Should you replace your roof before solar panels? Sequencing the two clocks

Your roof's remaining years decide it

Everything in this decision reduces to one number you may have to dig for: how many years your current roof covering has left. If you do not know the roof's age, the Department of Energy's step-by-step consumer guide notes that your local government should have a record of when it issued the building permit for the last installation or replacement, and a roofer's written remaining-life opinion is the other half of the evidence. With that number in hand, the table is the decision in one screen.

Reading your roof the way the decision requires

Remaining life is an estimate with a method behind it, and it is worth knowing what the roofer is looking at so you can judge the opinion you are handed. The covering's material sets the outer frame: the Department of Energy puts roof life anywhere from 20 to 50 years depending on the materials used, and the material's own warranty and the roofer's opinion place your covering inside that range. On tile, the underlayment beneath the tiles keeps its own clock, which the tile roof solar guide covers. Within the frame, condition moves the number: curling or missing shingles, granule loss, soft decking, stained ceilings below, and flashing that has already been patched twice all pull a roof toward the short end of its material's range regardless of its birthday.

Two features of the assessment matter specifically for solar. First, layers: if the roof already carries a second layer of shingles over the first, ask the roofer whether the next replacement means a full tear-off, because a heavier re-roof makes the detach-and-reinstall sequence heavier too. Second, the array plane itself: a roofer quoting remaining life for the whole roof may average in planes the array will never touch, while the plane that matters is the one the panels will cover, since its failure alone forces the event. Ask for the opinion plane by plane, in writing, with the array plane named. That sentence in the roofer's letter is the input the whole worksheet stands on, and it costs nothing extra to obtain while they are already on the ladder.

Where the roof clock crosses the array clock

The opening table compresses the decision into bands; the chart below shows why the bands sit where they do. Put the array's service life of about 25 to 30 years, the Department of Energy's figure, on one line, and three roofs with different remaining lives beneath it. A roof with 25 years left is replaced about when the array retires, and the two jobs fold into one future project. A roof with 15 years left is replaced in the middle of the array's life, which is the detach-and-reinstall event this page exists to price. A roof with 8 years left forces the same event inside the first decade, when the array has most of its production still ahead of it and the shutdown costs the most energy.

Notice what the chart does not show: a reason to replace the 25-year roof today. Its replacement already lands at the array's end, so pulling it forward buys nothing but an early roofing bill. The chart is the visual version of the section further down on when replacing first is the wrong call.

The array usually covers less of the roof than owners picture, which matters for the partial-replacement question later on this page. Lawrence Berkeley National Laboratory's 2026 data update estimated that residential arrays installed in 2025 typically covered 16 to 39 percent of total roof area across the percentile band it reports, with a median of 31 percent on multi-story homes and 22 percent on single-story homes. The lab approximates roof area from each building's ground-floor square footage adjusted for an actual or assumed 20-degree pitch, so treat the figures as a national pattern rather than a measurement of your house. The practical reading: on a typical 2025 installation the panels covered well under half the total roof, so the planes they sit on are a subset of the roof, and those planes are the ones whose failure forces a detach.

The event you are trying to avoid: detach and reinstall

Name the mid-life event precisely, because installers sometimes sell past it. Detach and reinstall means a crew removes every panel and the racking, stores the equipment while roofers strip and replace the covering, then remounts, rewires, and recommissions the system, with fresh flashing at every penetration and, where the permit office requires it, a new permit and inspection. What it costs is labour twice for the same array, and it is entirely absent from the original quote unless you asked.

So ask, in the quote, in writing. The Solar Energy Industries Association's Residential Consumer Guide to Solar Power (version 8.0, January 2025) lists the question for lease and PPA signers in so many words: what happens if I need to replace my roof after the system is installed, and how much will you charge to remove and reinstall the system? The same guide's general list asks whether you will be required to make any changes to your home, such as roofing upgrades. Cash buyers deserve the identical line from their own installer. A detach-and-reinstall figure given verbally at a kitchen table is not a figure; the written one is what goes into the worksheet below, and an installer who will not write it has priced your risk at whatever the future charges.

