Shales and Other Degradable Materials

LChapter 21 JOHN L. WALKINSHAW AND PAUL M. SANTI SHALES AND OTHER DEGRADABLE MATERIALS

1. INTRODUCTION time frame that may be as short as only a few hours. Such a rapid breakdown is easily identified by he designers of earthwork must take precau- straightforward laboratory tests. Other rock mate- T tions when the materials at hand cannot be rials show no appreciable change in strength over classified as rock or as soils in terms of their behav- many years. Unfortunately for engineers, the prob- ior in slopes or in civil engineering works in gen- lem materials fall somewhere between these eral. In their in situ form, the geologic formations extremes of rock behavior. may have names or appearances that imply rocklike Predicting the behavior of degradable materials behavior. Once disturbed, however, some of these has been the subject of much research since the formations retain the character of rock, but others early 1960s, and this research thrust continues may degrade to soil-size particles in a time frame today. Much of this research can be tied to the that is relevant tà the long-term performance of construction of large transportation facilities. In slopes built in, on, or with these materials. the United States, the development of the The currently available methods of identifying, Interstate highway system required much higher classifying, and treating these degradable materials cuts and embankments than had been éomrnon in so as to reduce the risk of slope failure are discussed the past. Problem geologic materials that previ- in this chapter. Sedimentary rocks, which consti- ously could be economically addressed by avoid- tute the bulk of degradable materials worldwide, ance, minor mitigation, or maintenance created are discussed first. Other degradable materials, in- the need for new engineering solutions. In Central cluding weathered igneous and metamorphic America the construction of the Panama Canal rocks, are discussed in less detail. Emphasis is resulted in slope problems that continued decades placed on the successful use of these materials in after construction and are also associated with the embankments and on their treatment in the for- properties of degradable materials (Berman 1991). mation of cut slopes. In geologic terms, all of the soils and rocks in the earth's crust are "degradable" 2. GEOLOGICAL CONSIDERATIONS materials, since all materials modify over geologic time. However, on the human time scale, only Degradable materials were not considered in detail comparatively rapid degradation of a strong or hard in the 1978 landslide report, but the basic princi- rocklike material into a weaker soil-like material is ples that govern them were identified: of concern to the designer of stable slopes. Certain sedimentary rocks can exhibit a loss of strength Before one can completely comprehend the that can be of several orders of magnitude within a particular problems of stability, one must under- 555 556 Landslides: Investigation and Mitigation

stand the lithology of the physical properties third miscellaneous substances. The principal not only of the rock mass itself but of all the ma- minerals of shales, such as quartz, clay miner- terials in the mass.... A sedimentary rock seals, and hydrated oxides (such as bauxite and quence, for example, is markedly different from limonite), are formed by the weathering of an igneous series or a metamorphic complex. feldspars and mafic igneous rocks. Some asso- Each particular type is characterized by a certain ciated minerals such as calcite, dolomite, texture, fabric, bonding strength, and macro pyrite, illite, and glauconite are formed during and micro structures. The most important rock and after deposition of the primary minerals. properties are the nature of the mineral assem- (Hopkins 1988, 8) blage and the strength of the constituent minerals; a rock material cannot be strong if its Unfortunately, many of the shale particles are less mineral constituents are weak or if the strength of the bonds between the minerals is weak. than 1 .im in diameter, and consequently study of (Piteau and Peckover 1978, 194) - their mineralogy is difficult or impossible by simple visual observation. The resulting geologic field The strength of the bonds between the miner- classification of shales does not reliably relate to als is also related to the geologic history of the for- engineering properties. mation of interest. The resulting hardness is Terzaghi and Peck described very clearly the generally due to long-term consolidation under geologic processes that lead to the problem prop- external pressures and not to cementing minerals erties of shales: (McCarthy 1988). Degradable materials can be grouped according to two broad geologic sources, As the thickness of the overburden increases those derived from sedimentary rocks and those from a few tens of feet to several thousands, derived from igneous and metamorphic rocks. the porosity of a clay or silt deposit decreases; an increasing number of cohesive bonds devel- 2.1Degradable Materials from ops between particles as a result of molecular Sedimentary Rocks interaction, but the mineralogical composition of the particles probably remains practically Shales constitute about one-half of the volume of unaltered. Finally, at very great depth, all the sedimentary rocks in the earth's crust. They are particles are connected by virtually perma- exposed or are under a thin veneer of soil over a nent, rigid bonds that impart to the material third of the land area (Franklin 1981). Shales are the properties of real rock. Yet, all the materi- by far the most pervasive and problematic degrad- als located between the zones of incipient and able material. As early as 1948, Taylor stated (53): complete bonding are called shale. Therefore, "Shale itself is sometimes considered a rock but, the engineering properties of any shale with a when it is exposed to the air or has the chance to given mineralogical composition may range take on water, it may rapidly decompose." between those of a soil and those of a real rock. In the American Geological Institute's Glossary (Terzaghi and Peck 1967, 425-426) of Geology, shale is defined as

a fine-grained detrital sedimentary rock, formed These two authors further suggested using an im- by the consolidation (esp. by compression) mersion test on intact samples to obtain the relative of clay, silt, or mud. It is characterized by performance of otherwise "identical sedimentary finely laminated structure, which imparts a fis- deposits." As will be discussed in Section 3, this sility approximately parallel to the bedding.... was the direction taken by many researchers of [It is composed of] an appreciable content of clay minerals and detrital quartz. [Shale in- that time. cludes rocks such as] claystone, siltstone, and Just as increasing loads over geologic time play mudstone. (Bates and Jackson 1980, 573) an important role in the interparticle bonding of shale formations, the reverse process, unloading, Referring to Huang (1962), Hopkins noted that has significant effects. During the removal of load, typically, shales are composed of about one- "the shale expands at practically èonstant horizon- third quartz, one-third clay minerals, and one- tal dimensions" (Terzaghi and Peck 1967, 426). Shales and Other Degradable Materials 557

