Regina Beach - a town on a landslide

A. WAYNECLIFTON,RICHARD
T. YOSHIDA,
AND ROYW. CHURSINOFF
Clifon Associates Ltd., 340 Maxwell Crescent, Regina, Sask., Canada S4N 5Y5
Received July 1, 1984
Accepted October 29, 1985
The town of Regina Beach is constructed on landslides along the Last Mountain Lake valley, a glacial meltwater channel in
south-central Saskatchewan, Canada. The landslides are retrogressive in nature and are seated in bentonitic clay shale of the
Bearpaw Formation. A study was conducted at Regina Beach to determine the sensitivity of the slopes to changes in slope
geometry as a result of regrading, or groundwater changes. Back-calculated shear strength was compared with values from the
laboratory and from other landslides in Cretaceous bedrock shales. Modest changes in grading of the toe areas would result in
significant reductions in the local stability and future movements of upslope portions of the landslide. Development should
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Calif Dig Lib - Davis on 12/26/14

minimize the amount of grading. Increased slope movement due to breaks in watermains was observed and measured. This
correlated well with analysis. Development in landslide areas must assume that differential vertical and horizontal movements
will occur. Proper site reconnaissance should identify optimum locations for buildings and services such that the risk of rupture
or damage is minimized.
Key words: landslide, retrogressive, back analysis, residual shear strength, Bearpaw Formation, inclinometer, slope
movement.

La ville de Regina Beach est construite sur des glissements le long de la vallCe de Last Mountain Lake qui est un canal
IormC par l'eau de fonte des glaciers dans le sud de la partie centrale de la Saskatchewan, Canada. Les glissements sont
rktrogressifs et se produisent dans le schiste argileux bentonitique de la formation Bearpaw. Une Ctude a CtC rCalisCe pour
Cvaluer la sensibilitC des talus aux changements de gComCtrie des pentes ou aux variations de pressions interstitielles. Les valeurs
de rksistance au cisaillement obtenues par calcul ii rebours ont CtC comparkes a celles obtenues en laboratoire ou provenant
d'autres sites de glissement dans les schistes du CrCtacC. Des changements modCrCs de la gComCtrie au pied des pentes
produiraient localement des reductions significatives de la stabilitC avec des mouvements possibles de la partie suptrieure du
glissement. L'amCnagement devrait minimiser la quantitC de rkgalage. L'augmentation du mouvement de la pente par suite
des bris des conduites d'eau a CtC observke et mesurCe; cette observation concorde avec les analyses. Les amknagements dans
For personal use only.

les zones de glissement doivent supposer que des mouvements differentiels verticaux et horizontaux se produiront. Une
reconnaissance adCquate du site devrait identifier des localisations optimales pour les bdtiments et services de telle sorte que
le risque de rupture ou de dornmage soit minimisC.
Mots clis: glissement, rkrogressif, analyse ii rebours, resistance residuelle, formation Bearpaw, inclinomktre, mouvement
de pente.
[Traduit par la revue]
Can. Geotech. J . 23, 60-68 (1986)

