ODTÜ KKK'deki güneş fotovoltaik santralinin uzun vadeli enerji verimi tahmini

Installing large scales of solar energy introduces some financial risks for the investors since the solar resource is variable. Thus, a long-term energy yield estimation is needed to see the variability of the solar resource. This study aims to see the relation between long-term trends of global horizontal irradiation (GHI) with the energy yield and the levelized cost of electricity (LCOE). Probability of exceedance (POE) values were found to assess the bankability of installing a solar PV power plant. Different datasets, such as typical meteorological year (TMY), satellite-based and ground-measured data, were used. Quality assessment was done to check the accuracy of ground-measured data using quality control tests. Erroneous GHI data was estimated using the Erbs model. Global tilted irradiation (GTI) was also estimated using the isotropic sky-diffuse model, and it was compared with measured GTI. The interannual variability of GHI was found as 5.94% and 2.21% for ground-measured and satellite-based data, respectively. The comparison of TMY and P50 values showed that TMY datasets underpredicted the annual GHI by about 5.15% and the energy yield by about 6.83% on average. When Normal distribution was compared with the empirical method, the P50 value was underpredicted by 0.93% for energy yield, whereas, Normal cumulative distribution function (CDF) overpredicted P90 value than the empirical CDF 3.93%. Doing stochastic simulations resulted in the highest POE values: P50 and P90 values increased by 3.30% and 6.10%, respectively, compared to the empirical method. Moreover, while the range of energy yield increased to about 1500-2000 kWh/kWp, LCOE range increased to 0.05-0.19 $/kWh in stochastic simulations. As the overall uncertainty of energy yield was found as about 7.08%, it can be further reduced by searching for other sources of error such as temperature and soiling.