38
Geotechnical News • March 2018
GEOTECHNICAL INSTRUMENTATION NEWS
stability. The Marsh funnel viscosity
of a stable and pumpable grout should
be around 50 s (Chapuis et al. 1984;
Contreras et al. 2007; Marefat et al.
2018). Since bentonite properties and
field conditions vary between projects,
it is usually not possible to specify a
definitive grout recipe with an exact
bentonite proportion.
The updated paper will be available
on the Geotechnical Testing Journal
website in the coming months (https://
/
JOURNALS/GEOTECH/). You can
also email us for a copy (francois.
). For an update on
timing, please contact François.
References
Chapuis, R. P., Paré, J.-J., Loiselle, A.
A. 1984. Laboratory tests results
on self-hardening grouts for flex-
ible cutoffs. Canadian Geotechni-
cal Journal, 21:185-191.
Contreras, I. A., Grosser, A. T., Ver-
Strate, R. H. 2007. The Use of the
Fully-grouted Method for Piezom-
eter Installation. Proceedings of
the 7th International Symposium
on Field Measurements in Geome-
chanics (FMGM), Boston.
Marefat V., Duhaime F., Chapuis R. P.,
Le Borgne, V. 2017. Fully grouted
piezometers in a soft Champlain
clay deposit – Part I: Piezometer
installation. Geotechnical News,
35(3): 35-38.
Marefat, V., Duhaime, F., Chapuis,
R. P., Le Borgne, V. 2018.
Performance of fully grouted
piezometers under transient flow
conditions: field study and numeri-
cal results. Geotechnical Testing
Journal, paper accepted with revi-
sion.
Francois Duhaime
Extract / Transform / Load (ETL) processes for
instrumentation data transfer
Garrett Bayrd
Introduction
Geotechnical instruments have
changed slowly over the past few
decades. With some notable excep-
tions, a geotechnical engineer practic-
ing in 2001 would be familiar with the
majority of instruments in 2018. For
example, the manufacturing process
used to make an inclinometer sensor
may have changed, but the principle
of the inclinometer is itself much the
same.
Conversely, the way we are able to
acquire and process data has dramati-
cally changed over the past decade in
several significant ways:
• Telemetry options are more plenti-
ful, cheaper, and smaller.
• Dataloggers are less expensive,
smaller, and many have built-in
telemetry options.
• Less expensive processing and data
storage tools allow for more data
storage and more diverse ways of
examining the data.
• Easier to use programming lan-
guages like Python process and
display large quantities of data.
These changes, combined with a
modern culture that emphasizes instant
availability and access to information,
have increased the number of monitor-
ing projects that require near-real time
data availability. Establishing this data
availability can be more difficult than
some realize when bidding on these
projects, as it requires systems for data
telemetry, Extract / Transform / Load
(ETL) processes, data storage, data
processing, and data visualization.
As a result, there have been more
budgetary and conceptual mistakes in
the design of these data acquisition
systems – frequently with regard to
ETL programs. It seems that many
project managers, cost estimators, and
engineers feel that if they establish a
connection between a computer and a
datalogger system, the data will easily
flow back to them and be displayed in
a timely manner. Instead, it takes care-
ful planning and organization to create
effective data acquisition systems. As
is the case with any instrumentation
project, designing data acquisition
processes (and designing an applicable
monitoring system) requires an under-
standing of the geotechnical setting
that is to be monitored and the purpose
of the system.
ETL processes
move the data from
an instrument or
datalogger to a
computer or server.
What is ETL?
Establishing communications with
a datalogger does not move the data
around. Extract / Transform / Load
(ETL) processes move the data from
an instrument or datalogger to a com-
puter or server. ETL processes can be
