Development of recreational water spray ground design regulations in New York State, an engineering approach

f

Post filter fraction of turbidity

Q

Water flow rate, gpm

Filter flow rate for bypass treatment system, gpm

Feature flow rate for bypass treatment system, gpm

S

Ratio of feature to filter flow rates

T

Turbidity, NTU

Turbidity added value, NTU

Mean residence time, V/Q

V

Volume of water, gallons

Spray grounds have gained popularity because they provide significant entertainment value and present a limited, or no, drowning hazard. In 2004, approximately 30 (and increasing) spray grounds existed in New York State at various lifecycle stages. New York State Department of Health (NYSDOH) recognized potential problems as existing pool and spa regulations failed to capture the unique treatment requirements of spray grounds. The literature at the time substantiated the concerns based on reported recreational water illness data and epidemiology.

Published statistics demonstrated an increasing frequency of recreational water-associated outbreaks between the years 1978 and 2002 (Blackburn et al. 2004). Of the 65 documented outbreaks between 2001 and 2002 (the highest frequency to date), 44 (68%) involved treated-water venues (e.g., pool, spas, spray grounds). The data demonstrated cryptosporidiosis outbreaks were becoming more frequent in recreational treated-water venues. Although cryptosporidiosis was associated with a modest fraction of all recreational water outbreaks, the number of individuals affected was disproportionately large. For example, during the period 2001–2002, only 14% of all outbreaks associated with treated-water venues involved cryptosporidiosis, however, at least 58% of all individuals affected by recreational water outbreaks were affected by Cryptosporidium. A single outbreak involving a spray ground in New York State in 2005 resulted in 2,307 documented cases of illness attributed to cryptosporidiosis (Yoder et al. 2008). The literature also documented several outbreaks caused by chlorine-sensitive organisms such as Escherichia coli and Shigella species, which were the result of poor facility maintenance and could have been prevented by greater diligence in operation and maintenance procedures (Blackburn et al. 2004).

To ensure public health would be protected, NYSDOH initiated a process to develop regulations for the design and operation of spray grounds with the final regulation taking effect on March 28, 2007. To address the unique characteristics of spray grounds, a proactive approach based on modeling and process engineering methods was introduced into the regulation development cycle. The focus of this work is to provide the public health community with the rationale underlying the development of the water treatment portion of the spray ground regulation, Subpart 6-3 of the New York State Sanitary Code (NYSSC) (NYSDOH 2007). Subsequent to the development and promulgation of Subpart 6-3, all spray grounds in New York State were required to conform to the new regulation. In the years since the regulation was enacted, spray grounds have been designed, approved for construction and used by patrons of New York State. The engineering plan review for both proposed and existing facilities, and eventual approval of these projects, provided valuable feedback that confirmed the regulation was well developed from engineering and regulatory perspectives.

Spray grounds are recreational water venues that spray water through features onto a play area and are attractive to a wide age range of patrons from very young children to adults seeking relief from heat. The typical design of spray grounds includes a hard play surface (spray pad) which may be constructed of brushed or polymer coated concrete. The spray pad has water features installed on it and a means of collecting the sprayed water for recycling. Modern spray grounds incorporate a variety of spray feature types including jets, fountains, dumping buckets, water cannons, and water falls. By definition, in New York State, a spray pad does not impound any water above grade so as to reduce the drowning hazard with respect to that of a pool. Spray pads may be scaled from very small (100 ft²) to much larger with surface areas measuring approximately 10,000 ft². As a result, spray pads have been placed in a variety of locations such as shopping malls, theme and municipal parks, and both indoor and outdoor recreational water complexes. Spray pads have also been proposed as serving dual duty by becoming decorative fountains during non-recreational hours.

The water treatment challenges that are unique to spray grounds (compared to pools) include: (1) high patron-to-water volume ratio (spray grounds have approximately 80–150 gallons of water per patron where pools may have approximately 400–500 gallons per patron); (2) high water–air interface that can impact water chemistry and quality with increased rates of evaporation; and (3) underground water storage limiting the ability for supervisory staff to make visual observations and decisions concerning the water quality.

