Geotechnical Characteristics of Sediments on the Western Central Scotia Slope
Author: Kevin MacKillop, Kate Jarrett, Perry Mitchelmore, Dave Mosher, Calvin Campbell
Sediment mass movements resulting from diverse triggering mechanisms have long been known to occur on continential margins with different slopes. The ability to determine potential failures and hazard mapping is becoming more significiant with the development of deep water oil and gas resources and the demand for submarine cable installitons. The Scotia Slope (Figure 1) provides many examples of large slumps and debris flows. A comprehensive geohazard assessment requires geotechnicial characeterization of surficial sediments together with identification of features or conditions that represent past and present geologic hazards. This study uses available geotechnicial data including index properties, shear strength, Atterberg limits and consoldation data to characterize the sediments and analysis slope stability in the western central Scotia Slope. The engineering analysis was limited to slope stability using the infinite slope method to determine factor of safety, critical slope inclination, critical sediment thickness and critical earthquake acceleration coefficient using a pseudostatic analysis.
The study area is located on the western central Scotia Slope (43o00’ to 42o18’ N latitude, 62o30 to 60o 30’ W longitude; Figure 2) in water depths from 500 to 2500 mbsl. The area is bounded on the west by Mohican Channel and on the east by the Acadian Valley system. The shelf break lies between 400-500 mbsl and the transition from slope to continental rise is between 2000-2500 mbsl. The morphology of the central scotia slope and rise are relativity smooth. The seabed is cut by shallow, downslope-trending gullies that begin at 500 mbsl. Seafloor slope angles range from 5o on the upper slope to 1.5o on the rise. On the lower slope head scarpes representing slide failure scars are visible (Figure2) that range in height from 10 to 80 m. and are continious for up to 20 km.
Slope instabilities occur just before the shelf break in depths of 400 to 500 meters and extend onto the rise in water depths of 2000m. These flows cover 600km2 and extend for 30 km from the source area.
2. GEOLOGIC SETTING
The Scotian Shelf is part of the Mesozoic rifted margin of the central North Atlantic Ocean, with thick Jurassic and Cretaceous strata overlying Triassic salt (Wade and Maclean, 1990). The upper few hundred metres of the sedimentary sequence beneath the scotian slope consist of a thick progradational Tertiary succession, overlain both conformably and unconformably by Quaternary sediments (Swift, 1985). There was rapid sedimentation on the Laurentian Fan and Scotia Slope with the onset of
terrestrial glaciation in the early Pleistocene. Widespread gully cutting took place in the early Pleistocene, but the overall style of sedimentation continued to be prodeltaic (Piper and Normark 1989; Newton et al.,2003). The first shelf-crossing glaciation event occurred about 0.5Ma and since that time, the continental slope has been dominated by proglacial sediment deposition, with little sediment accumulation at high-stands. Ice sheets extended close to the shelf edge at the last glacial maximum (18 ka) and had retreated to the present shoreline by about 12 ka (Stea et al., 1998). Since that time, and with subsequent sea level rise, continental slope sedimentation has been slow, dominated by pelagic and hemipelagic deposition (Mosher et al., 1994). The relative sea level through geological history will have varied from the present level. The Holocene (past 10,000 years) is believed to be a period of transgression and relatively light, hemiphalagaic deposition. Regression periods during the late and middle Pleistocene, where relative sea levels may have been 110 to 120 m below present, are believed to be periods of more active deposition.
The surficial sediment in the study area is understood to be transported glacial moraine, deposited either hemiphalagically or by mass transport and debris flows. The cause of mass transport is not well understood but earthquakes are generally thought to be a factor (e.g. 1929 Grand Bank earthquake). In this slope environment, ground-shaking can remobilize sediment and cause massive landslides as it did in the 1929 event (Piper and Normark, 1982; Piper et al. 1988, 1999a). There is little evidence of sediment failures in the last 10 ka and slope stability analysis has shown the surface sediment to be statically stable, except on steep escarpments and canyon walls. There is, however, abundant evidence of sediment failures that approximately correlate to glacial advances (25-12 ka, ~ 75 ka, ~130 ka) suggesting a potential loading situation likely to cause failure (Mosher et al., 1994).
