TRANSPORTATION RESEARCH RECORD 1169

Deformation Measurements with

Inclinometers
GORDON E. GREEN AND P. ERIK MIKKELSEN

used to determine the absolute position of boreholes prior to

Inclinometers have steadily gained widespread use for measur-
ing deformations In landslides, natural slope creep, temporary installing other inst:rllments, for example, geophones or tieback
excavations, earth and rock embankments, slurry walls, shafts, tendons. They may also be used indirectly to measure moments
tunnels, lateral pile movements, and settlements beneath tanks, and stresses in structures and can give good results with
fills, and foundations. Inclinometer data can be used to deter- flexible steel structures where the deflection is large and the
mine displacement magnitude and rate and the location of the section modulus is known. Their value in composite concrete
zone of displacement, as well as the absolute p!)Sition of the structures where the moment of inertia is uncertain is more
inclinometer casing In the ground. High-precision Inclinome-
ters today utlllze servo-accelerometer sensors In a unlaxlal or limited.
biaxial configuration, mounted Inside a waterproof wheel caiJ- The inclinometer commonly used today developed from a
riage. The most common Is the traversing borehole type, where device built in 1952 by S. D. Wilson at Harvard University. It
the wheels are oriented and guided by specially made grooved first became available commercially in the late 1950s from the
casing. Fixed-In-place inclinometers employ the same guide Slope Indicator Company. Such an inclinometer consists of a
casing but are used infrequently due to high cost. Inclinometer probe, containing a gravity-actuated transducer, which is fitted
monitoring Is one of the most labor- and data-Intensive geo-
with wheels and lowered on an electrical cable down a grooved
technlcal measuring activities. In addition to straightforward
data reduction, many errors have to be Identified and dealt casing to control orientation (1, 2) (see Figure 1). The cable is
with before results can be presented and correctly Interpreted. connected to a readout unit, and data can be recorded manually
In the authors' experience, many Inclinometer monitoring or automatically. The inclination of the casing with respect to
programs have failed to yield good results because errors were gravity is measured at incremental depths, and the entire casing
not recognized. In this paper, guidance Is given to the reader profile is obtained by numerical integration. Sets of readings
regarding Inclinometer system design, Installation, monitoring, taken periodically enable both the magnitude and rate of lateral
data processing, and sources of errors and their resolution.
Accuracy of the Inclinometer Is highly dependent on the casing movement to be calculated. The casing is normally set in
equipment selected, the method by which it Is installed in the vertical drillholes or attached to a structure in a vertical
ground, monitoring techniques, correct scrutiny of the data, position to measure horizontal movements (Figure 2). It can
and ability to correct instrument errors. A servo- also be set horizontally to measure heave or settlement (Figures
accelerometer-type system Is capable of a precision of ±0.05 to 3-5) but will not measure movements in a horizontal plane.
0.25 in. over 100 ft when the casing Is vertical or horizontal. Movements of casings inclined up to 45 degrees may also be
monitored .but with considerably less accuracy (7).
The best results are achieved when cooperation and under-
standing between engineers, clients, and contractors provide
continuity in the total monitoring process.

The measurement of ground movements is an essential part of

many civil and mining engineering operations. Monitoring
subsurface movements with inclinometers can be considered a
direct extension of normal surface survey techniques. fu-
clinometer systems are used extensively to monitor the dis-
placement magnitude, rate of displacement (i.e., accelerating or
decelerating), and location of the movement zone. fu many
situations, where, how much, and whether or not the movement
is increasing or decreasing are key issues. fuclinometer systems
can be used to monitor displacements in landslides, natural
slope creep, temporary excavations, earth and rockfill dams,
slurry walls, shafts, tunnels, lateral pile movements, and settle-
ment under tanks, fills, and foundations. fuclinometers are also

G. E. Green, 9835 4 lst Avenue, N.E., Seattle, Wash. 98115. P. E.

Mikkelsen, Slope Indicator Company, 3668 Albion Place North,
Seattle, Wash. 98103. FIGURE I Principles of operation of probe Inclinometer.
2 TRANSPORTATION RESEARCH RECORD 1169

DEFLECTION (cm)
0

~
.,----..,""",.,,.,_- ,_ o[ FIFTH AVE •

• • • ~t: 1 •
SETTLEMENT
• .. • •
w
IJ)

- · -12/82
l
.. . .. .... . . 9/82

MEASURED
CLAY

SAND l GRAVEL

TILL

SANO/SIL Ti CLAY I
I
I
------ 3/83
---8/83
FEM DEFLECTIONS AT
/
I
.
:ib''\\'CIJ\l.,.-3/83 EXCAVATION COMPLETED
,8/83 El. 13
END OF EXCAVATION

EL. 7 (TIP OF SOLDIER PILE)

10 15 20

SCALE IN METERS El , -3 BOTTOM OF INCLINOMETER

FIGURE 2 Shoring wall movements on the 34-m-deep Columbia Center excavation

In Seattle, Washington (3).

have recently become available appear to be more promising

(9).
Inclinometers may also be combined with probe extensome-
tcr systems that measure exhmsion or compression along the
axis of the inclinometer casing, a mechanical probe that locates
casing joints. or an induction coil or m!!gnet/re~-<l switch se:nsor.
When these devices are installed on vertical casings, they
permit settlement or heave monitoring so that subsurface three-
dimensional movements are monitored (JO).
Inclinometer monitoring is one of the most labor-intensive
30 JO
geotechnical measurement activities. It generates large vol-
umes of numerical data that must be correctly recorded,
40
processed, scrutinized for errors, ploued, and interpreted. Ex-
perience indicates that many inclinometer measurement proj-
ects fai l to achieve their intended aim because of a lack of
60 appreciation of the many factors, both human and instrument,
that need to be correctly implemented. Thus, in many cases,
FIGURE 3 Vertical and borlzontal monitoring of an oil
thick files of unplotted data are generated, or depth/
tank foundation during water test loading (4).
displacement profiles wander mysteriously back and forth
across a graph, or large movements are believed to be occurring
Fixed-in-place inclinometers are available. They consist of a when there may actually be lillle or none. More information
series of sensors with guide wheels, each containing a uniaxial concerning inclinometer systems and their use may be found
or biaxial gravity-operated transducer that generates electrical elsewhere (10-16).
signals. The sensors are joined by articulated rods and are
suspended permanently in vertical grooved casing. Continuous INCLINOMETER EQUIPMENT
or remote monitoring or both, by direct wire link, radio, or The accuracy and reliability of the measured position or
telephone modem, permit real-time data to be obtained in displacement profile is dependent on the quality of the casing,
critical situations, in contrast to the manually read probe probe, cable, readout, and accessories selected. A poorly engi-
inclinometer system. neered probe, stretchy cable, faulty readout, or inferior casing
Other types of drillhole survey tools are also available. The will resltll in poor quality data at best, and an unhappy user.
oil well and geophysical logging industry use a variety of Inclinometer casings may be plastic, aluminum alloy, or
photographic, gyroscopic, and gravity-based electronic sensors steel, with rigid or telescopic couplings (see Figure 6). Plastic
that operate in Wllined vertical drillholes and do not require casings may be ABS, PVC, or fiberglass and the grooves may
casing with oriented grooves. At1cmpts have been made to be Conned by broaching, extrusion, or moulding. ABS is
develop versions of these tools (8) that are suitable for civil and fiex.ible and easily cemented, PVC less o. The casing groove
mining use in boreholes 50- to 300-ft deep as an alternative to ~piral should be less than 1 degree per 10 fl, and for plastics, a
the probe inclinometer used inside grooved casings, but these broached tube appears to be betcer than an extruded casing.
efforts have been relatively unsuccessful to date due to cost, Even so, hot sun and improper storage can warp and twist an
convenience, size, and accuracy. The rate gyro-based tools that initially str:tight casing. Casing diameter rartge from 1.9 LO 3.8
Green and Milli/sen 3

.
.. ~ " "
= ~ ·::::: == -----------;----~----:----=----;----;----~----~----~-- --3----~~~~~~~~~~~~~~~~~==========·
I
~!

