THESE TERMS GOVERN YOUR USE OF THIS DOCUMENT

Your use of this Ontario Geological Survey document (the Content) is governed by the
terms set out on this page (Terms of Use). By downloading this Content, you (the
User) have accepted, and have agreed to be bound by, the Terms of Use.
Content: This Content is offered by the Province of Ontarios Ministry of Northern Development and
Mines (MNDM) as a public service, on an as-is basis. Recommendations and statements of opinion
expressed in the Content are those of the author or authors and are not to be construed as statement of
government policy. You are solely responsible for your use of the Content. You should not rely on the
Content for legal advice nor as authoritative in your particular circumstances. Users should verify the
accuracy and applicability of any Content before acting on it. MNDM does not guarantee, or make any
warranty express or implied, that the Content is current, accurate, complete or reliable. MNDM is not
responsible for any damage however caused, which results, directly or indirectly, from your use of the
Content. MNDM assumes no legal liability or responsibility for the Content whatsoever.
Links to Other Web Sites: This Content may contain links, to Web sites that are not operated by MNDM.
Linked Web sites may not be available in French. MNDM neither endorses nor assumes any
responsibility for the safety, accuracy or availability of linked Web sites or the information contained on
them. The linked Web sites, their operation and content are the responsibility of the person or entity for
which they were created or maintained (the Owner). Both your use of a linked Web site, and your right
to use or reproduce information or materials from a linked Web site, are subject to the terms of use
governing that particular Web site. Any comments or inquiries regarding a linked Web site must be
directed to its Owner.
Copyright: Canadian and international intellectual property laws protect the Content. Unless otherwise
indicated, copyright is held by the Queens Printer for Ontario.
It is recommended that reference to the Content be made in the following form: <Authors last name>,
<Initials> <year of publication>. <Content title>; Ontario Geological Survey, <Content publication series
and number>, <total number of pages>p.
Use and Reproduction of Content: The Content may be used and reproduced only in accordance with
applicable intellectual property laws. Non-commercial use of unsubstantial excerpts of the Content is
permitted provided that appropriate credit is given and Crown copyright is acknowledged. Any substantial
reproduction of the Content or any commercial use of all or part of the Content is prohibited without the
prior written permission of MNDM. Substantial reproduction includes the reproduction of any illustration or
figure, such as, but not limited to graphs, charts and maps. Commercial use includes commercial
distribution of the Content, the reproduction of multiple copies of the Content for any purpose whether or
not commercial, use of the Content in commercial publications, and the creation of value-added products
using the Content.
Contact:
FOR FURTHER
INFORMATION ON

PLEASE CONTACT:

The Reproduction of
Content

MNDM Publication
Services

The Purchase of
MNDM Publications

MNDM Publication
Sales

Crown Copyright

Queens Printer

BY TELEPHONE:
Local: (705) 670-5691
Toll Free: 1-888-415-9845, ext.
5691 (inside Canada,
United States)
Local: (705) 670-5691
Toll Free: 1-888-415-9845, ext.
5691 (inside Canada,
United States)
Local: (416) 326-2678
Toll Free: 1-800-668-9938
(inside Canada,
United States)

BY E-MAIL:
Pubsales@ndm.gov.on.ca

Pubsales@ndm.gov.on.ca

Copyright@gov.on.ca

LES CONDITIONS CI-DESSOUS RGISSENT L'UTILISATION DU PRSENT DOCUMENT.