3.1 Sediment Coring and Physical Properties
There have been a number of expeditions on the scotian shelf and slope to collect geotechnical data and recover samples for laboratory analysis. Synthesis of the data is ongoing for many of the expeditions. The scope of work for this paper pertains to data from 12 cores collected during 2000 (cruise 2000036) on the Scotia Slope in water depths from 532 to 2490 mbsl (Figure 2). Cores were collected from the seabed using the Long Corer Facility (LCF). The device collects a 100mm diameter, 7m to 15 m long relatively undisturbed sediment core. The seabed inclination for the core locations varied from 1.01 degrees to 10.71 degrees. The seabed inclination at each core location was determined from high resolution multibeam data. (These data are confidential and cannot be used explicitly in this report).
Figure 2. Core location of cores analysed for this study.
Core processing was done on both piston and trigger weight cores at the BIO core-processing laboratory. The core is initially processed in the Multi Sensor Track logger (MST). The MST measures bulk density, magnetic susibility and compressional wave velocity at 1 cm intervals. The core is then split with one half designated as the archive and one half as the working half. The archive half is digitally photographed and visually described. Discrete index property samples, shear strength and acoustic velocity measurements are taken at 10 cm intervals on the working half.
Piston cores from the continental slope have a distinctive Holocene and late Pleistocene sedimentary sequence (Figure 3). Holocene olive-gray muds pass down into early Holocene–late Pleistocene silty muds and thin sand beds that are believed to reflect lowered sea level at that time on the outer shelf. Distinctive ice rafted marker horizons date from 12 and 14 ka. Older sediment is proglacial mud with dropstones on the continental slope and related mud turbidites on the continental rise, with distinctive color changes probably related to changing glacial sources.
Figure 3. Type lithostratigraphy derived from a composite of shallow piston cores, with representative grain-size-distribution (after Mosher et al. 2004).
3.2 Consolidation testing
Consolidation testing was conducted on four samples from cores 036_11, 036_28 and 036_29. The tests were preformed in a back-pressured consolidometer at the GSC(Atlantic) geomechnical lab,. The application of back pressure is to ensure 100% saturation. The samples were allowed to adjust to the back pressure for a minimum of twelve hours before incremental loading was started. A load increment ratio of 1 was used for the first two tests and a ratio of 0.5 was used for the last two tests.
3.3 Slope Stability
A successful engineering model for slope stability characterizes the driving forces and the available shear strength. Available shear strength is assessed based on geotechnical characterization of soils, their stress history and the type of loading estimated. The driving forces or loads can be eiter static or dynamic.
This assessment of slope stability used an infinite slope method of analysis (Figure 4) assuming undrained loading conditions. The infinite slope method uses force equilibrium theory to evaluate both the resisting and driving forces on an assumed sliding surface. For simplicity of analysis, the end and side restraining conditions of the sliding mass are ignored. The Factor of Safety (FS) for a potential failure plane is defined as
It is appropriate where the failure surface, or assumed slip surface, is plane and parallel to the ground surface over a significant length, and the depth is small in comparison to the length (Skempton and Delroy, 1957). The infinite slope method will generally give a very conservative assessment of stability at a site and is limited in its application, but may be used for slopes in cohesionless soil and for relatively flat slopes in which stability is governed by the residual shear strength (Skempton and Delroy, 1957). Others have used the infinite slope method to develop relationships for cohesive slopes (Lambe and Whitman, 1979).
Figure 4. Stress condition in the Infinite slope model. (a) Diagram of slope. Force diagrams (a) for earthquake. (c) Earthquake with uplift, and (d) surcharge load.
