Determination of optimal separation well distance for single borehole ATES systems, in the Netherlands, implementing an an axi-symmetric numerical model

Abstract

Aquifer thermal energy storage (ATES) is a growing technology in the Netherlands. There are two kinds of ATES configurations the doublet and the single borehole ATES (SB-ATES) layout. The limited subsurface space in combination with the lower construction cost and the lower performance of the SB-ATES lead to the need to optimize their design. This master thesis focuses on gaining a better insight in the processes that occur around this configuration. Specifically, how anisotropy influences the efficiency of the design and what the optimal distance between filter screens is, in order to limit the interference between the screens, as it has a negative impact on system performance. To meet these two objectives, an axisymmetric numerical model was developed in a MATLAB environment, using MODFLOW and MT3DMS or SEAWAT groundwater flow and transport simulators. The simulation of heat advection was conducted applying the finite different method (FD method), it was the only method compatible with axial symmetric models that produced consistent results. As the FD method is subjected to numerical dispersion, three different grid resolutions were tested that were the 0.25, 0.50 and 1.00 m, respectively. The finest grid was decided to be used in the elaboration, as it gave the most accurate results compared to larger thickness width grid cells. Capacity test and borehole profile data were used to calibrate the overall vertical anisotropy of the case-studies. The capacity test allowed the calibration of one hydraulic parameter, for which the overall vertical anisotropy was chosen. The Kozeny-Carmen equation was used to calculate the hydraulic conductivity of each soil layer. This overall vertical anisotropy even though was estimated roughly, can be used to determine the presence of overlooked clay layers during the drilling of the borehole (anisotropy 10). A sensitivity analysis was applied to estimate the optimum distance between the filter screens. Three different types of SB-ATES, called GT15xx, GT20xx and GT25xx, were examined separately. The numbers indicate the installed pump capacity in m3/h while their mean and representative screen lengths are 5, 7 and 10m respectively. For the sensitivity analysis, three discharge fractions, Qfrac, were tested with values 0,25, 0.50 and 1.00 and three anisotropy values of 2, 5 and 10, these values are representative for sandy soils. The simulation time was 5 years, which was sufficient to the recommended efficiencies. The results showed that the maximum efficiency is practical independent of system type. The optimum separation distance for an anisotropy of 2 is respectively, 25, 30 and 35 m for Qfrac 0.25, 0.50 and 1.00. The evaluation of the sensitivity analysis was conducted, using real-scale case-studies and taking into account the distribution of conductivities along the layers. The available 18 case-studies were examined in terms of efficiency; it was found that a thin resistance layer between the screens, like a clay or peat layer, influences positively the performance of SB-ATES systems. On the other hands, when there is a high conductive zone between the screens, the efficiency drops. Finally, it was found that temperature induced differences in density and viscosity have a negligible effect on SB-ATES systems, at least with injection temperature differences between warm and cold wells up to 120C.