A recent thesis from the Department of Industrial Economics and Technology at NTNU investigates two different optimization-based system control strategies on a local electricity market level. These strategies are evaluated based on two outputs: the total cost of electricity during operation [for both the energy sharing region (ESR) as a whole and for each end-user within the ESR] and peak grid power demand, which is determined by the rate of self-consumption of distributed energy resources (DERs) within the ESR. To investigate the relationship between peak power demand and total electricity costs, a multi-objective optimization (MOO) approach based on the ϵ-constraint method was also implemented.
The first optimization strategy was the decentralized control system strategy, which minimized electricity costs for each end-user within the ESR, assuming they could only utilize their own local production, storage units, and the grid to meet their demands (visualized in the figure above). The second optimization strategy, contrarily, minimizes electricity costs for the ESR as a whole, and enables peer-to-peer (P2P) trading amongst end-users. These strategies were applied to two different communities – 25 residential buildings in London, UK, and three large industrial end-users at Forus, Norway. The results for one house in Week 24 using approach 1 are visualized below.
The results of the thesis show that the centralized optimization strategy with P2P electricity trading gave the lowest total costs for the ESR. A cost reduction of 1.0-8.0% was found when compared to the decentralized strategy, as well as a reduction of grid energy consumed of 1.4-18.9%. Finally, the results from the MOO show that there is a dependency between total electricity costs and peak power demand for the cases studied and that a small increase in cost can reduce the peak power demand by a significant amount.