By Dave Peterson
I hadn’t planned on breaking away from my coverage of the National CAD Standard anytime soon, but recent events – including the arrest of a pool builder on charges of manslaughter in a suction-entrapment incident – compelled me to do otherwise.
As I started composing this column, my plan was to call it “Entrapment Rundown” and make it a straightforward, positive summary of recent changes in codes and systems related to suction entrapment. As I dug more deeply into the topic, however, I found the issues and solutions to be much more confusing than I’d anticipated – so much so that describing the situation as a “meltdown” seemed far more accurate.
Much of my digging came as a result of being called in by a couple of well-known hotel chains to perform safety audits on their facilities and help them figure out what compliance with the new Virginia Graeme Baker Pool & Spa Safety Act really meant. Simultaneously, I began receiving calls from builders and even inspection officials wondering the same thing.
All I can say at this point is that nobody has all the answers – and that, every day, more and more misinformation seems to end up swirling its way around a grand sump of conjecture.
Based on the phone calls we get, I’m guessing that few people have actually purchased (let alone read) the relevant documents from the American National Standards Institure (ANSI) or downloaded copies of any relevant local codes or regulations. To improve that situation, my ambition here is to review information from all currently available sources and do what I can to clarify the situation.
Put on your seatbelts: It’s going to be a bumpy ride.
FINDING THE WAY
So far, I’ve found five sources of information that must be consulted to make sense of these issues:
[ ] ASME/ANSI A112.19.8-2007: On February 5, 2007, ANSI and the American Society of Mechanical Engineers approved a document officially titled, “Suction Fittings for Use in Swimming Pools, Wading Pools, Spas and Hot Tubs.” Most of this text describes specific requirements for suction outlet covers/grates, covering materials, testing procedures and markings.
According to the standard, all suction-outlet covers/grates are required to be labeled with information on the maximum flow rating and product life span as well as guidance on approved mounting location(s) and a statement on the fitting’s suitability for use on single- or multiple-drain systems. (There are other requirements, but these are the important ones.)
Another key element of ASME/ANSI A112.19.8 is its establishment of dimensional requirements for field-built sumps. In my operation, we’ve always favored using injection-molded sumps because they guarantee compliance with known requirements, but we’re also aware that many builders leave sump design to a shotcrete crew that does little more than dish out a small divot around a pipe. By the time that pipe is sealed with hydraulic cement and the plasterers have lined the sump, there is literally no clearance below the cover/grate.
Building it right without an injection-molded fitting, of course, requires the steel to be adjusted and the sump formed. If you take this path, what you need to know is that the horizontal dimensions need to match or exceed the dimensions of the suction outlet cover/grate and that the vertical dimension measured from the top of the pipe to the lowest point of the suction outlet cover/grate should be one-and-a-half times the pipe’s inside diameter.
To illustrate: If you use four-inch plumbing, the sump must be at least 3.99 inches times 1.5 – that is, six inches deep if the pipe enters vertically through the floor. If the pipe enters horizontally through the sidewall of the sump, at least 4.5 additional inches will needed and the total depth of the sump will need to be 10.5 inches.
One last point: ASME/ANSI A112.19.8 addresses only four of the five types of entrapment, mentioning body, limb, hair and mechanical entrapment but not evisceration. This is primarily because evisceration is easily avoided as an issue by installing virtually any drain cover and keeping the flow rates at reasonable levels.
[ ] ANSI/APSP-7 2006: ANSI and the Association of Pool & Spa Professionals approved this standard, called “Suction Entrapment Avoidance in Swimming Pools, Wading Pools, Spas, Hot Tubs and Catch Basins,” on September 11, 2006, several months before ASME/ANSI A112.19.8 was approved. The two standards were in simultaneous development, and the committees shared many members. (I myself was briefly involved in developing ANSI/APSP-7 at a time when I managed the engineering department for a manufacturer).
So what does ANSI/APSP-7 add to the subject? For new construction, the standard reiterates what every watershaper should already know: Split all suction outlets and keep the plumbing velocities low. It also offers guidance for retrofitting existing pools and spas that have single-outlet configurations, but some suggestions (discussed just below) are better than others and each pool and spa will likely require a different solution.
The first option is to retrofit the pool or spa with split suction outlets either separated by 3 feet or placed on different planes (that is, for example, on a spa floor and on a vertical wall in the bench) so that a model torso covering 18 by 23 inches cannot cover both suction outlets at once. Here, the plumbing between the suction outlets must be sized so that if one outlet is covered, the remaining outlet will handle the full flow rate without generating a high vacuum force on the covered suction outlet. In other words, the plumbing cannot be downsized between the suction outlets.
