 |
| Titel |
Analysing Thermal Response Test Data Affected by Groundwater Flow and Surface Temperature Change |
| VerfasserIn |
Massimo Verdoya, Gianmario Imitazione, Paolo Chiozzi, Marco Orsi, Egidio Armadillo |
| Konferenz |
EGU General Assembly 2014
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 16 (2014) |
| Datensatznummer |
250091412
|
| Publikation (Nr.) |
 EGU/EGU2014-5704.pdf |
|
|
|
|
|
| Zusammenfassung |
| Tests that record the underground temperature variation due to a constant heat injected into a
borehole (or extracted from it) by means of a carrier fluid are routinely performed to infer
subsurface thermal conductivity and borehole thermal resistance, which are needed to size
geothermal heat pump systems. The most popular model to analyse temperature-time
curves obtained from these tests is the infinite line source (ILS). This model gives
appropriate estimations of thermal parameters only if particular hydro-geological
conditions are fulfilled. Several flaws can however affect data interpretation with
ILS, which is based on strong assumptions like those of a purely conductive heat
transfer regime in a homogeneous medium, no vertical heat flow and infinite length of
the borehole. Other drawbacks can arise from the difficulty in the proper thermal
insulation of the test equipment, and consequently with oscillations of the carrier fluid
temperature due to surface temperature changes. In this paper, we focused on the
treatment of thermal response test data when both advection and periodic changes
of surface temperature occur. We used a moving line source model to simulate
temperature-time signals under different hypothesis of Darcy velocity and thermal properties.
A random noise was added to the signal in order to mimic high frequency disturbances,
possibly caused by equipment operating conditions and/or geological variability.
The subsurface thermal conductivity, the Darcy velocity and the borehole thermal
resistance were inferred by minimising the root mean square error between the
synthetic dataset and the theoretical model. The optimisation was carried out with the
Nelder-Mead algorithm, and thermal and hydraulic properties were determined by iterative
reprocessing according to a trial-and-error procedure. The inferred thermal and hydraulic
parameters are well consistent with the “a priory” values, and the presence of noise in the
synthetic data does not produce instability. The same optimisation procedure was
also applied to interpret the synthetic signal with the ILS model. In case of Darcy
velocity of the order of 10-6 m s-1, ILS largely overestimates thermal conductivity.
The optimisation analysis was then applied to real thermal response tests carried
out in boreholes drilled in sedimentary formation aquifers, whose volume heat
capacity was assumed to be known. These data showed a periodic offset in the
recorded temperature-time series. A Fourier analysis allowed the recognition of
harmonic contributions with period of about 12 and 24 hours. These oscillations
appeared to be coherent with the data of air temperature change recorded at the
test sites. The temperature-time curves were then filtered to remove the disturbing
spectral components associated to a non-optimal thermostatic behaviour of the
apparatus. This produced reliable estimates of thermal conductivity, Darcy velocity
and borehole thermal resistance. The magnitude of the inferred groundwater flow
was also checked by means of an independent method based on the analysis of
temperature-depth logs recorded prior to the thermal tests, under thermal equilibrium
conditions. |
|
|
|
|
|
|