美國冷凍空調協會ASHRAE在Guideline 14 中曾提出一冰水機耗能經驗式做為預測冰水機耗能模式,由於冰水機組耗能預測結果對於建築物能源分析的準確性影響很大,且自該模式提出至今十幾年來,陸續有學者提出其他形式耗能測模式,使得該模式的適用性有待驗證。此外,隨著近年來VPF (variable- primary- flow systems) 一次側變冰水量系統應用案例之逐漸增多,尋找可供應用於預測變冰水量冰水機組在不同運轉參數下耗能經驗模式之研究值得吾人作進一步探討。 本文主要研究目的在於針對目前文獻中包含ASHRAE Guideline 14、DOE-2 與Modified DOE-2在內之九種具代表性之冰水機耗能經驗模式,利用八組廠商提供變冰水量操作之實機測試數據進行分析與評估比較。除了檢討ASHRAE Guideline 14 所提模式在預測變冰水量冰水機耗能之適用性外,更進一步探討得到可供應用於預測變冰水量冰水機組在不同操作參數下之最佳耗能經驗模式,俾以作為建築節能分析應用的依據。 綜合本文針對變冰水量冰水機耗能經驗模式預測準確度之研究結果發現:以Modified DOE-2 (Model 9)與Model 3的預測準確度最高,前者平均CVRMSE = 0.403 %,而後者平均CVRMSE = 0.476 %。二者差距甚微,但是Modified DOE-2 在數據分析上比Model 3 較為煩瑣。至於ASHRAE Guideline 14 (Model 6)是所有模式中最差的,平均CVRMSE = 14.298 %。
The ASHRAE Guideline-14 has proposed an empirical model for predicting energy consumption of liquid chillers. Because the prediction of the result of the chiller consume of energy greatly affects the accuracy of the energy analysis of buildings; and since this particular model has been put out for more than ten years; gradually there has been many scholars proposing other types of the consume of energy models; the applicability of the proposed model is yet to be tested. In addition, cases of applications of VPF (variable- primary- flow systems) has gradually increased in recent years, the study of searching for a suitable empirically based model to predict power consumption for VPF chillers under various operating parameters is greatly considerable for this research. The purpose of this research is aimed at nine typically empirical models, including ASHRAE Guideline 14、DOE-2 and Modified DOE-2, etc., using the eight sets of in-field testing data for VPF chillers to do feasibility evaluation and accuracy comparison. Besides the examination of the applicability of performance prediction models for VPF chillers proposed by ASHRAE Guideline 14, and further discuss to conclude the best empirical model for predicting performance of VPF chillers under various operating parameters, in order to use as a basis for application and analysis of energy saving for buildings. To sum up this research of accuracy of predictions for variable chiller water flowrate of consume of energy empirical model for chiller, one can conclude that: the accuracy of the prediction is the highest for Modified DOE-2 (Model 9) and Model 3, the former having an average CVRMSE = 0.403 %, and the latter having an average CVRMSE = 0.476 %. The amount of differences is slim, but the data analysis process for Modified DOE-2 is more complicated than that of Model 3. As for ASHRAE Guideline 14 (Model 6), it is the worst among all models, having an average CVRMSE = 14.298 %.