CHAPTER

14
Alternatives to Full Remediation
of a Landslide

U
nder normal circumstances a landslide is stabilized 14.2 MAINTENANCE
with a margin of safety that meets the standard of
practice. However, there are many situations where Railroad owners and highway agencies have maintenance costs
full stabilization is impractical due to factors such as the built into their annual budgets. Since the staff already is includ-
large size of the landslide, environmental and ownership ed in the budget, additional maintenance is usually a marginal
restrictions, excessive cost, etc. In these cases, it is often con- extra cost, i.e., for materials or possibly emergency/overtime
cluded that nothing can be done, but some action is often situations. It can be scheduled for low-demand periods. For a
preferable to no action and can achieve worthwhile objec- landslide that is moving slowly, or moving periodically during
tives. This chapter discusses some alternative options to full wet weather, maintenance can be the best option. It may be the
remediation. only option on very large landslides where remedial costs are
extremely high.
When compared to the capital cost of landslide remedia-
tion (often exceeding $1 million), the cost of maintenance is
14.1 NO ACTION
usually an economical alternative. However, there is the con-
siderable nuisance effect of annual road closures on the public,
The option of “do nothing” is, of course, plausible in many sit- especially in a high-traffic area. Of more concern is the risk
uations. It is sometimes presented in a geotechnical report as if that a major movement may occur that could endanger public
the writer has just discovered Archimedes Principle! When it safety and property.Therefore, as a minimum, a landslide that
is included as a technical recommendation, there is a need to is moving sufficiently quickly to require continual road resur-
discuss the technical issues and likely consequences that may facing or shoulder reconstruction should be examined in
not be readily apparent to the owner. some depth to determine the causative factor involved. If cau-
An example of taking no action might be an owner who sation is likely to become more dangerous in the future (such
has monitored a landslide over a number of years. A review as river erosion, sea erosion, springs, etc.), then maintenance
of the data may show a slowing, constant, or accelerating may not be a suitable option because of the increasing risk
trend. The landslide consultant can offer to review the data with time. Similarly, if the amount and frequency of repairs are
and discuss the likelihood of damaging or dangerous conse- increasing for no apparent reason, it may be indicative of
quences in the future. If the rate of movement is slowing or delayed failure developing in a stiff clay, and some action will
constant, and the technical reasons are understood, no action need to be taken.
may need to be taken (see Case History 1). By contrast, the A good candidate for maintenance is a slow-moving land-
owner may not be aware that an accelerating rate of move- slide in which the rate of movement does not change signifi-
ment in a stiff clay may be the prelude to more catastrophic cantly with time. An example might be a colluvial slope
(delayed) failure later (Case Histories 6 and 8). moving along an ancient slip surface at residual strength.