Another retrofit option involves eliminating the single suction outlet altogether. Last year, for example, we completed a major remodel on a 50-by-20-foot pool where we eliminated the single suction outlet on the floor of the eight-foot deep end. We aligned the return inlets to promote complete mixing and had no concern about the pool’s hydraulic performance. (For more on this approach, click here to read Ray Cronise’s article, "Considering an Option.")
A variation on this outlet-elimination scheme involves converting a single suction outlet into a return inlet. This can be easy if the outlet is plumbed to the equipment area and valved to the skimmer line before the pump, because all the work can be done at the equipment area. Here again, completely eliminating the suction outlet hazard works better than simply reducing the hazard with some other solution.
Another (more complicated) option involves adding a vented reservoir between an existing single suction outlet and the pump. This gravity flow system ensures that the pump sucks only on the vented reservoir and not directly from the outlets in the pool. As the pump draws water from the reservoir, the water level drops relative to the pool and creates a pressure differential that causes water to flow from the suction outlet(s) to the reservoir.
Engineered vent systems are also a possibility, but they require implementation by a licensed engineer and are not as safe as other methods. (I’m a licensed engineer and would give a job a pass if I were asked to design one of these systems. My reasoning: I’m not sure how I could defend anyone in a potential lawsuit once the jury found out that complete elimination of the hazard had been a possible retrofit option.)
The last possibility offered in ANSI/APSP-7 involves equipping an existing single outlet system with a safety vacuum-release system (SVRS) – another choice I would reject mainly because, where the five retrofitting options listed above are plumbing solutions that involve no moving parts, current SVRS systems require electronic and/or mechanical systems to function. It’s a plain fact that all electronic and mechanical devices eventually fail – that is, they all have mean times between failures (MTBF) or mean times to failure (MTTF) and eventually will break down.
Temperature variations play a role in such breakdowns, while stainless steel rusts in saline water, springs fail, and ultraviolet light and heat damage plastic. I’m certain suppliers of these systems have tested them and selected components for reliability, but such systems have defined life spans and experience tells me that if something can go wrong with electronic controllers and mechanical devices used with watershapes, it inevitably will go wrong.
It’s also true that SVRS systems are only a partial entrapment solution: Hair and mechanical entrapment are still issues with them. Bottom line: We have better options.
[ ] Virginia Graeme Baker Pool & Spa Safety Act: This legislation is federal, nationwide law, not a standard. Signed on December 19, 2007, it’s most significant requirement is that by December 19, 2008, all public pools must be equipped with ASME/ANSI A112.19.8-listed covers/grates and any with single suction outlets must be retrofitted using one of six options, some of which are proposed in ANSI/APSP-7.
The Baker Act, interestingly enough, doesn’t refer directly to the ANSI/APSP-7 standard at all. Instead, it lists its six solutions for single blockable suction outlets as SVRS systems, vent systems, gravity systems, automatic pump shut-off systems, drain disablement and “any other system determined by the [Consumer Product Safety] Commission to be equally effective as, or better than, the systems described.”
The last option – that is, “any other system” – allows new devices and methods to be developed, but what’s missing here are three obvious ANSI/APSP-7 solutions: splitting suction outlets, disabling single suction outlets or replumbing the suction outlet to make it a return line.
Another failure of the Baker Act is that there is no requirement for retrofitting private pools. All it says is that all covers/grates manufactured, distributed and sold after December 19, 2008, must comply with ASME/ANSI A112.19.8 and does not require homeowners to upgrade their pools. It does, however, require that all pools built after December 19, 2008, must have split suction outlets or single or multiple unblockable outlets or covers – or no suction outlets at all.
[ ] Consumer Product Safety Commission Guidelines: CPSC drove the Baker Act and were involved in development of both ANSI/APSP-7 as well as revisions to ASME/ANSI A112.19.8. In fact, CPSC made its positions clear as far back as January 1998 with publication of “Guidelines for Entrapment Hazards: Making Pools and Spas Safer” – a document it updated in March 2005.
Many of CPSC’s guidelines are carried in ANSI/APSP-7, but there are additional suggestions about inspection requirements that didn’t make it into the standard, and there’s also a lot of background information and statistics on entrapment-related deaths and injuries.
[ ] International Code Council: ICC publishes both the “International Residential Code” and the “International Building Code” – and this is where I found the biggest mess of all with respect to entrapment-related recommendation.
Simply put, SVRS manufacturers managed to get their devices written into these codes as requirements in a context that creates a false sense of security about the devices. In fact, the codes start by requiring split suction outlets and then specify use of SVRS systems that will activate only if both suction outlets are completely blocked.
