Optimal Energy and Flexibility Dispatch of Grid-Connected Microgrids

Doktorsavhandling, 2022

This thesis proposes an optimization model to efficiently schedule energy and flexibilities of a grid-connected microgrid (MG) with non-dispatchable renewable energy sources and battery energy storages (BESs). The model can also be used to coordinate the MG operation with the connected upstream distribution grid and to assess the MG flexibility considering economic viability, technical feasibility, and BES degradation. The performance of the model was tested for both deterministic and stochastic formulations using two solution approaches i.e., day-ahead and rolling horizon, in different simulation and demonstration test cases. In these test cases, the model optimizes the schedule of the MG resources and the energy exchange with the connected main grid, while satisfying the constraints and operational objectives of the MG. The flexibilities from the MG would also be optimized when the MG provided flexibility services (FSs) to the distribution systems. The coordination with the distribution system operator (DSO) was proposed to ensure that the microgrid operation would not violate the technical constraints of the distribution grid.

 

Two types of test systems were used for the simulations studies: 1) distribution grids with grid-connected MGs and 2) building MGs (BMGs). The distribution test systems included the 12-kV electrical distribution grid of the Chalmers campus and a 12.6-kV 33-bus standard test system, while the BMGs were based on real residential buildings i.e., the HSB LL building and the Brf Viva buildings. Results of the Chalmers’ test case showed that the MGs’ economic optimization could reduce the annual cost for the DSO by up to 2%. Centralized coordination, where the MG resources were scheduled by the DSO, led to an even higher reduction, although it also led to sub-optimal solutions for the MGs. Decentralized coordination was applied on the 33-bus network with a bilevel optimization framework for energy and flexibility dispatch. Two types of FSs were integrated in the bilevel model i.e., the baseline (FS-B) and the capacity limitation (FS-C). The latter has found to be more promising, as it could offer economic incentives for both the DSO and the MGs. In the studies of the BMGs, the BESs were modeled considering both degradation and real-life operation characteristics derived from measurements conducted at the buildings. Results showed that the annual building energy and BES degradation cost could be reduced by up to 3% compared to when the impact of BES degradation was neglected in the energy scheduling. With the participation of the BMG in FS-C provision, the building’s operation cost could be further reduced depending on the flexibility price. A 24-h simulation of the BMG’s operation yielded an economic value of flexibility of at least 7% of its daily energy and peak power cost, while the DSO could benefit from the FS assuming that the dispatched flexibility could be used to reduce the subscription fee that guarantees a certain power level. For frequent flexibility provision i.e., multiple times within a year, the value of flexibility for the MG operator could be reduced due to the BES degradation.

 

To demonstrate the practical use of the proposed model, an energy management system was designed to integrate the model and employ it to optimize the energy schedule of the BMGs’ BESs and energy exchange with the main grid. The energy dispatch was performed in real-time based on the model’s decisions in real demonstration cases. The demonstration results showed the benefits of the model in that it helped reduce the energy cost of the BMG both in short term and in long term. The model can also be used by the MG operators to quantify the potential and assess the value of microgrid flexibility. Moreover, with the help of this model, the MG can be employed as a flexible resource and reduce the operation cost of the connected distribution grid.