[RE-wrenches] PV and Roof Flammability

Thu Sep 23 20:38:48 PDT 2010

Wrenches,

Building fires are serious stuff. The Northbrook UL labs fire tests are 
awesome. Imagine a gale-force wind driving a wide, roaring natural gas flame 
at the roof eave to see how far the fire will spread as it consumes the 
roofing material and anything else combustible. Tremco has a photo of the 
spread-of-flame test at http://www.tremcoroofing.com/qa/fire_test.asp Some 
roofing material is self-extinguishing. Some roofing material burns to 
ashes. PV module glass does not feed the fire from the top but module 
backsheets combust. Non-glass modules combust readily but can be rated Class 
A in low-pitch applications.

For a photo of the burning brand test see 
http://www.extension.org/pages/Fire_Ratings_for_Roofing_Material Glass solar 
modules on stand-off mounts have saved homes from flying burning brands.

I think that most wrenches warn their customers about fire hazards. Stay 
safe.

Joel Davidson

----- Original Message ----- 
From: "Matt Lafferty" <gilligan06 at gmail.com>
To: "'RE-wrenches'" <re-wrenches at lists.re-wrenches.org>
Sent: Thursday, September 23, 2010 6:22 PM
Subject: Re: [RE-wrenches] PV and Roof Flammability

