For building owners, facilities and portfolio managers, and commercial roofing contractors

TPO is the membrane most often installed and the one least safely judged by its field record

Low-slope · single-ply thermoplastic · commercial and industrial

The sheet is welded rather than glued, reflective rather than black, and cheaper per square foot than the alternatives. All three of those facts have a condition attached, and the conditions are where roofs are won and lost.

30-second answer

What should an owner actually know about TPO before specifying it?

TPO is a hot-air-welded thermoplastic single ply, and on the last publicly verifiable industry figures the leading low-slope membrane in North America. Three things decide whether yours performs: the weld quality a crew achieves on site, the coating thickness over the scrim you specified, and how reflective the sheet stays once it is soiled. Its material standard has been revised twelve times since 2003, which dates its field record fast.

Safety first

A commercial roof is a workplace before it is a roof, and a white one has two hazards a black one does not

Roof access on a commercial building is occupational work. Under the federal construction standard, each employee engaged in roofing activities on a low-slope roof with unprotected sides and edges six feet or more above a lower level must be protected by guardrails, safety nets, personal fall arrest, or one of the defined warning-line combinations (29 CFR 1926.501). That obligation is not waived because a survey is quick or because the parapet looks tall enough.

Two hazards are specific to a reflective membrane. Glare off a clean white sheet in direct sun is severe enough to matter for eye protection and for seeing a change in plane. And a soiled white membrane wet with dew or light rain is slick — the same biological and particulate film that costs you reflectance costs you traction.

Escalate rather than investigate if you see a deck deflecting under standing water, insulation squelching underfoot, or open conduit and energised rooftop equipment near a leak. Saturated insulation is dead load that was never designed for, and water plus a live rooftop unit is an electrical hazard, not a roofing question.

If you manage the building rather than the roofing work, your job on this page is to specify and verify from the ground and from the documents. Send a contractor whose crew is trained, equipped and covered by a written fall-protection plan.

Learning paths and saved lessons

At a glance

The short versionSection link

Every figure here has a stated basis. None of them is a warranty term, and none of them describes a product until a submittal names one.

What the sheet is: Reinforced thermoplastic polyolefinASTM D6878/D6878M requires reinforcing fabric or scrim. An unreinforced TPO sheet is not what the standard covers.
How seams are made: Hot-air welded, no adhesiveThe two sheets are melted into one material. The weld is made on your roof, by your crew, on the day.
Governing material standard: ASTM D6878/D6878M-26Active edition as listed by ASTM. Eleven earlier editions exist, back to D6878-03.
Thickness that matters: Coating over scrim, not nominalPer NRCA, the 2013 revision required at least 30% of nominal on the weathering side; the 2011 revision raised the floor from 12 to 15 mils.
Reflectance, new: About 0.80 for a clean white membraneLevinson et al., Lawrence Berkeley National Laboratory, 2005 — measured on fifteen white and light-grey PVC membrane samples, not on TPO. A product's own rating comes from the Cool Roof Rating Council directory.
Reflectance, soiled: Can fall to about 0.60Same study, same PVC samples. Soot, dust and biomass deposition. The paper gives 0.30 to 0.80 as the range of degraded reflectance for white membrane roofs, depending on exposure.
Market position: 40% of new construction, 30% of reroofingNRCA 2015–16 market survey, for 2015, as reported by IIBEC. The most recent figure this page could verify publicly; treat it as a decade old.
Service life: This page publishes no numberA planning range would have to rest on field record, and TPO field record belongs to formulations that have since been revised.

Tradeoffs

Where TPO is the right answer, and where it is a habitSection link

TPO is specified by default on a very large share of North American low-slope work. Default is not the same as correct, and the failure cases are specific enough to name.

Best when

The roof is a large, simple field — warehouse, distribution, big-box retail — where welded seams can be run long and straight and a robotic welder can do most of the work.
Cooling load dominates the building's energy use and the roof is visible to the sun for most of the day.
You are buying reflectivity that you also intend to maintain, with a cleaning line in the maintenance budget rather than a hope.
The specification can name a thickness over scrim, an attachment listing for each roof zone, and a cover board — and you have someone who will check that the submittal matches.
Local crews weld TPO daily. Weld quality is a practised skill, and practice is regional.

Think twice if

The roof drains grease, solvent or animal-fat-bearing exhaust. Chemical resistance is where PVC earns its premium, and a restaurant or rendering exhaust plume is the classic case.
Rooftop equipment, glazing or a reflective wall focuses heat onto a patch of membrane. Writing in Professional Roofing in 2015, three GAF technical staff described high-heat problems on TPO and named directly attached solar panels, heavy dark dirt buildup, high-altitude locations and nearby highly reflective surfaces such as parapet walls as what sustains those temperatures. That is a manufacturer's account of its own industry's problem, not an independent finding — but it is the specific version of the concern.
The roof carries heavy, uncontrolled service traffic and nobody will fund walkway pads. A welded seam is only as good as the sheet either side of it.
You are relying on a bidder's twenty years of TPO experience as evidence. The product has been revised repeatedly over that period; the experience is real, the extrapolation is not.
Nobody on your side will be on the roof during installation. Weld quality is the dominant variable and it is invisible from a photograph.

What changes the answer

The exhaust inventory. What the building puts on its own roof narrows the membrane shortlist faster than price does.
The heating-to-cooling balance. A reflective roof that pays in Phoenix can cost you in Duluth, and DOE says so plainly.
The attachment method available for your deck and your wind zones — that decision changes cost, air-tightness and the interior noise of the roof more than the polymer does.
Whether the building will be sold or held. Reflectance loss, weld quality and warranty transferability all price differently against a five-year horizon than a twenty-five-year one.
Whether anyone will read the maintenance obligation in the warranty before signing it. On a no-dollar-limit warranty that obligation becomes yours on the day the roof is accepted.

How it works

The sheet is three layers, and only one of them weathersSection link

Almost every argument about TPO quality — thickness, formulation, crazing, warranty class — comes back to a single band of material a fraction of a millimetre thick.

Magnified cross-section through two 60-mil reinforced TPO sheets. Both are 60-mil sheets, both comply with the thickness-over-scrim rule NRCA describes, and they differ by a third in the only layer that faces the sun. The dimension figures are illustrations of the rule, not measurements of any product.Original diagram, Understanding Roofing.

A reinforced TPO sheet is a sandwich. In the middle is a scrim — a woven or knitted fabric that gives the sheet its breaking and tearing strength. Above and below it are layers of TPO compound. ASTM D6878/D6878M, the material specification the industry writes to, states that the sheet “shall contain reinforcing fabrics or scrims”; the reinforcement is not an upgrade, it is the definition.

The scrim contributes nothing to weathering. Neither does the backing layer, which faces the cover board and never sees sun. Every hour of ultraviolet exposure, every thermal cycle and every wash of ponded water is spent on one layer: the coating above the scrim on the weathering side. When that layer is consumed, the scrim prints through, the surface crazes, and the sheet begins to crack along the weave. That is why the industry measures thickness over scrim rather than only nominal thickness.

What the standard requires, and what your bid says

Writing in NRCA’s Professional Roofing in 2017, Mark S. Graham recorded the history. ASTM D6878 was originally published in 2003 and was based on 45-mil-thick products. The 2011 revision raised the minimum thickness over scrim from 12 mils to 15 mils. The 2013 edition, acknowledging the widespread use of 60-mil and thicker sheets, added a second rule: the coating over scrim on the weathering side must be at least 30 percent of the sheet’s nominal thickness. Graham gives the worked cases — a 60-mil sheet requires at least 18 mils over scrim, an 80-mil sheet at least 24.

Put the two rules together and something useful falls out. On a 45-mil sheet, 30 percent is 13.5 mils, which is below the 15-mil floor — so on thin sheets the floor governs and on 60-mil and thicker sheets the percentage governs. That is our arithmetic on the two rules as NRCA describes them, not a quotation from the standard; we have read ASTM’s published scope and edition record, not the body of the standard, and the active edition is now D6878/D6878M-26, which may have moved either number again.