Introduction process was repeated until the river valley reached its maximum
The community of Regina Beach is located 50 km northwest depth. Subsequently, alluvial and lacustrine sediments were
of Regina, Saskatchewan (Fig. 1). Virtually all of the town is deposited to form the present-day valley bottom.
situated on a landslide on the flanks of Last Mountain Lake, a The valley slopes at Regina Beach display distinctive features
100 m deep valley eroded by glacial meltwater through glacial of old landslide slopes. On aerial photographs (Fig. 2) the slope
sediments and into underlying Cretaceous clay shale bedrock. exhibits a distinctive series of arcuate, interconnected rear
Development on the valley slopes began in the early 1900's headscarps and a pattern of subparallel ridges down the slope.
with construction of a railway and mainly nonpermanent Other identifying features are a steep headscarp and a hum-
cottages on temporary foundations. Upgrading of municipal mocky broken slope. The overall pattern of the slope tends to be
services followed. At the time of this investigation, redevelop- slightly concave upward. Commonly, the slopes tend to bulge at
ment was continuing with conversion of the town to an urban the toe. Ponded water or luxuriant vegetation may be found
community with permanent residences. Like similar areas in the behind the slump blocks. The angle of the slope is dependent
southern Canadian prairies, river valleys are highly valued as upon the geology, material properties, and the groundwater
recreational land and thus have high property values. regime.
Landsliding along the valley dates back to erosion of the The valley walls are moderately steep with an average slope
valley following deglaciation about 12 000 years ago (Chris- ranging from 4" to 7'. The valley is approximately 40m in
tiansen 1979). Although the valley has been partially backfilled height from the upland to lake level. The slopes are comprised
with sediments, slope movements are continuing. of numerous slump blocks ranging in width from 15 to 60m.
The natural drainage courses have developed along the scarps of
Geomorphology individual slump blocks. Construction of houses and streets has
The Qu'Appelle River valley was one of the major spillways resulted in grading of the slump blocks, altering historic flow
for glacial meltwater during deglaciation of southern Saskatch- patterns of both surface water and groundwater. Slopes are well
ewan (Christiansen 1979). As a consequence, the valley carried vegetated along natural drainage courses, especially where
huge volumes of water and was eroded to a depth of at least 60 m seepage is occumng.
below the present-day level of Last Mountain Lake. Valley
widening of the Last Mountain Lake spillway occurred by a Geology
series of successive landslides, with the landslide debris eroded The bedrock within the study area is Cretaceous marine clay
by streamflow. Oversteepening of the valley walls caused shale of the Bearpaw Formation. This formation has a thickness
further landsliding. Morphology of the valley indicates that the of over 100 m. The Bearpaw Formation clay has high plasticity
 ET AL. 61

Stratigraphy
Subsurface conditions at Regina Beach were investigated by
a series of test borings at the locations shown in Fig. 3. Rotary
drilling methods were used and continuous logs of spontaneous
potential and resistivity were obtained. Figure 4 illustrates
typical strata encountered in the landslide.
Considerable distortion and displacement of stratigraphic
units have occurred as a result of slope movement. Also, erosion
near the toe has removed the majority of the upper stratigraphic
units. For these reasons, it is not possible to extrapolate the
stratigraphy across the landslides.
Definition of the failure plane along which previous slope
movement has occurred is desirable information for slope
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Calif Dig Lib - Davis on 12/26/14

stability analysis. Continuous sampling was used to recover the

failure plane in undisturbed cores. The failure zone generally
coincided with a thin seam of bentonite approximately 80 mm
thick within the shale. The bentonite was softened, very highly
plastic, and moist. Conditions at a depth of about 3 1.5 m in the
test hole log in Fig. 4 are typical of this zone. The shale above
the failure plane was noticeably softer, disturbed, highly
slickensided, and brecciated, compared with the very hard
intact shale below the failure plane.
Subsurface data were compiled into two stratigraphic cross
sections included as Figs. 5 and 6. The stratigraphy beneath Last
Mountain Lake was established by drilling two test holes during
the winter months from the ice (Fig. 6). The bedrock surface at
test hole 117, drilled in the lake, was approximately 50 m lower
For personal use only.

than the bedrock surface at test hole 102, on the upland

representing the depth to which the valley was eroded during its
formation. The upper surface of the bedrock was highly
weathered below the lake, evidence of its previous exposure.
Up to 23 m of fluvial-lacustrine sediment was encountered
above the shale at test hole 117. Valley infilling has occurred to
FIG. 1. Location of Regina Beach. such an extent that the water depths were very shallow.