This is a clear case of the code writers not understanding the technology or perceiving the consensus of the standards that listed grates/covers and split suction outlets to get the job done. All ICC has done is mandate a third device – a backup that provides scant benefit if the first two safety measures are in place.
More specifically, the “International Residential Code” includes its SVRS requirement as an appendix that won’t necessarily be adopted by all states, but the “International Building Code” jumps right into the thick of things in section 3109.5.2 by requiring “Atmospheric vacuum relief systems” in the event grates/covers break or go missing and, in section 3109.5.3, by requiring that dual drain systems must be piped to SVRS systems.
So ICC requires builders to add SVRS devices onto pools with split suction outlets where they would only react if both outlets were completely blocked and where the SVRS devices do nothing to solve the problem of hair and mechanical entrapment. Even ANSI A112.19.17-2002, “Standard for Manufactured SVRS Systems,” acknowledges this limitation and requires manufacturers to state the following: “Warning: Due to the lack of physiological data, it cannot be concluded that a device of this type referred to hereafter as a Safety Vacuum Release System (SVRS) will eliminate the potential for disembowelment.”
At this writing, APSP is sparring with ICC over the SVRS issue, but it’s clear that ICC doesn’t understand the engineering issues and is too caught up in recent headlines about suction-entrapment accidents to make negotiations as direct as they should be.
To make as much sense as I can out of this alphabet soup of standards, codes and recommendations, allow me to offer some guidelines related to new construction, retrofits and more.
Before I begin, however, let me provide the following disclaimer: Please note that I do not represent any of the standards-promulgating bodies listed above and am not a member of any of their committees or supervising organizations or of any code writing or compliance agency. For specific clarifications on the documents, consult the authoring organization(s): These are my own opinions, and it is your responsibility as designers and builders to ensure safety at every level
This column does not address every situation, nor even all the opinions of this writer. In sum, this subject matter is exceedingly complicated, and I recommend seeking very specific, case-driven professional advice when a solution is not obvious or been proved to be effective.
With all that in mind:
[ ] New Construction: First, consider whether you even need suction outlets. For a basic pool, the skimmer(s) may be enough – and elimination of the hazard is the safest solution. If suction outlets are required, however, always split the outlets even if you are using unblockable covers/grates. This will significantly reduce or eliminate entrapment hazards if the cover/grate is broken or missing.
In separating suction outlets, set them up with three feet of distance between the cover/grate fittings – or locate them on different planes, as suggested previously. Some codes require a spacing of three feet, center to center, as measured at the pipes, but others don’t mention the pipes. Although this standard is not law, you can bet that plaintiff’s attorneys will present these standards as exhibits.
I would also suggestion using ASME/ANSI A112.19.8-listed covers/grates and making sure that the listed flow rates are not exceeded. Also, verify that the fitting is approved for its position (on a wall, in the floor or suitable for both). In addition, you should limit the velocity in the line between the two sumps to four feet per second, as determined by considering the maximum flow rate.
For example, a spa jet pump running eight jets at 15 gallons per minute each will require 120 gpm total. This will require a four- inch diameter PVC line with a potential velocity of 3.1 feet per second (if one drain is blocked). In normal operation when neither drain is blocked, you would expect about half of the total flow to pass through each suction outlet – a normal velocity of one-and-a-half feet per second.
On a personal note, I proposed a two-foot-per-second limit during an ANSI/APSP-7 committee meeting in 2004, and the room burst into laughter and groans as if that was something that simply couldn’t be done. That’s true if you use two-inch pipes and two horsepower pumps, but with half-horsepower pump on an unfiltered waterfeature, it’s possible to develop a flow rate of more than 80 gpm if you use three-inch diameter suction pipes (as I would recommend) or two-and-a-half-inch pipes using ANSI/APSP-7 criteria.
Another step I recommend is making a note with a permanent marker on the time clock or controller of a planned replacement date for the covers, per ASME/ANSI A112.19.8. As is the case with SVRS devices, covers/grates are beaten up by sunlight, chemicals, being stepped on and other factors that result in a finite life span. They should be replaced before they fail, not after an injury or death has occurred!
[ ] Retrofit Construction: Always split the outlets as in new construction (see above) or eliminate the hazard completely by abandoning the suction outlet and relying on the skimmer alone for circulation needs – or by reversing the flow so the outlet becomes an inlet. In addition, validate the size of all pumps: Chances are good that they were oversized by the original builder and that you can use this as an opportunity to guide your client to energy-efficient variable-speed pumps.