During expansion, the interparticle bonds are bro- chemical. Oilier (1969) described in detail several ken, and joints form at fairly regular spacings. At types of physical weathering, including sheeting or depths on the order of 30 m, the joints are spaced spalling (fracturing parallel to a free surface created meters apart and are closed. Closer to the surface, by erosion, excavation, tunneling, etc.), frost weath- intermediate joints form because of differential ering (extension of fractures by expansion of freez- movements between the blocks. These joints open, ing water), salt weathering (extension of fractures by allowing moisture to penetrate. The increase in the growth of salt crystals), and isolation (partial dis- moisture content may reduce the shear strength, integration of the rock caused by the volume and, if so, new fissures are formed. The final result changes accompanying temperature changes). and slope of any exposed face depend on the inter- Oilier (1969) also described types of chemical particle bonding remaining in the shale formation. weathering, including solution (dissolution of sol- uble minerals, particularly salts and carbonates), oxidation and reduction (chemici alteration of 2.2 Degradable Materials from Igneous minerals to form oxides or hydroxides), hydration and Metamorphic Rocks (incorporation of water to create a new mineral), Because sedimentary rocks (and shales in particu- chelation (leaching of ions such as metals), and lar) are typically formed relatively near the earth's hydrolysis (reactions between minerals and the surface and without the extreme heat and pressure component ions of water). that occur at depth, they tend to be mineralogi- As noted by Macias and Chesworth (1992), cally stable near the surface. Weathering of these chemical weathering, which brings about miner- materials then involves either a reversal of the alogical changes in igneous and metamorphic consolidation pressure or a dissolution of cement rocks, is usually more crucial than physical weath- bonds holding the grains or mineral groups to- ering in defining the strength properties of the gether. In contrast, igneous and metamorphic materials. Physical weathering, however, does rocks are created under temperature and pressure provide avenues for water to enter the rock by the conditions that are drastically different from con- creation and extension of fractures and subse- ditions at the surface. Macias and Chesworth de- quently encourages the more rapid progress of scribed the implications of the difference: chemical weathering by an increase in surface area exposed to water. One might therefore expect that they would Obviously some minerals, and therefore some weather more readily than sedimentary materi- rocks, are more susceptible than others to the als.... Generally however, expectations in this weathering processes described above. For in- regard are not flulfihled. Soils form more readily stance, Bowen's Reaction Series (Goldich 1938), on sedimentary rocks than on other trpes and which describes crystallization of magma, may be the reason is obviously hydrodynamic. For reversed to model the weathering process: calcium chemical weathering to take place to any signif- plagioclase weathers more readily than sodium icant degree, water must circulate through the plagioclase, and olivine weathers more readily rock, and the open structure of most sedimentary than biotite, which weathers more readily than materials is more conducive to this than is the muscovite, which weathers more readily than restricted porosity of most igneous and meta- quartz. Thus, rocks containing high percentages of morphic rocks. . . the igneous rocks most suscep- calcium plagioclase or olivine will weather faster tible (to weathering) are those with an open than rocks containing high percentages of sodium structure such as the non-welded pyroclastics. plagioclase or quartz. Again, in the metamorphic regime the impor- Detailed descriptions of the weathering prod- tance of hydrodynamics is shown in that a verti- cal disposition of fihiation encourages a more ucts of minerals have been provided by Macias facile descent of aqueous solutions and a more and Chesworth (1992), Oilier (1969), and Carroll rapid weathering, than a horizontal foliation. (1970). In general, most silicates (feidspars and (Macias and Chesworth 1992, 283) micas in particular) weather into clay minerals. Under more extreme conditions, such as those Weathering of igneous and metamorphic rocks is found in tropical or humid climates, or after long generally divided into two categories: physical and periods of time in a geologic sense, they break 558 Landslides: Investigation and Mitigation

down further into oxides and hydroxides of alu- The weathered rock product referred to as minum and iron. The specifiè types of clay miner- saprolite is of particular interest in evaluating engi- als formed depend to a great degree on the parent neering properties. Saprolites are "rotten rocks," materials, the pH, and the extent of saturation. or rocks in which the rock structure is preserved Because the weathering products of rocks are but many of the less durable minerals have altered largely a function of the mineralogical composi- to clay. Saprolites generally form less stable slopes tion, certain igneous and metamorphic rock types than their parent rocks because of the increased that share common minerals may share similar amount of clay and loss of interlocking structure. weathering characteristics. The following obser- They also maintain zones of weakness by preserv- vations may assist in predicting general weathering the general rock structure, or new zones of ing characteristics of igneous and metamorphic weakness may be created by preferential weather- rocks (OIlier 1969; Macias and Chesworth 1992): ing along bands of less stable minerals. Saprolites that preserve zones of weakness from Granite and diorite: Because granites typically the original rock structure or contain intact, exhibit massive structure, they also typically de- unweathered blocks may be expected to behave velop unloading fractures when exposed at the like degradable materials. Saprolites that do not surface. Continued physical weathering increases have these characteristics may be expected to be- the surface area exposed to chemical weathering: have like deeply weathered soils, whose properties Chemical weathering alters feldspars and micas are better described by a system that addresses into clays, whereas quartz persists as a sand. tropical soils, as discussed in Chapter 19. According to OIlier (1969, 81), "weathering This brief geological background clearly estab- often follows the joints, and isolated joint blocks lishes that the evaluation of degradable materials is weather spheroidally, leaving 'corestones' of un- complex and that no single approach to determin- altered granite in the center." ing the long-term behavior is likely to work for every Gneiss and amphibolite: In igneous and meta- formation. Thus, many researchers have concen- morphic rocks, feldspars and pyroxenes tend to trated on developing identification and classifica- weather rapidly, amphiboles weather at an in- tion methods that have regional applications. Local termediate rate, and quartz and accessory min- experience and understanding are keys to success erals are persistent. 011ier (1969, 82) noted that when building through, on, or with these materials. gneiss, in particular, "is rarely as well jointed as granite, so unloading is not common, or at least harder to detect. Minerals are segregated into bands, and bands of the most weatherable min- The investigative procedures for identifying or rec- eral affect the total rock strength—a property ognizing potential slope stability problems in shale that often proved troublesome in engineering." formations are similar to those described in other Schist, slate, and phyllite: OIlier (1969, 82) chapters of this report. The focus in this section will noted that "these [schists] have marked fissility be on reviewing laboratory and field tests developed along the 'schistosity' and this is very important for sedimentary rocks, shales in particular. in weathering. They contain some very resistant minerals but weathering is moderately easy. 3.1 Shales Frost weathering can rapidly break up schist." Basalt and peridotite: According to Oilier 3.1.1 Identification (1969, 81), "Basalt is attacked first along joint planes, leading eventually to spheroidal weath- From a visual reference, the natural topography of ering. All the minerals are eventually converted regions underlain by shales displays certain char- to clay and iron oxides, with bases released in acteristics. Terzaghi and Peck (1967, 426) stated: solution, and as there is no quartz in the origi- "On shales of any kind, the decrease of the slope nal rock, the ultimate weathering product is angle to its final equilibrium value takes place pri- often a brown base rich, heavy soil." Peridotite marily by intermittent sliding. The scars of the shares mineralogical characteristics with basalt slides give the slopes the hummocky, warped ap- and may be anticipated to weather similarly. pearance known as 'landslide topography." Shales and Other Degradable Materials 559

In the 1978 landslide report, Rib and Liang mon laboratory tests (moisture content, density, (1978) described the typical landforms of shale void ratio, permeability). He recognized, however, landscapes and their interpretation from aerial that developing such a scheme would be a signifi- photography. If available, aerial photographs are cant effort and he recommended (1967, 116) that excellent tools for identifying potential problem "a comprehensive study involving the cooperation sites provided that the user is trained to recognize of government agencies, private engineering firms, certain characteristics associated with the diag- and universities, is needed to produce a satisfac- nostic features. tory engineering classification for compaction For thick, uniform shale beds, Rib and Liang shales, especially the clay shales." described the associated landforms (1978, 57): In the early 1970s, major research efforts were "Clay shales are noted for their low rounded hills, under way at Purdue University sponsored by the well-integrated treelike drainage system, medium Indiana Highway Department and at the U.S. tones, and gullies of the gentle swale type." Ray Army Engineer Waterways Experiment Station (1960, 16) noted that shales "have relatively dark sponsored by the Federal Highway Administration. photographic tones, a fine-textured drainage, and Numerous reports resulted from this research, in- relatively closely and regularly spaced joints." cluding those by Bragg and Zeigler (1975), However, it has also been observed that shales Shamburger et al. (1975), Strohm et al. (1978), and are particularly susceptible to landsliding when in- Strohm (1978). Because of the widespread occur- terbedded with pervious rocks such as sandstones rence of problem shales and their almost infinite or limestones. In this case, Rib and Liang noted: variation of behavior, researchers in a number of state transportation departments, other public Interbedded sedimentary rocks show a combina- agencies, private companies, and universities have tion of the characteristics of their component continued to refine these earlier studies, to revise beds. When horizontally bedded, they are recog- the proposed tests, and to apply them to their re- nized by their uniformly dissected topography, spective areas. contourlike stratification lines, and treelike Although it has been known for decades that drainage; when tilted, parallel ridge-and-valley certain shales deteriorate rapidly upon immersion topography, inclined but parallel stratification in water, Franklin and Chandra (1972), Lutton lines, and trellis drainage are evident. (Rib and (1977), and Franklin (1981) have made signifi- Liang 1978, 57) cant contributions to establishing specific tests for slaking of shales. The tests from these studies are described briefly below. 3.1.2 Laboratory Tests and Classification Systems 3.1.2.1 Slake Test Since the late 1960s, there have been numerous at- The slake test was originally developed to provide tempts to develop tests to assist design engineers in an indication of material behavior during the the difficult task of classifying argillaceous shales stresses of alternate wetting and drying, which, to and predicting their performance in embankment some degree, simulate the effects of weathering. or cut slopes. These issues are of interest to the The test procedure and applications have been dis- mining industry as well as to transportation engi- cussed by numerous authors, including Chapman et neers. The main objective has been to find tests al. (1976), Withiam and Andrews (1982), and that will reliably differentiate between durable Hopkins and Deen (1984). The procedure is briefly shales that may be treated as rock and those with described below: limited durability that are degradable on a human time scale. Choose six pieces of shale each weighing about Underwood (1967) discussed in detail the lim- 150 g or the largest pieces available to have a itatibns of the various geological, chemical, and total of 150 g in each group. mineralogical classification methods of that time. Identify and photograph each piece beside a He suggested grouping shales according to signifi- millimeter scale. cant engineering properties (strength, modulus of Dry shale to a constant weight at 105°C and elasticity, potential swell) and according to com- record dry weight. (Note: drying the sample is 560 Landslides: Investigation and Mitigation

