FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics © 2007 ASCE

Why Monitor Performance?

W. Allen Marr, P.E., PhD, ASCE Fellow
President and CEO, Geocomp Corporation
1145 Massachusetts Ave., Boxborough, MA, USA, 01719. [email protected].

ABSTRACT

This paper provides an overview of the many potential benefits of an effective

performance monitoring program for constructed facilities. It also outlines some specific
methodologies to quantify these benefits in ways that non-specialists can understand and use.
These include quantifying the benefits of monitoring and comparing them to the cost of
monitoring. A case is described where the benefits of performance monitoring exceeded the
cost by more than eight times. Another example shows that performance monitoring
decreased the probability of failure by 40-fold. The goal of the paper is to provide others with
the means to better define and quantify the benefits of proposed performance monitoring
programs to other members of the Project Team so that effective monitoring can be
performed throughout the project life.

INTRODUCTION

Many geotechnical engineers have experienced the benefits of using geotechnical

instrumentation to monitor structural performance somewhere in his or her career.
However, many of us struggle to justify the expense of geotechnical instrumentation
to our clients. The purpose of this paper is to provide a resource to help define and
quantify the benefits of a geotechnical instrumentation program for a project.

Construction of transportation facilities must deal with many unknowns and

limited data. This is especially true for those projects in urban areas that involved
construction on or in soil and rock. We are working in materials with properties that
can change instantly and significantly from one point to the next. These changes may
result from the actions of nature in laying down the earth, from prior activities of man
on the site, or from actions of the contractor as he works with the site. Further
complications may come from uncertainties in the loads that the new facility must
withstand during construction and operation. These various uncertainties combine to
produce substantial uncertainty in how the completed facility will perform throughout
its life.

Compounding the importance of these uncertain conditions are the potential large
consequences of unexpected performance by the facility. Unexpected performance
may adversely impact the project, neighboring structures and utilities, and people.

1
FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics © 2007 ASCE

Unexpected performance may delay the project, increase its cost, and lead to lengthy
and expensive litigations.

Urban work amplifies these issues because there are more structures within the
potential influence zone, urban structures tend to be more significant, there are more
people to be impacted, the population tends to be less tolerant, and more unknowns
exist due to previous activities at the site. Additionally, one may be working in and
around existing structures that must stay in operation or joining new construction to
existing facilities and completed sections of the work.

Monitoring provides us with quantitative information on actual performance. We

compare the measured performance with the predicted or expected performance.
Differences indicate the effects of uncertainties in our design. We need to evaluate
those differences to determine what they indicate for future performance. If the
anticipated future performance is unacceptable, we look for changes, modifications,
and remediation that can be made to alter the future performance.

During design we have data that represent an indication of the true state of nature.
We use our knowledge and judgment to combine these data with models to predict
ultimate performance. If the predicted ultimate performance is unacceptable the
Engineer alters the design. Traditional design treated predictions as discrete values
but in fact every prediction has uncertainty. Measured performance is nature’s
indication of the true condition. Measured performance reduces the range of
uncertainty caused by all the unknowns present during design.

Traditional approaches attributed unexpected performance to an act of god; this

defense has become increasingly useless as lawyers and experts seek relief for those
who are allegedly damaged. More recently the blame has shifted to acts of the
contractor or acts of the design professional. By measuring performance and taking
action, the goal is to reduce unexpected performance and take the blame game out of
the project equation.

BENEFITS OF PERFORMANCE MONITORING

Geotechnical instrumentation programs are used to save lives, save money and/or
reduce risks by giving advanced notice of unexpected, undesirable performance. In
concept, these are simple and easy to understand benefits. In practice, they may be
benefits that are difficult to quantify or substantiate.

Table No.1 summarizes the principle technical reasons one might recommend a
geotechnical instrumentation program for a project. Dunnicliff (1988, 1993)
discusses some of these points. Each of these is discussed below in the context of
today’s practice of geotechnical engineering.

In this paper, “geotechnical instrumentation program” is used to describe the

complete effort required to obtain an effective instrumentation program. This
complete effort includes planning the program, specifying the instruments, procuring

2
FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics © 2007 ASCE

the hardware, collecting data, interpreting results, preparing reports and acting on the
conclusions.

Indicate Impending Failure

Geotechnical facilities can fail with catastrophic consequences to life and property.
Such failures may be the result of excessive loads, design errors, construction
deficiencies, unknown or different conditions, deterioration, operational errors or
intentional action. Geotechnical instrumentation has been widely used to detect the
onset of failure in dams, slopes, embankments and excavations. Such monitoring
may have different purposes. It may be to issue a warning to evacuate people and
move equipment. It may be to initiate action to forestall the failure. It may provide
feedback when causing an
intentional failure, such as for a Table No. 1: Reasons to Monitor Performance
mining operation or a field test. 1. Indicate impending failure.
2. Provide a warning.
Geotechnical 3. Reveal unknowns.
instrumentation programs may 4. Evaluate critical design assumptions.
save lives by giving advanced 5. Assess contractor’s means and methods.
6. Minimize damage to adjacent structures.
warning in time for people to 7. Control Construction.
get to a safe area. A good 8. Control Operations.
instrumentation program may 9. Devise remedial measures to fix problems.
reveal an unknown condition 10. Improve performance.
early enough that changes can 11. Advance state-of-knowledge.
12. Document performance for assessing damages.
be made that greatly reduce the 13. Inform stakeholders.
risk of failure. Instrumentation 14. Satisfy regulators.
can save money and reduce risk 15. Reduce litigation.
by decreasing the likelihood of 16. Show that everything is OK.
an unexpected failure that
destroys or delays the project.

Provide a Warning

Geotechnical instrumentation systems can warn that some indicator of

performance is exceeding acceptable limits. These instruments may be made a part
of an automated system that automatically initiates the warning. A tiltmeter can warn
of a sudden movement across an existing shear zone. A piezometer can warn of
excessive pore pressures in the downstream toe area of a dam that might become
unstable and threaten the stability of the dam. Flow meters might warn of significant
changes in the volume of flow.

In these cases, the geotechnical instruments are a vital part of a warning system
that is used to get people out of harm’s way or initiate preemptive actions to avoid an
undesirable event. The instrumentation saves money by reducing the risk of a loss of
life and/or property, or by reducing delays.

3
FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics © 2007 ASCE

Reveal Unknowns

Geotechnical engineers constantly work with unknowns. Sometimes these

unknowns can cause a catastrophic failure that destroys the entire project, takes lives,
or ruins careers. Other times they cause delays, which increasingly lead to expensive
claims for Differing Site Conditions.

The foundations of the geotechnical discipline were built on the use of field
measurements to reveal unknowns during construction and head off disaster. The
work of Terzaghi and Peck in Chicago to measure the forces on excavation support
systems is a classic example. In fact, one might argue that the driving force that leads
to the development of most of the instrumentation we use today was a need to
measure something to reveal unknowns.