Hydrogeology
and is montmorillonitic with frequent layers and lenses of The potential major aquifer systems in the Regina Beach
bentonite. In this area, the upper few metres of the Bearpaw region include surficial aquifers and intertill sand and gravel
Formation is highly fractured. With depth, the shale becomes deposits. Outwash sand and gravel form a surficial aquifer on
hard and intact. The clay shale has a very low residual shear much of the upland. Where the base of this aquifer is exposed
strength; consequently, most of the landslides within the study along valley slopes and coulees, springs are abundant. This is
area are seated within this formation. the source of several perennial springs with considerable flow
The bedrock at Regina Beach is overlain by up to 60 m of within the Town of Regina Beach. Other smaller springs are
glacial deposits. A layer of sand and gravel commonly lies evident along the headscarp of the major landslides. The amount
between the bedrock and the glacial deposits. This layer is of seepage from such springs is controlled by water levels in the
known as the Empress Group and regionally may vary in aquifer that are directly related to climatic conditions.
thickness from nonexistent or very thin up to 100 m. At Regina Seepage from the surficial aquifer is responsible for local
Beach, this stratum is generally less than a metre thick. ponding and areas of high groundwater levels noted in the
The glacial strata lying above the Empress Group consist of landslide areas. These are more common in the depressions
till and intertill stratified drift. The till is a heterogeneous adjoining major slump blocks. Further seepage may be coming
mixture of clay, sand, silt, gravel, and boulders with a Unified from sand and gravel layers existing within the till strata. These
Classification of low plasticity clay (CL). The stratified drift aquifers tend to be drained along the valleys and truncated
units are strata that have been sorted by water. These are within the landslide mass. They therefore have low yields but
commonly composed of sand or gravel with minor silt or clay can exhibit sufficient head to affect the stability of slopes.
strata. Groundwater conditions at the site were monitored by open
Proglacial stratified drift overlies the till in much of the area standpipe piezometers. In general, the highly fractured nature of
surrounding Regina Beach and forms the present ground surface the landslide mass functions as a surficial aquifer over the
on the upland. These strata are commonly either glacial- landslide area. A high groundwater table elevation was indi-
lacustrine clay, outwash sand and gravel, or an eroded till cated by the piezometric data. Refemng to Figs. 5 and 6, the
surface with a thin residual mantle of boulders and silt. Outwash groundwater table approximately followed the slope of the
sand and gravel and eroded till are the most prevalent surficial ground, at an average depth of 5 m below the ground surface.
sediments near Regina Beach. Piezometric levels were approximately hydrostatic with depth.
 62 CAN. GEOTECH. J. VOL. 23, 1986
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Calif Dig Lib - Davis on 12/26/14
For personal use only.

FIG.2. Aerial photograph of Regina Beach.

I / I //
II
II

----- I I
I
I
TH117

1st STREET I '

LAST
MOUNTAIN
LAKE
SLOPE INDICATOR MOVEMENT

L 36
3 mmm)
m JULY 8 2 to OCT.83

LEGEND:
PIEZOMETER
SLOPE INDICATOR !.i
0 25 50

metres

FIG. 3. Test hole and cross section locations.

 CLIFTON ET AL. 63

- -
PLASTIC WATER LlOUlD
L I M I T CONTENT L I M I T
SHEAR STRENOTH
UNCONF
- kPa
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Calif Dig Lib - Davis on 12/26/14

L A DRY DENSITY - k(t/m3

REOINA BEACH ORILLEO: NOV 2 1 / 6 0
TEST HOLE - I04 SURFACE ELEVATION: 5 2 1.22 n
DRILLER: M. MILLER SR CONO. WATER: 8 4 O n r h o s / c n AT 25.C
E LOO BY: E SZAFRON SR COXD. MUD- 1300mmhos/cr AT 25.C
For personal use only.

SASK. DEPT OF HWYS.

FIG.4. Log and engineering properties of soil at test hole 104.

L60
1
I
LEGEND
TILL - o x ~ d ~ r e d
TILL - unoxidized
GRAVEL
CLAY
:f:
SHALE
I I I
C"--':l
[W
E
PIEZOMETER T I P
ELEVATION

4 I
102A -
I I I
VERTICAL EXAGGERATION L : 1
I I I I I
0 100 200 300 LOO 500 600
METRES

FIG.5. Stratigraphic cross section A-A.

Engineering properties increased, and strength and dry density decreased, with depth as
The properties of the soils were determined in a laboratory the shear zone was approached. Shear strength dropped from
test program. Typical results of laboratory tests have been 400 to 20kPa and density from 1700 to 1400 kg/m3 as the
plotted in Fig. 4. A summary of index properties and engineer- natural water content increased. The slickensided bentonite
ing properties is given in Table 1. layer at a depth of 31.5 m (Fig. 4) exhibited an elevated natural
The variation in properties of the clay shale are typical of water content and liquidity index as a result of softening by large
those encountered in similar landslide investigations by the shear strains. Below the shear zone, the shale was hard and
authors. Figure 4 shows the variation in the engineering intact.
properties of the shale with depth. The moisture content Multistage direct shear tests were conducted on shale samples
 CAN. GEOTECH. J . VOL. 23, 1986

- 540
-
0
Y)

- 520

-
EDGE OF WATER
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Calif Dig Lib - Davis on 12/26/14

I I I
-
E
ELEVATION

I I I I I I I I I
VERTICAL EXAGGERATION L t
I I I I I
0 100 200 300 LOO 500 600 700 800
METRES

FIG. 6. Stratigraphic cross section B-B.

