In this article, Amandine Martins of Reuniwatt, Bruno Daugrois of Naldeo Technlogies & Industries and Amélie Belfort of Synergîle detail the interest of PV + storage with forecasting for the french overseas territories, with the exemple of a 986 kWp photovoltaic power plant installed on the roof of a supermarket on Reunion Island and coupled with a lithium-ion battery with a storage capacity of 1200 kWh.
Non-interconnected zones (ZNI) are territories that are either not connected or only partially connected to the continental electricity grid. This is specifically the case for Corsica, the overseas departments and regions (Guadeloupe, La Réunion, Mayotte), the single territorial collectivities (Martinique, Guyana), certain island collectivities (Saint-Pierre-et-Miquelon, Wallis and Futuna), and the Channel Island of Chausey.
These territories have small electricity systems (50 GWh – 3000 GWh) and have climatic and geographical specificities which present major challenges for the massive development of renewable energies and for the stability of their power systems. Currently, the ZNI produce most of their electricity through the use of imported fossil fuels.
The Law on Energy Transition for Green Growth has enabled the implementation of an energy policy adapted to each ZNI by means of multi-annual energy programming (PPE), thus constituting a strong tool for local management, with the common objective of energy autonomy for these territories in 2030.
The ZNI will experience a strong change in their electrical consumption, the three main causes of which are: demographic change, the increase in the rate of electrical appliances in households ( especially air-conditioning) and the development of electric vehicles suited to the scale of the territories and their usage.
The strong development of non-synchronous generation, essentially composed of wind and photovoltaics, will affect the energy supply in these territories significantly by 2030. In this context, projects integrating renewable energies have additional constraints in that they must use proven and reliable technologies, either to reduce public expenditure or to guarantee the electrical grid security .
Modelling the means of generation and the requirements for the safe and stable operation of these projects is therefore a major enabler for the energy transition in the ZNI.
Spotlight on Reunion Island
In Reunion Island, renewable energies account for 752 GWh to 1 079 GWh per year and have represented between 30.2% and 46.7% of the electricity mix since 2000. In 2019, the share of electricity production from photovoltaics was 8.5% . Reunion Island is therefore a territory committed to the comprehensive use of renewable energies.
In this context, the E. Leclerc shopping centre, on the Le Portail site, was the first installation in Reunion using a combined photovoltaic / storage system. Le Portail, operated by Albioma, consists of a 986 kWp photovoltaic plant installed on the roof of a supermarket, coupled with a lithium-ion battery with a storage capacity of 1,200 kWh.
The Le Portail plant is required to adhere to a trapezoidal daily power profile, indicating the maximum power output in order to comply with the requirements for non-interconnected areas. Because of its location in the middle of the Indian Ocean, however, cloud cover at this site can vary greatly during the day, which has a massive impact on solar irradiance levels and therefore on the power produced by the photovoltaic panels. Albioma has thus selected an Energy Management System (EMS) and a forecasting solution to determine the level of energy production commitment and to control the production facilities integrated in the plant.
The Energy Management System (EMS)
The Energy Management System (EMS) is used to determine the power plant’s production commitment in order to maximise revenue, and to make the best possible real-time trade-offs to control the power plant. Several hours in advance, a trapezoidal production schedule is set by the EMS and communicated to the grid operator. Moreover, every 15 minutes, the control settings applied to the photovoltaic power plant and the battery storage are adjusted to adapt their response to the expected differences between the forecast and the actual conditions, by means of the flexibilities around the template.
These decisions are based on weather forecasting, which allows to anticipate the solar resource available. These forecasts, combined with numerical models, make it possible to predict the behaviour of the photovoltaic + storage system so as to maximise performance at several levels: minimising the amount of solar energy curtailment, avoiding template violations which are accompanied by penalties, avoiding excessive storage cycles, maintaining the batteries at appropriate charge states…
The combination of solar forecasts and advanced energy management tools enables real-time control decisions that respect the technical and contractual constraints of the plant while anticipating their implications for the rest of the day.
The Portail plant is a pioneer for projects of this type. The calls for tenders launched more recently by the French Energy Regulatory Commission (CRE) make even better use of the forecasting capacity of the EMS/forecasting combination, since they offer a premium for the grid injection of electricity after sunset, in periods of peak demand. The key is to fully charge the batteries during the day in order to maintain high production in the evening. During the day, the EMS will look for a compromise between respecting gauges and recharging to supply electricity to the grid in the evening. The solar forecasts provide crucial information for finding the right balance.
The forecasts are fed into the EMS, which performs the necessary analyses to inject the required power profile into the grid. On this power plant, forecasts are made at several time horizons, from fifteen minutes to several days ahead, using a combination of three complementary services:
- The day-ahead forecast (6 hours to 24 hours ahead) uses numerical weather prediction models. This allows a commitment to be made in accordance with the grid operator’s specifications for 24 hours ahead.
- The intraday forecast (15 minutes to 6 hours in advance) uses satellite images to predict the cloud motion and calculate the solar irradiance in the area. This allows for intraday balancing of supply and demand and for the grid operator to starting up back-up power plants if necessary.
- Intrahour forecasting (from 1 minute to 30 minutes in advance) is based on observation of the sky with an all-sky imager. It allows for the optimisation of battery life by limiting microcycles (rapid alternation between charging and discharging linked to a short and sudden variation in PV production).
The “PV + Storage” model, a vector of energy transition
The “PV + storage” model with solar forecasting was born from the ambitious idea to maximize the share of renewable energies in the energy mix of the ZNI. This model has been successfully implemented, particularly in areas with multiple constraints (limitation of building spaces, fragility of the electric power grid, intense meteorological phenomena, complex climatic conditions, etc.). The performance and therefore the profitability of this type of plant depends in particular on the quality of the forecasts and the performance of the EMS. These intelligent systems make it possible to maximize the energy injected into the grid, limit penalties (in the event of non-compliance with the production template) and slow down the aging of storage batteries.
In a “smart grid” logic, it is possible to replicate this model on a larger scale, by having several photovoltaic plants and central or distributed storage units: in this case, information networks are necessary to gather and process information about the the current state of production resources, obtain the production forecast for each of the plants, and manage the flexibility levers.
The article was originally published on pv magazine France (26/07/2021).