Generally speaking, geotechnical engineers cannot control the materials in which

they work. Nature created those materials in random processes that produced non-
uniform and highly variable conditions. A seam of weak material, a zone of high
compressibility, or a pocket of high pore water pressure may go undetected in the
exploration work and not be considered in the design. Yet, these hard-to-detect
details may become the primary cause of undesirable performance.

There will always be uncertainty in geotechnical work. As a result, geotechnical

engineers cannot accurately predict the performance for their designs. Society cannot
afford very conservative designs to minimize the potential effects of these
uncertainties; nor will society accept the risks from large uncertainties.

Where the consequence of these unknowns might threaten the success of a project,
we instrument to measure the actual performance of our design. We use the
measurements to identify potential undesirable outcomes, including failure, and make
plans to take preemptive action early. The measurements help us answer questions
and reduce uncertainty. Terzaghi was a strong advocate of this approach. Peck
(1969) defined and illustrated its use as the Observational Method, a concept used in
underground construction world-wide.

Data from geotechnical instrumentation are valuable in assessing Differing Site

Condition claims that result from unknown conditions. Data from a good
geotechnical instrumentation program may help prove the existence or absence of a
DSC and lead to an equitable arrangement between the Contractor and the Owner.
Without such data, proponents are left to argue opinions without supporting facts
leading on considerable expense and unpredictable outcomes for both sides of a DSC
dispute.

4
FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics © 2007 ASCE

In my own experience the lowest overall cost to a project from unknown

conditions is to use procedures which reveal those unknown conditions as early as
possible and engage remedial work as soon as possible. A good geotechnical
instrumentation program is vital to this approach. The alternative of delay, denial,
and blame almost always costs more. Added costs come from the expenses incurred
to determine who pays the added cost.

Evaluate Critical Design Assumptions

Usually we cannot justify the expense of investigations and studies required to

remove all uncertainty about the conditions and parameters that affect geotechnical
design. We make simplifying assumptions about ground conditions and choose
conservative parameters to prepare a design. If these assumptions could be wrong
and the consequences would be unacceptable, we may use geotechnical
instrumentation to gather data with which to evaluate our critical assumptions. For
this approach to work effectively, we need a design that can be altered if the
instrumentation shows our assumptions to be wrong.

We might for example assume that a sand layer at the middle of a clay deposit will
provide drainage to hasten consolidation of the clay under the weight of a new
embankment. If our assumption is wrong, the project could be delayed by years. A
single piezometer placed in the sand layer beneath the fill would tell us how good our
assumption was early enough to take corrective action and minimize adverse
consequences.

Instrumentation saves money by permitting the designer to choose cost effective

solutions with reasonable design assumptions and avoid expensive conservatism.
Data from the instrumentation are used to prove that actual behavior is within the
limits permissible for the design, or that actual behavior is different than anticipated
and further consideration is warranted.

Assess Contractor’s Means and Methods

The outcome of some geotechnical projects depends on the means and methods of
the contractor. The job requirements may be in the form of a performance
specification where the contractor is required to provide the design and complete the
work. Maintaining the stability of the bottom of a deep excavation against uplift is
one example. The specifications might require that the contractor maintain a
minimum factor of safety against bottom heave due to uplift of at least 1.1.
Piezometers installed to measure pore water pressures beneath the excavation would
indicate whether the contractor is meeting this important requirement.

5
FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics © 2007 ASCE

Geotechnical instrumentation is used to determine whether the contractor’s means

and methods meet the specified performance requirements. A good instrumentation
program can provide sufficient data of the right type to show the potential for
undesirable performance early in the work. Data from the instrumentation may show
why the contractor’s means and methods are not working. The means and methods
can then be adjusted to reduce their impact on the project.

Instrumentation saves money by helping to reduce the consequences of

undesirable performance. Data from the instrumentation may also help identify
ineffective or inefficient aspects of the contractor’s means or methods.

Minimize Damage to Adjacent Structures

Underground construction can have adverse consequences that reach beyond the
project boundary. These consequences may affect adjacent property with undesirable
results. Expensive repairs, bad relations and protracted litigation can result.

Movement of the ground outside a supported excavation is one example. The

specifications might require the contractor to provide an excavation support system
that limits horizontal and vertical movements outside the excavation to less than 1
inch so that adjacent structures are not damaged by the work. Geotechnical
instrumentation to measure vertical and horizontal movement outside the support
system is used to determine whether the contractor meets this requirement.

Instrumentation saves money by providing data on performance of adjacent

facilities early enough that damage to those facilities can be avoided or minimized by
changing the construction operations. In doing so, we save the costs to fix the actual
damages. In addition, we may avoid or greatly reduce the costs that come from
inflated claims and protracted litigation resulting from the damages. Such savings
can be of great significance, especially in urban areas.

Control Construction

Instrumentation may be used to monitor the progress of geotechnical performance

to control a construction activity. For example, an embankment might be placed over
a soft soil stratum by constructing it in stages. Placed all at once, the embankment
would cause a foundation failure. Placing the embankment in stages with time
between each stage allows the soft soil to strengthen by consolidation between each
stage. Instruments to measure movements and pore water pressures could be used to
determine when enough consolidation of the clay has occurred that the next stage of
fill can be safely added. A delicate balance may be sought between adding the next
stage as quickly as possible to minimize construction time but not so quickly that a
stability failure is created. Other examples include using instrumentation to
determine how deep to drive piles to attain a required capacity, controlling the
excavation and supporting sequence for a deep excavation, controlling the advance
rate for soft ground tunneling, and controlling the sequence for compaction grouting.

6
FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics © 2007 ASCE

Instrumentation saves money by helping us determine the fastest and most

expeditious way to proceed with construction without creating undesirable
performance. Having data from instrumentation may permit more economical design
approaches, such as staged construction instead of other means of ground
improvement.

Control Operations

Geotechnical instrumentation may be used to help control the operation of a

facility. The rate of drawdown of a reservoir for a pump-storage power facility might
be tied to readings of pore pressure in the embankment dam or side slopes to avoid
stability failures due to draw down. Readings from piezometers might be used to
control the amount of ore that can be safely stockpiled over a soft foundation.

In these situations, data from the instrumentation permit the operations of the
facility to be pushed closer to their limits without causing a failure. As a result, the
owner realizes an economic gain from the higher utilization or more efficient
operation of the facility.

Devise Remedial Methods to Fix Problems

Things sometimes go wrong in geotechnical construction that must be fixed.

Finding the best fix requires understanding what went wrong. Data from
geotechnical instrumentation can help one figure out what caused the problem. Then
one can devise a remedial action that addresses the specific cause rather than masks
the symptoms.

Instrumentation saves money by helping us tailor the remedy to the specific

cause of the problem. Otherwise we may face repeated efforts of trial and error
actions until something finally works.

Improve Performance

Modern concepts of business management stress continual improvement and the

need for measurements to gage success. A common saw in business practice is “that
which is measured improves, while things not measured eventually fail.” The mere
process of measuring performance coupled with normal human behavior leads to
improved performance.

