Field layout of a tiltmeter array to monitor micro-deformation induced by pumping through a horizontal collector well
H.C. Chen, H.-J. Kümpel and C.M. Krawczyk
Journal name: Near Surface Geophysics
Issue: Vol 8, No 4, August 2010 pp. 321 - 330
Info: Article, PDF ( 8.41Mb )
Price: € 30
Pump-induced tilt deformation in the subsurface has repeatedly been observed as a hydromechanical, poroelastic response to changes in pump rates in the vicinity of fluid-producing wells. Within the Fuhrberg waterworks area, north of Hannover, Germany, groundwater is withdrawn from 22 m depth through a system of horizontal wells out of a sandy, unconfined aquifer. To measure the pump-induced tilt, a specific field layout was designed and a unique data set recorded. The monitoring experiment was set in place by installing 12 borehole tiltmeters at depths of approximately 2 m below surface. We monitored micro-deformation signals in situ for up to two years, so that long-term, short-term and event-type tilt signals could be identified. During the study, the pump rate of the producing well system varied from 50–400 m3/h in an irregular but controlled manner. The field layout of the tiltmeter array allowed to trace both the overall signature of the induced micro-deformations around the well system as well as the tilt responses at individual tiltmeter positions. For a variation of the pump rate by 100 m3/h we found the average amplitude of the induced signal to be around 0.4 μrad. There is a general trend for the induced tilt signals to build up in a concentric way around the central pit. Nevertheless, there are significant deviations from the expected tilt response, namely of order ±10° in azimuth or ±0.2 μrad in amplitude per 100 m,sup>3/h change in pump rate for roughly half of all the 42 tiltmeter sites monitored. The deviating tilt azimuths are believed to reflect the presence of heterogeneities in the glacio-fluviatile sediments at this location. While influence of the local topography is of minor importance, as well as noise from meteorological and hydrological causes, effects due to the nearby vegetation can mask the pump-induced tilt signal by up to 30%. Here, the distance of the observation points away from larger trees is essential, rather than the overall amount of vegetation.