The event also has a calendar cost the quote rarely shows. The system is down from detach to recommissioning, which means lost production across the roofing schedule, and roofing schedules slip: weather, material deliveries, and the roofer's queue all sit between the two halves of your array's life. A detach planned for a low-production month loses less than one forced by a spring leak into the sunniest quarter. None of this argues that a detach is catastrophic; it argues that when the worksheet totals land close, the sequence that never schedules the event deserves the tie-break, because its rival's true price includes weeks nobody itemised.

Reading a detach-and-reinstall quote line by line

When the installer does put the figure in writing, a single lump sum is still hard to judge, and it is easy for a later invoice to grow lines the original figure never had. Ask for the detach-and-reinstall price broken into the steps the crew actually performs. We found no national price data for these lines, so this site prints none; what follows is the structure to request and the question that each line should answer before you rely on the total.

Two lines deserve the closest reading. The flashing line decides whether a future leak is the roofer's problem or the solar crew's, and the roof warranty guide shows how shingle makers' documents treat penetrations. The warranty line decides whether the reinstall quietly resets your protection, and it matters most when the original installer is not the one doing the work, the situation the guide to an installer that has closed covers. Neither line has a standard national answer, which is exactly why the question belongs in writing before the roofer is booked.

Doing both together: the published figures

The case for combining the jobs is not folklore; the Department of Energy published the arithmetic, in a July 28, 2021 article, so read every dollar figure in this section as a 2021 national average. It put the average roof replacement at about $10,000, depending on the roof's size and other factors, with roofs lasting anywhere from 20 to 50 years depending on the materials. It cited National Renewable Energy Laboratory figures for an average residential rooftop array of $19,000, so the two bought separately came to nearly $30,000, and NREL researchers estimated that done together the average cost could be $25,000, a potential average saving of about $4,000. The same article gives one mechanism, that roofers and solar installers often have partnerships, and another that needs no estimate at all: panels last about 25 to 30 years, similar to a roof, so doing both together avoids having the panels re-installed after a later roof replacement. The same article also estimated that an average of 2.3 million single-family detached homes a year would need a roof replacement over the following decade, which is why the question comes up so often.

Treat those as dated 2021 national averages from a named source, which is how this site treats every benchmark: they frame the size of the prize, and your quotes replace them. What survives at any price level is the structure. One scaffold, one permit sequence, one roofing warranty written with the array in mind, flashing installed by the people who will warrant the roof, and a roof covering chosen to run the array's full 25 to 30 years rather than the cheapest covering that passes inspection this spring.

Coordination is the combined job's real content, so specify it: who flashes the mounts, whose warranty covers a leak at a penetration (the roofer's material warranty and the installer's workmanship warranty meet exactly there), and in what order the trades work. Get both warranties in writing with the penetration question answered in each (the roof warranty guide reads what the shingle makers' documents say). A combined job with uncoordinated warranties can be worse than two separate jobs with clear ones.

A combined job is also the one moment certain preparations are nearly free. While the deck is exposed, the trades can confirm the rafter layout the mounting design assumes (the roof load arithmetic for solar shows why roof-foot spacing on those rafters matters), place blocking or backing where the design wants it, agree the conduit path before the covering closes it off, and photograph the deck condition for both warranty files. Ask for those photographs explicitly; a dated picture of sound decking under the array plane settles arguments a decade later that no memory wins. This is the unglamorous substance of the combined-job saving: not a discount, but a dozen small coordinations that are cheap exactly once, on the day both crews are already there.

What to ask the roofer and the installer before you sign

• Establish the roof's age from the permit record if the paperwork is missing, and get a roofer's remaining-life opinion in writing.
• Get the detach-and-reinstall figure in the solar quote, in writing, before signing anything.
• Price all three sequences: solar now, both together, roof first; the opening table says which band you are in.
• On a combined job, get both warranties in writing with the penetration question answered in each.
• Choose a roof covering rated to run the array's 25 to 30 years, not the cheapest covering available.
• If only the array plane is tired, ask the roofer for a written partial-replacement quote before assuming a full one.

