-1, BNP-2, and BNP-3 in the upper and lower slopes of a village (Figure 4).

Measurement of soil shear strength

The shear strength of soil mass is an internal resistance per unit area. Soil mass can offer to resist failure and sliding along any planes inside it. One must understand the nature of shearing resistance in order to analyze soil stability problems such as bearing capacity, slope stability, and lateral pressure on earth-retain structure [13]. The unsaturated shear strength characteristic of residual soils as function of the degree of saturation, was studied in the laboratory using the multi-stage direct shear test. The shear strength was completed by using multi-stage loading over a range of net normal stresses and shear stress values. The physical properties of residual soil samples were analyzed in geotechnical engineering laboratory by using the testing method of American Society for Testing and Materials (ASTM) including total unit weight, water content test (ASTM D 2216), specific gravity test (ASTM D 854), sieve analysis test (ASTM D 422), hydrometer test (ASTM D 4221-11), Atterberg’s limits test (ASTM D 4318), standard compaction test (ASTM D 698-78) and permeability test (ASTM D 2434-68). Undisturbed soil samples were collected at the interface level between bed rock and residual soil for testing the shear strength parameter, cohesion and friction angle including consolidation drained test from total 8 pits (also see locations in Figure 4). Consolidation undrained test was made only from BNP ST9 because only one multi-stageAnalysis direct ofshear residual test is soil for forewarning landslide from highland area Ë 639 needed to find shear strength parameter.

After geomorphological map has been created, test

pits and shallow excavation along the prospected transects were conducted. These combined methods provide an academic mean of acquiring a very detailed record of

the complex soil conditions which often exist near to the ground surface. It is worth remembering, however, that

test pits and other outcrops can also be used for in situ testing and to obtain high quality samples. Disturbed and

undisturbed soil samples from these methods were analyzed for both physical and engineering properties.

3.2 Two-dimensional electrical resistivity

survey Figure 4. Two-dimensional resistivity line surveys (dark lines) and locations of pits and soil Figuresampling 4: (Two-dimensionalred dots) resistivity line surveys (dark lines) and locations of pits and soil sampling (red dots). Two-dimensional electrical resistivity measurement has been carried out aiming to distinguish a layer of residual soil from rock basement. Basic modes of operation include to 1,200 m above the mean sea level. Streams within sub- the profiling (mapping), vertical electrical sounding (VES), watershed flow from southeastern to northwestern direc- combined sounding and profiling (two-dimensional re- tion into the Huai Nam Puea main stream (Figure 2). Land sistivity imaging), and electrical resistivity tomography use in this area can be classified mostly as temporary agri- (ERT). The images which are obtained (apparent resistiv- culture such as short-life rice, lychee, and corn for animal ity pseudo sections) are processed by inversion software. feeding. Some areas are still dense forest without landslide The multi-electrode resistivity technique consist a multi- scar (Figure 3). Considering that the failure of slope from core cable with as many conductors (24, 48, 72, 96 and so the highland or steep terrain in northern Thailand is often on) as electrodes plugged into the ground at a fixed spac- reported to occur mostly in conjunction with flash flooding ing, every 5 m. Three resistivity lines were carried out in from heavy rainfall and in those areas where human activ- this study including BNP-1, BNP-2, and BNP-3 in the upper ities such as exploitation of land by changing soil slope for and lower slopes of a village (Figure 4). agriculture purpose are present. The Ban Nong Pla area is considered to be situated in one of the landslide risk zones. 3.3 Measurement of soil shear strength

