Application of univariate search methods to the determination of HVAC plant capacity

For nearly a century, the sizing of building heating and cooling systems has been carried out by calculation procedures based on matching the output of the plant to a peak load design figure. The design conditions are normally based on historical records of weather data and give a probabilistic 'worst case' scenario. These calculations, although often regarded as a routine stage in the design process, are critical in the sense that they affect the overall efficiency and energy consumption of the plant, the installed cost and the space occupied within the building. It is argued in this paper that advances in the simulation of the dynamic thermal performance of buildings and HVAC plant afford an opportunity to change the basis of F-mnc plant sizing procedures from the load calculation to an assessment of the ability of the plant to meet an installed performance specification. Two potential performance specifications are identified in this paper and an example zone/HVAC plant model is used to illustrate how performance-based sizing might work. Simulations have been run for conditions in the UK and in Hong Kong. Plant sizes determined by simulation have so far been smaller than those given by the conventional sizing calculation procedures.