The FS can be determined using either a Total Stress Stability Analysis (TSSA) or an Effective Stress Stability Analysis (ESSA). For TSSA, the available shear strength is equal to the undrained shear strength mobilized when failure occurs before any significant dissipation of shearing induced porewater pressures takes place. For ESSA, the available shear strength is the drained shear strength mobilized at relatively large strains and stresses. For present purposes, a TSSA was performed assuming a completely undrained condition insitu. The FS is defined and is determined by
Where Su is the undrained shear strength of the soil, measured using a laboratory shear vane, is the slope inclination and is the buoyant unit weight of the soil above the potential failure plane and H is the vertical distance from the potential failure plane to the surface.
For static loading conditions, the measured value of shear strength was used for the resisting force. For cyclic loading only 80% of the measured shear strength was used for the resisting force. Seed has shown that soils subjected to cyclic loads may exhibit progressive accumulation of strain accompanied by softening and subsequent loss of strength. A value of 80% strength loss appears to represent 10 cycles of loading/unloading and was recommended for clayey soils (Madeski and Seed, 1978).
The critical slope inclination is determined as the angle where the factor of safety is unity. To obtain the critical slope inclination, eq. (2) is solved for . The critical slope angle is determined by
The critical thickness is the thickness where the factor of safety is unity. To obtain the critical slope thickness, eq. (2) is solved for Hc, the critical slope thickness. The resulting equation is
The critical earthquake induced load is considered using a pseudostatic analysis. In a pseudostatic analysis, a coeficient (kh) is multiplied to the weight of the soil and the resulting load is applied horizontally outwards towards the surface of the slope. The critical earthquake acceleration coefficient Kh, is the acceleration coefficient where the factor of safety is unity. To obtain the critical critical earthquake acceleration coefficient Kh, a modified version of eq. (2) is solved for Kh, the critical acceleration coefficient. The resulting equation is
From equation 5 it is obvious that for a slope with a static FS less than 1.25 will result in an unstable slope under earthquake loads (i.e., FSE <1).
The conservatism of the assumptions in the infinite slope method was checked using the slope stability package
XStabl™ and block surfaces. The principal difference between the two methods is block stability factors the end conditions and allows for a bottom failure surface that may not be parallel to the slope.
A wedge analysis was performed in XSTABL™ using rankine wedges at the upslope and downslope ends of the central wedge. The wedge analysis assumes three blocks contribute to the driving force, an upstream passive wedge, a central wedge and a downstream wedge. The wedges are generated based on the assumption that the limiting surface is a function of the angle of internal friction for the soil as defined by the equation
For an undrained case, and the angle on the upstream and downstream is 45 degrees. The central segment length was arbitrarily selected as 300 metres, or approximately 30 times the depth. The Xstabl software allows the user to specify search blocks for the end positions of the critical surface. For this analysis, the middle depth was chosen with a search box 50 metres long and 8 metres high at both ends of the section. Two hundred (200) trial surfaces were generated. The wedge analysis was performed on only one core, 2000_036_027 , due to time constraints on the project.
4.1 Sedimentology and index properties
A geotechnical model is proposed consisting of five geotechnical stratigraphic units. The subsurface lithostratigraphy facies for the cores was adopted from Mosher (1994, et al 2004) (2004) and Gauley (2001). Unit 1 is the hemiphelagic Holocene soils. Unit 2 corresponds to the Heinrich layers that were deposited between 10 ka and 14 ka. Units 3 and 4 are late Pleistocene soils deposited by ice margin processes during glacials and hemiphelagically during inter-glacials (Mosher et al., 1994). Unit 5 consists of an undifferentiated mass transport deposit. It shoulld be noted that there is limited data for units 2 and 5. The physical property data for the lithostragraphic units are summarized in Table 1. All soils are classified according to their Unified Soil Classification System symbol.