3 The methodologies The shear strength of soil mass is an internal resistance per unit area. Soil mass can offer to resist failure and slid- 3.1 Satellite image and interpretation ing along any planes inside it. One must understand the nature of shearing resistance in order to analyze soil stability problems such as bearing capacity, slope stability, The THEOS satellite image from Geo-Informatics and and lateral pressure on earth-retain structure [13]. Space Technology Development Agency is firstly applied The unsaturated shear strength characteristic of resid- in this study for interpreting and classifying patterns of ual soils as function of the degree of saturation, was stud- landslide scars, land-use/land cover, type of common and ied in the laboratory using the multi-stage direct shear unusual agricultures, and boundary of existing forest. Ge- test. The shear strength was completed by using multi- ographical reference data were digitized with the coordi- stage loading over a range of net normal stresses and shear nates either in terms of latitude and longitude or columns stress values. The physical properties of residual soil sam- and rows. Attribute data associate a numerical code to ples were analyzed in geotechnical engineering laboratory each cell or set of coordinates and for each variable, or to by using the testing method of American Society for Test- represent actual values (900 m elevation, 25 degrees slope ing and Materials (ASTM) including total unit weight, wa- gradient) or to connote categorical data types (land-uses, ter content test (ASTM D 2216), specific gravity test (ASTM vegetation type, land cover, rock type) [12]. D 854), sieve analysis test (ASTM D 422), hydrometer test (ASTM D 4221-11), Atterberg’s limits test (ASTM D 4318), 640 Ë Thanakrit Thongkhao et al.

using both consolidated and unconsolidated, undrained

triaxial tests with pore pressure measurements. At each

stage, a single soil sample was sheared until failure, fol- lowed by an increase in confining pressure leading to

shearing [14]. Multi-stage direct shear test should be lim- ited to soil that is not sensitive to change in structure. Con-

solidated drained triaxial and consolidated undrained tri-

axial tests were performed with pore pressure measurement on both undisturbed and remolded soils [15]. This

test is not recommended for drained compression test on sensitive soils. However, there is benefit in using multi-

stage triaxial test on undisturbed residual soils mainly

composed of weathered rhyolite and granite, which have great variation in their properties [16]. The results are prac-

tically indistinguishable from conventional test.

Unlike conventional shear strength test for soils that

use several soil specimens, a multi-stage test uses a single

soil specimen and shears a sample in stage with increasing confining stress. The multi-stage test is not an ASTM

standard method for obtaining total or effective stress pa-

rameters, but has been widely used in practice. The multi-

stage test has been adapted to both triaxial and direct

shear tests, especially when there are difficulties in sam-

pling and sample preparation. It has been applied to rocks,

undisturbed submarine soils, undisturbed silty sand and

undisturbed residual soils [17]. The benefit of multi-stage

test we expected in this study includes (1) the effects of sample variability are eliminated, (2) the time required for

sample preparation and testing is minimized, and (3) the overall cost for the tests is reduced [18].

Figure 5. Boundary of residual soil and rock basements (dark dash line) from resistivity line FigureBNP 5:-1Boundary (A), BNP-2 of(B) residual and BNP- soil3 (C). and Soil rock thicknesses basements range from (dark 2- dash9 m line) from resistivity line BNP-1 (A), BNP-2 (B) and BNP-3 (C). Soil thicknesses range from 2–9 m. 4 Results

4.1 Soil thickness from two-dimensional standard compaction test (ASTM D 698-78) and permeability test (ASTM D 2434-68). electrical resistivity tomography (ERT) Undisturbed soil samples were collected at the interface level between bed rock and residual soil for testing the Ban Nong Pla village is located in a small hill bounded by shear strength parameter, cohesion and friction angle in- steep slope at higher elevation from a village and is sur- cluding consolidation drained test from total 8 pits (also rounded by deep valleys. Dense forest has been largely see locations in Figure 4). Consolidation undrained test removed for agriculture purposes. The development of was made only from BNP ST9 because only one multi-stage land exploitation tends to move to higher elevation of the direct shear test is needed to find shear strength parame- mountains, causing a reduction of dense forestry. In this ter. study, we operated two-dimensional resistivity survey to distinguish layers of residual soil from bed rock. The evaluation of the thickness of soil overlying on rock basement 3.4 Multi-stage direct shear test is important for calculating the possible volume of soil that can collapse. Three ERT lines we preformed (BNP-1, BNP-2, Multi-stage direct shear test is a method to measure shear and BNP-3, Figure 5). BNP-1 is located from the southeast- strength of undisturbed, partially saturated silty clay by ern to northwestern direction (SE-NW), BNP-2 is designed Analysis of residual soil for forewarning landslide from highland area Ë 641

Table 1: Classification of soil types according to size and grain size distribution from 11 disturbed soil samples.

