Thread: Straw breaks camels back.

Or ,,, So you want to be a roofer.


http://komonews.com/news/local...wnhouses

"Just after 3pm Tuesday, a pile of roofing equipment and materials suddenly fell through the roof of the Lynnwood apartment building. Police do not know why the structure failed, adding there’s no obvious sign of negligence. They will investigate the cause further."


Really, don't know why structure failed and no sign of negligence. I think it was a bird landing on the stack of shingles.

Similar Threads:

• Straw bale construction
• fire breaks
• Straw mixing question
• Jerry Breaks 6000!
• Breaks in Duct Chase

It does not indicate whether the material may have been stacked, thus creating a concentrated load, thus contributing to the failure. I have seen a few situations where the load may have exceeded the building design load for a floor or roof. Building codes typically provide design loads for structural conditions.

Based on the photo it appears that most of the material was concentrated together in one general area.

Looks like a well stacked mass sitting in the attic.

Look at 1:02 of video it looks like they have 9 to 12 square stacked in two locations that they are removing. I would imagine that the load that fell through was similar.

On most house roofing jobs, the supplier delivers the shingles 'on the roof', on some smaller house roofs, the roofer may have the shingles dropped off on the ground and hand carry the shingles up to the roof ... but ... on a job that size - almost guaranteed that the supplier delivered the shingles 'on the roof'.

That makes the roofer on the liability ladder, but it puts the supplier fully and firmly 'holding and supporting' that liability ladder - and it is the supplier who would likely have the most insurance and deepest pockets.

That's the only 'good thing' to that ... if that can be classified as a 'good thing'.

Someone apparently forgot about 'distributing the weight around the roof' when they loaded that roof with shingles.

If we go by your 9-12 squares, and just use 10 squares (easy on the math that way) and presume that the shingles weight between 200 to 350 pounds per square (depending on the quality of the shingles), so at 275 per square (spitting the weight difference) for 10 squarea is 2,750 pounds placed in an area the size of two pallets (a standard pallet is 48" x 40", thus two pallets would be 8' x 40 ", or, simply 8' x 4' to allow for spacing of pallets) - 2,750 pounds on 64 sf = 42 psf. A Google search shows that a typical pallet of asphalt shingles weight about 3,000 pounds, so that estimation is likely close.

That is 43 psf IN ADDITION to existing dead load, plus whatever live load may have been applied overnight, and live loads are not calculated for the duration that that 43 psf of additional dead load was there.

Let's just say 'See, that is a perfect example of what NOT to do.'

In my neck of the woods the delivery to the roof does not mean that the supply house is on the roof. The roofer has to have someone to unload/accept the shingle to the roof surface. Some delivery is by conveyor belt and others are boomed up for roof delivery. Otherwise it is onsite (ground) delivery which the supply house will do, drop and run.

I do not think that the supply house is liable, just can not believe they would take on stacking shingle on the roof. But stupid is as stupid does sometimes.

I can hear it from here. "We have been stacking shingle like that for 20 years and never had a problem." Though you do have to take into account that it is the Left Coast and they may have a more liberal attitude about such things.

It appears they did not redistribute the shingles evenly across the trusses.
Typically 67 pounds per pack. 3 packs per square.

If I remember correctly a square of 3 tab shingles, including felt and nails, was about 210 pounds.
I use to carry my own and nail ><7 square on a plan roof.

Originally Posted by Garry Sorrells

In my neck of the woods the delivery to the roof does not mean that the supply house is on the roof. The roofer has to have someone to unload/accept the shingle to the roof surface. Some delivery is by conveyor belt and others are boomed up for roof delivery. Otherwise it is onsite (ground) delivery which the supply house will do, drop and run.

I do not think that the supply house is liable, just can not believe they would take on stacking shingle on the roof. But stupid is as stupid does sometimes.

I can hear it from here. "We have been stacking shingle like that for 20 years and never had a problem." Though you do have to take into account that it is the Left Coast and they may have a more liberal attitude about such things.

The contractor has the ability to place entire pallets on the roof?

And the supplier does not know how heavy a pallet of shingles is?

I still think the supplier (or whoever placed the shingles on the roof like that, and I suspect it is the supplier) will be found to be the responsible party.

Originally Posted by Jerry Peck

The contractor has the ability to place entire pallets on the roof?

And the supplier does not know how heavy a pallet of shingles is?

I still think the supplier (or whoever placed the shingles on the roof like that, and I suspect it is the supplier) will be found to be the responsible party.

There was no pallet on the roof, just shingles and a 2x with 2 jacks.

The shingles on the roof were stacked using roof jacks. Roof jacks that the roofer installed. A supplier would not set the jacks, at lease in my neck of the woods. In my roofing world the supplier never takes the responsibility/liability to put their men on the roof or the liability of the product. The roofer must have set the jacks and stacked the shingle, but again it is the Left Coast.