4.1.1 Unit 1 – olive grey clay (CL to CH)
An olive grey clay forms a drape like layer up to 2 meters (Figure 4) thick over much of the Scotia Slope. The clay is classified as a lean (CL) to fat (CH) clay and consists of 13.2 % sand 59.9 % silt and 28.9 % clay. The Su/Po’ averages 1.76 which is well above Skempton’s (1970) range of 0.2 to 0.5 for normally consolidated soil. This over consolidation is typicial of the upper 2 meters of marine sediments and is termed apparent overconsolidation.
4.1.2 Unit 2 – brick red mud (CL)
The mud is described as silty with abundant fine and coarse sand, consisting of thin horizontal layers, and interbedded with brown to dark greyish brown clay. It was deposited during the last glaciation (from 12ka to 14 ka) by rapid iceberg discharge from the Gulf of St. Lawrence (Piper and Skene, 1998).
One atterberg limits test, from core 036_029, indicates the deposit is a lean clay (CL). The Su/Po’ (2 measurements) calculated using laboratory shear vane and MST density data averages 0.22. The measured increse in shear strenght with depth is 1.39 kPa per meter.
4.1.3 Unit 3 - red brown mud (CL)
Units 3 is a red brown clay with ice rafted dedris and sand. It was deposited by ice margin processes during glacials and hemiphelagically during inter-glacials (Mosher et al., 1994). This unit is classified as a lean clay (CL) to lean clay with sand. The grain size varies from 2
to 5.0 % gravel 10 to 21 % sand 31 to 34 % silt and 45.0 to 59.0 % clay. The Su/Po’ calculated using laboratory shear vane and MST density data averages 0.36. The measured increse in shear strenght with depth is 2.18 kPa per meter.
4.1.4 Unit 4 – grey brown mud (CL)
This facies is either massive or, more commonly, contains distinctive interlaminated and interbedded very fine to fine sand and grey brown mud up to 1 m thick. Grey brown mud beds occasionally contain sandy blebs and random granules and is interlayered with Unit 3 (Figure 3). This unit is classified as a lean clay (CL) to lean clay with sand. The grain size varies from 0.4 to 5.0 % gravel 8.0 to 16.8 % sand 31.0 to 35.5 % silt and 44.0 to 59.0 % clay. The Su/Po’ calculated using laboratory shear vane and MST density data averages 0.46. The increse in shear strenght with depth is is 3.76 kPa per meter.
4.1.5 Mass transport deposits (CL)
Mass Transport Deposits (MTD’s) were recovered in cores 036_023 and 036_029. One Atterberg limits test from core 036_029 was classified as a lean clay (CL). The Su/Po’ calculated using laboratory shear vane and MST density data averages 0.40. The increse in shear strenght with depth is 2.51 kPa per meter.
Table 1. Physical properties for the geotechnical units. Density data are MST and discrete values.
4.2 Consolidation testing
Consoliation results are presented in Table 2. The tests were conducted on samples from Units 1 (036_011) and 4 (036_027, 029). The compression indices range from 0.25 to 0.56 (low to moderate compressibility). The high OCR value of 5.5 is from Unit 1 and is within the zone of apparent overconsolidation for marine clays. The results for the two consolidation tests from 036_029 sugests that the sediment is underconsolidated to normally consolidated. The geotechnicial profile (Figure 5) from core 036_029 indicates that the sediments below the MTD are underconsolidated. There is an decrease offset in the density profile and the natural water content are greater then the liquid limits.
Table 2. Consolidation test results.
4.3 Slope Stability Analysis
The minimum FS for each core was determined by calculating the factor of safety for each shear strength measurement ( ≈10 cm intervals). The minimum factor of safety was then selected as the actual factor of safety for slope stability at that core location. The effective buoyant weight was calculated from MST bulk density values. A summary of results are presented in Table 3. for the twelve cores comprising the sample set for this study. Five of the cores, had a factor of safety less than unity with a minimum of 0.44. A result suggesting that 5 of the 12 core loactions is unstable under static loading conditions is very conservative.
