Flood Testing of Roofs

Log Out | Topics | Search Moderators | Register | Edit Profile

Author Message Ronald L. Geren, RA, CSI, CCS, CCCA, MAI Senior Member Username: specman Post Number: 235 Registered: 03-2003 Posted on Monday, April 17, 2006 - 03:32 pm:    I have a client that wants a "flood test" of the new roof system, and the contractor is willing to do it, but is wanting a procedure on how to go about conducting such a test. I know NRCA doesn't recommend flood testing (no help there), and ASTM has D5957 for waterproof membranes, but it specifically excludes roofs. Does anybody have a procedure they're willing to share that is moderately successful? Thanks! William C. Pegues, FCSI, CCS Senior Member Username: wpegues Post Number: 592 Registered: 10-2002 Posted on Monday, April 17, 2006 - 08:40 pm:    Ron, I don't know that it has to be an ASTM test. I have always required a flood test. Most of the time though, just an example area, but many times the full roof (we have noticed that now as more owner's hire their own waterproofing consultants, these consultants want the entire roof area flood tested). Typically what I say is (this is for an IRMA configuration)... Prior to placement of insulation and ballast, impound water in an area selected by the Architect, of not less than [[whatever area you require]] to a minimum depth of 1-1/2 inches. Plug any drains in the area and maintain water levels for a period of 48 hours. I note that some manufacturers actually advise that this be done, no standard is listed. I note the some of them state 2 inches minimum depth. Then you check for leaks. and repair if necessary. Mark Gilligan SE, CSI Senior Member Username: markgilligan Post Number: 73 Registered: 05-2005 Posted on Tuesday, April 18, 2006 - 04:45 am:    If you contemplate flooding your roof PLEASE check with your Structural Engineer. It is possible you will have to pick him up off of the floor. In many parts of the country roof design live loads are 20 psf which for large tributary areas can be reduced to 12 psf. 20psf corresponds to just under 4 inches of water. Given that the minimum roof slope is typically 1/4" per foot and many are built with 1/2" per foot slope it would not take a very big area to fail some roof members. Doing this before placement of ballast helps but still there needs to be some limits established. You may want to consider increasing the design load on the roof members if flooding is required. William C. Pegues, FCSI, CCS Senior Member Username: wpegues Post Number: 595 Registered: 10-2002 Posted on Tuesday, April 18, 2006 - 09:26 am:    Checking with structural is always wise. But if your project is in an area where you are required to have overflow scuppers and those are locaetd 4 to 6 inches above the top of the ballast top, then flooding to 2 inches over the membrane is not going to be a problem. Hopefully the structural engineer knows that the roof has to potentially carry that much load as it must by code due to the code scupper requirement. Which brings up a point that one probably better tell the engineer that anyway. We had a project just after that requirement sarted. The membrane type was in question and we convinced a condo developer that they would be better off with an IRMA configuration than with a fully adhered single ply system with insulation under the membrane and no ballast. The strucutral engineer sent back an email to the project manager that he would have to design for the increased ballast if we did that. I looked at the PM and said, "maybe before he starts embarrasing himself you should show him these scuppers that are going to make the roof hold a lot of water potentially". The engineer was not aware of that code change, and he had a lot of last minute redesign to do. William John Bunzick, CCS, CCCA Senior Member Username: bunzick Post Number: 507 Registered: 03-2002 Posted on Tuesday, April 18, 2006 - 09:58 am:    William makes a good point, but it is worth noting that the requirement for secondary drainage is not always met with the use of scuppers (could be secondary internal drains) so the roof structure may be designed properly for a much lower amount of water depth than would be the case if scuppers were providing the secondary drainage. While this may seem obvious, if a flood test is performed it must be done before ceiling finishes are installed so you can monitor for leaks. If you have a concrete deck, you may not be able to get any useful information from a flood test as a slow leak may not show up during the time of the test. One may be able to isolate limited areas of the roof with sandbags to keep water depths relatively low, but I don't know how practical this really is. Does anyone know? William C. Pegues, FCSI, CCS Senior Member Username: wpegues Post Number: 596 Registered: 10-2002 Posted on Tuesday, April 18, 2006 - 10:21 am:    I was just putting up the comment because of the 'interesting' situation that it created. I have also had to give a swift look to a PM when the curb/sill to the stairs from the roof were a couple inches lower than the height of the scupper too -grin! Fortunately, none of those have made it out of the office. Anyway, yes, good to check on structural, good to assure finishes are not in yet. Also, much more critical for purposes of water weight when the roof is an exposed membrane with no ballast. If no scuppers, then it is likely that the roof deck may not carry a flood test load...at least not for the whole deck, and maybe not even in selected areas for impounded areas. Impounding, I see it done all the time - a good number of our projects are done with impounded areas. I have not visited one in years, but I recall from photos that our CA that they used sand bags to build up a bulkhead, and then did a temporary flashing to the membrane that draped up the face of the sandbags. So the impounded area did not seep out onto the deck. We have noticed that with the increase in owner's hiring their own waterproofing consultants for document review and CA monitoring that all of them want a flood test of some kind, about 70% of them want to flood the whole roof. That practice includes the Washington DC region (out to the mountains, over to the eastern shore, north to Deleware and south to Richmond. The same is true in our Dallas office which has encountered this requirement in the major cities across Texas and Colorado. William