Media Filtration

Media-type filters are commonly used in LARFs as either stand-alone systems or pretreatment modules in more complex water treatment systems. There are many types of media filters used. The number and type of filters to use is dependent on the volume of water to be treated, whether water must be filtered continuously or in batches, the amount and sizes of suspended particles in the source water, the concentration of organic compounds, and the acceptable amount of contaminants in the treated water. A comprehensive water analysis (Curran and Smart 2007) should be conducted on the facility’s domestic water supply in order to determine the need for media filtration. Once the need has been determined, a water quality consultant or engineer can help determine the type of media filtration that would be best suited for removing particular water contaminants of concern. A water quality consultant or engineer will also be able to assist with the sizing of the media filtration and the correct order that the domestic water supply should pass through the media filtration units in order to achieve the facility’s water quality goal.

Media filters are made from a wide variety of materials and can filter gross particulates from the domestic water supply down to particles as small as 0.1 μ. Media filters are designed to entrap suspended particles as the water passes through the media. Much like high-efficiency particulate air (HEPA) filters, as the media loads with particles, the flow of water through the media surface becomes restricted, resulting in a reduced flow rate and a reduced volume of treated water produced over time. With cartridge media filters, the filter membrane can be easily removed from the filter housing and replaced with a new cartridge. Sand filters, which are designed for the removal of gross particulates from the domestic water supply, require periodic back-flushing in order to maintain filtration efficacy. The periodic back-flushing, which is normally accomplished automatically with a timer–valve assembly, flushes the particulates from the sand media to a drain.

Activated carbon filtration is another type of media filtration that is commonly used to treat domestic water supplies. Activated carbon filtration is specifically designed to remove organic compounds. Activated carbon filtration differs from sand or cartridge-type filtration by utilizing the adsorptive properties of carbon rather than the particle entrapment process utilized with sand or cartridge-type filtration processes. Through the adsorptive process, organic compounds that are dissolved in the domestic water supply are attracted to the surface of the activated carbon particles and thereby removed from solution. There are many factors that influence the efficiency of this process, including the carbon particle size and pore structure, the amount of actual time that the incoming water is exposed to the carbon media, the incoming water flow rate, and the amount of activated carbon remaining in the filter media housing. Once carbon adsorption sites are occupied by contaminants removed from the incoming water supply, the ability of activated carbon to adsorb additional organic compounds is lost, resulting in “breakthrough.” Breakthrough is the process by which previously bound organic contaminants are released back into the water by the activated carbon media. Once the breakthrough point has been reached, the filtering capacity of the activated carbon filtration media will be less than optimum (Dvorak and Skipton 2013). Similar to sand media filtration, periodic back-flushing to remove captured contaminants from the activated carbon media or replacement of the carbon media is required in order to maintain adequate filtration of dissolved organic compounds from the domestic water supply.

Reverse Osmosis and Ultrafiltration

In order to reduce the potential for experimental variables in research protocols, many LARFs have invested in water purification systems that reduce or eliminate the risk of using domestic water supplies that may have contaminants remaining in the drinking water after having been treated with simple media filtration. RO and ultrafiltration (UF) are two water purification methods that are commonly used to purify animal drinking water destined for laboratory animal consumption.

RO is a process that utilizes high-feed water pressure to overcome osmotic pressure and produce a product that is predominantly free of low-molecular-weight contaminants. RO membranes are capable of removing contaminants larger than 0.001 μ, including monovalent and multivalent ions, pyrogens, proteins, colloidal substances, and microorganisms. In order to understand how RO systems work, it is worthwhile understanding the theory of RO, the design and function of the RO membrane, and the related components that produce a purified water product that is suitable for consumption by laboratory animals.

The RO membrane is the primary element of an RO system. The RO membrane is semipermeable, allowing purified water to pass through the membrane when high pressure is applied to the incoming water. The purified water that passes through the semipermeable membrane and is suitable for laboratory animal consumption is aptly named the “permeate,” whereas the water that remains on the inlet side of the membrane and is highly concentrated with dissolved ions and precipitated heavy metals is aptly called the “concentrate.” There are several different RO membranes from which to choose. The choice should be based on the water quality goals of the LARF, the water quality test of the domestic water supply, and the operating environment of the RO system. The materials commonly used to make RO membranes include cellulose acetate, polyamide, polyether sulfone, polyacrylonitrile, and polyvinylidiene fluoride. Polyamide and cellulose acetate are the most common membrane materials used in laboratory animal drinking water applications. Polyamide membranes have a slightly higher salt rejection rate than cellulose acetate membranes. However, polyamide membranes will degrade and become ineffective at removing water contaminants if exposed to chlorine. If polyamide membranes are selected, chlorine must be removed from the domestic water supply prior to the RO process, and should be injected into the RO permeate water after the membrane of the LARF is interested in affording protection against possible biofilm formation.

RO membranes are very thin (≈0.2 μ), with a pore size as small as 0.0001 μ. These membranes are supported by a polymeric microporous thermoplastic layer that adds strength and protects the integrity of the membrane. The RO membrane is composed of several of these composite layers, which increases the filtration surface area available for the feed water.