269
270 REMEDIAL AND PREVENTATIVE OPTIONS

Before selecting maintenance as the option of choice, the measurements. The elapsed time between observation and
following issues need to be considered: interpretation will depend on the quantity of data and need
for the results, but it is important that the process takes place as
• Is there a moderate to high risk of significant and abrupt
soon as practicable. There are some owner-clients who want
loss of ground that could endanger the public? An exam-
their own staff to take the readings for various reasons: cost-
ple might be a landslide caused by erosion at the outside
saving, availability of staff at the site, training, etc. With a few
bend of a river. During flood, more rapid erosion could
exceptions, this is not a good practice and should be discour-
undermine the road (or railroad) and quickly cause signif-
aged.Too often, the assigned personnel are disinterested in the
icant loss of ground at a time when remediation would be
tasks and make careless errors. Worse, the raw data are often
very difficult.
put in a desk drawer and forgotten about for long periods; it is
• Is continued movement likely to regress and endanger
then too late to correct for missing or questionable readings.
upslope structures or facilities?
Therefore, the consultant responsible for the geotechnical
• Will continued movement break culverts, water,
work should also be responsible for collecting and analyzing
stormwater, or sewer pipes within or close to the land-
the field observations.
slide? Such pipes can often tolerate small movements, but
Observations of movements or potential movements of
major breaks may cause flow slides during storms.
hillsides are sometimes required in construction contracts and
• Are there other sensitive utilities within or close to the
are included in the bid price of each contractor. In the United
slide area? Breaks in underground cables, high voltage
States (and elsewhere), a contractor may collect bid prices
lines, pressure gas pipelines, etc. can be dangerous to the
from several engineering consultants and include the lowest
public and very damaging to the local economy should
price (or an even lower price of his own) in the bid docu-
failure occur.
ment. The successful bidder may then re-open the “bidding”
The decision to favor maintenance over remediation is for the instrumentation subcontract to obtain a lower price.
usually dictated by cost. However, the economic and poten- From the perspective of past experience, the author can relate
tially life-threatening risks listed above need to be considered that the results have varied from generally unsatisfactory to
because the potential for litigation from unsafe conditions and disastrous. Many of the more reputable geotechnical specialists
economic losses to third parties are relevant to engineering simply refuse to participate in a bidding format.The best way
cost comparisons. to handle these requirements is to remove this requirement
from the contract and treat field instrumentation as a neces-
sary part of the project’s design. In this way, the tasks of read-
14.3 OBSERVATIONS ing and interpreting are left with the design team. However,
contractual language and insurance require careful attention
Monitoring Landslides to avoid potential problems.
Observations in the form of inclinometer measurements or Monitoring is especially useful for landslides subject to
surface survey points are commonly employed in landslide third-party oversight, such as regulatory authorities or adjoin-
analysis. Owners may elect to continue taking readings rather ing landowners, to show that the rates of movement are non-
than to undertake remediation. This is a cost-effective threatening. It is also a good practice to take additional
approach because controlled monitoring may be all that is readings on inclinometers after a landslide has been stabilized
needed for a slow-moving landslide. Once the pattern of to demonstrate that remediation has been successful. Such an
movement has been established by monthly or quarterly read- offer can sometimes facilitate a speedier resolution of a lawsuit
ings, for example, the frequency of readings can be progres- involving a landslide, especially to parties skeptical of the
sively reduced to yearly or longer intervals. remediation outcome.
There needs to be a sufficient number of control points
such that a loss of one or more points will not materially affect Examples of Observation Usage
the landslide control measure. For example, on medium to 1. At the Washington Park Reservoirs Slide, Portland (Case
large landslides, three or more lines of survey hubs could be History 1), the hillside has been monitored for more
installed, each line having 4 to 10 hubs. The reference point than 100 years, possibly a record for any landslide in the
itself must be stable, which is self-evident but a critical require- United States. Some survey hubs have been damaged or
ment in landslide areas. An engineering geologist should destroyed over this period, but most have survived.The
check that the reference point is on stable ground, such as an ancient slide moves intermittently in response to winter
in-place rock outcrop or stable ground at some distance from rainfall. However, when the data were subdivided into
the landslide. Where inclinometers are being used, the stable 22-year intervals, the rate of movement over these peri-
reference point is the casing segment anchored below the ods was approximately uniform, indicating that the
landslide slip surface. underlying clay is at residual strength and is unlikely to
Readings taken on inclinometers, piezometers, survey change in the future unless there are manmade changes.
points and any other on-site devices should have the data cal- Although there has been extensive damage to a pump
culated, plotted, and interpreted within 1 to 7 days of the field house at the base of the slope, the landslide movements
 ALTERNATIVES TO FULL REMEDIATION OF A LANDSLIDE 271