Votre utilisation de ce document de la Commission gologique de l'Ontario (le contenu )
est rgie par les conditions dcrites sur cette page ( conditions d'utilisation ). En
tlchargeant ce contenu, vous (l' utilisateur ) signifiez que vous avez accept d'tre li
par les prsentes conditions d'utilisation.
Contenu : Ce contenu est offert en l'tat comme service public par le ministre du Dveloppement du Nord
et des Mines (MDNM) de la province de l'Ontario. Les recommandations et les opinions exprimes dans le
contenu sont celles de l'auteur ou des auteurs et ne doivent pas tre interprtes comme des noncs
officiels de politique gouvernementale. Vous tes entirement responsable de l'utilisation que vous en faites.
Le contenu ne constitue pas une source fiable de conseils juridiques et ne peut en aucun cas faire autorit
dans votre situation particulire. Les utilisateurs sont tenus de vrifier l'exactitude et l'applicabilit de tout
contenu avant de l'utiliser. Le MDNM n'offre aucune garantie expresse ou implicite relativement la mise
jour, l'exactitude, l'intgralit ou la fiabilit du contenu. Le MDNM ne peut tre tenu responsable de tout
dommage, quelle qu'en soit la cause, rsultant directement ou indirectement de l'utilisation du contenu. Le
MDNM n'assume aucune responsabilit lgale de quelque nature que ce soit en ce qui a trait au contenu.
Liens vers d'autres sites Web : Ce contenu peut comporter des liens vers des sites Web qui ne sont pas
exploits par le MDNM. Certains de ces sites pourraient ne pas tre offerts en franais. Le MDNM se
dgage de toute responsabilit quant la sret, l'exactitude ou la disponibilit des sites Web ainsi relis
ou l'information qu'ils contiennent. La responsabilit des sites Web ainsi relis, de leur exploitation et de
leur contenu incombe la personne ou l'entit pour lesquelles ils ont t crs ou sont entretenus (le
propritaire ). Votre utilisation de ces sites Web ainsi que votre droit d'utiliser ou de reproduire leur
contenu sont assujettis aux conditions d'utilisation propres chacun de ces sites. Tout commentaire ou toute
question concernant l'un de ces sites doivent tre adresss au propritaire du site.
Droits d'auteur : Le contenu est protg par les lois canadiennes et internationales sur la proprit
intellectuelle. Sauf indication contraire, les droits d'auteurs appartiennent l'Imprimeur de la Reine pour
l'Ontario.
Nous recommandons de faire paratre ainsi toute rfrence au contenu : nom de famille de l'auteur, initiales,
anne de publication, titre du document, Commission gologique de l'Ontario, srie et numro de
publication, nombre de pages.
Utilisation et reproduction du contenu : Le contenu ne peut tre utilis et reproduit qu'en conformit avec
les lois sur la proprit intellectuelle applicables. L'utilisation de courts extraits du contenu des fins non
commerciales est autoris, condition de faire une mention de source approprie reconnaissant les droits
d'auteurs de la Couronne. Toute reproduction importante du contenu ou toute utilisation, en tout ou en partie,
du contenu des fins commerciales est interdite sans l'autorisation crite pralable du MDNM. Une
reproduction juge importante comprend la reproduction de toute illustration ou figure comme les
graphiques, les diagrammes, les cartes, etc. L'utilisation commerciale comprend la distribution du contenu
des fins commerciales, la reproduction de copies multiples du contenu des fins commerciales ou non,
l'utilisation du contenu dans des publications commerciales et la cration de produits valeur ajoute l'aide
du contenu.
Renseignements :
POUR PLUS DE
RENSEIGNEMENTS SUR

VEUILLEZ VOUS
ADRESSER :

la reproduction du
contenu

Services de
publication du MDNM

l'achat des
publications du MDNM

Vente de publications
du MDNM

les droits d'auteurs de

la Couronne

Imprimeur de la
Reine

PAR TLPHONE :
Local : (705) 670-5691
Numro sans frais : 1 888 415-9845,
poste 5691 (au Canada et aux
tats-Unis)
Local : (705) 670-5691
Numro sans frais : 1 888 415-9845,
poste 5691 (au Canada et aux
tats-Unis)
Local : 416 326-2678
Numro sans frais : 1 800 668-9938
(au Canada et aux
tats-Unis)

PAR COURRIEL :
Pubsales@ndm.gov.on.ca

Pubsales@ndm.gov.on.ca

Copyright@gov.on.ca

fc) CMNR-OGS 1982

MiniStn/Of
J

Hon AlanW.Pope
Minister

n6SOUrC6S

Deputy Minister

Natural
Ontario

ONTARIO GEOLOGICAL SIJRVEY

Open File Report 5374

Grant No. 7 Horizontal Deep Drains to Stabilize Clay Slopes

by
T.C. Kenney and K.C. Lau

1982

Parts of this publication may be quoted if credit

is. given. it is recommended that reference to this
report be made in the following form:
Kenney, T.C.,and Lau, K.C.
1982: Horizontal Deep Drains to Stablilize Clay
Slopes, Ontario Geological Survey OFR 5374,
23 p., 10 figures.

Ontario Geological Survey

OPEN FILE REPORT

Open file reports are made available to the public subject to the following conditions:
This report is unedited. Discrepancies may occur for which the Ontario Geological Survey does
not assume liability. Recommendations and statements of opinion expressed are those of the author or
authors and are not to be construed as statements of government policy.
Open file copies may be read at the following locations:
Mines Library
Ontario Ministy of Natural Resources
8th Floor, 77 Grenville Street, Toronto
The office of the Regional or Resident Geologist in whose district the area covered by
this report is located.
Handwritten notes and sketches may be made from this report. Check with the Library or Region
al or Resident Geologist's office as to whether there is a copy of this report that may- be borrowed.
The Library or Regional or Resident Geologist's office will also give you information on copying ar
rangements. A copy of this report is available for Inter-Library Loan.
This report is on file in the Regional or Resident Geologists' office(s) located at:

All Regional and

Resident Geologists

The right to reproduce this report is reserved by the Ontario Ministry of Natural Resources.
Permission for other reproductions must be obtained in writing from the Director, Ontario Geological
Survey.