Over the years, there have been many advances in the design of RO membrane modules. Today, RO membranes are available as hollow fiber modules, spiral wound modules, and plate or frame modules. Each module design has application-specific characteristics. Modules are assembled in special pressure vessels of various dimensions, diameters, and lengths and fitted with feed water, permeate, and rejection water fittings. Earlier RO membrane designs recovered less than 10% of the water used. With continued advances in RO membrane technology and the ability to interconnect multiple RO membrane modules in series, significantly higher recovery rates of greater than 90% (Stoughton et al. 2013) can be achieved while providing higher flow rates and flux to meet any demand.

Although a high capital investment, RO systems offer animal resources the distinct advantage of providing a consistent and pure source of drinking water regardless of changes in supply water quality. Commercial RO systems are available in a variety of different configurations and options that can provide variable permeate flow rates, percent recovery, and percent rejection. Depending on the specific needs of the institution, RO systems can be sized as a single centralized site plant with distribution to all campus animal resource centers or designed to handle the specific needs of a dedicated point of use.

UF can provide a cost-effective alternative to RO water purification. Similar to RO purification, UF has the capability to remove microorganisms, particulates, and macromolecules that are greater than 0.01 μ. Unlike RO membranes, UF membranes do not have the ability to remove low-molecular-weight organics and elemental ions such as calcium and sodium. UF membranes have greater pore sizes than their RO counterparts, and therefore require less applied pressure to overcome osmotic pressure. The lower applied feed water pressure, in combination with the greater pore size of the UF membrane, translates into a higher flux rate (>90%) of purified water (Pilutti and Nemeth 2003).

The domestic water supply that feeds an RO or UF system will likely contain concentrations of suspended solids, dissolved ions, and organic compounds that have the potential to prematurely foul RO or UF membranes. It is important to have a water quality test performed to determine what pretreatment methods should be used to reduce these water contaminants so that maximum membrane life is achieved, and to also ensure that the pretreated supply water is compatible with the intended membrane type. As the supply feed water interfaces with the semipermeable membrane barrier, contaminants from the supply feed water settle on the membrane surface and load the various microporous membrane layers and feed water channels. This process is referred to as membrane fouling, which can negatively impact the performance of an RO or UF system (Amy 2008). Under the right conditions, dissolved salts may also precipitate from the feed water stream and accumulate on the membrane surface in the form of scale.

To circumvent these conditions and prevent premature fouling of RO and UF membranes, pretreatment of the feed water is commonly used to improve the quality of feed water. As described in the previous section, media filtration may be used to remove suspended particles, colloidal materials, and microorganisms. Activated carbon media filters will remove organic compounds and chlorine. Residual chlorine used as a disinfectant in domestic water supplies may damage certain RO or UF membrane materials (polyamide). Newer membrane technology is far more tolerant of residual chlorine (Lee et al. 2011). To protect the RO or UF membranes from scale formation, water softeners may be used to remove Ca²⁺ and Mg²⁺ ions.

Although pretreatment provides some membrane fouling protection for RO and UF systems, over time the total volume of water passing through the membrane vessel will ultimately lead to membrane fouling. Periodic back-flushing of RO or UF membranes with detergents is an additional procedure that can be utilized to extend membrane life.

Most RO and UF systems cannot produce purified water at a rate that will keep up with the peak demand experienced in most LARFs. This is particularly true in facilities that utilize water bottles or disposable water pouches. If the LARF is utilizing an automated animal drinking water system, a smaller-volume tank may be sufficient, but the storage tank will need to be adequately sized to accommodate the daily flushing of the room distribution piping and the manifold piping on the individual animal holding racks. Large purified water storage tanks allow the RO and UF systems to spread production over time, yet still meet the immediate demand for animal drinking water during peak hours of facility operation. Large purified water storage tanks also provide the LARF with a purified water reserve, which can prove essential during disaster scenarios, during periods when the RO or UF system is down for maintenance, and for other contingency planning.

Posttreatment of RO or UF purified water is essential to maintaining animal drinking water that is free from microorganisms. If uncontrolled, bacteria will begin to proliferate in the purified water medium and will attach to the surfaces of the tank and to the water distribution piping and manifolds, forming biofilms (Molk et al. 2013). There are a number of ways to proactively approach protection, including the use of low-level chlorine or acid injection after RO or UF purification. When using purified water recirculating loops as a method of distributing the purified water to bottle filling systems or automated animal drinking water systems, ultraviolet C (UV-C) radiation, in conjunction with additional microfiltration, is often used to keep bacteria from proliferating and forming biofilms within the purified water holding tanks and distribution piping. However, in this application UV-C has limitations (described later) that should be given due consideration.

Monitoring RO or UF system performance is a critical function to ensure that water is always available and consistent in quality and to ensure that system performance is within specifications. Most modern RO and UF purification systems offer continuous monitoring and notification features that will monitor key parameters of the system and proactively notify designated institutional staff when parameters are out of bounds. It also serves as a means of timely remediation of potentially significant issues without compromise to continuation of service or system integrity.