Type of soil Parent rock Test Pit Gravel Sand Silt Clay Cu Cc > 2.0 2.0–0.075 0.075–0.002 < 0.002 Shale BNP-ST1 15.34% 12.64% 41.74% 30.28% Shale BNP-ST2 14.34% 13.62% 41.46% 30.58% Shale BNP-ST3 27.12% 8.92% 34.95% 29.01% Shale BNP-ST4 11.86% 9.49% 42.56% 36.09% Siltstone BNP-ST5 6.14% 14.74% 35.63% 43.49% Siltstone BNP-ST6 0.37% 13.49% 35.89% 50.25% Shale BNP-ST7 19.19% 13.94% 34.06% 32.81% Siltstone BNP-ST8 3.32% 30.55% 35.07% 31.06% Siltstone BNP-ST9 0.12% 37.71% 29.92% 32.25% Shale BNP-ST10 17.51% 11.48% 38.64% 32.37% Sandstone BNP-ST11 0.37% 65.82% 25.91% 7.9% 51.42 0.0083 Cu is Uniformity Coeflcient, Cc is Coeflcient of gradation

Figure 6. Grain size distribution of residual soil from weathered parent rocks in the area Figure 6: Grain size distribution of residual soil from weathered parent rocks in the area.

The coefficient permeability values of residual soil from sandstone, shale and siltstone from the east to theare west between (E-W) 10 and-6 to BNP-3 10-7 cm/sec. is oriented Therefore, in wheneversified basedwater from on plasticityrain fall to soilcharts surface by Unified Soil Classifica- southeastern to northwesternwith higher direction rate than (SE-NW).permeability of soil, watertion level System in soil as layer ML-CL, will CL-ML increase and and SM. then They indicate that causing a decrease in stability of soil slope, and subsequently increase the risk for landslide. The classification in type of soil from 11 disturbedthe toughness soil samples properties according to of size soil and can grain change when the mois- size distribution is summarized in Table 1. ture content increases [19]. 4.2 Physical laboratoryFrom Table tests 1, residual soil from shale consistsThe 15.22% coefficient of gravel, permeability 17.36% of sand, values of residual soil 30.98% of silt and 36.42% of clay. Residual soil from siltstone includes quantity of 7.84 % gravel, 15.95% of sand, 38.65% of silt and 37.55from% of sandstone, clay. Residual shale soil from and sandstone siltstone has are between 10–6 to Residual soil from weathered0.37% of gravel, parent 65.82% rocks of such sand, as 7.9% shale, of silt 10–7and 25.91% cm/sec. of clay. Therefore, It can be whenever concluded waterthat from rain fall to siltstone and sandstoneresidual in soil the from study weathered area includes parent rocks high such assoil shale, surface siltstone with and higher sandstone rate recognized than permeability in of soil, wa- this study area has high percentage of fine grained texture and can be qualified to poor percentage of fine drainagegrained that texture can a (Figurebsorb high 6). amount It is of charac- waterter levelduring in rain. soil Furthermore, layer will increase shear strength and then causing a de- terized as gap and uniformdecreases grades., increasing Most the ofpos residualsibility of soilsthe soil tocrease slide. in stability of soil slope, and subsequently increase in the study area showsResults a low of plasticity shear strength (Liquid test Limit, of soil atthe natural risk for moisture landslide. content The condition classification using in type of soil from consolidated drained test are shown in Table 2. The consolidated drained direct shear test in LL < 47% and Plasticity Index, PI = 15.58%). Soils are clas- natural water content condition shows a cohesion value that tends to increase when a degree of saturation is low. The cohesion of soil is between 0.223 to 0.714 ksc. Friction angle ranges from 11.51 to 25.30 degrees.