E.G. Pye, Director

Ontario Geological Survey

Report to
MINISTRY OF NATURAL RESOURCES, ONTARIO
Regarding Geoscience Research Grant GR-7

"Horizontal Deep Drains to Stabilize Clay Slopes"

by
T.C. Kenney and K.C. Lau
Department of Civil Engineering
University of Toronto

January 1982

Abstract
Introduction
Research Program
Test Site .
Ground Water Conditions
Field Test

Parametric Study
Design of Drainage Systems
Conclusions
Acknowledgements
References
Figures

Horizontal Deep Drains to Stabilize Clay Slopes

by

T.C. Kenney and K.C. Lau

Department of Civil Engineering
University of Toronto

ABSTRACT

Horizontal drains were installed in a natural clay slope and

changes of piezometric level were monitored by means of piezometers.
These measurements showed that piezometric levels decreased in the
vicinity of the drains, demonstrating that horizontal drains in clay
slopes would increase the stability of the slopes.
On the basis of the theory of consolidation estimates were made
of the changes of piezometric level resulting from installation of
horizontal drains. A comparison of these estimates and the measured
*
values showed very good agreement, indicating that the method of
prediction was sufficiently accurate to use in design.
For purposes of designing drainage systems for slopes computer
programs are available for the following calculations. Estimates can
be made of piezometric levels and their changes with time on the basis
of soil properties, hydraulic boundary conditions and theory of
consolidation. The three-dimensional water pressure conditions are
then averaged to give equivalent two-dimensional conditions, and these
are used in stability calculations to determine the safety factor and
changes of safety factor of the slope.
The effectiveness of horizontal drains depends on the spacing and
diameter of the drains and their location relative to the zone
containing the critical slip surfaces.
This study has shown that horizontal drains can be designed to
improve the stability of clay slopes. Additional work is required to
develop low-cost methods of installing the drains.

-2-

I INTRODUCTION

Extensive regions of Ontario have problems related to instability

of slopes in weak clay soils. The objective of this research project
was to investigate the effectiveness of deep horizontal pipe drains to
reduce piezometric levels and ground water pressures within clay slopes
and thereby to improve the stability conditions of clay slopes. The
rationale is that clay soils are frictional materials and reduction of
ground water pressures will increase contact forces between mineral
particles and hence will increase the shear strength of the soil and
the stability of the slope.
Details of the research project are reported by Lau (1982) and
the object of this report is to describe in brief the main findings.
Deep horizontal drains have been extensively and successfully
used in the past in slopes of soil and rock containing isolated zones
and layers of pervious, water-bearing materials. A review of this
experience has been made by Hutchinson (1977). Only one case has been
described in the literature of the use of horizontal pipe drains in
slopes in massive deposits of soft clay (La Rochelle, Lefebvre i
Bilodeau 1977). In this case the drains were judged to have improved
i
the stability of the slope, but evidence of reduction of piezometric
levels was not obtained by field measurement.
An important case, concerning the misuse of drains in a clay slope
has been reported by Lefebvre, Lafleur S Chagnon (1976). The drains
were vertical, and measurements of piezometric levels indicated that
these drains caused ground water pressures within the slope to increase
They promoted water infiltration and, thereby, caused the stability of
the slope to decrease.

RESEARCH PROGRAM

There were three goals for the research program: (i) to install
drains in a clay slope in order to confirm by monitoring piezometric
levels that horizontal pipe drains will cause decreases of ground water
pressure, (ii) to develop an analytical method for predicting changes

-3-

of piezometric level which result from the installation of drains,

and to test the accuracy of the method by predicting results for the
test site and comparing them to the results of measurement, and (iii)
to develop a method of using the predicted changes of piezometric
level in slope stability calculations in order to determine the
improvement of slope stability caused by the drains.
All three goals were achieved.

