PV Commissioning Tips and Best Practices: Page 3 of 5
Inside this Article
Allocate resources wisely. When implementing the test plan, consider how you will allocate resources and look for opportunities to streamline the workflow to improve operational efficiencies. Some tests are a one-person job, meaning that multiple people can perform these tasks at various locations in the system. Other tests are better suited to a tag-team approach, meaning they are most efficient when two or more people work together. Some tests, for example, require that technicians take readings at multiple locations simultaneously. In this scenario, it is best to designate one person as the lead documenter. Both technicians can still take notes, but one person is specifically responsible for ensuring that all tests are completed and documented properly.
All else being equal, we recommend a workflow that starts with independent testing activities designed to identify the most obvious potential issues and then transitions to simultaneous testing activities after team members establish a rhythm. Testing location is another consideration. Teams can lose a lot of time when they have to travel from a rooftop array down to a basement panelboard and back to the roof again. Look for opportunities to minimize downtime by optimizing test activities around specific locations.
Know your targets. To avoid unnecessary callback visits, commissioning technicians need to verify that performance test data are within expected ranges. If the measured data do not make sense based on the anticipated results, it is important to determine whether something went wrong during the test process or whether something is wrong with the system. It is much easier to determine the root cause of an unexpected measurement in the field than back in the office.
When gathering performance test data in the field, make sure that you are documenting all of the requisite information. If you find outliers or suspect values in the data, verify that the measurement is representative of the system and not a problem with the testing tools or methods. If an I-V curve looks strange, run another trace. If string voltage measurements do not make sense, make sure that the multimeter is not accidentally set to measure ac voltage.
In most cases, a quick investigation will turn up the cause of an erroneous or out-of-range measurement. If the problem is indeed an installation error, do your best to identify the nature of the problem in the commissioning notes. For example, you can add a lot of value to a commissioning report by noting that someone misidentified and incorrectly terminated a pair of conductors rather than simply reporting that you did not measure any voltage on strings 1 and 2.
Verify performance. The simplest performance verification tests start with the nameplate power rating of the system and calculate the effects of real-world irradiance and temperature measurements, as well as the estimated system-level efficiency. The aforementioned Gleason article outlines a five-step performance verification process that calculates expected power (PE) based on the following equation:
PE = PSTC × KI × KT × KS
where PSTC is the nameplate rating of the array under standard test conditions, KI is the irradiance factor, KT is the module cell temperature factor and KS is a system derating factor.
It is not difficult to calculate the irradiance, temperature and system derate factors. To find KI, simply divide the measured irradiance by the irradiance at STC (1,000 W/m2). To estimate KS, multiply system-level efficiencies together to account for power tolerance, soiling losses, age of system, inverter efficiency, and ac and dc wiring losses. The calculation for KT is slightly more involved:
KT = 1 + (CT × (TC − TSTC))
where CT is the module temperature coefficient, TC is the measured cell temperature and TSTC is the cell temperature under STC conditions (25°C).
The SunSpec Alliance’s best practices guide, “Commissioning for PV Performance” (see Resources), details an initial commissioning-capacity test method, known as the power temperature coefficient model, similar to Gleason’s method. To evaluate inverter- or system-level performance in this fashion, technicians require accurate plane-of-array and module temperature measurements. Teamwork is helpful and sometimes required to capture these measurements simultaneously with a power-output reading.
While these types of instantaneous performance tests are relatively straightforward to execute and reasonably accurate, the process does require concentration and attention to detail. Rather than running numbers while taking measurements on the roof, have a partner sit down with a pencil and calculator. If you use a spreadsheet to automate the process, technicians simply have to enter field measurements instead of performing calculations manually.