The concept of how spray grounds fit into the recreational water environment was still being developed in 2005. The idea that a recreational water venue could meet patron needs by providing significant entertainment value without a drowning hazard was attractive to facility owners and planners. As a result, spray grounds were being proposed at an unprecedented rate in New York State. Swimming pools had been in use for decades, providing designers and operators with a large amount of empirical data with which to refine future designs. However, spray ground treatment design criteria were not yet developed. To address this, NYSDOH decided to develop a new regulation before the industry proceeded much further. To facilitate the time constraints, a first-principles approach was used to develop design requirements that would provide the basis for a predictable and scalable water treatment process. During the development of the design component of the regulation, industry experts and design professionals were consulted to ensure the new regulation would be tenable from an industry perspective.

Effective and efficient water treatment must remove particles and supply adequate disinfection. Particulate solids have the potential to shield organisms and consume disinfectant. The way interactive fountains/features distribute water over a play surface and the risk presented by Cryptosporidium-contaminated water warranted an examination of the methods used to develop and specify the treatment of spray ground water.

An ‘order of magnitude’ analysis was used to determine the appropriate water disinfection efficacy. For this analysis, the treatment tank volume was assumed to be 5,000 (of the order 10³) gallons and the minimum infectious dose for cryptosporidiosis set at ten oocysts (Chappell et al. 2006). In addition to this information, an estimate of the maximum number of oocysts deposited by an ill patron was needed to make the calculation. Goodgame et al. (1993) measured the concentration of excreted oocysts in stool of AIDS patients with cryptosporidiosis while counting the diarrhea events and measuring the total volume of feces. A statistical analysis was performed on the data to provide the 99% confidence level, which was 4.6 × 10⁸ oocysts in a single diarrheal event (of the order 10⁸). Therefore, in the case of a fecal accident at a spray ground, the resulting maximum average concentration of oocysts in the water is of the order 10⁵ per gallon. Using the minimum infectious dose of ten oocysts (Chappell et al. 2006), an individual could become ill by consuming less than 2.6 × 10⁻⁴ gallon (1 mL) of contaminated water. To reduce the probability of infection after a fecal (diarrheal) contamination event it was determined that a disinfection method would be required that could achieve at least four logs of inactivation of Cryptosporidium prior to the water being delivered to the spray pad for contact with patrons. Four logs of inactivation would result in a patron having to consume as much as a gallon of water (on average) to obtain ten oocysts immediately after a fecal accident. Consumption of this much spray ground water by a patron is unlikely.

Chemical disinfection methods were examined to determine their feasibility for this type of application. An analysis was performed assuming a hydraulically efficient treatment tank design (e.g., plug flow) and pseudo-first order disinfection kinetics to determine whether chemical disinfection could be adequate. The treatment tank was modeled as a chemical reactor to determine the volume and mean residence time that corresponds to 4-log reduction of viable Cryptosporidium oocysts. The analysis showed the volume would need to be unreasonably large (i.e., large ⁠) even for a hydraulically efficient treatment tank with elevated concentrations of disinfectant such as 10 ppm of free available chlorine (FAC). Ozone disinfection was considered as well; however, it was determined ozone disinfection had similar barriers to effective use as does chlorine. From this analysis it was determined chlorine-resistant organisms such as Cryptosporidium and Giardia (Korich et al. 1990) present a significant challenge to the use of chemical disinfection for their control. Clearly, alternative methods would need to be considered in order to meet the water treatment goals outlined previously.