The underground construction industry is searching for ways to improve their

operations to produce facilities that perform better and cost less. Like other business
processes, improvement can only be assessed by measurement. Geotechnical
instrumentation systems can play a central role in providing these measurements.
This is especially the case for projects that use performance-based specifications.
Future contracts may reward contractors and engineers for good performance and
penalized them for poor performance of the completed facility. A good
instrumentation system will be a central part of determining the quality of the work.

7
FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics © 2007 ASCE

Advance State-of-Knowledge

Many of the advances in the theories of geotechnical engineering have their roots
in data from geotechnical instrumentation on full-scale projects. The data give us
insight into how things are performing and causal relationships. Historically, a
significant amount of geotechnical instrumentation was performed as part of a
research effort to improve our state of knowledge. Much of this was paid for by
governmental agencies with a mission to improve practice.

Instrumentation to improve the state of knowledge saves money by leading to

improvements in our design and construction methods. On some projects,
instrumentation of the early phases of the job may lead to an improved understanding
of site conditions and geotechnical performance such that the design and/or
construction methods can be altered to reduce costs and risks on later phases of the
project. Manufacturers of specialty materials may instrument projects to demonstrate
the performance advantages of their products for future projects or to find ways to
improve their product for future applications.

Document Performance for Assessing Damages

Claims for damages by third parties represent one of the substantial risks
encountered in geotechnical projects. Some claims may include charges for damages
unrelated to the construction. Others may be inflated, such as a claim for structural
damage when only minor architectural damage has occurred.

Data from geotechnical instrumentation can help establish the validity of such
claims. For example, if the instrumentation shows that an adjacent building has not
moved during construction, it becomes more difficult for the owner to claim that
cracks in the building resulted from the construction activity.

Instrumentation saves money by helping to identify bogus or inflated claims. It

may also indicate the potential severity of any damages so that a fair settlement can
be established. The mere presence of data from geotechnical instrumentation may
help discourage the filing of frivolous claims. Some insurance companies have
started to use the data from geotechnical instrumentation programs to help them
determine whether to settle a claim and for how much. As we undertake more
demanding projects in developed areas and litigation grows more sophisticated, I
anticipate geotechnical instrumentation to experience more widespread use to help
limit and settle damage claims.

8
FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics © 2007 ASCE

Inform Stakeholders

Construction in developed areas may affect numerous parties, all of who seek a
role in controlling the adverse impacts of the project. People tend to anticipate the
worst outcomes and fear construction impacts. Data from geotechnical
instrumentation can provide solid evidence of the true construction impacts. It can
provide powerful responses to the questions and fears of stakeholders.

A good example of this is people’s sensitivity to construction-induced vibrations.

People inside buildings may become concerned with the level of vibrations caused by
nearby pile driving. Humans typically sense the presence of vibrations at a level less
than 10% of those that begin to cause minor architectural damage to the building.
Building owners may become concerned for the safety of their building when they
sense these relatively low level vibrations. Data from a good geotechnical
instrumentation program can demonstrate to these people that the vibration levels are
well below those that might cause damage. Alternatively, the measurements may
show vibrations that approach unacceptable levels so construction methods can be
changed before damage occurs.

Instrumentation saves money by keeping stakeholders informed of the actual

situation. This reduces the potential for bad relations, costly disputes and work
stoppages.

Satisfy Regulators

Some facilities must be instrumented to meet the requirements of specific

regulations. For example, some states require piezometers be installed in all earthen
dams over a specified height. Some cities require seismographs be installed in tall
buildings to record earthquake response. In these cases the governmental agencies
have determined that a public good is served by requiring an instrumentation
program. The instrumentation may be required to help protect public safety, or it
may be required to provide data with which to improve the state of knowledge about
a particular problem.

It is not always easy to see how instrumentation saves money when installed to
meet a regulatory requirement. For the specific project, it may not save money,
especially if the only reason the equipment was installed was to satisfy the regulatory
requirements. Unfortunately, many of those involved see such instrumentation only
as an added cost. With the instrumentation properly installed and the data carefully
collected and evaluated, it can become a valuable resource in maintaining and
rehabilitating the facility at some later time.

Reduce Litigation

Data from geotechnical instrumentation can be a powerful deterrent to litigation.

Contractors may claim differing site conditions. Abutters may claim for damages
caused by construction. Owners may claim poor performance of the completed

9
FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics © 2007 ASCE

facility. Where subsurface conditions are involved, data from a good geotechnical
instrumentation program may provide powerful evidence to help get to a fair
resolution of such claims. I have been involved in a number of cases where the entire
basis for a differing site condition claim could have been refuted if only a few key
measurements had been taken during construction. A common claim is that
excessive water was encountered due to a differing site condition. Unfortunately,
measurements of actual quantities of flow, or the flow conditions in the vicinity of the
contested region are rarely taken. A few key measurements could quickly establish
the validity of the contractor’s claim.

Instrumentation has the potential to save considerable money in reducing the

frequency of litigation and the size of the claims. Good geotechnical instrumentation
programs may reduce unexpected performance and thereby avoid the cause of the
dispute. The instrumentation may reveal the presence of a differing site condition
and permit the construction operations to be altered to minimize the impact of the
change and result in a smaller claim. Data from the instrumentation may help
establish the actual impacts of differing site conditions or adverse performance so that
an equitable adjustment can be made fairly and quickly.

I’m waiting for the corollary to this idea to develop as well, that is the case where
someone sues a contractor or engineer for not using geotechnical instrumentation to
warn of undesirable performance. Doctors have been sued for not requesting certain
diagnostic tests or collecting certain data when the methods for making such
measurements were readily available. Considering the parallels between medical and
engineering practice, how long can it be before a similar concept is applied to sue an
engineer for failing to require instrumentation to identify and avoid undesirable
performance?

Show That Everything is OK.

Increasingly we use instrumentation programs to demonstrate the actual

performance is within the bounds anticipated by the designers. The presumption is
there will be no surprises or unexpected consequences to cost and schedule, and that
unexpected behavior can be identified early enough to maintain control of the project
cost and schedule.

In this use, data from an instrumentation programs helps maintain the various
parties’ confidence in the performance of the work and frees them to focus on other
issues. I find more clients desiring performance monitoring systems that are
comprehensive and robust but with instant reporting as simple as a green light to
indicate that everything is in an acceptable state.

QUANTIFYING BENEFITS OF GEOTECHNICAL INSTRUMENTATION

To this point, the paper has discussed the engineering reasons for why we monitor
performance. These are the talking points in engineer-to-engineer exchanges.
They’re the language we put into proposals and demonstrate in publications. But

10
FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics © 2007 ASCE

they’re not the language of business. They’re not the language of most owners or
contractors, or designers, or regulators. They can read and understand the rationale
outlined above but they have trouble translating the words into perceived benefits that
give them value.

This section provides an approximate method to quantity these benefits. While the
suggested method is not precise, it is generally sufficient to decide how much of an
instrumentation program is worthwhile for many situations. The suggested method is
based on concepts of decision theory and risk analysis.