[ ] Unblockable Suction Outlets: Both ASME/ANSI A112.19.8 and ANSI/APSP-7 use a simulated torso 18 inches wide by 23 inches long to determine if a suction outlet is blockable – a sizing meant to encompass the fact that 99 percent of men, women and children have torsos smaller than that and would not block the drain. (I guess we all should pity the one percent of outsized men.)
|This spa has split suction outlets on the floor for the filter pump and split suction outlets in the bench walls for the jet pump.|
The important thing to note here is that standard grates/covers measuring 18 inches by 18 inches are blockable and should not be used singly. I’ve heard builders argue that if you measure these grates diagonally, they measure 25.5 inches and the model torso will not be entrapped. But that’s bogus reasoning, since it fails to consider those who might become entrapped while aligned with the grate. In this context, only 24-inch-square grates do the job of providing the minimum unblockable suction outlet dimensions.
(An alternative would be a channel drain more than 29 inches long designed so that, when challenged by the model torso, the remaining open channel drain is still capable of handling the maximum designed flow rate.)
One retrofit solution that is not outlined in any of the codes is changing the single suction outlet to an unblockable single suction outlet. For example, if an old pool is unnecessarily deep, it’s easy enough to dowel in some reinforced concrete, build a proper 24-square-inch sump and install an approved cover/grate with the new pool finish. In this way, you fix the drain by filling in an unused part of the pool and never even touch the old plumbing system.
[ ] Skimmer Equalizer Lines: When I’m asked about what to do with skimmer equalizer lines, I always suggest splitting them. If a float valve is installed, it has the capacity to shut off the skimmer when the water level drops below the weir; in doing so, it eliminates the skimmer as a vent and the pump is then pulling on the submerged suction outlet as though is was directly plumbed to it.
|In this project, the skimmer above the sun shelf has a pair of equalizer lines located on the wall between the sun shelf and the spa, spaced 36 inches apart.|
Look at it this way: When the skimmer no longer serves as a suction outlet, the equalizer line effectively becomes a suction outlet and should be split. (Some builders don’t install the float valves, but you can’t ensure that an owner or service technician won’t drop one in the skimmer to adjust the skimmer/suction outlet flow ratio.)
[ ] Combined Suction Outlets Under Shared Covers/Grates: We always separate our spa jets from the filtration system, which results in two pairs of split suction outlets. Some builders gang these pipes together so that the finished spa only has two covers/grates, but I don’t like that detail for two reasons: First, the filtration outlets should be on the floor while the jet pump outlets should not; second, this arrangement can be dangerous if the math hasn’t been done correctly.
An eight-jet system running on a two horsepower pump may flow at 125 gallons per minute. Combine that with a filter pump moving 65 gpm at 40 feet of total dynamic head (with a half-horsepower pump!) and you have a combined total of 190 gpm, which exceeds the recommended rates for many of the ASME/ANSI A112.19.8-listed covers/grates.
Under normal operation, the flow rate is split with only 95 gpm flowing through each cover/grate, but the design needs to be redundant and capable of handling the full flow through just one side. This is why we always split the filter pump’s suction outlets and keep them on the floor where the heavy debris settles – then we split the jet pump outlets on the bench walls, putting them below the seat where they are less likely to pull in sand or other settling debris. With this configuration, our velocities through the covers/grates are lower and the whole system is improved in terms of both safety and energy efficiency.
The good news in all of this is that many pool designers and builders are aware of the issues underlying hydraulic and safety practices and have little to worry about. The bad news is that other builders who may have been lucky to date are walking on thin ice these days, as it seems serious legal jeopardy awaits some of those who don’t follow the rules.
It is simply unwise in this climate to run the risk of having a cover/grate get broken or come off a single-port suction outlet – the greatest risks in suction-entrapment incidents. All it takes is simple action: The last time I checked, there had yet to be an entrapment death resulting from contact with a properly secured, correctly flow rated, ASME/ANSI A112.19.8-listed cover/grate on a split-suction pair of outlets.
To be sure, there’s some confusion out there. The key at this point is for the industry to get organized and be certain that health departments and code writers know the full story and can be persuaded to accept the fact that, in many cases, suction outlets aren’t even needed.
Dave Peterson is president of Watershape Consulting of San Diego, Calif. He’s been part of the watershaping industry since 1994, starting his own firm in 2004 after stints with an aquatic-engineering firm and a manufacturer. A registered civil engineer, he now supports other watershape professionals worldwide with design, engineering and construction-management services and may be reached via his web site, www.watershapeconsulting.com.