The OMNIBUS project aims to develop a new generation of ultrasonic
monitoring technology that can be used to monitor the rock barrier at
potential and operational radioactive waste disposal sites, taking existing
technology and addressing some of its current limitations. These include
the need for a robust and versatile tool, the need to integrate hardware
and software to maximise data accessibility and the requirement to be able
to interpret the results of processed velocity and AE data in terms of
useful engineering parameters. It is anticipated that this technology would
be used by a wide variety of organisations charged with evaluating,
selecting and operating deep geological repositories for nuclear waste, as
well as having applications in other fields of civil engineering.
The progress achieved at the end of the
second year is summarised below.
Hardware developments

A complete data-acquisition tool has been developed and has been
successfully tested at a soft-rock site. The hardware uses lower frequency
ultrasonic sensors tuned to the highly attenuating transmission
characteristics of argillaceous rocks. Switching electronics has been
developed that allows each sensor in the array to act as both a transmitter
and a receiver during ultrasonic surveys (right figure). The electronics is
a vast improvement over available instrumentation. These components have
been interfaced to a 16-channel high-frequency data acquisition system. The
system operates at up to 10MHz with 16-bit resolution and uses eight
two-channel PC cards installed in a purpose built mainframe chassis. The
down-hole equipment has been specifically designed for permanent
installation in an argillaceous rock mass and the complete system has been
successfully tested at Tressange Iron Mine in France.
Software
developments
Software functions have been written into the InSite Seismic
Processor that integrates the developed hardware with data
acquisition, management, processing and visualisation (figure
below). The software provides configuration parameters to the
data acquisition hardware, controls the switching electronics
for the ultrasonic surveys, and automatically captures full
waveform information. The data is then passed through real-time
processing functions.
In
situ experiments
In order to test the new down-hole acquisition technology and
associated hardware, a field test has been undertaken at Tressange
Mine in eastern France. This site has provided the opportunity to
test the system in a lithology comparable with that observed at
several of the soft-rock sites proposed for waste storage (such as
ANDRA’s Bure en Meuse site). The experiment involved
installing an array of sensors in boreholes around a small rock
volume (figure below), making measurements to determine the
background properties of the rock and then perturbing this rock by
using an expanding grout at the base of several narrow, closely
spaced boreholes.
Laboratory
experiments
Laboratory experiments have involved a series of controlled
tests on argillaceous rock samples from the
Bure en Meuse/Haute
Marne
site. These are the only ones of their kind in Bure rock which
combined acoustic emission (AE), mechanical measurements and
ultrasonic velocity surveys. They have demonstrated that the
damaging of this rock can generate AEs. Measured parameters
have been used directly in developing numerical models.
Numerical
modelling experiments
Numerical modelling is being used to study wave propagation
through rock and the effects of various physical attributes
(e.g. fracture density, fracture size/distribution and fluid
content) on the propagation of ultrasonic waves. The study
makes use of the WAVE finite difference modelling approach.
This allows amplitude and phase spectra to be derived for a
simulated cracked sample (right figure) that can then be
related to results obtained from measurements in laboratory and
in situ experiments. A campaign of numerical modelling
experiments has been performed to assess the sensitivity of
ultrasonic waves passing through rock to variations in applied
stress, crack population and fluid content in different
experimental scenarios. In total, results for 667 models have
been obtained totalling more than a year’s worth of
computer time, with different models run in parallel on a
super-cluster at. Dedicated processing and visualisation
software will now be written that allows users to correlate
observed laboratory and in situ ultrasonic survey results with
the different modelled scenarios