Ultraviolet Irradiation

Used in water treatment and wastewater treatment facilities to disinfect and enhance water quality, ultraviolet (UV) irradiation is also a viable solution available for disinfecting laboratory animal drinking water in LARFs. As mentioned previously, UV irradiation is typically used as a treatment method for reducing the proliferation of bacteria and biofilms in recirculating purified water loops. UV radiation is electromagnetic radiation that lies between the visible light spectrum and x-rays and typically falls between 100 and 400 nm. UV-C radiation, with a wavelength falling between 245 and 285 nm on the UV spectrum, is the most effective radiation range for inactivating microorganisms in water systems. This wavelength is produced by special low-pressure mercury vapor lamps housed within a UV-C reactor that emit electromagnetic radiation within this wavelength. The UV-generating lamps do not come in contact with the water. The lamps are either surrounded by a quartz sleeve within the lumen of the UV-C reactor or arranged around UV-C-penetrable tubes. The purified water source must constantly pass through the reactor in order for the water to receive the UV-C treatment, thereby destroying microorganisms as they come into contact with the UV-C radiation. UV-C energy does not add anything to the water, nor does it alter water chemically or remove other water contaminants (Benjamin and Lawler 2013). UV systems used in conjunction with recirculated or single-pass systems do offer the benefit of reducing microbial loads; however, due to design limitations, these systems may not totally eliminate the threat of microbial growth. Biofilm-producing colonies at any point in the recirculating loop or in the water distribution system will be unaffected by UV unless these colonies break free and pass through the UV reactor, where they will be exposed to the UV-C radiation. Therefore, additional water treatment options should be considered depending on the water quality goals of the laboratory animal research program.

UV-C energy penetrates and inactivates microorganisms by photochemically transforming UV-sensitive nucleotides, particularly pyrimidines. When exposed to UV-C radiation, the RNA or DNA transcription and replication process within the microorganism is impaired, compromising the essential cellular processes that eventually lead to the death of the microorganism. Certain microorganisms have the ability to repair the photochemical damage caused by the UV-C radiation through a process known as photoreactivation or dark repair. The ability of certain microorganisms to survive UV-C exposure is usually due to the effect of radiation doses that are too low. It has been shown that photoreactivated microorganisms show a greater resistance to UV radiation than microorganisms that have not been exposed and survived (Hijnen 2010).

There are several factors that will determine the efficacy of UV-C irradiation as a reliable method for disinfection of laboratory animal drinking water systems. These factors include the UV sensitivity of microbial species, the UV-C radiation wavelength or dose, the presence of water contaminants that absorb UV-C radiation, the presence of suspended particles that can block UV-C penetration, the age of the UV-C lamps, the physical distance between the UV-C radiation source and the microorganisms, and the presence of radiation-blocking film on the interior surfaces of the quartz sleeves or tubes. These factors must be monitored for effect, and periodic maintenance must be performed on the system to maintain optimal performance.

In general, most viruses, yeasts, and bacteria are sensitive to UV-C irradiation and require much lower UV radiation doses than molds and protozoa. Giardia and Cryptosporidium, which are typical protozoa found in water, require much higher doses of radiation for effective treatment (Chen et al. 2006).

Dosage, intensity, and exposure time are interrelated and important factors of UV system efficacy. Dosage is the product of intensity and exposure time expressed in seconds. Intensity is the output energy from the UV-C lamp and expressed as microwatts per square centimeter (μW/cm²). Organisms passing through a UV reactor at a faster and more direct rate will receive a much lower dose than those that take a more circuitous path and slower rate.

UV-C lamp intensity does not remain constant over the life of the lamp. New lamp intensity declines rapidly at first, but the rate of decline lessens over time. As intensity declines, a point is reached where lamps should be replaced. Replacement frequency will largely be dictated by the type of lamp used and the minimum intensity determined empirically. UV-C intensity meters are available to assist in this regard. Generally speaking, low-pressure mercury-type lamps have a life of roughly 1 year.

The presence of UV-absorbing contaminants like sulfites, nitrites, phenols, humic and fulvic acid, iron, and turbidity reduces the UV radiation intensity that is available for microbiological control. Suspended matter harboring microbes may block UV energy and also reduce UV system efficacy. Contaminant deposits on the internal surfaces of the reactor or lamp sleeves also affect system efficacy. To mitigate the affect water quality has on UV systems, establishing a program of assessing intensity, cleaning UV reactors, and replacing lamps is warranted to ensure optimal performance.

UV systems used in conjunction with recirculated or single-pass animal drinking water systems do offer the benefit of reducing microbial loads; however, they may not totally eliminate them. Biofilm that may be growing downstream of the UV reactor or at any point in the recirculating loop or water distribution system will be unaffected by UV-C radiation. Bacterial colonies that break away from biofilms can be neutralized as they pass through the UV reactor; however, the inactivated bacteria will provide a rich nutrient source for biofilms growing on surfaces downstream. Additional water treatment options should be considered depending on the water quality goals of the LARF.

Acidified Water

Inorganic acid has long been used in laboratory animal drinking water systems as a means of controlling bacterial contamination. Acidified water has been particularly effective at eliminating gram-negative opportunistic pathogens like Pseudomonas aeruginosa, a common organism found in domestic water supplies. Although effective against P. aeruginosa and other gram-negative microbes, some microorganisms (e.g., acid-resistant fungi) (Edstrom Industries 2003) can be unaffected and survive (Meltzer 1993).

