Performance ratio is the figure of merit of a solar plant for its conversion efficiency of the incident light energy into electrical energy. It is dependent on the efficiency of the equipment used in plant construction, design, engineering and operational practice. Selection of equipment, such as solar module, inverter, inverter transformer and main transformer is key to achieving high PR.
Besides, selection of equipment the inter-row spacing, module tilt and azimuth angle, plant layout, cable length and gauge, pooling voltage, DC/AC ratio etc. are the design parameters which influence PR. Needless to say, integration and workmanship in plant construction though not overtly but covertly play a role in PR. After the plant is designed and engineered the operational practice involving frequency of module cleaning impacts the PR.Since irradiation is not within our control invariably the EPC companies guarantee PR as one of the Facility Acceptance Test (FAT) criteria.
It may be mentioned that the parameters which enter the PR calculation are the electrical energy yield at the pre-defined location, Irradiation received on the inclined plane and the total wattage of the plant (sum total of the nominal wattage of all PV modules). The electrical meters are quite accurate and have a maximum measurement uncertainty of 0.2%. The irradiation is measured using a thermopile based secondary pyranometer which has measurement uncertainty of 2%. Similarly, the wattage of the modules is measured by a flash tester (AAA class) which has a measurement uncertainty of 2-3%. The combined uncertainty is the uncertainty in the PR measurement which turns out to be 3%. In order to compare the temperature corrected PR, high accuracy temperature sensor is used which along with the signal conditioning equipment shall have an uncertainty better than 1%. The placement of sensors and their maintenance is very critical from the point of view of correct and reproducible and accurate PR measurement. The pyranometer shall be coplanar with the array and it should be ensured that there is no shadow on the pyranometer particularly in the morning and evening. The dome of the pyranometer shall be cleaned daily and it should be calibrated every year. Placement of temperature sensor for measuring the module temperature is important. Temperature sensor shall be installed in the middle of the array as edge of the array have different air velocity and will influence the temperature measurement. Temperature sensor shall be stuck to the centre of the cell but away from the junction box using a thermal conducting glue. The length of cable between sensor and data logger shall be as per the specification of the measurement system supplier. It is suggested that data shall be recorded in 5 minutes interval. In fact, most plants record data in 15 minutes interval which is good for monthly reporting but finer interval of recording is better in the sense of capturing events.
Generally, developers have PV plants in more than one state and often performance of one plant in one state is com- pared with another. For the same radiation, temperature corrected PR in one state shall be equal to another plant design-, equipment and operation-specific parameters being same. If the PRs are identical then it is expected that generation for a given radiation across states shall also be equal. We collected weather as well as generation data for 5 plants: 4 plants in Telangana and 1 plant in Rajasthan having almost similar capacity, module supplier and inverters. In fact, 3 plants of Telangana and 1 plant of Rajasthan had module and inverter makes from the same sup- plier. The generation data for the month of April for all the plants was collected and data of days there was any break down due to grid or plant was excluded. All plants have the same cleaning cycle yet there may be slight difference in cleanness of modules in different plants and that is a possible noise in the data. Based on the collected data PR, Temperature Corrected PR and Specific Generation (kWh/MWp) were compiled at one irradiance, namely, 7.45 kWh/m2 on the PoA. The data are given in table 1.
It is evident from the table that for the given radiation the PR values and specific generation are quite comparable. The deviations are within the range of 3% expected measurement uncertainty of PR. Deviation in specific generation is more than the PR numbers. This deviation can be explained on the basis of spectral matching of module as well as the solar spectrum varying across locations because of atmospheric factors such as aerosol and various other gases. The pyranometer measures both the direct and diffused light incident on it. However, for a given energy incident on it the fractions of diffused and direct irradiance may me different and response of same module under two conditions will be different. So, even if the PR of two different plants located in two different locations may be within the range of measurement uncertainty, the energy output variation may be more than the PR variation.
The question remains whether PR can be a good indicator for the energy yield of a plant. Given all the uncertainties of measurement PR can be used as a performance monitoring tool of the solar plants. Given below is the correlation curve between the specific generation and performance ratio of plants in Telangana on a given day. It is shown that correlation of PR with the specific energy output of a solar plant is very good and to the extent of more than 98.5% plant output can be explained on the basis of PR.
It is shown that PR of solar plant is a very good indicator of the plant performance within a state. The data of Rajasthan plant does not lie on the regression curve of Telangana. It is possible only when the pyranometer is under-measuring the irradiation within the uncertainty band or the nominal module wattage is higher than actual. However, to expect that 100% energy yield can be explained on the basis of performance ratio is fortuitous. While energy yield actually measured is the reality with high degree of certainty as electrical meters are quite accurate, the PR calculation involve measurement of irradiation, temperature, module wattages, etc having inherent measurement uncertainties. Thus, 100% correlation with goodness of fit (R2) as 1 cannot be achieved. There are other solar spectrum related factors which impact generation but do not directly enter in the PR calculation. So, it is possible that besides other design factor differences the plant in location A may perform better or worse than plant B because of better or worse spectral match of the modules with the local solar spectrum. This impact shows up in the energy output implicitly. Often O&M contractors are incentivized based on achieving a PR above a thresh- old. This practice per se has no lacuna but one must bear in mind that one-to-one correlation between PR and yield does not exist and accordingly the incentives shall be worked out.
Credits: Mr Amitabh Verma, CTO, Aditya Birla Group