The OWA has launched an Invitation to Tender for companies, academics and consortia to standardise the processing of wind farm SCADA data for use in model validation.

Turbine interaction modelling for offshore wind farms is reaching a critical stage as wind farm clusters become larger and larger, and the potential for modelling biases increases. Wind farm SCADA data is still the primary measured data used for testing turbine interaction models and assessing modelled vs measured pattern of production (PoP, the distribution of energy production across turbines) and Annual Energy Production (AEP, energy yield per year) array/turbine interaction efficiency.

SCADA should be the source of the truth, however this truth can change depending on how the data are interpreted and analysed. There is no standardised way of processing SCADA which means no wake model benchmark can be compared to another even if the model is identical as the outcomes could be different. These differences can be the same if not larger than some effects that we are trying to model and substantiate with real world evidence.

The OWA would like to investigate the statistical fundamentals of analysing pre-processed SCADA data to ensure that data processing and filtering are appropriate to the physical processes taking place and models being compared. It is not attempting to redefine how individual turbine SCADA data should be cleaned as this is already well established for the purposes of power performance tests etc. Instead, it should be assumed that pre-existing and available processes and, ideally, codes can be leveraged in the first instance to produce a SCADA dataset that can be interrogated for the purposes of model testing.

The deadline for clarification questions is 11 April.

The closing date to receive tender submissions is 2 May at 2pm.

All clarification questions and tender submissions should be sent electronically, by their respective deadlines to neil.adams@carbontrust.com.