Ponding on flat gym roof 1

[JoeBaseplate]

I have a 68x100 gym with flat roof (metal deck on LH joists). Support walls are precast with 2’ parapet (may change). Ground snow load in the area is 25 psf. Per ASCE-7, as I understand it, I am going to check the lowest point on the roof for ponding with primary drainage system blocked. But what is the maximum water height that I should be using. Is it supposed to be height of the parapet?

 

[JedClampett]

Your drain detail should have an accomodation for the drain being blocked. Usually, there's an overflow feature like a scupper or an independent drain two inches higher than your main roof drain.
I would say to check for that two inches of water on the roof. If the roof is sloped, you shouldn't have to check for the load on the whole roof, just the region unitl the roof slopes up two inches.
If you have to check the roof for the two feet of water (125 psf)on it, it's going to fail. But it does solve any ponding issues!

 

[Ron]

Not sure of your location, but most codes in the US require checking for a 100 year storm if no supplementary drainage on the roof.

JC's example is conservative and certainly works.

 

[hokie66]

A flat roof will leak. I don't know why they continue to be built in the US. Guess it must be that the steel joist lobby is strong.

 

[dhengr]

Hokie.... I think it has more to do with the fact that we all need a good leak every so often, but I prefer not to do it in the steel joist lobby. I’m a rather private person.

 

[hokie66]

Best explanation I've heard.

 

[msquared48]

They usually don'tr serve beer on the roof, so I wouldn't worry. Just stay in the lobby.

Mike McCann
MMC Engineering

 

[paddingtongreen]

A war story. I did the design for a mall, 4" of water, a fifty year storm,, the roof drains were limited throat so that the water would take 24 hours to drain, so I put scuppers at the 4" level to protect it. I left that company soon after, and after a few months at the new job, I got a call from the field inspector for the mall. He said the contractor didn't like the scuppers at 4", he was going to put them in at 12" because it was more convenient, was that okay?.



Michael.
Timing has a lot to do with the outcome of a rain dance.

 

[csd72]

At least they asked you! I have heard of these things just being changed as they are an architectural detail.

The calculation for ponding is quite simple, if your deflection under dead load and applied loads is more than about 40% of the total fall over that span then you have a problem.

 

[JoeBaseplate]

Thanks for all the info. I have done pondnig calcs a few times and the way I approach is that I assume that the primary drainage system is blocked and then assume water height over the maximum deflected roof to the secondary drainage point.

Csd: sorry I didn’t get what you meant by “40% of the total fall over that span”.

 

[csd72]

The total fall over the span is the difference in height between one end of the span and the other. Rather than the total roof fall which may be over multiple spans.

I am assuming you are using rafters that fall with the span rather than purlins which run across.

 

[TXStructural]

Why are you using a flat roof rather than low slope (1/4"+ per ft)? A low slope roof with a free drain is not subject to ASCE 7 ponding instability. Pitch it to the sides, raise the peak to 34/4" (8-1/2"), put scuppers in the parapet as secondary drains, and call it a day.

As for the ponding, you would consider the height of water which could occur if the primary drain is blocked. Add that weight to the roof, look at deflections, again look at the added water weight resulting from the deflection, add that weight if it is significant, and so on, hence "ponding instability". This local deflection due to local impounding of water is the mechanism for excessive load development. Look at the commentary to ASCE 7 for examples and guidance.

 

[JoeBaseplate]

TX, I have actually talked the arch out of flat roof. I am thinking of doing specing a LH joist with double pithced top chord, I think they can do that without a major additional cost.

But out of curiosity, from what I read in chapter 8 of ASCE-7, ponding applies only to relatively flat roofs and the ponding load does not have to be combined with snow. Correct? Sorry I am writing this from the phone while in a boring meeting, so I could be wrong.

 

[MiketheEngineer]

DO NOT DESIGN OR BUILD A FLAT ROOF. It will haunt you for years to come...

 

[JoeBaseplate]

Mike, I am actually going to be a frequent visitor to the facility in question so I'll take your word for it. No flat roof.

 

[MiketheEngineer]

Case in point. We leased out a new warehouse and built a 2,000 sq ft "office space" inside and under existing 18' tall ceiling. No signs of leaks.

June 2010 - 7'' of rain that month - no problems.

Jan 2011 - 3-4'' of rain - much colder - leaks like a sieve???

Driving us crazy....

 

[TXStructural]

"Relatively flat" and there must be a free-draining arrangement, like edge scuppers as a primary or secondary drain. Just be sure the drains are sufficient (the job of the mechanical engineer.)

I investigated a collapse where there were very strong winds which may have impounded water under the sustained wind pressure, or may have flipped some TPO over all of the drains at the corner of the building, or may have damaged joists causing either ponding or directly collapsing the joists. The roof was 1/4"/ft pitch and we figured that the water could possibly have been 10+ inches deep.