Most large battery storage projects do not fail due to equipment, they fail due to commissioning being treated as a formality, rather than an ordered, phased approach that holds and verifies each step of the prior phase. It is a phased plan that, when followed correctly, drip feeds energization and operation in a safe and controlled manner.
What’s the real danger of doing it out of order? Industrial electrical battery systems contain enough stored energy that, if they are not energized in the correct sequence, or worse, if a partial energization test is done after a more advanced integration test, enough energy is present to demonstrate a failure for the system under test. The results in the grid test can then be misinterpreted as a successful individual system test, leaving weeks to months of rework to retest and repair the system during warranty.
Why commissioning gets rushed, and what it costs later
Commissioning schedules are shortened on almost every project. The equipment arrives late, the interconnection date is firm, and the EPC has to turn the site over. So you skip a couple of steps, energize before your insulation resistance test is 100% and/or before your commissioning plan is at final draft status. The EPC wants to pre-function the equipment to save time, relieving the pressure on your critical path items as long as you take them at their word that they are keeping an eye on things. Perhaps even more dangerously, they’ll run a few performance tests before your BMS has been properly validated at every level.
Building the pre-commissioning test package
Before anyone puts on PPE and opens a panel, you should have your test package ready, with all tests designed to safely and reliably prove the equipment functions as intended, and those tests organized into an official process. This means someone with responsibility for the delivery of the plant (not necessarily the same company you bought the equipment from) sits down with the factory acceptance tests and works out which ones need repeating, adapting, extending, or skipping. Then that person schedules every test by site location and by date in logical order for minimum rework, ensuring any pre-requirements (cleaning a room, temperature stabilization, pre-powering a different system, etc) are listed. If you turn to a new page of your test package and the “pre-reqs” aren’t already done, you aren’t going to have a successful test that day.
Most teams building this package from scratch waste weeks reinventing checklists that already exist in a more mature form. Platforms like BESSBASE.com give engineers and project managers a curated set of commissioning checklists, test templates, and technical specification references built specifically for battery storage, rather than adapted from generic switchgear or solar commissioning docs. Starting from a proven template and adjusting it for your site’s findings is a lot faster than building the structure yourself, and it reduces the odds of missing a test case that only becomes obvious after something breaks.
The pre-commissioning audit
This is the unglamorous stage that catches the most problems per hour spent. Before any panel gets energized, walk the entire system and verify:
- Mechanical torque on every busbar connection, terminal, and structural fastener against the manufacturer’s torque spec.
2. Insulation resistance testing on every DC and AC circuit, with results logged against a baseline.
3. Cable gland integrity at every enclosure penetration, especially outdoor containers exposed to weather cycling.
4. Grounding continuity from every rack frame back to the site ground grid, confirmed with a low-resistance ohmmeter rather than a visual check.
A loose torque connection or a compromised cable gland won’t show up in a functional test right away. It shows up six months later as an intermittent fault that takes days to trace. Catching it now, while everything is still de-energized and accessible, is the entire point of this stage.
Validating the BMS before paralleling strings
The battery management system is the most important thing to protect everything else in the system, so it has to work right before you start stringing things together. Test at three levels: cell, module, and rack.
At the cell level, make sure that the balancing comes on at the correct voltage and that the over/undervoltage trips come on at the spec values, not whichever ones the parts arrive set to at the factory. At the module level, make sure the temperature alarms come on at the right temp and that the module-level BMS reports the correct state of charge to the rack controller. At the rack level, make sure the rack-level protection logic isolates a faulted string without taking down perfectly good strings that happen to be wired to the same bus.
This is a sequence because if you don’t validate the BMS and protection logic before you start paralleling strings, a fault in one string can propagate into the others before the protection logic even gets a smell at it. So, try to do this testing one string at a time, not as one big batch.
Testing the power conversion system
The PCS is responsible for converting DC battery power to AC power that can be used, and it should be tested for both grid-following and grid-forming operation if the PCS will be used in applications that require one or both. In addition to this, reactive power output helps run the site and should be tested with the PCS under full load rather than merely trusting the specification sheet. Finally, the PCS is likely your first line of defense in staying connected if the grid has a short-duration disturbance simply because it’s less sensitive than a synchronous generator. This should be checked by directly applying faults and seeing if the PCS stays connected.