> Wrenches,
>
> I'm coming from a perspective of comparing "beneath array" areas to "not
> beneath array" areas of the same roof. Regarding flammability, I don't
> really care what the source of ignition is in this conversation. I prefer 
> to
> think that the PV isn't the cause, frankly. Would rather think in terms of
> flying embers from a fire in the area (wild fire, neighbor's house on 
> fire,
> chimney, etc.), but I don't think we can discount any cause out of hand.
>
> I think it is very important to understand that the UL tests are designed 
> to
> emulate pre and early-stage combustion, as opposed to emulating a
> fully-involved fire situation. This is important because it marks the
> difference between "fire resistant" and "fire proof" materials. Only fire
> proof materials would survive to pass a full-on fire test. These tests
> essentially test the fire resistance of roofing materials.
>
> Another element to understand is that these tests only test from the top.
> They don't emulate an attic-fire burning thru the sheathing and engulfing
> the roofing materials from the back.
>
> Based on observation and farm boy commons, it seems to me that the biggest
> problem area is low-profile, flush mounted arrays over petroleum-based
> roofing materials. These are parallel to and close to the roof surface. 
> Most
> commonly with horizontal rails below the surface of the modules, thereby
> reducing the effective clearance to the roof by the height of the rail.
> Which dramatically reduces convection. This mounting configuration
> concentrates heat onto the roof surface and traps roof gasses beneath the
> array from otherwise normal ventilation.
>
> Gassing is normal and occurs throughout the life of most roofing 
> materials.
> Petroleum based roofing products tend to gas at a declining rate 
> throughout
> their life cycle. This gassing is not generally combustible in these
> concentrations. Combustible gasses are created when a roofing material
> begins to burn. Early-stage combustion. If these combustible gasses are 
> not
> allowed to evacuate quickly, the fire is exacerbated.... Gaining heat and
> intensity quickly. Flame spread.
>
> Once the roofing material is fully involved, meaning self-sustained 
> burning,
> it's probably not gonna matter much whether there is an array above it or
> not unless someone is there to fight the fire.
>
> The "chimney" effect behind a tilted array would have three primary 
> effects
> on this cycle. The first effect is that normal gassing would occur 
> similarly
> to adjacent roof sections not covered by the array. The second is that,
> during pre and early-stage combustion, the combustible gasses will not
> concentrate between the module and the roof, effectively reducing the fuel
> and heat concentration by some amount. Possibly to a level that is equal 
> to
> adjacent roof sections not covered by the array. Possibly even 
> extinguishing
> an ember or small flame in a manner similar to blowing on a lit match. The
> third effect would occur once the roofing material is fully involved. In
> this part of the cycle, the chimney effect could certainly draw more air
> into the fire, possibly increasing its intensity, much like blowing into 
> the
> base of a campfire.
>
> From a pure flammability perspective, I like the tilted array better than
> the flush array. My reasoning is primarily that normal aging will be more
> likely and, if an ember were to reach the roof beneath the module, it is
> more likely to self-extinguish or at least act like the adjacent roof. 
> These
> characteristics reduce the overall liklihood of actually catching fire in
> the first place. Which I like.
>
> In the event that the roof beneath a tilted array were to become involved
> and induce an adverse chimney effect, the fire would be WAY more 
> accessible
> to a water hose than with a flush mounted array.
>
> Another negative that a flush mounted array has in the fully involved
> scenario, is that it's gonna melt down for sure. If this is happening 
> during
> daylight hours, it may contribute more ignition source from the DC arcing
> and sparking in close proximity to the already flaming roof.
>
> The real question at hand is to determine the affect(s) on roofing
> flammability caused by mounting PV above the roof materials and compare 
> that
> to the flammability of identical roofing material, on an adjacent surface
> without an array above it. Part of a comprehensive study will include
> evaluation of various mounting methods, on different slopes, above various
> roofing materials. Anything less is suspect and must be discounted.
>
> For example, if the tests are only performed on a flat, level plane, they
> won't properly model real-world sloped-roof installations. Similarly, if 
> the
> tests are only performed at a single low-angle pitch, they are incomplete.
>
> Ryan is right. At some point, we have to accept that collateral damage 
> will
> occur. As part of that, we must determine how
> much/many/frequent/catastrophic is acceptable. My definition of
> acceptability goes like this: Collateral damage is acceptable as long as
> it's somebody else's customer.
>
> I would guess that we all share a similar definition when we get right 
> down
> to it. Which sort of causes a problem, doesn't it? This is at the heart of
> why I am in favor of responsible testing to determine the truth in this
> matter. Because we all want it to be somebody else's customer and we can't
> make sure it's not our customer unless we understand how to do that. We
> won't really understand how to do that unless somebody goes thru the
> exercise of testing stuff and telling us what the results are.
>
> I suspect that testing will show that most common sloped-roof materials 
> will
> maintain their UL rating as long as there is something like a 6" clearance
> up-slope for convection. Meaning 6" between the roof surface and any
> horizontal rails. I don't think one can honestly argue that "there's an
> array over that piece of roof so hot embers can't get onto the roof 
> surface
> in that area". Sort of like having a disaster response plan composed of 
> "It
> will never happen".
>
> I say be proactive on this issue or somebody else will make decisions for
> us. We just might learn something important along the way and avoid some
> really hard lessons in the future. Even if it's somebody else's customer.
>
> Like all of us, I am concerned about further limitations on PV 
> deployments.
> The CalFire guidelines pretty much strangle a lot of potential residential