of around 0.3 inch per year are considered to be minor (i) unacceptable horizontal curves, (ii) right-of-way purchase,
and no remediation has been undertaken. In addition (iii) relocation of utilities, and (iv) destabilizing the uphill side
to surface survey hubs, there are eight inclinometers of the cut.
that are read quarterly. Relocating the alignment into the slope is often accompa-
2. The Washington Park Station Slide (Case History 4) is nied by preventative measures such as a wall to support the
on a hillside close to the Washington Park Reservoirs inner slope. Another consideration is that the landslide will
Slide, and is built over similar ancient landslide terrain. regress back into the realigned road which, with the curves
High-speed elevator shafts pass through the landslide introduced by the original realignment, would increase the
into the underlying bedrock; the two shafts have very risk to road users.
limited tolerances for lateral movements at the land- Realigning the road onto the valley floor takes the road off the
slide slip plane. Although marginal stabilization (see hillside to flat ground.This alternative has become less attrac-
later) has been provided, the owner is relying on an in- tive in recent years because the flat ground is often floodplain
place inclinometer and other conventional inclinome- or wetlands. Permits to build over wetlands can take years to
ter installations to monitor the rates of movement. If obtain or may be refused.
the rate of movement exceeds 1⁄4 inch in 10 years, addi- When Interstate 5 was built through Kelso,Washington, in
tional remedial treatment will be implemented. the mid-1960s, a 1-mile stretch of the proposed freeway align-
However, there is a good chance that this threshold ment passed along the base of the hillside identified as ancient
will not be reached, in which case there will have been landslide terrain. Four separate active landslides were identi-
a substantial cost saving. fied as impacting the alignment.The state decided to relocate
3. The Leaburg Canal Hydroelectric project at Leaburg, the freeway onto the floodplain of the Coweeman River
Oregon, covers a hillside of weathered andesite on below the slope rather than to stabilize the four slides.
which six landslides, ranging from medium to large, Relocating the facility is a possible alternative in some cases.
were mapped. One of these landslides, Percy Slide, However, many projects are built on or adjacent to active
continued to move at an increasing velocity and has landslides or ancient landslide terrain through necessity.
been stabilized. The other five landslides have been Natural gas pipelines, water transmission pipelines, and
monitored since 1987 using inclinometers. All five high voltage electrical transmission lines can be routed to
were read at 3-month intervals initially. After three avoid landslide areas. A thorough geological reconnaissance,
years, two slides with intermittent movements were including ground traverses, topographic and terrain map
reduced to annual readings; three landslides showing reviews, aerial photograph reviews, etc., can avoid these prob-
no movement were reduced to 3-year intervals, and lem areas during the planning stage. Established pipelines in
later to 5-year intervals. rural areas can often be relocated at a reasonable cost because
Observations are a cost-effective way of managing pipelines are usually buried at a shallow depth.
this landslide complex because it keeps the owner Bridging over a landslide, when only the headscarp affects
informed of the slope stability status and satisfies feder- the road, seems attractive to many owners because it does not
al regulators that the project is not endangered.To sta- involve adjoining land and there are few environmental con-
bilize all the landslides would have rendered the straints. In most cases, however, it is a risky option because the
project uneconomic, and it probably would have been untreated landslide may widen in the future, jeopardizing the
closed down. abutments of the bridge. If selected as a viable option, the
bridge generally will need to have pile-supported abutments
and no intermediate supports within the landslide.As the cost
of a bridge structure increases rapidly with increasing span
14.4 AVOIDANCE length, the cost of a bridge can be uneconomic except for rel-
atively short lengths (typically less than 150 feet).
Avoiding a landslide can be less expensive than remediation. Tunneling through the stable hillside behind the slide. This
Five possible ways of achieving this objective for landslides expensive option is usually a “last resort” choice for roads and
affecting roads are: requires unusual circumstances. The huge Thistle Slide in
Utah (April 1983) required construction of a tunnel to
• Moving the road into the stable hillside
reestablish the Denver & Rio Grande Western Railroad tracks.
• Realigning the road onto the valley floor
• Relocating the facility elsewhere
• Bridging over the landslide
14.5 SELECTIVE STABILIZATION
• Tunneling through the stable hillside behind the slide
Cutting into the stable hillside is often performed on a side- Technique
hill cut-and-fill road where the outer edge of fill becomes The upper reaches of a landslide can be selectively stabilized
unstable.The road alignment is moved into the hillside to put to protect a particular facility while allowing the remainder of
the road section onto cut. Possible disadvantages include: the landslide to be untreated.This option can be chosen when
272 REMEDIAL AND PREVENTATIVE OPTIONS

