Jack Poole, P.E.
Poole Fire Protection Engineering, Inc.
1317 S. Fountain Drive
There has been much debate on the degree of backflow protection on Class I and Class II wet-pipe fire sprinkler systems. Much of the controversy centers on whether the water quality in a wet-pipe fire sprinklers system would pose a health hazard if a backflow incident were to occur. The potential requirement for backflow prevention has caused concern in the fire protection industry because of possible hydraulic problems associated with retrofitting existing Class I and Class II wet-pipe fire sprinkler systems.
To address this issue, the American Water Works Association Research Foundation funded this study to evaluate the water quality in Class I and Class II wet-pipe fire sprinkler systems, determine if a public health hazard exists, and identify methods to effectively safeguard the public in such a case. This paper presents a summary of the results of an 18-month study that was performed with the participation of 26 American and 4 Canadian water purveyors. Information and findings related to backflow simulation studies, conceptual risk evaluations, and retrofitting costs impacts are not presented in this paper, but are available in the final report of this project.
TYPES OF SYSTEMS EVALUATED
This project specifically determined the quality of water within Class I and Class II wet-pipe fire sprinkler systems constructed of black steel (black iron) pipe materials. Class I and Class II is defined in accordance with the American Water Works Association (AWWA), M-14 Manual, Recommended Practice for Backflow Prevention and Cross- Connection Control. The definitions as stated in the M-14 Manual are as follows:
Class 1. Direct connections from public mains only; no pumps, tanks, or reservoirs; no physical connection from other water supplies; no antifreeze or other additives of any kind; all sprinkler drains discharge to atmosphere, dry wells, or other safe outlets.
Class 2. Same as Class 1, except that booster pumps may be installed in the connections from the street mains (booster pumps do not affect the potability of the system). It is necessary; however, to avoid drawing so much water that pressure in the water mains is reduced below 10 psi. It should be noted that the typical industry practice is to not reduce the pressure in the mains to less than 20 psi, which is also referenced in the AWWA M31 Manual.
In order to assess the magnitude of the problem, the research program included the voluntary participation of 30 water purveyors geographically dispersed throughout North America. Participating utilities are listed in Table 1 and their geographical distribution is shown in Figure 1. The participating utilities include large and medium water systems with a range of climatic conditions, as well as urban and rural situations and provide a basis for comparison of information and data.
The information-gathering phase included a review of the identified literature and compilation of utility and fire department internal information. Utility information was collected primarily by the means of a four part, check-the-box, mail-in survey of to determine the extent and type of problems and concerns associated with wet-pipe fire sprinkler systems and backflow into distribution systems, and by on-site interviews at fifteen utilities.
Field-testing to determine the water quality within Class I and Class II fire sprinkler systems included evaluation of metals, general water quality, and microbiological parameters. In order to standardize the sample collection of wet-pipe fire sprinkler systems at the participating utilities, a sampling protocol was developed and supplemented with an instructional video to demonstrate the water sample collection and analyses methods. The field sampling and laboratory analysis phase took approximately nine months to perform and was completed in November 1995.
WATER QUALITY PARAMETERS
The major contaminant groups to-be evaluated as the most appropriate water quality parameters to be tested in the laboratory included metals, general water quality parameters, and biological parameters. The development of the field testing protocol assumed that the cost for analytical work for one fire-sprinkler testing location would approximate $2,500, and included the following parameters:
1. Significant Transition and Earth Metals: Lead (Pb), Copper (Cu), Zinc (Zn), Iron (Fe), Manganese (Mn), Cadmium (Cd), Chromium (Cr), Sodium (Na), Calcium (Ca).