an important step; the following step of the test Rotate the drum at 20 revolutions per minute must not be started with a field-moist sample.) for 10 mm. Place each specimen in a separate jar and cover Remove the drum from the water, rinse, dry in with distilled water. The condition of the spec- oven, and weigh drum and remaining shale. imens should be checked for the first 10 mm, Repeat the cycle four more times to produce five then at 1, 2, 4 or 8, and 24 hr. cycles, but calculate the slake durability index Remove the specimens from the water and (Ia) after each cycle. Photograph as necessary. check for any change in pH of the water. Calculate the durability index as follows: Dry to a constant weight and record weight of shale specimens retained on 2-mm (No. 10) = weight of shale remaining inside drum < 100 sieve. (Note: recording weight of intermediate D original weight of sample cycles is desirable so that the results may be compared with those of the slake durability test, Run at least two specimens from each sample of described next.) shale and take the average of their durability indexes. Photograph specimens if significant degradation The test proposed by Franklin has been stan- has occurred or at the end of the last test cycle. dardized and is described in ASTM D-4644-87 Repeat procedure four additional times, or until (1992). In this newer test it is recommended that total degradation, to make five cycles. only two cycles be performed before the slake Calculate the slake index for each of the six durabilityindex is computed. samples and take the average: From these two tests several agencies have developed classification systems that allow them to deter- = (original weight - final weight) >< 100 mine the method or methods by which they will original weight treat shales in embankment construction. These treatments are discussed in Section 4. In addition, This simple test will usually identify the poorly several researchers have proposed classification sys- performing shales in a matter of hours. If the spec- tems and slope stability evaluations that depend on imens are quite resistant, however, this test is time a number of other laboratory tests, including jar consuming and requires qualitative judgment as to slake, rate of slaking, Atterberg limits, free-swell its performance. tests, and point-load strength. The procedures for jar slake, point-load strength, and free-swell tests are 3.1.2.2 Slake Durability Test described below. Atterberg-limit tests are common Franklin (1981) suggested a more severe and less in current engineering practice, so no procedural de- time-consuming test known as the slake durability scription is given. However, a word of caution test, which is summarized below. Obviously, shales should be expressed. Chapman et al. (1976) have that fall apart in the slake test need not be sub- noted that Atterberg limits when evaluated for jected to the slake durability test. degradable materials are often a function of the en- For the slake durability test, a wire-mesh drum ergy input and mode of preparation; thus variation made with 2-mm (No. 10) mesh is rotated while in the test can be introduced by the operator. partially submerged in a trough of water. The axis of the 140-mm-diameter drum is 20 mm above the 3.1.2.3 Jar Slake Test water surface. The following procedure for the jar slake test was described by Wood and Deo (1975): Select 10 pieces of shale (40 to 60 g each) with a total weight of approximately 500 g. A piece of oven-dried shale is immersed in Identify and photograph the group beside a mil- enough water to cover it by 15 mm. [It is im- limeter scale. portant that the shale sample be oven dried. Place the shale fragments in the drum. Weigh Lutten (1977) reported that damp material is drum and shale together. Place drum in an oven relatively insensitive to degradation in this test and dry the shale to a constant weight at 110°C. when compared with dry material.] Compute natural moisture content; then mount After immersion, the piece is observed contin- drum in trough. uously for the first 10 min and carefully during Shales and Other Degradable Materials 561

the first 30 mm. When a reaction occurs, it The reproducibility of the jar slake test was eval- happens primarily during this time frame. A uated by Dusseault et al. (1983). final observation is made after 24 hr. The U.S. Office of Surface Mining Reclamation 3. The condition of the piece is categorized (com- and Enforcement (1991) has defined "durable rock" plete breakdown, partial breakdown, no change), as rock that does not slake in water as in the jar as follows: slake, test. Welsh et al. (1991) proposed a strength- durability classification system that includes the Jar Slake point-load test to "clearly differentiate between Behavior strong-durable and weak or nondurable materials." Index Ii 1 Degrades to a pile of flakes or The principal advantages of the test are that the mud equipment can be taken to the field, irregular sam- 2 Breaks rapidly, forms many ples can be used, and it is inexpensive to perform. chips, or both On the basis of work by Olivier (1979), Welsh 3 Breaks slowly, forms few chips, et al. (1991) selected a dual-index system to cate- FIGURE 21-1 or both gorize rock into three classes (Class I, nondurable Strength-durability classification of 4 Breaks rapidly, develops several and weak; Class II, conditionally stable; and Class III, durable and strong). In addition to the jar jar slaking fractures, or both (Welsh et al. 1991). 5 Breaks slowly, develops few slake test, a graph (Figure 2 1 - 1 ) is used that plots REPRINTED WITH PERMISSION OF THE fractures, or both the results of the point-load versus the free-swell AMERICAN SOCIETY OF 6 No change test. The two tests are described briefly next. CIVIL ENGINEERS

MEN ____... ii•u•i

u•iuIUUAWHi4!UiI. I-I.I1 it' 111111 ______IIIIIIIVJ1IUI ;1it,TLI1 -,MaMa.MvMua • WMMDJ7AM4M ri•• •i 1UtT1 I ••U•U.__,F4!IJi&JAVA kUI• 1'II W,,MI1AhWA 11.1 IM UIIII MMI wzir#aiuuuI 1111111 1t14III

1111111 .VAflIiIIIrMMNWNAVNMW MVMNAWMagMUMA !• vaa'MijMM4VNi. M• ••• ___..MMMMOOOO11Ir:mvwiaru•ii 5fffffffflJ uuuuu rirnririmuuiu ___ uiiiui WIIJMII/JIfl