the cost of stabilizing the entire landslide is prohibitive and many other situations, total loss of support is highly unlikely to
selective stabilization can be done economically. It also occur. It is suggested that the calculated factor of safety be
becomes a viable option when downslope landowners are dropped to 1.1 or 1.15 for this extreme case and the main
unwilling to cooperate in stabilization of the entire slide. design calculation be based on some judgment of the probable
The main benefits of selective stabilization to the owner loss of support over the design life of the remedial structure.
are that the treatment usually involves no purchase of adjoin- Another option is to require field inspections at prescribed
ing property and requires less-extensive remediation than time intervals after construction and to take appropriate
would be needed to provide a sufficient degree of safety to the action to reinstate support if needed. However, future inspec-
entire landslide. It can also avoid environmental issues that tions are not generally recommended because they can
cause delays to remediation. Using selective stabilization, the become overlooked as personnel change jobs and memories
cost of providing remediation to a specific facility is signifi- fade. Unless failure inspections are mandated by a regulatory
cantly lower. agency, reliance upon future inspections can lead to problems
Selective stabilization is often chosen for roads and rail- from neglect.
ways where the top of a landslide is at or near the right-of- A separate stability calculation of the lower unremediated
way. A potential disadvantage of the selective stabilization landslide may show that the ground will remain stable. For
option is that the slope below may continue to move and example, a wall supporting the upper portion of a circular arc
reduce any support it provides.Thus, the design of a retaining slope failure may remove most of the original driving
wall has to assume that some or all of the downslope ground moment of the landslide. The curved slip surface at the base
will move away from the wall, leaving a vertical crack at the of the slide may have sufficient shear resistance to remain sta-
front of the wall. The depth of such a crack depends on the ble indefinitely once the upper part of the slide has been
soil type, its strength, and the depth to the slip surface. One remediated.
possible remedy to this effect is to perform a staged partial
excavation in front of the wall and backfill the excavation Examples of Selective Stabilization
with rounded gravel. If a crack develops in the ground, the 1. A riverbank landslide at Mohler, Oregon, reached the
rounded gravel will fall into the crack and provide continued shoulder of the county road, the headscarp being 5 feet
support on the downslope side of the wall (Figure 14.1). high (Figure 14.2a). Full remediation would have
The recommended design procedure is to assume that required some type of support at the river’s edge over a
some support will be lost on the downslope side of the remedi- width of 180 feet. At the highway however, the affect-
ation.In the worst case,all the support from the landslide soils is ed width was less than 90 feet. In addition, to stabilize
lost, and this can happen for shallow landslides or stiff clays. In the ground below the highway would have required

Figure 14.1 Use of cohesionless Crack develops Road

Tieback anchor
Continued
movement
Slip surface

Rounded gravel Road

Drain to remove
collected water

Multi-level anchors
can be added
Slip surface
 ALTERNATIVES TO FULL REMEDIATION OF A LANDSLIDE 273

