A new strategy based on power demand forecasting to the management of multi-energy district boilers equipped with hot water tanks

作者：Labidi Mouchira; Eynard Julien; Faugeroux Olivier; Grieu Stephane^*

来源：Applied Thermal Engineering, 2017, 113: 1366-1380.

DOI：10.1016/j.applthermaleng.2016.11.151

• Multi-energy district boiler
• Thermal energy storage
• Predictive strategy
• MPC
• Power demand forecasting

As part of the second phase of the research project "OptiEnR", aiming to improve the operation of multi energy district boilers by adding optimally-designed and managed hot water tanks to the plants, a new and generalized predictive strategy is proposed. This phase of the project was first dedicated to developing both a design approach, based on a parametric analysis, and a sequential management strategy. Basically, the excess thermal energy produced by the wood boiler unit(s) during low-demand periods can then be stored and released when demand is high, instead of engaging a gas boiler unit. However, this preliminary study has allowed us to point out possible improvements in the management of a tank. That is why we have decided to develop a flexible and generalized strategy based on anticipating changes in power demand. So, the present paper focuses on the optimal management, using a generalized model based predictive controller, of an ENGIE Cofely's multi-energy district boiler equipped with a hot water tank. The plant is located in northeast France, in the Grand Est region (Haut-Rhin). The controller makes use of a generic model of the district boiler and power demand is accurately forecasted over the next 24 h thanks to the MRA-ANN approach (i.e. a wavelet-based Multi-Resolution Analysis combined with Artificial Neural Networks). With the predictive strategy, the consumption of gas and carbon dioxide (CO2) emissions are significantly reduced. In the same time, the overall economic gain is increased. Overall performance is the best with a 200 m(3) hot water tank added to the plant.

• 出版日期2017-2-25