Actual Versus Predicted Well Cooling Kinetics: Implications on Reservoir Modelling Approaches

Abstract

Predicting production well cooling kinetics induced by (re)injection of heat depleted brines into the source reservoir is a vital segment of any geothermal district heating (GDH) doublet design and operation. Actually the exercise is often biased by oversimplifying the reservoir structure and the heat transfer process assumed single layered and convective alone (the so called breakthrough time) respectively. It can be somewhat misleading while assessing reservoir life and designing well architectures. The present paper favours instead an approach reflecting more closely actual multi-layered reservoir structures a distinctive attribute of the most low enthalpy sedimentary environments. It accounts for the heat stored in the confining hydraulically impervious but thermally conductive, aquitards stacked in a single heat storage volume surrounded on the top and bottom by symmetric stacked pervious layers. This configuration achieves a “3D sandwich” reservoir equivalent which not only better matched monitored temperature histories but also allowed to validate a modified convective thermal breakthrough formula. The approach proved rewarding while modelling thin layered foliated structures, fast moving contrasted injection cycles and, last but not least, sub-horizontal well doublet designs.