Decision theory provides a framework for managers to make decisions when faced
with incomplete and uncertain information. It uses probabilistic analyses to estimate
likely outcomes. Decisions are based on the desirable outcomes with the highest
likelihood of success or lowest chance of failure. Most graduate level business
programs teach decision theory as a recognized decision making tool.

Risk analysis embodies a wide range of scientific theory and engineering analyses
to identify potential sources of risk, determine the probability of each source, and
estimate the consequences from each source of risk. Total risk is the summation of
the probability of each source of risk occurring times the consequences of that
occurrence. Risk can be decreased by actions that reduce the probability of a source
of risk occurring or reduce the consequence of that event occurring. As an example,
consider two dams of similar construction in a similar setting. Both dams might have
the same probability of failure. But suppose Dam A is located 10 miles upstream of a
major city sited on the banks of the river and Dam B outlets directly to the ocean 10
miles away. Clearly Dam A poses a much higher risk than Dam B even though they
have a similar probability of failure. Dam B could have an even higher probability
of failure than Dam A; yet pose less overall risk. However, risk is in the wallet of its
recipient. While failure of Dam B might present much less societal risk, its risk of
failure might still be unbearable to its shareholders who would suffer from the
physical loss of the facility.

Risk analysis provides input for decisions using decision theory. A manager may
choose a course that minimizes risk, or the manager may choose a course in which
the benefits achieved by lowering risk outweigh the costs of achieving that reduction.
In its simplest form, the approach outlined here to help quantify the benefits of
instrumentation is as follows:

1. Determine all ways by which the project can fail or experience undesirable
performance from a geotechnical cause.
2. Estimate the probability of occurrence of each of these events during the
period of interest, Pi.
3. Define the consequences of each event and estimate the potential cost of each
consequence, Ci.
4. Estimate the reduction in probability of occurrence of each event that a
geotechnical instrumentation program could produce, Pi.

11
FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics © 2007 ASCE

5. Determine the expected reduction in risk produced by instrumentation by

computing the sum of the reduction on probability of occurrence of each event
times the cost the consequence of that event, RR = Σ(Pi * Ci).
6. Use instrumentation as long as the cost is less than the estimated reduction in
risk, RR.

As described above, geotechnical instrumentation can be used to help reduce risks,

minimize damages and avoid delays. Each of these elements can be assigned a cost.
Consequences may include added construction costs, damages to adjacent facilities,
delays, litigation, etc. While formal methods exist to quantify risk, usually they are to
complex to apply in decision making about geotechnical instrumentation. One
approach is to use approximate subjective estimates of risk. In this approach one
seeks to identify all significant undesirable outcomes and estimate the likelihood of
their occurrence.

It is helpful to simplify the likelihood of occurrence to a few possible states that

are defined sufficiently to give useful results but simply enough to avoid unnecessary
complication. Table No.2 gives an example of one set of risk states that is sufficient
for most evaluations of geotechnical instrumentation. The values in Table No.2
represent the author’s own interpretation of what numerical probabilities to assign to
subjective adjectives in the context of developing geotechnical instrumentation
programs. These values are intended only to provide a tool to aid in making
approximate assessments of risk. Decisions based on these values should take into
account their approximate nature.

Table No.2: Risk Classification Scheme

LIKELIHOOD PROBABILITY OF RISK

OCCURRENCE PROBABILITY
Zero, none, impossible <0.0001 0.01%
Virtually impossible, very unlikely 0.00011 to 0.001 0.1%
Unlikely, improbable, barely possible 0.0011 to 0.01 1%
Small, limited, marginal 0.011 to 0.1 10%
Moderate, considerable, somewhat 0.11 to 0.5 50%
unlikely
Likely, probable 0.51 to 0.9 90%
Highly likely, very probable >0.9 100%

Engineers seem to be able to use adjectives to describe their judgment about how
much uncertainty exists in their design. Table No.2 assigns probability ranges to
some of the more common adjectives used to describe uncertainty or risk. To
simplify risk calculations, the ranges given in Table No.2 for probability of

12
FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics © 2007 ASCE

Fig. 1: Failure Modes for Deep Excavation

occurrence are rounded to the highest value associated with each group of descriptive
adjectives. These simplified probability states are sufficient to produce reasonable
estimates of risk cost for most geotechnical instrumentation applications. Of course,
this simplified procedure can always be replaced with a more rigorous risk
assessment if the project conditions justify the extra expense.

It is easiest to illustrate how to proceed with an example. Figure 1 shows a

proposed 14 m (45-ft) deep excavation in the center of a city. Several historic
buildings 4 to 8 floors high are founded on shallow foundations just outside the
excavation. Excavation support will be a tied-back wall. Table No.3 lists the
significant potential adverse consequences associated with making the excavation. If
the soils are too weak, we may cause a global stability failure that could collapse one
or more of the buildings, severely damage the work and delay the project for 2 years.
Shear deformations caused by making the excavation could cause cracking and
possible structural damage to the adjacent buildings. Dewatering of the excavation
might produce dewatering outside of the excavation and result in consolidation
settlements of buildings, streets and utilities. Loss of ground due to water flowing
and carrying soil into the excavation might cause sinkholes that undermine and
damage significant utilities buried in the street just beyond the excavation. Other
consequences are possible, but the design engineers consider these the ones of most
importance and consequence.

13
FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics © 2007 ASCE

Table 3: Potential Adverse Consequences from Deep Excavation

Undesirable outcome Likelihood Consequence

Global stability failure 0.03 $50,000,000 to fix and 2 year
delay, possible loss of life
Excessive Deformation from shear Likely $400,000 to fix
Excessive Deformation from Unlikely $200,000 to fix
dewatering
Ground loss from flow into Small $500,000 to repair
excavation

The middle column of Table No.3 gives the design engineer’s subjective estimate
of the likelihood of the undesirable outcome occurring during the life of the project.
A factor of safety for global stability of 1.3 was computed using strengths estimated
from SPT data. These conditions indicate a Lambe-Silva-Marr Engineering Level III
condition (Silva, et. al., 2007). The corresponding probability of failure is 0.03.

To complete the evaluation, we need to assign monetary values to all

consequences. As an example, additional study shows a small probability of a global
stability failure that could involve a catastrophic collapse of the adjoining building
with consequential estimated costs of $50,000,000 from total loss of the building,
damage to the project and delays. Combining this information with the information
in Tables No. 2 and No. 3 leads to Table No. 4.