The practical consequence is the diagram above. “60-mil TPO” on a bid line is a nominal total. It is compatible with 18 mils of weathering reserve and with 24. Two compliant sheets can differ by a third in the only layer that ages. The submittal, not the bid line, is where that number lives — and asking for it is a two-sentence email that no honest supplier objects to.

Why the seam is a different kind of risk from the sheet

TPO is a thermoplastic: heat it and it softens, cool it and it sets, repeatedly. That is what makes a hot-air weld possible. A robotic or hand welder drives hot air between two overlapping sheets while a roller presses them together, and the two become one material rather than two materials with an adhesive between them.

This is genuinely an advantage. Writing in Professional Roofing in 2017, Mike Ennis noted that thermoplastic membranes “remain capable of hot-air welding throughout their service lives” — a fifteen-year-old TPO roof can, in principle, take a welded patch that becomes continuous with the original sheet.

It is also the largest single quality risk on the job, for one structural reason: the seam is manufactured on your roof. An EPDM seam depends on cleaning and priming; a TPO seam depends on temperature, speed and roller pressure, all three set by a person on a day with weather. A cold weld and a correct weld look identical from three feet away and behave identically until water finds the difference. That is why the rest of this page spends more time on quality control than on polymer chemistry.

The original asset on this page

Seam weld quality control: what is checked, how, and what a failure looks likeSection link

On a TPO roof the seams are manufactured on site, by hand or by machine, in whatever weather the day supplies. This table is the whole quality-control programme in one place — and the honest reason to publish it is that most owners have never been shown one.

Seam weld quality control on a hot-air welded TPO roof. Frequencies and dimensions shown as examples are Carlisle SynTec's published guidance for its own products, dated 27 March 2023, used here to show what such a procedure looks like; the governing numbers are those in the bulletin for the product actually specified.
What is checked	How it is done	When	What a failure looks like	Whose document sets the number
Trial weld (destructive)	A coupon is welded with the production machine at production settings, allowed to cool, then peeled apart by hand.	Carlisle's guidance: before beginning work each morning and each afternoon — and again whenever conditions or machines change.	The two sheets separate cleanly at the weld interface. A sound weld tears the membrane away from its own scrim rather than opening at the lap.	Manufacturer bulletin for the specified product.
Weld width	Measured across the fused band on the trial-weld coupon and at any suspect run.	Every trial weld, and whenever the machine speed is changed.	A band narrower than the specified minimum — Carlisle looks for at least 1½ inches — or a width that wanders along the run as speed drifts.	Manufacturer bulletin; often repeated in the project specification.
Probe of every finished seam (non-destructive)	A blunt seam probe is drawn along the seam edge under firm pressure after the weld has cooled.	Carlisle's guidance: a minimum of 30 minutes after welding, and before the crew leaves the area that day.	The probe enters the lap. Carlisle states the tool will not penetrate a properly welded seam, so penetration is the defect, not the test.	Manufacturer bulletin. The probing record belongs in the project QC submittal.
Machine settings log	Temperature, travel speed, nozzle and roller pressure recorded per welder, per session, against ambient conditions.	At each trial weld, and whenever air temperature, wind, cloud cover or membrane surface temperature moves.	One set of settings recorded for a whole day on which the weather changed. That is a guess wearing a log's clothing.	Project specification. No standard requires it; the owner does.
Surface condition before welding	Visual check and wipe. On aged or overnight-exposed membrane, scrubbing with an abrasive pad and the manufacturer's solvent.	Before any weld onto membrane that has been exposed — which on a multi-week job includes most day-joint tie-ins.	A weld attempted through an embedded dirt film. NRCA's report of the WSRCA guidelines describes scrubbing precisely in order to restore weldability.	Manufacturer bulletin, plus WSRCA repair and maintenance guidance as reported by NRCA.
Overheating	Visual, continuously: smoke, bleed-out beyond the weld, scorching, local distortion.	Continuous during welding.	A darkened glossy bead and squeezed-out compound. Peer-reviewed testing found peel strength rising to a peak with temperature and then falling — hotter is a defect, not a margin.	Manufacturer bulletin for the limit; the research explains why the limit exists.
Test cut (destructive, investigative)	A sample of completed work is cut out and examined, then the opening is repaired.	Not routine. Used to investigate a problem already identified by observation.	Using cuts instead of observation. NRCA states test cuts are not part of a routine QC or QA programme and must not substitute for continuous visual examination.	NRCA guidance, and the project specification if it calls for cuts at all.
The record itself	A dated daily log of everything above, keyed to locations on a roof plan, signed, and handed over at closeout.	Daily, and delivered as part of the closeout package.	No log exists. Then the only remaining evidence about weld quality is the roof, and it reports late, in water, after the retention has been released.	The owner's specification. This is the single line most often missing from a low-slope specification.

Read this table one item at a time

Trial weld (destructive)

How it is done: A coupon is welded with the production machine at production settings, allowed to cool, then peeled apart by hand.
When: Carlisle's guidance: before beginning work each morning and each afternoon — and again whenever conditions or machines change.
What a failure looks like: The two sheets separate cleanly at the weld interface. A sound weld tears the membrane away from its own scrim rather than opening at the lap.
Whose document sets the number: Manufacturer bulletin for the specified product.

Weld width

How it is done: Measured across the fused band on the trial-weld coupon and at any suspect run.
When: Every trial weld, and whenever the machine speed is changed.
What a failure looks like: A band narrower than the specified minimum — Carlisle looks for at least 1½ inches — or a width that wanders along the run as speed drifts.
Whose document sets the number: Manufacturer bulletin; often repeated in the project specification.

Probe of every finished seam (non-destructive)

How it is done: A blunt seam probe is drawn along the seam edge under firm pressure after the weld has cooled.
When: Carlisle's guidance: a minimum of 30 minutes after welding, and before the crew leaves the area that day.
What a failure looks like: The probe enters the lap. Carlisle states the tool will not penetrate a properly welded seam, so penetration is the defect, not the test.
Whose document sets the number: Manufacturer bulletin. The probing record belongs in the project QC submittal.

Machine settings log

How it is done: Temperature, travel speed, nozzle and roller pressure recorded per welder, per session, against ambient conditions.
When: At each trial weld, and whenever air temperature, wind, cloud cover or membrane surface temperature moves.
What a failure looks like: One set of settings recorded for a whole day on which the weather changed. That is a guess wearing a log's clothing.
Whose document sets the number: Project specification. No standard requires it; the owner does.

Surface condition before welding

How it is done: Visual check and wipe. On aged or overnight-exposed membrane, scrubbing with an abrasive pad and the manufacturer's solvent.
When: Before any weld onto membrane that has been exposed — which on a multi-week job includes most day-joint tie-ins.
What a failure looks like: A weld attempted through an embedded dirt film. NRCA's report of the WSRCA guidelines describes scrubbing precisely in order to restore weldability.
Whose document sets the number: Manufacturer bulletin, plus WSRCA repair and maintenance guidance as reported by NRCA.

Overheating

How it is done: Visual, continuously: smoke, bleed-out beyond the weld, scorching, local distortion.
When: Continuous during welding.
What a failure looks like: A darkened glossy bead and squeezed-out compound. Peer-reviewed testing found peel strength rising to a peak with temperature and then falling — hotter is a defect, not a margin.
Whose document sets the number: Manufacturer bulletin for the limit; the research explains why the limit exists.

Test cut (destructive, investigative)

How it is done: A sample of completed work is cut out and examined, then the opening is repaired.
When: Not routine. Used to investigate a problem already identified by observation.
What a failure looks like: Using cuts instead of observation. NRCA states test cuts are not part of a routine QC or QA programme and must not substitute for continuous visual examination.
Whose document sets the number: NRCA guidance, and the project specification if it calls for cuts at all.

