Ponding 2

[JedClampett]

I'm looking at the Ponding design requirements in AISC (9th Edition, Section K2). I don't understand where this would be a problem. It states that ponding must be checked "... unless the roof surface is provided with sufficient slope ... to prevent the accummulation of rainwater." Who wouldn't do this? All architects I know of are aware that it rains and slope roofs, using tapered insulation or just sloping the roof members. I would dare say they would be negligent if they didn't. I know there have been roof failures during heavy rainfalls, but I don't think that not providing a roof slope is the reason.
So when would this be an issue? Is there some slope that isn't sufficient?
The reason I'm asking is that I'm doing a review and formula K2-1 is about .4, but the roof has a 3/8 inch per foot slope.

 

[RARWOOD]

I would never assume that the Architect is designing the roof for proper drainage. They should be but sometimes they don't.

One case that you need to watch is beam camber. It might not cause problems in steel design, but with glulams it can be a problem. Also sometimes on jobs there are areas of roof transitions that get over looked or errors in bearing elevations.

If you have 3/8" per foot slope you should be Ok.

 

[UcfSE]

Not all rain water can freely run off of the roof. Most of the time I see it has to go through scruppers or internal drains. Even unclogged the water will develop some head as it builds up before it can get out. If the scuppers are clogged then you'll have both static head and hydraulic head as the water builds up to get out of the overflow scuppers. That's plenty reason to design for ponding. Many times when I'm looking at steel joists and girders ponding stability per AISC K2 will control the design.

 

[jike]

Years ago, there used to be a philosophy to use flat roofs and "controlled flow" roof drains which would restrict the flow of water to the sewer so that a larger sewer pipe did not need to be provided. The idea was to pond water on the roofs until the sewer had time enough carry the water away. Unfortunately, there are still those adjustable flow drains which are sometimes specified by plumbing designer that may not know the structural implications.

I recommend that you have a heart to heart talk with the plumbing designer to make sure that he has an non- adjustable drain specified and a proper overflow or scupper at the right height.

 

[DonPhillips]

You must also consider maintenance - or lack thereof - of the roof drains to avoid a failure, or assume some ice damming. And I have seem the plumbing and architectrual drawings not being coordinated and the roof drains are in the wrong location relative to the slopes of the flat roof. All are potential sources of evenutal failure.

Don Phillips

http://worthingtonengineering.com

 

[JedClampett]

My problem is I'm reviewing a project for other engineers who are pretty sharp. The AISC doesn't say to assume lack of maintenance or clogged drains and the scuppers are not flow restricted. They're going to trump me with the roof slope and say it's sufficient.
If AISC wants us to check ponding, they should give a minimum slope where this might be an issue instead of using terms like "sufficient slope." Maybe the new AISC does this.

 

[UcfSE]

If the roof in constructed such that there is no way for water to pond on the roof, then you're fine. Think of a typical plain house with a 4:12 sloped roof. The water just falls off the sides with no where to collect on the roof. No ponding design needed. If you have parapets or anything that can hold water if your drains become blocked, then you need to check ponding stability. It doesn't matter how much slope you have if the water cannot get off the roof as fast as it can get on the roof. The AISC should not have to tell us to assume the drains are clogged. We should know to do that. At some point to have to use judgment. They can't do it all for us. Check also chapter 16 of your building code. The FBC, based on the IBC, has am updated section for rain loads. You may have this also.

 

[DaveAtkins]

I read somewhere (the IBC?) that a roof need not be checked for PONDING if it slopes more than 1/4" per foot. That is the criterion I use.

That being said, let's get our terminology straight. Ponding has nothing to do with blocked drains or a lack of scuppers. If a roof has large bays and is sloped 1/8" per foot, the joists may deflect so much that they POND water, deflect, POND more water, deflect more, and so on, until failure.

If you have no scuppers or secondary drains, the IBC requires you to check for BLOCKED drains. This is much worse than ponding, since you must assume the roof fills up with water until the water spills over the top of the parapet. A ponding check will NOT prevent failure if you have no scuppers or secondary drains.

DaveAtkins

 

[Lutfi]

You may want to perform rigorous analysis on the primary and secondary members utilizing the equations for:

1. Simplified design for ponding
2. Improved design for ponding

These are contained in Appendix 2 “Design for Ponding” in AISC specifications for steel buildings dated March 9, 2005. This version of specifications supersedes all previous editions.

AISC members can download it from the web site in PDF format.


Regards,
Lutfi

 

[JAE]

DaveAtkins - could you please clarify your criteria for me....you said "if it slopes MORE than 1/4" per foot" you don't check ponding.

Most "flat" roofs in the US are set at 1/4" per foot. I'm just making sure you didn't mean "equal to or more than" 1/4" per foot.

With most "flat" roofs at 1/4" per foot, the question then is whether or not the ponding check in AISC has any significant meaning for those roofs. I think that DaveAtkins point that you are really checking the predisposition of the roof to be susceptible to ponding and that is really the whole point.

