2002 Volume 24 Issue 1 Pages 29-46
In European and North American countries, Ground-Coupled Heat Pump (GCHP) systems have been widely used for space heating and cooling purposes of houses and buildings. In Japan, however, the rate of increase in the number of GCHP system installation is still slow mainly due to the high drilling cost of ground heat exchange wells. The cost-effective design of GCHP systems, therefore, will be of crucial importance for promoting the application of GCHP systems in Japan. In this paper, an approach in optimizing the well designs and operating conditions of heat exchange wells in GCHP systems was proposed. In the optimization process, the net present value of expenditures for the constructions and operations of GCHP systems were minimized as the profitbased objective function. As optimization methods, quasi-Newton method, the Polytope method and Genetic Algorithms (GA) were applied and their efficiency and stability were investigated. Among the optimization methods, GA showed their advantage in optimizing discrete variables, i.e., well numbers. GA also showed the high efficiency and stability in simultaneously optimizing well designs and GCHP operating conditions. The Polytope method optimized well depth and GCHP operating conditions efficiently though the method could not optimize well numbers. The performances of quasi-Newton method were unstable since the method required appropriate initial guess for convergence to the optimum solution.