Short-term monitoring long-term prediction of energy use in commercial and institutional buildings: The SMLP method.

摘要

The Short-term Monitoring Long-term Prediction (SMLP) method is a new method for baselining, predicting and evaluating the energy performance of commercial and institutional buildings when only short-term data are available. The new method overcomes complexities of existing methods, and provides an accurate baseline model from a two-week period of hourly energy data. The SMLP method is based on sound statistical procedures to obtain reliable results within acceptable margins of uncertainty as is the case with any prediction method. In developing the SMLP method, the level of accuracy of the predictions which result from using different time periods for in-situ monitoring was ascertained; the findings can be used to find a suitable trade off between accuracy and cost in time and effort that would be incurred by considering any time period for conducting the monitoring.; In the baselining of building energy use, energy analysts in general show reluctance in using laborious and complicated methods. Moreover, whenever long-term high frequency monitored data are not available, the need for a simple yet accurate method for energy predictions based on short-term monitoring becomes obvious. The new method goes beyond the previous work in terms of accuracy of long-term prediction obtained with models developed from short-term monitored data, optimum length of the monitoring period, optimum time of the monitoring period, the exclusive and necessary variables to monitor, and the most appropriate modeling technique and its ease of use. The SMLP method requires the monitoring of the building energy consumption and weather conditions for a short period of time (two-week period), while most inverse methods need long-term monitored data (yearlong data sets). The measured data along with established intensities, load shapes and schedules for occupancy and building internal loads enable modeling the building energy performance for baselining applications and long term predictions. Moreover, the new method promises to reduce the time and effort that should be spent, had the comprehensive calibrated simulations (DOE-2, BLAST) been used for baselining.; The new method can be used by energy analysts, utility companies, ESCO's, researchers, academics, and students who will profit from its capabilities, and yet its simplicity.