Relation between physical and engineering properties of residual soil

Total cohesion of soil and degree of saturation

From figure 7, it can be recognized that shear strength of residual soil decreases when moisture content increases. The consolidated drained test of a cohesion in soil ranges from 0.096 to 1.196 ksc of residual soil derived from shale and siltstone and 0.18 to 0.617 ksc from sandstone. Therefore, a total cohesion of soil decreases when volume moisture content in soil increases. Also, it is believed that moisture content is an important factor affecting shear strength of soil. Shear strength decreases when moisture content in soil increases. An increase in volume of water in soil will disturb surface tension, making cohesion of soil to decrease and the volume of water, then, increases. A friction between soil particles reduces to a decrease in soil shear strength that can impact directly to a stability of soil slope.

Shear strength of soil and volume of moisture content in soil mass

Analysis of engineering properties of residual soil from parent rocks was carried out for evaluating the possibility of landslide. Shear strength of soil in natural condition decreases 642 Ë Thanakrit Thongkhao et al. when moisture content increases [25]. Based on shear strength from unconsolidated drained test (from 8 test pits), the cohesion of soil ranges from 0.096 to 1.196 ksc. The angle of internal friction is Tablebetween 2: Shear 11.51 strength to 35.78 of soil atdegrees. natural moisture Consolidated content condition undrained using test the consolidated(from 1 test drained test. pits) ranges from 0.18 to 0.617 ksc and angle of internal friction is between 19.45 to 29.80 ′ degrees. Therefore, theTest monitoring pit of soil WC shear strengthDepth moisture content Degree in natural of c′ Φ conditions is necessary for analyzing soil % stability problems. (m) In addition, weatheringshear strength (ksc) (degree) parameter (i.e., cohesionTP and 1 friction angle) 22.02 should be used to2.0 analyze stability Grade model V in 0.223 14.77 order to find a plane of failureTP 2 and a safety 23.75 factor of soil slope 2.5 and this information Grade may V be 0.315 20.47 used in combination withTP geophysical 3 survey 22.85 (i.e. ERT survey). 2.5 Grade V 0.501 25.30 In this study, it has TPbeen 4 shown how shear 33.04 strength of soil 0.8 decreases when Gradethe moisture VI 0.433 23.08 content changes. Rate ofTP shear 5 strength decreases 27.57 in a range of 0.6 72.11% to 98.78%, Grade 83% V to 0.455 11.51 91.20% and 79.6% to TP 95.6% 6 for residual 26.87 soil from shale, 1.0 siltstone and Grade sandstone, VI 0.525 17.82 respectively. Strength of soil decreases when the moisture content in soil mass increases TP 7 19.98 0.8 Grade VI 0.714 15.17 continuously. Therefore, it can be concluded that a failure of natural slope in the study area TP 8 15.81 1.2 Grade VI 0.408 21.90 can occur when the moisture content between soil particles increases leading to a decrease in strength of soil.

is low. The cohesion of soil is between 0.223 to 0.714 ksc.

Friction angle ranges from 11.51 to 25.30 degrees.

5 Discussions

5.1 Types of landslide and their controlling factors

Criteria for classifying possible landslide and erosion of hillside in highland area are based on several mecha- nisms such as types of sediments, trace and shape of land-