2. General Water Quality Parameters: pH, Temperature, Sulfate, Total Dissolved Solids (TDS), Total Suspended Solids (TSS), Conductivity, Alkalinity, Turbidity, Oil and Grease (O&G)
3. Microbiological Parameters: Heterotrophic Plate Count, Total Coliform.
WET-PIPE FIRE SPRINKLER SAMPLING LOCATIONS
Based on information and experiences in developing the video and written protocol, the written protocol was enhanced to identify rapid, uniform and consistent sampling approaches. Since the primary goal of the project relied on sampling water to profile quality within the wet-pipe fire sprinkler system, it was reasoned that to avoid moving field collection and sampling equipment, use of the main drain connection on the fire sprinkler riser would allow access to sampling the internal water quality of the fire sprinkler project, simply by flowing the water from the fire sprinkler for specific time intervals prior to withdrawing samples.
The sample locations were identified as follows:
WATER QUALITY RESULTS
A total of 81 wet-pipe sprinkler systems were sampled using an established written and video protocol. The number of systems sampled ranged from 13 by the City of Winnipeg to 1 by the Pinellas Utilities. Sampling was conducted between March and November, 1995.
Average lead and copper concentrations for all data collected are presented in Figure 2 as a function of each sample location of the fire sprinkler system. The average lead concentration levels ranged from 0.52 mg/L on the fire sprinkler side of riser control valve, to 0.007 mg/L at the potable water main. The average copper concentration levels ranged from 0.287 mg/L at the fire sprinkler side of the backflow prevention assembly to 0.016 mg/L after the system had drained for 120 seconds.
Average concentration of zinc and manganese with respect to their sample location is shown in Figure 2. Zinc was detected at average concentrations ranging from 3.033 mg/L to 0.63 mg/L. The lowest zinc values encountered were taken from the potable water main. Manganese is an important contributor to alloys and therefore can be leached into the water system from deterioration of pipes or valves. The average manganese levels ranged from 0.386 mg/L to 0.080 mg/L.
Average concentrations of cadmium and chromium are also presented (see Figure 3). Cadmium was discovered to range in average concentrations of 6 ug/L at 4-US to 0.23 ug/L at 1-PU. Chromium was detected at average concentrations of 5.7 ug/L at 4-US to 2.4 ug/L at 1-PU. Average iron concentration levels ranged from 0.869 mg/L at the potable water main to 25 mg/L after the system was flushed for 60 seconds. The sodium concentration levels did not vary appreciably. Average sodium concentration levels ranged from 33.3 to 28.2 mg/L. Average calcium concentration levels ranged from 40.6 mg/L at the utility side of the backflow preventer to 19.2 mg/L after the system was drained for 120 seconds. Distribution of the average concentrations of iron, sodium and calcium with respect to their sample location can be found in Figure 4. Figure 5 presents a summary of average total organic carbon (TOC) concentration as a function of sample location. Significant amounts of TOC are present in the samples taken on the fire sprinkler side of the alarm check valve, and is believed to be primarily a result of the cutting oils used in the construction of the fire sprinkler system, and/or the manufacturing process of black steel pipe.
WATER QUALITY CONCLUSIONS
Based on an evaluation of the data collected during the testing program, lead was measured within Class I and Class II wet-pipe fire sprinklers at certain fire sprinkler locations that were at levels exceeding the Lead and Copper Rule (LCR) action level_ of 0.015 mg/L. Other water quality findings included:
1. Water quality within existing wet-pipe fire sprinkler systems exceed the primary standard for lead and cadmium, and secondary standards for iron, manganese, total dissolved solids, sulfate, color, and taste and odor. Soluble lead appeared to be originating from check valves that had lead-weighted clappers, leaded fittings, machined leaded brass valve bodies, and other accessories.
2. Coliform (an indication of pathogens or disease causing organisms) were predominantly absent in wet-pipe fire sprinkler systems; however, heterotrophic plate counts exceed recommended guidelines for potable Consumption. Of the 84 wet-pipe fire sprinkler systems evaluated, coliform was found to be present in 4 of the sprinkler systems, and was attributed to recent unsanitary construction activities recently performed on the sprinkler systems.
3. Metal and total organic carbon concentrations were found to be highest in proximity of the fire sprinkler check valve on the sprinkler system side of the valve, and tended to decrease with horizontal pipe distance within the fire sprinkler pipe.