'X'P1i IflhiIIU4WIIEVIIMIIIII!IIUIU I,

zirni rr ii II'tAIj 562 Landslides: Investigation and Mitigation

3.1.2.4 Point-Load Test To reduce scatter in the results, the samples The history and development of the equipment should respect certain length- to-d iameter (lid) ra- and the suggested method for determining the tios: lid ~! 1.4 for cores and lid between 1.0 and 1.4 point-load index were described more completely for rock lumps when measured perpendicular to by Broch and Franklin (1972). The test was de- the loaded axis. In addition, when samples indi- veloped principally to be used in the field on rock cate to the geologist the potential for significant core or irregular lumps ranging in size from 25 to anisotropic mechanical behavior, samples should 100mm. be divided and tested in groups to measure the The point-load apparatus (Figure 21 -2) com- strengths in each direction. presses the rock sampled between the two points Since strength test results are influenced to of cone-shaped platens. The shape of the cone has some degree by the size of the specimen tested, been standardized. The radius of curvature of the Broch and Franklin (1972) proposed that the re- cone tip (5 mm) is the most critical dimension. sults be adjusted to a reference diameter of 50 mm. The angle of the cone (60 degrees) is of impor- In their paper, a number of graphs showed the size tance only if significant penetration of the cone effects reported from numerous tests on different occurs during testing. rock types. They preferred using graphs to adjust The apparatus must be equipped to measure the the equation results rather than using factors in distance, D, between the platens at failure to the formula because the formula is in stress units within an accuracy of ±0.5 mm. The load is ap- and has some theoretical justification. The graphs plied hydraulically using a small hand pump and showed that the point-load strength decreases high-pressure ram with low-pressure seals to re- with increasing diameter. Consequently, when duce inaccuracies of load measurement. The load, lumps are tested, the authors recommended using P, is determined from a gauge monitoring the hy- or preparing samples so that the initial dimension draulic pressure in the jack. A maximum-pressure d be as close as possible to 50 mm. indicator needle on the face of the gauge is neces- In order to obtain a statistically valid average sary to accurately record the maximum pressure or value of I, at least 20 samples from the same forma- load at failure. The apparatus should have a cation should be tested (Oakland and Lovell 1982). pacity of 50 kN. After both the distance D and the failure load P have been measured, the point-load 3.1.2.5 Free-Swell Test index, is determined: The free-swell test is performed on intact core or from a bulk sample sawed into a rectangular prism. is = PID2 The minimum size should be 10 times the maximum grain size or 15 mm, whichever is greater. For where P is the point load at failure and D is the direct testing purposes, NX-diameter core (54 distance between platens at failure of the sample. mm) is generally acceptable. To measure the max- FIGURE 21-2 It should be noted that for hard rock, the initial imum swell, an axis perpendicular to the bedding Point-load tested diameter d and the measured distance D are laminations is chosen. The sample is oven-dried, apparatus. essentially the same. carefully measured, and placed in water, and the volumetric strain is computed from measurements taken after 12 hr of soaking. Olivier (1979) reported that at least 75 percent of the maximum free swelling occurs within the first 2 to 4 hr of the test. Welsh et al. ( 199 1) reported that for Appala- chian shales the proportion of swelling ranged from 80 to 99.1 percent, with an average of 90.8 percent by the end of 4 hr. Consequently, they recommended using the shorter-term test for those shales and approximating the 12-hr results by mul- tiplying the 4-hr results by 1.1. fl' 1flfl?I ., - L In the proposed classification shown in Figure 21-1, durable and strong rock (Class III) has a Shales and Other Degradable_Materials 563 swell of 4 percent or less and a rock strength equal 3.2 Igneous and Metamorphic Rock to or greater than 6 MPa and exhibits rocklike behavior during the jar slake test (ranking higher Little research has been undertaken to quantify than 2). how the degree of weathering of igneous and meta- In this classification system, any shales with morphic rock affects their engineering properties. properties less than those of Class III should not Cawsey and Mellon (1983) provided an overview of research in experimental weathering of basic ig- be used in drain applications. Therefore, particular neous rocks. They noted the merits and weaknesses care must be taken to avoid placing these materi-. of various tests to reproduce the effects of weather; als in or near drainage features used in landslide ing; Dearmän (1976) discussed the use of a weath- mitigation works. ering classification in the characterization of rock, Class I material, nondurable and weak rock, has and Dearman et al. (1978) provided an evaluation the following characteristics: of engineering properties based on a visual classification applied to granites (Table 21-1). Fails the jar slake test (behaves like a soil), To some degree weathered igneous and meta- Fails during sample preparation for either the morphic materials may be characterized by the free-swell test or point-load test, same tests used to characterize shales (Section Produces a value less than 2 MPa in the point- 3.1.2). This is particularly true for igneous and load test, or metamorphic rocks, which have an abundance of Has a free swell greater than 4 percent. clay minerals and few core stones or unweathered blocks. However, a number of differences should "Hard" shales, as defined by these simple tests, be highlighted: are not all without problems. As discussed by Strohm et al. (1978), the water in the slake dura- Researchers of degradable rocks have investigated bility and jar slake tests should be checked for pH. shales and designed their classification and test- A pH less than 6.0 indicates an acid condition, ing systems to apply to shales. Correlations with and the shale mineralogy should be checked for tests applied to weathered igneous and metamor- minerals that can cause chemical deterioration phic rocks have not been established. (Shamburger et al. 1975). Chemical deterioration Weathering rates, and therefore long-term en- of hard shales in Virginia with I > 90 percent was gineering behavior, depend highly on the original mineralogy, all other factors, including studied by Noble (1977), who soaked samples in climate, being equal. Consequently, fresh rock dilute solutions of concentrated sulfuric acid (18 cuts may weather dramatically near the surface. M) and distilled water as a classification test. He Similarly, increasing the exposed surface area of found that a 25 percent solution was more reac- the rock fragments by excavation and crushing tive and gave the same ranking in degree of dete- before their placement as fill may also acceler- rioration as the modified sulfate soundness test ate subsequent weathering rates. (ASTM C88). Control of water not only will improve pore; Noble recommended that hard, dark-colored pressure characteristics (as in shales) but will shales be checked for iron sulfides and chlorite as also reduce weathering rates, further improving a clay mineral, since this combination can have longer-term behavior. great potential for rapid weathering. Upon oxidation and access to water, shales containing iron ENGINEERING DESIGN sulfides (e.g., as pyrite) produce sulfuric acid, CONSIDERATIONS which dissolves the chlorite. These chemical soaking tests should be considered in classifying Using the laboratory tests described above, several hard shales contemplated for rock fill on impor- researchers have proposed procedures for cut-slope tant projects where long-term settlement must be and embankment design and construction using kept to a minimum. In contrast to acid reaction, degradable materials. Although these proposals do some shales in the. western United States have dis- not have the benefit of a wealth of engineering persive tendencies (Shamburger et al. 1975) and precedent and experience, they provide general may react adversely in alkaline (high pH) water. guidelines. 564 Lands/ides: Investigation and Mitigation

A preliminary evaluation of the characteristics 4.1 Embankment Design of shales as indicated by a number of standard laboratory tests has been described by Underwood Numerous workers have sought to correlate labo- (1965) and is shown in Table 21-2. This evalua- ratory tests with construction design parameters tion may be applied with care to nonshale degrad- for degradable materials. Much of the early work able rock materials, done at the U.S. Army Engineer Waterways

Table 21-1 Rock Mass Properties of Weathered Granites and Gneisses (modified from Dearman et al. 1978)

ENGINEERING FRESH, SLIGHTLY MODERATELY HIGHLY COMPLETELY RESIDUAL SOIL, PROPERTY I WEATHERED, II WEATHERED, III WEATHERED, IV WEATHERED, V VI Foundation Suitable for Suitable for Suitable for small Suitable for Suitable for low Generally conditions concrete and concrete and concrete structures, earthfill dams earthfill dams unsuitable earthfill dams earthfill dams earthfill dams Excavatability In general, In general, Generally blasting Generally Scraping Scraping blasting blasting needed, but ripping ripping and/or necessary necessary may be possible scraping depending upon the necessary jointing intensity Slope designa 1/4:1 H:V 1/2:1 to 1:1 H:V 1:1 H:V 1:1 to 1.5:1 H:V 1.5:1 to 2:1 H:V 1.5:1 to2:1 H:V Tunnel Not required Not required Light steel sets on Steel sets, partial Heavy steel sets, Heavy steel support unless joints are unless joints are 0.6- to 1.2-rn lagging, 0.6-to complete lagging sets, complete closely spaced closely spaced centers 0.9-rn centers on 0.6- to 0.9-rn lagging on or adversely or adversely centers; if tunneling 0.6- to 0.9-rn oriented oriented below water table, centers; if possibility of soil tunneling flow into tunnel below water table, possibility of soil flow into tunnel Drilling 75, usually 90 75, usually 90 50-75 0-50 0 or does not 0 or does not rock quality apply apply designation (RQD), % Core recovery 90 90 90 Up to 70 if a 15 as sand 15 as sand (NX), % high percent of core stones; as low as 15 if none Drilling rates 2-4 2-4 8-10 8-10 10-13 10-13 (m/hr) 5-7 8 12-15 12-15 17 17 (diamond NX), 2'/z-in.