slide and estimated amount of water involved in the land- Figure 7Figure. Total 7: Totalcohesion cohesion and and degree degree of saturationsaturation (TP ( 1,TP TP 1 4, andTP TP2 andslide TP 3) [20]. Field mapping and landslide classification were 8). applied in this study, and they are based on a principle of classification of avalanche material and type of move- 11 disturbed soil samples according to size and grain size ment [1]. Possible types of landslide in the study area can distribution is summarized in Table 1. be classified as creep and lateral spread. The two possible From Table 1, residual soil from shale consists 15.22% types reflect the diversity of factors which are responsible of gravel, 17.36% of sand, 30.98% of silt and 36.42% of clay. for their origin. Following a detailed geological field map- Residual soil from siltstone includes quantity of 7.84% ping, possible factors producing slope movements in this gravel, 15.95% of sand, 38.65% of silt and 37.55% of clay. area are classified as (1) change of slope gradient [22], (2) Residual soil from sandstone has 0.37% of gravel, 65.82% rock types and degree of weathering [23] and (3) changes of sand, 7.9% of silt and 25.91% of clay. It can be con- of water content in soil due to heavy rainfall [24]. cluded that residual soil from weathered parent rocks such as shale, siltstone and sandstone recognized in this study area has high percentage of fine grained texture and can be 5.2 Relation between physical and qualified to poor drainage that can absorb high amount of engineering properties of residual soil water during rain. Furthermore, shear strength decreases, increasing the possibility of the soil to slide. 5.2.1 Total cohesion of soil and degree of saturation Results of shear strength test of soil at natural moisture content condition using consolidated drained test are From Figure 7, it can be recognized that shear strength of shown in Table 2. The consolidated drained direct shear residual soil decreases when moisture content increases. test in natural water content condition shows a cohesion The consolidated drained test of a cohesion in soil ranges value that tends to increase when a degree of saturation from 0.096 to 1.196 ksc of residual soil derived from shale and siltstone and 0.18 to 0.617 ksc from sandstone. There-

Factor of safety, slope angle, soil thickness and degree of saturation

The behavior for soil slope failure in upstream area of Huai Nam Puea basin, Ban Nong Pla village is discussed in terms of factor of safety (FS) and slope angle. According to the undisturbed soil samples test, a failure of slope can occur where slope angle of the area is greater than or equal 25 degrees. Based on analysis in stability of soil slope using a method of infinite slope, values of slope angle that can cause a failure of slope is between 25-45 degrees. This range of slope angle is in agreement with literature values [26]. Factor of safety of cut slope is also lower than a natural slope. Critical angle of slope that can trigger landslide is 17.1 degrees, which corresponds to FS equal to 1.3. Therefore, any construction having slope angle greater than 17.1 degrees may be at risk of failure slope. Analysis of slope stability by method of infinite slope shows that factor of safety decreases when soil slope reaches a critical thickness (equal or greater than 1 m). In this area, a degree of saturation is between 60-70%. This range of saturation makes factor of safety greater than 1. However, if a degree of saturation is greater than or equal 80%, factor of safety of soil slope will be less than 1 and this can generate failure (FigureAnalysis of8). residual soil for forewarning landslide from highland area Ë 643

Figure 8: Plots ofFigure factor of safety8. Plot versuss of slope fact angleor of (left) safety and thickness versus (right).slope angle (left) and thickness (right) fore, a total cohesion of soil decreases when volume mois- 83% to 91.20% and 79.6% to 95.6% for residual soil from ture contentCONCLUSION in soil increases. Also, it is believed that mois- shale, siltstone and sandstone, respectively. Strength of ture content is an important factor affecting shear strength soil decreases when the moisture content in soil mass in- of soil. ShearLandslide strength decreases is a severe when natural moisture disaster content thatcreases often continuously.devastates highland Therefore, areas it can of be Thailand. concluded that in soilLandslide increases. Anoften increase occurs in volume after heavy of water rainfall in soil resultinga failure in of natural a loss slope of stability in the study and area equilibrium can occur when will disturbamong surface residual tension, soil making and weathered cohesion ofrock soil tobasements. de- the moisture The geological content between properties soil particles of residual increases soil lead- creasecollected and the volume in this of water, area then,includes increases. specific A friction gravity,ing grain to a decrease size distribution, in strength of liquid soil. limit, plastic betweenlimit, soil plastic particles index, reduces tota to al decreaseunit weight in soil and shear moisture content. Engineering properties includes strengthshear that strength can impact and directly multi to-state a stability direct of soilshear slope. tests. Also, a geophysical survey by means of ERT has been carried out. Results can be summarized5.2.3 as follows.Factor of safety, slope angle, soil thickness and (1) Residual soil in the study area is mainly degreederived of saturation from weathering processes of 5.2.2 dominantShear strength clastic of soilsedimentary and volume rocks of moisture including shale, siltstone and sandstone. Most of residual soilcontent is incharacterized soil mass by low plasticity (LL