TABLE1. Summary of soil properties T E S T No. c'lkPol +Yo)

27.0 8.5
IP 2 53.6 8.8
For personal use only.

W WP Pd cu
Soil ( 1 (%I (76) (kg/m3) (kPa)

Till 20 19 25 1750 100-400"

Clay 23 35 15 1650 50- 180b
Softenedc
shale 30 32 60 1600 20-80
Softenedc
bentonite 50 35 100 1400
IntacP NORMAL STRESS ( kPa
shale 20 32 60 1700 120-400
T E S T No. c' I kPol +' I D )
"Intact till on the upland exhibited the higher values; till from the landslide
0.0 15.8
mass, the lower. 18.6 12.L
bIntact clay exhibited the higher values; slickensided clay, the lower.
'Sediments obviously reworked by landsliding were termed "softened or
"disturbed."
d ~ h hard
e shale below the bentonite shear zone was termed "intact."

taken from the major shear zone. The objective of these tests
was to determine the residual shear strength parameters for the
failure surface. Five multistage residual shear tests were per-
formed on four undisturbed and one slurried sample of shale NORMAL STRESS ( kPa)
recovered from the region of the failure plane. Multistage FIG. 7. Direct shear test results: c' = 0-50 kPa, +,' = 8-15"
testing was conducted to eliminate inaccuracies associated with
sample variability. The results of the tests, shown in Fig. 7, showed some scatter.
The maximum stress at which the samples were tested was A range of cohesion intercepts from 0 to 50 kPa and a range of
equal to the maximum overburden pressure. Two or more residual friction angles from 8" to 15O were measured. Similar
normal loads were then chosen by dividing the maximum load residual shear strength parameters have been measured for
into thirds. The specimen was trimmed to fit into a 50 mm direct Cretaceous clay shale bedrock samples by others (Insley et al.
shear box and a normal stress applied. The specimen was 1977; Krahn et al. 1979; Clifton Associates Ltd. 1983). Table 2
immersed in water and allowed to swell or consolidate. A strain compares shear strength data from these sources.
rate of approximately 1.5 mm displacement in 1 h was used. On
the first normal load, the cycling was continued until a total History of movements
displacement of approximately 25-40 mm occurred. Cycling Slope movements have been in evidence on various dwellings
was continued until the strength no longer changed by a and improvements in Regina Beach for several decades. The
significant amount. The next load increment was applied and the most obvious movements occurred along a railway grade
sample was again sheared once consolidation or swelling was constructed along the shoreline of the lake. Regular and
completed. extensive maintenance was required to repair slips that occurred
 CLIFTON ET AL.

TABLE
2. Comparison of residual shear strength parameters
I
WL Wp
Laboratory
tests

C' c'
Slope
analysis
OUTCROP OF SCARP
LmE 1
Case (% (% (kPa) 4' (kPa) 4'
-

~ i ~ a w i n " . ~72.0 25.0 0 8-15" 0 10.0"

Smoky 2b 126.8 38.7 0 7.5" 0 7.5"
aids stone^ 68.9 24.1 0 10.0" 0 10.0" SECTION
66.8 26.7 0 8.5"
Maymontc 50.7 33.7 0 8-28" 0 8.2" FIG. 10. Movement of dwelling foundation over a 1 week period.
or den^ 121.8 21.5 10 6.5" 3.5-5.0 6.5-10.5"
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Calif Dig Lib - Davis on 12/26/14

97.5 36.5 10 9.0"

Regina
Beach 135.0 35.0 0-50 9-15" 0 5-7"

"Glacial clay.
b I n s l e y et al. (1977).
'Krahn et al. ( 1 9 7 9 ) .
d C l i f t o n A s s o c i a t e s Ltd. ( 1 9 8 3 ) .
~ OUTCROP OF SCARP
SLIDE

r SCARP

SECTION
SLUMP BLOCK
FIG. 11. Movement of modem dwelling foundation over a 4 year
,-SLUMP BLOCK
period.
\ \
For personal use only.