Hydrochloric acid is the predominant type of acid used to treat water supplied to laboratory animals, although sulfuric acid provides an effective option. A pH of between 2.5 and 3.0 is the recommended concentration of acid in animal drinking water systems. A pH below 2.5 has been shown to affect weight gains and water consumption in male mice (Hall et al. 1980). The use of acidified water may have some undesirable side effects on laboratory animal research, which should be considered prior to use. In some instances, acidified water may react with water bottle stoppers and release undesirable substances (Kennedy and Beal 1991). Acidified water may also alter the excretion of phenol red, and result in lowered proteinuria and decreased urine volume in rats (Clausing and Gottschalk 1989). Acidified water has also been shown to alter the gut microbiome and the incidence and onset rate of diabetes (Sofi et al. 2014; Wolf et al. 2014).

Acids used to treat water are corrosive by nature, and can cause damage to animal drinking water treatment and delivery systems, and even injury to facility personnel. Care should be exercised in the handling and storing of acid reagents following facility standard operating procedures (SOPs), using appropriate personal protective equipment (PPE), and referencing the appropriate safety data sheet (SDS). Materials used in the design and manufacture of animal drinking water treatment and distribution systems should be resistant to the corrosive properties of these acids. Stainless steel and certain plastic piping materials will resist the corrosive nature of inorganic acids. Silicone elastomers and Teflon used to join and seal piping materials are also resistant to the corrosive effects of acid. Brass and other metallic materials should be avoided. For automated animal drinking water systems, type 316L stainless steel is the material of choice. Type 316L stainless steel is resistant to prolonged exposure to low pH drinking water; however, it is incumbent upon the LARF staff to monitor the pH for maintaining the acceptable range of 2.5–3.0. Severe and irreversible damage has occurred to systems using 316L stainless steel when acidified animal drinking water has been maintained at levels below 2.5 pH. The anticorrosive properties of 316L stainless steel are due primarily to its unique concentrations of nickel and chromium and the addition of up to 3% molybdenum. Type 316L stainless steel utilized in animal drinking water systems is commonly exposed to a “passivation” process during its manufacture. Passivation of stainless steel further enhances the corrosion-resistant properties of the material by removing free iron from the surface and leaving a protective chromium oxide layer. Damage to this chromium oxide layer can occur when acidified water is maintained below a pH of 2.5, allowing the acidified water exposure to the ferrous material underneath the protective layer. When this protective layer becomes breached, pitting of the stainless steel will occur, allowing for pinhole leaks and eventual failure of the material.

Disposal of acidified water into the municipal waste stream may require special handling (e.g., neutralization) prior to its release. Therefore, it is suggested that facility administrators give this point due consideration should acidification of the animal drinking water be implemented.

Chlorination

The use of chlorine compounds to treat animal drinking water is an effective and acceptable means of disinfection. Chlorine compounds added to the animal drinking water supply preserve the quality of the water by preventing microorganisms sensitive to chlorine from colonizing the water distribution system. Even in low concentrations, chlorine compounds in an aqueous solution have a profound effect on eliminating episodic microbial contaminations. Understanding that animal drinking water can become contaminated with microbes at any given time, consideration should be given to the use of chlorine compounds as an adjunct treatment of the water provided to laboratory animals.

There are many chlorine compounds available to treat potable water supplies. Chloramines, sodium hypochlorite (NaOCl), and chlorine dioxide (ClO₂) are among the most predominant compounds used today for water disinfection. At the correct concentrations, chlorine compounds are safe for human and animal consumption and may be a means for eradicating endemic chlorine-sensitive microorganisms from rodent colonies (Takimoto et al. 2013). The chlorine compound and final concentration used for biocidal activity, whether it is for preventative measures or treating a water supply, should be determined based on the biocidal results at a given concentration and understanding that certain contaminants in the source water can react with chlorine, making its biocidal characteristics less effective.

The chlorine compound used for treatment and the final concentration of chlorine used in animal drinking water should be given due consideration. Although somewhat variable depending on the contamination level of the water supply, concentrations of free chlorine at <10 ppm for a sufficient period of contact time are generally adequate to treat for most microbial contaminants. However, microbiological assessment of the water supply and determination of the optimal chlorine concentration are warranted. Although laboratory rodents can tolerate high concentrations in drinking water, previous studies carried out by the National Toxicology Program (NTP 1992) in F344/N male and female rats and B6C3F1 male and female mice indicated there was an inverse relationship between increasing concentrations of two chlorine compounds (up to 275 ppm available atomic chlorine and up to 200 ppm chloramine administered to both species) and decreased water consumption and body weights over a 2-year period.

