Why a Solar Installer's Permitting Delay Is a Document Problem, Not a Jurisdiction Problem

The county is a convenient defendant

Your sales team closed the job in March. The homeowner signed, the financing cleared, the crew has an open week. And the project is sitting in a plan review queue at a county you have filed in a hundred times, waiting on a correction letter that will ask for something you have supplied before.

Ask anyone in the office why and you will hear about jurisdictions. Every county wants something different. One reviewer wants the fire setbacks dimensioned on the roof plan, another wants them in a note, a third wants a separate sheet. The patchwork is real, and the industry has built an entire service category on top of it. Search for the problem and page one is permit design firms explaining that AHJs are inconsistent and offering to absorb the inconsistency for a fee per plan set.

We think that framing is mostly wrong, and not because the counties are fast. It is wrong because it puts the variable you cannot control at the centre of the problem and leaves the one you can control out of the picture entirely. The best public data on residential solar permitting, collected by a national lab in the same jurisdictions over several years, says the thing that moves the timeline is the submission, not the reviewer.

What happened when the packet was standardised

SolarAPP+ is a web-based permitting tool developed by the National Renewable Energy Laboratory, now the National Laboratory of the Rockies, and the building code community. An installer answers a fixed set of questions about a system. If the answers describe a code-compliant design within the tool's eligibility rules, the permit issues on the spot. Nobody at the county reads a plan set.

That setup is a rare natural experiment. The same counties process some solar permits through SolarAPP+ and others the traditional way, with the same code, the same inspectors and the same local politics. The lab has published the comparison every year since the pilot, and the direction has not changed.

In the 2024 data, covering 37,393 permits from 861 installers, the median time from permit submission to passed inspection was 37 business days through SolarAPP+ and 49 business days through traditional review, across the 29 jurisdictions with complete data. That is about 12 days, or roughly 24 percent. The 2022 review found 13 business days. The 2023 review found 14.5. The pilot, run in five jurisdictions from late 2020 into 2021, found an average of 12.

Look at where that time comes from. Traditional permits in participating jurisdictions had a median review time of seven business days, which is not the horror story installers tell. But around 12 percent of traditional permits took more than 30 business days to issue, and around 4 percent took more than 60. Fewer than 0.1 percent of SolarAPP+ permits took longer than 30. Same counties. Same code. The long tail belongs to the traditional packet, not to the jurisdiction.

Where the days actually go

A jurisdiction problem would show up as a fixed delay. Every permit in a slow county would be slow by roughly the same amount. What the data shows instead is a median that is tolerable and a tail that is brutal, and a tail like that is the signature of a loop.

The loop is the revision cycle. The lab's pilot report describes it plainly: when a design is deemed non-compliant, or when the installer changes the system after submission, the revision can send the applicant and the county back to the beginning of review. Each round is a letter, a redraw, a resubmission and a new place in the queue. A county that takes a week to review takes three weeks to review the same job twice.

What SolarAPP+ removes is not the reviewer. It removes the ambiguity the reviewer exists to resolve. The tool will not accept a design it cannot evaluate. It asks for inputs in a fixed structure, and a missing value is a missing value rather than a judgment call about whether sheet four implies what sheet two left out.

The pilot contains a detail every installer should sit with. During its first phase, about 37 percent of SolarAPP+ projects had revisions, against 8 percent of control projects. Part of that was a software quirk the team later fixed. Part of it was contractors learning what the input questions actually meant. The standardised format did not create sloppy inputs. It exposed them, early, where a plan reviewer would otherwise have found them later.

That is the argument in one number. When the format is fixed and complete, the errors surface at the start and the review collapses to nothing. When the format is whatever your drafter produced this week, the errors surface in a correction letter and the review becomes a negotiation.

What the data does not say

It is worth being careful here, because this is the point where most writing on SolarAPP+ oversells.

