Inside Operations

Solar Monitoring & Production Verification

The step between an authorised system and a funded one. Monitoring provisioned to the right owner, per-device reporting confirmed, early output checked against the design, and the evidence a final draw asks for.

Why this is where projects sit unfunded

Permission to operate is the visible milestone, and once it lands attention moves to the next project. But a financier's final-draw conditions commonly require confirmation that the system is actually producing, and a monitoring account that was created under an installer's email address, reporting a site total nobody has looked at, is not that confirmation.

So the pattern repeats: a project that is physically complete, inspected, authorised, and waiting on a screenshot. The same gap has a second cost further out — partial equipment failures that a site total hides for a year, discovered when the parties who owed the fix are harder to reach and the production is already lost.

Two different jobs, usually confused

Provisioning is a commissioning task. It happens once, on the day, and it decides whether the record that follows is attached to the right property and the right person. Verification is a closeout task: comparing what the system produces against what the design said it would, and turning that into the document a financier will accept.

Neither is service coverage, and that boundary is worth keeping sharp. Monitoring tells you a system is underperforming; it does not tell you which obligation is engaged or who will attend. That is a separate question, answered in warranty and service. What monitoring does is convert a problem the homeowner would have found in a year into one somebody can act on this week.

The long-form treatments of both halves sit on the blog: who is responsible for solar monitoring after PTO on the ownership question, and who pays when a system underproduces on where accountability lands once a shortfall is real.

Six steps, in order

The first two happen at commissioning and cost minutes. The rest are closeout, and every one of them is harder a month later than on the day.

Provision at commissioning

The account is created against the correct property address and the correct owner of record, not against whichever email was to hand on the roof. Getting this right on the day is trivial; correcting it later means a support ticket with a manufacturer.

Confirm per-device reporting

Every inverter, and every module where the equipment supports it, reporting individually. A site total conceals exactly the partial failures that go unnoticed for a year and look like weather in a monthly figure.

Record the design reference

The estimate, and the site inputs it was built from, filed with the project. Without the reference, a later shortfall conversation has no baseline and becomes an argument about recollection.

Verify early output

Initial production compared against the estimate with season and weather allowed for. The purpose is not to prove the estimate right but to catch a system that is authorised and not actually producing.

Produce the funding evidence

The production record in the form the financier's final-draw conditions ask for. This is the item most often missing from a physically complete project whose last draw has not landed.

Transfer to the homeowner

Account access handed over with the homeowner confirming a working login, so the record survives a service call, a roof replacement and a sale of the home.

Why a site total is not enough

The failure mode worth designing against is not a system that stops. A system that stops gets noticed, because the homeowner's bill moves. The expensive failure is partial: one failed optimiser, one string offline, one microinverter that quietly stopped reporting. Each takes a slice off the total that is smaller than the difference between a cloudy month and a clear one.

Nobody investigates weather. So the shortfall persists, and it persists specifically through the window in which somebody still owes a fix — the manufacturer under a product warranty, the installing contractor under a workmanship warranty. Per-device visibility, where the equipment provides it, is what makes the difference between a warranty claim and a lost year.

The same argument applies to the design reference. A shortfall conversation without the original estimate and the site inputs behind it has no baseline, and an estimate built on imagery rather than measured shading is a different problem from an installation defect. Both produce the same complaint and they have different owners, which is exactly why the reference has to be filed at the time rather than reconstructed under pressure.

Where it sits in the sequence

After inspection and interconnection, not instead of them. The final inspection closes the permit and interconnection produces the authorisation to operate. Only then is there anything to monitor, which is why this work reliably lands at the point where everyone assumes the project is finished.

Inside closeout, as one of nine items. Production evidence and the monitoring handover sit alongside the as-built set, the serial number record, warranty registration, punch list closure and lien waivers. What closeout consists of sets out the full document set, and the PTO and interconnection checklist carries monitoring as a tracked item rather than an assumption.

Before the last draw, not after it. Which is the commercial point. Final-draw conditions are set by the financing product, and milestone funding means the money arrives against evidence rather than against completion. A project waiting on production data is working capital sitting still, and the cash conversion cycle is where that shows up.

And it does not end at handover. On a third-party-owned system the asset owner carries production risk for the whole term, which is why TPO arrangements have heavier evidence requirements than a loan. On an owned system the obligation to watch belongs to nobody by default, and that default is the problem.