I have to admit that JedClampett's question has been in my mind in the past.

 

[haynewp]

AISC 2005:

"Ponding: Retention of water due solely to the deflection of flat roof framing"

"The roof structure system shall be investigated through structural analysis to assure adequate strength and stability under ponding conditions, unless the roof surface is provided with a slope of 1/4 in per ft or greater toward points of free drainage or an adequate system of drainage is provided to prevent the accumulation of water."


We make a living by what we get, we make a life by what we give.
Sir Winston Churchill

 

[Ron]

UcfSE has the key that we would use here in the Deluge State, otherwise known as the Sunshine State. As haynewp noted, the criterion is "free" drainage. Scuppers and drains don't usually constitute free drainage if the rainfall intensity is high and the time of concentration is low (both of which can occur on a roof!). I have seen buildups behind scuppers on roofs with adequate slope to as much as 15 inches above the roof surface, resulting in so much deflection of the perimeter support at a tilt-up wall that it pulled the roof flashing off the wall and dumped the water inside the building (nice safety valve, but a bit surprising to the secretaries). This happened more than once in Central Florida during a COMMON summer afternoon downpour.

A 3/8" per foot slope is great for drainage, and if the drainage edge is free (gutter or direct overflow) then the structural impact is minimal. If there are internal drains or scuppers, I would check it.

 

[DaveAtkins]

Ron,

I would blame that problem on the scupper design, not the lack of a ponding check. Again, what you are describing is more akin to a "blocked drain" than a ponding problem.

DaveAtkins

 

[UcfSE]

Lutfi, can we use the new steel spec for ponding design when it has nto been adopted by the state building code yet?

I think the point Ron and I are trying to make is that water can pile up on a roof even with a 1/4-inch per foot slope. Where water can pile up, I don't see how you can avoid checking for ponding stability, in addition to other things. If a blocked drain causes the water to pile up and you didn't check for stability, what will you tell your lawyer?

 

[JedClampett]

Thank you for all your responses. I think you see my problem. With a vague statement like the one in AISC concerning ponding design, the burden is placed on the designers. If I was a plaintiff's lawyer with a collapsed section of roof, this would be a dream come true. If anything, it makes our jobs harder. I think the new AISC looks a little better.
As far as my review project that started this thread, I looked at the architectural drainage plan and didn't like what I saw. Small scuppers, parapet roofs, one level of roof draining onto another, etc. I noted that I'm concerned about ponding issues and bounced it back to the design engineer. Luckily this is in southern California, so we don't have Florida or Texas type rains. But every once in a while....

 

[Ron]

Dave,
I agree it's a scupper design problem, but since most building design is not well integrated, we often see that disconnect.....the structural engineer doesn't design the roof drainage, but sometimes has to consider that whomever did, might not have done such a great job.

 

[DaveAtkins]

I feel the need to beat this dead horse a little more...

If a roof is sloped at 1/4" per foot AND has proper drainage, then ponding CANNOT occur. The deflection of the joist at this roof slope is not enough to allow water to build up.

If a roof does NOT have proper drainage, the issue is not ponding. Rather, the live load must be increased (sometimes tremendously) to make the roof safe. The average roof live load in my part of the country is about 30 psf (for snow). If you don't have proper scuppers or secondary drains, and the roof is designed for this live load, then it is only good for about 6" of water. Parapets can be alot taller than 6" sometimes--for example, a 2' high parapet with no scuppers or secondary drains will result in 125 psf live load before the water spills over the top of the parapet.

My point? Unless you want to seriously overkill the roof member sizes, make sure there are adequate scuppers or secondary drains.

DaveAtkins

 

[UcfSE]

Dave, to use steel joists for an example, let's say they are sloped 1/4" per foot. The girders holding them up will be perpendicular to the slope of the roof. In essence the girders will be flat. With an ultra-mojo storm that can dump 24 inches of water in one sitting, wouldn't you have water on the girder causing the unsloped girder to deflect and collect more water as it deflects? Yes it depends on how far away the girder is, but we'll say it's about 4-8 inches above the low end of the joists. Maybe I'm missing something.

Properly designed drains don't account for the improper placement of dead animals, sticks or leaves. I can put whatever notes on the drawings I want and that dead raccoon won't move out of the way. You're still back to primary drains blocked. Someone else sizes the drains where I'm from, so how can I ensure proper drains? We're not even in that equation.

 

[DaveAtkins]

Good points, UcfSE--but again, a ponding check will not suffice in the situation you describe. If the drains are blocked, you have to design for all the water that can collect on the roof.

DaveAtkins

 

[JedClampett]

And then you're back to my question. Are you liable for a roof failure due to blocked or insufficient drains based on the AISC code? I'd say it's going to cost your insurer some money, even though it's an impossible design criteria.