Distributed MPC for Power-to-Gas facilities embedded in the energy grids

This paper studies the optimal supply levels of Power-to-Gas (PtG) facilities equipped with local storage devices and fuel cells. The produced energy can be directly injected in the gas grid, utilized in chemical/mobility industry, and stored in a storage device and later on reconverted back into electrical energy using fuel cells before injecting it to the power grid. The aim of the control process is to maximize the PtG facilities´ profit from the produced energy while avoiding overloading grids. The problem is solved by a distributed model predictive control (MPC) approach in which each PtG facility makes decisions based on its local information, yet still coordinates its supply bids to the grid operators. In this way, grid operators can manage PtG facilities via shadow prices associated with the grid constraints in order to avoid overloading. We perform simulations to investigate the evolution of shadow prices modified by grid operators and to study the economic feasibility of PtG facilities under realistic estimates of the selling price and demand patterns on the energy grids.