Inorder to remove extreme and/or errorous shear strength measurements the shear strength data was smoothed by adopting an averaging technique. A variety of averaging ranges were attempted but it was found that by simply using the average of the discrete shear strength value with the one shear strength above and below that value (i.e., moving average every 300mm) produced adequate results. With the exception of core 036_027pc where a 2500 percent increase was noticed, the average increase in the FS was 46 percent and the number of cores with a FS less than Table 3. Summary of results for FS infinite slope analysis.
Factor of Safety Critical slope (deg) Critical Earthquake Coefficient Critical Thickness (m)
unity decreased to two. A summary of the results are presented in Table 4.
Sensitivity of seabed inclination to the calculated factor of safety was assessed for cores 036_ 023, 027, 028 and 029. In general, the FS increases rapidly for slopes less than two degrees and seems to be at unity for slopes between six and nine degrees.
The FS for core 036_027,calculated with the wedge method of analysis and infinite slope method are presented Figure 5. As can be seen, the FS is improved for the core section, from less than unity to 1.84. As well, the critical seabed inclination improved from approximately four degrees to greater than nine degrees. The dramatic improvement in FS for a wedge analysis indicated the infinite slope method required laterally homogeneous soils at greater than 300 m extent under static conditions to become unstable.
Table 4. Summary of results for FS infinite slope analysis using averaged shear strength measurements.
Factor of Safety Critical slope (deg) Critical Earthquake Coefficient Critical Thickness (m)
The critical acceleration coefficient for a pseudostatic analysis was negative for core 036_027 since the calculated infinite slope FS was less than unity. However, using the wedge analysis and improved FS, the critical acceleration coefficient was still determined to be relatively low at 0.04. A summary of slope stability analysis are presented in Figure 3.3. Further, the critical height as calculated in xstabl was much greater than determined from the infinite slope method.
Engineering analysis of slope stability was performed using the infinite slope method for twelve locations in the central region of the scotian slope. Geotechnical stratigraphic units were established based on deposition history, color differences and lithostratigraphic interpretation. A geotechnical model is proposed consisting of four geotechnical stratigraphic units
Each of the clay units (Units 3 and 4) below the upper two meters of soil was characterized as lean clays to lean clays with sand, both of low plasticity. Regression analysis of shear strength based on the four geotechnical stratigraphic units and each sub-strata did not provide any pattern. A more general qualitative assessment of the data indicated a shear strength increase with depth of between 2 to 4 kPa per metre.
Consolidation tests and physical property data suggests that the sediments are overconsolidated in the upper 2 meters and become slightly underconsolidated to slightly overconsolidated with depth. The near surface overconsolidation is typicial of most marine clays and be termed apparent overconsolidation. Deeper underconsolidated sediments may result from increases in pore pressure due to rapid loading by debris flows (Figure 4).
An analysis of slope stability was performed using the infinite slope method of analysis. It uses force equilibrium theory to evaluate both the resisting and driving forces on an assumed sliding surface. The FS was calculated for each core using a TSSA and the undrained shear strength, assuming a frictionless soil (i.e., u = 0). Sensitivity to unrepresentative values was assessed using a moving average technique for shear strength that gave a 46 percent increase in FoS. As a result, the number od cores with a FS<1 was reduced from five to two, out of a total of twelve sites.
The critical seabed inclination was assessed by calculating the factor of safety for various slopes. The results indicate a critical slope angle of between six to nine degrees for the central scotian slope region. These values were reasonable consistent for discrete and averaged shear strength profiles.
The lateral homogeneity required for the infinite slope method and the appropriateness of ignoring end forces was assessed preliminarily by using a wedge method of analysis. The FS improved by approximately 300 percent, as did the critical angle, critical acceleration and critical height. The wedge analysis was performed in XSTABL and assumed a base width of 300 m. The results suggest that the infinite slope method requires more than 300 m lateral distance to develop enough driving force to overcome resistance.