FIGURE 4 Horlwntal inclinometer monitoring of a rockfill dam (5).

in., and the larger sizes are preferable where movements are casing is extruded with grooves, and a close-fitting, similar
large, thin shear 7.0nes are present, and drilling costs permit. second extrusion is employed for rigid or telescopic joints.
Rigid couplings are available, with self-aligning features that Joints are riveted and must be sealed with mastic and tape. This
include flush couplings for use where borehole space is Light aluminum alloy casing is subject Lo corrosion in alkaline
Connections may be made with rive1s, plastic cement, or a self environmenls. Some records have shown that in the presence of
aligning Wcstbay coupling that incorporates an 0 -ring and steel, electrolysis has caused a total loss in a few monlhs.
nylon shear key that is very convenient. The aluminum alloy Baked-on epoxy paint offers the best protection, but even so,

30 FT. .,.£---.,-.--., SETILEMENT, IN.

1.0 2.0
20 FT.

EXTENSOMETER
EX-1

INCLINOMETER 80-3 (JI

g
ITi
;:
\ 0

~
'2.0 ~

9 IN.
EXTENSOMETER
·., ...'II LEGEND

-·
EX-4 1
1-271 DATE, LOCATION. ANO DIRECTION
I OF TBM MOVEMENT
10 FT. 9 IN
NOTE : TBM WAS WITHDRAWN FROM HEADING
BETWEEN 2·4 ANO 2-17 AND THE
HEADING FILLED WITH GROUT.
THE HEADING WAS READVANCED
20 FT ON 3-20·83.
..... 1.0
30 FT. -'----J HEAVE, IN.
360+90 360+70 360+50 360+30
STATION, FT.

FIGURE S Vertical rock deformations In a tunnel In response to tunnel boring

machine advance (6).
4 TRANSPORTATION RESEARCH RECORD 1169

measure angle. A very small, damped pendulous mass sus-

pended by an elastic flexure is electrically nulled, with the
voltage proportional to angle. The servo-accelerometer is accu-
rate, reliable, reasonably robust, and compact, and two of them
can be mounted in a slender biaxial probe for vertical casings.
The probe body (Figure 6) should be waterproof and prefera-
bly of stainless steel, wiih a well-engineered wheel carriage
that has little or no side play in the wheels, axle, and swing arm
assemblies. Preferably, the wheels should have double ball
bearings and a wheel rim profile that is compatible with the
casing groove geometry. Self-centering wheels are better than
the earlier fixed-spring wheel assemblies. Wheel geometry and
carriage design control the precision with which the probe can
be relocated each time in the casing. This factor is particularly
important with the currently used biaxial probes, which utilize
only one pair of grooves and rely on the sides of the grooves for
B-direction location. A sealed electrical connector should be
located at the top of the vertical probe and at either end of a
horizontal probe to permit it to be turned end for end. Inclined
probes are available in which the sensor is set with its axis
vertical and at an angle to the inclined probe (18), but these
cannot be reversed end for end to eliminate zero errors and are
less convenient and accurate in use. At the same time, it should
be recognized that the accuracy of vertical or horizontal probes
degrades severely when they are used in steeply inclined
casings (7).
FIGURE 6 Digitilt inclinometer, manual readout, 1 ne eiectricai cao1e shuuiu ul.' llexible, waterproof,
Inclinometer casing, and couplings. untwisted, durable, easy to grip, and permanently and accu-
rately graduated. It must not stretch under load or with time,
and the outer sheath must not creep over the inner core or
serious errors will result. These requirements are stringent and
corrosion can occur, and ground conditions need to be carefully
assessed before an aluminum alloy casing is selected. In benign some inclinometer cables do not meet all of them. Connection
environments, some users prefer aluminum casings over plastic to the probe requires 11 heavy -duty waterproof electrical con-
(17). In some instances, seamless welded square steel tube can nector, but a nonwaterproof connector is adequate at the
be used where great strength is required (e.g., auached to readout end of the cable. For cable lengths up to 200 ft, a cable
driven steel piles) or where the user wishes to attach strain reel is not required because the cable is most easily handled in a
gauges to the casing for axial strain measurements (e.g., in manner similar to that used with a climbing rope. For cable
concrete piles). Otherwise, the steel tube is subject to corrosion lengths of 200-1,000 ft, a manual or powered reel is essential.
and to loss of precision of probe location within the casing. The readout unit should be robust, reliable, easy to read,
Extruded steel tubing is usually twisted excessively and should portable, insensitive to temperature, and weather resistant
not be used. (Figure 7). Traditionally, manual data recording on a field sheet
The gravity-actuated sensor may employ a rotary poten- is employed. Battery-powered digital readouts that display one
tiometer, bonded resistance strain gauges, vibrating wire strain or two sensor axes sequentially or simultaneously in a boxed or
gauges, an electrolytic level, or a servo-accelerometer. The now clipboard package are available. The amount of data that must
discontinued Slope Indicator 200B employed a rotary poten- be recorded is voluminous, and errors are easily made when
tiometer with a relay-operated contact to reduce friction. This manual transcription or computation is employed. Direct field
sensor was stable and reliable but bulky. The Soil Instruments reading to a solid-state memory unit with field data checking
Mark II inclinometer, also discontinued, incorporated a strain facilities is preferred. Field or office computer processing of the
gauge steel leaf spring with an oil-damped pendulum. The data recorded on magnetic tape or disk, including transmission
sensor was reasonably stable and reliable but subject to damage by telephone modem (if needed) and automatic plotting reduce
if it was not immobilized in transit. European manufacturers the labor, cost, and chance of errors. Most manufacturers now
have incorporated the vibrating wire strain gauge on dual-axis provide automated readouts of various types, of which the
pendulum devices, which are reliable but cumbersome. The Slope Indicator Digitilt RPP (Recorder, Processor, Printer) is an
electrolytic level has only been used in horizontal inclinome- example (see Figure 7). This instrument can record, store, file,
ters because (a) it is currently too large to mount laterally in a and reduce inclinometer data, spiral data, and azimuth angle
vertical probe, (b) it has a more limited angular range, and (c) it input. Built-in software allows the data to be corrected for
is relatively sensitive to temperature. The servo-accelerometer systematic errors and permits spiral and inclination readings to
is the sensor most widely used today. A product of the space be combined. This software also allows coordinate system
program, the accelerometer can also be used in a static mode to rotation, a.lld all of these procedures can be done on the spot, in
Green and Mikkelsen 5

Il
llNSH:il IUMEN1

FIGURE 7 Dlgltllt Inclinometer In use with recorder,

processor, and printer (RPP).

the field, if needed. Results are tabulated and graphed on a

built-in electrostatic printer, and data files are stored on mag-
netic tape. The Digitilt RPP can operate as a computer terminal
and can send data via a telephone modem to a remote com-
puter, or it can be connected to an external printer or plotter.
Fixed-in-place inclinometers (13) consist of a series of long- FIGURE 8 Fixed-in-place inclinometer.
gauge length probes (Figure 8) permanently installed to
provide continuous automatic and remote deformation data at
critical locations and where the relatively high cost is justified.
These insLrnments will not provide a detailed profile of ground
movements unless sensors are closely spaced. Sensor zero drift
with time can be difficult to detect and is not eliminated as it is
with the portable probe inclinometer, which can be reversed
through 180 degrees. Available accelerometer sensors are fairly
reliable, and fixed-in-place inclinometer systems have been
reasonably successful (19). Fixed-in-place inclinometers can be
installed permanently in casings adjacent to probe inclinometer
casings or may be installed during a critical period in a probe
inclinometer casing and then removed. Even a single sensor
fixed-in-place unit can be installed across an identified thin FIGURE 9 Installation tools and accessories.
shear zone.
As noted, a casing spiral or twist can occur due to manufac-
turing, storage, or installation deficiencies. Serious spiraling Accessories are needed to install and operate inclinometer
has, on occasion, been identified in the field. Spiral should be systems (Figure 9), including installation tools, pulley wheel
measured (a) on deep casings, (b) where measured movements and cable clamp, grout valves, and end caps. Protective
are in suspect directions, (c) where absolute casing position is enclosures, a calibration facility, instructional material includ-
required to good accuracy, or (d) where installation difficulties ing videotapes, data sheets, software package, calculator or
have occurred. A spiral survey is normally required only once, computer, telephone modem, and printer are also available. The
and computerized data reduction is essential. One type of spiral calibration of the probe inclinometer may change with time,
survey tool consists of a long torque rod mounted between and calibration frames that provide a functional check are
wheel assemblies with a rotary displacement transducer. available from some manufacturers. On-site limited calibration
6 TRANSPORTATION RESEARCH RECORD 1169

