Comparing the energy use of centralized and decentralized pumping systems

The research article 'Experimental and numerical investigation of energy saving potential of centralized and decentralized pumping systems' has been published in Elsevier journal Applied Energy (Volume 251, 1 October 2019).


In energy distribution systems, thermal energy is usually transferred by a heat carrier fluid via pumps. Improper design and unreasonable control of pumping systems result in inefficient operation which accounts for a significant part of electricity consumption in the industry. The need to save energy has been sharpened the focus on improving energy efficiency in pumping systems. The application of a decentralized pumping system with the variable-frequency drive can be considered a technological improvement that has potential in saving energy compared to the conventional centralized pumping system. In this paper, a reduced-scale experimental apparatus and computational fluid dynamic model are used to investigate the energy saving potential of decentralized and centralized pumping systems. The energy-saving potential of decentralized configuration and two types of centralized configurations are then compared.

The results showed that the decentralized pumping system consumes less power than centralized pumping systems under the same conditions. When the flow rate is reduced to 80%, the power consumption of the decentralized configuration decreases by 47% while the consumption for a centralized configuration with constant pressure control decreases by only 19%. The decentralized pumping system can offer higher energy-saving potential under variable flow rate conditions, which is expected to extend to other fluid delivery systems for improving efficiency. Moreover, the computational fluid dynamic simulation results correspond well with experimental results. The maximum discrepancies of the developed model for prediction of gauge pressure and system total pressure loss are 7.2% and 9% respectively, which confirms the accuracy and applicability of this model.

Access the complete article on ScienceDirect.