Skip to main content
European Commission logo print header

Grid Architecture for Wind Power Production with Energy Storage through load shifting in Refrigerated Warehouses

Final Report Summary - NIGHT WIND (Grid Architecture for Wind Power Production with Energy Storage through load shifting in Refrigerated Warehouses)

The NIGHT WIND project's full title was 'Grid architecture for wind power production with energy storage through load shifting in refrigerated warehouses' which basically reflects its aims. The idea was to use a refrigerated warehouse as a giant battery for wind energy, to store all electricity produced during night time by windmills all over Europe, and to release this energy again during the peak electricity demand hours in daytime.

The wind energy is produced at rather random times, whereas the energy use pattern shows distinct demand peaks during day time and office hours, and low levels during the night. The random production of wind energy cannot easily be accommodated on the grid by switching on and off conventional energy suppliers, like coal fired power plants, which would lead to an increase of CO2 emissions, rather than the reduction of CO2 emissions which is desired. In order to accommodate the random production of wind energy in the grid, it would be most convenient when alternative (renewable and conventional) electricity producers could balance out the difference between production of wind energy and electricity demand. The NIGHT WIND project aims to store wind energy produced at night in refrigerated warehouses, and to release this energy during daytime peak hours.

NIGHT WIND made use of existing technology, extended with novel control strategies. The new control strategies are needed to set the temperature level in refrigerated warehouses to a level that is derived from the actual balance between wind energy production and actual electricity demand. This was the case for 'island operation' with delivery of surplus energy to the grid, and also for the case of distributed energy resources (DER), where windmills are physically located elsewhere than the (existing) cold stores, but controlled in an interdependent way to support the European energy service network. Design of control strategies, with the help of powerful simulation tools was the main task in the NIGHT WIND project.

The research stage of the project included the following topics:
- potential, economic and trade aspects of wind power DER + cold store DSM;
- design and modelling of infrastructures for island operation of wind energy + cold store DSM;
- control concepts and algorithms for wind energy + cold store DSM grid integration;
- quality preservation of frozen products during minor temperature fluctuations;
- legal issues;
- demonstration and introduction outline plan.

The projects objectives can be summarised as follows:
1. Integrating renewable energy resources into the European energy service network by providing new facilities for energy storage.
2. Increasing the economic value of wind energy by providing means to deliver the energy at peak demand hours.
3. Increasing the competitiveness of SME cold storage facilities by providing adding 'energy storage' as an additional service to be provided for the European energy service network.
4. Offer a solution to integrate wind energy with energy storage in the European electrical grid, giving space to a further growth in the use of wind energy worldwide and a contribution to the Kyoto targets at the same time.

In Bulgaria the experiments on food quality deterioration due to temperature cycling were prepared and carried out by the Plovdiv Canning Institute. In these experiments, foodstuffs were tested physically for a number of important quality parameters, and by a human test panel for sensory evaluation.

Partners CENER and RISOE have bundled their efforts in order to produce a controller for the refrigeration system, in which the NIGHT WIND function is incorporated. Unfortunately, such a control system could not be realised in the time available in the project. A system functioning partly was produced, but the time required to produce the time series (three hours) was impractical for implementation in a real cold store.