checks may also be performed by using short lengths of fixed surveys are proportional to the product of casing inclination
casing cast into an immobile concrete block or by extending and angular changes in sensor alignment. For inclined casings,
some casings to well below the movement zone so that regular sensor alignment changes of 1 to 2 degrees may produce errors
surveys of immobile casing are performed. Otherwise, calibra- in the measured displacement of several inches per 100 ft of
tion should be done periodically by the manufacturer on a casing (Figure 11). Sensor alignment change occurs with time
dividing head. Inclinometer system accuracy is dependent on a due to wheel play in a groove, wheel carriage wear, internal
number of factors, including sensor design and construction, changes in the sensor, and changes in alignment between the
installation technique, casing quality and orientation, care and sensor and wheel carriage. Thus tight specifications on bore-
attention given when taking readings, and instrument mainte- hole verticality and quality drilling techniques are preferable.
nance (10). True independent checks on system accuracy are CASING DRIFT (m)
difficult to perform and rarely done. An inclinometer system 007 0.--~---.-
1...-~2...--.--..3~-,,.
check test setup (12) is shown in Figure 10, but this excludes
field related factors that can degrade measurement accuracy
(20).

0
3"

SCAFFOLD TUBE
....-...-- SINGO OR SIL
CASING
VEE -BLOCK
2·
PLUMB· LINE

1"

TRUE N-S
PLANE

NOTE CASING DRIFT OVER ANY DEPTH

N·S INCREMENT IS MEASURED AS HORIZONTAL
KEYWAY DEVIATION FROM TRUE VERTICAL AXIS IN
PLANE TWO MUTUALLY PERPENDICULAR PL.ANES
DEFLECTION ERROR IN ONE PLANE IS
RESULT OF DRIFT IN OTHER PEAPENDIC
OFFSET ULAR PLANE
MEASUREMENT
WATER
GEOMETRY FIGURE 11 Measurement
error as a result of casing
FIGURE 10 Inclinometer casing test set up In a Inclination and sensor rotation
stairwell (12). (U).

In selecting appropriate equipment for inclinometer mea- The largest inclinometer casing size should be used where
surement, instrument accuracy, quality, reliability, manufac- hole size permits, because this size will accommodate the
turer reputation, and availability of service, manuals, and largest movement before probe access is obstructed. Drilling
software should be the prime considerations. Hardware costs costs or structural constraints may necessitate using smaller
are only a small proportion of the total measurement cost, sizes, for example, the 1.9-in. O.D. plastic flush coupled casing
which includes drilling, installation, data processing, and eval- in an NX-borehole. The casing is coupled together (typically in
uation. Accurate, reliable instruments are essential. Many have 5- to 10- ft lengths), the joints are sealed (including the bottom
discovered this too late, and low cost is a poor basis of choice. of the casing), and the casing is lowered into the borehole. If
the borehole is filled with water or drilling mud, water must be
INSTALLATION PRACTICE added inside the casing, and extra weight may be needed to
over come buoyancy. The casing should be oriented so that
Casing installation procedures are described in detail in man- grooves are aligned in the anticipated predominant movement
ufacturer's instruction manuals (13, 15) and discussed direction, and twisting of the casing should be avoided during
elsewhere (10, 14, 21). The inclinometer casing is commonly insertion. Groove alignment at couplings is maintained by key-
installed in soil or rock in vertical boreholes drilled by a variety ways of varying types (Figure 12). Where settlement or heave
of methods, depending on the material type, borehole stability, in excess of 1 percent is anticipated, telescopic couplings are
amount of water present, drilling equipment available, casing required and should be set during installation at the appropriate
size, and cost. For observational accuracy, the borehole should position for the anticipated movement direction. Telescopic
be as close to vertical as practicable. Errors in inclinometer couplings complicate both installation and reading procedures
Green and Mikkelsen 7

~ GROOVES TO ALIGN
INCLINOMETER SENSOR

152 mm (6 on)
ALUMINUM CASING ALUMINUM
COUPLING
BUTT JOINT AT

305mmfl2on) CASINGENIO POP R !VETS

AT ENO OF
-r- ALUMINUM COUPLING
POINT OF CONTACT
COUPLING

1 FOR SETTLEMENT STANDARD

152 mm (6 m) PROBE WHEN CASING
MAKING SETTLEMENT LENGTHS
MEASUREMENTS I 5m 15hl ALUMINUM
or3m(10ft) CASING
CASING ENO
:~~~~~up~~ I
POP RIVETS AT
ENO OF COUPLING 1 POP RIVETS
AT.,, POINTS

FIGURE 12 Inclinometer casing detalls (14).

and should be used only when required. Care is needed during also serves to keep the casing straight and helps overcome
installation to avoid extending or collapsing the preset assem- buoyancy effects unLil the grout sets. Where large movements
bled joints. All casings should extend 10 to 20 ft below the on well-defined shear zones are anticipated, a large borehole
zone of anticipated movement to provide a stable reference and weak grout backfill should be used so that the casing wiH
section. Inclinometer measurements are usually more accurate shear locally through the backfill, maintain probe access longer,
and reliable than surface survey check measurements made on and provide continuity of readings. The localized shear move-
the tops of the casings, and reliable casing base fixity is vital. ments will be redistributed over a larger casing length, but this
A casing installation in a borehole must be backfilled around is preferable to getLing no data. Completed installations should
the casing with sand, pea gravel, or grout (Figure 1) to ensure
be wasb.ed ou1 10 clean the casing, and adequa1e surface
conformity with the surrounding ground movements. In-
protection should be installed and locked, if appropriate, to
complete backfilling or backfill settlement causes spurious
guard against damage or vandalisnL Casings can be suc-
casing movements that arc best avoided. In some cases, for
cessfully installed in deep holes (200 co 1,000 fl) by using a
instance, soft ground, the backfill strength should be matched
safety cable attached to the bottom of the casing and multistage
10 the ground strength so that conformity is achieved. Uniform,
coarse clean sand or pea gravel can be used to backfill the grouting with an external tremie pipe.
annular space in stable shallow holes, and clean sand can be Vertical inclinometer casings installed during construction of
sedimented through water or flushed through a tremie pipe. earth or rockfill dams must (a) incorporate telescopic coup-
Granular backfill is more prone to bridging and seulement. lings, (b) be adequately protected from damage by construciion
Grout pumped through a tremie pipe extending to the bottom of traffic, (c) be extended seccion by section as the embankment
the borehole is the preferred method, but this may not work in rises, (d) avoid providing a zone of increased penneability, and
open granular materials. An external grout pipe along side the (e) settle by the same amo1mt as the surrounding production
casing can be used if the borehole is large. It is more machine-placed fill. These needs are often difficult to achieve,
convenient to grout through a drill rod inside the inclinometer and great care and quality control are required. This is par-
casing, connected temporarily to a one-way grout valve at- ticularly true if other instmments are also being installed, as is
tached to the bottom of the casing (Figure 13). The drill rod often the case (7).
8 TRANSPORTATION RESEARCH RECORD 1169

some, if not all, of the casings beyond the base of the wall. This
usually requires drilling 10 to 15 ft beyond the base of the
installed wall through a steel pipe.
Inclined inclinometer installations are used in sloping cores
MOOIFIEO FEMALE
OUICK.CONNECT
COUPLING WITH
and concrete membranes in dams and for drillhole alignment
CHECK VALVE surveys with temporarily installed casing. Examples and
_; Qf.
RUBBER GASKET
:; !'. methods for dams are discussed by Penman and Hussain (22)
MAl.E QUICK-CONNECT
: :. COUPllNG WITH and errors are analyzed by Mikkelsen and Wilson (7). Inclined
:: : CHECK VA.LYE
,;: . installations in drillholes should be avoided where possible
CHECf< VALVE
{: O J;: because measurement errors are too great, due to casing groove
HEAVYOUTY