The record itself

How it is done: A dated daily log of everything above, keyed to locations on a roof plan, signed, and handed over at closeout.
When: Daily, and delivered as part of the closeout package.
What a failure looks like: No log exists. Then the only remaining evidence about weld quality is the roof, and it reports late, in water, after the retention has been released.
Whose document sets the number: The owner's specification. This is the single line most often missing from a low-slope specification.

Nothing in this table detects a weld that was sound at installation and failed later. That is what an inspection programme is for. What it does detect is the defect that was there on the day — which is the population that produces most early-life seam leaks.

Why this is the workmanship variable that matters most

A commercial low-slope roof has a great many things that can be done badly. Most of them are visible: a missed fastener row, a sloppy termination, an unsupported pipe. Weld quality is not. A cold weld and a correct weld are the same colour, the same texture and the same thickness, and the difference between them is a few degrees or a few metres per minute of travel speed several hours ago.

The 2025 study in Materials is useful precisely because it quantifies how narrow the window is. Welding across 300 to 600 °C, the authors found peel strength that rose to a peak and then fell, with the strongest welds at 400 °C, 4 metres per minute and about 15 newtons of pressure. Move off that peak in either direction and you lose strength. That is a laboratory result on one product with one machine — do not take those numbers to a roof — but it establishes the shape of the problem: there is an optimum, it is not “as hot as possible”, and the machine has to be re-tuned as conditions move it off the peak.

Which is why the trial weld exists. It is not paperwork. It is the only way to find the peak on this membrane, with this machine, on this morning — and the reason to require it twice a day is that the morning answer is wrong by two o’clock.

What an owner can actually do

You are not going to stand on the roof with a probe. What you can do is three things, none of which requires being a roofer. Put the quality-control procedure in the specification so it is priced. Ask for the daily log as a closeout deliverable, alongside the warranty registration. And on a large job, retain an independent observer — an enclosure consultant, not the contractor’s own foreman — for the periods when seams are being run. The cost of that observation is small against the cost of finding out in year four.

The track-record caveat

Twelve editions in twenty-three years, and what that does to the evidenceSection link

The strongest thing anyone says about TPO is that there is a great deal of it on roofs and most of it is fine. The weakest link in that argument is that the sheet being quoted today is not the sheet that produced the record.

ASTM’s own catalogue for D6878/D6878M lists twelve editions: 2003, 2006, 2006a, 2008, 2008e01, 2011, 2011a, 2013, 2017, 2019, 2021, and the active 2026. That is not a criticism of the committee — a standard that is revised is a standard that is being maintained, and every one of those revisions presumably made the specification harder to meet or more accurate about what is being sold. It is a statement about what a field observation can support.

Two of those revisions are documented in useful detail. Per NRCA’s Mark Graham, the 2011 edition raised the minimum thickness over scrim from 12 mils to 15 and extended specimen heat aging before aged physical property testing from 28 days at 240 °F to 224 days at the same temperature — four weeks to thirty-two. The 2013 edition added the requirement that coating over scrim on the weathering side be at least 30 percent of nominal thickness, in explicit acknowledgement that 60-mil-and-thicker sheets had become the norm while the standard was still built around 45-mil products.

Set those two facts beside each other. A sheet installed in 2009 met a specification requiring 12 mils over scrim and four weeks of heat aging. A sheet installed today meets one requiring 15 mils or 30 percent, whichever governs, and thirty-two weeks. Those are not the same product class, and the twenty-year-old roof you can go and look at is evidence about the first one.

How to ask the question so it produces an answer

The useless version is “how long does TPO last?” The useful version is four questions that a supplier can answer in writing:

Which edition of ASTM D6878/D6878M is this specific sheet certified to, and by whom?
What is the thickness over scrim on the weathering side, in mils, as stated on the product data sheet?
Has this formulation changed since the reference projects you are offering me — and if so, when and in what respect?
What will you put in writing about the specific exposures on this roof — the exhaust, the reflected heat from that glazed wall, the ponding at the north drain?

The last one is the most revealing. A supplier who will write an exposure-specific statement is telling you their technical department has looked at your building. A supplier who will only send a brochure is telling you something too.

Where this page’s own advice is wrong

Take this argument too far and you arrive somewhere silly: that no evidence about any membrane means anything, so choose on price. That is not the conclusion. Three things push the other way, and an honest page has to say so.

First, the revisions have been in one direction — more heat aging, more weathering reserve, more differentiation between products. A current sheet is being held to a harder test than a 2009 sheet was. Second, the same instability argument applies, in weaker form, to every membrane on the market; EPDM’s seams and PVC’s plasticiser packages have also changed. The difference is one of degree. Third, and most practically: on a very large share of real commercial roofs, weld quality, cover board, attachment and drainage will decide the outcome long before formulation does. If you have limited attention to spend, spend it on the quality-control table above rather than on the polymer.

The place the formulation argument genuinely governs is narrower: when a bidder is using long field record as the reason to choose their product over another, and when the roof has an unusual thermal or chemical exposure that makes generic experience a poor guide. In those two cases, ask the four questions.

Reflectivity

What you buy when you buy a white roof, and how much of it you keepSection link

Reflectivity is the most-marketed property of TPO and the most misunderstood. It is real, it is measurable, it is climate-dependent, and it declines.

Two numbers describe a cool roof. Solar reflectance is the fraction of incoming sunlight the surface bounces back, from 0 to 1. Thermal emittance is how readily the surface sheds the heat it did absorb, also 0 to 1. A membrane can be reflective and a poor emitter, or the reverse; the combined index the industry quotes, SRI, is derived from both.

The magnitude is genuinely large. Lawrence Berkeley National Laboratory’s Heat Island Group puts a clean white roof reflecting 80 percent of sunlight at about 31 °C (55 °F) cooler on a summer afternoon than a grey roof reflecting 20 percent. That temperature difference is not just an energy number; it is a durability number for everything under the membrane and everything sitting on it.

The number on the data sheet is a clean number

The Cool Roof Rating Council rates products twice: initially, and again after three years of outdoor weathering at approved test farms in Arizona, Ohio and Florida, chosen to span hot-dry, cold-temperate and hot-humid conditions. The reason the aged rating exists is that roofs get dirty.

How dirty is measurable, though the best measurement available is not a measurement of TPO. In a 2005 study in Atmospheric Environment, Levinson, Berdahl, Berhe and Akbari examined fifteen initially white or light-grey PVC membrane samples taken from roofs across the United States, and reported that the high initial reflectance of a white membrane roof — around 0.80 — can be lowered to about 0.60 by deposition of soot, dust and biomass, with degraded values spanning 0.30 to 0.80 depending on exposure. In round terms about a quarter of the reflectance can go. Soiling is a surface-deposition process rather than a polymer process, which is why the mechanism transfers to a TPO roof — but the magnitude on your sheet is a question for that product’s own aged CRRC rating, not for this study.

The same study measured what cleaning brings back at the most soiled location on each sample, as a ratio of cleaned to unsoiled reflectance: wiping recovered 0.53 to 0.95, rinsing 0.74 to 0.98, washing 0.79 to 1.00, and bleaching 0.94 to 1.02. The gap between rinsing and washing is the finding a facilities manager can use. Running a hose over the roof is not a wash. Neither is rain.

How fast the loss arrives is not something either source cited here establishes for any particular roof. What can be said is what the industry does about it: the CRRC rates at three years of outdoor exposure, and that aged number — not the initial one — is the one to put beside an energy calculation.

Whether it pays is a question about your building

DOE’s Federal Energy Management Program is unusually blunt about this: cool roofs “achieve the greatest cooling savings in hot climates, but can increase energy costs in colder climates due to reduced beneficial wintertime heat gains,” and a qualifying product “may not be life cycle cost-effective in certain low-use applications or in locations with very low rates for electricity or natural gas.” LBNL’s researchers add the counterweight: because winter sun is low, days are short and skies are often cloudy, the winter penalty is typically small compared with the summer benefit in most U.S. locations.