\ AI SCARPS OF to location on the landslide mass. Figure 9 illustrates movement

\ \ SLUMP BLOCKS
\ \ of approximately 250 mm that occurred uniformly over a period
\ \
+ of approximately 30 years, resulting in a rate of horizontal
movement of about 8 mm/year with very little vertical move-
FAILURE PLANE
ment. Figure 10 illustrates foundation movement of about
FIG. 8. Position of structure on slump block affects degree of 200 mm horizontally and vertically that occurred rapidly,
damage experienced: (a) structure fully on one slump block, less probably in less than a week. In this instance, the structure had
susceptible to damage; ( b ) structure across scarp between two slump been on the foundation for more than 10 years and had not
blocks, very susceptible to damage. experienced foundation movement until the rapid movement
occurred. Figure 11 illustrates continual annual movement of
about 50 mm vertically and 50 mm horizontally experienced by
a modem dwelling on a pile foundation, since its construction.
In another location it was estimated, by the movement required
to break a municipal water line, that movements were in the
order of 100 mm/year.
Inclinometers were installed at five locations in 1982 to
accurately measure the magnitude of movement and the depth at
which it occurred. The magnitude and direction of movements
are shown in Fig. 3. They confirmed the location of the failure
-I SECTION
plane noted during site investigations and that movement was
continuing. In general, movement accelerated during the spring
PLAN
and early summer and slowed during the winter months.
FIG.9. Movement of dwelling foundation over a 30 year period. Movements measured at three inclinometers are shown in
Fig. 12. Average rates of movement ranged from less than
along the track within the town limits. However, the railway 1 mm/month to 9 mm/month.
was abandoned and maintenance ceased. Hence, the locations The highest rates of movement were observed in the vicinity
of the landslide could be clearly identified. of test hole 108. In this area, evidence of cracking was
The approximate traces of the scarps of some of the more abundant. Rupture of services was a continuous problem and
active landslides are shown in Fig. 2. Movement along these ground cracking was visible along scarps. Along Centre Street,
scarps manifests itself in various ways. Breaks in water mains the rates of movement were smaller and there was little apparent
and gas lines, tightening of overhead power lines, and misalign- surface evidence of landslide movement. Even a 6 year old
ment of poles are typical. The effect on structures is very concrete sidewalk running the length of Centre Street showed no
dependent upon their location as indicated schematically in Fig. evidence of cracking attributable to slope movement.
8. Structures located fully on an intact slump block are much As expected, movement occurred in a direction perpendicular
less susceptible to damage than those located across scarps to the existing scarp and the shoreline. Movement was detected
separating adjoining slump blocks. at several levels in the slope movement indicators, but the major
The observed rate of movement varied greatly from location zone of movement was in the bentonite seam. The direction of
 66 CAN. GEOTECH. J . VOL. 23, 1986

The simplified Bishop method of analysis, a limit equilibrium

technique utilizing vertical slices, was used. The computer
program SLOPE-II was utilized. The formulation of SLOPE-I1
allows the analysis of composite slip surfaces (Fredlund and
Krahn 1977).
Two cross sections at locations shown in Fig. 3 were
examined. Cross section A-A (Fig. 13) is in a direction
approximately parallel to movements measured at test hole 108.
Cross section B-B (Fig. 14) is skewed to the actual measured
direction of movement, but parallels the direction of many of the
services within the town. The piezometric level used in the
analysis was measured during field investigations. The soil
parameters used are shown in Table 3.
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Calif Dig Lib - Davis on 12/26/14