.'
PROTECTIVE CAP --.., , .: r. alignment uncertainties and sensor errors. With care, short
. '
holes can be surveyed with modest accuracy, and position
surveys of tieback holes can sometimes be obtained with
sufficient accuracy. Where inclined installations on sloping
cores are required, groove alignment must be very closely
controlled. The trolley and track described by Penman and
Hussain (22) is an elegant, if complicated, technique that
PIPE CAP
uses a conventional vertical probe that can be reversed
(Figure 14).
Fixed-in-place inclinometers are installed in vertical casing,
GASKET SEAL TYPE which should be installed as described previously. The fixed-in-
place inclinometer assembly is suspended in the casing, and a
FIGURE 13 Grout valve for Inclinometer casing waterproof electrical cable from each sensor is connected to a
Installation. junction box at the surface. A conveniently located automatic
data acquisition system (ADAS) monitors each sensor sequen-
Horizontal inclinometer casing is usually placed in com- tially. The system may be hard wired or radio telemetered to the
pacted fill beneath an embankment or structure (4, 5). The ADAS, w·hich can be accessed by te!ephon~ ~~ requiTed. Fixed-
casing must be straight, with one pair of grooves vertical and in-place inclinometers can be transferred from one casing to
preferably on a constant flat, free-draining grade so that the another and can be salvaged at the end of the project.
casing can be washed out. The instnunent is pulled into the
casing with a steel cable. If access is available only at one end, MONITORING PROCEDURES
the cable should be passed around a dead-end pulley and
retur:ned in a second small PVC pipe ..alongside the casing. Installed casings should be monitored regularly by a trained
Inclinometer casings can be cast into concrete piles, grouted two-person crew who are thoroughly familiar with correct
into hollow core concrete piles after driving, attached to steel procedures and have read and digested the manufacturer's
sheet, H, or pipe piles, installed after driving in steel tubes, or instruction manual. Inappropriate procedures produce poor-
grouted into slurry walls. When the inclinometer is being used quality data and a disillusioned client. A reading interval equal
to monitor wall deflections of any type, it is advisable to extend lo the distance between the probe wheel carriages is usually

inclinometer

latch ing mechan ism

stain less s t e~I insert

p~ll out pl u g

grooved wheel

mai n den se asphalt layer

running rail

latching mechanism
st riking plate
su b-frame

FIGURE 14 Inclinometer trolley/track used on an Inclined upstream

membrane on a dam (22).
Green and Mikkelsen 9

most appropriate, allhough a greater interval may sometimes be include equipment model and serial number, data set, a dinry of
used with little loss of accuracy, provided that thin shear zones field activities, and observations related to the installation. Jobs
are absent. Preferably, the same instrument should be used may extend over many months or years, and detailed records
throughout the job because probes are not interchangeable can be crucial when personnel changes or problems arise.
wilhout generating systematic errors. Instrument damage or Where manual readings are taken, data are recorded on a
manufacturer recalibration may necessitate reinitialization of field sheet formatted for computation or computer key entry
Lhe readings. Readings should be taken from the bottom of Lhe (Figure 15). Face errors for pairs of corresponding readings
casing up, with close depth control (± 1/• in.), by using a pulley (i.e., A+, A- and B+, B-) must be det.ermined and should be
wheel and cable clamp or similar device anached Lo the casing relatively constant at each depth. This provides a critical field
collar. In vertical casings, a uniaxial probe requires four passes check on data reliability. A similar procedure muse be followed
up the casing, whereas only two passes are needed with a with an electronic notebook or a computer-based readout, such
biaxial probe, which saves field time. The B-sensor data as the Digililt RPP. Field equipment or procedures that do not
obtained wilh the biaxial probe are less accurate than those of permiL on-th~spot data quality checks should never be used.
the A-sensor (i.e., parallel to the wheels) because the side of the Two or three initial data sets should be obtained on a casing
casing groove controls the B-axis sensor alignment. It is before any construction activity to ensure reliable baseline data
essential to obtain readings on both faces, that is, with the and confirm that backfill movements are absent.
probe turned 180 degrees, to eliminate zero shift errors and
enable data quality "checksums" to be computed. Checksums
must always be computed and scrutinized for errors in the field.
~~ .
INCLINOMETER DATA SHEET
•
IW .. q !i ,. .a. -
• • ..
A. I IH
.....
I
' ....
~ I p
.
e.
.. -
Slnoo::::..--:-.:=:;m
..
_A. II ,,.
.. '
·-
.. ft -
,,. .,
, ... ·-·
.Ji M.l. IAti F-.
Q).ICllHlll Q)M:>altlU lllil.l('lltPC)T.-UCOl.. tOI MOV..,lr
Where errors are suspected, repeal readings must be taken on
the spot. In horizontal casings, Lhe uniaxial probe must be ~) tQ.l.Na
- 1 ~7 , 'di•.o< T, 16~~
4 S(tHo >DA'(
.1.ofu.
•TOK t Sl&T
'19
l~ f
·I
disconnected from the cable and reconnected to the second
socket at Lhe other end of Lhe probe to reverse the probe, and
only two passes are needed.
H casing spiral is suspected to be of significance, a spiral
survey tool should be used to survey the casings. All portable
1-•·.-----.----.-----.----.------,
equipment should be handled with care and protected during DEPTH DIR A+ OIR A- OIR 8+ DI R 8-

day-to-day use. The back of a pickup truck is no place for a 11' ~ 1--~~;~~~~~~u.2~~"'!-.......c~~.?~"'t-~~-~~.~ol--~-~~~·':"'-i'
~"""'°"' ~ ·D 'Z, 7 .D i t!J ~ ~ () 'l J q ~
delicate instrument on a rough site; instruments require "tender -C( ~ .,. ...., 'Z '- '1
-
- 2. -" _4 ,., -... ~
_.
- 2 3- a;.
./.}. Z..'- ~ -
~oving care" for good results. The cable should be protected '1 6 1 " "" - 1. !I

I .d .o 7 ~ _.d ~ .d ~ J
from nicks, and connectors must be kept clean and dry. - '1 I '.!!a. 0 • t

Although most probes and readouts are rated for a reasonably t t} ..n 't' .:Z. .. "1 ~ 0 I n I D d
-i. n . o '2. 4;o - '1 "> - :J

broad temperature range, neither should be left unshaded in hot OICITUKJlDfl-'---'"'·~2~-""'l-~~~·o=w~>l-'-~-~~~b.,,....1--~~""l-~--'~

MOOt:L~'" 1--~•~•~~"'f.-~~·~~'·.ii-~~-~•u~ ~ o 1--......._~.~
' A'!-I-......_.....:......,_,
1 .h . o ; 4 b ... _4 . .-:7_6 - J .fl .(,
climates or exposed to prolonged temperatures close Lo or ~ ... o .. ._~ - J ~ • - " G < I
below freezing point. Battery life at low temperatures is
considerably reduced, dependent on type (23). Electronic com-
ponent malfunction is more likely under extreme environmen- . 3 -~-o a. 7 e - .z J e 1 L'J - , 1
h"' Z6.1 - '2.i .l! I .t=. 11 - 1.~.7

tal conditions. 4~.n I.A.~ -16 - ~ 1~0 - J 9_,

FlUlction checks on the instrument are necessary. Hanging t-1
r;y
POL.AAITY
..d . li,O
.II 0 .r.:l1 -l
.ll ~ - .I
.,. I
llF' 0
'1 I
I . "1
4
J
j
...
- /
~

t. . ~
the probe freely on 6 ft of cable gives a quick daily zero check, TU AHGll 1--~~
....L')"''""o1--~•,_.,_.
1 "04-...-.:-"""""'="<;'1--'-~........·<>:.j..
·..._.~
- .o...
' ""L""lo

