How an electrical mistake can cause 10% energy loss
By Paul Grana, co-founder, Folsom Labs
East-west arrays are becoming increasingly common for flat-roof applications. These systems offer greater packing density – fitting more modules into a space-constrained roof to maximize the total energy generated. However, this new design topology comes with one important additional requirement: the modules on the east and west orientations need to be wired to different inverter MPP zones, or different inverters entirely. This can slow down the electrical installation process, and more importantly, it creates opportunities for installation mistakes. These mistakes – specifically, when east- and west-facing modules are wired in the same series circuit – can drop the array’s energy yield by 8-10%.
Why east-west wiring is hard to get right (and even harder to diagnose)
In east-west arrays, series circuits need to skip a row at the end of each run as adjacent rows are facing opposite directions. Skipping rows becomes doubly challenging when the rooftops are irregular (see the example below for a simple illustration of this). While these wiring layouts are fine in the hands of a capable installation crew, it can lead to mistakes if the crew is less experienced, or simply moving too fast.
Incorrect wiring can be difficult to detect in the commissioning process. Open-circuit voltage checks will not show potential problems at all, as the string Voc will be identical whether the stringing was done correctly or not. An IV curve trace will only sometimes identify the problem. While the average energy loss is 6-10%, the losses will depend significantly on the time of day – when the sun is directly overhead, an incorrectly-strung array will look exactly like a properly-strung array.
So how bad are the losses?
Mismatch losses from an incorrectly-strung east-west array at a 10° tilt are roughly 8% over the course of the year. The tilt of the modules drives most of the mismatch losses. If the modules are tilted at 8° east & west, the mismatch losses drop to 6%, while a 15° tilt will result in 13% mismatch losses. Additional smaller differences can be seen based on the modules used (~1% lower losses for more mismatch-tolerant modules), the building orientation (~1% lower losses if the building isn’t perfectly east-west), and location (1-2% larger losses at higher latitudes).
It is important to note that mismatch losses will vary a lot based on the time of day and time of year. In the example below, while the annual mismatch losses are 8%, the mismatch ranges from 1% or less from 11am-1pm (when the sun is overhead), to 15-25% in the morning & afternoon.
What can be done?
There are a few steps installers can take to avoid these problems:
Design for Installability. Use design blocks (such as “Frames” in HelioScope) to reduce the complexity of the installation by reducing the number of irregular stringing runs.
Take extra care during installation. This is a case where extra planning and thoughtfulness on-site can eliminate (or dramatically reduce) the odds of running into this problem. This might also include color-coding the conductors for the two module faces, or even wiring up one set of modules before installing the second set of modules.
Upgrade the commissioning process. Since a failed string might only be 5% off from a healthy string (depending on the time of day), extra case should be taken when tracing the string IV curves. And because any string could be at risk of being mis-wired, it may make sense to trace each source circuit, rather than a sample.
Run IV curve traces before 10am or after 2pm. Running IV traces in for the morning or evening when the sun is significantly stronger on one part of the array versus the other is the only way to spot errors in wiring. Otherwise, there won’t be any mismatch at high noon, even in an incorrectly-wired system.
East-west racking can significantly improve the economics of rooftop solar in space-constrained applications. But as with any new technology, we must be careful to apply it correctly.