Changes in slope geometry that might be caused by typical

property development were examined, as were changes in water
levels that might be expected from a break in a water service or
FIG.12. Slope indicator monitoring Regina Beach. with variations in climatic conditions. Fills of 1 and 2 m in
height located near the toe of the landslide, expected to be the
movement at test hole 107 was anomalous. However, the most sensitive area, were analyzed. General water table rises of
amount of movement was relatively small and the direction of 2 m, which may be associated with increased infiltration, or a
movement may not be accurately represented. No inclinometers local rise of 4 m, modelling a local water service break, were
were installed upslope and thus there were no firm data to also analyzed.
determine the rate of movement of the remainder of the slope. A shear strength equivalent to c' = 0 and 4' = 7" was
However, from observations in the upper portions of the more required for a factor of safety of unity on cross section A-A
active slopes, it is suggested that movement occurred at a rate (Fig. 13). For cross section B-B (Fig. 14), the mobilized shear
much less than that measured near the toe, in a manner common strength parameters were c' = 0 and 4' = 5". The low friction
to a retrogressive landslide. angle associated with cross section B-B is attributed to the fact
For personal use only.

that the section is skewed, resulting in a much flatter slope and a

Stability analysis longer failure plane. A friction angle of 7" was lower than values
The objective of the study conducted at Regina Beach was to of 8-15" measured in laboratory tests (Table 2) but was
estimate the sensitivity of the slopes to changes in slope consistent with values measured by others in similar materials.
geometry or groundwater conditions. An effective angle of internal friction of 7" was used in further
It was known that the slopes of Regina Beach were unstable analyses of the slopes.
and extremely sensitive to changes in the system of forces on the Trial slip surfaces used in the analyses are shown in Figs. 13
slopes. Hence, a factor of safety of unity was assumed. A back and 14. The slip surfaces were assumed to be reasonable
analysis technique was utilized to determine the mobilized shear representations of existing slip surfaces in the landslide mass as
strength along the slip surface. The back-calculated shear determined from positions of various slump blocks. The
strength was compared with shear strength parameters deter- calculated factors of safety are shown in Tables 4 and 5. The
mined in the laboratory from this study and from other studies addition of a 1 or 2 m fill near the scarp of any block resulted in
of landslides in Cretaceous bedrock shales to evaluate the as much as a 9% decrease in the factor of safety for landslide
validity of the analysis. blocks below the fill. As expected, the fill had the effect of

520

500

-I
W
LBO LBO
LEGEND
TlLL - o x i d ~ z e d
TlLL - unoxldlzed
GRAVEL FZza
r----1

1
SILT
SHALE
I I
IW
B
1 I I I
VERTICAL EXAGGERATION L 1
I I
1I
460

0 100 200 300 LOO 500 600

METRES

FIG.13. Slope stability analysis along cross section A-A.

 CLIFTON ET AL.
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Calif Dig Lib - Davis on 12/26/14

I I I I I I I I I 4 I I I I I I I
0 100 200 300 LOO 500 600 700 BOO
METRES

FIG. 14. Slope stability analysis along cross section B-B.

TABLE
3. Soil properties used in stability analysis increasing the stability when slip surfaces upslope of the fill
were examined. On cross section B-B (Fig. 14), a fill across test
hole 107 tended to increase stability owing to the large area of
For personal use only.