The variation between observations and the calculated FS may be attributed to several factors including sample disturbance, erroneous measurement of available shear strength, analysis methods for earthquake loading and errors in assuming a frictionless soil (i.e., u = 0). The potential for sample disturbance is greater when the soils are non-cohesive (i.e., sandy) or low plasticity clays. The degree of sample disturbance for sediments recovered using the LCF is not well defined, but correlations suggest the lab shear strength may be as much as 40% lower than insitu shear strength for Lean clays (10
- Lambe, T.W. and Whitman, R.V., 1979. Soil mechanics, SI version, MIT, John Wiley & Sons, New York
- Lee, H.G, 1980. Offshore Soil Sampling and Geotechnical Parameter Determination. Journal of Petroleum Technology, Society of Petroleum Engineers Publication No. 8667.
- Makdisi, F.I. and Seed, H.B., 1978. Simplified procedure for estimating dam and embankment earthquake-induced deformations”, Proc. of the American Society of Civil Engineers, Journal of Geotechnical Engineering, Vol.104, n.GT7.
- Mosher, D.C., Moran K., and Hiscott R.N., 1994. Late Quaternary sediment, sediment mass flow processes and slope stability on the scotian slope, Canada. Sedimentology, Vol, 41, pp. 1039-1061.
- Mosher, D.C., Piper, D. J. W., Campbell, C.D. and Jenner K. A. 2004. Near-surface geology and sediment-failuure geohazards of the central scotia slope. AAPG Bulletin, Vol. 88, no. 6, pp. 703–723.
- Newton, C. S., Mosher, D.C., Wach, G., Piper, D. J. W., and Campbell, D.C., 2003. Quaternary and late Pliocene seismic stratigraphy of the central scotian slope from high-resolution reflection seismic (abs.): Geological Society of America Program with Abstracts, Vol. 35, no. 3, pp. 90.
- Piper, D. J. W., and Normark, W. R., 1989. Late Cenozoic sea level changes and the onset of glaciation: Impact on continental slope progradation off eastern Canada. Marine and Petroleum Geology, Vol. 6, pp. 336–348.
- Piper, D. J. W., and Skene, K.I., 1998. Latest Pleistocene ice-rafting on the Scotian margin (eastern Canada) and their relationship to Heinrich events. Paleoceanography, Vol. 13, pp.205-214.
- Piper, D. J. W., Cochonat, P., and Morrison, M. L., 1999a. The sequence of events around the epicentre of the 1929 Grand Banks earthquake: Initiation of the debris flows and turbidity current inferred from side-scan sonar. Sedimentology, Vol. 46, pp. 79– 97.
- Skempton, A.W., 1970. The consolidation of clays from gravitional compaction. Journal, Geological Survey of London, Vol. 125, pp.13-30.
- Stea, R. R., Piper, D. J. W., Fader, G. B. J., and Boyd, R., 1998. Wisconsinan glacial and sea level history of maritime Canada and adjacent continental shelf: A correlation of land and sea events. Geological Society of America Bulletin, Vol. 110, pp. 821–845.
- Swift, S.A. 1985. Late Pleistone sedimentation on the continental slope and rise off western Nova Scotia. Geological Society of America Bulletin, Vol, 96, pp. 832-841.
- Wade, J. A., and MacLean, B. C., 1990. Aspects of the geology of the scotian basin from recent seismic and well data, in M. J. Keen and G. L. Williams, eds., Geology of the continental margin off eastern Canada: Geological Survey of Canada, Geology of Canada, no. 2 (also Geological Society of America, The Geology of North America, Vol. I-1), pp. 190– 238.
109 Ilsley Avenue, Unit #14
Dartmouth, Nova Scotia, B3B 1S8
Toll Free: 1.877.434.3131