A large-scale PV project’s transition from construction to operations is one of the busiest, highest-stakes stretches of the entire project timeline. Stakeholders have to coordinate schedules, materials, trades, troubleshooting and testing while holding the line on design documents, contractual requirements and project milestones — all racing toward the commercial operations date (COD), the point at which the asset finally starts generating revenue.
That challenge isn’t new. Back in 2018, independent engineers Anastasios Hionis and Mat Taylor wrote an influential piece for SolarPro magazine, “Achieving Commercial Operations in Large-Scale PV Power Systems,” distilling lessons learned from steering utility-scale solar projects to COD — some of them smooth, some of them “unmitigated disasters” that left every stakeholder frustrated and financially exposed. Their core message still holds up: a collaborative, transparent performance-evaluation process gets high-value energy assets across the finish line far faster than an antagonistic one.
What has changed is almost everything around that process. The codes have been rewritten, the interconnection rulebook has been replaced, the test standards have been revised, and most commercial-scale projects now pair PV with battery storage — which means COD is no longer a single test, it’s two. Here’s how to apply that original playbook to a 2026 project.
What Is the Commercial Operations Date (COD), Really?
The commercial operations date is the contractually defined moment a PV (or PV-plus-storage) plant has demonstrated it performs as promised and is cleared to start generating revenue. Getting there requires the EPC, owner, financier, SCADA integrator, inverter vendor and independent engineer to agree — usually under time pressure — that the plant was built correctly, is wired and mapped correctly, and produces the energy yield the contract promised.
The mechanism for proving that is performance testing: benchmarking system capacity, efficiency (performance ratio) and energy yield against a set of contractually mandated targets. Done well, it’s a fast, evidence-based process that also produces the operations team’s baseline data. Done poorly — with unrealistic expectations, proprietary “black box” models, or one-sided contract terms that require chronic overperformance just to pass — it becomes exactly the kind of conference-table (or courtroom) standoff the original article warned about.
The Critical Negotiations: Get These Agreements in Writing Early
The 2018 article’s central recommendation remains the single best piece of advice in this space: settle testing methodology, risk allocation and definitions at project inception, not during the sprint to COD. A handful of topics come up on nearly every project and should be nailed down early.
Build One Testing Model Everyone Agrees On
The performance test needs its own energy model — related to, but distinct from, the annual energy model — tuned to reflect actual conditions at the time of testing. It has to be dynamic enough to adapt as design, implementation and site conditions change, and the contract language should drive the modeling assumptions rather than the other way around.
Don’t Pretend Measurement Uncertainty Doesn’t Exist
Every operational measurement carries uncertainty. Ignoring it — or assuming it always favors the owner — is inequitable. A 2% measurement-uncertainty allowance is a reasonable starting point, to be refined once equipment and the test plan are finalized.
Assign Module Power-Tolerance Risk to Whoever Controls It
Whichever party procures the modules should own the risk tied to nameplate power tolerance, since that party controls how much positive power tolerance gets backstopped.
Build In a Real Soiling Allowance
Zero-percent soiling is a myth, especially over a multi-day or multi-week test window. Contracts need either direct soiling measurement at test time or a documented estimate based on the wash cycle — and it should be assessed before testing starts, not argued about after a shortfall.
Use Open, Standards-Based Test Methods
Keeping methods and evaluation tools proprietary doesn’t protect anyone — it just slows troubleshooting when a test result comes in low. An open-book approach, where every at-risk party can review the assumptions, inputs and formulas, resolves disputes faster and builds trust across the closeout team.
Test Preparation: From Pre-commissioning to Sign-Off
Planning starts with a clear-eyed read of the contract and performance-test requirements, since that’s what determines SCADA specifications, sensor placement and documentation strategy long before crews are on site — the same planning that determines whether the commercial operations date holds.
- Pre-commissioning. Assemble the full closeout team, generate the testing documents, build a testing model separate from the annual model, and verify the SCADA and sensor installation plans against the applicable test standard.
- Start-up and commissioning. Commission and validate SCADA and sensor accuracy, troubleshoot inverters and field wiring, close out punch-list items, and run practice tests before the clock starts on the real one.
- Performance testing. Run the test, disseminate data in real time, compare evaluations, and determine results. Given reasonable weather and a transparent process, the outcome should be unambiguous — pass, or a clearly identified reason for failure.
A pre-populated commissioning folder — contracts, the testing model, the technical standard, combiner-box as-builts, calibration certificates, SCADA log-in credentials, and the permission-to-operate paperwork — turns project closeout from a fire drill into a checklist.
Sensors and SCADA: Where Good Projects Still Go Wrong
Reliable, accurate measurement of primary data — irradiance, power, temperature — is what performance assessment stands or falls on, whether the plant is being tested or already operating.
