DC Bus Overvoltage: 196 Spikes, One PCS Unit
Ingro identified 196 DC bus overvoltage incidents at a grid-scale BESS deployment in India, where voltage spiked from a nominal 720–750V to as high as 881V, well above the safe threshold of 840V. Here's how we diagnosed and resolved it.
6 slides — scroll or use the arrows
At one of India’s earliest first-of-its-kind grid-scale BESS deployments, Ingro EMS logged 196 DC bus overvoltage events on a single PCS unit. Voltage spiking from a nominal 720–750V to as high as 881V. The safe operating ceiling: 840V.
We looked at the data and here is what the data actually showed.
The fault
One PCS unit at a grid-scale BESS deployment.
- DC bus voltage — spiking from 720V to 881V.
- Safe ceiling — 840V.
- Trip events — 196.
- Shutdowns in a single day — 57.
Each shutdown = node offline = lost grid service revenue.
One data point that ruled out the obvious
Every time voltage spiked — current dropped to zero. Simultaneously.
- Short circuit: Voltage ↓ | Current ↑↑↑
- What we saw: Voltage ↑↑↑ | Current = 0
These are physically opposite signatures.
This is not an isolated problem
NERC 2022 California BESS Disturbance Report: 622 MW of BESS capacity lost to DC bus overvoltage in a single incident.
Root cause? Unknown. Reason: none of the affected facilities could perform sequencing-level root cause analysis.
Ingro EMS could. That is what made the diagnosis possible.
After diagnostic testing — three most probable causes
1. DC bus capacitor degradation
Voltage-smoothing components degrade with age and heat. In Delhi’s conditions — faster than rated lifetimes suggest.
2. Precharge circuit failure
The startup sequence may have broken down. On most modern PCS platforms, a software-level compensatory fix exists. Corrective options constrained by OEM firmware on this platform.
3. AC contactor timing mismatch
Wear causes timing drift. Even milliseconds matter at these voltages. This unit: 2.5x more switching operations than its healthy peer.
A voltage spike is easy to see
DC bus overvoltage is not a fringe fault. It is a systemic risk — and at scale, the consequences are severe.
Knowing why it happened requires data most deployments don’t have.
Opens the original post on LinkedIn — like, comment, or share it there.