Both are true, and the resolution is arithmetic rather than allegiance. The variables are your cooling-to-heating balance, your electricity and gas prices, how much insulation is under the membrane — a well-insulated roof transmits less of either effect — and how many weeks a year the roof is under snow, which reflects perfectly well on its own.

One more caution that is specific to a commercial retrofit: a reflective membrane runs colder than the black roof it replaced, and the temperature of the layers below is an input to vapour control. That is not a reason to avoid a cool roof. It is a reason for the reflectance decision and the insulation and vapour-control design to be made by the same person, at the same time.

What to require in the submittal

The CRRC listing for the exact product and colour — initial and three-year aged solar reflectance and thermal emittance, not a generic “white TPO” figure.
The manufacturer’s permitted cleaning methods and chemistry, in writing, before the first cleaning rather than after it.
Confirmation of whether any energy-code, utility-incentive or certification requirement your project is claiming uses the initial or the aged value — they are different numbers and the requirement names one.

Attachment

Four ways to hold the sheet down, and three zones that need different amounts of itSection link

Attachment decides wind performance, acoustics, air movement through the assembly, and a surprising amount of the price. It is also the part of a TPO specification most often reduced to a single number that cannot be correct.

There are four common ways a TPO sheet is held to the building, and each changes more than its name suggests.

Mechanically attached — fasteners and plates driven through the sheet along the seam laps and into the deck, with the next sheet welded over them. It is the fastest and usually the cheapest. The membrane between fastener rows is free to deflect, so the sheet moves under wind suction, and the assembly below it is not sealed by the membrane.

Induction welded — fasteners with coated plates are installed through the insulation first, then a hand-held induction tool bonds the membrane down to each plate from above, with no penetration of the sheet at all. It removes the fastener rows from the seams and gives a much denser bond pattern than seam fastening does.

Fully adhered — the sheet is bonded across its whole area to the cover board with adhesive. The membrane does not move. It is normally the most expensive way to hold the sheet down, and the most sensitive to substrate condition, moisture and application temperature, because an adhesive has a temperature window and a roof does not always cooperate.

Ballasted — the sheet is held by weight, typically stone or pavers. It is a structural-load decision before it is a roofing decision, and it makes leak location genuinely difficult, because the water travels under ballast you have to remove to look. It is uncommon on new TPO work.

The three zones

Wind does not load a roof evenly. Writing in IIBEC’s Interface, Karl A. Schaack notes that ASCE 7 provides methods for determining wind uplift pressures for the field, perimeter and corner wind zones — three zones on one roof, with progressively higher pressures as you move outward and into the corners. The attachment density has to follow.

Schaack also describes what “more attachment” can mean in practice at the edges. Reporting FM Global’s Property Loss Prevention Data Sheet 1-29, he notes that prescriptive enhancements for FM’s Zones 2 and 3 — the perimeter and the corners — may be used for relatively low wind-pressure designs where the rating needed in Zone 1, the field, is 90 lb/ft² or less. Above that, meeting the elevated pressures can require simultaneous attachment of the cover board and the underlying insulation layers, or of a polyester-reinforced base ply together with those layers — not simply more fasteners through the membrane. That threshold is an insurer’s criterion for its own approvals, not a code determination and not a rule that applies where FM is not the standard being designed to.

This is the reason a bid that states one fastener rate for the whole roof deserves a question, and the reason edge metal and perimeter securement belong in the same conversation as the membrane. Where the roof fails first is almost always the edge; what is at the edge is the subject of drainage and edge systems.

What an approval actually approves

FM and UL listings attach to tested assemblies, not to sheets. A listing describes a specific deck, a specific insulation type and thickness, a specific cover board, a specific fastener at a specific density, and a specific membrane — and it holds only for that combination. Substituting a different cover board, or a thicker insulation, or a different fastener, produces an assembly that was not tested, whatever the membrane on top of it says.

None of that is a wind design. The design pressures for this building come from a qualified design professional working from the building geometry, height, exposure, enclosure classification and risk category, under the code edition your jurisdiction has actually adopted. This page explains the concepts so you can read a proposal. It does not determine anything about your roof.

Cost and lifecycle

The five specification decisions that move the price of a TPO roofSection link

This page publishes no dollars per square foot for TPO. What it can do is name the decisions that separate a cheap bid from an expensive one, so that two proposals can be compared on scope before they are compared on price.

Specification decisions on a TPO roof, what the cheaper and more expensive choices actually are, and where each one shows up in a proposal. No prices are published; the point is comparability.
Decision	The cheaper choice	The more expensive choice	What the extra money buys	Where to look in the bid
Sheet thickness over scrim	A 60-mil sheet at the 30% minimum — about 18 mils of weathering reserve.	A thicker sheet, or a 60-mil sheet with more coating above the scrim.	More of the only layer that ages. It does not buy puncture resistance on its own; that belongs to the cover board and the assembly.	The product data sheet in the submittal, not the bid line. If it states only nominal thickness, ask.
Attachment method	Mechanically attached: fasteners and plates through the seam laps into the deck.	Induction welded to coated plates, or fully adhered to the cover board.	A quieter, less mobile membrane and a different air-leakage path through the assembly; adhered systems remove the fastener rows from the seams entirely.	The line naming the FM or UL assembly listing, and the fastener or adhesive rate for each roof zone.
Cover board	Membrane laid directly over the insulation facer.	A gypsum, high-density polyiso or other cover board between insulation and membrane.	A hard, continuous substrate under the sheet. This is what most changes how the roof tolerates foot traffic, dropped tools and hail.	A separate line item with a named product and thickness. Its absence is often the reason one bid is lower.
Traffic protection	None. Technicians walk the shortest route to each unit.	Walkway pads on the actual service routes, plus pads at the access hatch.	A defined path that takes the abrasion instead of the membrane. Cheap, routinely deleted, and the reason for a specific class of small leaks.	A quantity of walkway pad in linear feet, keyed to a roof plan with the equipment marked.
Weld quality control and its record	Welding done to the crew's judgement; nothing written down.	Documented trial welds, seam probing of every seam, and a daily log tied to a roof plan.	Evidence. Without a record, the only proof of weld quality is the roof itself, and it reports late and in water.	A quality-control submittal, and a line in the schedule of values that pays for the time it takes.

Read this table one item at a time

Sheet thickness over scrim

The cheaper choice: A 60-mil sheet at the 30% minimum — about 18 mils of weathering reserve.
The more expensive choice: A thicker sheet, or a 60-mil sheet with more coating above the scrim.
What the extra money buys: More of the only layer that ages. It does not buy puncture resistance on its own; that belongs to the cover board and the assembly.
Where to look in the bid: The product data sheet in the submittal, not the bid line. If it states only nominal thickness, ask.

Attachment method

The cheaper choice: Mechanically attached: fasteners and plates through the seam laps into the deck.
The more expensive choice: Induction welded to coated plates, or fully adhered to the cover board.
What the extra money buys: A quieter, less mobile membrane and a different air-leakage path through the assembly; adhered systems remove the fastener rows from the seams entirely.
Where to look in the bid: The line naming the FM or UL assembly listing, and the fastener or adhesive rate for each roof zone.

Cover board

The cheaper choice: Membrane laid directly over the insulation facer.
The more expensive choice: A gypsum, high-density polyiso or other cover board between insulation and membrane.
What the extra money buys: A hard, continuous substrate under the sheet. This is what most changes how the roof tolerates foot traffic, dropped tools and hail.
Where to look in the bid: A separate line item with a named product and thickness. Its absence is often the reason one bid is lower.

Traffic protection

The cheaper choice: None. Technicians walk the shortest route to each unit.
The more expensive choice: Walkway pads on the actual service routes, plus pads at the access hatch.
What the extra money buys: A defined path that takes the abrasion instead of the membrane. Cheap, routinely deleted, and the reason for a specific class of small leaks.
Where to look in the bid: A quantity of walkway pad in linear feet, keyed to a roof plan with the equipment marked.

