Page 68

energetica-india-55

RENEWABLE ENERGY KIRAN NAIR HEAD : WIND RESOURCE ASSESSMENT & FORECASTING MYTRAH ENERGY (INDIA) LTD The number game - P50, P75, P90; witch-science or mathematical prodigy? Energy predictions and investor(s) confidence Key statistics for any financer for his model is CAPEX, OPEX, DOC (date of commissioning), tariff and the expected energy yield. Apart from the expected energy yield and to a certain degree DoC (but the degree of uncertainty associated with DoC is negligible when compared to the energy yield prediction) all other figures can go relatively straight to the model without much of a confusion. The financial state of a wind farm or portfolio strongly influenced by the expected energy generation over the lifetime of the project and obviously an accurate prediction of long term energy yield is the most important footing for any wind power project. As the investor (debt) community becomes more conversant, they tend to use P75, P90 or even P99 values as a safe bet to calculate the Debt Service Coverage Ratio (DSCR1). Some of the underperforming (or over predicted??) operational assets are fuelling their arguments by a decent manner. Meanwhile in the other side of the community, as the technology/market getting matured the windfarm owners and operators improved their technical awareness and are more cognizant about the prediction, operation & maintenance and the performance. So it’s equally vital and important to investors as well as developers to have ‘close to the actual’ energy estimates and the risks associated with it to build a sound business model around it. An uncertainty analysis is often performed as part of a standard wind farm energy yield prediction and is of particular interest for risk profiling in the financial community. When the uncertainty in the wind power project is high, the desire for flexibility in investment tends to be higher and it’s no surprise that an investor wants to know the level of uncertainty before investing in a project. Once the long term net energy yield or P50 (It is called the P50 as there is a 50 % chance the result will be lower and a 50 % chance the result will be higher than the predicted long term energy yield) of the wind farm is calculated, uncertainty associated to the predicted long term net energy yield over varying averaging periods is calculated and presented across various confidence levels and unless if the project falls under a certain degree of confidence level, it is impossible to construct a sound financial model for an investment. Energy Estimates, Uncertainties and Probability levels The uncertainties in the energy prediction (and windfarm operation) have two main ancestries: 1. Uncertainty in wind measurement and prediction methodologies. 2. Uncertainty due to natural wind variability. While the first one can be abridged with careful planning through the wind monitoring and energy yield assessment techniques by limiting the errors originating while selection of equipment, measurement, analysis and modelling the second one i.e.; natural wind variability is an independent factor which leads to uncertainty in long-term energy yield predictions is more trickier among the two to handle. There are several sub classes and sources of uncertainty. The most common contributors and the means to diminish them are listed below; Wind measurement uncertainty – The wind measurement uncertainty is usually linked to three sources: • Calibration of the instruments • Operational characteristics of the instruments • Flow distortion due to instrument mounting and the effect of the terrain International technical guidelines/ standards are available for wind anemometry to reduce these effects, and adhering to these guidelines is recommended to improve wind measurement quality. The guidelines do not ensure the effects are eliminated, but are generally designed to limit the effects to acceptable limits. However a good understanding of the mechanisms is essential in order to correct them during the final data analysis. Vertical extrapolation - If wind speed measurements are undertaken at a height different from the hub height, an extrapolation technique is required to estimate the wind speed at hub height at the monitoring location. Two most popular methods used for this (Power Law and Logarithmic Law) are pragmatic in nature and hence carries a certain amount of uncertainty. Performing a wind speed extrapolation introduces uncertainties, whose magnitude depends on the methodology used, number of sensors available, amount of height through which data being extrapolated/ interpolated, height of sensors from the ground etc. Uncertainty in vertical extrapolation can be eliminated by monitoring the wind speed at hub height. Wind flow model uncertainty – Arguably the biggest contributor from assessment side ; a wind flow model is used to horizontally extrapolate wind speeds from the monitoring site over the wind farm area, in order to predict the average wind speed and energy output at different turbine locations. The complexity of the site, number of measurement campaigns available, the model (linear/Computational Fluid Dynamics) used for the calculation, quality of the map used etc. are the factors assessing the uncertainty involved in the process. Unlike the other wind speed related uncertainties, this one is a more of a user depended one. Using multiple masts, having a ground surveyed map, advanced modelling techniques are some of the techniques to reduce the uncertainty here. Power curve and the performance - Turbine manufacturers provide power and thrust data for one or several specific air densities against every wind speed values. These curves (power/ thrust) are often calculated from a model or derived from a tunnel. Once the proto type testing of the specific model is done, manufacture can provide a measured power curve to support the calculated power curve. But many a times the actual operational fleet conditions are different from that of test conditions and there could be deviations in the power curve imparted by site specific 68 energetica INDIA · JAN | FEB16


energetica-india-55
To see the actual publication please follow the link above