6 Conclusion some significant relations among their properties. This integrated analysis of both perspectives, definitely, can be used to forewarn future landslide, not only in this area, Landslide is a severe natural disaster that often devastates but can also be extended to some other landslide high risk highland areas of Thailand. Landslide often occurs after areas that own a similarity in terms of geological and geo- heavy rainfall resulting in a loss of stability and equilib- graphical conditions. rium among residual soil and weathered rock basements. The geological properties of residual soil collected in this Acknowledgement: National Research University Project, area includes specific gravity, grain size distribution, liq- Office of Higher Education Commission provided fund to uid limit, plastic limit, plastic index, total unit weight and MC and SP (WCU-045-CC-57). Graduate School of Chula- moisture content. Engineering properties includes shear longkorn University provided M.Sc. research fund to TT. strength and multi-state direct shear tests. Also, a geo- We thanks Professor S. Soralump for discussing soil engi- physical survey by means of ERT has been carried out. Re- neering properties. Thanks are also to Editor-in-chief and sults can be summarized as follows. anonymous reviewers that improve this manuscript signif- (1) Residual soil in the study area is mainly derived from icantly. weathering processes of dominant clastic sedimen- tary rocks including shale, siltstone and sandstone. Most of residual soil is characterized by low plasticity (LL < 47% and PI = 15.58%). This means that References properties of soil changes when moisture content in- [1] Cruden D.M., Varnes D.J., Landslide types and processes. In creases. This indicates that the area may have a high landslides investigation and mitigation, USA. Transportation risk of landslide. Research Board, US National Research Council, Special Report (2) Result from two-dimensional electrical resistivity 1996, 247. survey (BNP-1, 2 and 3) shows that soil layer overly- [2] Carrara A., Guzzetti F., Cardinali M., Reichenbach P., Use of GIS ing on rock basements has a thickness between 2 to technology in the prediction and monitoring of landslide hazard, Natural Hazard 1999, 20, 117–135. 9 m. [3] Tantiwanit W., Landslide geohazard at Ban Kathun Nua, Nakhon (3) Size of soils in this area is characterized mostly as Srithammarat Province, southern Thailand, Department of Min- fine-grained, which is sensitive for sliding whenever eral Resources, Technical Report 1992, pp. 25 water is induced. [4] Soralump S., Landslide hazard investigation and evaluation (4) Based on unconsolidated undrained test, cohesion of Doi Tung Palace: Example of soil creep landslide, EIT-JSCE of soil ranges from 0.096 to 1.196 ksc and angle of in- Joint International Symposium, Monitoring & Modeling in Geo- Engineering, 15–16 September, 2008, Imperial Queen’s Park ternal friction is between 11.51 to 35.78 degrees. Con- Hotel, Bangkok, Thailand. solidated undrained test shows that cohesion of soil [5] Soralump S., Kunsuwan B., Landslide risk prioritization of ranges from 0.18 to 0.617 ksc and angle of internal tsunami affected area in Thailand, Regional Symposium on In- friction is between 19.45 to 29.80 degrees. frastructure Development in Civil Engineering, 6–9 December, (5) Shear strength of soil will reach high decrease rate 2006, Manila, Philippines. [6] Tepparnich J., Jotisankasa A., Behavior of soil slope and shear in a range of 72.11% to 98.78%, 83% to 91.20% and strength properties in the landslide-prone area of Laplae, 79.6% to 95.6% for residual soil from shale, siltstone Uttaradit, The 15th National Convention on Civil Engineering and sandstone, respectively. It reveals a relation- (NCCE15) 2009, Ubon Ratchathani, Thailand. ship between geological and engineering properties [7] Yumuang S., 2001 debris flow and debris flood in Nam Kor area, that shear strength decreases when moisture con- Phetchabun Province, central Thailand, Environmental Geology, tent changes. 2006, 51, 4, 545–564. [8] Kunsuwan B., Study of behavior for landslides on upstream ar- (6) Strength of soil decreases when moisture content in eas in Mae Nam Chantaburi Basin, Thailand, M. Eng. Thesis, soil increases. Therefore, a natural soil slope in the 2005, Kasetsart University. study area can be stable when moisture content in [9] Akkrawintawong K., Landslide hazard investigation in Chang- soil level is equal to one, but when moisture content wat Nan, M.Sc Thesis, 2008, Geology, Chulalongkorn Univer- in soil particles increases, strength of soil can be de- sity. [10] Department of Mineral Resources, Geologic map of Thung creased, then, resulting in soil strength decreases. Chang district, Nan province 1: 50,000, Thailand, 2006, Depart- An integration of geological and engineering field ment of Mineral Resources. method and laboratory testing in this paper highlighted [11] Land Development Department, Land use map and soil group map of Chiang Klang District, Nan Province scale 1: 25,000. Analysis of residual soil for forewarning landslide from highland area Ë 645