Where Seamless Home fits

Seamless Home is a licensed contractor. It stands between the companies that sell home energy systems and the crews that install them, and monitoring provisioning and production verification are part of the inside operations it runs on their behalf — alongside design and permitting, inspections and funding operations. Installation is performed by vetted installing partners engaged as its subcontractors.

Concretely: the account is provisioned to the property and the owner of record at commissioning rather than to whoever was on the roof, per-device reporting is confirmed where the equipment supports it, the design estimate and its site inputs are filed with the project, early output is checked against that reference, the production evidence is assembled in the form the financing product's final draw asks for, and account access is transferred to the homeowner with a login they have confirmed works.

What this is not is a service contract. Diagnosing a fault, dispatching a technician and administering a claim are obligations of the parties that carry them, and which party that is depends on the component. Coverage is confirmed per service area rather than promised as blanket availability.

Frequently Asked Questions

What does production verification on a solar project actually involve?+

Confirming that the system that was authorised is reporting, and that what it reports is consistent with the design that was sold. In practice that is four checks: the monitoring account exists and is provisioned to the right property and the right owner; every inverter and, where the equipment supports it, every module is reporting rather than just the site total; the output over an initial window is consistent with the design estimate once weather and season are allowed for; and the account access has been handed to the homeowner with credentials that work. None of it is difficult. All of it is easy to skip, because the visible milestone — permission to operate — has already passed by the time anyone would notice.

Why does a funding milestone depend on monitoring?+

Because for a financier the monitoring record is the cheapest available proof that the asset it funded exists and works. Final-draw conditions on residential PV solar commonly include confirmation that the system is energised and producing, and on third-party-owned systems the owner's exposure runs for twenty years or more, so its evidence requirements are heavier still. An installation that is physically complete, inspected and authorised can still sit unfunded because nobody produced a production screenshot, which is an expensive way to discover that closeout is a document exercise.

Who owns the monitoring account after permission to operate?+

It should be the homeowner, and by default it frequently is not. Accounts are created during commissioning by whoever was on site, under whichever email address was convenient, and they stay there. The consequence surfaces later: a homeowner who cannot see their own system, a service technician who cannot get diagnostics without contacting a company that may no longer exist, and a sale of the home where the monitoring does not transfer. Transferring the account at closeout, with the homeowner confirming they can log in, is the whole fix.

Is a system total enough, or does each inverter need to report?+

A site total will hide the failure mode that actually costs money. A single failed optimiser, a string offline, or one microinverter that stopped reporting reduces output by a fraction that looks like weather in a monthly figure. Nobody investigates weather. Partial failures are the reason systems underproduce quietly for a year or more, and per-device reporting — where the equipment provides it — is what makes them visible at the point they are still somebody's obligation.

How is a production shortfall distinguished from a bad estimate?+

By separating the reference from the result, which has to be done before anyone is asked to pay for anything. A design estimate is a modelled expectation, not a guarantee, and it is only as good as the site inputs behind it — shading measured on site rather than from imagery, correct pitch and azimuth per plane, the right consumption baseline. A system can produce exactly what a correctly built installation should produce and still fall short of an estimate that was optimistic. Establishing which of the two happened is a documentation question, and it is far easier to answer when the design inputs and the first months of data both sit in the project file.

Does monitoring count as service or warranty coverage?+

No, and the distinction is worth keeping sharp. Monitoring tells you a system is underperforming. It does not tell you which obligation is engaged, it does not dispatch anyone, and it does not pay for anything. What it does is convert a problem the homeowner would have found in a year into one somebody can act on this week, while the parties who owe the fix are all still identifiable. Coverage itself is a separate subject with separate answers.

Does Seamless Home handle monitoring provisioning for its partners?+

Yes, as part of inside operations at closeout. Provisioning the account against the correct property and owner, confirming per-device reporting rather than a site total, checking early output against the design estimate, assembling the production evidence a final draw requires, and transferring account access to the homeowner with a confirmed login are handled alongside inspections, interconnection and funding operations. Ongoing service dispatch and warranty administration are separate obligations owed by the parties that carry them. Coverage is confirmed per service area rather than promised as blanket availability.

Stop leaving the last draw on a screenshot nobody took

Seamless Home provisions monitoring at commissioning, verifies early output against the design reference, and assembles the production evidence a final draw asks for — as part of inside operations.

Get In Touch