The efficacy of UV disinfection is determined by the amount of ultraviolet radiation received by the water, which is measured as dose (in units of mJ/cm²). Full scale UV reactor validation procedures determine the operating conditions required to achieve disinfection efficacy with respect to the reduction equivalent dose (RED) for specific organisms (USEPA 2006). At the time the spray ground regulations were being developed, a common RED to which UV reactors were being validated was 40 mJ/cm² using MS2 (bacteriophage) as the surrogate. For more information on UV reactor validation procedures, the reader is referred to Masschelein (2002) and the draft and final UV disinfection guidance manuals published by the United States Environmental Protection Agency (USEPA 2003, 2006). The analysis contained in the draft UV Disinfection Guidance Manual indicated a MS2 RED of 40 mJ/cm² is likely to produce 4-log inactivation of Cryptosporidium, which was determined to be protective of public health for spray ground water. The final UV Disinfection Guidance Manual (UVDGM) (USEPA 2006) includes additional data that indicated a MS2 RED of 40 mJ/cm² would likely produce 4-log inactivation on the condition the water had a UV transmissivity greater than or equal to 85% (see Appendix G of the UVDGM (USEPA 2006)). Measurements made by NYSDOH personnel at several recreational water facilities indicated spray ground water typically has ≥85% UV transmissivity. In the event the water transmissivity or lamp output becomes too low to produce the required UV intensity to deliver 40 mJ/cm², the regulation required the treatment system to automatically divert the water from the features to prohibit inadequately treated water from coming in contact with patrons. For more information regarding UV reactor installation guidance, the reader is referred to Chapters 3 and 4 of the UVDGM (USEPA 2006).

UV disinfection is very effective for most microorganisms; however, it does not leave a residual in the water and therefore does not provide disinfection to the interior of the treatment tank, recirculation system plumbing, or the spray pad itself. Therefore, a requirement for chemical disinfection was included in the regulation. This requirement would provide a means for preventing the growth of biofilms on the interior wetted surfaces. It is important to maintain equipment free from biofilm because clusters of organisms can break free from surfaces and pass through the UV reactor without receiving adequate UV radiation to fully inactivate them. Halogen-based (chlorine or bromine) compounds have been determined to be effective for preventing the growth of biofilms and for inactivating many UV-resistant pathogens (in particular, some adenoviruses (Gerba et al. 2002)) at concentrations that had been approved for use in drinking water, and in then-current recreational water regulations (NYSDOH 2000). Additionally, chlorine and bromine compounds are effective at oxidizing organic materials that may otherwise accumulate in the water. The most commonly used disinfectants are liquid sodium hypochlorite solution applied using a chemical feed pump or pelletized calcium hypochlorite fed by an erosion feeder. Bromine compounds may also be used; however, the minimum residual is required to be 2.2 times the concentration required for chlorine (reported as Cl₂) in order to obtain the same disinfection efficacy. Bromine is fed to the spray ground water in stick form (bromochlorodimethylhydantoin) using an erosion feeder.

Maintenance of pH is important because it impacts the disinfection efficacy of chlorine (White 1992). A pH range of 7.2–7.8 was specified in the regulation to ensure chlorine efficacy while protecting the integrity of some spray pad surface finishes (e.g., plaster). At the upper pH limit of 7.8, the fraction of FAC that exists as hypochlorous acid, the more effective form (from a disinfection perspective), is less than 40% and decreasing (White 1992). The relatively small pH range was determined to be acceptable because automatic control systems can maintain the pH within limits.

The range of acceptable FAC concentrations was based on the need to maintain disinfection of the spray pad and other wetted surfaces. For example, an unnoticed fecal accident could leave large numbers of bacteria, viruses, and protozoa on the spray pad surface without the benefit of UV disinfection. Therefore, the minimum FAC concentration was set at 2.0 ppm. This is greater than the lower limit of 0.6 ppm required for swimming pools in New York State (NYSDOH 2000) because the patron-to-water volume ratio is much larger than for pools and the greater air–water interface could dissipate disinfectant in relatively short periods of time. The upper limit of the FAC concentration was set to 10 ppm, which represents a significant increase over the 5.0 ppm upper limit for pools and spas in New York State (NYSDOH 2000). An additional benefit of using elevated concentrations of FAC is that it allows a disinfectant residual to remain in the water for a longer period of time after the spray features are turned off. This helps minimize the growth of biofilms in the spray feature supply plumbing, which may develop in the absence of a disinfectant residual. The public health risk of biofilms was investigated by Rose et al. (1998), who demonstrated a link between plumbing biofilms and ‘lifeguard lung’ disease due to the inhalation of endotoxins from bacteria that may amplify in stagnant water and on pipe walls.

