Long term analysis of CSP plant output must include the probability of exceedance.
Exceedance probabilities estimation should include all the uncertainty sources.
Long term statistics on solar resource could differ than statistic on CSP plant output.
Feasibility analysis for CSP projects requires quantifying the economic risk associated with the inter-annual variability of the solar resource and uncertainties in the annual energy production estimates. To assess this risk, statistical methods may be applied either before the system simulation (e.g. to create solar resource data in terms of exceedance probabilities such a P50 or P90) or on the system performance to determine the likelihood that a power plant will generate a certain amount of energy in any given year over the plant’s life (e.g. the P50 or P90 annual productions). This paper presents and illustrates the long-term analysis of four CSP plants (parabolic trough and central receiver technology) in four sites within the Mediterranean and North Africa region; the analysis includes the exceedance probabilities estimations. Satellite-derived data for 20 years have been used for solar resource datasets in hourly basis, and multiyear simulations of the plant output have been performed with System Advisor Model. The associated uncertainties have been included in the long-term analysis of both DNI and CSP plant production. The results remark the advantages of multiyear analysis of CSP output for long-term analysis against the use of artificial meteorological years which represent the long-term.
- Concentrating solar power;
- Solar resource assessment;
- Long-term CSP generation;
- Probability of exceedance
Copyright © 2015 Published by Elsevier Ltd.