20
Geotechnical News • June 2016
GEOTECHNICAL INSTRUMENTATION NEWS
Eight common sense rules for successful monitoring
Martin Beth
Introduction
Geotechnical, structural and environ-
mental monitoring is becoming a stan-
dard requirement on civil engineering
construction and mining projects, and
the amount of recorded data increases
rapidly. It is very important to under-
stand that if the monitored data are not
of sufficient quality, or if the data are
just data instead of useful informa-
tion to help reduce risks and enhance
operations, then monitoring is a waste
of money.
The writer would like, though this
article, to help convince decision mak-
ers to make the right choices when
dealing with monitoring.
This article is based on “The chal-
lenges of supplying good quality and
useful data for significant projects”,
presented at the Symposium on Field
Measurements in GeoMechanics
(FMGM), Sydney 2015.
Monitoring program and
specifications
Rule number 1: The monitoring pro-
gram must be designed specifically
for the project, and justified by the
project needs.
(See Figure 1)
.
The key point is to understand the geo-
technical and structural behaviour of
the site. Each instrument or group of
instruments must be aimed at answer-
ing at least one specific question,
or one specific problem. A common
mistake is when the monitoring design
of a past project is copied, totally or in
part, to other projects. Consequences
will be either:
1. Under design: For example, weekly
manual survey is still found in
some specifications, in cases
where the risks and the potential
onset of occurrence of the risks
would suggest that hourly readings
would be more appropriate. As a
consequence the risks on site are
not adequately covered, giving rise
to potential incident or accident.
Or
2. Over design: Contrary to what one
might think, this is seen nearly
as often as under design. In such
cases, instruments are installed
that were not really needed. The
end result is that the site stakehold-
ers will view the monitoring as an
unnecessary expense, and not as
it should be - a risk minimisation
tool.
Rule number 2: Specifications must
be clear, listing clear objectives
including accuracy (see note at the
bottom of the page about the word
accuracy), and leaving some degrees
of liberty regarding the methods to be
used.
The major considerations that can help
increase the quality of the measure-
ments are as follows:
1. List clear objectives, if possible
listing the engineering values to
be obtained, the frequency and the
required accuracy.
2. If possible, liberty should be given
to the monitoring contractors to
select the monitoring system they
will use to answer these objectives.
The designers/specifiers cannot be
experts in all the techniques that
they may specify. By leaving some
liberty to the monitoring contrac-
tors the best value for money will
be achieved.
3. The required accuracy of instru-
ments should be achievable on-site
– this is
not
the same thing as ac-
curacy determined in the laborato-
ry. If possible, the definition of the
accuracy should be detailed in the
specification, and also the way that
it can be measured. Some liberty
can be taken with the official inter-
national vocabulary of metrology
in order to define something that
can be estimated. For example, the
accuracy might be defined as “the
band containing 80% of the values
during 12 consecutive hours with
no work and temperature variation
of less than 10°C”. A full article
could be written on this subject!
4. The required accuracy should be
at a level that is necessary and
reasonable. Do not over-specify
here, as those monitoring contrac-
tors who wish to comply with the
specifications will see increased
cost, and those who disregard the
specifications will end up winning
the job. Sometimes we see require-
ments for +/-0.1 mm accuracy
when 0.5 mm or 1 mm would be
sufficient.
Note. The word “accuracy” is used
throughout the text. Depending on
the instruments used, “accuracy” is
correct, but for some instruments a
more appropriate word is “precision”
or “accuracy of change”.
Figure 1. Unsuitable design, under design and over design.