As for the weight, the roofer knows the weight of the single. The supplier know the weight of the shingle. The roofer should care and the supplier could care less since the supplier is just delivering the materials to the roofer. What the roofer does is the roofer's problem not the supplier's. But again it is the Lest Coast so who knows.

As to my last post, "It appears they did not redistribute the shingles evenly across the trusses."

Look at the shingle bundles below. They are grouped and not distributed evenly across the entire deck.
One side of the roof deck is stripped of old shingles. No barrier is in place.
I guess the roofer was preparing to install new shingles on side.

roof collapse.JPG

Originally Posted by Garry Sorrells

There was no pallet on the roof, just shingles and a 2x with 2 jacks.

I do not think roof jacks were installed to support the shingle bundles, Garry.
The bundles appear gravity bundled on end and not on the sides.
Looks like a 4/12 pitch deck, so jacks would not be necessary. Safe for walking.

On 5/12's I used 2x4s with galvanized metal straps at the ends and only jacked 6/12 and up.

Originally Posted by Jerry Peck

On most house roofing jobs, the supplier delivers the shingles 'on the roof', on some smaller house roofs, the roofer may have the shingles dropped off on the ground and hand carry the shingles up to the roof ... but ... on a job that size - almost guaranteed that the supplier delivered the shingles 'on the roof'.

That makes the roofer on the liability ladder, but it puts the supplier fully and firmly 'holding and supporting' that liability ladder - and it is the supplier who would likely have the most insurance and deepest pockets.

That's the only 'good thing' to that ... if that can be classified as a 'good thing'.

Someone apparently forgot about 'distributing the weight around the roof' when they loaded that roof with shingles.

If we go by your 9-12 squares, and just use 10 squares (easy on the math that way) and presume that the shingles weight between 200 to 350 pounds per square (depending on the quality of the shingles), so at 275 per square (spitting the weight difference) for 10 squarea is 2,750 pounds placed in an area the size of two pallets (a standard pallet is 48" x 40", thus two pallets would be 8' x 40 ", or, simply 8' x 4' to allow for spacing of pallets) - 2,750 pounds on 64 sf = 42 psf. A Google search shows that a typical pallet of asphalt shingles weight about 3,000 pounds, so that estimation is likely close.

That is 43 psf IN ADDITION to existing dead load, plus whatever live load may have been applied overnight, and live loads are not calculated for the duration that that 43 psf of additional dead load was there.

Let's just say 'See, that is a perfect example of what NOT to do.'

Jerry, I agree. I cringe every time I see roof contractors put the product on the roof because that will be a guaranteed roof job as the homeowner waits sometimes for the insurance check to arrive. To much weight as you explained it. I just wonder how much GL the contractor can carry. Structural Engineer is not having a good day. Sad to see. The shingle supplier should only deliver when the job is started. There is a home owner in my neighborhood that has had shingles on his roof since October of 2016. No joke. I guess it's all about the dollar.

Originally Posted by ROBERT YOUNG

I do not think roof jacks were installed to support the shingle bundles, Garry.
The bundles appear gravity bundled on end and not on the sides.
Looks like a 4/12 pitch deck, so jacks would not be necessary. Safe for walking.

On 5/12's I used 2x4s with galvanized metal straps at the ends and only jacked 6/12 and up.

Look at 1:02 of video, the red things are roof jacks. There is a 2x across them to stop the shingle from sliding. Again, red things on bottom edge of pile of shingle are roof jacks.

I have seen local suppliers, including big box stores, dump trusses off a flatbed by tilting the bed and driving away. The trusses (tied together and laid flat) are 'bent' one way as they slide off the truck and the other way as they bounce onto the ground. I have seen loose gussets on every truss in attics, and assumed it could be from this type of delivery method. I've call this and had engineers write 'reports' that the loose gussets are not a problem.

I'm wondering if something like that is what happened here.

- - - Updated - - -

I have seen local suppliers, including big box stores, dump trusses off a flatbed by tilting the bed and driving away. The trusses (tied together and laid flat) are 'bent' one way as they slide off the truck and the other way as they bounce onto the ground. I have seen loose gussets on every truss in attics, and assumed it could be from this type of delivery method. I've call this and had engineers write 'reports' that the loose gussets are not a problem.

I'm wondering if something like that is what happened here.

See where the trusses broke? Right at the splice in the top chord, which is a weak spot, because they never put a web under that splice.

JMO.

Upper cords and/or nail plates.
What came/gave first, the chicken or the scrambled egg.

Originally Posted by ROBERT YOUNG

Upper cords and/or nail plates.
What came/gave first, the chicken or the scrambled egg.