Chlorine compounds used as water disinfectants are most effective in solution when water chemistry conditions are optimal. The effectiveness of chlorine compounds in solution is influenced by pH, the presence of organic matter, water hardness, temperature, and chlorine concentration. The pH of the water has the greatest influence on the antimicrobial properties of chlorine (Dychdala 2001). In the case of sodium hypochlorite, an increase in pH significantly decreases the disinfecting ability of chlorine, whereas a decrease in pH has the opposite effect. Sodium hypochlorite is the active ingredient in household bleach and is perhaps the least expensive source of sodium hypochlorite available for use as a water disinfectant. When added to animal drinking water supplies, sodium hypochlorite molecules react with the water molecules to form hypochlorous acid (HOCl) and sodium hydroxide (NaOH). HOCl dissociates in water to form hydrogen (H⁺) and hypochlorite (OCl^–) ions. Of the two compounds, HOCl exhibits far more germicidal activity than OCl^–. The relationship of the HOCl and OCl^– ions in solution is pH dependent. As pH increases, the concentration of OCl^– increases. Conversely, as the pH decreases, the concentration of HOCl increases (Dychdala 2001). When used as a disinfectant in water supplies, sodium hypochlorite can react with organic compounds dissolved in the water, forming trihalomethanes. Trihalomethanes are suspected to be carcinogenic and may add an additional variable when consumed by animal research subjects under research protocols.

Chlorine dioxide is more frequently being used for drinking water disinfection in municipal water supplies. Unlike other chlorine compounds, chlorine dioxide is unaffected by high pH and does not form trihalomethanes in the presence of organic material. Chlorine dioxide has a higher oxidative capacity than sodium hypochlorite and is therefore much more effective as a biocide at lower concentrations.

Chlorine is a powerful oxidizing agent and corrosive by nature. Stainless steel is susceptible to the corrosive effects of chlorine; however, various grades of stainless steel are more resistant to the effects of chlorine corrosion than others. As is the case with acids, type 316L is more resistant to chlorine corrosion damage than other stainless steel grades commonly used in the industry.

Steam Sterilization

Steam sterilization is one of the oldest and most effective methods of treating laboratory animal drinking water. Under the right conditions, steam sterilization destroys all microorganisms present in the water medium, including viruses, bacteria, spores, fungi, molds, and protozoa. Steam sterilization is effective at deactivating these microorganisms both in the water and when they are attached to water storage and containment surfaces. Steam is a penetrating sterilization medium, able to overcome biofilms and the cellular defense mechanisms of the individual microorganisms.

Achieving sterility is a function of adequate exposure time to both temperature and pressure. The quality of the steam produced by the sterilizer is also very important. Dry saturated steam or superheated steam is preferred over wet steam. Superheated or “dry” steam has less entrained water droplets than wet steam and, as a result, can transfer more heat energy during the sterilization process. Transferrable heat energy is important, as it promotes sterilization efficiency and efficacy. Steam sterilization is a nontoxic approach to treating animal drinking water, is relatively inexpensive, and kills microorganisms rapidly.

Steam sterilization is typically used to sterilize animal drinking water bottles that are either filled or empty. The sterilizer chambers that are common on most steam sterilizers used in LARFs are significant in size, allowing the facility to sterilize hundreds of bottles per sterilization cycle. Steam sterilization of filled and empty bottles is handled in very different ways. Empty bottles can be sterilized very quickly using a high-vacuum sterilization cycle. In contrast, bottles filled with animal drinking water must utilize a slower processing cycle in order to prevent the drinking water from boiling out of the bottles. This process exchanges vacuum pulses for steam pulses to evacuate air from the sterilization chamber. A slow exhaust cycle is used to cool the chamber, rather than the dry cycle that is used for nonliquid materials. Purging air from the chamber with steam is a function of the density of the steam versus the air. Since air is denser than steam and therefore heavier, it settles to the bottom of the chamber and is forced to the sterilizer drain as steam is added to the chamber. The exhaust process involves slowly lowering the chamber temperature and water temperature in the bottles to a point where the water temperature falls below the boiling point so that the water remains in a liquid state and does not boil over. Many sterilizer manufacturers offer load probes as a means of confirming product temperature throughout the sterilization process. However, a load probe provides only one reference point within the chamber, so other means of verification may be required.

It is important to establish a process of evaluating sterilization efficacy and validation of sterilization cycle parameters after the sterilization cycle is complete. Validation can be accomplished through the use of biological indicators for liquid cycles. It is important to place biological indicators at multiple sites throughout the entire load to provide greater assurance that the load has been sterilized.

Steam sterilization is an effective means of sterilizing animal drinking valves that are an integral component of automated animal drinking water systems and some disposable water pouch systems (e.g., stainless steel animal drinking valves). Since animal drinking valves are the farthest point away in terms of distance from where the water was initially treated, and since these valves come in direct contact with the research animals, animal drinking valves can harbor microbial contaminants that can cross-contaminate through the cage replacement process. The stainless steel manifold pipes that are integral to automated animal drinking water systems used to supply water to individually ventilated caging (IVC) racks and other laboratory animal housing systems are designed to be sanitized either chemically or through the use of steam sterilization. Attention to the manufacturer’s recommendations of cycle parameters, including temperature, vacuum depth, and dry cycle, is important in order to prevent damage to this equipment.