EMS integration and SCADA validation
The EMS is the brain that makes the real-time trading and dispatch decisions for a battery that big, based on market signals and forecasts. It’s what decides when the battery charges, discharges, or just sits idle, and none of that happens if the EMS isn’t working correctly. You won’t know if the EMS is functional until commissioning time, but by then it’s too late to find that out.
Similarly, you don’t know if your battery can actually deliver 20 MW when the EMS says so, because you’ve never tried that either. Want to commission a large BESS? Verify dispatch setpoint accuracy – if the EMS commands 500 kW, confirm the PCS actually delivers something close to 500 kW, not 460 or 540.
If the clocks in any of the BMS, PCS, or EMS are wrong or drift over time, the root-cause failure analysis you’ll do a couple of months after commissioning will be a nightmare to trace back. Check time-synchronized alarms across all BMS, PCS, and EMS layers.
Grid interconnection and fire safety testing
Coordination for interconnection testing must be scheduled with the utility in advance and call for relay settings to match both in your system and on their side of the fence. IEEE 1547 compliance testing also demonstrates abnormal frequency and voltage ride-through and abnormal grid condition stress testing to prove to the utility that your system responds properly to disconnect and reconnect commands. Have your relay coordination study from the utility finished up before you book this one – inviting the test company before you have that study in hand ensures you’ll overhear the most commonly overheard words of the commissioning effort: “Hurry up and wait.”
Fire protection commissioning is run as a parallel workstream to power commissioning. Instead of crossing your fingers and waiting for a bad day, test that the thermal runaway detection system works as intended by falsely triggering it. Once it indicates stimulation, check the suppression release paths to ensure they are clear, then test that the airflow and gas detection initiate and react in the order they were designed to using the pre-set testing requirements under the thermal propagation testing in UL 9540A and the installation requirements in NFPA 855. If you have one, cross-reference IEC 62933 for additional design and safety requirements. Make that same check with the added gas testing all over again after the entire system is energized for the first time.
Staged energization and performance verification
The capacity test is the last pre-operational check: when you pass that, you’re ready to go commercial. Post-operational warranty provisions typically call for a 72-hour continuous operation test. This is where the engineering team gets to shine, maintaining essential load even while cycling auxiliary loads on and off to test step-load-following, frequency and voltage regulation, and discharge supply ramping.
Prepare your team for a high-stress 72 hours while the vendor’s engineer is evaluating you (by this time they’re as anxious to get the handover over with as you are!); the battery’s biggest operational risk is over-charging or over-discharging trying to compensate for auxiliary system ramping, so you want to hit these transitions with the fastest possible balancing power inputs. Energy systems don’t like being rushed; during your final construction punch list, spend time on your auxiliary system manufacturer’s limits of ramping and on the interface limits that are defined in UL 9540. Some systems are very slow ramping and you will have no end of grief if you discover this for the first time during your continuous operation.
Closing out the commissioning package
Handover isn’t finished until you have red-line ‘as built’ drawings that reflect what’s actually installed not just what was originally designed, a full test report package that covers everything from the pre-commissioning audit through performance verification, and documented operator training that includes alarm response, manual override procedures, and routine monitoring tasks.
The training part is often skipped and it’s not a good corner to cut. An operations team that takes control of a system and has no idea what the alarms are or how to respond will either be continuing to run up the score on nuisance alarms or will be way too blithe about the ones that matter. Informed, appropriate reactions are never going to happen without appropriate training.
The kind of commissioning process characterized by all that – hold points, test cases, energization steps – will take longer up front than an ‘everybody back they’re good to go’ approach pushed by management in a hurry (and given in to by submissive engineers). However, the schedule that doesn’t have time to do something properly is not much different from the one that can’t afford it but the resulting ramp-up period of troubleshooting and bug-fixes is an order of magnitude more painful.
For more information click here.