When replacing first is the wrong call

The sequencing rule has a mirror image, and a contractor who sells both roofs and solar has no reason to volunteer it. A sound roof with twenty or more good years left is not a reason to delay solar, and it is not a reason to replace early: you would be discarding a decade or more of paid-for roofing to synchronise two clocks that were never going to collide inside the array's productive life. The years of production lost while waiting for a roof that does not need replacing are a real cost, priced by the same worksheet, and they usually dwarf the tidy feeling of matched ages.

There is a middle scope worth one question in the judgement zone: replacing only the roof plane the array will cover, if your roofer will quote it and warrant it. Planes weather at different rates, the array plane is the one whose failure forces the detach, and a partial replacement resets the clock that matters while leaving sound planes alone. Not every roofer will split the job or warrant the seam, which is exactly why it is a question for the roofer's written quote and not an assumption in yours.

One more sequence exists for the right property, and it dissolves the problem instead of pricing it: skip the roof entirely. If the roof is tired but the land is open, a ground-mounted array never touches the covering, the roof ages on its own schedule, and the re-roof, when it comes, is a plain roofing job with no array in the way. The mount comparison guide prices that trade in full. It will not suit a small or shaded lot, but on the properties it fits, it converts this page's whole three-sequence worksheet into a footnote, which is the best outcome a sequencing problem can have.

The worksheet: three sequences, three inputs

Settle your roof with three quotes and one division-free comparison. Sequence A, solar now and the roof when it fails: the solar price plus the roof price plus your installer's written detach-and-reinstall figure. Sequence B, both together now: the combined quote, which should show the saving the Department of Energy's figures predict, in whatever size your market gives it. Sequence C, roof now and solar right after: the roof price plus the solar price, no saving, no detach, and the array starting a few months later. The inputs are your roof quote, your written detach-and-reinstall figure, and the combined quote; everything else is addition.

The worked example below runs the worksheet on published and labelled figures so you can see its shape before your quotes arrive. Two judgements ride on top of the totals. First, Sequence A's detach lands in a future year's prices and in a future year's disruption, which the raw sum understates. Second, a roof in the under-ten band is not really offering you Sequence A at all: the detach inside the first decade is close enough to certain that A and C differ mainly in whether you pay for the removal. Run your own three numbers and let the band from the opening table check the result.

File the worksheet when you finish it: the roofer's letter, the three quotes, and the sequence you chose, reason recorded in one line. That small file is the one a future buyer, insurer, or warranty claim will ask for, and it is unanswerable from memory in year nine. Sequencing decisions look obvious the week they are made and arbitrary a decade later; the paper is what makes yours look considered, because it was, and it quietly supports the price of the house when the array's remaining life is part of what is being sold.

Pricing the calendar in kilowatt-hours

The dollar worksheet leaves out time, and time has a unit you can count. Take the same 8 kW system and a planning example of 1,300 kilowatt-hours per kilowatt a year, a figure for illustration, not a forecast for your roof: the system makes about 10,400 kilowatt-hours a year, or about 867 a month if the year were split evenly. A real year is not even, so a delay that lands in summer costs more than these averages and one in winter costs less. Your own production estimate, read month by month, replaces the even split.

Now price the two delays the sequences carry. Sequence C starts the array a few months late while the roof goes on: three months at the even split is about 2,600 kilowatt-hours that never get made, a one-time loss. Sequence A carries the detach instead: if the array is down for 14 days while the roof is replaced, that is about 399 kilowatt-hours at the even split, and a schedule that slips to 30 days more than doubles it to about 855. Neither number decides the question on its own, but both belong next to the dollar totals, valued at whatever your electricity costs per kilowatt-hour, because they are the part of each sequence that no quote will show you. The production estimate guide shows where the monthly figures sit on a PVWatts report.