Water Bottles

Arguably one of the most versatile water delivery systems, water bottles have withstood the test of time and remain a viable option for water delivery to laboratory research animals. Water bottles also provide a means of customizing water treatments for subsets of laboratory animals for which it is much more difficult to provide using other water delivery systems. Water bottles come in a variety of shapes and volumes to meet the needs of the animals, husbandry schedules, and research protocols. Volume is a particularly important aspect, as it relates to its service life and recycle frequency. The bottle volume choice should be compatible with the animal housing type, with ample consideration given to the recycle (changing) frequency of the cage. Operational costs should also be considered in the selection of the bottle volume and materials. The change frequency of the cage and the bottle will affect the labor associated with the bottle handling process, so larger bottle volumes that will extend bottle change-out will reduce operational costs.

The materials used in the manufacture of animal drinking water bottles and sipper caps should also be considered. Water bottles are commercially fabricated from a variety of materials, including glass and various plastic polymers. Exposure of bottles and sipper caps to sanitizing chemicals and the steam sterilization process will degrade the material composition over time.

Glass is a fragile material long used in the industry, and although still used today, glass represents handling issues for husbandry and scientific staff due to its relative weight compared with plastic polymers, and due to operational and personnel injury costs due to frequent breakage. Research has shown that under certain circumstances, glass bottles may contribute silicon to the drinking water, which can introduce a research variable (Lohmiller and Lipman 1998).

Polymer-type bottles offer the advantage of being lightweight, durable, transparent, and resistant to heat and chemical degradation. They are available in a variety of volumes and shapes to suit the considerable variety of laboratory animal research species and the corresponding variety of animal housing systems. Although polymer-type bottles offer many positive features for use as a water containment vessel in LARFs, one type of plastic polymer, polycarbonate, has been associated with the leaching of bisphenol A (Howdeshell et al. 2003; Hunt et al. 2003), a synthetic organic compound used in the production of polycarbonates and other thermoplastics. Bisphenol A and analogs bisphenol S and bisphenol F (Rochester and Bolden 2015) are characterized as endocrine disruptors, exhibiting estrogen-like properties. In contrast, polypropylene, polyethylene, and polyethylene terephthalate (PET) are not made with bisphenols or other known endocrine disruptors, suggesting that these materials represent a viable alternative to the use of affected thermoplastics. The introduction of endocrine disruptors through the supplied drinking water can add a variable to research, and perhaps disturb normal breeding and reproductive cycles in laboratory animal species.

Stoppers or sipper caps that are affixed to the open end of water bottles are also available in a wide range of materials. Black rubber and neoprene stoppers are the predominant types. Of these two types, neoprene is more resistant to degradation from repetitive steam sterilization cycles; however, hardening of the stoppers, as a consequence of steam sterilization, occurs with both types. Hardening of the stoppers can lead to the breakdown and release of particles from the stopper into the drinking water. These particles will ultimately be consumed by research animals with potentially negative consequences. Rubber and neoprene stoppers are also known to release minerals and heavy metals (Kennedy and Beal 1991; Nunamaker et al. 2013) into the drinking water.

An alternative to conventional stopper and sipper tubes is capped bottles with a “pinhole” orifice. The design characteristic of this approach relies on the inherent surface tension of water to prevent it from leaking out of bottles while providing free access to water. However, accumulation of debris in the orifice can lead to water delivery failure. Therefore, careful attention to appropriate cleaning and inspection of the bottle and orifice is warranted.

Automated Watering Systems

Automated watering systems in LARFs have been in existence for more than 50 years. These systems provide a viable and reliable alternative to using water bottles. Automated watering is a centralized approach to delivering water to laboratory animals. These systems begin with centralized water purification and treatment, and then use a network of pressure reduction stations and distribution piping to deliver the purified and treated animal drinking water to the animal holding rooms within the LARF. Manifolds on large animal enclosures, or racks for smaller species can connect to the animal room drinking water distribution system to provide a quality drinking water product at the cage or enclosure level. Drinking valves have been designed to accommodate the different water volume requirements and natural water access instincts of most laboratory animal species. Modern automated watering systems are designed with control systems that allow for automated daily flushing of the room, rack, or enclosure distribution system. These systems may also be equipped with advanced monitoring capabilities to enable notification of facility staff in the event of a system failure or leak. Automated watering systems can be designed to address specific facility layouts and budgets, and to meet institutional water quality standards.

Animal drinking valves are adequately designed to reliably and repeatedly deliver high-quality water to laboratory animals with minimal effort and maintenance. With the wide adaptation of IVC systems for rodents, drinking valves were required to be located on the inside of the individual cages. The design of these new valves had to effectively prevent water leakage into the cage caused by particulates and other contaminants in the drinking water, or by the natural behavior of rodents pushing bedding into the frontal opening of the valve. Although the initial designs of these new rodent drinking valves did not perform well, the various manufacturers have refined and improved valve designs to the point where rodent automated drinking water systems are reliable, acceptable, and safe for these valuable research subjects.

