by JACK POOLE, P.E., POOLE AND ASSOCIATES ROLAND HUGGINS, P.E., AFSA
There has been a long-standing conflict regarding the use of backflow preventers on sprinkler systems due to many questioning if they are really needed but mainly due to the fire protection problems they can create. A study by the AWWA Research Foundation (AWWARF) called "Impact of Wet-Pipe Fire Sprinkler Systems on Drinking Water Quality" (but referred to by multiple names such as the Boyle, Poole, or AWWARF study) clearly indicated that backflow prevention is required. What is particularly noteworthy from this report, though, is that on Class I and II sprinkler systems, an alarm check valve should be capable of providing adequate backflow protection for new installations provided it contains specific features. These features were not available at the time the report was written. Such a device is now available.
Let's start by providing some definitions and briefly discussing the objectives for both sides of the issues and the fire protection problems with these devices. This discussion is applicable only to Class I and II wet-pipe sprinkler systems. AWWA defines a Class I system as having "Direct connections from public mains only; no pumps, tanks, or reservoirs; no physical connections from other water supplies; no antifreeze or other additives of any kind; all sprinkler drains discharge to atmosphere, dry wells, or other safe outlets." A Class II system is the same as a Class I but with a fire pump installed in the connection from the street main. This encompasses a majority of the systems.
The objective of the water community is to provide clean, safe water as required under the Safe Drinking Water Act. In order to accomplish this, they not only have to treat the water they supply but also ensure that any connection to their distribution system does not contaminate the supply. Such contamination is not restricted to just items that cause immediate, severe health problems but also long-term health issues, such as exposure to heavy metals. Water quality, such as color (or the lack thereof), taste, and odor, are also controlled. If a system connected to the water supply contains any characteristic that violates the specified allowed limits, some type of backflow prevention assembly is required. Whether these assemblies interfere with forward flow or other functions of the connected system is secondary compared to stopping the potential for backflow. This is where the conflicts start. The objective of the sprinkler community is to maximize the reliability of water supply in the direction of flow. Additionally, minimizing the potential for water damage is extremely important, thus the need for water flow alarms. It's widely known that the current backflow preventers have failed regarding forward flow and can interfere with the ability of the water flow alarm to activate. So what's the answer?
One answer is to simply live with things as they are. This is not acceptable, since all parties are interested in the overall well-being of society, not to mention avoiding unnecessary liability. There is another option that can meet the performance objectives of both communities. The AWWARF study sheds much more light on this issue.
The need for monitoring
Failure rates in regard to backflow, referenced by the AWWARF study, showed an 8.7 percent failure rate for double check valve assemblies with a 2 percent failure rate for both check valves. If such a failure occurs right after an inspection, backflow can occur for a year or more before it is caught. Since there is no visual indication of a problem, in order to provide a truly adequate level of protection against such high failure rates, there needs to be a means to automatically monitor when both checks are fouled. For comparative purposes, the failure rate of sprinkler systems was identified as 4 percent, but these were standard check valves that had apparently not been maintained. There were no identified failures of the alarm check valve itself.
Additional applicable information is that a standard alarm check valve on an existing system is considered to provide adequate backflow protection. As currently designed, though, these devices are not deemed adequate for a new installation. This is not a performance issue but stems from the base requirement that only double check valve assemblies can be approved.
Need for a new device
The AWWARF study made some extremely pertinent statements on this issue. These are:
1. "Alarm check valves are reliable devices" but more study is needed to determine the number of failures (to prevent backflow). As already pointed out, the identified failures have all been standard check valves with no maintenance.
2. There exists a need for a new cross-connection assembly for Class I and II sprinkler systems. It was described as "the proposed features of this new and unique backflow alarm assembly are based on the inventive ideas developed by the authors. These devices should have similar operating features to that of an alarm check valve and be equipped with an alarm feature that would indicate when the main clapper is in the open position, but the device should be provided with test cocks for testing purposes, and the check valve in the bypass line should be replaced with a rubber-seated check valve. The tolerances of the main check should be similar to the tolerances of a check valve in a backflow preventer."
3. "A standardized cross-connection control testing protocol for wet-pipe fire sprinkler alarm check valves does not currently exist but is essential for monitoring the performance of alarm check valves and hence should be developed."
Such a device now exists. It's a standard alarm check valve monitored by the AFCON Model 1100 component. The combined assembly does not affect the operational features of the alarm check valve. Rubber-seated check valves in the by-pass line are no longer an issue, since their use is now standard practice. The tolerances of the main check of an alarm valve is not an issue, since, if anything, [it is] designed and manufactured to a higher standard. This assembly monitors the integrity of the alarm valve seal and not just whether the clapper is on the seat. This assures us that, if water is leaking past a damaged gasket, it, too, is detected.
The only real hurdle is the false premise that two check valves are better than one. It's obvious that the performance of two devices of the same mechanical design is superior to that of one, especially if evaluating only one operating characteristic. This is shown by the much smaller failure rate of both checks versus that of one check. The fallacy is in equating the design of these checks to that used in the alarm check valve.
Check valves are designed to lift when there is a pressure difference and close when there is none. The backflow preventers use either a poppet-type, where a valve stem travels up and down through a hole, or a swing-type, where components glide up and down a track. The alarm check valve is simply hinged at the side and freely swings. The design of the backflow preventers reflects the expectation that it's used in a dynamic system where there is a regular flow of water. Sprinkler systems, though, sit dormant for long periods of time with no water flowing, and the alarm check valve was designed to operate in such static conditions. These mechanical designs are as different as night and day, and the presumption that two devices with more moving parts and greater susceptibility to failure are superior to a different single device is incorrect. This is especially | true when requiring the devices to satisfy | all safety functions, such as full forward flow, and not just backflow.
In conclusion, the AWWARF study has identified that both the water and sprinkler communities are tasked with safeguarding society. When a single device is attempting to serve both communities, all objectives must be met. It has clearly identified an option that would improve the overall ability for a single device in fulfilling this dual capacity. We now have such a device for new installations, and it's up to both communities to study it and modify their standards and practices as needed if they are truly interested in properly serving the safety needs of society.
?2004. Reprinted with permission from Sprinkler Age, the official publication of the American Fire Sprinkler Association.