Weld quality control and its record

The cheaper choice: Welding done to the crew's judgement; nothing written down.
The more expensive choice: Documented trial welds, seam probing of every seam, and a daily log tied to a roof plan.
What the extra money buys: Evidence. Without a record, the only proof of weld quality is the roof itself, and it reports late and in water.
Where to look in the bid: A quality-control submittal, and a line in the schedule of values that pays for the time it takes.

The membrane itself is usually a minority of an installed low-slope price. Insulation, cover board, attachment, tear-off or recover, deck repair, drainage work and details normally dominate — which is why comparing bids on the membrane line is the wrong comparison.

Units: No dollar figure is published in this section. The units named here are the ones two TPO proposals have to be normalised to before their prices mean anything: U.S. dollars per square foot of roof area, itemised by line, with quantities stated.
Scope included: Membrane, attachment, cover board, traffic protection, and the quality-control effort — the five decisions above, taken separately rather than as one number.
Not included: Deck repair, structural work, drainage correction, tapered insulation design, permits, disposal of the existing roof, and anything a condition survey has not yet found.
Geography: United States, national. Nothing in this section describes a price in any market.
Data as of: 26 August 2026 — the date this section was written and its sources re-read. No price dataset underlies it, because none separating TPO by installed cost at an acceptable tier could be verified.
Confidence: The decisions and their direction are well supported. The magnitude of each is not published here, and any figure a reader attaches to them has to come from their own market.
Method: How this figure is built

A TPO bid is easy to make cheap without lying. Drop the cover board, delete the walkway pads, take the sheet to the thickness-over-scrim minimum, choose the least expensive attachment the listing allows, and price no time for quality control. Every one of those is a real choice a competent contractor might defend on the right building. Together, on the wrong one, they are the difference between a roof that reaches its warranty term and one that does not.

The instrument that fixes this is not a price negotiation. It is a specification that names all five decisions before the bids go out, so that the cheapest proposal is the cheapest way of doing the same work rather than the cheapest way of doing less of it. That is the subject of normalising proposals before comparing them.

A planning range is not a quote. It is a number to argue with a proposal about — the only price that binds anyone is the one in a signed scope of work for this building.

Considerations

What changes this on a real buildingSection link

Fire

A TPO roof does not carry a fire class. The tested assembly does. On low-slope work the deck, the insulation, the cover board, any vapour retarder and the membrane are all part of what was tested, and swapping the insulation or deleting the cover board can move the assembly outside its listing even though the membrane is unchanged.

Fire classification applies to a tested roof assembly, not to a membrane in isolation. Ask for the listing number for the assembly actually being installed on this deck, and check that the submittal matches it layer by layer.

Wind

Uplift performance is a property of an assembly on a building on a site, not of a sheet. Writing in IIBEC’s Interface in April 2022, Karl A. Schaack noted that ASCE 7 provides methods for determining wind uplift pressures for field, perimeter and corner wind zones on the roof — three zones, three different pressures, three different attachment requirements on the same roof.

That is why an attachment rate quoted as one number for the whole roof is a warning sign. Reporting FM Global’s Property Loss Prevention Data Sheet 1-29, Schaack describes prescriptive enhancements for the perimeter and corner zones, and notes that meeting elevated pressures there can require simultaneous attachment of the cover board and the underlying insulation layers, or of a polyester-reinforced base ply, rather than simply adding fasteners to the membrane. FM’s criteria are an insurer’s, not your jurisdiction’s.

A membrane does not carry a wind rating by itself. FM approvals and UL listings apply to specific tested assemblies at specific attachment densities, and the design pressure for this building comes from a qualified design professional working from the building's geometry, height, exposure and risk category — not from a product brochure and not from this page.

Climate

Reflectivity is the reason many owners choose white TPO, and its value is climate-dependent in a direction DOE states plainly. The Federal Energy Management Program’s cool-roof guidance says that cool roofs “achieve the greatest cooling savings in hot climates, but can increase energy costs in colder climates due to reduced beneficial wintertime heat gains.” It also warns that a qualifying product “may not be life cycle cost-effective in certain low-use applications or in locations with very low rates for electricity or natural gas.”

Lawrence Berkeley National Laboratory’s Heat Island Group puts the other side of it: because winter sun is low, days are short and skies are often cloudy, the winter heating penalty is typically small compared with the summer cooling benefit in most U.S. locations. Both statements can be true. Which one governs your building depends on your heating-to-cooling balance, your fuel prices, your insulation, and how much of the year the roof is under snow.

Cool-roof benefit varies by climate, insulation level, roof type and HVAC system, and can carry a cold-climate heating tradeoff. A reflective membrane also runs colder than a black one, which is a vapour-control input for the assembly below it — that belongs to the insulation and air-vapour design, not to the membrane choice.

Maintenance

Reflectance is not a permanent property. In a 2005 study in Atmospheric Environment, Levinson, Berdahl, Berhe and Akbari examined fifteen initially white or light-grey PVC roofing membrane samples taken from roofs across the United States, and reported that the high initial reflectance of a white membrane roof — around 0.80 — can be lowered to about 0.60 by deposition of soot, dust and biomass, with degraded values ranging from 0.30 to 0.80 depending on exposure. Note the material: those samples were PVC, not TPO.

The same study measured recovery by cleaning method at the most soiled location on each sample, expressed as the ratio of cleaned to unsoiled reflectance: wiping restored 0.53 to 0.95, rinsing 0.74 to 0.98, washing 0.79 to 1.00, and bleaching 0.94 to 1.02. The gap between rinsing and washing is the practical finding — a hose is not a wash, and neither is a rainstorm.

Those are measurements on fifteen light-coloured PVC membrane samples in one twenty-year-old study. They are not measurements on TPO and not a schedule for your roof. What transfers is the mechanism and the direction: soiling costs reflectance, and cleaning recovers most but not all of it. The magnitude on a TPO sheet has to come from that product's own CRRC aged rating. Any cleaning method has to be one the membrane manufacturer permits, because the wrong chemistry voids coverage.

Code and jurisdiction

Whether your jurisdiction requires a reflective roof, how much insulation it requires, and whether an existing roof may be recovered rather than torn off are all set by the energy and building codes your jurisdiction has actually adopted, in the edition it adopted, with its local amendments. There is no nationwide building code for site-built construction. Model IECC and IBC text is model text; it is evidence about what your code probably resembles, not about what it requires.

This page cites no code provision as a requirement anywhere, because a requirement is jurisdiction-specific. Confirm the adopted edition, its effective date, and any local amendments with your authority having jurisdiction before a specification relies on any of it.

Access and site conditions

Every inspection, every survey and every service call on this roof is work at height. Rooftop equipment brings technicians who are not roofers onto the membrane, on routes nobody planned, carrying sheet metal and dropping screws. Walkway pads and a marked service route are a membrane-protection measure and a housekeeping measure at the same time, and they cost less than the leaks they prevent.

Warranty and repair

What a TPO warranty covers, and the obligation it hands back to youSection link

A membrane warranty is a contract, not a service-life statement. On a commercial roof it is also, in practice, a maintenance agreement you are signing on behalf of your own staff.

