Pumping power prediction in low temperature district heating networks

Abstract

ABSTRACT A low temperature network providing heat for domestic buildings and cold for a large data centre was modelled in the software Polysun based on simplified loads. The emphasis was set on the electric energy consumption of the decentralized circulation pumps. For this reason, the pump and hydraulic resistance model of the software Polysun was adapted in order to reproduce the characteristics of modulating high efficiency pumps. Energy consumption of the decentralized circulation pumps of less than 1% of the transferred energy was predicted from dynamic simulation with the new models. In order to challenge these predictions, the first four months of operation of a network in Zurich were monitored and analysed. Launched in December 2014, the network is equipped with an extended monitoring system, which records amongst others the network temperature evolution, the heat pump performance and the pumping energy consumption. The main heat pump providing heat to over 400 households at a mean temperature of 68°C reached an overall coefficient of performance (COP) of 3.7 in initial winter operation. This value corresponds to a remarkable overall efficiency factor (including primary and secondary circulation pumps) of 65%. Due to intense usage in winter months, but also slightly higher pressure drops than predicted, the pumping energy reached 1.6 % of the low temperature energy delivered to the heat pump and 1.1% of the cooling energy supplied to a data centre. Even though not all modelled consumers started operation during the monitoring period, the measured data was used to validate dynamic simulations and pump models. By introducing the measured loads and the effective hydraulic resistance to the simulation model, the measured pumping energy fractions could be closely reproduced with the simulation model.