TEST SITE

There was a suitable clay slope for the field test at New Liskeard,
Ontario where piezometric levels had been monitored in detail over the
previous eight years.
Details concerning the geotechnical conditions at the test site
have been presented by Kenney and Chan (19721. The site is on the
eastern bank of Wabi Creek on the property of the New Liskeard College
k
of Agriculture Technology, about l km north of the city of New Liskeard.
Topography is presented in Fig. 1. The site is located within the
boundaries of glacial Lake Barlow, and the soils are lacustrine sediments.
Fig. 2 records the stratigraphy and the properties of the soils,
determined by Chan and Kenney (1972), Kenney and Chan (1972) and Lau (1982)
For purposes of an earlier research project 70 piezometers had
been installed throughout the slope to determine piezometric levels and
their temporal variations. A representative sample of the results is
contained in Fig. 3.
Wabi Creek flows into Lake Temiskaming and the river level at
the test site is essentially equal to the lake level. Lake level is
controlled for purposes of hydro-electric power generation, falling
during the winter period when there are large demands for electric
power, rising rapidly during spring melt and remaining relatively
constant during the summer and autumn periods (Fig. 3). Stability of
the slope at the test site changes with time as the result of two
variables; that is, river level and piezometric levels. Results of
stability calculations are presented in Fig. 3 and it is apparent that

-4-

the most important factor is river level. For this particular slope
the least stable condition occurs during late winter when the river
level is at its minimum annual level.
The relationship between location of calculated critical slip
surface and river elevation is given in Fig. 4, indicating that the
bottom of the critical slip surface is usually about 6 to 7 metres
below the river level. It is clear that horizontal drains could not
be installed at this level but would have to be installed at a higher
level at which they could not be expected to cause large improvements
of the stability of the slope. Therefore, the purpose of the field
test was solely to determine the effect of the drains on piezometric
levels in the clay in the vicinity of the drains.
At the test site precipitation averages about 760 mm of water
per year, distributed as 50 mm per month during winter and spring and
75 mm per month during summer and autumn.
*
GROUND WATER CONDITIONS

In this report consideration will be given to slopes in deep deposits

of clay which do not contain aquifer or drainage layers. However, the
principles to be discussed can be used to deal with such conditions.
Ground water flow through clay can be considered as a case of flow
of an incompressible fluid through a compressible porous medium. Where
the soil remains saturated the change of total flux of water being
conducted through the soil plus the rate of change of volume of void
space occupied by water must be zero (conservation of mass). This is
the process of consolidation or swelling and changes of piezometric
level occurring during the process can be estimated approximately
through the use of the consolidation equation:

ly.
3t ~

Hi

Tw 3t

...l

-5-

where u is ground water pressure, t is time, v w is unit weight of water,

h is piezometric elevation (elevation of water surface in a piezometer
standpipe), a is total stress, a 1 is effective stress, k is coefficient
of hydraulic conductivity, m is coefficient of volume change of soil
expressed as change of porosity per unit change of effective stress
(m - -r^r),
and x, y and z are coordinate axes.
do
Within a slope in a deep deposit of clay change* orf pie.zometsu.c
teveJL wUUL onty ocean, a* the. Jie*uJtt o fi change* ojj apptied *tne** and
change* ofi hyd/iau&ic condition* on the. *LLn.^ace, Ojj the. *tope.. Attention
will first be focused on conditions which exist at the ground surfaces,
and then consideration will be given to how changes of these conditions
would influence piezometric levels within the soil mass.
Fig. 5 presents a typical pattern of ground water flow in the Wabi
Creek slope. Piezometer measurements made on 15 April 1974 were used
to determine locations of the equipotential lines. Flow lines have been
sketched in to account for the fact that the soils have anisotropic
permeability properties (k horizontal > k vertical).
It is shown in Fig. 3 that piezometric levels vary with time and
therefore the pattern of ground water flow will also vary with time.
The areas where changes are initiated are the area of infiltration at
the head of the slope and the area of discharge at the river. In the
area of infiltration hydraulic conditions are governed by rate of
infiltration, which is influenced by the availability of water, the
Sh
hydraulic gradient at the ground surface (i * ~~)
and the coefficient
dZ
of hydraulic conductivity of the sotl. During any period of time when
the infiltration rate for a slope is less than the discharge rate
piezometric levels will decrease and the phreatic surface (locus of
points at which u s 0) will fall. Conversley, when infiltration rate
exceeds discharge rate, piezometric levels will increase and the phreatic
surface will rise. The highest possible position of the phreatic
surface is the ground surface.
In Fig. 6 are presented the measured ranges of piezometric elevation
and position of phreatic surface below the area of infiltration over a
time span of several years. The measurements show the influence of
both annual changes and longer-term changes. Minimum piezometric