and regular readings in fixed short casings are desirable. ~-~~~~---~~~ ........~~ .~~~..........
~~~ ~·+-"~~---t~~ .~~~~-'-1
Manufacturer adjustment, maintenance, and recalibration are
advisable once per year or if instnlment performance is suspect.
The instrument must be cleaned and wheels checked for FIGURE 15 Field data sheet formatted for keypunch data
excessive wear and oiled daily. Where practicable, periodically entry.
washing out the casing will remove grit that can collect in the
grooves and cause increased groove wear, degrading data Computer-based field readouts are being used more widely
quality. Periodic surface surveys of both line and level at the because they provide convenience, reliability, elimination of
casing collar provide an important check on measured gr01md data transcription errors, on-the-spot computation of casing
movements and should always be made. If possible, where high position or displacement, and cost savings. A recent cost
accuracy is needed, duplicate sets of inclinometer readings comparison (24) indicated a tenfold reduction in monitoring
should be taken and an average data set determined. costs when use of a Digirilt RPP was compared with use of
Data should be reported in an initial installation report, manual equipment.
followed by monitoring reports (13). The installation report Monitoring telescopic casing where large axial displace-
should include equipment model and serial numbers , calibra- ments occur poses difficulties in repositioning the probe in the
tion, description of installation, site plan with casing local.ion, casing. If readings are made in the normal way at uniform
elevation, groove orientation, convention adopted for Lhe sign depth intervals, the absolute profiles must be compared graph-
of the movement and probe orientation, initial data sets, and a ically. If a separate settlement survey is performed, computa-
spiral survey, if one was taken. Each monitoring report should tional techniques can be used to compare the two data sets.
10 TRANSPORTATION RESEARCH RECORD 1169

1120
Alternatively, the probe may be relocated at the same position ...mr1120
w - DOWNSTREAM
in a casing section by "feeling" the casing coupling as the "'2
w
wheels pass through. Again, a separate seLLlement survey is a:
w
required Additional bookkeeping complications occur when
casing is built into an earth or rocldill dam because the refer-
..
"'a:
0
u
1080

ence point from which measurements are made is constantly Elev

w 1040
changing (25) and reliance is placed on surface surveys. As yet, I
"'2 lee\
there is no better solution to this problem. w
a:
Fixed-in-place inclinometers should be read daily after in- 4 w
z
3 u: 1000 1TI 1 I
stallation to check for zero drift of a bad sensor. Depending on 2 EL 99 5
EL 990
the readout arrangements, threshold readings and alarms may Inclinometer
1111!
sensors wtlh
be set and telemetry systems may be checked. Automatically connecting 960
collected data should be scrutinized periodically for errors, rods
15 10 05 0 05
01 splacement 1n inches
particularly drift, and periodic probe inclinometer surveys
should be made on an adjacent casing (Figure 16). FIGURE 16 Horizontal movements measured with
probe Inclinometer and with fixed-In-place Inclinometer
In adjacent casings (26).
DATA PROCESSING, ERROR SOURCES, AND
ACCURACY reduction is tedious even with the help of a programmable
calculator, and many users recognize the advantages of com-
Data checksums (i.e., algebraic sum of pairs of readings 180 puterized analysis (24, 27). IBM-PC software is now available;
degrees apart) must be computed and scrutinized in the field, for example, Slope Indicator/Geo-Slope PC-SLIN, with key-
and data should be processed as soon as possible. Manual data board or RS-232C serial input, tabular and graphic output, error

P'ocld Data ml D;.placemonta

------·- PC-SL IN
- -~- ------ -~ -~-------~--------------------80~26------------------~
IE~ EXAl'IPLE NO. I
DIGITILT DATA
PAST PRESENT

FIL£ -S A:Sll.SN At Silt. SN

JOB M.tUIER W-25S3-0l W-25S3-0l
HOU: IUUIEA a-1 1-1
DA TA SCT NUl'1BER 7 7
DATE SEPT 20173 OCT. ll/711
TlllE 13110 12:30
READIN89 ~R DIRCCTION 2 2
INST"RUENT ~ 039 039
CONST~ 20000. 20000.
ERROR llNGLE - A COl1P. .ooo .000 Total Dellectlon belweMI
ERROR MIGl.E - 8 C011P. .000 .000 the PAST and PRESENT sets
ZERO SHIFT - A COl1P. 0 or data
ZERO SHIFT - 8 C~. 0

PAST DATA

PAST
PRESENT DATA

PRESENT
~
Df:"FL. DEPTH
A+ A- OIFF. A+ A- DIFF. CHANGE IN. FT.

in -179 ~ 124 -12:5 249 -107 I. 0149 2.0

202 -l9S 397 210 -188 398 1. l 112 4,0
211b -293 S49 270 -283 ss= 4 I. 1106 6.0
2S9 -261 S20 267 -264 S3l 11 l . 1082 8.o
254 -249 S03 263 -2:51 Sl4 II l.1016 10.0
247 -248 49:5 246 -24S 4'1 -4 J.09:50 12.0
237 -236 473 234 -234 468 -:5 l .0974 14.0
234 -247 481 241 -24S 4911 s I. 1004 111.0
25S -256 Sil 252 -2:50 S02 -9 l.0974 19.0
243 -2:57 soo 251 -2S9 SIO 10 1. lO::!B 20.0
26.7 -279 S46 21.3 -:?66 s~ -17 l .0969 22.0
313 -319 632 319 -320 639 7 1 . 1070 24.0
33S -328 61.3 346 -3Z!I 1.74 II l . 10::!8 26.0
3112 -362 724 ~55 -!54 709 -1:5 1.0962 29.0
288 -294 :582 299 -:?SS S84 2 l.10~2 30.0
1872
-·~
211S -211b S3l 936 1341 l .1040 32.0
283 -279 :562 S49 -SS3 1102 S40 .:?994 34.0
2'90 -234 S24 299 -277 Slob 42 -.0241. 36.0
273 -281 SS4 278 -283 :5111 7 -.0498 38.0
271 -2119 S40 2113 -::.?58 :521 -19 -.OS40 40.0
194 -19:5 389 194 -193 387 -::! -.o4:6 42.0
112 -187 3119 193 -183 3116 -3 -.0414 44.0
190 -192 382 190 -190 390 -::! -.0396 411.0
1711 -17S 3Sl 170 -171 341 -10 -. 0384 48.0
200 -203 403 210 -20s 41S I:! -.0324 so.o
2:?1 -223 444 219 -219 439 -6 -.03911 :52.0
224 -237 4111 :!..."':5 -239 4114 3 -.03110 S4.0
2:54 -21>0 S14 253 -2:50 S03 -ll -.0378 Sb.O
192 -197 389 189 -191 379 -10 -.0312 se.o
240 -2111 SOI ::!:53 -267 S20 19 -.025:? 110.0
224 -231 4:5:5 213 -215 4::!9 -27 -.03116 112.0

FIGURE 17 Tabulated computer output.

Green and Milclulsen 11

Data Qualit.y Chock ond Statisticol Routino

M.IL1,.SIO( A-..DUCTtA..CNOlt . . Wllf

OltlTILT DAU

&•T•n.lll(NT ~ltFOllll•aNt[ CtCCll.

1... nui.. ~' ...,..,,. I C\Mlltlfrlf Ut . . . . . . 7

MOl.l ......-r.• ·-· l"ltlAL DUE 1051:~.n C~UNT OAH UOCTtft
_________
....CUHllNf
........ --------
.------ , . ,_ ., .
........ ... -. . . ........
.
INITIAL DAU OIU ltlllTIA6. DATA '-"lNT DUI
1111111

... ......
DI .. '"
•• _,,.
INJr
•• . ....
CIMO Ol"M

-·. ......... . "·

Dllll'F Dllll'lll'

-••!t _.,-·' ... .... ..• ...

.....•••••••••, -•"
117

•l•l

-· ..
• ...
.
'",
•U!t

•Hl -u
•
1

... .... ... .....

-111
,
-s
_, ,., -us
I
Ill

...
•HS ...• ••
..•
•••
•••
....
••••
.... , _,,. -·-· ...•••..., -n•
-ao• '
-1•1 •Ult
·1•9
... ,... ·2!tl u'
•'
u1.w
•••
••• .... ....
•Ho
-an -u
•i&I
•lH •••
....
-·-n_,'' _,u•
-ai.!t 11.u
1 -u-
-· -·•• l••U
••·a
.•• .....,,.....
U•
...•• ..
uo •UI
••••
....,.,. ,, ....,.,. -··_,,...•' ........, ........
..., .. 1 •IJ -i·~ 111 -uo II• -111

-·_, • .- .......... ......

-.uu

.
ZS~ •l!tlb 2~1 1 11.u IOI •189 •&01 11.0
;!Ill -u1
... •• ID•\J -11
-u u• ••••
-··.••• .....•• ._,-·, •
11.D ID 1

-·.. •
......••• ,.. -·-·. ...•••••• ....
-sat •.Hll , •• a -11 u
, •• o
•••••• _,,. .....-· u
..... -u

-·
·)21 •HI 11 11
...

..
I
••
I 11.0

. .. ... ...
_,.-·
........... ..., -·-·• "• .........,.,,,.,,..
......,.,"' ..... .... -··-··..
•.ll!t JD.lot

