REF's Wind Power Flow Model
Experience in both Germany (see the E.ON Netz Wind Report and Denmark, where there are substantial deployments of wind power, are not encouraging in this regard. REF believes that comprehensive modelling of the envisaged wind power development is a logical precaution to avoid suboptimal system design. However, no such modelling currently exists. In order to move towards supplying this deficit the Renewable Energy Foundation has commissioned Oswald Consultancy Ltd to build a power flow model employing hourly Met Office wind speed data to estimate the aggregate performance of a large wind carpet spread over the UK. Weather data has been kindly provided by the Met Office and the British Atmospheric Data Centre and is Crown Copyright. An interim report on the model is now available, and shows that while some degree of smoothing does occur, the aggregate output is still volatile to a significant degree, with many important implications for the overall system.
This ongoing work has already attracted considerable interest, and a full version will be publishing further elements in the study as they become available.
Other Related Studies
While this model is, at present, unique, there is a considerable related literature around the subject. For example REF co-commissioned and funded with the Institution of Civil Engineering (ICE) two studies of the Danish wind experience. These were conducted by Hugh Sharman (of the consultancy Incoteco), and published in the Proceedings of the ICE, Civil Engineering in 2005:
These two acclaimed papers were awarded the Institution of Civil Engineering's Telford Gold Medal in 2006, and have been the cause of much debate.
Others, by contrast, have argued that the United Kingdom's superior wind resource and the possibility for geographical dispersion would give greater yields and render the aggregated output smoother. A prominent example, the work of Mr Graham Sinden at Oxford University's Climate Change Institute, which performs a statistical analysis of the frequency of wind speeds.
However, this work, while interesting, is far from conclusive since it does not model actual power flows from a distributed wind carpet. In a report for the Tyndall Centre, Professor Strbac and his colleagues observed:
We believe that serious statistical analysis needs to be performed on a considerable amount of historical data in order to assess the probability of having zero or near zero wind power output from all wind farms. The most serious incidence of calms could be anticyclone 'cold snaps' which give high demand but little wind anywhere in the country. Due to lack of data we were not able to estimate the likelihood of such events.
Dusko Nedic, Anser Shakoor, Goran Strbac, Mary Black, Jim Watson and Catherine Mitchell, Security assessment of future UK electricity scenarios (July 2005) Tyndall Centre for Climate Change Research Technical Report 30. Available from: http://www.tyndall.ac.uk/publications/tech_reports/tech_reports.shtm
Interested readers may also wish to consult the studies conducted by the United Kingdom Energy Research Centre, The Costs and Impacts of Intermittency (2006).