ase study of real-time microseismic monitoring of solution
mines prone to ground collapses
NERIS performs a routine application of automatic near -to-real time microseismic monitoring of
solution-mining cavities to detect and locate microseismic swarms, and identify potential areas of sud-
den surface collapse. It has been successfully integrated into an early-warning system implemented on
an active brine field, where the geometry of the solution cavities is unknown, bringing major outcomes in
terms of safety achievements, mine planning and land use.
ontext. Solution extraction of soluble minerals such as salt
entails the creation of underground cavities, injecting fresh wa-
ter into the salt to dissolve it, and then recovering brine. Depending on
the geological conditions, a number of these cavities can result in local-
ized land-surface subsidence and sinkhole formation. As these phenom-
ena can happen very rapidly, they present a pervasive threat to miners
and a high potential for environmental damages.
ite description. In this active brine field site located out
of France, the salt deposit occurs in heterogeneously distributed
lenses at varying depth, over an area of several km2. The site is
equipped with a surface microseismic network that is technically man-
aged by the mine operator while INERIS provides near-to-real time data
processing capabilities as well as early-warning services.
Scheme of producton procedure and
potental sinkhole creaton
olution. The data flow from the field monitoring network is transferred to INERIS monitor-
ing Center, in France, according to a secure communication protocol that allows also remote
control of the system. The data is then managed through the cloud monitoring infrastructure e.cenaris.
First, an automatic event detection and classification approach has been implemented to distinguish
microseismic events from noise related to exploitation, such as pumping, trucks, etc., thunder/
lightening and regional tectonic earthquakes.
Second, a fully automatic amplitude based location approach has been implemented to deal with low
signal-to-noise ratios (resulting from strong local attenuation effects, low energetic events and strong
surface seismic noise related to local exploitation).
Third, an estimation of the seismic source sizes is integrated in the data pro-
cessing tool by calculating the seismic source with the seismic moment, Mo.
esults. Several local sinkhole events
could be anticipated within a few weeks
or days before collapse was observed at surface,
allowing the mine staff to take preventive safety
decisions. It is likely, that these events have been
provoked by increased production rate, what
generally increased the potential to build under-
ground cavities, sufficiently large in size to propa-
gate to surface.
Example of sinkhole formatons (stars) and associated seismic actvity
icroseismicity is mainly related to caving process-
es in the overburden and can be used as an indi-
rect measure of the total cavity size. Accordingly, seis-
mic source volume (cumulative Mo) is generally pro-
portional to the total extracted salt mass. In addition,
the spatial distribution of the seismic source volume
and fresh water injection volumes are in very good
agreement. This good spatial fit indicates that solution
and caving processes proceed in a controlled manner,
i.e. that solution processes are focused on those injec-
tion wells where most fresh water is injected.
Cumulatve seismic moment Mo (gray bars) shown as a functon
of tme compared to the mass of extracted salt
learned. In this
complex environment com-
posed of operating and aban-
doned wells and cavities of
unknown geometries, micro-
seismic monitoring has be-
come a new tool to identify
possible situations of uncon-
Spatal distributon of seismicity (a) and actve fresh water injecton wells (b)
trolled dissolution.
he early warning system turns to be also rapidly adaptable to every changes in the microseismic
sensors layout on site, which evolves with the extension of the production area. The system simply
has to be recalibrated performing some surface rock falls to adjust the data management and pro-
cessing parameters to the new sensors layout. In this way, the system continuously keeps its capacity to
provide timey and meaningful information to reduce the possibility of harm or loss on site.