Material-only versus no-dollar-limit: These differ materially. A material-only warranty covers the sheet against manufacturing defect and typically nothing else — not the labour to find the leak, not the labour to fix it, and not the workmanship that caused it. A no-dollar-limit (NDL) system warranty covers a defined assembly, usually including labour, usually only when the installer is a manufacturer-approved applicator and the assembly is built from that manufacturer’s listed components. Read which one you are being offered before you compare the terms.
The maintenance condition: Commercial membrane warranties commonly condition coverage on documented inspection and maintenance, on prompt reporting of damage, and on the owner not permitting alterations by others. That last one is the trap on a busy roof: an HVAC contractor setting a new curb, a telecom crew running a sled, or a solar installer adding attachments can all be “alterations by others.” Whether a particular act voids a particular warranty is decided by that document and by the law where the building is — not by a summary, including this one.
The formulation problem, in warranty terms: A twenty-year warranty issued in 2010 covered the formulation sold in 2010. A twenty-year warranty issued today covers a sheet made to a materially different standard edition. The warranty term is the same number in both cases, which is precisely why the term is not evidence about the product. Ask what edition of ASTM D6878/D6878M the sheet is certified to and what the certification document actually says.
What a warranty is not: It is not a service-life prediction, it is not insurance, and it is not a substitute for a condition survey. Nothing on this page promises that any warranty will be honoured, will transfer, or will be enforceable — those are legal questions about a specific document in a specific jurisdiction.

Repairability

TPO repairs well, with one condition. A welded patch becomes continuous with the original sheet rather than sitting on top of it, and a competent crew with a hand welder and a probe can make a permanent repair in an afternoon. That is a genuine advantage over systems whose repairs are adhesive-dependent.

The condition is surface preparation. Reporting on the Western States Roofing Contractors Association’s TPO repair and maintenance guidelines in 2011, Professional Roofing described scrubbing repair areas with a relatively abrasive pad and the membrane manufacturer’s solvent, in order to more thoroughly remove embedded dirt and restore the membrane’s weldability — and noted that where a surface weld cannot be achieved on aged membrane, underside patch placement and reverse-side welding are suggested. Read that as the honest version of “TPO welds for life”: the polymer remains weldable, the weathered surface does not, and the difference is work.

The practical planning consequence: a repair on a fifteen-year-old TPO roof costs more per patch than the same repair at year three, and a crew that skips the scrubbing produces a patch that looks finished and is not.

A warranty is a contract between a reader and whoever wrote it. What it covers, what voids it, whether it transfers, and how it is enforced are set by that document and by the law where the reader lives. Read the actual warranty for the product and the installer in front of you — not a summary of one, including this one.

Ask before you sign

Questions for a TPO bidder, and what a weak answer tells youSection link

These are deliberately specific. A contractor who welds TPO every week answers all eight without looking anything up.

What is the thickness over scrim on the weathering side of the sheet you are proposing, in mils, and where does the submittal state it?
The right answer is a number and a document reference. An answer that repeats the nominal thickness — “it’s a 60-mil sheet” — has not answered the question, and the difference between a compliant 18 mils and a compliant 24 mils is a third of the layer that ages.
Which edition of ASTM D6878/D6878M is this sheet certified to, and can I see the certification?
The standard has been revised repeatedly since 2003 and the active edition is D6878/D6878M-26. “It meets ASTM” without an edition is not a claim you can check.
Describe your trial weld procedure. How often, what sample, and what are you looking for when you peel it?
A practised crew describes a routine without prompting. As an example of what one manufacturer publishes, Carlisle SynTec recommends a test weld before beginning work each morning and each afternoon, an approximately one-inch-wide sample cut across the seam, roughly ten minutes of cooling, and a peel that shows delamination of the membrane from its scrim reinforcement across a minimum inch-and-a-half weld width. Your specified product’s own bulletin sets your numbers — the point is whether the bidder has one.
Who probes the seams, with what tool, how long after welding, and what do they do with the record?
Probing is the primary non-destructive check on a welded seam. Carlisle’s published guidance is to probe once welds have thoroughly cooled — a minimum of thirty minutes — drawing the probe along the seam edge under firm pressure, and states that the tool will not penetrate a properly welded seam. “We probe as we go” with no log is half an answer.
How do your welder settings change between eight in the morning and two in the afternoon?
They must change. Air temperature, wind, cloud cover and the membrane’s own surface temperature all move the correct setting. A bidder who says the machine is set once for the day is telling you their quality control is a guess.
What is the fastener or adhesive rate in the field, at the perimeter, and at the corners, and which listed assembly are those rates from?
Three numbers and a listing reference. One number for the whole roof means the zones have not been designed, and the perimeter and corner are where progressive failures start.
What cover board is under the membrane, and what happens to this price if I delete it?
The answer tells you whether the cover board is in the bid at all. It also tells you whether the bidder will explain a tradeoff or simply take the deletion.
Show me the roof plan with walkway pads on it and the service routes marked.
If no such plan exists, traffic protection was priced as a quantity rather than designed as a route — and the route technicians actually walk will not be the one that got padded.

Require these in writing

Membrane manufacturer, product name, colour, nominal thickness, and thickness over scrim on the weathering side.
The ASTM D6878/D6878M edition the sheet is certified to.
The FM or UL assembly listing number for the assembly as built on this deck, layer by layer.
Fastener or adhesive rates stated separately for field, perimeter and corner zones.
Cover board manufacturer, product and thickness — or an explicit statement that there is none.
Insulation type, thickness and R-value, and how it is attached.
A written weld quality-control procedure: trial weld frequency, acceptance criteria, probing method and timing, and the daily record that will be handed over.
Walkway pad quantity keyed to a roof plan showing rooftop equipment and the access hatch.
The specimen warranty document itself — not a description of it — for both the material and the system.
The maintenance obligations that begin on the day of acceptance, and who is responsible for each.

What goes wrong

What owners get wrong about TPO, and how it actually failsSection link

Common misconceptions

Common belief
60-mil is 60-mil. A thicker sheet is a better sheet.
What is actually true
Nominal thickness is the total. The layer that weathers is the coating above the scrim on the sun side, and the rule NRCA describes sets that at a minimum — at least 30 percent of nominal since the 2013 revision — not at a fixed value. Two compliant 60-mil sheets can differ by a third in weathering reserve. Thicker is generally better, but the number to compare is over scrim.
Common belief
TPO has twenty-five years of proven performance.
What is actually true
TPO has twenty-five years of field exposure, which is not the same claim. ASTM’s own catalogue lists twelve editions of D6878/D6878M since 2003 — 03, 06, 06a, 08, 08e01, 11, 11a, 13, 17, 19, 21 and the active 26. NRCA has documented what changed at two of those revisions: the 2011 edition raised the thickness-over-scrim floor and extended specimen heat aging from 28 days to 224 days at 240 °F, and the 2013 edition added the 30-percent rule. A roof installed in 2008 is evidence about a sheet that had to survive four weeks of heat aging, not thirty-two. Field record for TPO is weaker evidence about what is being quoted today than it is for the older single plies.
Common belief
You can tell a good weld by looking at it.
What is actually true
You cannot. A cold weld — too little heat, too much speed, too little roller pressure — is visually indistinguishable from a sound one and holds water until it does not. That is exactly why the trade uses a destructive trial weld to set the machine and a blunt probe to check the production seam. Neither is optional and neither is a photograph.
Common belief
Hotter is a stronger weld.
What is actually true
Over-welding is a defect. In a 2025 study in Materials, Wang, Zang, Li and colleagues welded TPO waterproofing membrane across 300–600 °C, 1–5 m/min and 7.47–15.57 N of pressure and found that peel strength “increases first and then decreases with the increase in welding temperature,” peaking at 400 °C. Above the peak, strength fell. On a roof the visible version of the same thing is smoke, bleed-out and a scorched, distorted edge.
Common belief
A white roof is a cool roof, so it saves energy anywhere.
What is actually true
DOE’s Federal Energy Management Program says cool roofs achieve the greatest cooling savings in hot climates but can increase energy costs in colder climates through lost wintertime heat gain, and that a qualifying product may not be life-cycle cost-effective in low-use applications or where energy rates are very low. Separately, the Cool Roof Rating Council states that a product’s presence on its directory “does not mean that the product is ‘cool,’ as defined by any particular code, standard, or program.” The rating is data. Whether it pays is arithmetic on your building.
Common belief
The reflectance on the data sheet is the reflectance of my roof.
What is actually true
It is the reflectance of a clean specimen. CRRC rates products both initially and after three years of outdoor weathering at approved test farms in Arizona, Ohio and Florida, chosen to represent hot-dry, cold-temperate and hot-humid conditions. Ask for the aged value, not the initial one — and note that FEMP’s own comparison table lists a three-year solar reflectance of 0.87 for the best available product against 0.50 for an ENERGY STAR-qualified one and 0.05 for a less efficient roof.
Common belief
A test cut will tell me whether the roof was installed properly.
What is actually true
Not on its own. Writing in Professional Roofing in 2017, Mark S. Graham was direct: test cuts “are not considered to be a part of a routine QC or QA program” and “should not be substituted for in-process QC or QA provided by continuous visual examination,” because small test areas likely do not accurately represent the overall membrane application over large roof system areas. A cut is a way to investigate a problem you have already found. Continuous observation is how you find it.