-6-

elevations occur during late winter and maximum values occur during the
summer/autumn period when rainfall is plentiful. Maximum changes occur
at the ground surface and, at this site, annual changes do not propogate
below a depth of about 5 metres. The phreatic surface is at the ground
surface during the summer/autumn months and, as shown in Fig. 6, falls
during the winter period, the maximum fall being about 2.5 metres in 1978.
In the area of discharge changes of boundary conditions are caused
by changes of the river level, and this has two effects; it causes (i)
changes of total stress on the ground surface and (ii) changes of
piezometric elevation along the submerged ground surface. As indicated
in Fig. 3 by measurements of piezometers 17, 18 and 19, the first
effect causes immediate changes of piezometric levels in the soil adjacent
to the river; increases of river level cause increases of piezometric
level and vice versa. The second effect causes slower changes by
consolidation or swelling which are complementary to the first effect.
Fig. 7 presents the results of an estimate made of changes of
piezometric level below the area of infiltration based on the tlteory
of consolidation (Equation 1). The calculations were made for the
winter period when normally the ground surface is frozen and there is
zero infiltration. Immediately before freeze-up piezometric levels are
at, or close to, their maximum annual values, whereas at the end of
winter they are at, or close to, their minimum annual values. Therefore,
the changes of piezometric level occurring during the winter period are
indicative of maximum annual changes, and estimates can be compared to
the results of field measurements summarized in Fig. 6. The calculations
involved the following assumptions.
(a) hydraulic gradient immediately before freeze-up was i s 0.5
(see Fig. 5),
(b) hydraulic gradient at the ground surface immediately after
freeze-up was i * O (zero infiltration),
(c) water movement was vertical (one-dimensional consolidation),
and
(d) c V - k/Y..W m * 0.10 m27day for the soil (see Fig. 2).
In Fig. 7 the period of time equal to 100 days approximates the
length of winter time when infiltration would be zero. The measured and

-7-

calculated results in Figs. 6 and 7 compare quite closely. This finding

supports the use of consolidation theory to estimate piezometric levels
in natural clay slopes and in slopes in which drains have been installed.
The following remarks can be made regarding ground water conditions
in the slope at the test site, and these remarks are generally applicable
to all slopes in clay soils. Reference is made to Tigs. 3 and 6.
(a) Fluctuations of piezometric levels are greatest near the
ground surface and decrease substantially at increasing
distances from the ground surface. Below the infiltration
area at the head of the slope significant fluctuations did
not occur beyond a depth of about 5 metres.
(b) The rate of fluctuation of piezometric level is maximum
near the ground surface and decreases substantially with
increasing depth.
(c) Rapid changes of piezometric level can be expected adjacent
to a river bank in response to rapid changes of river level.
In the absence of a fluctuating water body it can be expected
at the toe of the slope in the area of discharge that the
annual changes of piezometric level will be small.
(d) Maximum values of piezometric levels usually occur during
late spring/early summer or during autumn/late autumn periods.

FIELD TEST

In April 1979 five horizontal drains were installed. The locations

are shown in Figs, l and 2. Each drain is 50 mm diameter and 35 m long
and the spacing between drains is 12 m. Values of length and spacing
were chosen on the basis of suggestions made by Kenney, Pazin and
Choi (1-977), and the choice of diameter was based on the drilling .
equipment to be used and on the sizes of available plastic pipe.
The drain pipes were constructed at the University of Toronto and
consisted of lengths of perforated plastic pipe covered with plasticfibre filter fabric. Each section of drain pipe was 3 m long. Slots
were cut with a ski 11-saw at 40 mm intervals along the pipes and the
slots were 3 mm wide and about 30 mm long. Filter fabric was sewn into
the form of tubes which were slipped over the pipes and held in position
by an adhesive tape. In the field the sections of pipe were connected
together by means of couplings and cement.

-8-

It was decided to install the drains at the lowest possible

elevation. The work of installation was done when the river level was
low during late winter. A working surface was constructed of packed
snow on the river ice. Holes for the drains were drilled with standard
diamond-drill equipment using BW casing and large quantities of wash
water. The holes were inclined upwards at an angle of about l degree.
The drilling rate averaged 4.5 m per hour. When the hole reached the
desired length of 35 m an AW drill rod with washing bit was used to
clean out the casing. Sections of drain pipe were then pushed into the
casing and coupled together. When the drain reached the end of the
hole it was pushed about l m into the clay so as to anchor it in position
The casing was then withdrawn, a screen placed over the end of the drain
and a length of metal pipe inserted over the end of the drain to protect
it. Installation of each drain was completed as quickly as possible to
reduce the possibility of the drill casing becoming stuck to the clay.
This indeed happened with the first hole after the casing had been left
in the slope overnight, and it was with great difficulty that it' was
finally freed.
Final alignment of each drain was determined with a bore-hole
directional inclinometer. The maximum deviation from the desired
location was approximately 0.6 m, which is surprizing!y small considering
the method used to make the hole.
Prior to installing the drains piezometers were installed along
Sections AA and BB in Fig. l on both sides of the centre drain.
The piezometers were positioned at 3 m intervals measured from the
drain in horizontal and vertical directions. The purpose of these
piezometers was to monitor changes of piezometric level which were
caused by the drains.
In fact, during one and a half years after installing the drains
piezometer measurements indicated that any changes of piezometric
level which occurred after the installation of the drains were
basically equal to those changes which had annually occurred before
the drains were installed. There were several possible explanations
of this result. One possibility was that the clay had squeezed
tightly around the drain pipes and had compressed the filter cloth

