The most expensive mistake in commercial rooftop solar is not a mounting error, an undersized inverter, or a badly modelled yield curve. It is a scheduling decision made in a meeting room two years before anyone climbs onto the roof: installing photovoltaics on a membrane that has six years of service life left.
On a warehouse or distribution centre, the membrane and the array are two assets with completely different clocks. A reinforced TPO or PVC membrane is typically warranted for fifteen to twenty years. A commercial PV array is financed, and expected to generate, for twenty-five. When those two timelines are laid on top of each other without adjustment, one of them has to be interrupted — and the interruption always costs more than the sequencing would have.
This guide sets out how owners, facility managers, and their consultants should decide the order of work: what to survey, how to read the remaining membrane life honestly, which sequencing scenario fits which building, and how to design the roof so that the next replacement does not require dismantling a functioning power plant.
Why the sequence decision is financial, not technical
Almost every membrane roof can physically carry a PV array. The engineering question — uplift, point load, seam integrity, drainage — is solvable, and it is solved on projects of every size. The question that actually determines project value is different: over the next twenty-five years, how many times will someone have to work on the waterproofing layer underneath this array?
The twenty-year mismatch
Take a building with a membrane installed in 2018, a fifteen-year warranty, and a realistic service life of eighteen to twenty years. If a 500 kWp array is installed in 2026, the membrane reaches the end of its life around 2036 — year ten of a twenty-five-year array. The owner then faces a choice between patching a failing membrane around several thousand modules or removing the array to replace the roof underneath it.
Neither option is cheap. Decommissioning and reinstalling a commercial array typically adds a meaningful percentage to the original installed cost — labour, crane or hoist time, new flashing details, re-commissioning, and lost generation during the outage. Doing the same work in the right order, before the array exists, costs nothing extra at all.
The cost of getting it backwards
There is also a hidden cost that rarely appears in a business case: the roof replacement itself becomes more expensive once solar is present. Access is obstructed, ballast has to be moved and re-spread, walkway and cable routes have to be protected, and the roofing contractor carries risk they will price into the quote. Roofing contractors consistently charge more for a roof that has an array on it, and they are right to.
Reading remaining membrane life honestly
Most sequencing failures start with an optimistic survey. A membrane that looks clean from the parapet may still be within five years of the end of its life, because membrane ageing is a chemical process that is largely invisible from above.
What the survey should actually include
A credible pre-solar roof assessment covers more than a visual walkover and a few photographs. It should include seam probe testing along a representative sample of welds, not just the ones that look suspect; an assessment of plasticiser migration and surface chalking on PVC; a check for ponding and the deflection that causes it; core samples or manufacturer records to confirm the membrane type and construction; and a review of every existing penetration, curb, and drain.
The output should be a written statement of estimated remaining service life with a stated basis, not a pass/fail verdict. If the roofing contractor will not commit to a number, that is itself information: it means the membrane’s condition is not well understood, and a twenty-five-year asset should not be placed on top of an unknown.
A practical threshold
The rule most asset managers converge on is straightforward. If the membrane has more than about eight years of reliable life remaining, install solar now and plan the roof replacement around the array. If it has fewer, re-roof first and treat the PV project as the second phase. The eight-year figure is not a code requirement; it is roughly the point at which the cost of a mid-life intervention under an array exceeds the cost of bringing the roof forward.
Three sequencing scenarios, and when each applies
Scenario A — re-roof first, then solar
This is the cleanest outcome and the easiest to warrant. A new membrane goes down with a full-term warranty, the array is installed on a known substrate, and both assets start their clocks on the same day. The array’s design can then inform the roof: reinforced details where rails will sit, drainage laid out to avoid future obstructions, and walkway routes that keep maintenance access clear of modules.
It is also the scenario that most often gets rejected on capital grounds, because it front-loads two large expenditures into one financial year. The counter-argument is that the alternative simply moves the second expenditure into a year when the owner has less control over it, and pays a premium for the privilege.
Scenario B — solar now, roof replacement planned around it
Where the membrane has real life left, this is the correct choice, but only with the replacement designed in from the start. That means a mounting layout that can be lifted in sections rather than dismantled entirely, cable management that can be disconnected and re-made at defined points, and a drawing set that records exactly where every base sits so the roofing contractor can work around it.
Scenario C — overlay or renovation with the array in place
Overlay systems — a new membrane installed over the existing one — are sometimes proposed as a way to avoid removing an array. They can work, but they raise the roof surface, change the drainage profile, and may void the original warranty. They also make future leak detection harder, because there are now two layers to trace. This scenario should be treated as a specialist solution, not a default.
Designing the roof so it can be replaced under an array
The single most valuable design decision is how the array touches the membrane. A non-penetrating mounting approach keeps the waterproofing continuous, which means the roof’s warranty position is unchanged and, critically, the membrane can be replaced later without repairing hundreds of penetration details.
A ballasted mounting system extends that logic further: ballast can be lifted, temporarily relocated, and re-spread in a way that mechanical fasteners cannot. Where structural capacity allows, the ballast approach turns a future roof replacement from a decommissioning project into a staged lifting operation — still disruptive, but a fraction of the cost and duration.
Both approaches depend on the same discipline at design stage: keeping the array off the drainage path, maintaining a clear maintenance walkway to every drain and rooftop unit, and specifying protection pads that will not react with the sheet over twenty years of heat cycling. That last point is where chemistry enters the sequence discussion, and our comparison of PVC and EPDM compatibility on solar roofs covers the material interactions worth checking before pads, walkway mats, and cable trays are specified.
The bottom line
Before the PV contract is signed, the owner’s team should be able to answer six questions in writing: the membrane’s type, age, and estimated remaining life with the basis for that estimate; the date and scope of the last roof survey; whether the membrane warranty is transferable and how solar affects it; the designed attachment method and the reasoning behind it; the estimated cost and duration of a future roof replacement under the proposed layout; and the phasing plan that would allow that replacement without a full decommission. If any answer is missing, the sequence has not been decided, and an undecided sequence on a membrane roof becomes a very expensive one.
Solar and roofing are usually procured by different people, on different budgets, against different asset registers. That organisational split is the real source of the problem. The buildings that perform well over twenty-five years are the ones where a single decision-maker looked at the roof and the array together, decided which one goes first, and wrote the answer into both contracts.
The rule is simple enough to apply on any project: if the membrane will not outlive the array, the roof goes first. Everything else — the mounting system, the warranty language, the phasing plan — follows from that one decision.