How it actually fails

Cold or partial weld: Too little heat, too much travel speed, insufficient roller pressure, or a machine set once in the morning and never adjusted as conditions changed. The lap holds mechanically for a while and then opens under thermal movement.What you can see: A seam probe entering the lap. Later, leaks that track a seam line rather than a penetration, often appearing well away from the stain inside.
Over-welded, scorched seam: Excess temperature or too slow a pass. The compound smokes, bleeds out and degrades locally; the peak-then-decline relationship between temperature and peel strength is exactly this.What you can see: A darkened, glossy or wrinkled bead along the seam edge, and material squeezed out beyond the weld.
Failed repair on aged membrane: Welding onto a weathered, soiled surface without scrubbing it back with an abrasive pad and the manufacturer’s solvent. The patch bonds to the dirt film rather than to the sheet.What you can see: Patches lifting at their edges, or the same leak returning after a repair that was signed off as complete.
Depletion of the coating over scrim: The end state of normal weathering, reached sooner where the weathering reserve was thin, where heat is concentrated, or where water stands. The scrim eventually governs the surface.What you can see: Chalking, then fine crazing, then the weave printing through, then cracking that follows the scrim lines rather than random paths.
Localised high-heat deterioration: Reflected and concentrated heat from adjacent glazing, a light wall, or rooftop equipment raises local surface temperature far above the field. Writing in Professional Roofing in 2015, three GAF technical staff named directly attached solar panels, heavy dark dirt buildup that blocks reflectivity, high-altitude locations and nearby highly reflective surfaces such as parapet walls as the circumstances that sustain such temperatures. Read that as a manufacturer’s account of its own industry’s problem rather than as an independent finding.What you can see: Surface loss confined to a band or patch near a reflective surface, with the surrounding field intact.
Puncture along unplanned service routes: Technicians take the shortest line to a rooftop unit, drop tools, and leave sheet-metal offcuts. Foot traffic over a hard object drives it through the sheet. Whether there is a cover board under the membrane changes how much of this the assembly absorbs.What you can see: Small isolated leaks in a straight line between the access hatch and the equipment, and debris found on the roof at inspection.
Progressive peel from the perimeter: Uplift pressures are highest at corners and along edges — the zones ASCE 7 treats separately. When perimeter securement disengages, wind gets under the membrane behind it and strips the field progressively inward.What you can see: Loose or rattling edge metal, fasteners backing out along the perimeter, and membrane that billows visibly in gusts. Usually observable before the failure.

Sources and further readingSection link

Understanding Roofing / Published August 26, 2026

Scope and limitations

It cannot tell you whether TPO is right for your roof.
That answer comes from a survey of this building — its exhaust inventory, its drainage, its deck, its insulation moisture content, its traffic, and its wind design — by a qualified consultant or contractor who has been on it.
It publishes no installed price per square foot for TPO, and no service-life range.
No transparent dataset separating TPO by installed cost at an acceptable source tier could be verified, and a service-life range would have to rest on field record that belongs to superseded formulations.
It has not read the text of ASTM D6878/D6878M.
The thickness-over-scrim and heat-aging rules described here are as reported by NRCA in 2017 about the editions current then.
ASTM's catalogue confirms the edition history and the reinforcement requirement in the scope; the active edition, D6878/D6878M-26, may have revised either rule again.
Buy or borrow the standard before writing it into a specification.
The trial-weld and probing numbers quoted here are one manufacturer's published instructions for its own products, used as a worked example of what such a procedure looks like.
They are not an industry requirement and they are not your numbers.
Your numbers come from the bulletin for the product actually specified.
The welding-parameter research quoted here is a laboratory study of a specific membrane and a specific welder.
It supports the shape of the relationship between temperature and weld strength, not a machine setting for any roof.
It cites no code provision as a requirement.
There is no nationwide building code for site-built construction; adopted edition, local amendments and the authority having jurisdiction govern every code question this page raises.
The reflectance and soiling figures come from a 2005 study of fifteen white and light-grey PVC membrane samples, not TPO.
Soiling is surface deposition rather than a polymer process, so the mechanism and the direction transfer; the magnitude does not.
Your product's initial and three-year aged values come from the Cool Roof Rating Council directory listing for that specific product and colour, and no equivalent independent soiling study of TPO specifically was found for this page.
The market-share figure is from a 2015 survey reported second-hand.
No more recent publicly verifiable figure was found.
Treat 'most-installed' as a decade-old finding rather than a current fact.
Nothing here is legal advice about a warranty.
Whether coverage exists, transfers, or is enforceable is decided by the specific document and the law where the building is.