Irradiance Sensors
Pyranometer misalignment is still the most common cause of a plant appearing to underperform when it isn’t. A plane-of-array (POA) sensor that isn’t mounted at the array’s actual tilt, or a global horizontal irradiance (GHI) sensor whose bubble level isn’t centered, will quietly skew results — and on single-axis trackers, checking that POA and GHI readings converge at solar noon is a fast sanity check on tracker angle, sensor installation and SCADA scaling all at once.
Power Measurements
Utility meters, check meters and inverter output data aren’t automatically accurate. Understand each meter’s accuracy spec and roll-up method before test day, not while the clock is running.
Temperature Sensors
Ambient temperature sensors tend to be reliable if installed correctly. Back-of-module (BOM) measurements are a poorer proxy for the array as a whole, and thermal-loss models built on BOM data still aren’t mature enough for evaluations where a tenth of a percent can be worth hundreds of thousands of dollars — a caution that remains just as true today.
Documentation: The Closeout Package That Protects the Asset
A standards-based performance-test package is effectively the factory acceptance certificate for a fielded PV plant. Owners, asset managers and future buyers will lean on it for years, and incomplete documentation is consistently the difference between a smooth transaction and a discounted one. At minimum, the closeout folder should include the contracts and addenda governing the test, the testing model and all assumptions, the applicable technical standards, combiner-box as-builts, inverter and string mapping, calibration certificates, SCADA credentials, and the permission-to-operate documentation — the paperwork every commercial operations date ultimately rests on.
Battery Energy Storage: The Parallel Path to Commercial Operations
The single biggest structural change since 2018 is that a large share of commercial and utility-scale projects reaching their commercial operations date today aren’t PV-only — they’re hybrid. Adding a battery energy storage system (BESS) doesn’t just add hardware; it adds an entire second commissioning track that has to be planned, scheduled and closed out alongside the PV performance test, not bolted on afterward.
Factory and Site Acceptance Testing (FAT/SAT)
Where PV commissioning leans on performance testing, battery commissioning leans on factory acceptance testing (FAT) at the manufacturer and site acceptance testing (SAT) once racks are installed and energized. FAT confirms the system was built and configured correctly before it ever ships; SAT confirms it survived transport and integration intact and communicates properly with the site’s SCADA and protection systems. Our breakdown of FAT and SAT procedures for battery systems walks through the specific checkpoints — cell balancing, insulation resistance, communication protocol verification and thermal management checks — that both tests need to cover.
Capacity and Round-Trip Efficiency Verification
Just as a PV plant’s performance test benchmarks capacity and yield against contractual targets, a BESS needs its own acceptance test benchmarking usable capacity, round-trip efficiency and response time against the specification the owner is paying for. These figures degrade over time even under normal operation, so the acceptance test also becomes the baseline every future degradation and warranty claim gets measured against — which makes accurate, well-documented baseline testing at COD just as valuable for storage as it is for the PV array.
Safety Certification: UL 9540A, UL 1973 and UL 9540
Battery systems carry a safety-certification burden that PV modules generally don’t. UL 1973 covers the battery and rack construction itself; UL 9540A evaluates thermal-runaway propagation and fire behavior; and UL 9540 certifies the energy storage system as a whole. See how MEGATRON’s commercial ESS racks and systems are certified to these standards — documentation the closeout team should have in hand well before the performance-test window, since a missing certificate can stall COD even after the batteries themselves pass every functional test.
BMS Integration and Continuous Monitoring
A battery management system (BMS) is doing, for the storage asset, roughly what SCADA and irradiance sensors do for the PV array: providing the cell-level voltage, temperature and state-of-charge data that both proves the system is healthy at COD and protects it in operation afterward. That data needs to be mapped into the same central, transparent data repository recommended for PV performance testing — not siloed in a separate vendor portal — so the same closeout team troubleshooting an inverter can just as easily troubleshoot a battery rack.
Aligning the Two Commissioning Tracks
The practical takeaway is scheduling. Battery FAT typically happens off-site, months before mechanical completion, while SAT and capacity testing need to land inside the same closeout window as the PV performance test. Building both tracks into a single master schedule — with a tiger team that includes storage-specific expertise from day one — is what keeps a hybrid project’s COD from getting held up by whichever system happens to test second.
What’s Changed Since 2018: Updating COD for Today’s Projects
The fundamentals above haven’t aged. The regulatory and technical landscape around them has moved substantially, though, and a 2026 closeout team needs to account for several developments the original article couldn’t have anticipated.