Table No.4: Potential Risk Costs for Deep Excavation

Outcome Consequence Probability Risk Cost

Global stability failure $50,000,000 to fix and 2 year 0.03 $1,500,000
delay, possible loss of life
Excessive Deformation $400,000 to fix 0.9 $360,000
from shear
Excessive Deformation $200,000 to fix 0.01 $20,000
from dewatering
Ground loss from flow $500,000 to repair 0.1 $50,000
into excavation

We can use these results to guide our selection of a geotechnical instrumentation

program. It is clear that the biggest exposure is with global stability. With additional
work, we determine that a geotechnical instrumentation program could provide
sufficient warning of an impending global stability failure that we could institute
preventative measures, i.e. fill the excavation with soil or water. We believe this
approach would reduce the likelihood of a global stability failure from 0.03 to 0.005.
This translates to a reduction in Risk Cost from $1,500,000 to $250,000 or reduction
of $1,250,000. From a straight decision making perspective, we can argue that we
are justified in spending up to $1,250,000 on a geotechnical instrumentation program

14
FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics © 2007 ASCE

that removes the likelihood of a global stability failure occurring. The same system
might help reduce some of the risk cost from excessive deformation due to shear.
The risk costs estimated for the other two outcomes are relatively small so it would
be difficult to justify much more than visual monitoring and precautionary
construction steps to control these two failure modes.

This example shows one simplified approach to evaluating how much to spend on
a geotechnical instrumentation program. Used consistently over a number of
projects, it provides a consistent way to estimate the monetary value of geotechnical
instrumentation programs. However, it is not the final answer to any particular
project. There may be factors that cause significant undesirable consequences that
cannot be easily monetized. Loss of life, political fallout from delays, loss of
reputation and bad press are examples that come to mind. Any of these may provide
sufficient cause to justify a more extensive geotechnical instrumentation program.

It is important to recognize that this approach only provides an organized way to

help make rational decisions based on quantified information that contains
uncertainty. It does not ensure outcomes. Geotechnical instrumentation by itself
does not change the outcome. Placing geotechnical instrumentation in a deep cut to
monitor stability does not alter the factor of safety of the cut. It is only through the
intelligent use of the data from the geotechnical instrumentation that engineers can
better foresee potential outcomes and take appropriate actions to alter the events, or
reduce the consequences. Engineers change outcomes and geotechnical
instrumentation is one of the tools they use to make effective changes.

The approach described above can be made much more complete using principles
and tools of probabilistic analysis and decision theory. Complete event trees and
fault trees can describe each likely failure scenario in considerable detail. Rigorous
activities can be used to establish probabilities. An excellent example of an event
tree for the failure of an earthen dam by cracking was provided by R.V. Whitman
(Whitman, 1984) in his Terzaghi lecture.

15
FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics © 2007 ASCE

Professor Whitman laid out an event tree for failure of an earthen dam due to
cracking. Figure 2 illustrates his event tree, which included the effects of
instrumentation on the probability of failure. He assigned probabilities to various
events based on his engineering judgment. His event tree was for the hypothetical
case of “good, conservative design practice” and it included monitoring to detect
oncoming failure. It shows an expected lifetime probability of failure of the dam by
cracking of 5*10-5.

Figure 2: Whitman's Event Tree for Good, Conservative Design Practice.

This event tree includes two branches where monitoring plays a role: “piping
detected” and “high pore pressure detected.” In his assessment, good conservative
design practice would have a 99% chance of detecting piping and a 100% chance of
detecting high pore pressure before they caused failure. These high probabilities
imply a comprehensive and vigilant monitoring program. They also provide a way
for us to assess what happens to the probability of failure for the dam with a relatively
ineffective monitoring program.

16
FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics © 2007 ASCE

Consider the situation where the design follows good, conservative practice but
there is little to no monitoring of construction and operation or the information from
the monitoring effort is ignored. This might reduce the chances of detecting piping

Figure 3: Whitman’s Event Tree with Ineffective Monitoring

and detecting high pore pressure to less than 10%. Figure 3 shows the event tree with
these revised probabilities. The lifetime probability of failure of the dam increases
from 5*10-5 to 0.002, a 40-fold increase. Looked at in reverse, an effective
monitoring program is reducing the lifetime probability of failure of a good,
conservative design by a factor of 40.

This quantified benefit from performance monitoring has immediate meaning to

most owners. Provided the consequences of failure are significant and understood, a
40 fold reduction in risk is a major improvement. In my experience, effective
monitoring programs can reduce the risk of undesirable performance by at least one
order of magnitude in most situations.

CASE HISTORY ILLUSTRATING BENEFITS OF MONITORING

It would be extremely helpful to the instrumentation community to have

documented cases that demonstrate the benefits of performance monitoring where
those benefits are explicitly defined and quantified. Unfortunately, the benefits can
be difficult to quantify and people are reluctant to do the extra required work. I

17
FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics © 2007 ASCE

sought to do this on the large Central Artery/Tunnel Project (CA/T Project) in

Boston.

The Central Artery/Tunnel Project is the largest, most complex highway project in
American history. The $15 billion project consists of 161 lane miles of urban
highway – about half underground in a 7.5-mile corridor – through the heart of one of
America's oldest and most historic cities. The main portion of the Project replaced
Boston's aging, elevated Central Artery (I-93) with a modern eight-to-ten lane
underground expressway. The Project also extended the Massachusetts Turnpike (I-
90) to Logan Airport, built a dramatic 10-lane cable-stayed bridge across the Charles
River, created a state-of-the-art incident response and traffic management system,
constructed five major highway interchanges and has the largest vehicular tunnel
ventilation system in the world. The work for the CA/T Project was unprecedented,
with the majority of the construction taking place in the middle of downtown Boston
which houses the main financial and historical areas that must remain open and
accessible for businesses, residents and tourists.

More than 150 significant structures exist within the potential construction impact
zone of the CA/T Project. These include several high-rise buildings located on mat
foundations within tens of feet of proposed deep excavations, many multi-story brick
and masonry structures of historic value, major transportation infrastructure, and
completed portions of the project.

Previous construction experience in Boston had shown that damage to existing

structures from new construction could create significant added costs to a project.
For example, in the 1980’s, the Trinity Church sued the John Hancock Insurance
Company for 40 million dollars in damages. They successfully argued that cracking
of masonry elements in the historic church was caused by construction of a nearby
excavation for the foundation of the John Hancock Building and were awarded $11.6
million (Lambe, 2003) Also, other prior construction experiences in the Boston area
had shown that the effects of dewatering for an excavation could extend thousands of
feet from the construction (Lambe and Whitman, 1969). The potential for damage to
structures adjacent to the CA/T Project was difficult to assess, but an amount of
several hundred million dollars to more than a billion dollars in claims and litigation
expenses seemed to be reasonable estimates of the possible damages. There was also
significant financial risk from the potential for delay claims by General Construction
Contractors if litigation related to damage claims halted construction at any point.

