Posters

Abstract

By 2030, a total of wind power installed capacity between 40 GW and 80 GW will have reached its useful lifetime in Europe. Repowering consists of replacing a wind plant with a newer one that either has a greater nameplate capacity or more efficiency which results into a net increase of annual energy production (AEP). Repowering wind plants has several benefits. First, it significantly improves the capacity factor. Second, when the rated power remains constant, a repowered wind plant is expected to generate less noise, to have lower visual impact and to have lower impact for birds and bats thus increasing local acceptance of projects. Repowering projects come with lower capital investments compared to greenfield projects since a part of the infrastructure can be reused. However, a cost-benefit analysis is required to determine the project profitability. The evaluation of the project profitability requires an estimate of the AEP. The estimation of the annual energy production (AEP) for a greenfield project is somehow "naive" in the case of repowering since it relies on onsite wind measurements and long-term weather data but not on existing wind plant data. For rather simple terrains, the "naive" AEP estimate uncertainty ranges between 6% to 12%. We present here a method called "Repowering coefficient method" that consists in scaling the existing wind plant AEP estimate from production data so that it represents the AEP of the projected wind plant. The scaling factor is evaluated with the "naive" method applied to both the existing and the projected wind plant. We observe that this method does better than the "naive" one for certain cases. The first part of the presentation consists in presenting the uncertainty evaluation of this new method: is it biased ? how the error estimate variance (uncertainty) is related to the uncertainty of the underlying processes? For example, if the repowering coefficient is estimated from a "naive" EYA that comes with 10% uncertainty, what is the uncertainty of the AEP estimate of the repowering coefficient method. With these answers, we will see that repowering coefficient only does better if there is a certain degree of similarity between the two wind plants. This similarity is evaluated with mathematical operators that should be handy to implement. The previous results come from simulation and we conclude with real-life evaluations of the method. These onsite results validate our uncertainty estimate. With this presentation, the audience will gain better understanding on whether to opt for a standard EYA process (that comes with the cost of installing wind measurement devices) or to opt for the repowering coefficient method that besides lowering uncertainty could save costly measurement campaigns.