-9-

to such an extent as to render it rather impervious. Another possibility

was that the zone of influence of the drains was less than 3 m in
radius and changes of piezometric level were not occurring at the
distances at which the piezometers were located.
To test the first possibility a pressure-packer device was
constructed which would seal off a 0.5 m length of drain, apply water
pressure to the inside of the drain and measure the rate of flow out
of the drain and along the filter cloth. The drains were tested over .
their entire lengths and it was found that there was no restriction
to flow and that the drains in the ground were as pervious as those
tested in air.
To test the second possibility piezometers were inserted in
Sections AA and BB to obtain vertical profiles of piezometric levels
at several horizontal distances from drain No. 4. The method of
measurement was to use vibrating-wire electrical piezometers attached
to the end of diamond-drill rods, to push each piezometer vertically
downwards into the soil to a specific depth, to leave the piezometer
at this location until a constant value of piezometric elevation was
measured (usually about 3 days) and to repeat the process by pushing
the piezometer down to a new location. The results of the measurements
are presented in Fig. 8. The accuracies of the determined elevations
(piezometers and piezometric levels) are excellent whereas the
accuracies of the horizontal distances between piezometers and the
vertical plane through the drain are uncertain. All measurements
correspond to a time interval after drain installation of 500 days.
The measurements in Fig. 8 indicate that the piezometric levels
in soil close to drain No. 4 were substantially reduced but that the
radius of influence of the drain was rather small; about 2 m at Section
AA and l m at Section BB. (This explains why the piezometers placed
3 m from drain No. 3 did not register any substantial decreases of
piezometric level). The size of the zone of influence is dependent
upon the original pressure head at the drain location (pressure head s
piezometric elevation - elevation of drain) and upon elapsed time.
The size of the zone will increase with time until steady-seepage
conditions exist.

-10-

Two broad conclusions can be drawn from these results.

(a) Horizontal drains in this clay slope caused reduction of
piezometric levels.
(b) The size of the zone around drain No. 4 in which piezometric
levels were reduced was rather small.
Also presented in Fig. 8 are the results of calculated changes of
piezometric level using consolidation theory. Initial ground water
conditions were determined from piezometer measurements made prior
to installation of the drains, and it was assumed that over the 500-day
time period the phreatic surface would remain at the ground surface.
The soil boundaries and properties shown in Fig. 2 were used. Calculations
were made with the computer program developed by Narasimhan (1975), which
utilizes an integrated finite-difference method to apply Equation 1.
In Fig. 8 the values of measured and calculated piezometric level
compare quite well, indicating that the method of calculation can be
used with some degree of confidence.
Fig. 9 gives an indication of the rate at which average ground
water pressures are expected to change as the result of installing the
drains and are expressed as fractions of the ultimate changes of
average ground water pressure. From the results of the field measurements
made 500 days after drain installation a value of average change of
ground water pressure was estimated. In this particular case it is
estimated that about 5 years will be .required for the influence of the
drains to be fully realized.
k

PARAMETRIC STUDY

Results are presented in Fig. 10 of a study made of the influence

of drain spacing and drain diameter on the effectiveness of a system of
horizontal drains to improve slope stability. The study used the
conditions of the Wabi Creek slope with constant river level at
elevation 176 m. The length of the drains was such that all slip
surfaces intersected the drains.
The study consisted of the following steps.
(a) The minimum safety factor of the slope was estimated for
ground water conditions corresponding to no drains.