The upper concentration limit of 10 ppm can be justified by ocular and dermal studies performed by Gnemi (1991a, 1991b) that were subsequently reviewed by Bruch (2007). The ocular study examined the effect of placing 1,100 ppm FAC solution in the eyes of rabbits, which showed no effects during the subsequent 72 hours of observation. The dermal study entailed placing 1.1% (11,000 ppm) FAC on gauze in direct contact with shaved rat skin for 24 hours with a subsequent 14 day period of observation. The studies revealed no ill-effects in the test animals. A compilation of anecdotal incidences of the consumption of highly chlorinated water suggests drinking water with up to 50 ppm is not deleterious to humans (Muegge 1956). These studies suggest the upper limit could possibly be greater than 10 ppm; however, many commercially available test kits produce erroneous results at greater FAC concentrations because the chlorine begins to bleach the indicator dye.

The control of particles in spray ground water is important because they can shield microorganisms and/or consume disinfectant. Modeling was employed to understand how spray grounds accumulate particulate materials. Pool water acquires particulate contaminants from patron use and atmospheric particles settling onto the water surface. The rate of particle accumulation is roughly proportional to the number of patrons that can fit in a pool which is limited by surface area. Spray grounds, however, distribute water over a play surface that may be disproportionately large compared to the volume of water stored in the treatment tank. This causes the water to become particle laden at a rate proportional to the spray feature flow rate, not the treatment tank volume. For example, consider two spray grounds both using a 5,000 gallon treatment tank to supply water to the features. The first spray ground has a feature flow rate (water provided to the spray pad) of 1,000 gallons per minute (gpm) as compared to another system, where the same volume of water is used to provide water to a spray pad at 100 gpm. The spray ground that has a feature flow rate of 1,000 gpm will expose the water to the environment at ten times the rate and therefore collect particles more rapidly as compared to the system with a 100 gpm feature flow rate. To address this problem, the water treatment process development effort applied modeling, filter equipment specifications (NSF/ANSI 50) and treatment standards to determine design and operation parameters (NSF International 2004).

The amount of turbidity acquired by water while it travels from the features toward the spray pad drain is a model parameter that was measured at an operating spray ground. The value of was determined by measuring the turbidity of the water flowing down the main drain and subtracting the turbidity of the water discharging from a feature nozzle. The value was determined to be ∼0.25 nephelometric turbidity units (NTU). The data used were collected on a breezy day when the air was laden with cottonwood seeds. Clearly, this model parameter may be a function of the type of features used and the population of patrons on the spray pad. Further measurements should be taken to confirm the value of for high-use spray grounds.

The model (see Figure 2(a) and 2(b)) used to describe the accumulation and removal of turbidity from the water uses the parameter f which represents the fraction of turbidity remaining in the water after it passes through the filter. This value was assumed to be a constant to simplify the analysis and provide a parametric study. The instantaneous filter efficiency was determined to be 24% from the minimum filter performance required for listing by NSF International (2004). From Equation (1), the only remaining quantity needed to determine the maximum feature to filter flow rate ratio (S) is the permissible upper limit for turbidity.