Soil the landslide block that was covered and the configuration of the
fill. However, raising the piezometric surface by 4 m reduced
Disturbed the stability by as much as 12%.
glacial sediments 1750 0 15"
Undisturbed Discussion and conclusions
glacial sediments 1750 0 30" There is a degree of uncertainty in the analysis of a
Bedrock shale 1600 0 5-7" retrogressive landslide particularly regarding consideration of
the effects of the many slump blocks that make up the slope.
TABLE
4. Factors of safety analysis of section A-A Others have analyzed a multiple retrogressive landslide by
assuming each block separately (Thomson and Hayley 1975;
Failure surfacea Haug et al. 1977). The method of analysis used here examines
any portion of the landslide beginning at the toe to an
Case 1 2 3 4 5 6 7 intermediate point in the landslide mass. For the slip surfaces
considered, the back-calculated values for the mobilized shear
Natural 0.96 0.85 0.89 1.07 1.09 0.98 0.93 strength along the slip surface fall within a small range. This
1rn fillb 0.86 0.88 0.91 1.08 1.10 0.98 0.93 simplifies the analysis.
2 m fillc 0.90 0.77 0.92 1.10 1.11 0.99 0.94
2 m rise in
The effects of development, such as site grading and changes
water table 0.91 0.79 0.83 1.02 1.05 0.95 0.91 in the piezometric levels, can be evaluated simply by examining
net changes in the calculated factors of safety. In a similar
"Refer to Fig. 13. manner, various remedial measures to arrest movement can be
bNear shore. examined. Experience on similar slopes has shown that a 10%
'Between test holes 108 and 109 change in the factor of safety can significantly affect the
TABLE
5. Factors of safety analysis of section B-B physical stability of the slopes.
For the area under investigation at Regina Beach, modest
Failure surfacea changes in the grading of the toe areas resulted in significant
reductions in the local stability. Although fills constructed at the
Case 1 2 3 4 5 6 7 toe of the slope resulted in increased stability of the landslide as
a whole, movement of a landslide block at the toe of the
Natural 1.13 1.04 0.95 1.00 1.06 1.04 0.94 retrogressive landslide could result in future movements of
4 m water rise 1.02 - 0.83 0.89 0.99 1.00 0.91 upslope portions of the landslide. Thus, any local development
FillatTH107 1.16 - 1.00 1.04 1.08 1.05 0.95 should require minimizing the amount of grading, thereby
Fill as in case 3 reducing the possibility of widespread movement.
plus 4 m
water rise A rise in the piezometric level can be the result of surface
from case 2b 1.06 - 0.88 0.93 1.02 1.01 0.92 infiltration, imgation of developed lots, or breaks in water
mains. Increased slope movement has been measured and
"Refer to Fig. 14. observed as a result of all of these events. This correlates well
bSimulating a water main break at test hole 106. with the results of the analyses.
 68 CAN. GEOTECH. J. VOL. 23, 1986

Construction of buildings in a landslide area must assume that sion to use data; and to Dr. J. Krahn, P.Eng., for his cooperation
some differential vertical and horizontal movement will occur. and assistance.
A proper site reconnaissance can usually choose locations for
buildings that do not straddle landslide scarps. Services such as CHRISTIANSEN,
E. A. 1979. The Wisconsinan deglaciation of southern
Saskatchewan and adjacent areas. Canadian Journal of Earth
water and gas lines should not cross landslide blocks in such a Sciences, 16, pp. 913-938.
manner that the risk of rupture is increased. CLIFTON ASSOCIATES LTD.1983. Site investigation, proposed Borden
The effects of slope movements on structures have been Bridge. Report prepared for Saskatchewan Department of Highways.
observed at Regina Beach for several decades. Movements FREDLUND, D. G., and KRAHN, J. 1977. Comparison of slope stability
occur at variable rates from less than 10 rnrn/year to more than methods of analysis. Canadian Geotechnical Journal, 14, pp.
200 mm in less than a week. Thus, relatively rapid movement is 429-439.
possible, but catastrophic movements threatening human life HAUG,M. D., SAUER,E. K., and FREDLUND, D. G. 1977.
have not been reported. However, the existing landslide mass is Retrogressive slope failures at Beaver Creek, south of Saskatoon,
very sensitive to changes in stress and groundwater conditions. Saskatchewan, Canada. Canadian Geotechnical Journal, 14, pp.
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Calif Dig Lib - Davis on 12/26/14

Unrestricted development of such a geotechnically sensitive 288-301.

INSLEY, A. E., CHATTERJI, P. K., and SMITH, L. B. 1977. Use of
area can result in accelerated movements and extensive property residual strength for stability analyses of embankment foundations
damage. containing preexisting failure surfaces. Canadian Geotechnical
Journal, 14, pp. 408-428.
Acknowledgements KRAHN,J., JOHNSON, R. F., FREDLUND, D. G., and CLIFTON, A. W.
Appreciation is expressed to the Town of Regina Beach and 1979. A highway cut failure in Cretaceous sediments at Maymont,
Saskatchewan Parks and Renewable Resources for their co- Saskatchewan. Canadian Geotechnicai Journal, 16, pp. 703-715.
operation is studies leading to this paper; to Saskatchewan THOMSON, S., and HAYLEY, D. W. 1975. The Little Smoky landslide.
Highways and Transportation for their assistance and permis- Canadian Geotechnical Journal, 12, pp. 379-392.
For personal use only.