Lessons Learned at RCI: Ponding Instability
A safe, streamlined process for analysis that is understood by all is required to ensure the risk of catastrophic failure is minimized.
Many interesting and important building envelope issues were discussed in the educational seminars at the recent RCI International Convention and Trade Show in Houston this March. Dr. Stephen L. Patterson, RRC, PE with Roof Technical Services Inc. and Dr. Madan Mehta, PhD, PE with the University of Texas at Arlington inspired us with their presentation on roof drainage design, roof collapses, and the related codes. At Zero/Six Consulting, we regularly work on re-roofing projects, where the issues being discussed in this seminar prove especially relevant. We believe any and all life safety concerns should be studied and simplified as much as possible. A safe, streamlined process for analysis that is understood by all is required to ensure the risk of catastrophic failure is minimized.
Areas without ÂŒâ per foot slope causing ponding
Controlling a situation, rather than responding to it, is the key to this process. Ponding on low-slope roofs is a life-safety hazard that can be anticipated and mitigated with proper analysis. An issue that has been observed in the AEC community is the building code doesnât clearly address the roof drainage situation for existing buildings. The different disciplines involved in dealing with roof drainage is  another issue. Architects, engineers, pluming consultants, and roofing contractors are all obligated to ensure the structure can handle the worst-case-scenario for rain loads. My opinion, as an engineer, is the engineer takes the lead and becomes responsible for the final roof drainage design. The engineer is understandably focused on structural design, but simply getting the rain load from the architect or plumbing consultant does not adequately address the potential issues. The structural engineer should likewise engage with the architect to ensure roof slopes will properly collect water at the primary drains. In the case of a primary drain backup, the overflow design handles water without ponding. Finally, coordination with the plumbing contractor is necessary to ensure the roof drain size and height will be set properly, and the conductors are designed and installed to be fully independent from primary drains to superintend rain to the overflow as required.
The most important factor in preventing ponding on the roof is providing adequate slope to quickly and efficiently get water to the primary and overflow drainage. The newer building codes require one quarter inch per foot of slope and define those areas without adequate slope as âsusceptible baysâ. Susceptible bays include areas where water is impounded â when the secondary drainage system is functional while the primary drainage system is blocked.
Progressive deflection or instability are worst case scenarios preventable with adequate drainage.
Two basic outcomes can occur with flexible roof supports and inadequate slope:
as ponding water gains depth and weight, it causes deflection in an area, which means an even greater water thickness will result. This progressive deflection will continue to expand until the ultimate bearing capacity is reached and the roof collapses;
if there is adequate stiffness, equilibrium will occur, and the deflection fails to increase and ponding maintains a constant depth while draining, which protects against progressive deflection and instability.
The orientation of joist members must be considered with different roof geometries and drainage options. The depth of water as it collects around the low areas on its way to primary and overflow drainage will increase the rain load in that area. Therefore, it is best to orient joists parallel to the drainage path so the highest rain loads are dispersed across multiple members at the lowest areas. Joist orientation isnât typically considered when designing roof slope and drainage, but if the engineer takes responsibility for coordinating with architects, it is more likely to be incorporated into the design.
The most important factor in preventing ponding on the roof is providing adequate slope to quickly and efficiently get water to the primary and overflow drainage.
Overflow Scuppers vs. Drains
Scupper roofed over creating a birds nest
When designing for secondary (emergency) drainage, otherwise known overflow drainage, typically the geometry and design of the primary drainage system drives the overflow design.
If cost is a consideration, scuppers are preferred versus overflow drains due to an increase in plumbing costs to discharge the overflow to an end point separate from the primary system. Section 1108 of the International Plumbing Code (IPC) also requires all overflow to be discharged to a location above grade that would be observed by building occupants or maintenance personnel. This is to alert all parties the primary system has backed up or become clogged, and the overflow drain now has to manage the water. With larger areas and bigger drains, it is a common mistake to create a sump for both drains with a hydraulic head that ends up being higher than the overflow collar. This arrangement can cause the overflow to activate during regular rain events without the primary drains being backed up creating staining on the walls. This also defeats the purpose of the visible discharge for the secondary drainage system because itâs now unclear if itâs activated due to the sump design or a primary system backup.
New vs. Existing Buildings
While new buildings provide roof drainage challenges of their own, factoring in proper analysis of rain loads during the design phase will result in roof members designed to carry the anticipated loads. For re-roofing and evaluating existing buildings, this analysis is often overlooked and can quickly become a public safety hazard. Many older buildings were either built without overflow drainage or have inadequate overflow capacity due to code requirements and guidance at the time being unclear.