Permeability Low to Medium Medium High Medium - Low medium to high to high Seismic velocity 3050-5500 2500-4000 1500-3000 1000-2000 500-1000 500-1000 (m/sec) Resistivity' 340 240-540 180-240 180-240 180 180 (ohm-rn)

NOTE: Weathering grades (shown here as column headings) are based on Dearman (1976). Benches and surface protection structures are advisable, particularly for more highly weathered material. The presence of through-going adversely oriented structures is not taken into account. b Tends to be determined by joint openness and water-table depth. Shales and Other_Degradable_Materials 565

Experiment Station for the Federal Highway 4.1.1 Benching and Drainage Administration was summarized into technical guidelines by Strohm et al. (1978), who divided Typical recommended benching and drainage pro- the design of shale embankments into five areas: visions are shown in Figures 21-3 and 2 1-4, where foundation benching, drainage provisions, mate- both longitudinal and transverse (cut-to-fill) situ- rial usage, compaction requirements, and slope in- ations are illustrated for shales interbedded with clination. The following discussion utilizes a sandstone. Obviously the benches must be made similar grouping. into stable ground and the drainage rock must be

Table 21-2 Engineering EvalUation of Shales (modified from Underwood1965 and Wood and Deo 1975)

PHYSICAL PROPERTIES PROBABLE IN Srw BEHAVIOR

LABORATORY Trs AND UNFAVORABLE FAVORABLE HIGH Low TENDENCY SLOPE Tua IN SITU RANGE OF RANGE OF PORE BEARING TO STABILITY RAPID RAPID SUPPORT OBSERVATIONS VALUES VALUES PRESSURE CAPACIIY REBOUND PROBLEMS SLAKING EROSION PROBLEMS Compressive 50 to 300 300 to 5,000 X X strength (psi) (0.3 to 2 MPa) (2 to 34 MPa) Modulus of 20,000 to 200,000 to X X elasticity (psi) 200,000 (140 2 x 10-6(1400 to 1400 MPa) to 14,000 MPa) Cohesive 5 to 100 (0.03 100 to> 1,500 X X X strength (psi) to 0.7 MPa) (0.7 to> 10 MPa) Angle of internal 10 to 20 20 to 65 X X X friction (degrees) Dry density (pcf) 70to110(1.1 110to160(1.8 X X(?) tol.8g/cm3) to2.6g/cmt) Potential 3 to 15 1 to 3 X X X X swell(%) Natural moisture 20 to 35 5 to 15 X X content (%) Coefficient of 10 to 10_10 >0_5 x x x permeability (3 x 10 to (>3 x 10 (cmjsec) 3 x 10 2 ft/sec) ftjsec) Predominant Montmorillonite Kaolinite or X X clay minerals or illite chlorite Activity ratio = 0.75 to >2.0 0.35 to 0.75 X (plasticity index/ clay content) Wetting and Reduces to grain Reduces to X X drying cycles sizes flakes Spacing of rock Closely spaced Widely spaced X X X(?) X defects Orientation of Adversely Favorably X X X rock defects oriented oriented State of stress > Existing = Overburden X X X overburden load load Expected problems indicated by X. FIGURE 21-3 Longitudinal bench drainage tailored to LONGITUDINAL stratification of DITCH UPSLOPE seepage layers (modified from EXCAVATION LINE Strohm et al. 1978).

SHALE EMBANKMENT SOIL AND ROCK NON-PERFORATED COLLECTOR PIPE .(IQCALINE ORIGINAL GROUND LINE - - - ..O r------. -- -. -- -- -:

- - - - I . • II - . -- - - - ------

ROCKUNEDOR S CONCRETE PAVED DITCH. Legend EEP LIMESTONE (OR SANDSTONE) SHALE 4— SEEPAGE FLOW

ORIGINAL GROUND LINE V -

ROCK LINE GRADE LINE PROFILE SUBGRADE

CROSS VALLEY (THRU) SHALE EMBANKMENT

ALTERNATEROCK DRAINAGE PAD 0 . (INSTEAD _-\t=--

- 4. __ - FIGURE 21-4 Transverse bench LEGEND drainage tailored to []] SANDSTONE (OR LIMESTONE) stratification of SHALE seepage layers, 4— SEEPAGE FLOW cut-to-fill (modified from Strohm et al. 1978).

Shales and Other Degradable Materials 567

durable and not degrade with time. Requirements ID > 901 I = 6: These materials can be used as for this rock might follow the recommendations of rock fill as long as soil- and gravel-size materials Welsh et al. (1991). A typical design is shown in do not exceed 20 to 30 percent of total lift. Too Figure 21-5, reproduced from a report by the much fine material prevents the rock-to-rock Oregon Department of Transportation (Machan contact necessary for stability and causes long- et al. 1989). Here the select durable rock embank- term settlement. ment is placed to a level above the flood stage of I = 60-90, I, = 3-5: These are hard, nondurable an adjacent river, essentially eliminating the po- intermediate shales that require special treat- tential for wet and drying cycles in the shale em- ment, typically including a high degree of com- bankment material 'placed above. paction by heavy rollers (see Section 4.1.3). 1< 60, 1, < 2: These materials need to be com- 4.1.2 Material Use pacted as soil in thin lifts.

During project development a material use plan A number of other authors have prepared should be prepared to cover excavation require- graphs or tables to help designers prepare the ma- ments and ensure that nondurable rocks are terial plans. Lutton (1977) provided an estimate placed as soils in thin lifts and that durable rocks of allowable lift thicknesses as a function of slake are placed as rock fill. The alternative is to place durability index (Figure 21-6). The shale rating all materials as soils while meeting maximum gra- system proposed by Franklin (1981) (Figure 2 1-7) dation size and minimum compaction criteria. groups rock materials according to slake durability Wasting of degradable materials, such as shales, index and either plasticity index or point-load can generally be avoided by proper treatment and strength. Various groupings are assigned a shale use. Exceptions might involve extremely wet clay rating, R, which is used to derive a number of shales, which may not be economically dried by slope design parameters. For instance, Figure 21-8 disking or other means. provides an estimate of allowable lift thicknesses Strohm et al. (1978) developed design criteria based on Franklin's shale rating. based on the slake durability index, 'D' and the jar Sand and Rice (1991) used a modification of slake index, I, as follows: the slake index test to provide a preliminary clas-

Embankment Foundation Excavation (Shown Hatched)

I , .\. Subgrade Elevation 309 m - im — - - - A.( Select Rock Embankment - - - - 1:1 14— 3 Existing Ground 4— 3 m— Excavate to FIGURE 21-5 Firm Materials 20 mm Drain Pipe Oregon benching Gravel Drain and drainage detail Filter Fabric (modified from Machan et al. 1989). 568 Landslides: Investigation and Mitigation