I meant the nail plates in the top chord, where we often see that splice. When I first started noticing them in trusses, I thought they were mistakes.

From the photo it looks like 30 to 40 bundles of shingles spread over a few 2x4 top chords that span 8 feet or so between supports. I would have to do the math, but the top chord could have fractured or the plate joints could have failed.

Jerry et al being right in the ball park on weights.

To add a little data to the discussion.
The shingle were Malarkey- Legacy a laminated shingle.
There are two versions of Legacy.
They come as 12sq/pallet, 4 bundles/sq, 275lb/sq or 325lb/sq.
My count came to 43 bundles or 10.75 squares = 2750lb or 3250lb of shingle.
There dies bit appear to be any other materials other than shingle in the collapsed area.

If you look at 1:02 in video to the right of what is being removed they have additional materials stacked on the shingles adding more weight though not at tipping point to fail.

It would be an interesting exercise to determine what the load forces were on the joints of the cord at the splice and at the ridge as they relate to connection methods.

Some reading that I like for the deeper thinkers.
Truss Structures - Springer

introduction structural design - BeamChek

Chapter 3: Design Loads for Residential Buildings - HUD User

Originally Posted by John Kogel

I meant the nail plates in the top chord, where we often see that splice. When I first started noticing them in trusses, I thought they were mistakes.

I know John.
I am sure many have observed fink roof truss plate deficiencies.
As I drive the roads I observe flatbeds with Long/wide and unusually engineered roof trusses taking unnatural forces being delivered to building sites due to the flat beds size.

While inspecting I regularly observe fink roof truss nail-plates poorly attached at the web to upper or lower cord.
Howe trusses gussets attachments are typically/usually correctly installed.

Thanks, Garry.
At the ridge, the chords are in compression like opposing rafters, very strong.
At mid-span, there is that butt joint with two little gusset plates. First a bit of sag, then the ends start pulling away.
Or maybe one 2 X 4 split as Mark suggests, when the bird landed, eh?

I have seen trusses sagging in townhouses, (long narrow roof spans with minimal pitch) but the sheathing usually spans that joint, which helps a lot. There are lots of nails in the sheathing if it is done right. Not as good if it is soggy OSB and elctro-plated nails instead of good old plywood with hot-dipped galvanized nails.

So stupid is as stupid does and they got away with this before, by being lucky.
I feel sorry for the folks that own the units below that mess.
I'll bet there's a partition wall there that stopped the pile from dropping into their living rooms.

John,
Like you mention the splicing plate on the cord without a web underneath that splice relies on the plate in conjunction with the butt hinging at the splice to transfer/distribute the force vector loads being applied to that joint/splice.

From the pictures the failure appears to be at a splice. Due to the uniformity of the failure point in the cord. What I am not able to see in the pictures is any evidence of a splice plate on the top cord. So everything involves some speculation on my rt.

What I am curious is the amount of load/force that the unsupported (no web) joint in the top cord is capable of handling before it fails using a nailing plate as it appears in the field as opposed under test controlled conditions. Specifically, looking that after assembly the effects of transportation and installation stresses to the plate by way of flexing at the joints as it is handled and moved multiple times.

When time and tide allows I hope to get a call into a fabricator to see what the answers are. In reality it is all academic as to exercising common sense to distributing the load across this type of roof. Basically knowing that roofs are not designed to park cars on.

BUT, if someone else has the time to chase down the answers I and others would love the information.

It appears the trusses are 16" OC.
10 broken truss = 160".

The bundles appear 5 wide. (13-3/4")=80")

The load spanned 5 trusses.

IMO, a concentrated load.

Now you know why GCs can make serious money, to cover incredibly stupid errors by subs. Effin' roofers, right up there with drywallers.

It's entirely possible the supply company stacked that concentrated load but I doubt it. Supply companies will deliver and stack without the roofing sub or GC/superintendent on the job, and I've seen that happen, but any roofer in his right mind (and there aren't many) would be there. You have to be a total idiot and/or a newbie to stack it like that, and there are always two or three people stacking the roof, one loading and one or two stacking. At least one of those people would have the requisite one week's worth of experience but that certainly doesn't preclude that it may have been the supply house. I don't see any other bundles spread out and as it's Washington state with constant rain this time of year, they were stripping only one section at a time. The roofers may have been trying to save time, and it may have been a roofing company with its own equipment, like most of the mid- to larger-sized ones have.

I vote for the likelihood of the roofer screwing up but this whole thing just reeks of stupidity.

Edit: In rewatching the news video to try to determine the roof pitch, I noticed the roofing company was using their own or rented equipment. This, IMO, makes it more likely that the roofing company did indeed stacked the roof itself. Roofers are a unique breed.