> installations and, in my opinion, are overkill that provides very limited
> benefit. That sucks. Bonehead AHJs cause a lot of grief. That sucks.
> Uncertainty about materials supply or incentives... That sucks too. Big 
> oak
> tree in the neighbor's yard. That sucks. Foreman quits to go into business
> for himself. That sucks. Third microinverter failure in a week. That 
> sucks.
> When it comes down to it, we have to deal with all kinds of things that 
> suck
> on a daily basis. Some avoidable and some not. Let's avoid the avoidable
> ones.
>
> $0.02001
>
> Solar Janitor
>
> PS... On a brainstorm note: It occurs to me that laying a corrugated metal
> roof skin over the top of an existing roof material should effectively
> increase the fire resistant characteristics to an acceptable level. A
> physical barrier. A roof condom. I figure that you wouldn't need to attach
> it to the roof... Just secure it to the standoffs and rails, attach some
> stiffener system to eliminate the wind-rattle, and you're golden. Just a
> thought.
>
>
> -----Original Message-----
> From: Michael Welch
> Sent: Thursday, September 23, 2010 12:50 PM
> To: RE-wrenches
> Subject: Re: [RE-wrenches] PV and Roof Flammability
>
> Hi gang. It seems to me that an upward sloping gap between the flammable
> roofing materials and the PV modules would result in a chimney effect for
> any fire in the roofing materials whether the fire is a result of the PV
> system or not.
>
> Combine that with the modules blocking access to a roofing fire, and maybe
> you have their concern.
>
> Just guessing here.
>
> Ryan LeBlanc wrote at 12:43 PM 9/23/2010:
>
>>Good points all,
>>
>>Has anyone had an experience where you've seen new shingles and PV
>>installed at the same time, where the shingles became prematurely
>>brittle beneath the array, where you could say for sure it was PV's
>>fault?  I too have tried to keep an eye on that, but I never can pin it
>>on PV, most are still retrofits, not lending any credible data.
>>
>>As long as the PV doesn't cause the fire, due to:
>>- Ground-faulted or otherwise compromised wiring
>>- Melted cheap-o junction boxes (lowest cost import products)
>>- Improperly wired roof mounted combiners (reverse polarity)
>>- Conduit/expansion fitting errors (like the TARGET fire)
>>- The fact that they simply will be in the way if firefighters have to
>>vent...
>>
>>Are we saying there is evidence that the "PV shade structure" can
>>increase the flammability of the roof product from combustion, due to
> proximity?
>>Trapping a combustible level of heat beneath the array that can dry out
>>and set fire to comp shingle? Hypothesis = 1 in a billion, necessary to
>>look at, sure.  Maybe 1 in a 1/2 billion for wood shake.
>>
>>Solutions I see then:
>>-They can be stuck to the roof, no air gap no problem.  Well, except no
>>possibility of removal for firefighters.
>>-Integrated into the roof, then maintenance and wiring is buried.
>>-0-5" (or whatever), no go due to too close.
>>-5" (or whatever) and higher, Ok, due to adequate airflow and lack of
>>proximity.
>>
>>So now we're going to need PV compatible roof product ratings?  PV has
>>a great track record ratio of installs to related fires, and this is
>>with a majority of installers NOT really knowing what their doing, me
>>included.  I hope we don't have to get more expensive as a result,
>>let's make the roofers tell us which ones are not compatible, and then
>>we can just tell them to stop installing it.  PV is too important. :o)
>>
>>I have to believe that having a roof covered by PV, especially in CA,
>>could also Help Prevent fires from falling embers from forest and field
>>fires, the occasional PG&E GAS LINE EXPLOSION BLOWING UP WHOLE
> NEIGHBORHOODS, etc.
>>
>>Most material science would indicate that shading of the roof, the
>>overwhelming majority of the time, will extend roofing product
>>lifespan, often significantly, and help keep the attic a bit cooler.
>>
>>Let's not let officials overdo it, as I get more frustrated and
>>educated about this overly passive and tolerant society of ours, I'm
>>beginning to believe a little collateral damage is completely
>>acceptable.  It obviously is for the big boys of energy, money,
> automobiles, policy, etc.
>>
>>*See BURNING THE FUTURE (Coal), GASLAND (Natural Gas), FLOW (For Love
>>of Water), all the oil movies (oil), The Cove (Dolphins)... etc.
>>
>>We could probably have a roof fire per day, and still be doing better
>>than these "A" for alternative holes, except we'd get blitzed by the bad
> guys.
>>Thanks again for keeping watch guys.
>>
>>$.0001
>>
>>P.S. WHERE IS OUR FEED-IN TARIFF?... oh what, the PPA's have it
>>covered, oh yeah, they don't want one, great for us.  Is anyone
>>watch-dogging these PV finance guys?  They're the ones that scare me
>>the most, you know, the ones that get lost at the "T" for
> "truth-in-lending".
>>
>>Thanks for your relentless service to the industry that could save
>>America, if only they would let us.
>>
>>Ryan J. LeBlanc
>>NABCEPT Certified Solar PV Installer
>>Cell: 707.591.1950
>>Direct: 707.536.9839
>>ryan at NaturalEnergyWorks.com
>>http://www.NaturalEnergyWorks.com
>>
>>
>>
>>
>>_______________________________________________
>>List sponsored by Home Power magazine
>>
>>List Address: RE-wrenches at lists.re-wrenches.org
>>
>>Options & settings:
>>http://lists.re-wrenches.org/options.cgi/re-wrenches-re-wrenches.org
>>
>>List-Archive:
>>http://lists.re-wrenches.org/pipermail/re-wrenches-re-wrenches.org
>>
>>List rules & etiquette:
>>www.re-wrenches.org/etiquette.htm
>>
>>Check out participant bios:
>>www.members.re-wrenches.org
>
>
> _______________________________________________
> List sponsored by Home Power magazine
>
> List Address: RE-wrenches at lists.re-wrenches.org
>
> Options & settings:
> http://lists.re-wrenches.org/options.cgi/re-wrenches-re-wrenches.org
>
> List-Archive:
> http://lists.re-wrenches.org/pipermail/re-wrenches-re-wrenches.org
>
> List rules & etiquette:
> www.re-wrenches.org/etiquette.htm
>
> Check out participant bios:
> www.members.re-wrenches.org
>
> _______________________________________________
> List sponsored by Home Power magazine
>
> List Address: RE-wrenches at lists.re-wrenches.org
>
> Options & settings:
> http://lists.re-wrenches.org/options.cgi/re-wrenches-re-wrenches.org
>
> List-Archive: 
> http://lists.re-wrenches.org/pipermail/re-wrenches-re-wrenches.org
>
> List rules & etiquette:
> www.re-wrenches.org/etiquette.htm
>
> Check out participant bios:
> www.members.re-wrenches.org
> 