SE17LEMENT ALTERNATIVE CRITERION CONSERVATIVE CUTOFF sification of degradable materials. As is shown in zI— / Figure 2 1-9, they suggested plotting the one-cycle Uj 100 l slake value (representing the current state of cc NO MLOR PFJOBLEM, I7 T weathering of a material) against the difference 80 JEWTMINORPçX4._EMS / ]II 0 between the five-cycle and one-cycle values (rep- x resenting the susceptibility to weathering of a ma- Lu 60 MAJOR - terial). They term this difference the slake PROBLEMS z differential. The resultant graph is then divided 40 into sections denoting expected material behav- —J m ior. Such a classification emphasizes the con- 20 tinuum between soil and rock behavior. It also a itr7g/ _ indicates subsequent laboratory tests that are w 0 likely to further characterize the material. 0 0.25 0.50 0.75 1.00 1.25 Hopkins (1984) tested Kentucky shales exten- (I) LIFT THICKNESS, m sively in order to correlate slake durability with the California bearing ratio (CBR) used by the state of LEGEND Kentucky for pavement design. Mathematical ex- WESTERN U.S. SHALES pressions were developed to define suitable rela- o EASTERN U.S. SHALES tionships for design. These expressions convert three different indexes for slake durability to pre- FIGURE 21-6 dictions of the Kentucky CBR values. (above) Criteria for evaluating embankment construction on basis of slaking behavior of 50 Shale Rating,R materials (modified ) 10 from Lutton 1977). 1 8 2 5 6'< 6 40 N 4 ;30

FIGURE 21-7 Shale rating chart. a- 20 Sample of shale is assigned rating depending on its 0.6 slake durability 10 and strength if slake durability 0.4 is > 80 percent, or depending on its slake durability 0 0.2 and plasticity if 20/\ 40 60 slake durability is Slake Durability Index (%, measured <80 percent after the Second cycle) (modified from Franklin 1981). 80 85 90 95 100 Shales and Other Degradable Materials 569

4.1.3 Compaction I I I I I I

In the contract documents for a construction proj - 0.8 ect, it is generally good practice to call for test sec- LIFT MAJOR tions or field-scale tests to evaluate construction THICKNESS 0.7 PROBLEMS (m) materials and methods. The test sections are used 0.6 to develop the required compaction methods and control procedures for nondurable shales before 0.5 major earthwork is started. 0.4 Most specifications require a minimum density . t FEWPROBLEMS MIN0R of 95 percent of the maximum dry density as de- 0.3 termined by AASHTO T-99 and a moisture content within 2 percent of optimum. Some PLASTIC ROCK-LIKE specifications make a point of indicating a value CLAY- SHALES 2.2 SHALES below optimum. However, by their very nature, RETAIN WATER these materials degrade excessively during labora- COMPACTED tory processing and compaction. This degradation FIELD 20 can result in unrealistically high dry densities. The DENSITY (tons/cubic tests should be started with the in situ or natural meter) water content, since that is what will be used on EASY TO BREAK 1.8 ///7 the project. It is also necessary to use fresh sample fWN material for each determination of moisture content because of material degradation. In the field it has been found that generally it is 1.6 necessary to use two different types of rollers to obtain the specified density. A static or vibratory 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 sheepsfoot roller weighing around 25 000 kg is SHALE RATING

FIGURE 21-8 (above) 100% Tentative correlations among shale quality, lift thicknesses, and 80% compacted densities SLAKE DIFFERENTIAL . (modified from (5 cycles- 1 cycle) SUSCEPTIBILITY TO WEATHERING Franklin 1981). 60% . : D. Weathering continuum (will lose strength with • continued weathering) F. Controlled by other factors .

B. Slake durability may better characterize (or apply rock 20% C. Treat as soil with classification system) high gravel content FIGURE 21-9 . Implication of slake differential partitioning. 0% 20% '\40% 60% 80% 100% Characteristics of A. Treat as rock (or apply rock\ classification system to account for fractures) 's_ E. Highly weatherable components each category and \ already broken down appropriate testing (little loss of strength with continued weathering) may be estimated 1 CYCLE SLAKE as shown (modified STATE OF WEATHERING from Santi and Rice 1991). 570 Landslides: Investigation and Mitigation

needed to break down large rock fragments. Two to dation to thick, blocky fragments]. Where chip four or more passes may be needed, typically fol- slaking was dominant [degradation to thin, flat lowed by a 46000-kg pneumatic-tired roller for an- segments], the mass appeared to be relatively other two to four passes, which compacts the stable. The chips form an interlocking matrix now-soil-like materials. Loose lift thicknesses are which is resistant to bulk movement. When normally specified in the 0.2- to 0.3-rn range. The slaking to inherent grain size [degradation to quantity of water to be added or dried off by disking fine-grained particles] was found to be the pri- must also be evaluated. All of the above considera- mary mode, stability problems were observed, tions support the test-section approach for deter- as evidenced by slips, slides, and similar features. (Perry and Andrews 1982, 27) mining the proper procedures to use in the field with the equipment provided by the contractor. In addition to the observations on mass stability, Perry and Andrews (1982) related slaking to ero- 4.1.4 Slope Design sion problems. Sheet, nIl, and gully erosion were observed to occur to varying degrees on all spoils. In this discussion of embankment slope design the Where a high proportion of materials that slaked conventional procedures outlined in Chapter 13 to their inherent grain size was encountered, the of this report have been modified. Two figures most severe erosion was observed, whereas spoils from Franklin (1981) have been included. Figure with a high percentage of slake-resistant rocks 21-10 uses the shale rating, R, obtained from were least affe4ted. A "pebble pavement" created Figure 21-7 to estimate a range of values for cohe-. by an armoring of the surface with resistant small sion and angle of internal friction. Figure 21-11 chunks was observed to be effective in controlling FIGURE 2 1-10 provides an estimate of allowable slope angles and sheet erosion. Trends in shear- embankment heights as a function of R. strength parameters As a continuation of his earlier study, Hopkins of compacted shale Perry and Andrews related the mode of slaking (1988) performed numerous tests on some 40 dif- fills as function of to slope stability problems observed in mine spoils ferent shales in an attempt to present predictive shale quality ranging in age from 2 to 10 years: equations of engineering parameters from various (modified from index tests. For instance, he found that the natural Strohm et al. 1978 Little or no stability problems were found water content of an unweathered shale was a good and Franklin 1981). where slab or block slaking dominated [degra- predictor of important engineering properties. SHALE RATING, R In addition, Hopkins selected nine types of tn shales for triaxial testing on remolded specimens 250 compacted to standard-, modified-, and low-energy compaction. The behavior of these complex materials required experience and engineering judg- 200 ment for interpretation of the results. Hopkins stated:

Since 4)' and c' values defined by the (a' 1 l(F'3) 150 failure criterion are generally higher and lower, 0 U) respectively, than values obtained from the w I ((;' - a' 3) failure criterion, then it is unclear 0 o 100 which set of 4)' and c' to use in a given stability analysis. (Hopkins 1988, 105)

The principal difference lies in the values of c' ob- 50 tained, which can have a significant influence on the value of the safety factor computed for the slope. Because of this, Hopkins r&commended that designers use both sets of parameters in their sta- 10 15 20 25 30 35 bility analyses to determine which is the most ANGLE OF INTERNAL FRICTION (DEGREES) conservative. In their reports, designers were en- Shales and Other Degradable Materials 571 couraged to give the method used in the defining Embankment Height (m) 4)' and c' obtained from their laboratory tests. 5 10 15. 20 25 30 3H:2V- 4.2 Cut-Slope Design

The design of cut slopes in degradable materials, particularly in clay shales, is as complex as their 30° R = 8-9 geology, which was briefly outlined in Sections 2.1 and 2.2. As described by Terzaghi and Peck (1967, 426), upon unloading, the shales expand horizon- 2H:1V- - tally, take on moisture, decrease in strength, and 25° consequently are notorious for delayed and pro- 77---R=6-8 gressive failures. Duncan and Dunlop (1969) examined the influence of the initial stresses on the stresses near the 20° excavated slope using the finite-element method. 3H:1V------R =4-6 In particular, their analyses were conducted to study a- the behavioral differences of the excavated slope in 0 (1) 15° materials with low and high initial horizontal 4H:1V- - stresses. Other studies have shown that in heavily overconsolidated clays, the horizontal stresses may 5H:1V-1 be 1.5 to 3 times the current overburden pressures 00 7 ------R=0-4 at the site and that in the Bearpaw shale and simi- 6H:1V- - lar rocks of western North America, horizontal 7H:1V- - stress has been measured at 1.5 times the overbur- 8H:1V- - den pressure at a depth of 20 m. 5° Duncan and Dunlop concluded:

