Optimization of the operating conditions of district heating networks is usually performed limiting the analysis to the primary energy related with heat production. An additional aspect that should be considered is the role played by the pumping system. Pumping may contribute to about 10% of the total primary energy consumption, especially in large networks or when small temperature levels are applied. Furthermore, the increasing share of waste heat or renewable energy sources from distributed producers requires a flexible and efficient pumping system. A further aspect which pumping strategy should face is system operation when malfunctions in the plants, pumps or pipes occur. Optimization of the pumping system requires the use of detailed simulation tools, which may need significant computational resources, especially in the case of large networks. A reduced model, based on Proper Orthogonal Decomposition combined with Radial basis functions (POD-RBF model) is proposed in this paper. This approach allows maintaining high level of accuracy despite reductions of more than 80% in the computational time. This make the approach effective tool for control strategy operations. An application to a large district heating network shows that reductions of about 20% in the pumping request and effective management of failures are possible. (C) 2016 Elsevier Ltd. All rights reserved.