The maximum turbidity of 3.0 NTU was obtained by comparing the prescribed limit for unfiltered drinking water and the results of UV disinfection research. The Surface Water Treatment Rule (SWTR) (USEPA 1989) referenced the work of LeChevallier et al. (1981) who suggested the upper turbidity limit of 5.0 NTU for unfiltered drinking water. UV disinfection research (Malley 2000) demonstrated, for certain types of turbidity-causing particles, the amount of power required to attain 2-log (99%) inactivation of MS2 bacteriophage increased significantly above 3.0 NTU. Based on this comparison, NYSDOH defaulted to the lower value of 3.0 NTU for spray ground water to assure effective UV disinfection. Using the permissible upper limit for turbidity of 3.0 NTU, the analysis resulted in a maximum feature to filter flow rate ratio of 3:1 for bypass systems (Figure 2(a)). Spray grounds that filter all the water prior to being conveyed to the spray pad (full-flow filtration) do not need to address the flow rate ratio as there is only one stream, and the maximum expected turbidity would be approximately 1.0 NTU (Figure 2(b)).

Basing the filtration flow rate on the flow provided to the spray features rather than the treatment tank volume represents a departure from traditional pool recirculation system designs that are focused on turnover time. As a result, the maximum flow rate ratio of 3:1 allows spray grounds to be scaled over a wide range of sizes with a ready means of determining the filter flow rate. The efficiency of the treatment process may be enhanced by considering treatment tank hydraulics that possess a smaller degree of mixing of newly arrived (from the spray pad drain and filter effluent) and existing water in the treatment tank.

The result of addressing combined pool–spray ground systems using the methods described above provides a means for treating the water to significantly reduce the likelihood of a waterborne disease outbreak.

The NYSDOH recognized the need to develop a new regulation (Subpart 6-3) to address the design and operation parameters for spray grounds in order to protect public health. The unique character of spray grounds (compared to pools) warranted a more rigorous analysis to develop the specifications of a water treatment system. As a result, the design component of Subpart 6-3 is based on process engineering methods that utilized modeling and the scientific literature. Subpart 6-3 also stipulates process automation requirements to reduce the need for operator intervention and the potential for operational excursions such as inadequate/excessive disinfectant concentration or improper pH.

The requirement for UV treatment is based on the inability of traditional chemical methods to adequately inactivate chlorine-resistant organisms in a treatment tank of reasonable volume (or ⁠). The design and operation of UV reactors focused on RED instead of log inactivation of a specific organism to increase the likelihood the regulation would be protective of public health in the event of the discovery of an emerging pathogen that may be more resistant than the organisms that were of immediate concern. Focus on RED also eliminated the need to measure the UV transmissivity, which is required to determine the log inactivation using the validation procedures that were published in the UVDGM (USEPA 2006).

The most important requirements for spray ground system design include the following:

• UV treatment (RED = 40 mJ/cm²) of all water provided to the features using a UV reactor installed in the spray feature supply plumbing.

• Chemical disinfection to maintain a biofilm-free system.

• Automatic control systems for treatment tank water level and chemistry (disinfectant and pH) maintenance.

• An automatic means of diverting water from the features in the event of a failure of the UV reactor to deliver at least 40 mJ/cm².

• A maximum feature to filter flow rate ratio (S) maintained by automatic flow control systems in the filter and feature supply plumbing.

• A means of diverting water to waste while the spray pad is not in use or undergoing cleaning.

Subpart 6-3 has been used as the standard for reviewing spray ground designs in New York State since the rule went into effect in 2006 and was finalized in 2007. During that time, the regulations have been shown to be defensible and readily applied to proposed designs. Future work on this code should include measurement of the maximum contamination burden for spray grounds to confirm the maximum value of the feature to filter flow rates (S). Further analysis of treatment tank design would also benefit particle removal and improve the overall process.

The authors would like to thank Ramesh K. Kapur, P.E. (retired) and Douglas C. Sackett (retired) of the New York State Department of Health for their support during the development of the water treatment portion of the regulation. The authors extend their appreciation to Paul Kaczmarczyk, Research Scientist of the New York State Department of Health, Bureau of Water Supply Protection for his assistance in editing the manuscript.