.,, ....
II ID
-1••
.-·• ....-·• . -u
I
,Ja.u
.... " U• -11 I

,,,
-119
•Ill
.. ,.
...... ..,., -·...-· ....,.... ....
•••
•Hl
-in
-au
11
, •• o
110 -u• •H•

...... ... ........, ......-· ••'

u

....••••........
II• •lH -11•
l
• •• •o.o •U7 11

......'" ...,............... ..._.,_,._,_,

•19! •lU I •l•D •lH •lU

...... ...., _.,-·-· ....,."''••• .....,,,..

U.t: •lll
•• •
-·•• ••
••• o
••• o
-11
-u

.....•.•• ....... -··_,-·. .'

•ltl ·&•u IU I

-· .........
.,. ..... ..
.
•lH ••• o -n
...... .._,,,,., ..-· ...
1 II

......... ........-u•....... _,.-·-·

•••
.
-10.s 11n •II! I !to.o •lll -a•o I
an -uo

...-··-·_, ...
17J
·.U•

...•••...,.,••• ......,.,_,,,...... _.,-·_,_,_, '"...••• .........,

-1•0 •HO _,' -·•
........
lti

..,
-1•1

...·•.. ••
I

....
''·"
............
-1•1

...... _., .
-an
...
......"' .... -··... _.,.'
-11 ••• o

.••••••, ....._,,.. ..._, •••...•••'" .... llJ -au

....-· -·• ....
U) 111
.,,
•lll
.. )

- ..
•lU • •• o -u I
-111

....,.,.. ....
•llt

,.''·'..
....
•IU •llO

... , •
-·· ...'"'" ..,.
-11 -1•0 1u.a Ill II

,., ....
-· ... ....
•llO •IH
.... • Ta.a -n -u ,, ,
••• ·101 -102

' -· ., •• o
,,.a ·l7•
•U7
-11 •110
•l

D"UILT oau
llt\,1111,,.lJill'f •UFOllllllf&""[ c~c~
~AlllT STIUS.TIC~

l"llfUL IU ~· l
1.. 111.aL o.nE aow•.n '"""'"' "' ....... '
tl.lllllllNT OAfl llDCftH

...
--------··-·SUll--·--
--·-·-··-··-··--··--·-·--···· -··-
...
···--···-····-··-- ·-···-.... --...........-. . ......-
.... ......... ........ ..... ...... ...... ............ ........ ....-··..
........ .......
...... INITIAL s.u--•--01,,r•rNCl·-·
lf.M•l~' I- HUlUL ~-I·- ,...,..,,., iUW••l---Ol,,l•[frilCC•-1

..........
"D• STD• U'l. sro. "O• ITDo
INl[lllW&L oc. • OC•• CAN IX••

-·· .....'· ....-·· ....'· .... "·...... ....... ...

1·0 ZDoD
l1o0
1112.a ,.'"'" -.o.o
•a.o
O•
). •llo
•a.o
'"
["TUl ._O...t.
l•.a -1.

-·· -··
-1.

-·· -··
-·· ... .. .. II•

FIGURE 17 continued.

detection routines, and systematic error correction features. decreasing deep-seated movement rate, that is, a more stable
Readings can be input manually from hand-recorded data condition developing.
sheets or by RS-232C serial input from magnetic tape when A wide range of random or systematic errors can occur, both
recorded automatically, as with the Digitilt RPP. obvious or subtle. Experience indicates that data interpretation
Processed data (Figure 17) should be scrutinized by using is often difficult and that wrong conclusions are drawn concern-
error detection and correction routines to optimize the data ing magnitude and locations of movement. Both authors have
quality and enable correct interpretation of engineering be- extensive experience in trying to diagnose errors and draw
havior. Graphical data presentation is essential and may include
rational engineering conclusions from conflicting and inexpli-
(a) slope change/depth, (b) displacement/depth, (c)
cable data, not always as successfully as desired. Errors arise
displacement/time for a specific depth interval, and (d) dis-
due to equipment faults, user misuse, or mistakes, as well as
placement vector plots on a site plan, in that order (Figures 18,
19, and 20). Construction activities and instrument field crew recognized system limitations.
observations should be noted on the plots to aid interpretation. Equipment problems include sensor malfunction, wheel
Ground movements are usually progressive and continuous in bearing wear, low batteries, moisture ingress to cable connec-
one direction, although the rate of displacement will generally tions or readout, mechanical shock damage, nicked cables,
not be constant. Erratic displacement/time plots should always improper casing installation, calibration changes inherent in
be carefully investigated. Figure 20 shows steady state creep in manufacturer servicing, and cable stretch or marker movement
a reactivated ancient slide, whereas Figure 21 presents a with use. User errors arise due to reading errors and data
 PC-ILIN PC-SLINI
---------90:S:?A----------- -
IEJCH'WtK £1AfPL..£ trCJ. I 8ENCMWae E:I~ NO. l
DIGITIL T DATA DIOlTILT DATA
H0L£ ~RI 9-1
MOL.£
PA9T ...,._,.,
F'ILE . . .·.- •
1A1911.SN PftESENT F'ILE NNE1A19111.SN P'AST FU•.£ HAl'C1A1Sll.SH PRESENT FILE HN'E:AtSlll . SN
PA91' 8S'T ~· PftESENT SET M..l'IBEfh 7 PAST SET ...... 8E1'1 1 PRESENT SET hUP'IKfh 1
PA9T DATE1
A-
1
8EP'T 2011'3

DtANBE JN READINS
...
PAl:SDfT DATE• CX:T. II /76 PAST OATE1
A-
9EP"T 20173 PftESENT DATE1 OCT. ll/76

DEF\..ECTION - IPCHE9
A•

-900. -i.oo. -300. o. 300.

----- --------·---------·-------·- - 900. -. 7~ .000 . 7SO
·-- - -- - ----·--·------------
I
1.~00

I
I
• I

""
12.0 11 . 0
I
I
"• 22.0
I
I

" 21.0
I
I
"• I
I
l2.0
I
II
3:;.--
11

"
•
42.0
."
II
II
41.0

"• ""
""
.."
52.0
" 51.0

"
,," 6.2.0
". 6.1.0
" "
" "
•I

""
72.0
"" 71.0
II
• ""

GEO-SLOPE PC-SL IN
SERIAL NO. 85019
HOLE -BER: S-1
PAST FILE NAl1E: A:BNl.SI PAST DATE: NOV.lb/82
FILE NAl1E: A:BN2 . SI DATE: HAY.18183

~!Lio

-·
FILE NAl1E: A:BN3.SI
A:llN4.91
FILE NAf'E : A: 8N~. SI
FLLE lllAP1E1 A:BNb.SI
DATE: JUNE.14/83
DATE: Jll Y 11 /93
DATE: AUG . 22/83
DATE: SEPT 27/93
x

A- A+
DEFLECTION - INCHES
.000 1.250 2.500 3.7:SO
-·------+---------·--------+ -- - -
'
I
11.0

l
2 1.0

I
I
l l.O

loI
:s!.o
I

I
bl.O

71.0

81.0

91.0

101.0

FIGURE 18 Graphical computer output: (top) single data set, (bottom) multiple data set.
Green and Mikkelsen 13

instrument performance rating. If opposite walls of the in-

clinometer casing are not quite parallel, if any wheel is in a
coupling, or if depth control is imprecise, the checksum may
vary randomly. Small variations are tolerable.
Ideally, checksums should remain constant with depth within
a given data set for both A and B sensors. If a change in the
zero offset occurs between sets of readings, the checksum will
change, but this does not matter. If a change occurs in the zero
offset during a particular set of readings, either during a
traverse or between opposite traverses due to temperature or
shock-induced changes to the sensor suspension, the result will

\
\A·
Jr--Ii\.-mim lllDI IDUIDlll 0 35
INCLINOMETER CASING GROUTED IN PLACE
e A DIRECTION
OI B DIRECTION
ID 11111111 DISPLACEMENT BETWEEN 31 & 33
0 30 FOOT DEPTH INTERVALS
11 11
Ul 0 25
llr!IUI ltlll II lltMll w
I 0 12 INCHES/YEAR
0
2
~ O.?.O
SUBSTITUTE
2
Q
t> 0 15
~ 11
K INSTRUMENT
USED •

~010 M
;:::~
00
WW
O OS g;g;
00

00
I
INSTALLED 4- 10-76

FIGURE 20 Creep movements on a thin shear

zone In a reactivated ancient slide.
'
\.-1111111

\t ·&

.. 11111111
IL IDI
lllllDUI
::;
INCLINOMETER DHS-1
LATERAL t.AOVEME NT
l0 4 n TO 122 Fl DEPTH

I ID I 1G
NORTHERLY ) M[.t.SUREi.IENTS OISCONTINUEO
DEFORMATION DUE TO BLOCKAGE IN CASING
1111 If IOll•11 Still II lltNll 1111
(: AS1EPP:
FIGUR.E 19 Vector plots or slide movements showing DE.F~MAT1Qp.;
INCLINOMETER DHS - 2

magnitude or movement January 1976-April 1979 and LAT ffiAL ""OVEMENT S20 FT TO 530 n DEPTH

average rate of movement January 1978-April 1979.

transcription errors where manual methods are used; incorrect

probe orientation in the casing grooves; probe depth position
errors due to careless reading techniques, surface casing distur- .... ....
bance, or incorrect matching of readings 180 degrees apart;
FIGURE 21 Deformations measured on deep shear zones
