Geothermal distribution network modeled as Heat Exchanger Network to be optimized with Mixed Integer Nonlinear Programming

Abstract

Geothermal energy is commonly harvested at either shallower depth (below 150ft/45.72m) for residential purposes (with ground source heat pumps), or deeper depths (beyond 8000ft/2.43 km) for Enhanced Geothermal Systems. The in-between depths are rarely visited due to high drilling costs, and the water harvested being unable to power turbines. Recent studies powered by the data released by the National Geothermal Data System (NGDS) opened a new opportunity of harvesting the geothermal potential in post-production oil/gas boreholes in Pennsylvania. We are interested therefore in whether it is feasible to connect the different heat sources with different temperature availabilities to distribute to spatially scattered end-users. Presented in this paper is a project that generates heat exchanger network configurations through mixed nonlinear programming (MINLP) problem formulation and optimization in Python. With the case intentionally simplified, the computational costs of the optimization was found to be marginal.