This is very interesting. Would not sand also drift like snow?

The building I am designing has no parapets. So even if it drifts, it falls off the building roof and gets collected in the gutter.

@BAGW-
Which code live load provision are you using to reduce sand accumulation loads below the 20psf you mentioned? How was the 20psf for the sand weight determined? Is your roof a flat roof, or a sloped roof? Some more details would be interesting, and informative, so we could chime in on possibilities.

Thaidavid

@thaidavid40

I am using ASCE7 Rooflive load section (Section 4.9). 20psf again comes from section 4.9 of ASCE and this was also confirmed by the contractor. This is an industrial building with sloped roof. The roof consists of trusses which have tributary more than 600 sq ft. So, we reduced the rooflive load to 12psf based on tributary area.

Thanks

A minimum live load of 20 psf is intended to provide for maintenance workers on the roof and should not be reduced.

Sand placed on the roof would normally be considered dead load if it could be kept in place. It is not clear how sand can be placed on a sloping roof without blowing away, but if there are any rooftop units, it seems to me that the sand could drift unevenly in some locations. It seems to be an inappropriate material to use on a roof.

BA

Loose dry sand typically weighs about 100 #/c.f.

Actually the 20 psf sounds about right. I was talking to someone just back from Afghanistan and he said there was a perpetual two inch layer of sand on everything. It didn't accumulate and it was forever moving but it was always present.

@ BAretired – The code says the roof live load can be reduced based on tributary area and the minimum needs to be 12psf for maintenance workers. I am reading this wrong? In our design we are using 12psf of roof live load and this was approved by the local municipality. Still for me 12psf of roof live load seems high for maintenance workers on this sort of building (Hangar consisting of trusses 200' long). And I feel the roof live load is not just the representation of construction workers because in load combinations its snow/roof live. So, I think sand is treated as roof live as well.

There are no roof top units on the roof and the roof is sloping. The building roof is 100ft high. I just can’t imagine how sand can be placed on sloping roof without being blown away. What I am getting to is what happens if the roof trusses are designed just for the dead load and not for roof live load? Or are we going to be conservative if we use 12psf of roof live in addition to dead for the design?

I just want to understand roof live load in such situations.

Thanks

@BAGW, For me, the problem was never a consideration because minimum snow load in my area is in the order of 30 psf. On some roofs, snow drifts so that some areas are designed for 100 psf or more plus roof dead load.

I don't know whether you are interpreting the code correctly or not because I don't know which code is applicable in the middle east but it seems to me that if sand can blow onto the roof as suggested by BUGGAR above, a little research is necessary to determine an appropriate live load.

The 200' length of trusses is a condition not stipulated in your original post and I tend to agree that for maintenance purposes, 20 psf may be excessive for such long spans but perhaps not for shorter span purlins between trusses and roof deck etc.

If your total design live load is 12 psf including sand plus maintenance workers, I believe it is too low but I have no special knowledge about how sand accumulates on roofs.

Another factor to consider is internal pressure resulting from hangar doors being left open during windstorms from various directions in combination with sand load. 12 psf leaves little room for error.

BA

We've done some low-slope roofs on barracks and admin buildings in Bagram and Baghdad, and used the 20psf unreduced to cover either the roof live load, or the sand load. You won't have both at the same time, so design for the single load. Good luck -
Dave

Thaidavid

I'm not sure why you won't have both at the same time. If there are people working on the roof, won't they be walking on the sand?

BA

@BAretired -
A good question, which we also asked. The Ops colonel said that they would only need to access the roofs to adjust or repair the antennas - thus a very minimal service live load. If any real roof work were to be needed, then the sand would have to be removed (swept off) in any event. We accepted his outline of how the roof loads would occur, and designed around that.
Dave

Thaidavid

@thaidavid,
I'm not convinced that the Ops colonel is a satisfactory reference for roofs in general. Sweeping sand off a large roof is likely to result in localized overloads; some areas will be swept bare while other areas will be carrying deeper layers as sweeping progresses.

If loose dry sand weighs 100 #/c.f. as suggested by oldestguy above, then 20 psf represents a thickness of 2.4" of sand and 12 psf represents a thickness of 1.44" of sand.

Heavy rainfall on a sloping roof would tend to wash the sand down the slope. Light, steady rainfall would tend to increase the unit weight of the sand from dry to damp or even saturated.

The 20 psf minimum live load specified by code is not intended to cover a layer of sand over the entire roof. It is a live load and perhaps it could be reduced on the basis of roof area. Snow load cannot be reduced on the basis of roof area and it seems to me sand should be treated in similar fashion.

BA

@BAretired -
I think that you and I may be looking at the environmental aspects surrounding our past design effort from differing perspectives. We were designing for a perpetually dry environment, where heavy rainstorms rarely occur, and light sprinkles are just that - very light, to the point of almost not leaving any trace on the ground. Many times, light rainfall would evaporate again in the air, before hitting the ground. Removing the sand was usually done by sweeping, or blowing, the sand downslope. As it was swept, it tended to continue moving downslope, and didn't accumulate in any appreciably thicker layers. I suppose there were probably some roof areas locally loaded heavier than others, but not globally so. We didn't use any area reductions on these loads for the very reasons you mentioned. As as for the Ops colonel, well, he outranked me, so we worked within his group's guidelines! None of this logic applied, obviously, to any of the flat roofs that were used.
Thanks,
Dave

Thaidavid