incorrect assumption of casing base fixity; unrecognized settle- on a landslide in Canada.
ment or heave; improper reading techniques; or lack of survey
control when telescopic couplings are used or when large
settlements or heave must be accounted for. This is a for- be errors that can only be eliminated by careful scrutiny and
midable list of problems that can and do occur, and good computerized error detection routines (Figure 22). When read-
project planning, equipment selection, training, data scrutiny, ings are obtained in slightly inclined casing, a correction for
and awareness of the pitfalls are needed to avoid them. changes in the servo-accelerometer azimuth orientation may be
If it is assumed that equipment and user related errors can be required. Slight changes in the sensor due to shock occur from
recognized and controlled, measurement errors also exist be- time to time. These changes show up as irregular rotation of the
cause of the inherent limitations of the best available instru- displacement/depth profile (Figure 22) and can mask shear
ment and casing. The data quality tabulation (Figure 17) is an occurring at discrete zones. The combination of sensor azimuth
essential aid and should always be used. Variation in check- rotation with inclined casing causes systematic errors. Individ-
sums can be quickly scanned, errors detected, and data quality ual correction angles can be applied to the A and B sensors of
assessed. The statistical analysis provides a useful comparative each data set by using computerized routines and shear zone
14 TRANSPORTATION RESEARCH RECORD 1169

movement correctly identified (Figure 21). Although sensor achieved is quite different from the quoted resolution or
zero shifts are a significant problem and must be properly dealt sensitivity of the probe and readout. The inclinometer measures
with during data processing, the scale span or slope of currently relative movements, so it may be more appropriate to discuss
available servo-accelerometers is fairly stable, and errors due to performance in terms of precision, that is, the repeatability with
this source appear to be less serious. They cannot be eliminated which the instrument can determine the position of one end of
by routine procedures and require instrument recalibration by the casing relative to the other. A servo-accelerometer-type
the manufacturer. Some manufacturers provide simple field inclinometer system is capable of a precision of ±0.05 to 0.25
calibration frames, but these appear to be of limited value. in. over 100 ft in vertical and horizontal casings but much less
in inclined casings (7). It is, of course, rare for an accurate
NO CORRECTION WITH ZERO SHIFT CORRECTION
CHANGE DEFLECTION CHANGE DEFLECTION independent check to be possible (12), and estimates of preci-
sion or accuracy achievable are based on a combination of
direct measurement, experience, and reasonableness of the
data.

PLANNING AND EXECUTION OF INCLINOMETER

MONITORING PROGRAMS

As with all geotechnical monitoring programs, inclinometer

NO COAAF.CTION CASING PROFILE 8 AXIS CORF~ECTEO measurements must be properly planned and executed. Too
DEFLECTION A AXIS !CROSS AXIS 1NCL1NATION1 OE:FLECllON A AXIS
often instruments are installed for ill-defined reasons, and data
are collected and filed away, unused (10, 31). Instrumentation
has become fashionable. In contrast, real needs should be
assessed, behavior predicted, and tasks assigned following a
systematic approach. What needs to be measured, the accuracy
required, and if and how this can be achieved should be
considered. Specialist advice on instrumentation system design
r v should be sought where in-house skills are lacking. The engi-
FIGURE 22 Typical zero shift and rotation errors. neer is usually the party most interested in obtaining reliable
monitoring data and should retain as great a control as is
possible. In general, instrumentation should not be left up to the
Errors can arise immediately after installation due to settle-
contractor.
ment of the backfill. Erratic casing movements develop but
Descriptive or performance specifications for instrument
usually cease with time. An extreme type of error can occur
procurement may be used (JO, 31, 32). Too often n single lump
where drilling and grouting upset the stress and thermal regime
sum bid item appears in a specification for supply and installa-
of the surrounding material. Casings installed on a permafrost
tion of a variety of instruments, and the low bidder wins (7).
slope to monitor slope creep have shown extraneous move-
Under these conditions the engineer often ends up with some-
ments due to recovery of thermal equilibrium around the
thing far less than is needed, even though considerable sums of
casing, effects of stratigraphy and settlement, and heave of the
money have been expended. It is preferable for the engineer to
casing (28, 29). High-quality monitoring procedures and
select the make and model number of instrument believed to be
painstaking analysis were employed that enabled definitive
most suitable and to specify with no substitution allowed. Sole
slope velocities of 0.30 cm/yr to be established. Other data
source justification may be required and can be used (33). The
filter techniques have been suggested (30); however, caution is
inclusion of "or equal" leads to excessive emphasis on low
urged in using statistical techniques.
cost so that high-quality instruments are excluded (34, 35).
Similar principles to the foregoing apply to horizontal in-
There is often no good, legally defensible way of excluding
clinometer casings. Where access is available at both ends of
substitute instruments. An entire inclinometer monitoring pro-
the casing, first-order leveling is used to convert the inclinome-
gram may fail if an inferior instrument is selected. This is not
ter survey to a closed traverse, and settlement measurement
an acceptable risk and hence must be avoided. Detailed pro-
accuracy is greatly increased. Inclined casings used to monitor
cedures for minimizing these risks are discussed elsewhere
displacement pose special problems (7) that require considera-
(10, 31).
tion of (a) groove spiral, (b) initial groove roll angle at the
casing collar, and (c) sensor azimuth changes. In addition, if
absolute position is desired, the azimuth of the vertical refer- CONCLUSION
ence plane is determined by surveying the casing collar. Fixed-
in-place inclinometers are subject primarily to sensor drift The inclinometer is widely and successfully used to measure
errors, and slope change/time plots for individual sensors ground movements and is capable of considerable accuracy.
should be scrutinized. Successful projects require good planning, high-quality and
Measurement accuracy achieved with inclinometer systems appropriate instruments, proper installation, diligent reading,
is highly dependent on (a) the equipment selected, (b) the correct data processing, and intelligent scrutiny and interpreta-
installation and monitoring procedures used, and (c) correct tion. Cooperation and understanding between engineers, cli-
data scrutiny and processing. The system accuracy or precision ents, and contractors is essential for success.
Gree11 and M iklulse11 15

ACKNOWLEDGMENTS Testing, AASIITO, Washington, D.C., 1978, T 254-78, Part II,