Automated watering systems utilize two typical design approaches to distribute treated animal drinking water to the racks and kennels located in animal holding rooms. Room or rack flushing systems rely on a principle that once the supply water is treated and distributed to the room or rack level, the water should be either utilized for animal drinking or flushed to drain. This system design eliminates the potential of the water dissolving contaminants downstream at the room or cage level and bringing those contaminants back to the treated water storage tank. Since animal drinking volumes rarely provide enough water exchange to keep the water fresh and free from bacterial growth, room or rack flushing systems rely on facility staff to either manually flush the distribution lines and manifolds to drain or utilize a sequencing controller to automatically flush solenoids at the end of the distribution lines. To reduce the risk of microbial growth on the internal surfaces of the distribution system (biofilm), modern room and rack flushing systems utilize a proportioning system to inject low levels of chlorine or inorganic acid into the drinking water. Automated flushing provides a mechanism to flush at prescribed intervals, for a given duration and at prescribed times. Modern flush sequencing and monitoring control panels have the added capability of providing documentation that the flushing actually occurred. They can also monitor key system parameters and provide alarm notification to facility personnel in the event of a system failure.

In contrast, recirculating room and rack distribution systems rely on a continuous flow of water from the animal drinking water holding tank to the animal holding rooms and back to the holding tank as a means of providing clean, fresh water to the research animals. These systems rely on additional microfiltration and the use of a UV-C reactor to reduce the potential for microbial contamination. Recirculating systems offer the advantage of conserving water; however, they employ the use of one or more continuously running pumps to move the water. For the reasons noted above, recirculating systems are more easily contaminated with microorganisms when relying solely on UV-C radiation and filtration. For this reason, recirculating systems often have a proportioning system that will allow for the injection of inorganic acid into the drinking water as a means of controlling microbial growth. Chlorine is not normally considered a suitable residual biocide with recirculating systems since it is difficult to control the concentration of free chlorine when the water is recirculating. However, routine testing for the presence of free chlorine or the use of chlorine monitoring and reporting equipment can adequately address the use of chlorine in the water delivery scenario. UV radiation also has a tendency to dissipate free chlorine concentrations.

Animal drinking water manifolds that are integral to animal holding racks or kennels are connected to the room distribution system by way of one or more detachable recoil hoses. These hoses allow for the holding racks or kennels to be detached from the system to facilitate room distribution system sanitizing without exposing animals to sanitizing agents, allowing for rack manifold and recoil hose sanitizing and for storing housing systems when no longer needed. If rack and kennel manifolds are designed to be flushed manually by facility staff, these manifolds are attached to the room distribution pipe using a single flexible recoil hose. Flushing of the rack and kennel manifolds is accomplished by manually opening a drain valve at the end of the manifold, allowing the water contained in the manifold to flow either to a room drain or to a bucket or catch pan that can be emptied into a local drain. Although manual flushing is effective if performed properly and on a daily basis, it relies on facility staff to accomplish this task. If facility personnel overlook this important maintenance step, there is increased risk of microbial contamination.

If rack or kennel manifolds are designed for automated rack flushing, or are part of a recirculating system that allows the water to flow through the rack manifolds during the water recirculating process, manifolds utilizing this design are connected to the room distribution system using two flexible recoil hoses. In rack flushing systems, one of the hoses will be attached to the water supply distribution line and one of the hoses will be attached to the water distribution drain line. There is an electrically controlled solenoid on the drain line associated with each rack or kennel manifold that can be sequenced to open during a flush cycle. The sequencing panel will also validate that the flushing cycle was completed. In recirculating flow-through-the-rack systems, one recoil hose is attached to the water supply line, and one recoil hose is attached to the water return line. Water will continuously flow through the room distribution piping, and each of the manifold pipes as the water makes its journey to and from the supply tank and through the UV-C reactor and microfiltration systems.

Manifold design is an important consideration when automated watering systems are purchased and utilized within a LARF. The design of the manifold has a direct bearing on animal drinking water quality and on the ability of the research animals to access drinking water ad libitum. A properly designed rack manifold will not only ensure that each cage on the rack has continuous access to the treated drinking water, but also ensure that air bubbles that become entrapped in the system are reduced and flushed to drain so that the research animals receive water from the drinking valves rather than air. A properly designed rack manifold will also ensure that all of the water in the manifold can be easily flushed and exchanged with clean treated water during the daily manual or automated flush sequences.

Daily flushing of rack and kennel manifolds is an important process for maintaining quality animal drinking water. At the manifold level, the animal drinking water is under low pressure, and the only exchange of the water is derived from animal drinking. The low and intermittent flow of water from animal drinking activity is not enough to thoroughly exchange the water in the manifold. If the water is not thoroughly exchanged on a daily basis, microorganisms will begin to proliferate and bacterial biofilms will begin to develop on the surfaces of the manifold and drinking valve components. To thoroughly and efficiently exchange the water in a manifold, the manifold must be designed as one continuous pipe from cage to cage and rack shelf to rack shelf (S configuration). To ensure that entrapped air is adequately removed during the flushing and filling of the manifold, a reverse S manifold design is preferred over an S configuration. With a reverse S–designed manifold, the water inlet flow begins with the horizontal pipe on the bottom of the rack and flows through the pipe on each row of cages as the water courses through to the last cage position on the top row of the rack. For automated systems that flush the manifold piping on each individual rack, reverse S manifolds are typically used.

