HYB25-36
Overview of Recent Microgrid Studies and Feedbacks : Evaluation of the Gain of Renewable Generation Forecast and Feedback on Frequency Control
01 HYB26-36
Presented by: Maxime VELAY
Maxime VELAY 1, Clement HUET 2, Emeric BRUN 1, Pierre HORSIN 1
1 EDF R&D, France
2 EDF SEI, France
EDF SEI operates over 20 isolated micro grids, notably in French overseas territories and off the coast of Brittany. So far, those micro grids have mostly been powered by diesel engines; renewable energy sources and storage systems are progressively installed to reduce fuel consumption and emissions.

Some grids already integrate those resources and the ambition of the French regulator and of EDF SEI is to go towards 100% renewable penetration in the next decade.

EDF SEI and EDF’s R&D propose in this paper an overview of recent studies carried out to enhance the operation of the microgrids operated by EDF.

First, to operate those new systems, Energy or Power Management Systems (PMS computes setpoints based on power measurements and thus do not consider forecasts, whereas EMS computes setpoints based on current status and forecast of consumption and generation) are needed to control all the producers. Before investigating how to get good quality forecast, EDF SEI and EDF R&D evaluated the benefits that the forecast could bring to those micro grids in terms of renewable energy penetration and diesel savings.

To answer this question, the methodology employed consisted in comparing the results obtained from simulations of:

The current control laws of the PMS with different settings

The current control laws optimized based on perfect forecast

The day ahead optimisation of the use of the asset based on perfect forecast, including relevant constraints (margins, cost of genset start, renewable curtailment limitation)

The simulations were conducted on the case of Sein Island (Brittany) that has a peak load of 330 kW, and considering 400 kWc PV systems and 300kW / 277 kWh of storage system.

In conclusion, the maximum gain obtained in those simulations was 6% fuel reduction with perfect forecast and day ahead optimisation. Optimizing the PMS settings based on perfect forecast brought 3% saving in fuel consumption.

With those results, EDF SEI decided not to add forecast for the moment because it would not bring enough benefits, and even more when considering that forecasts are not perfect.

The second topic shared in this paper regards the periods when gensets and storage systems operate at the same time. In one of EDF SEI microgrid, the slow but large fluctuations of PV systems generation were not compensated by the storage system that was in grid forming mode with a droop frequency control. The expected behaviour of the storage system would have been to smoothen the fluctuations of gensets generation by following the PMS forecasts. It was not the case even when changing primary frequency control settings (offset, droop, dynamic of the control) which ended up damaging the gensets. From our understanding, the frequency variations and the dynamics of the PMS regulation did not work fast enough or the isochronous frequency regulation of the gensets was faster.

The solution implemented was to set the storage system in grid forming isochronous mode and to set the thermal plant in PQ mode with droop. This solution worked as expected.