Shear stresses around excavated slopes are much larger for conditions representing a 0° heavily overconsolidated clay (high initial 0 10 20 30 40 50 60 70 80 90 100 horizontal stresses) than for conditions repre- Embankment height (ft) senting a normally consolidated clay (low initial horizontal stresses). The results indicate that shear stresses large enough to cause failure may develop at some points within the slopes, The number of variables is obviously quite FIGURE 21-11 Trends in stable even though the factor of safety according to large, and only a few examples are illustrated here. Franklin (1981) addressed this topic. Figure 21-12, embankment slope the usual 4 = 0 method is considerably greater angle as a function reproduced from Franklin's paper, provides an es- than unity. (Duncan and Dunlop 1969, 489) of embankment timate of allowable cut-slope angles as a function height and quality Great care must therefore be used when determin- of the shale rating R and the bedding and joint of shale fill ing shear strengths of intact shale specimens and orientations. State highway departments typically (modified from when using these values in slope stability calcula- address cut-slope design by using standard plans Franklin 1981). tions. The shear-strength envelopes must be based included in their contracts. These are used to on residual strength values from tests advanced to guide the excavation process and locate benches. large strains. Typically, the top surface of each bench is located In view of the difficulties in obtaining and test- at the top of the degradable materials. Several ex- ing samples representative of the materials in a amples from the Kentucky Department of large slope, many engineering firms and trans- Highways Geotechnical Manual (1993) are shown portation departments invoke the local "experi- in Figures 21-13 through 21-19. If rock fall is an- ence factor" in the design of slopes for minimal ticipated, the reader should refer to Chapter 18, maintenance over the long term. Stabilization of Rock Slopes. 90 HIOV Original Groundline 80 --- H:3V Base of Weathered Rock Zone ,4.6 m OB 70

H:2V 60 Class I Nondurable Shale 0) 0 5.5 m lB 0 l:IV

1)) H:2V 30 5.5 m IS H:1 V Coal 20 IH:1V Note: H:1 V lB Intermediate Bench 10 OB = Overburden Bench H:1 V

0 meters 0 1 2 3 4 5 6 7 8 9 Roadside Ditch Bench Shale Rating FIGURE 21-12 Trends in stable cut-slope angle as function of character of FIGURE 21-14 shale (modified from Franklin 1981). Typical slope configuration in Class II nondurable shale (modified from Kentucky Department of Highways 1993).

Original Groundline Original Groundline Base of Weathered Rock Zone _- -;-- Base of Weathered Rock Zone 4.6 m 05 0 2 Class Ill Nondurable Shale meters 1 Class II Nondurable Shale 5.5 m lB FIGURE 21-13 Typical slope configuration in Class Ill nondurable E shale (modified from Kentucky Department of Highways 1993). 5.5 m lB Coal

Note: lB Intermediate Bench OB Overburden Bench 0 4 LJ Roadside Ditch Bench meters

FIGURE 21-15 Typical slope configuration in Class I nondurable shale (modified from Kentucky Department of Highways 1993).

Shales and Other Degradable Materials 573

FIGURE 21-16 (left) Original Groundline Typical slope configuration in durable shale (modified from Kentucky Department of Highways 1993). Base of Weathered Rock Zone 4.6 m OB

_It FIGURE 21-17 5.5mB (below left) Typical slope configuration in massive limestone or sandstone / Durable Shale (modified from Kentucky Department of Highways 1993).

FIGURE 2 1-18 (below) Typical slope configuration for serrated slopes, which are utilized as means of controlling erosion and establishing Note: vegetation on material that can be excavated by bulldozing lB Intermediate Bench OB Overburden Bench or ripping (modified from Kentucky Department of 5.5 m lB Highways 1993). Coal o 4 meters m

Roadside Ditch Bench Slope

Original Groundline Topftof Rock 1 mStep Riser

1 rn Step Tread Staked Slope Line Note: 1:1 slope configuration shown. For Original Groundline a 3H:2V slope (not shown) use 0.6-rn riser with a 1-rn tread or a 1.2-rn riser Base of Weathered Rock Zone 4.6 m OB with a 2-rn tread. 0 2 meters

0 . . . . .•.•.• Sandstone

6mlB Coat L 32 m

Note: lB Intermediate Bench 08 = Overburden Bench > S. CONCLUSION CY 4 Limestone 16 m meters It should be clear that the topic of constructing with, in, or through degradable materials is a complex one. There are few absolutes, and one must 6 m lB trust the experience of others. A great deal of re- Shale liance is placed on knowing the behavior of slopes

m in the vicinity and the geologic conditions at the Limestonecy site. It is hoped that the information in this chapter along with that in Chapter 15 on rOck slope stability analysis will provide guidance for a suc- Roadside Ditch Bench cessful project. 574 Landslides: Investigation and Mitigation

FIGURE 21-19 600 Through cut with CUT S ABILF V Sta. 1 +546 Core Log STA 1 54681 rn At. dipping bedding ORIGI AL SC LE: 1:200 planes (modified 580 from Kentucky C re Log ST 1+546 2 m Lt. . - c- Department of - .-çI iterpreted Weathere Highways 1993). ockZone - -. 560

ae of 540

Vu 520 ° R ck= hmentrDitO. 500 100 80 60 40 20 0 20 40 . 60 80 100 NOTE: Centerline Bearing = N 64°W N 55°W' Strike of Bedding = /ECROD - True Dip = 24°SE 7 -_ Apparent Dip X-Section = 24°SE / , 96(8) Apparent Dip Centerline +40 = .L,.. - is 03 20(1)

10 00 41 (2)VerlicoJF,00ture ROD: Rock Quality Designation @ 8.7/n REC: % Recovery from 50 94 L) 80(2) / Geologist's Log 71(3) / SDI: Slake Durability Index 8 96 45(2(/ ). -

REFERENCES Carroll, D. 1970. Rock Weathering. Plenum Press, New York, 203 pp. ASTM. 1992. Standard Test Method for Slake Cawsey, D.C., and P. Mellon. 1983. A Review of Durability of Shales and Other Similar Weak Experimental Weathering of Basic Igneous Rocks. Rocks. ASTM Designation D-4644-87. In ASTM In Residual Deposits: Su7face Related Weathering Book of Standards, Volume 4.08, Soil and Rock; Processes and Materials (R.C.L. Wilson, ed.), Black- Dimension Stone; Geosynthetics, ASTM, Philadel- well Scientific Publications, Oxford, pp. 19-24. phia, Pa., pp. 951-953. Chapman, D.R., L.E. Wood, C.W. Lovell, and W.J. Bates, R.L., and J.A. Jackson, eds. 1980. Glossary of Sisiliano. 1976. A Comparative Study of Shale Geology, 2nd ed. American Geological Institute, Classification Tests and Systems. Bulletin of the Falls Church, Va., 751 pp. Association of Engineering Geologists, Vol. 13, No. 4, Berman, G. 1991. Landslides on the Panama Canal. pp. 247-266. Dearman, W.R. 1976. Weathering Classification in the Landslide News (International Newsletter of the Characterization of Rock: A Revision. Bulletin of the Japan Landslide Society), No. 5, Aug., pp. 10-14. International Association of Engineering Geologists, Bragg, G.H., Jr., and T.H. Zeigler. 1975. Evaluation Vol. 13, pp. 373-38 1. and Remedial Treatment of Shale Embankments Dearman, W.R., F.J. Baynes, and T.Y. Irfan. 1978. (Report FHWA-RD-75-62). In Desiga and Engineering Grading of Weathered Granite. Constmction of Compacted Shale Embankments, Engineering Geology, Vol. 12, pp. 345-374. U.S. Department of Transportation, Vol. 2, 245 pp. Duncan, J.M, and P. Dunlop. 1969. Slopes in Stiff- Broch, E., and J.A. Franklin. 1972. The Point-Load Fissured Clays and Shales. Journal of the Soil Strength Test. International Journal of Rock Me- Mechanics and Foundations Division, ASCE, Vol. chanics and Mining Sciences, Vol. 9, pp. 669-697. 95, No. SM2 (March), pp. 467-492. Shales and Other Degradable Materials 575