Rain loads may or may not have been properly accounted for with existing buildings. The guidance for calculating rain loads when sizing roof members has been in ASCE 7 since its beginning in 1988, but it may not have been considered or used properly. An often-overlooked issue with existing building roofs is a false sense of security can result based on how the roof has performed so far. The assumption is â since the roof hasnât collapsed over the past several years of heavy rains, surely the roof structure is adequately designed to hold roof loads. The problem with making this assumption is the roof may not have had issues with any primary drain backups to this point. If the overflow drainage isnât properly designed, all it takes is a âperfect stormâ of primary drains collecting excessive debris, clogging or backing up, or having inadequate overflow drainage.
Overflow scupper has been roofed over
Some older versions of the building code adequately addressed the need for re-roofs to create adequate roof slope and analyze structural stability for rain loads. For example, in 1988, the Uniform Building Code (UBC) required re-roofing to conform to the roof drainage design for new roofs. This includes minimum one-quarter inch  per foot slope and guidance for minimum overflow drains and scuppers that will adequately drain the roof in case of primary drain backup. This same 1988 UBC also includes a requirement for an initial inspection from a building official where the official inspects the roof prior to roofing to determine if any evidence of extensive ponding is apparent. The note for this requirement adds that if extensive ponding of water is apparent, an analysis is made of the roof structure and the appropriate corrective measures are undertaken. These may include relocation of roof drains and scuppers, re-sloping the roof, or structural changes..
Unfortunately, these requirements were relaxed in the most recent building code for existing buildings. In the International Building Code (IBC) and the International Existing Building Code (IEBC) 2015, there is an exception for re-roofs which changes the requirement to âpositive roof drainageâ rather than the one-quarter inch per foot minimum slope. The issue with this change is positive roof drainage is defined as âthe drainage condition in which consideration has been made for all loading deflections of the roof deck, and additional slope has been provided to ensure drainage of roof within 48 hours of precipitation.â There are a few problems with this type of imprecise requirement; even if the roof meets this definition, it could still be prone to ponding instability without adequate slope to the drains and overflow.
The new code requirements have also relaxed the prerequisite for overflow drainage to be added to existing buildings. There is an exception in the IBC 2015 (but not the IEBC 2015) for re-roofing that says recovers or re-roofs on existing buildings arenât required to meet the requirement for overflow drains. This is exactly the recipe for catastrophic disaster as it is assuming roof performance will continue as constructed, with or without overflow design. This can be dangerous as it only takes one primary drain to become blocked with debris, without any overflow, to create roof ponding instability due to excessive rain loads. For example, without the primary drain functioning during a rain event; itâs possible for 12â of water depth to accumulate in a roof area. Twelve inches of standing water translates into 62.4 psf of load. The roof members may have been designed to resist excessive deflection and stress at a 20 psf total live load. This additional load of 3 times more than what the roof members were designed to hold can lead to runaway deflection (deflection causing more ponding depth and rain load) and catastrophic failure.
While new buildings provide roof drainage challenges of their own, factoring in proper analysis of rain loads during the design phase will result in roof members designed to carry the anticipated loads.
Warehouse roof collapse â FEFPA 2007 Goolsby
For new construction, the IBC and IPC 2015 handles roof slope, drain leader and conductor sizing, and overflow requirements adequately, but remains vague on rain load calculations found in Section 1611. The roof slope is generally required to be a minimum of one-quarter inch per foot as previously defined. The drain leader and conductor sizing is handled in IPC Section 1106, where tables can be found that provide the maximum flow that can be handled by various sized pipe and gutters. Overflow requirements are discussed in Section 1503.4 of the IBC 2015, where the secondary drains or scuppers are required to comply with IPC Sections 1106 and 1108. These sections all require the secondary roof drainage be designed to prevent the depth of ponding water from exceeding that for which the roof was designed for in the event primary drains allow buildup for any reason.
Rain loads are handled in the ASCE 7-16 Minimum Design Loads for Buildings, chapter 8 and C8. Rain loads are calculated using a process which begins with the rainfall intensity, defined as i. An important difference from the IPC 2015 is commentary in ASCE 7-16 Chapter C8, where it suggests using the 15-minute duration/100-year instead of the 1-hour duration storm for roof overflow drain design. This is due to the fact that the 15-minute duration, when converted to inches per hour for i, equates to nearly double the intensity (For Galveston, i = 4.6 inches/hour for 1-hour duration, and i  = 8.0 inches/hour for 15-min duration). Anyone familiar with the heavy, short duration storms that frequently occur along the gulf coast in South East Texas can understand why it makes sense to use this higher rainfall rate that can occur within a shorter amount of time.