Figure 14.2 Mohler Slide,

negotiations with three property owners and difficult 14.6 MARGINAL STABILIZATION
environmental negotiations for construction work in
the river. Instead, the county chose selective stabiliza- Technique
tion of the road. A tied-back soldier pile was built to Marginal stabilization is the implementation of stabilization
stabilize only the county road. measures in the knowledge that stability will be improved but
2. Hagg Lake Perimeter Road Slide 4 (Case History 10). at a level that is lower than the desirable factors of safety dis-
Here, the road is near the head of a long landslide cussed in Chapter 10.
extending down to the lake. Instead of attempting to The goal of marginal stabilization is to slow down or stop
stabilize the entire slide, a keytrench was excavated landslide movements, thereby improving safety to the facility
through about 20 feet of slide debris to bedrock, back- and its users. In most cases, marginal stabilization is a decision
filled with shot rock, and the road was reconstructed to do some improvement over “do nothing.” It is accepted that
on top of the rockfilled keytrench.The ground downs- further ground movements may occur, possibly requiring
lope of the road has continued to move slowly towards additional treatment. If the facility is left open to the public,
the lake, but the road above the keytrench (and ground the risk to personal safety has to be minimal.
further upslope) has remained stable. Projects for marginal stabilization have to be selected with
3. Goat Lick Slide in Glacier National Park, Montana care and with the full consent of the owner (and regulatory
(Case History 9).The road is at the top of a deep-seat- agency, if applicable). The proposed treatment and its limita-
ed landslide extending down to the Flathead River. tions need to be fully explained and documented so that there
The headscarp of the slide crossed both lanes of the can be no possible confusion later that it was the result of cor-
road over a width of more than 200 feet. Stabilization ner cutting or ignorance.Typically, marginal remediation proj-
of the road was achieved by constructing a tied-back ects require continued monitoring to check on the adequacy
shear pile retaining wall near the outer shoulder. and effectiveness of the remedial measures.
274 REMEDIAL AND PREVENTATIVE OPTIONS

Lest the subject of low factors of safety appear onerous, it ed factor of safety improvement was inadequate but
can be pointed out that there are many thousands of slopes achieved the objective. This “temporary” stabilization
with marginal slope stability that provide little risk to the pub- lasted for more than 30 years, with minor creep move-
lic. These include natural slopes of talus, colluvium, and ments, before an extensive program of groundwater
ancient landslide terrain, plus manmade slopes such as roads lowering was implemented to further improve long-
built from spread rockfill and gravel where the outer slopes term stability. Ditch Camp Slide is an example (and
stand at the angle of repose.These slopes are stable or moving first use of) Original Profile Analysis (Chapter 10,
so slowly that they do not constitute a safety threat. Section 10.2).
The author’s experiences with marginal stabilization have 2. Pelton Park Slide, near Madras, Oregon (Case History
been excellent. Where applied carefully and appropriately, 5).This very large landslide is a public recreation park
many worthwhile improvements have been achieved at a rea- adjoining the reservoir behind Pelton Dam in central
sonable cost. Some examples are given below: Oregon. The deep-seated slip plane is horizontal and
the toe is well below the reservoir. It is a double
Examples of Marginal Stabilization wedge configuration, and the upper driving wedge is
1. Ditch Camp Slide near Sandy, Oregon. The water defined by ground cracks. Slow movements were
pipelines serving Portland, Oregon, cross a hillside of occurring, and there was no realistic way for provid-
ancient landslide terrain.The old slide debris is a het- ing a normal design factor of safety at a reasonable
erogeneous mixture of stiff clay and waterbearing cost.
gravels. In 1965, the local county highway department Marginal stabilization was undertaken by excavating
straightened out the road at the base of the slide area a trench at the head of the landslide to reduce the lat-
by excavating a large “hump” of ground. In the follow- eral loading provided by the upper (driving) wedge.
ing winter, the slope above the excavated area moved The calculated improvement in the factor of safety was
downslope by several feet during a prolonged period 2%. Although insufficient by conventional standards, it
of wet weather, requiring emergency measures to pre- was enough to stop the ongoing movements, and only
serve the alignment (on trestles) of the pipelines above minor creep has occurred subsequently. This has
the road. allowed the park to remain open to the public.
Stabilization was achieved by rebuilding the “hump” However, movements were subsequently reactivated
with a free-draining rockfill that replaced the original by an earthquake. Further excavations at the head of
weight and provided improved drainage.The calculat- the slide stopped the movements.