2009, Land Development Department, Bangkok. [20] Cruden D.M., A simple definition of a landslide, Bulletin of the [12] Montavani F., Soeters R., Western V., Remote sensing tech- international Association on Engineering Geology 1991, 43, 27– niques for landslide studies and hazard zonation in Europe, Ge- 29. omorphology 1996, 15, 213–225. [21] Varnes D.J., Slope movement: types and processes, In land- [13] Das B.M., Principles of geotechnical engineering, 3rd edition, slides analysis and control: National Academy of Sciences, 1994, United States of America: International Thomson Publish- Transportation Research Board, and Special Report 1978, 176, ing. 11–33. [14] Taylor D.W., A triaxial shear investigation on a partially satu- [22] Mairaing W., Landslide problems and warning by geotechnical rated soil, In ASTM. Spec.Tech. Pub. 1950, 106, 180–191. methods, 26th Conference of the ASEAN Federation of Engineer- [15] Kenney T.C., Watson G.H., Multiple-stage triaxial test for deter- ing Organization 2008, Sofitel Centara Grand Hotel, Thailand. mining c’ and f’ of saturated soils, In 5th Int. Conf. Soil Mech. [23] Mencl V., Záruba Q., Landslides and their control, 1969, Nether- 1961, 1, 191–195. lands, Elsevier. [16] Lamb T.W., Soil testing for engineers, 1951, John Wiley & Sons, [24] Thaijeamaree N., Stability of slope in Nam Kor watershed area Inc, Singapore. from soil engineering properties, northern Thailand, M. Eng. [17] Giao P.H., Weller A., Hien D.H., Adisornsupawat K., An approach Thesis 2003, Kasetsart University. to construct the weathering profile in a hilly granitic terrain [25] Soralump S., Thowiwat W., Shear strength-moisture behavior based on electrical imaging, Journal of Applied Geophysics of residual soils of landslide sensitive rocks group in Thailand, 2008, 65, 30–38. The Fifteenth National Convention on Civil Engineering (NCCE15) [18] Soralump S., Thowiwat W., Shear strength behavior of resid- 2007, Ubon Ratchathani, Thailand. ual soils in Thailand for supporting landslide warning analysis [26] Soralump S., Pungsuwan D., Chantasorn M., Inmala N., Alam- and geotechnical engineering design, The Fourteenth National bepola N.M.S.I., Landslide risk management of Patong City: Convention on Civil Engineering (NCCE14), 13–15 May, 2009, demonstration geotechnical engineering approach, Interna- Nakhon Ratchasima, Thailand. tional Conference on Slope 2010, Geotechnique and Geosyn- [19] Hazelton P., Murphy B., Interpreting soil test results: What do thetics for Slopes, 27–30 July, 2010, Shangri-La Hotel, Chiang all the numbers mean? European Journal of Soil Sciences 2007, Mai, Thailand, Organize by Department of Highways Ministry of 58, 5, 1219–1220. Transport, Thailand.