As the world broadens its portfolio of power options to meet growing energy demands and increasingly stringent environmental concerns, solar power is emerging as an attractive option. However, the current lockdown has taken a toll on the operations and maintenance activities of our solar power plants. This situation is more adverse for the rooftop plants located on the commercial and industrial locations because these power plants have also experienced several months of inactivity. Such prolonged isolation coupled with high summer temperatures can impact the health of the plant. In this situation, it is important to get a thorough diagnostics study done on the plant before re-commissioning and mobilizing the operations team post lockdown.
To prepare your plant to restart (for inactive rooftop plants) or start maintenance activities (for plants already in operation during the lockdown), you can consider getting remote diagnostics to identify the critical concern areas. Such analyses can help in ensuring high performance, low downtime, and proactive fault detection in a solar PV power plant. On-site weather data, production data from the panel strings, inverters, and transformers, data from AC and DC control panels and historical breakdown data can be studied to create a future maintenance strategy for your plant.
In this article, we are giving you a glimpse of the depth of analysis possible remotely and how it can be beneficial for the overall health of your solar power plant.
Historical Generation and Performance analysis
In this analysis, trailing twelve months and year on year data is analyzed at different plant levels like modules, inverter, string, etc. This analysis further helps in calculating the AC losses, auxiliary losses, plant unavailability, and inverter efficiency losses and provide a better understanding of the different plant components.
- Error in WMS data logging indicating abnormal PR: At a 15.6 MWp plant in the state of Tamil Nadu, we evaluated the historical data and found that normalized PR was 0.70% lower than estimated PR. In Jun-16 and May-17, the PR was abnormally high while in June, July, and August of 2017, PR and generation were extremely low.
The reason behind low PR was high load shedding along with mismatch in data logging and high PR was observed because of frequent radiation sensor shutdowns during the interval which resulted in fewer recordings of radiation data than normal. These abnormal trends can be avoided by better O&M practices on-site.
- Degradation trend based on PR analysis: At another plant of 5.8 MW in Maharashtra, we observed that the normalized generation of the plant increased by approximately 2.9% per year, and radiation increased by approx. 6.9% yearly between 2012-2015. For the period 2015-2018, the normalized generation was decreasing at a rate of 2.9% per year, and radiation was also decreasing by 1.8% per year.
From these figures, it is evident that the amount of radiation available for the plants is higher compared to the rate of increase in specific generation over the years 2012-2015 which clearly suggests that the plant is experiencing higher levels of degradation. The same result can be concluded by the decreasing PR (approx 1.5% per year) in the above figure.
Similarly, during 2015-2019, the specific generation is decreasing more rapidly as compared to the rate of decrease of POA radiation which again confirms high levels of degradation in the power plant.
There are various losses like shading losses, incidence angle losses, soiling losses, irradiance losses, thermal losses, ohmic wiring losses, auxiliaries consumption, external transformer losses, and system unavailability losses which can be evaluated remotely. This evaluated data can further be utilized to identify specific components in the system responsible for the degradation and can therefore save time as well as labor to identify and rectify the faults.
- Low plant availability due to nearby disturbances and inverter failure: In a 134.92 MW plant in Karnataka, we observed that normalized PR (expected PR if there was 100% plant and grid availability) was 1.1% lesser than the assumed normalized PR and the inverter efficiency losses were higher by 1.9% than assumed losses. These results suggested that plant availability was much lesser than expected. On further inquiring about the site, we were told that a canal was built between the plant after the installation of the plant which could have resulted in cable punctures which in turn reduced the plant availability.
Also, on further analyzing the inverter efficiency of the plant, we observed that the overall average efficiency of the inverter was 96.22% which was 2.88% less than specified in the datasheet. This suggested that there could be some internal issues with the inverter.
- Poor insulation impacting plant availability: In another plant of 26.3 MWp in Punjab, we observed that the inverter efficiency losses were higher by 1.25% than PVsyst estimation. Also, the plant’s unavailability was higher by 1.32%.
On conducting the breakdown analysis on the plant, we observed that the total downtime was 2.31% out of which SCB contributed to 1.309% time which is very high for any component. We concluded that this must have contributed to high unavailability during the period. Also, the inverter was unavailable for 0.51% of the time which would have also contributed to high unavailability rates. The reason behind frequent SCB shutdowns was insulation breakage which could further be tested by IR testing on the strings. With the combination of these two studies, we were able to identify specific components responsible for high unavailability without physically mobilizing the team and working on the site.
This loss analysis can be easily extended to different nodes like strings, transformers, HT & LT panels, transmission lines, availability-based tariff meters, etc. which could enhance this diagnostics study. The only thing holding us back is the availability of relevant data in the plant.
Force breakdown/unwanted breakdown can occur because of poor device quality, poor installation quality, inadequate cable dressing & loose termination, voltage surges due to lightning, etc. Some of the main reasons behind breakdowns are inverter failure, HT and LT cable termination failure, cable insulation failure, transformer failure, CT and PT failure, circuit breaker failure.Breakdown analysis to identify component failures: In a plant of 14.4 MWp in Tamil Nadu, we observed the overall breakdown period of 0.84% in the plant which is much higher than the accepted period. Major contributors to this were breakdowns in outgoing feeders and auxiliary/control and earth faults in the inverters. We concluded that on the inverter level, there could be breakage in cables or reduction in insulation on the DC side which resulted in multiple inverter tripping scenarios. We recommended to conduct IR testing for the strings and proceed further as per achieved results.
Inverter Efficiency analysis
- Inverter fault rectification to improve generation by 0.7%: In a 5.75 MW plant in Jodhpur, we conducted an inverter efficiency analysis of 7 inverters. We observed that the overall average efficiency of all inverters was 97.6% indicating approx. 1% loss.
One interesting observation was that a particular inverter that had a maximum specific generation was also showing the lowest efficiency of 95.7%. This was leading to a loss of 0.7% in the overall generation. A deeper study into the load v.s. efficiency curve showed an abnormal pattern suggesting that the efficiency of the inverter was reducing as the load increased. A proper root-cause-analysis suggested that there was an increase in internal or contact resistance which was the reason for this abnormal pattern, which was then taken up with the inverter manufacturer for rectification.
In the absence of a detailed desktop study, it would have taken us a significant amount of time to figure out the cause behind the poor generation.
Hence, you can see that remote diagnostics help in saving time and money while helping you in identifying key concerns areas in your plant, and thereby improving generation. Send us your plant data and optimize your analysis and maintenance process.
Source: PV Diagnostics