Dusseault, M.B., P. Cimolini, H. Soderberg, and D.W. Olivier, H.J. 1979. A New Engineering-Geological Scafe. 1983. Rapid Index Test for Transitional Rock Durability Classification. Engineering Geol- Material. Geotechnical Testing Journal, ASCE, ogy, Vol. 14, pp. 255-279. Vol. 6, No. 2, pp. 64-72. 011ier, C.D. 1969. Weathering. American Elsevier, Franklin, J.A. 1981. A Shale Rating System and New York, 304 pp. Tentative Applications to Shale Performance. In Perry, E.E, and D.E. Andrews. 1982. Slaking Modes Transportation Research Record 790, TRB, National of Geologic Materials and Their Impacts on Research Council, Washington, D.C., pp. 2-12. Embankment Stability. In Transportation Research Franklin, J.A., and R. Chandra. 1972. The Slake Dura- Record 873, TRB, National Research Council, bility Test. International Journal of Rock Mechanics Washington, D.C., pp. 22-28. and Mining Science, Vol.9, pp. 325-341. Piteau, D.R., and EL. Peckover. 1978. Engineering of Goldich, S.A. 1938. A Study in Rock Weathering. Rock Slopes. In Special Report 176: Landslides: Journal of Geology, Vol. 46, pp. 17-23. Analysis and Control (R.L. Schuster and R.J. Hopkins, T.C. 1984. Relationship Between Kentucky Krizek, eds.), TRB, National Research Council, CBR and Slake Durability. University of Kentucky Washington, D.C., pp. 192-225. Research Report UKTRP-84-24, Lexington. Ray, A.G. 1960. Aerial Photographs in Geologic Hopkins, T.C. 1988. Shear Strengths of Compacted Interpretation and Mapping. U.S. Geological Sur- Shales. University of Kentucky Research Report vey Professional Paper 373, 230 pp. UKTRP-88- 1, Lexington. Rib, H.T., and Ta Liang. 1978. Recognition and Hopkins, T.C., and R.C. Deen. 1984. Identification Identification. In Special Report 176: Landslides: of Shales. Geotechnical Testing Journal, ASTM, Analysis and Control (R.L. Schuster and R.J. Vol. 7, No. 1, pp. 10-18. Krizek, eds.), TRB, National Research Council, Huang, WT. 1962. Petrology. McGraw-Hill, New York. Washington, D.C., pp. 34-80. Kentucky Department of Highways. 1993. Geotech- Santi, P.M., and R.H. Rice. 1991. Preliminary nical Manual. Department of Highways, Trans- Classification of Transitional Materials. In Proc., portation Cabinet, Commonweath of Kentucky, 27th Symposium on Engineering Geology and Frankfort, 157 pp. Geotechnical Engineering, Logan, Utah, April 10-12, Lutton, R.J. 1977. Slaking Indexes for Design (Report Idaho State University, Pocatello, pp. 8-1 to 8-13. FHWA-RD-77-1). In Design and Construction of Shamburger, J.H., D.M. Patrick, and R.J. Lutten. Compacted Shale Embankments, U.S. Department 1975. Survey of Problem Areas and Current of Transportation, Vol. 3, 94 pp. Practices (Report FHWA-RD-75-61). In Design Machan, G., T Szymoniak, and B. Siel. 1989. Evalu- and Construction of Compacted Shale Embank- ation of Shale Embankment Construction Criteria. ments, U.S. Department of Transportation, Vol. 1, Experimental Feature Final Report OR 83-02. 292. pp. Oregon Department of Transportation, Salem. Strohm, WE., Jr. 1978. Field and Laboratory Investi- Macias, F., and W. Chesworth. 1992. Weathering in gations, Phase III (Report FHWA-RD-78-140). Humid Regions, with Emphasis on Igneous Rocks In Design and Construction of Compacted Shale and Their Metamorphic Equivalents. In Weath- Embankments, U.S. Department of Transpor- ering Soils and Paleosols (P. Martini and W. Ches- tation, Vol. 4, 154 pp. worth, eds.), Elsevier, Amsterdam, pp. 283-306. Strohm, WE., Jr., G.H. Bragg, Jr., and TH. Zeigler. McCarthy, D.E 1988. Essentials of Soil Mechanics and 1978. Technical Guidelines (Report FHWA-RD- Foundations: Basic Geotechnics, 3rd ed., Prentice- 78-141). In Design and Construction of Compacted Hall, Englewood Cliffs, N.J., 614 pp. Shale Embankments, U.S. Department of Trans- Noble, D.F. 1977. Identifying Shales That Weather portation, Vol. 5, 216 pp. Rapidly When Disturbed by Construction-Accelerated Taylor, D.W. 1948. Fundamentals of Soil Mechanics. Weathering of Tough Shales. Final Report VHTRC John Wiley & Sons, New York, 700 pp. 78-R20. Virginia Highway and Transportation Terzaghi, K., and R. Peck. 1967. Soil Mechanics in Research Council, Charlottesville. Engineering Practice. John Wiley & Sons, New Oakland, M.W., and C.W. Lovell. 1982. Standardized York, 729 pp. Tests for Compacted Shale Highway Embank- Underwood, L.B. 1965. Machine Tunneling on ments. In Transportation Research Record 873, TRB, Missouri Riverdams. Journal of the Construction National Research Council, Washington, D.C., Division, ASCE, Vol. 91, No. CO1 (May), pp. 15-22. pp. 1-28. 576 Landslides: Investigation and Mitigation

Underwood, L.B. 1967. Classification and Identifi- cation System. In Proc., Symposium on Detection of cation of Shales. Journal of the Soil Mechanics and and Construction at the Soil/Rock Interface (W.F. Foundations Division, ASCE, Vol. 23, No. SM6 Kane and B. Amadei, eds.), ASCE Geotechnical (Nov.), pp. 97-1 16. Special Publication No. 28, American Society of U.S. Office of Surface Mining Reclaimation and Civil Engineers, New York. Enforcement. 1991. Permanent Program Perfor- Withiam, J.L., and D.E. Andrews. 1982. Relevancy mance Standards—Surface Mining Activities. In of Durability Testing of Shales to Field Behavior. Code of Federal Regulations (7/1/91ed.), Title 30, In Transportation Research Record 873, TRB, Chapter VII, Part 816, Section 816.71: Disposal of National Research Council, Washington, D.C., excess spoil: General requirements, and Section pp. 36-42. 816.73: Disposal of excess spoil: Durable rock fills. Wood, L.E., and P. Deo. 1975. A Suggested System Welsh, R.A., Jr., L.E. Vallejo, L.W. Lovell, and M.K. for Classifying Shale Materials for Embankments. Robinson. 1991. The U.S. Office of Surface Min- Bulletin of the Association of Engineering Geologists, ing (OSM) Proposed Strength-Durability Classifi- Vol. 12, No. 1, pp. 39-55.