COD — commercial operation date — is a legal event. Commissioning is the months of engineering that earn it. It’s also where every earlier decision gets audited by reality: the EMS integration you deferred, the aux power design nobody reviewed, the capacity guarantee someone signed without defining the measurement boundary. Here’s the sequence, and where projects actually get stuck.

For a visual companion, explore the interactive PCS Dispatch Efficiency diagram on BESS.Engineer.

FAT: catch it at the factory

Factory acceptance testing happens before shipment — on cells, DC blocks, and PCS units. Typical scope: capacity and internal-resistance sampling on cells, functional tests of container systems (BMS, cooling, fire detection), insulation and functional tests on PCS. Two pieces of hard-won advice:

  • Witness (or send a third party to witness) FAT for at least the first units of each type. A defect found in the factory costs a retest; the same defect found on site costs a schedule.
  • FAT data is your beginning-of-life fingerprint. Archive it — cell capacity distributions, impedance data — because warranty arguments a decade from now will reference it.

Site: from mechanical completion to energization

On site, work moves through mechanical completion (everything installed, torqued, terminated) into pre-energization testing: insulation resistance and hipot tests, point-to-point wiring verification, protection relay testing with secondary injection, CT/VT polarity checks, grounding verification, and functional checks of every safety interlock. This is conventional power-plant commissioning discipline applied to hundreds of identical containers — the volume is the challenge, so checklist rigor and sampling strategy matter.

Don’t forget the fire code has a seat at this table: NFPA 855 expects a documented commissioning plan and acceptance testing for the storage system, and your AHJ may want to witness key steps — the fire safety article covers what they’ll be working from.

Energization — backfeeding the site from the grid — is a formal, permitted event coordinated with the utility. From there, hot commissioning begins: first charge and discharge on each block, control-loop tuning, verification that EMS, PPC, and SCADA actually do what the point list promised.

Grid compliance testing

Every serious grid now requires demonstrated compliance with connection standards before full operation — frequency and voltage ride-through, active and reactive power control, response times. In Australia’s NEM this is a formal staged process against the plant’s Generator Performance Standards, with hold points as the plant progressively proves itself; the Waratah Super Battery’s registration and testing campaign is the highest-profile recent example. If your plant is grid-forming, expect additional model validation and testing — the GB compliance process pioneered at Blackhillock under GC0137 is the template others are studying.

The single most common cause of compliance-test pain in my experience: the plant-level controller. Individual inverters pass; the coordinated plant response at the POI doesn’t match the registered model. Budget tuning time.

The capacity test: where the money is

The capacity test proves the contracted usable energy. Everything about it should be defined in the contract before anyone builds anything:

  • Boundary: measured at the POI revenue meter, not at PCS terminals.
  • Conditions: reference temperature, aux loads included, state of charge window as operated.
  • Protocol: typically full charge, a rest period, then discharge at rated power to the minimum operating SoC, with energy integrated over the discharge.

Worked example: the contract guarantees 400 MWh usable at the POI. The test discharges at 100 MW and sustains it for 3.96 hours before hitting minimum SoC: measured energy = 396 MWh, or 99.0% of guarantee. Whether that 4 MWh shortfall means a cure period, liquidated damages, or a price adjustment is a question your supply contract answered years earlier — or didn’t.

And the COD test is not the last one: capacity tests typically recur through project life — annually or per warranty and market-qualification requirements — with the COD result serving as the baseline every later dispute references.

Round-trip efficiency test: meter the AC energy in during the charge and the AC energy out during the discharge at the same boundary. If the plant absorbed 440 MWh and returned 396 MWh, RTE = 396 ÷ 440 = 90.0%. Ambient temperature and aux consumption move this number, which is why the reference conditions matter — a point I’ve hammered in the sizing article.

Trial operation and COD

Most contracts end with a reliability run: continuous or near-continuous operation for a defined period (commonly measured in days to a few weeks) meeting an availability threshold, with defined rules for what interruptions reset the clock. Pass it, close the punch list to an agreed level, hand over documentation — as-builts, settings files, cyber security baselines, O&M manuals, warranty registrations — and COD is declared.

Where commissioning schedules actually die

  1. EMS/PPC/SCADA integration. The perennial number one. Hundreds of data points, three vendors, one point list nobody owns.
  2. Auxiliary power design. Undersized or poorly sequenced station service that can’t run cooling during commissioning modes.
  3. Firmware version chaos. Containers arriving across months with different BMS firmware; align versions before performance testing.
  4. Communications. Fiber breaks, mislabeled patches, timeouts — mundane and merciless.
  5. Test windows. Grid operators grant compliance-test windows; miss one and the next may be weeks away.

FAQ

How long does BESS commissioning take? From first energization to COD, commonly two to six months for a utility-scale site depending on size, grid-compliance complexity, and how much integration was genuinely finished before energization.

What is a BESS capacity test? A contractual demonstration that the system delivers its guaranteed usable energy, measured at a defined boundary under defined conditions — typically a full discharge at rated power with energy integrated at the revenue meter.

What’s the difference between FAT and SAT? FAT proves equipment in the factory before shipment; SAT (site acceptance testing) proves the installed, integrated system on site. Both matter; neither substitutes for the other.


My course walks the full commissioning sequence with real test protocols and the contract language that should sit behind them — Grid-Scale BESS: Complete Guide.