• 설비공학논문집
• /
• 제26권6호
• /
• pp.263-268
• /
• 2014

Eskilson's g-function, a well-known geothermal heat response factor, is widely used for sizing of the ground heat exchangers. Unlike the Eskilson's original model that uses common temperature boundaries for all boreholes and along the borehole height, an analytical-solution-based g-function uses a uniform heat transfer rate over the height with variable heat transfer rates for respective boreholes. To evaluate the impact of such a boundary difference on g-function and the design length, a simple case study was carried out on the cooling-dominant commercial buildings. The results show that the design lengths given by the boundary of uniform heat transfer rates are longer than those given by Eskilson's boundary for all cases tested. The difference in length is more important when the bore field is composed of more boreholes with shorter length of each borehole.

• 설비공학논문집
• /
• 제28권10호
• /
• pp.402-407
• /
• 2016

Ground-coupled heat pump systems have been widely used, as they are regarded as a renewable energy source and ensure a high annual efficiency. Among the system components, borehole heat exchangers (BHE) play an important role in decreasing the entering water temperature (EWT) to heat pumps in the cooling season, and consequently improve the COP. The optimal sizing of the BHEs is crucial for a successful project. Other than the existing sizing methods, a simulation-based design tool is more applicable for modern complex geothermal systems, and it may also be useful since design and engineering works operate on the same platform. A simulation-based sizing method is proposed in this study using the well-known Duct STorage (DST) model in Trnsys. TRNOPT, the Trnsys optimization tool, is used to search for an optimal value of the length of BHEs under given ground loads and ground properties. The result shows that a maximum EWT of BHEs during a design period (10 years) successfully approaches the design EWT while providing an optimal BHE length. Compared to the existing design tool, very similar lengths are calculated by both methods with a small error of 1.07%.