The next step is to take the tributary area of the overflow drain being analyzed and multiply it by this rainfall intensity with the appropriate empirical factor in Equation C8.3-1, Q = 0.0104Ai. With the flowrate needed to drain this area, the tables in this C8.3 chapter are then used to calculate the hydraulic head for any type of overflow scuppers or drains.
Finally, using the calculated static head, dh (the depth above the primary drain to the overflow) and Equation 8.3-1, R = 5.2(ds+dh) the design rain load, R is established, This design rain load R can be combined with other loads to properly design the structural members.
Providing a systematic study of existing buildings is a great idea to rapidly assess multiple flat roofs and provide an assessment of the ponding instability.
In Europe, studies have been carried out that reveal that ponding instability is an issue there as well. In the Netherlands, about 20 flat roofs collapse every year from heavy rain showers causing ponding on roofs. The government there has taken action, with a supplement to the NEN 6702 building code (NPR 6703) to simplify the calculations and make a ponding assessment on flat roofs.
Consulting engineering firms were engaged, and all municipal buildings open to the public were assessed. The roofs were inspected and the relevant characteristics were reported on, then if necessary calculations were performed to ensure the roof meets the requirements to properly drain rainwater. For flat roofs assessed, 130 of the 231 roofs didnât comply, and had emergency overflow drains placed to prevent ponding.
Providing a systematic study of existing buildings is a great idea to rapidly assess multiple flat roofs and provide an assessment of the ponding instability. Since it is a life safety issue as well as a huge cost when a roof collapses due to ponding instability, this method could be employed on a state-wide level in America beginning with public buildings as well.
Lack of adequate slope (minimum ÂŒâ per foot) causing ponding which
destroys the membrane
The code commentary in IBC 2015 continues to give valuable information on ponding instability in C8.4, which includes references to many other standards and studies for additional direction on rain load analysis.
A great resource is the Structural Joist Institute (SJI) Technical Digest 3, updated in February of 2018. This document summarizes all the different code previsions for ponding and assists with roof design for joist to properly handle the rain loads. SJI has created an in-depth Excel file which can be used to double check calculations for the roof joists as well.
Our hats are off to Dr. Patterson, Dr. Mehta and others who understand and educate our industry on the vital importance of proper design in regard to roof drainage. Though catastrophic roof failures are rare, the associated consequences on your building or a building you may occupy should force this topic to the forefront of any acceptable building design.
Crowe, Joan P. âWater, Water Everywhere: Flood Testing Is Appropriate Only for Waterproofing
Systems.â Professional Roofing, Feb. 2006.
Duncan Stark. âInvestigation, Analysis and Design of an Experiment to Test Ponding Loads on
Flexible Roof Systems.â Oregon State University, 2008.
Edwards, Wanda. âSecondary Drainage and Ponding Requirements in the IBC and IEBC.â RCI
Publications, Nov. 2017.
Fisher, J.W. and Pugh, C.W. (2007). âTechnical Digest 3: Structural Design of Steel Joist Roofs to Resist
Ponding Loads.â Steel Joist Institute, Myrtle Beach, South Carolina.
Graham, Mark S. âConcerns with Roof Drains: Recent Code Changes May Affect Roof Drainage
System Design for Low-Slope Roof Systems.â Professional Roofing, Apr. 2018.
Graham, Mark S. âThe Importance of Proper Roof Slope.â Professional Roofing, Mar. 2005, p. 64.
Goolsby, M and J. Gamoneda (2007). âRoof Drainage and the Florida Building Code,â FEFPA 2007
Winter Conference.
Lawson J. (2012). âRoof Drainage: Not my problem ⊠Maybeâ 2012 SEAOC-SEANM Convention
Proceedings: Santa Fe, NM (2012) p. 136 â 151.
Stark, D., Higgens, C. and Green, Perry S. (2010). âStability of open web steel joists subjected to
ponding loads.â Steel Joist Institute, Myrtle Beach, South Carolina.
Vambersky, J.N.J.A. (2006). âRoof failures due to ponding â A symptom of underestimated
development. Heron.â 51 (2/3).
van Herwijnen, F & H. Snijder, H & J. Fijneman, H. (2006). âStructural design for ponding of
rainwater on roof structures.â 51.
Verhulst, Stewart M., Deleon, Marco A., East, Bradley L. âThe Roof Drainage Epidemic.â Forensic
Engineering 2009, 2009, doi:10.1061/41082(362)21.
Wijte, S.N.M. (2006). Ponding collapse analyses of light weight roof structures by water raising
capacity. Heron, 51. 97-113.
Wilen, Jason P. âStill Water Runs Deep: Proper Roof Slope and Drainage Are Important to Prevent
Excessive Water Accumulation.â Professional Roofing, July 2012, pp. 34â38.