Potential damages to adjacent structures associated with the CA/T Project

construction were anticipated to be very significant if careful steps were not taken to
prevent damage. To my knowledge, no attempt was made during the planning phases
of the Project to estimate potential damages (note that it is not typical practice to
quantify such estimates). The design engineers had the following main concerns for
major problems that could have occurred:

• Collapse of one or more major structures adjacent to the excavations due to a

failure of the excavation support system. Examples of major structures

18
FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics © 2007 ASCE

adjacent to the Project included the 33 floor main tower of the Federal
Reserve Bank Building, the 46 floor high rise building at One Financial
Center, the historic MBTA South Station and the AMTRAK South Station
rail yard system.
• Major damage to structures due to a failure of the excavation support system
resulting in business disruptions lasting for a significant period of time.
• Architectural damage to the many historic structures along the work
alignment due to excessive deformations from excavation or ground water
lowering.
• Claims from contractors for construction delays related to damage to adjacent
structures, which resulted in work stoppage and delays.

A total collapse of a major building was unlikely to occur, but with the type of
construction performed on the CA/T Project, the possibility of a total collapse had to
be considered. The consequences of such a disaster are hard to quantify but based on
costs of other major disasters, a value of several billion dollars in damage and
economic losses is a reasonable estimate. I used $5 billion to represent the average
loss that might be anticipated if one of the several major facilities close to the project
totally collapsed. This number does not include the potential for loss of hundreds of
lives and many injuries. For the CA/T Project it seemed very unlikely that such an
event could occur before significant distress was noticed and work was stopped.
However, collapses of excavations have been known to occur with little to no
warning. Given the number of major structures that could be affected and the
extended length of time of the Project, the likelihood of occurrence of a total collapse
was estimated to be about 0.03 without a monitoring program in place. This value
was determined by using an approximation of the number of significant structures
within the construction impact zone whose complete collapse would cause billions of
dollars in damage (10) times the average number of years each structure would be
exposed to subsurface work (3) times the annual probability of a major structure
collapsing due to underground work occurring without monitoring. I estimated this
last value to be approximately 0.001 based on his knowledge of the likelihood of a
major collapse of a significant structure exposed to excavation close to its foundation
without a monitoring program. Based on risk theory, one measure of risk is
consequence multiplied by likelihood. Therefore, the risk cost to the Project of a
collapse of a major building could be estimated as the value of $5 billion multiplied
by 0.03 or about $150,000,000, not including the potential for substantial loss of life
and injury.

Major damage and disruption to one or more structures would have less
consequence than a complete collapse but a higher likelihood of occurrence. I
estimated that the potential consequences of major damage to a structure like the
MBTA South Station would be in the $700 million range accompanied with some
loss of life and injuries. This event is estimated to have a likelihood of one
occurrence during the life of the project of about 0.3 without a monitoring program in
place. The probability of occurrence of major damage to one significant structure
was estimated with the same approach used to obtain the probability of a total

19
FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics © 2007 ASCE

collapse. I estimated the annual probability of one structure experiencing major

damage when exposed to excavation close to its foundation without a monitoring
program in place to be about 0.01. Therefore, the risk from major damage and
disruption might be on the order of $200,000,000.

As previously discussed, the original project designers were very concerned about
the potential consequences from architectural and/or structural damage caused by
construction. It was very likely that structures along the alignment could be damaged
by the construction if careful controls were not applied. I estimated that the potential
consequences of architectural and minor structural damage could be in the range of
$100,000,000 with a likelihood of almost one without a monitoring program in place.
This gives an estimated risk cost of $100,000,000 from architectural and/or structural
damage to adjacent structures.

Significant damage occurring to multiple structures along the alignment would

likely result in a major work stoppage until the business community and public could
be convinced that corrective actions were in place to prevent future damage.
Unexpected performance within a construction zone could delay a construction
contractor and cause delays to adjacent contracts. These events could lead to
significant claims from the General Contractors resulting in delays and requests for
extensions producing a domino effect of additional claims in other parts of the
Project. I estimated the cost of potential construction contractor delay claims from
this cause at $500 million and the likelihood as 1% per contract over the project life.
There were 18 construction contracts with heavy civil work. Summing over these 18
projects, there is an estimated 20% chance of at least one major work stoppage due to
damage if no monitoring occurred. This gives an estimated risk cost of $100 million
from construction contractor claims for delays that might result from damage to
adjacent structures. Table No. 5 summarizes these estimates.

The results of these very approximate estimates give a total risk associated with
damages to adjacent structures from construction related ground movements of about
$550 million, not including the potential for significant loss of life and injury. There
is some evidence to suggest that a good performance monitoring program can
decrease geotechnical risk by about one order of magnitude as documented by

20
FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics © 2007 ASCE

Table 5: Summary of Risk Cost Analysis for CA/T

Outcome Consequence Probability of Risk Cost

Occurrence
without
Monitoring
Collapse $5B .03 ~$150M

Major damage and $700 M 0.3 ~$200M

disruption

Construction delays from $500 M 0.2 ~$100M

unexpected performance
Architectural to structural $100M ~1 ~$100M
damage
Total: $550,000,000

Lambe, Silva-Tulla & Marr (1981). Therefore, a strong performance monitoring

program could reduce this estimated risk of $550 million to $55 million, for a
reduction in risk cost over the project life of approximately $500 million. A more
detailed risk assessment might produce a smaller or a larger estimate but I do not
expect the difference to be more than a factor of 2 above or below this estimate (e.g.
ranging from $300 million to $1 billion).

This simple risk analysis indicates a rough estimate of risk to the Project resulting
from damage to adjacent structures and delays from unexpected ground performance,
of about $550 million dollars, not including the potential for significant loss of life
and injury, but $500 million of this could be removed with a good geotechnical
performance monitoring program. Each entry in Table No. 3 involves subjective
engineering judgment. Some will argue that any one of these entries is too low while
others will argue that they are too high. Each entry is based on the subjective
estimate of experienced engineers knowledgeable with the geotechnical aspects of the
project and the consequences of underground construction on existing structures. A
thorough risk analysis would change every entry in Table No. 5; however I firmly
believe that the conclusion drawn from the results in the table will not materially
differ.

The approximate total cost for instruments, installation, monitoring and evaluation
was about $60 million. At the time of substantial completion of the underground
work, payments for repairs and damages to adjacent structures total approximately $9
million. There were no major failures from ground movements. Engineers working
on the project experienced numerous instances where the monitoring program
showed problems and deficiencies in time for corrective action to be taken. My
estimate of the monetary benefit to the project from the geotechnical performance
monitoring program is about $500 million. The estimated financial benefits from
avoiding ground failures, damage to adjacent structures and delays to the Project
exceeded the cost of the geotechnical monitoring program by about eight times.

21
FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics © 2007 ASCE

FUTURE ROLE OF PERFORMANCE MONITORING

We must find better ways to convey the benefits and value of performance
monitoring to the larger community in concepts and terms they understand. Too
many good plans for performance monitoring are reduced or removed at higher
decision levels or poorly implemented by disinterested contractors with the task of
installing and monitoring instruments. As described and illustrated above,
performance monitoring has an important roll on many projects. We must find better
ways to educate the project team on the role and benefits of performance monitoring
to the success of a project.

