Roof Drain Restrictors

[SteelPE]

I’ll try to keep this short.

I am designing a building where there are some drainage problems. To save money the civil engineer/contractor would like to reduce the amount of drainage discharged into the street drainage system by installing some “flow restrictors” on the main drains for the roof. It appears that the flow restrictors are a 6” high PVC with holes drilled in the side to restrict the amount of water flowing into the drainage system. On the top of the PVC pipe is a grate that would allow the water to freely flow into the drainage system (if the water ever got to be that high). Please see attached.

Due to the location, the building has already been designed for a roof snow load of 35psf (code requirements).

Question, should the roof design load be increased to reflect the potential of 6” of standing water on the roof together with the snow load? Or is it unrealistic to say that you would have the full 6” of water together with the full snow load?

 

[msquared48]

I would be worried about icing and blockage of the drain. How is this scenario to be handled?

I am not a big fan of rooftop detention, but if it must be - another thought. I would be nice, that if the works got plugged, the top of the proposed drain would be large enough, with enough air, to lift the restrictor out of the pipe, allowing the water to escape - unless the Architect is providing overflow scuppers at the outside walls.

Mike McCann
MMC Engineering

 

[SteelPE]

They had mentioned over flow scuppers at the edge of the roof in case the drain was plugged. However, due to roof pitch, I believe they would have to install an auxiliary drainage system.

This whole thing seems weird to intentionally retain water on a roof.

 

[jike]

If you are going to retain water on the roof then you need to go thru a ponding analysis.

If the roof pitch is significant then you will need separate interior overflow drains as you stated. We always ran those thru the outside building wall so hopefully someone will notice the "plugged condition".

We used to use undersized drains years ago with flat roofs but got away from them because the ponded water would cause the roof to wear out prematurely.

 

[Teguci]

ASCE 7 calls for Roof Live Load, Snow Load or Rain Load.

Rain on Snow surcharge is addressed under snow loads but will not apply in your situation, pg<20 psf.

Controlled drainage systems are addressed under rain loads in ASCE 7.

Most of this information is copied in IBC without much change.

 

[cvg]

the first design will easily plug with leaves. the second design may actually be less likely to plug up. However I believe intentionally ponding water on the roof is foolhardy.

A beehive grate would be my recommendation.

http://www.pacificmarine.net/pdf files/construction castings/Drainage Grates.pdf

 

[UcfSE]

Don't forget that it's 6" of standing water (static head) plus the hydraulic head of the water over the drain inlet.

 

[msquared48]

UcfSE:

From the sketch, I think the dead storage is only a couple of inches locally at the inlet, with the max live storage of around 4" for a head to the orifice of 3 to 4", plus, as you mentioned, any hydraulic charging or backwater effect to attain the orifice design flow.

Also like like this could be a multiple orifice condition.

If this is an interior drain, don't forget to install multiple cleanouts at the bends to mitigate future headaches.

Mike McCann
MMC Engineering

 

[cvg]

from my experience, these small grates clog very easily with leaves, pine needles, rocks and other debris. this can reduce the flow down to a fraction of the unclogged capacity and can even block the drain entirely. The head you will need to drain the water from the roof is dependant on the capacity of the "partially-blocked" drains and secondarily on the level of any overflow scuppers that may be provided. This could easily exceed your 6 inches. A significant safety factor should be used to account for at least 50% blockage of the grates.

 

[Teguci]

Structuraly you will have to design for the ponding water up to the auxilliary drains plus a little for flow.

Realistically, you are going to clog everytime it rains and have no problem locating all of the leaks in the roof.

Why not use a standard roof drain?

 

[plasgears]

I built a lab addition that had rain retention in the parking lot. The Chgo Metropolitan Sanitary District obliged us to do this.

This leads to the idea of remote retention off the roof; free flow off the roof into a retention structure like a tank rupture retention structure. Let that flow down thru a restrictor. Favor the roof against collapse.

 

[lrhg]

Did you consider a green roof to help solve the detention issue? Rather than all the water collecting at the low point of a clogged drain, water (at least some of it) will be retained more uniformly in the soil on the roof and eventually evaporate.

 

[tngolfer]

SteelPE,

Whether you have the flow retention device or no device, if you have 35 psf of snow you are going to have more than 6" of snow on the roof. Since that is the case, I believe it doesn't matter if you have the flow retention device because the rain load will be on top of the snow and not be able to get to the drain. Because of this, I do not believe the 35 psf snow and 31+ psf rain load are additive.

 

[jheidt2543]

Don't forget that if the roof drains are plugge and as the water gets deeper, the roof deflects more so the water gets deeper and the roof deflects more.... I just finished reading a paper where this happened and even though there were overflow scuppers, they were set too high. The roof nearly went down.

 

[hokie66]

jheidt2543,

That is what jike was referring to. It is called ponding.

 

[msquared48]

I have the fourth edition of the AITC, and there is a very good write-up on ponding and the requirred amplification factor on pages 5-315 through 5-325.

As I stated before, I do not like roof ponding. Just think about the sole benefit though...when the fire burns through the roof, it will be put out, so the fire insurance will be less. This should also negate any need for sprinklers.

Seriously though, you probably should consider adding at least a portion of the water weight to the seismic calculations. The limit for including the snow is 30 psf or around 5.5" of water. Although I have not personally found this to be a code requirement, I believe it would be the intent of the code, since the ponding can happen any time of the year and will generally be around more than snow depending on your location, I would seriously consider adding at least a portion of the total weight of the water. Just a thought.

Mike McCann
MMC Engineering

 

[dik]

Control flow drains have an advantage that they reduce the load on the storm sewer during a rainfall.

I generally provide scuppers around the perimeter and design the roof for the full amount of contained rainfall, including provision for ponding caused by deflection of the OWSJ as well as the structural steel frame.

Dik

 

[cvg]

speaking from experience (my house has a flat roof with scuppers) - the scuppers can and do also plug up. Maintenance on the roof is difficult as it involves a trip up a ladder, across a section of mansard roof, hop over the parapet wall and then hope you brought all your tools up with you or you are going back down the ladder to get them! Providing on-site storm water retention can be done in many different ways and on the roof is the method with the highest risk to the owner. yes it can be designed and constructed well. With good maintenance and some luck it will work without leaking. However, in an imperfect world, expect that it will leak.

 

[SteelPE]

I realize that a roof detention system is undesirable. Unfortunately I believe that we are going to be pushed into the system. As much as I don’t like contractors pushing me around it appears that even if I’m ultra conservative the project will still save big $ (from talking to some steel suppliers).

The way that it looks right now, I have 7” of pitch from the edge of the building to the drain. No parapets or gravel stops are shown at the edge of the roof… So this means the maximum amount of water the system can hold is 7”. I will have to increase the roof members to take this load and then check the ponding serviceability criteria. If required I will increase the members stiffness to accommodate this additional criteria. I will just end up being a little conservative in my loading criteria. From what I can see in the code there it is not required to consider both rain loads and snow loads acting together.

Interestingly enough, in talking to the civil engineer (who recommended this type of system) he says that they have designed several systems similar to this one. However, every single time the project switches back to a conventional detention pond (on the ground) system.

 

[connect2]

In our municpality specified rain load has been a design load for roofs for many years as they try to deal with their storm water capacity/capability problems. It has to be considered, with ponding re deflection, as a one-day rainfall in accordance with local climatic conditions regardless of installed drainage system flow restricted or otherwise. It is not combined with snow or its associated rainfall. In the case of scuppers its the lesser of the previous one day rainfall including ponding/deflection or rain water 30 mm above the scuppers applied horizontally over the surface(s)and trib. areas in question.