-li

A combination of drain diameter and drain spacing was

selected. Using the method of calculation described in
the previous section piezometric levels within the slope
were estimated for several time intervals. The threedimensional distribution of piezometric levels was averaged
to give an equivalent two-dimensional distribution for use
in a two-dimensional slope stability analysis. Stability
calculations were made and the minimum safety factor
determined. This result was expressed as a percentage
increase of safety factor relative to the "no-drain"
condition.
(c) Additional combinations of drain spacing and drain diameter
were used in calculations of piezometric levels and safety
factor of the slope.
The results of this study, presented in Fig. 10, are for two
values of drain diameter, the 5-cm size representing perhaps the
smallest practical size of drain and the 15-cm size representing the.
maximum practical size. The results clearly indicate that reduction of
drain spacing and/or increase of drain diameter will increase the
improvement of safety factor. Although not shown in Fig. 10, the
results also clearly indicate that by decreasing drain spacing and/or
increasing drain diameter more ra.pid changes of piezometric level and
safety factor will occur.
The results of this study cannot be viewed as anything else but
a general indication of the effects of drain spacing and drain diameter.
Each specific slope would behave somewhat differently but would be
expected to follow the general pattern indicated in Fig. 10.

The method which is recommended for selecting the spacing, diameter

and elevation for a system of horizontal drains is a trial-and-error
process of calculation and consideration. Specific drainage systems
are analysed to determine the improvement they cause to the safety
factor of the slope, and this is then considered in relation to the
estimated costs of the drainage systems.
The accuracy of the calculations is largely dependent on the
accuracy of the information concerning the site; topography, boundaries
of soil units, soil properties, piezometric levels and hydraulic

-12-

boundary conditions. Without reliable information the results of

calculations are not dependable.
Computer programs are available for making the calculations
(Lau, 1982).
The following steps are required for the design calculations.
(a) Initial conditions of the slope are established based on
field explorations, field measurements of piezometric level
and laboratory determinations of soil properties. Worst
conditions for the hydraulic boundary conditions should be
assumed.
(b) Stability calculations are made to determine the variation
of safety factor for slip surfaces at different depths for
the conditions of "no drains".
(c) A choice is made of a combination of drain spacing and
diameter. A drain elevation is also chosen. Because the
critical slip surface usually cuts below the toe of the
slope the most effective location of the drains is usually
at the level of the toe of the slope.
Ultimate changes of piezometric level are calculated using
consolidation theory. The three-dimensional distribution
of piezometric level is converted to the equivalent twodimensional distribution.
Slope stability calculations are made to determine the
variation of safety factor for slip surfaces at different
depths.
(d) Step (c) is repeated for different values of spacing,
diameter and drain elevation.
(e) Selection of the drain system from the results of these
calculations is based on the estimates of improvement of
slope stability resulting from the drains and on the
estimates of cost of the drains.

CONCLUSIONS
1.

2.
3.

Horizontal drains installed in a slope of natural, soft clay

caused measured reductions of piezometric level in the clay
adjacent to the drains. It was thus demonstrated that horizontal
drains can improve the stability of clay slopes.
Based on theory of consolidation, estimates were made of changes
of piezometric level which would be caused by the drains, and
these changes were similar to the measured values.
Computer programs have been developed for calculating improvements
of slope stability which result from installation of systems of
horizontal drains.

13-

By considering the beneficial results and the costs of a number

of different combinations of spacing, diameter and elevation,
selection of an effective drainage system can be made.

ACKNOWLEDGEMENTS

This research project was financially supported by Geoscience

Research Grant No. 113 of the Ministry of Natural Resources, Ontario
during the period 1978-81.
Sincere thanks are extended to Mr. J. Butler, Principal of the
New Liakeard College of Agriculture Technology, for permission to use
Wabi Creek slope as a test site.
Professor J. Curran provided considerable help with the theoretical
work. Mr. A. Godmaire made the field measurements and was of invaluable
assistance.

REFERENCES

Chan, H.T. and Kenney, T.C., 1973. "Laboratory investigation of

permeability ratio of New Liskeard varved soil". Canadian. Geotechnical
Journal, Vol. 10, No. 3, pp. 453-472.
Hutchinson, J., T977. "Assessment of the effectiveness of corrective
measures in relation to geological conditions and types of slope
movement". Symposium on Landslides and Other Mass Movement, Theme 3;
General Report. Bulletin AIGI, No. 16, pp. 131-155.
Kenney, T.C. and Chan, H.T., 1973. "Field investigation of permeability
ratio of New Liskeard varved soil". Canadian Geotechnical Journal,
Vol. 10, No. 3, pp. 473-488.
Kenney, T.C., Pazin, M. and Choi, W.S., 1977. "Design of horizontal
drains for soil slopes". Journal of Geotechnical Engineering Division,
ASCE, Vol. 193, No. GT11, pp. 1311-1323.
La Rochelle, P., Lefebvre, G. and Bilodeau, P.M., 1977. "The stabilization
of a slide in Saint-Jerome, Lac Saint-Jean". Canadian Geotechnical
Journal, Vol. 14, No. 3, pp. 340-356.
Lau, K.C., 1982. "Horizontal drains in clay slopes", Ph.D. thesis,
University of Toronto. Submitted for examination.
Lefebure, G., Lafleur, J., and Chagnon, J.Y., 1976. "Evaluation of
vertical drainage as a stabilizing agent in a clay slope at Hull, Quebec".
29th Canadian Geotechnical Conference, Vancouver VI 16- VI 30.
Narasimhan, T.N., 1975. "A unified numerical model for saturated unsaturaged groundwater flow". Ph.D. thesis, University of California,
Berkeley.