I believe a promising way to do this is within the context of project risk

management. Risk management has become an important element of owners’ project
management agendas over the last decade. Most large projects establish a risk
management team during the early phases of the project to explore the best means to
identify and manage project risks. Most owners, regulators and contractors
understand the basic concepts of risk assessment, risk allocation and risk management
well enough to meaningfully communicate among themselves and with their legal
and financial partners.

From my perspective, the role of performance monitoring of infrastructure is to

save owners money. These savings result from the benefits that an effective
performance monitoring system can provide. These benefits include avoiding
surprise behavior, reducing the likelihood of undesirable performance and providing
early warnings of unexpected performance so that remedial actions can be taken to
reduce the undesirable consequences. These benefits reduce the potential for delays
to the project from unexpected performance. They reduce the possibilities that
construction will adversely affect neighboring people and facilities. They also reduce
the opportunities for claims arising from unexpected performance.

Looked at in a broader perspective, almost all of the reasons given in the first part
of this paper for performance monitoring seek to answer questions about uncertain
things. We monitor performance because we are uncertain what actual behavior of
the site will be. Uncertainty produces risk. Performance monitoring offers the
project team a key tool to help monitor and manage risk during the construction
process.

Uncertainties and large consequences produce risk. Owners and contractors don’t
like risk. They are increasingly employing ways to manage and reduce risk to control
budget and completion time. Figure 4 illustrates the process of risk management.
Many of today’s so-called risk management programs for infrastructure projects
identify and assess risks, then seek to lay them off on someone else, usually the
Contractor or the insurer. This is risk allocation and not risk management. In the
long run, the Owner pays a higher price through higher insurance premiums and more
costly construction. True risk management adds steps to plan strategies that
minimize likelihood and control consequences, measure anything that can indicate

22
FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics © 2007 ASCE

risk, and take action to reduce

risk at every opportunity. As
illustrated in Figure 4,
monitoring is an essential part of
any true risk management
program. For heavy civil
construction, performance
monitoring has a central role in
risk management.

The traditional philosophy of

most engineers has been to deal
with unknowns and uncertainties
by designing with caution and
conservatism. The actual risks
Fig. 4: Risk Management Process
are arbitrarily masked by a factor
of safety. Their aim is to get the
facility big enough and strong enough that all the uncertainties don’t matter.
However this tact leads to higher costs for the owner without his knowledge of what
those costs are or what they are buying.

An alternate approach is to do more investigations and studies in the pre-

construction phase to reduce uncertainty and conservatism. At some point, the cost
of additional investigations and studies becomes high relative to the reduction in
uncertainty obtained with the additional work and some uncertainty remains. Figure
5 illustrates these points in a conceptual diagram. The horizontal axis depicts level of
effort expended in investigations and design. As this level increases, the
conservatism in the design should decrease leading to lower construction costs.
Curve 1 indicates the total of investigation and construction costs. Point A indicates
the optimal level of investigation that produces the minimum of investigation and
construction costs.

The costs under Curve 1 give no consideration to risk costs. Risk costs result from
unexpected events and conditions not anticipated in the design that increase the cost
of construction, produce delays and/or cause damages to people and property. These
costs are higher for low levels of investigation and design because the probability of
failure is higher due to the large uncertainty in the information used for design.
These costs decrease with increasing level of investigation and design, largely
because the uncertainty in predicted performance decreases due to the additional
investigation. When we add the possible risk costs to the total cost picture, we get
Curve 2. The optimal level of investigation and design increases to Point B. The
optimal total cost has increased because we have increased the level of investigation
and we have included risk costs; however the optimal total cost is significantly less
than at the level of investigation for Point A, if we include risk costs, e.g. Point C.

The really interesting observation is what happens when we consider the effect on
total cost of performance monitoring. There is some evidence to indicate that an

23
FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics © 2007 ASCE

“effective” monitoring
program can reduce risk
by an order of
magnitude (Lambe,
Silva and Marr, 1981).
This reduction comes
from reduced
uncertainty in predicted
performance and
reduced consequences.
Curve 3 shows total
costs when the costs of
a monitoring program
are added to
investigation,
construction and
reduced risk costs. The
optimal level of
investigation shifts to
Point D. The optimal
level of investigation is
Fig. 5: Total Cost Optimization
somewhat higher than
for the case where we
ignore risk costs (Point A), but much less than the case where we include risk costs
but do not monitor (Point C). The total costs with “effective” monitoring (Point D)
are significantly less than those without monitoring (Point B). Point D represents true
optimization of the design-construction process by employing an appropriate level of
investigation and design to remove costly conservatism and using “effective”
monitoring to reduce uncertainty about performance and better control the
consequences of unacceptable performance.

Risk can be quantified. For a specific event, risk is the likelihood that the event
will occur times the consequences of that event. We can manage risk by controlling
the likelihood, altering the consequence or doing both. For constructed facilities, the
consequences can be quantified in terms of money, lives lost, damage level, etc.
Likelihood is more difficult to obtain. It may come from historical data or frequency
or failures of similar projects with similar events. It may come from a formal risk
assessment where experts identify and characterize all sources of uncertainty and
develop a model of the project event to calculate a probability of the event. It may
come from expert opinion by which someone very familiar with the specific issues
opines on the likelihood that the event will occur or possibly provides a numerical
estimate of the probability of the event occurring.

The problem is that each of these techniques provides an estimated likelihood of

the event that is still uncertain unless the value is 0 – it won’t occur, or 1 – it will
definitely occur. Here is where performance monitoring enters. Most significant
events that produce risk to construction projects develop over time. They give off

24
FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics © 2007 ASCE

indicators as they develop that something is happening. This presents the opportunity
before the event actually happens for the process to be disrupted and altered so that
the probability changes, or precautions can be taken to minimize the consequences.
Performance monitoring seeks to monitor these key indicators to detect early
warnings before the event happens.

With performance monitoring, our aim is to get sufficient warning that we can
alter the probability of an event to 0 if the consequences are undesirable, or 1 if the
consequences are desirable, or else manage the consequences. Since altering
probabilities and consequences consume resources, we only want to intervene when
the monitoring system gives indicators that an uncertain event is becoming more
likely to happen. To the pure probabilists, performance monitoring provides the
factual data set to do Bayesian updating of the probability of an event occurring.

Surprising performance can have significant impacts on project cost and schedule,
(plus and minus). Performance monitoring can be a useful construction management
tool to detect surprise performance sufficiently early to manage its impacts on the
project. For performance monitoring to work as an effective component of risk
management, the information must be timely, reliable and presented so others can
understand what the data mean quickly, correctly and efficiently. By working with a
higher level in the project organization charged with construction management, we
have a better opportunity to influence the outcome and add more value with the
results of performance monitoring.

The increasingly important role of performance monitoring to managing risk on a

project should make us consider the best delivery method for performance
monitoring. There is a strong tendency on infrastructure projects to make
performance monitoring a part of the contractor’s work. In general this is akin to
requiring the contractor to do the quality assurance. Most general contractors are not
motivated to make performance monitoring systems work. They generally see
instruments as things that get in their way and they think that measured performance
only brings bad news for them.

