This was quite an interesting fault on an ATESS energy storage system as it initially looked like a PCS fault, but eventually turned out to be a problem inside the bypass cabinet. The site itself worked perfectly while the utility grid was available. The problem only started when the grid failed and the system had to run in island mode.
The PCS would successfully take over the load when the grid supply was removed, but then roughly every two minutes the load would suddenly drop out for a few seconds before reconnecting again. The PCS itself never actually shut down and, even more confusing was that there were no fault codes being generated. One thing I also noticed was that if I isolated the grid supply first and then tried starting the PCS, it wouldn’t start in island mode at all. Instead, it immediately came up with a Grid Phase Sequence Error, which didn’t really make much sense considering there wasn’t even a grid supply connected.
Next thought, while a bit of a shot in the dark as it never presented itself as a DC volt low fault or anything like that on the PCS, was to monitor the battery output both with a multimeter and connecting to the batteries BMS and see what was reported at the time of the PCS shutting itself down and was immediately able to rule that out as a point of failure as this tested fine.
Without any fault codes to work from, it wasn’t really possible to narrow it down much further, so I sent all the test results through to ATESS support. They asked me to completely isolate each SCR module and perform resistance measurements across them. According to ATESS, the resistance should measure greater than 0.3MĪ© or simply read OL (Open Loop) on the meter.
The SCRs in this system weren’t even close. They were measuring down in the kilo-ohm range. Based on those readings, ATESS confirmed the SCR modules had failed.
Although only one module appeared to be causing the issue, I recommended replacing the entire set. It’s much the same as replacing wheel bearings on a vehicle. They’ve all experienced the same electrical and thermal stresses throughout their life, so replacing only the failed unit often isn’t the best long-term solution. What made this fault particularly interesting was how misleading the symptoms were.
Normally, when you think of an SCR failure, you expect something fairly dramatic. In this case, the PCS continued operating normally while grid connected, there were no fault codes, and the fault only appeared when the system was forced into island mode. The failed SCRs were effectively allowing the PCS to detect a bypass voltage that shouldn’t have existed. Because the PCS believed voltage was still present on the bypass input, it resulted in repeated transfer attempts which caused the brief interruptions to the load.
This is definitely not a fault I’ve come across very often, but it’s one I’ll remember.
If you ever come across an ATESS system that operates normally while grid connected, drops the load every few minutes in island mode, indicates bypass voltage with the grid isolated, reports a Grid Phase Sequence Error when started without utility or has no obvious PCS fault codes it’s well worth checking the SCR modules in the bypass cabinet. It could save you spending hours chasing a fault elsewhere in the system.