4. Wet-pipe fire sprinkler systems contain water that can be aerobic, anoxic, and anaerobic. Dissolved oxygen concentration decreases with horizontal pipe distance, with the highest concentration in proximity of the sprinkler valve, and the lowest concentrations in proximity of the remote portion of the sprinkler system.
Our sincere thanks and appreciation go to the field crews, foremen and managers of the participating utilities, local fire department personnel, and the customers and business owners who allowed the authors to sample water quality in their Class I and II wet-pipe fire sprinkler systems. In addition, the advice and help provided by the Project Advisory Committee (PAC) and AWWARF project officers, Roy Marcinez and Ann Scarritt, were sincerely appreciated. The PAC consisted of Phillipe Boissonneault, Quality Assurance Supervisor, Portland Water District, Portland, Maine; Lou Allyn Byus, Assistant Manager of Field Operations, State of Illinois Environmental Protection Agency; Ken Isman, Technical Service Manager, National Fire Sprinkler Association, Inc., Patterson, New York; and Donald F. Newnham, PE, Direct of Public Works, City of Altamonte Springs, Altamonte Springs, Florida.
Alleman. I.E.: Milke, J.A.; and Hickey, H.E. 1981. An Investigation of the Water Quality and Condition of Pipe in Existing Automatic Sprinkler Systems for the Analysis of Design Options with Residential Sprinkler Systems. University of Maryland: Department of Fire Protection Engineering and Civil Engineering. NBS-GCR-82-399, August.
American Water Works Association. 1987. Distribution System Requirements for Fire Protection. (AWWA Manual M31). American Water Works Association. Denver, CO.
American Water Works Association. 1990. "Recommended Practice for Backflow Prevention and Cross-Connection Control." Manual of Water Supply Practices M-14. Denver: AWWA.
Black, J.T. 1992 "Backflow Prevention vs. Fire Suppression Systems, A Health Perspective." Proc. Distribution System Symposium . Atlanta, GA: American Water Works Association. September 8-11, 181-194.
Canadian Automatic Sprinkler Association. 1986. Report on the Health Risk of Public Consumption of Fire Protection, Sprinkler System Water. Willowdale, Ontario: SENES Consultants Limited. 499 McNicoll Avenue, Willowdale, Ontario M2H 2C9.
FCCCHR. 1988. Manual of Cross-connection Control. Los Angeles: University of Southern California Foundation for Cross-connection Control and Hydraulic Research, Eighth Edition.
Hart, F.L. et al. 1993. Backflow Protection for Residential Sprinkler Systems. United States Fire Administration and Federal Emergency Management Association. Cerritos, CA: Southern California Fire Sprinkler Advisory Board.
John, S.E; Kane, D.N.; and Skelton, L.W. 1990. Economic Impact Study of Proposed Cross-connection Regulations, R87-37. Springfield, IL: Environmental Planning & Economics, Inc., 1005 North Seventh Street, Springfield, IL, 62702, prepared for Illinois Department of Energy and Natural Resources, 325 West Adams, Springfield, IL 67202.
NFPA. 1992. Standard for the Installation of Private Fire Service Mains and Their Appurtenances. National Fire Protection Association - Standard No. 24 (NFPA 24), 1992 Edition
NFPA. 1992. Standard for the Inspection., Testing, and Maintenance of Water-Based Fire Protection Systems. Quincy, MA: National Fire Protection Association - Standard No. 25 (NFPA 25), 1992 Edition.
National Fire Sprinkler Association. 1987. Backflow Prevention for Fire Sprinkler Systems. Patterson, NY: National Fire Sprinkler Association.
Underwriter Laboratories. 1993. UL Fire Protection Equipment Directory. Northbrook, IL: Underwriter Laboratories, Inc. West, S.L. 1991. "Water Quality Sampling of Fire Protection Systems." Drinking Water & Backflow Prevention, August, 16-18.