Recent studies have shown that the intermittency of wind energy can be mitigated by means of an energy storage system (ESS). Energy can be stored during periods of low energy demand and high wind availability to then be utilised during periods of high energy demand. Measure-Correlate-Predict (MCP) methodologies are used to predict the wind speed and direction at a wind farm candidate site, hence enabling the estimation of the power output from the wind farm. Once energy storage is integrated with the wind farm, it is no longer only a matter of estimating the power output from the windfarm, but it is also important to model the behaviour of the ESS in conjunction with the energy demand. The latter is expected to depend, amongst other factors, on the reliability of the MCP methodology used.
This paper investigates how different MCP methodologies influence the projected time series behaviour and the capacity requirements of ESS systems coupled to offshore wind farms. The analysis is based on wind data captured by a LiDAR system installed at a coastal location and from the Meteorological Office at Malta International Airport in the Maltese Islands. Different MCP methodologies are used to generate wind speed and direction time series at a candidate offshore wind farm site for various array layouts. The latter are then used in WindPRO® to estimate the time series power production for each MCP methodology and wind farm layout. This is repeated with actual wind data, such that the percentage error in energy yield from each MCP methodology is quantified, and the more reliable methodology could be identified. While it is evident that the integration of storage will reduce the need for wind energy curtailment, the reliability of the MCP methodology used is found to be crucial for proper estimation of the behaviour of the ESS.