pp. 941- 950.
16. T. H. Hanna. Field Instrumentation in Geotechnical Engineering.
We are indebted to our colleagues for support. We are also Trans Tech Publications, Clausthal-Zellerfield, Federal Republic
grateful to our clients, through whose projects we have come to of Germany, 1985.
better understand many of the problems of instrument perfor- 17. C. L. Bartholomew, B. C. Murray, and D. L. Goins. Embankment
mance and subtleties of daca analysis. These experiences led to Dam Instrumentation Manual. BW'eau of Reclamation, U.S. De-
partment of lhe Interior, 1987.
the improvements djscussed in !his paper. The substance of this 18. K. Y. Nilsen, E. DiBiagio, and A. Andresen. Norwegian Practice
paper was originally presented at a symposium on field instru- in Insttumenting Dams. Water Power and Dam Construction, Vol.
mentation organized by the Nanyang Technological Institute, 34, No. 11. Nov. 1982, pp. 34-38.
Singapore, in November 1986. 19. D. J. Bailey. Land Movement Monitoring Sysiem. Bulletin of the
Ass.xiation of Engineering Geologists, Vol. 17, No. 4, 1980,
pp. 21 3-221.
20. 0. H. Cornforth. Perfomlllrlce Characteristics of the Slope lndica-
REFERENCES tor Series 2008 Inclinometer. In Proceedings, Symposium on Field
Instrumentation in Geotechnical Engineering, Buttcrwonhs,
L S. D. Wilson. The Application of Soil Mechanics to the Stability London, 1974, pp 126-135.
of Open-Pit Mining. Quarterly of the Colorado School of Mines , 21. S. D. Wilson and P. E. Mikkelsen. Foundation Instrumentation:
Vol. 54, No. 3, pp. 93- 133, 1959. Inclinometers. Report FHWA TS -77-219, FHWA, U.S. Depart-
2. S. D. Wilson. The Use of Slope Measuring Devices to Detennine ment of Transportation, 1977.
Movements in Earth Masses. Jn Special Technical Publication 22. A. D. M. Penman and A. Hussain. Deflection Measurements of
322: Field Testing of Soils, American Society for Testing and lhe Upstream Asphallic Membrane of Marchlyn Dam. Waler
Maierials, Philadelphia, Pa., 1962, pp. 187-198. Power and Dam Construction, Vol. 36, No. 9, Sept. 1984,
3. W. P. Grant, G. Yamane, and R. P. Miller. Design and Performance pp. 33-37.
of Columbia Center Shoring Wall, Seattle, Washington. Proc., 23. D. Shoup and H. Dutro. Battery Powe.r for lnslJUmentation. The
International Co'lference on Tall Buildings, Institution of Engi- Jndicator, Vol. 14, No. 1, Deo. 1985, pp. 19- 22.
neers, Singapore, Oct 1984, pp. 651--661. 24. R. Thomas. Time and Cost Comparisons Between Slope Indicator
4. K. Omori, K. Komatsu, S. Ohya, and H. Nakayama. Measurement Company Inclinometer Indicators. 1'he Indicator, Vol. 14, No. I,
Results of Vertical and Horizontal Deformation of Large Oil Tunk. Dee. 1985, pp. 4-7.
Proc .. 12th Annual Conference on Soi/Mechanics and Foundolion 25. S. D. Wilson and C. W. Hancock. Instrumentation for Movements
Engineering, Japanese Society of Soil Mechanics and Foundation Within Rocldill Dams. lo Special Technical Publication 392: ln-
Engineering, Tokyo, 1977, pp. 579- 582. strumenls and ApparaJus for Soil and Rock Mechanics. American
5. P. E. Mikkelsen. Horizontal Inclinometer Monitoring of Balsam Society for Testing and Materials, Philadelphia, Pa., 1965,
Meadow Dam. The Indicator, Vol. 15, No. 1, November 1986, pp. 115-130.
p. 25. 26. G. E. Green and O. A. Roberts. Remote Monitoring of a Coal
6. R. A. Robinson, E. J. Cording, D. A. Roberu, N. Phienwcja, I . W. Waste Embankment Proceedings, International Symposium on
Mahar, and 1-1. W. Parker. Ground Deformations Ahead of and Field Measurements in Geomechanics, Vol. 2 (K. Kovari, ed.)
Adjacent to a TBM in Sheared Shales in Stillwater Tunnel, Utah. A. A. Balkema, Rotterdam, The Netherlands, 1983, pp. 671-682.
Proc., 1985 Rapid ExcavaJion and Tunneling Conference, New 27. D. A. Barratt. Rapid Data Reduction and Ground Movement Mea-
York, Vol. I, American Institute of Mining, Met;illurgy and Pc- surements. Tunnels and Tunneling, Vol. IO, No. 3, April 1978,
crolcum Engineers and American Society of Ci vii Engineers, 1985, pp. 46-47.
pp. 34-53. 28. N. R. MOJ&enstern. Geotechnical Engineering and Frontier Re-
7. P. E. Mikke.lsen and S. D. Wilson. Field lnsbUmenration: Ac- sources Development Geotechnique, Vol. 31, No. 3, Sept. 1981,
curacy, Performance, Automation, and Procurement In Proceed- pp. 305-365.
ings, International Symposium on Field Measuremenrs in Geo- 29. K. W. Savjgny and N. R . Morgenstern. In Sini Creep Properties in
rru:chanics, Vol. 1 (K. Kovari, ed.), A. A. Balkema, Rotierdam, lee-Rich Permafrosr Soil Canadian Geotechnical Journal, Vol.
The Netherlands, 1983, pp. 251- 272. 23, No. 4, Nov. 1986, pp. 504-514.
8. W. Tesch. AutomaJed Borehole SUTVeying System for Steeply In- 30. E. C. Kalkanl. Filtering Probe Inclinometer Data to Identify
clined Cased Boreholes. Rcpon of Investigations 8156. U.S. Bu- Characteristics of Slope Movement Rock MecJ1anics, Vol. 13, No.
reau of Mines, U.S. Department of lhe Interior, 1976. 2, 1980, pp. 57-69.
9. G. W. Uttecht and J. P. deWardt. Survey Accuracy is Improved by 31. J. Dunoiclif[. NCHRP Synthesis of Highway Practice 89: Geo-
a New, Small OD Gyro. World Oil, Vol. 196, No. 4, March 1983, technical Instrumentation for Monitoring Field Performance.
pp. 61-64. TRB, National Research Council, Washinglon, D.C., 1982, 46 pp.
10. J. Dunn.icliff. Geotechnical Instrumentation for Monitoring Field 32. M. F. Kennard. Field Instrumentation Within a Civil Engineering
Performonce. John Wiley, New York, 1988. Contract. Proceedings, Symposium on Field Instrumentation in
1 I. I. P. Gould and J. Ounnicliff. Accuracy of Field Measurements. Geotechnical En.g ineering, Butterworths , London, 1974,
Proc., 4th Panamerican Conference on Soil Mechanics and Foun- pp. 220-228.
dation Engineering. San Juan, Puerto Rico, Vol. I. 1971, 33. R. S. Perlman. Sole Source Contracts , Basic Principles and
pp. 313-366. Guidelines. Briefing Papers 83 -7 , Vol. 6, Briefing Papers Collec-
12. G. E. Green. Principles and Performance of Two Inclinome ters for tion 187. Federal Publications, 1.nc., Washington, D.C .. July 1985,
Measuring Ground Movements. In Proceedings, Symposium on pp. 187-222.
Field Instrumentation in Geotechnical Engineering, Butte rwonhs, 34. P. E. Mikkelsen. Discussion on "Piezometcrs in Eart h Dam Imper-
London, 1974, pp. 166-179. vious Sections,'' by J. L. Sherard. Journal of the Geotechm'cal
13. JSRM. Suggesied Methods for Monitoring Rock Movements Engineering Division, ASCE, Vol. 108, No. GT8, Aug. 1982,
Using Inclinometers and Tiltmeters. Rock Mechanics, Vol. 10, pp. 1095- 1098.
Nos. 1 and 2, 1977, pp. 81- 106. 35. G. E. Green. Discussio n on "Piezometers in Earth Dam Imper-
14. S. D. Wilson and P. E. Mikkelsen. Field Instrumentation. In Spe- vious Sections,'' by J. L. Sherard. Journal of the Geolechnical
cial Report 176: landslides: Analysis and Control. TRB, National Engineering Division, ASCE, Vol. 108, No. GTl l, Nov. 1982,
Research Council, Washington, D.C., 1978, pp. 112- 138. pp. 1526- 1528.
15. Standard Method for Installing, Monitoring, and Processing Data
of the Traveling Type Slope lnclinomete.r. In Standard Specifica- Publication of this paper sponsored fry Committee on Soils and Rock
tions for 1'ransportaJion MaJerials and Methods of Sampling and Instrume11tation.