There are other manifold designs that are used on rodent housing racks and on large animal enclosures. These other manifold designs may require manual flushing by the animal husbandry staff. Other common manifold designs are named after the letter shapes that they resemble. H manifolds, I manifolds, U manifolds, and upside-down U manifolds are just some of the more common designs that are utilized for different species. Animal husbandry staff should be aware that these manifolds may work equally well in distributing animal drinking water to the individual cages as the S and reverse S manifold designs; however, greater care and attention must be devoted to these manifold types to ensure that the animal drinking water is thoroughly exchanged on a daily basis and that entrapped air has been removed during the manifold filling and flushing process.

Most LARFs that utilize automated animal drinking water systems have dedicated rack manifold and recoil hose flushing stations. These stations are typically located within the clean side of the cage wash facility and also serve as a solution for flushing rack manifolds after sanitizing. Although these flush stations can be purchased as either manual or with microprocessor controls for an automated flush, they have the ability to flush rack manifolds and recoil hoses with a hyperchlorinated solution for greater sanitizing potential. The automated flush stations have flush and soak cycles to provide adequate contact time with the chlorine solution, and a final rinse cycle to rid the manifold of any residual chlorine solution. In addition to the flush, soak, and rinse cycles, the automated recoil hose flush stations have a compressed air cycle to dry the internal surfaces of the recoil hoses so that they can be either put back into immediate use or placed in storage with the quick-disconnect ends connected.

Daily flushing, or the continuous flow of animal drinking water through an automated watering system, is an integral component of maintaining an institution’s water quality standards. Periodic testing of the animal drinking water at the cage or enclosure level is considered a good practice to mitigate the risk of microbial contamination, and to ensure that the LARF operation is adhering to the institution’s water quality standards.

If test results indicate that bacteria are present in the room distribution system, sanitizing the system with a hyperchlorinated water solution may be indicated. A hyperchlorinated water solution of 20 ppm of free chlorine and 30 minutes of contact time is recommended by the manufacturers of these systems. Concentrations higher than 20 ppm and in combination with longer contact times may actually damage the system materials and should be avoided. Portable sanitizing units for mixing and pumping hyperchlorinated water solutions through the room distribution system are available from the manufacturers of automated watering systems. It is advisable to disconnect rack manifolds from the automated watering system while performing a hyperchlorinated sanitation process to avoid introducing a research variable and, moreover, to avoid potential animal health-related consequences as a result of an atypical exposure to a higher level of chlorine. Animals must be moved to other housing areas or provided with another source of treated animal drinking water during a hyperchlorinated sanitation process. After exposing the distribution system to the hyperchlorinated water solution for 30 minutes of contact time, the distribution system should be completely flushed with the treated drinking water supply to completely rid the system of the hyperchlorinated water solution.

One of the more valuable features of automated watering systems is the integration of monitoring capabilities into the system. The current technologies not only provide real-time information on system performance and status, but also, more importantly, provide a feedback system when the system integrity is compromised. Remote notification keeps key personnel informed of system breaches at all hours of the day, facilitating immediate response and remediation by responsible facility staff.

Automated watering systems typically require a higher initial capital investment over water bottle systems; however, automated systems will usually recover the initial capital within 3–5 years due to low operating costs and labor savings.

Disposable and Semidisposable Systems

Perhaps the most recent developments in water delivery solutions are disposable and semidisposable systems. They complement and provide a viable alternative to existing solutions for water delivery. The fundamental properties of these systems are that they provide a defined volume, provide for a variety of water treatment options, and can be used with most animal housing systems with the use of specialized adapters or for housing systems designed to use them. The principal theory of their use is substituting the capital and maintenance costs of the various processing equipment, labor costs, and operational costs for refined maintenance, operational, and labor costs.

Gels

Used by commercial breeders and animal exporting institutions for hydration during animal transportation, gels have the unique characteristic of being largely composed of water available in a semisolid form. When provided in sufficient quantity during transportation, studies suggest that animals remain well hydrated and experience weight gains when provided with both hydration gel and diet (Fredenburg et al. 2009; Pruet et al. 2010).

Although water is the major component, incorporation of nutrients in a balanced manner allows for meeting both the water and nutritional needs of the animal. This approach is particularly useful in circumstances where conventional means of delivering water is not effective or where animals cannot physically access the water source. In a balanced combination with other feed nutrients, nutritive gels are useful for those circumstances where both water and conventional diets do not provide for the special needs of the animal. Nutritive gels are highly palatable and offer a solution for transitioning weanling animals to standard diets and, more importantly, where traditional methods of transitioning offspring to a water source may not be effective.

Gels may also be considered for use in those circumstances where supportive care is indicated in a research protocol (Khaing et al. 2012). Gels may also provide a convenient medium for postoperative analgesic delivery (Christy et al. 2014) or for continuous delivery of medications and compounds (Overk et al. 2011).

Consideration of the use of gels as an adjunct hydration method or nutritive supplement should take into account not only the benefits derived from their use, but also whether they are compatible with the objectives and predicted outcomes of existing and proposed research protocols. Discussion and coordination with prospective researchers is therefore highly recommended.