I believe that performance monitoring should become the responsibility of the

construction management team. An effective performance monitoring system
provides them with solid facts about the performance of the engineer’s design, the
contractor’s work, and the effects of site conditions.

CONCLUSIONS

Geotechnical instrumentation can reduce the undesirable consequences from

construction surprises. These consequences may be the result of adverse
performance, damage to adjacent facilities, and/or delays. Increasingly, geotechnical
instrumentation is becoming more important by helping us reduce the costs
associated with damages and delays. These costs are becoming very significant
elements of projects located in urban areas.

25
FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics © 2007 ASCE

Performance monitoring should be a part of any infrastructure project that involves

significant uncertainty or large consequences from unexpected performance. Results
from a performance monitoring program can help avoid undesirable performance and
reduce consequences of unexpected performance. The benefits are broader than the
narrow geotechnical issues we have traditionally considered. On projects that involve
uncertainties about the existing conditions, new construction methods or materials,
low margins of safety, high consequences of adverse performance, or tight
restrictions, performance monitoring can provide benefits that may be several times
the cost of the monitoring program.

We need a better vocabulary to explain the value of performance monitoring to

owners and the project organization. This paper has described three approaches to
doing this. One approach is the more traditional style where the technical merits of
performance monitoring are provided. Sixteen valuable reasons to do performance
monitoring were identified and discussed. Some of the reasons to monitor
performance represent relatively new roles for instrumentation (documenting
performance for assessing damages, keeping stakeholders beyond the Project Team
informed, satisfying regulators and helping to reduce litigation). Performance
monitoring may provide substantial benefits to the project in these areas in ways not
previously envisioned.

The second approach to valuing performance monitoring is to estimate how likely

a facility is to perform in different modes and what the consequences of the various
performance modes might be. One then looks at the potential impacts of performance
monitoring on the probabilities of undesirable performance and/or the consequences.
This approach provides numerical estimates of the benefits of monitoring that can be
compared to the costs of monitoring to indicate what monitoring is worthwhile. The
techniques taught in decision theory can help us estimate the potential monetary
benefits of a geotechnical instrumentation program. By applying these techniques,
we can estimate how much money we can justify spending on a project to reduce
potential risk costs from undesirable consequences. These techniques may also show
us where to focus our instrumentation efforts to have the most benefit.

The third approach looks at the project in broader terms to identify all events that
could impact the schedule and cost for the project, thereby producing risk. The aim is
to find ways to manage risk. The level of risk posed by each event is quantified by
some means. Key risk indicators are identified and ways to measure the level of risk
for each are found. A real-time monitoring program is used to monitor the key risk
indicators as the project progresses. Corrective actions are taken when the
measurement system indicates approaching undesirable performance. This approach
is an expansion of Peck’s Observational Method.

As discussed above, performance monitoring must be an important part of any

effective risk management strategy for a constructed facility. As more owners
develop their risk management strategies, I expect to see performance monitoring as a
key component of the risk management program. We might even go so far to

26
FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics © 2007 ASCE

consider performance monitoring as risk monitoring; that is a real-time quantitative

measure of whatever elements of risk that can be measured.

Performance monitoring best belongs to those responsible for risk management on

the project. This is generally with the Owner or its representative and not with the
contractor. Hopefully, this paper helps engineers and owners understand the value of
performance monitoring as an integral part of an overall risk management strategy.

The increasingly important role of performance monitoring to managing risk on a

project should make us consider the best delivery method for performance
monitoring. There is a strong tendency on infrastructure projects to make
performance monitoring a part of the contractor’s work. In general this is akin to
requiring the contractor to do the quality assurance. Most general contractors are not
motivated to make performance monitoring systems work. They generally see
instruments as things that get in their way and they think that measured performance
only brings bad news for them. There are strong financial motivations for contractors
to decimate the monitoring program once they have a signed contract.

Conservative designs based on limited information add significant costs to

repairing and constructing infrastructure. Delays and claims resulting from
unexpected performance add to these costs. I see conditions favorable for
performance monitoring to become a more integral part of the project management
process. When more people understand that data from real-time performance
monitoring systems can alert them to unexpected performance and allow them to take
evasive action early, saving money and time in the process, we will see performance
monitoring joining schedule and cost control as parts of the construction manager’s
resource kit.

Today, far too many well-thought-out monitoring programs are being decimated
during execution because those managing the project don’t have a clue about why the
performance monitoring is important. We must do a more thorough job in the early
phases of a project to clearly define the purposes and importance of the performance
monitoring program. The definitions must be given in terms that all members of the
Project Team can understand how each instrument is important to the success of the
project. I advocate the development of a Performance Monitoring Manual for each
instrumentation program that describes in lay terms the purpose(s) of the monitoring
program, the anticipated benefits, the types of instruments and why they are
important to the success of the project, the significant elements of the instrumentation
specifications and why they are important to the success of the monitoring program, a
listing of each instrument, its purpose and why it is important to the project, limits
values for each instrument, frequency of monitoring and why the required frequencies
are important, and any other material that justifies why a particular aspect of the
monitoring program is important and should not be waived or removed.

I believe that performance monitoring should become the responsibility of the

construction management team. An effective performance monitoring system
provides them with solid facts about the engineer’s design, the contractor’s work and

27
FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics © 2007 ASCE

the effects of site conditions. The construction management team can be in the
position to provide incentives and impose penalties on contractors so they support a
successful performance monitoring program.

Performance monitoring must become more real-time. Consequences of some

surprises have increased to unacceptable levels. Monitoring intervals must be
sufficiently frequent to reveal sudden changes in trends of performance. Results must
be reliable (believable). One has to be able react quickly without doubt about data
quality. Results must be timely. One must have sufficient time to engage corrective
actions. Results must be comprehensive. They must include enough information to
support the next action to be undertaken. Periodic written assessment of the quality
of the monitoring program and its actual benefits compared to anticipated benefits
should be made to keep the program relevant and its benefits understood by all
members of the project team.

REFERENCES

Dunnicliff, John (1988, 1993) Geotechnical Instrumentation for Monitoring Field

Performance, John Wiley & Sons, New York.

Lambe, T.W. (2003) (personal communication)

Lambe, T.W. and Whitman, R.V. (1969), Soil Mechanics, Wiley & Sons, NY

Lambe, T.W., Silva-Tulla. F. & Marr, W.A (1981). “Key Features of the
Geotechnical Safety Program of the Amuay Cliffside,” Geotechnical
Engineering, Vol. II, pp.97-121

Peck, R.B. (1969) “Advantages and Limitations of the Observational Method In

Applied Soil Mechanics,” Geotechnique, June, pp 173-187.

Silva, F., Lambe, T.W., and Marr, W.A. (2007). “Probability and Risk of Slope
Failure,” accepted for publication in ASCE:JGGE.

Whitman, R.V. (1984) “Evaluating Calculated Risk in Geotechnical Engineering,”

ASCE:JGE 110:2, February, pp 145-188.

28