Interconnection Now Runs Through IEEE 1547-2018
The interconnection standard referenced implicitly in 2018 has since been superseded by IEEE 1547-2018, which overhauled voltage and frequency ride-through, grid-support functions and reactive power requirements for distributed energy resources, with Amendment 1547a-2020 adding further flexibility around abnormal operating performance categories. Utility interconnection agreements — and the ride-through settings validated during commissioning — should be checked against the current revision rather than the pre-2018 baseline.
The National Electrical Code Has Moved Through Two More Cycles
NEC 690.8(A)(1)(2), the maximum-current calculation method the 2018 article’s sidebar highlighted as a source-circuit sizing improvement, is now a well-established part of the code. Since then, the 2023 NEC cycle has reworked rapid-shutdown requirements under Article 690.12, and Article 705.12 now governs how solar-plus-storage systems interconnect on the load side of the service, including the widely used 120% busbar rule. We covered what that means for hybrid projects in NEC 705.12: What the Solar-Plus-Storage Boom Means in 2026.
The Test Standards Themselves Have Been Revised
ASTM’s performance-test standard, cited in 2018 as E2848-13, has since been reapproved as ASTM E2848-13(2023). IEC’s monitoring standard has gone further: IEC 61724-1:2021 is now a full second edition, with updated data-quality and monitoring-system requirements that closeout teams should be testing against rather than the first edition referenced in older contracts. Sandia’s PV Performance Modeling Collaborative remains a solid, vendor-neutral reference for the modeling assumptions behind any testing model.
Drones and Remote Diagnostics Now Do Work Handheld Tools Used To
Aerial thermography and automated IV-curve tracing have gone from novelty to standard practice for locating underperforming strings and modules across large arrays in a fraction of the time handheld tools require, which shortens the troubleshooting cycle that used to eat into the run-up to a performance test (see this overview of drone-based thermal inspection).
1500V Architectures and Bifacial Modules Change the Performance Model
1500V DC architecture, still an emerging option in 2018, is now the utility-scale default, changing string counts, combiner-box design and balance-of-system cost assumptions that feed the testing model (background on the evolution of 1500V plant architecture). Bifacial modules have followed a similar path from niche to mainstream, which means testing models now need to account for rear-side gain using frameworks like those documented by Sandia’s bifacial characterization and rating work — a variable that simply didn’t exist in most 2018-era testing models.
Strategies for a Smoother Path to the Commercial Operations Date
Updated for today’s projects, the closing recommendations from the original article still frame the right approach:
- Say no to secrets. An open, transparent testing and closeout process finds solutions faster than a proprietary one — full stop.
- Centralize data. One repository, accessible to every at-risk party, should contain everything a completely uninformed third party would need to validate the test from scratch.
- Establish a tiger team. Assemble owner, EPC, SCADA, inverter, storage and independent-engineering representatives at project inception, and keep membership continuous through COD.
- Keep a backup squad ready. Projects approaching COD can’t afford to wait for a vendor to assemble an ad hoc troubleshooting team — have engineers and field personnel on call for the performance-test window.
- Plan the storage track separately, in parallel. Don’t let battery FAT/SAT and safety certification become the item that stalls an otherwise-passing PV performance test.
Frequently Asked Questions
What is the commercial operations date (COD) for a solar project?
COD is the contractually defined date on which a PV or PV-plus-storage plant has passed performance testing and is cleared to begin generating revenue for its owner. It marks the transition from construction to operations.
How long does PV performance testing take?
It varies by contract and system size, but most utility-scale performance tests run from several days to a few weeks, since the testing model needs enough data points — under acceptable weather and soiling conditions — to produce a statistically defensible result.
What’s different about COD for a solar-plus-storage project?
Hybrid projects effectively run two commissioning tracks in parallel: the PV performance test described above, plus battery factory acceptance testing (FAT), site acceptance testing (SAT), round-trip efficiency verification and safety certification (UL 9540A, UL 1973) for the storage system. Both have to close out before the full asset reaches COD.
Final Thoughts
The technology, the codes and the standards have all moved on since 2018, but the underlying lesson from the original SolarPro article hasn’t: projects that treat performance testing as a shared, transparent problem-solving exercise reach their commercial operations date faster and with far less financial risk than projects that treat it as a contractual weapon. Build the tiger team early, agree on the testing model and risk allocation before ground is even broken, and plan for storage as its own commissioning track rather than an afterthought — and COD becomes a milestone to hit, not a standoff to survive.
Whether you’re bringing a rooftop commercial array like our Walmart rooftop project or a ground-mount system like the Atlas Series 100kW commercial PV installation to commercial operations, Symtech Solar’s EPC team can help you plan a closeout process built around today’s codes, standards and hybrid-system requirements.