It is tempting to claim that a standardised packet also produces better installations. The inspection data does not support that cleanly. In the pilot, 19 percent of SolarAPP+ projects failed at least one inspection against 12 percent of control projects. In 2022 the result flipped, 17 percent against 24 percent, and the lab flagged that the gap might reflect which installers chose to use the tool rather than the tool itself. In 2023, PV-only failure rates were slightly higher for SolarAPP+ across the jurisdictions with data. In 2024 they were slightly lower, 31 percent against 35. The fair summary is that automated permitting does not hurt inspection outcomes and has not been shown to fix them.

The more useful number is the one that barely moves. In 2024, roughly a third of PV-only projects failed at least one inspection, whichever pathway they took. And in the lab's 2024 breakdown of SolarAPP+ inspection failures, 6 percent were due to a missing on-site inspection checklist and 5 percent to an installation that did not match the permitted plans.

Those are not code problems. They are document problems that survived the permit. The approved design did not make it to the roof, or the roof did not match the approved design, and the inspector found the gap. Every failed inspection is a second trip for the crew and a new slot on the inspector's calendar, which is the same economics we described in what field scheduling actually costs an oil and gas services firm. The expensive trip is the one you make twice.

The packet is a controlled document

Put the two halves together and the pattern is clear. Permitting delay for a residential installer is mostly a document control problem that starts at the sales appointment and ends at final inspection. The county sees the middle of it.

Consider what happens to a single design in a typical growing installer. Sales produces a proposal from a satellite image. Design produces a plan set from the proposal, sometimes before the site survey has come back. Procurement substitutes an inverter because the specified one is on backorder. The crew installs from a printout that predates the substitution. Each of those handoffs is a place where the document the county approved and the system on the roof can drift apart, and nobody owns the drift because each team owns only its own step.

A better relationship at the county counter does not fix this. What fixes it is treating the design as a single controlled record with an owner, a revision history, and a rule for what triggers resubmission. Then the first-pass packet is complete because completeness was defined in advance rather than rediscovered by each reviewer, and an equipment substitution triggers a revision before the crew rolls, not after the inspector asks. The paperwork, as it were, is the part of solar that most needs to be grounded.

This is the same shape we found in public sector procurement timelines, where the rules take the blame for calendar time that is actually spent drafting and routing documents the agency already controls. And it is the mirror image of the question a city faces when it automates permit intake: software on either side of the counter speeds up a process whose inputs are already clean, and speeds up the rework loop when they are not.

Where the jurisdiction still matters

None of this means the county is irrelevant. Some jurisdictions are genuinely slow, understaffed or idiosyncratic, and a perfect packet will still wait in their queue. SolarAPP+ itself is only available where a jurisdiction has adopted it: 264 had adopted or were piloting it by the end of 2024, across 13 states. California made the question mandatory with SB 379, which required cities and counties above set population thresholds to offer automated solar permitting by September 2023 or September 2024 depending on their size. Most installers outside those markets are still dealing with a human reviewer and a PDF.

But the lesson of the SolarAPP+ data does not depend on the tool being available to you. The tool is evidence of what happens when an installer's submission is complete, structured and internally consistent on the first pass. You can build that discipline into your own operation in a county that has never heard of SolarAPP+. What you cannot do is wait for the county to reform itself and call that a strategy.

Where this article stops

What we cannot tell you from here is where your own documents drift. That depends on how your sales proposals become plan sets, who owns the design after it leaves engineering, how equipment substitutions reach the permit record, and which of your correction letters repeat. The answer is sitting in your last year of permit history and your correction letters, and it usually takes a week with your design and operations leads to find the three handoffs doing most of the damage.

If your permitting timelines are the thing slowing your install calendar and the explanation in every meeting is the county, talk to us before you buy another permit service. We will trace your designs from signed proposal to passed inspection, show you where the record breaks, and tell you which fixes belong to your process and which ones genuinely belong to the jurisdiction.