ASTM D6878/D6878M-26, Standard Specification for Thermoplastic Polyolefin-Based Sheet Roofing
ASTM International / 2026 edition, listed as active
The active edition designation and its scope: the specification “covers flexible sheet made from thermoplastic polyolefin (TPO) as the principal polymer, intended for use in single-ply roofing membranes exposed to the weather. The sheet shall contain reinforcing fabrics or scrims.” The catalogue listing is also the source for the full edition history quoted on this page: 03, 06, 06a, 08, 08e01, 11, 11a, 13, 17, 19, 21 and the active 26.
This is ASTM's catalogue listing and published scope, not the text of the standard. The specific requirements — thickness over scrim, heat aging duration, physical properties — are behind purchase and were not read for this page.
TPOs face a changing landscape
Mark S. Graham, Professional Roofing (National Roofing Contractors Association) / 1 August 2017
That ASTM D6878 was originally published in 2003 based on 45-mil-thick products; that the 2011 revision raised the minimum thickness over scrim from 12 mils to 15 mils and extended specimen heat aging from 28 days at 240 °F to 224 days at 240 °F — four weeks to 32 weeks; and that the 2013 edition added the requirement that coating over scrim on the weathering side be at least 30 percent of nominal thickness, giving 18 mils on a 60-mil sheet and 24 mils on an 80-mil sheet.
Written in 2017; its statement that D6878-13 is the current edition is superseded. It is a trade technical article describing the standard, not the standard itself.
Tech Today — WSRCA TPO repair and maintenance guidelines
Professional Roofing (National Roofing Contractors Association) / 1 August 2011
That repairs to aged TPO call for scrubbing repair areas with a relatively abrasive pad and the membrane manufacturer's solvent to more thoroughly remove embedded dirt and restore the membrane's weldability, and that underside patch placement and reverse-side welding are suggested where a surface weld cannot be achieved.
It reports the contents of a Western States Roofing Contractors Association document that was not read directly for this page. WSRCA guidance is trade best practice, not a requirement anywhere.
Testing TPOs
Helene Hardy Pierce, Christopher McGroarty and Thomas J. Taylor, PhD — all of GAF — in Professional Roofing (National Roofing Contractors Association) / 1 August 2015
The localised high-heat concern described on this page: that “some manufacturers have had issues when membranes were installed in high-heat situations,” typically “on roofs located in extremely hot, sunny climates or on roofs that had an unusual amount of dirt buildup, leading to loss of reflectivity,” and that the circumstances driving sustained high membrane temperature “include membranes with directly attached solar panels, significant dark dirt buildup that blocks reflectivity, high-altitude locations and nearby highly reflective surfaces such as parapet walls.”
All three authors were employed by GAF, a TPO manufacturer, when this was written. It is a manufacturer's technical account of its own industry's field problems, published in a trade magazine — treated here as tier 6 despite the NRCA masthead, because the authorship, not the venue, sets the tier. It is not an independent finding and it quantifies nothing.
Assessing application
Mark S. Graham, Professional Roofing (National Roofing Contractors Association) / 1 October 2017
That “test cuts are not considered to be a part of a routine QC or QA program” and “should not be substituted for in-process QC or QA provided by continuous visual examination,” and that small test areas likely do not accurately represent the overall membrane application over large roof system areas.
Describes NRCA's position and its published guidance; it is not a code or contract requirement.
The rise of TPO
Mike Ennis, RRC, Professional Roofing (National Roofing Contractors Association) / 1 October 2017
That thermoplastic membranes “remain capable of hot-air welding throughout their service lives,” and the market context that PVC and TPO together represented about 30 percent of the commercial roofing market as of 2006 per Ducker Research.
The author writes from a single-ply industry perspective and the market figures are attributed to a third-party research firm whose report was not read. The 2006 figure is context, not a current market share.
Experimental Study and Application of TPO Waterproofing Membrane Lapping Process Parameters
Keyong Wang, Zhenhua Zang, Jie Li and colleagues, Materials (Basel), via PubMed Central / 2025
That welding temperature is the dominant parameter in TPO lap weld strength and that the relationship is not monotonic: peel strength “increases first and then decreases with the increase in welding temperature.” The study tested 300–600 °C, 1–5 m/min and 7.47–15.57 N, and reported a peak peel strength of 674.92 N per 50 mm at 400 °C, 4 m/min and 14.97 N.
A laboratory study of a specific membrane product and a specific welding apparatus. The numbers are evidence about the shape of the relationship, not machine settings for any roof, any product, or any weather.
TecTopics: Focus on Test Welds and Seam Probing
Carlisle SynTec Systems / 27 March 2023
One manufacturer's published trial-weld and probing procedure, used on this page as a worked example: a test weld before beginning work each morning and each afternoon; an approximately 1-inch-wide sample cut across the seam; roughly 10 minutes of cooling before peeling; delamination of the membrane from its scrim reinforcement across a minimum 1½-inch weld width; probing after welds have thoroughly cooled, a minimum of 30 minutes, drawing the probe along the seam edge under firm pressure; and that the tool will not penetrate into the lap area of a properly welded seam.
Product-specific manufacturer guidance for Carlisle SynTec products. Another manufacturer's requirements differ in frequency, dimensions and acceptance criteria. The governing document is always the bulletin for the product actually specified.
Roof Rating Program
Cool Roof Rating Council
That products are rated both initially and after three years of outdoor weathering at CRRC-approved test farms in Arizona (hot/dry), Ohio (cold/temperate) and Florida (hot/humid), and that “a product's placement on the Directory does not mean that the product is ‘cool,’ as defined by any particular code, standard, or program.”
A rating and directory programme, not a performance threshold and not a code. Whether a rated value satisfies a requirement is decided by the code or programme that sets the requirement.
Purchasing Energy-Efficient Cool Roof Products
U.S. Department of Energy, Federal Energy Management Program
The climate caveat quoted on this page: cool roofs “achieve the greatest cooling savings in hot climates, but can increase energy costs in colder climates due to reduced beneficial wintertime heat gains,” and that qualifying products “may not be life cycle cost-effective in certain low-use applications or in locations with very low rates for electricity or natural gas.” Also the three-year solar reflectance row of its comparison table: 0.87 best available, 0.5 ENERGY STAR, 0.05 less efficient.
Federal purchasing guidance written for federal buyers. The savings columns of its table rest on FEMP's own usage and price assumptions, which are not reproduced on this page and should be checked at the source before any figure is relied on.
Cool Roofs
Heat Island Group, Lawrence Berkeley National Laboratory
That a clean white roof reflecting 80 percent of sunlight stays about 31 °C (55 °F) cooler on a summer afternoon than a grey roof reflecting 20 percent, and that because winter sun is low, days are short and skies are often cloudy, the winter heating penalty is typically small compared with the summer cooling benefit in most U.S. locations.
A national laboratory research summary, not a prediction for any building. It states a general finding about U.S. locations; it does not replace an energy model for a specific roof, climate and fuel mix.
Effects of soiling and cleaning on the reflectance and solar heat gain of a light-colored roofing membrane
Ronnen M. Levinson, Paul H. Berdahl, Asmeret A. Berhe and Hashem Akbari, Atmospheric Environment, via the Heat Island Group, Lawrence Berkeley National Laboratory / 2005
That the high initial solar reflectance of a white membrane roof (circa 0.8) can be lowered by deposition of soot, dust and biomass to about 0.6, with degraded values ranging 0.3 to 0.8 depending on exposure; and the recovery ratios of cleaned to unsoiled reflectance, measured at the most soiled location on each sample — wiped 0.53–0.95, rinsed 0.74–0.98, washed 0.79–1.00, bleached 0.94–1.02.
The samples were fifteen initially white or light-grey POLYVINYL CHLORIDE membranes taken from roofs across the United States. This is not a measurement of TPO, and it is twenty years old. It supports the mechanism and the direction of soiling loss and cleaning recovery — soiling is surface deposition rather than a polymer process — but not a magnitude, a schedule, or a figure for any TPO product. The 0.3-to-0.8 span is the paper's statement about degraded white membrane roofs generally, not a measured spread across its own fifteen samples.
Roof Wind Load Design
Karl A. Schaack, RRC, PE, IIBEC Interface (International Institute of Building Enclosure Consultants) / April 2022
That ASCE 7 “provides methods for determining the wind uplift pressures for field, perimeter, and corner wind zones on the roof”; and, reporting FM Global's Property Loss Prevention Data Sheet 1-29, that “prescriptive enhancements for Zones 2 and 3 may be used for relatively low wind-pressure designs where the needed wind rating in Zone 1 is 90 lb/ft2 (4.3 kPa) or less,” and that meeting elevated perimeter and corner pressures can require simultaneous attachment of the cover board and the underlying insulation layers, or of a polyester-reinforced base ply in conjunction with those layers.
A technical journal article for enclosure consultants. The 90 lb/ft² threshold and the Zone 1/2/3 scheme it belongs to are FM Global's criteria for FM-approved assemblies — an insurer's requirements, not adopted law and not a code determination. It is not a wind design for any building, and nothing on this page substitutes for a design by a qualified professional working from the adopted code and ASCE 7 edition in force where the building is.
29 CFR 1926.501 — Duty to have fall protection
U.S. Department of Labor, Occupational Safety and Health Administration
That each employee engaged in roofing activities on a low-slope roof with unprotected sides and edges 6 feet or more above lower levels must be protected by guardrail systems, safety net systems, personal fall arrest systems, or a defined combination involving a warning line system — and that the same 6-foot trigger applies to unprotected sides and edges generally under 1926.501(b)(1).
This is the federal construction standard. Work by a building's own staff may fall under the general industry standard at 29 CFR 1910.28 instead, and OSHA-approved state plans may impose additional requirements. Which standard applies to a given task is a question for a competent safety professional.
NRCA Releases 2015-16 Market Survey
International Institute of Building Enclosure Consultants (IIBEC), reporting the NRCA market survey / 2016
The only market-position figure on this page: “Thermoplastic olefin (TPO) remains the market leader with a 40% share of new construction and 30% of the reroofing market for 2015.”
A second-hand report of an NRCA survey that was not read directly, describing 2015 data. No more recent publicly verifiable market-share figure was found. It supports 'was the market leader in 2015', not 'is the market leader now'.

Suggest a correction