Solar PV Performance Analysis: 3.1. Meter vs Sum of Inverters

Your solar site has all irradiance and production data accessible and all inverters and strings are available and producing normally. But you’ve still got a “low” performance ratio, what to do?

Performance Ratio (PR) varies from site to site based on module choice, DC/AC sizing, design, interrow shading, installation type (ground mount, rooftop, floating) etc, but generally a reasonable PR is around 80% for a new site. This figure reduces approximately 0.5% per year taking module degradation into account, so as an example, a 10-year-old site should still be achieving around 75% PR levels. A well simulated PV Syst will be a good benchmark to compare an operational site's performance to.

Throughout the year the PR of the site will vary based on seasonal change:

• In winter the sun is lower in the sky so there will typically be more module shading which reduces PR;

• In dry conditions, a lot of dust is created which increases module soiling and reduces PR;

• In a hot summer, the PR will generally decrease as module output reduces with increased module temperature (negative temperature coefficient);

• Inversely during low irradiation days <3kWh/m2/day (e.g. rainy season, humid winter), the PR will increase as modules perform better at lower temperatures.

So an analysis of the PR shouldn’t just be done on a 1 hour or 1 day basis. It should span multiple days (at least 1 month and ideally more) to get an idea of seasonality and how the site is really performing.

Case Study

Coming back to our example (which was recently encountered) ... we have a newly installed solar site, all modules and inverters are new, all strings and inverters are 100% available and all are generating energy normally. There are no alerts indicating any issue with the DC field. But somehow the PR is 55% and causing a major headache.

One critical aspect to check is the comparison between the output at the meter and the sum of the inverter output. As previously mentioned in episode 1, inverter output is normally around 1-2% higher than the meter output, this is because from the inverter to the meter there are AC cable losses so the meter output will be slightly lower.

The delta between inverter and meter output should look something like this:

Here we see that the meter and inverter curves follow each other well during the day and the sum of inverter output is 1.2% higher than the meter (EDMI). This is as expected.

However when comparing the sum of inverter output and the meter output on this site, the meter output was 32% lower than the sum of inverter output!

Clearly something was wrong at the meter. A team was immediately sent to site and it was found that one of the phases at the EDMI meter had blown. So all the inverter production on that phase was just being lost!

The fault was repaired and the site PR jumped instantly from 55% to 80%.. the client was happy! It was a simple solution which had a massive impact on site performance. Also such a fault isn't necessarily picked up by monitoring alerts, so it's a good check to do.

During the next post we’ll go through another potential source of underperformance!