metres
O ~20 4fJ 60
Level G round
(Elev. 189.5-190 m)
Horizontal drains
1
2 ,3 .4 ,5
A

Edge of water

Wabi River

Figure 1. Planet site

196

C/)

l.
"55

E
c

33

186
i^-^^-"^^
~*^^

Horizontal drain
__ __

^ - 0.10 mVd
kv - 2 x 10"0 m/s

__ __ __ __ ^cv z 0.07 m27d

Grey'varved soil
Thick-layered zone

JO

"S

Q)
111

Homogeneous silty clay

^^^^^"^ Grey varved soil

^^^'^
Thin-layered zone

Wabi

176

^^^-^""^

166

156

kv - 3 x 10-" m/s
kn * 6 x 10-10 m/s

20

30

40

50

60

70

80

90

Distance normal to the river bank D, metres

Figure 2. Cross section of slope

below ground surface, m

190c 180
o
^ 170

Distance from Wabi Creek, D, m

O 10 20 30 40 50 60 70 80 90
^IIlTjllil
M

Wabi Creek
.18

***^

.22
.32
.70

! 160

ULJ

150

Locations of Piezometers

Figure 3. Record of field measurements

200

to

178

CO

ffi
"S 190

180
co

170

.2
UJ

160

o d"

Max. river level

.j ^

176
174

O CO

Zone containing critical slip surfaces

O 10 20 30 40 50 60 70
Distance from Wabi Creek, metres
Figure 4. Relationship between calculated
location of criticial slip surfaces
and river level.

55-172
5; CO

.2 170
Bottom of critical
Ufi
slip surfaces at
u
different times
168
1
174 176 178
River elevation, metres

25 35 45 55 65 75 85 95
Distance from Wabi Creek, metres
Figure 5. Flow pattern from piezometer measurements
on 15 April 1974.

Piezometric Elevation, metres

170

175

190

180
185
190
Ground Surface

Range during 1972-77

Measurements August 198

co

Q)

185

^'
(D

l 180
V-"
CO
jl)

175

170

scoo ^25g
O

Ground Surface

Winter period
Zero infiltration

JC
"d CO

(0

ASONDJFMAMJJ

Month of year
Figure 6. Measured changes of piezometric elevation and
phreatic surface below infiltration area.

Piezometric Elevation, metres

O

10
l

metres
2

Ground Surface

Assumed conditions
at beginning of winter

r^-

o(D

of
Lowering SurPhrefaatce,ic

Time, days
20 40 60 80 100

Figure 7. Predicted changes of piezometric

elevation and phreatic surface for
condition of zero infiltration.

Ground surface
189

CO
(D

Measured values
A before drains installed
186 - D 0.0 m from drain no. 4
* 1.5 m from drain no. 4
Predicted values
1.5 m from drain no. 4
0.0 m from drain no. 4

drain elevation

c
.g
v*

CO

.92

180
177

UJ

Location A - Figure 1
Time 500 days

174
171

Measured values
A before arams installed
D 0.3 m from drain no. 4
9 1.8 m from dram no. 4

184
CO

d)
g
o

182
drain elevation

180

03

LU

Predicted values
1.8 m from drain no. 4
0.3 from drain no. 4

178
176

176

Location B - Fig. 1
Time 500 days

178

180
182
184
Total head, metres

186

Figure 8. Changes of piezometric elevations

after installation of drains.

CD
D
CO
CO
Q)

Q)
CO

4-*

T3
C
D

CD

1100
CD

73 75
D
C

From field measurements

and indicating the range
of uncertainty

o 50
O
o c

2
o

O)

co

CD

CD Ok
D) "
C
CO

.c
O
O)

(D

rt

CO

JC

Time in years

Figure 9. Rate of change of average ground water pressure in

Wabi Creek slope due to installation of drains.
o

CO
LL

20

co

t/)
o

CO
CO
CD
O

CD
O)
CO
4-J
C
CD
O
CD

QL

Limits of bottom of slip surface

15

Test site condition

s z 12 m
d s 5 cm

10

369

12

Drain Spacing S, metres

Figure 10. Results of parametric study using Wabi Creek slope
as an example.