By Duane Dunk & Gregg McCarthy

Many of us have seen neighborhood kids zipping around at 25 mph on electric-powered scooters (hopefully wearing safety helmets). But none of us would expect that the same tiny scooter motor could power a one and one-half ton automobile down the street at the same 25 mph.

Undersizing a motor for a given task is a sure way to produce disappointing results. It is no different with designing POE or POU water treatment devices.

The importance of sizing a device appropriately for its intended purpose is critical. And there are also many additional parameters to consider other than just sizing when properly engineering a device.

These functions include prefiltration, water quality, efficacy, flow rate, post treatment time, post treatment and user interface with the treatment device. All of these are interrelated and no single function can be altered without affecting the others (see Figure 1).

Water quality
The severity of the water conditions to be treated is a major factor to be considered in treatment device design. Will the device effectively treat worst-case challenge water scenarios or is it to be marketed to a more narrow market with predictably less severe contamination?

There is certainly going to be a significant cost impact in designing a fail-safe product for worst-case environments. Alternately, one might select a modular approach with a base product design to address the majority of the market’s needs, offering add-on accessory modules for more extreme water quality conditions.

Temperature can be a particularly important factor, especially when assessing halogens for disinfection, which have a slower reactivity in colder water and an increasing volatility at higher temperature. With iodine resins in particular, higher temperature can significantly increase the amount of iodine leached into the water. This will generate a need for more downstream scavenging and polishing media.

Remember, everything is interrelated.

Based upon the concentration and variability of harmful organisms in the incoming water, the water quality you intend to disinfect will drive how much efficacy is needed and the other functions of flow rate, disinfection medium sizing, prefiltration and post treatment dwell or residence times will all influence if/how that efficacy is achieved.

Prefiltration can take many forms, including sediment filters, membranes, activated carbon, ion exchange resin or other specialty media for a variety of specific contaminants.

The need for prefiltration or removal of certain contaminants prior to disinfection is primarily driven by two factors.

Will prior reduction of certain contaminants enable the remaining treatment device or ‘train’ to operate more efficiently and/or for longer periods of time with greater capacity? High turbidity or TDS, for example, can lead to downstream scaling and blinding-over of subsequent treatment media, rendering it less efficient and shortening its useful life. Organics left in the water will compete with pathogens for consumption of disinfection potency, whether using halogens, ozone or UV, not to mention the potential of formation of disinfection byproducts.

Will the downstream treatment system be able to adequately address residual contaminants encountered in the feed water? For example, if arsenic or fluoride is present, without appropriate adsorptive capability in a pre-filter, one cannot expect a device with only an ultrafiltration membrane to reduce levels of these harmful contaminants.

Multiple options are likely available for any given prefiltration requirement. Should you use carbon or a specialty ion exchange resin? For carbon, should it be granular or block, coconut or bituminous? From a purely scientific approach, one could simply choose the media that offers the highest percentage removal of targeted contaminants.

However, selection criteria will be affected by the other parameters to follow, such as pressure drop, desired performance, regulatory requirements, capacity and price. Everything is interrelated and no single portion of a water treatment system operates in isolation from the other design considerations.

If you skimp on the prefiltration up front, the system will be taxed more downstream, thus requiring more disinfection medium, more post treatment time and slower flow rates to achieve the same level of efficacy. One must take an integrated treatment approach to product design.

The term efficacy designates the performance (i.e. reduction of targeted organisms) over a given period of time (the product life of the device), including an adequate safety margin.

How much bacteria and virus reduction is required? If you are operating in the US (in California for example, as well as many other states) and want to make a bacteria claim, you will also have to provide virus reduction. The US EPA Guide Standard and Protocol for Microbiological Purifiers and NSF P-231 require a six-log bacteria and four-log virus reduction under harsh challenge conditions. This relates directly to US EPA regulation of the word ‘purify,’ along with its three-log cyst reduction requirement.

What if you are selling overseas in a market without specific antimicrobial requirements? As little as a one-log (90 percent) reduction of bacteria has been shown to have a positive impact on the incidence of diarrheal disease.

Brazil requires a two-log (99 percent) reduction of E. coli. In the absence of specific regulations, many industry players (such as in India and China) look for a three-log (99.9 percent) reduction. How is this to be measured? By periodically spiking otherwise benign tap water with challenge organisms or by passing a continuous three-log challenge through the system? Standard accepted protocols do not exist in all markets, so do not overlook establishing how the device will be tested to demonstrate achievement of desired product water quality.

Figure 1. Example Disinfection System Design Parameters

Again, everything is interrelated. And if you want greater efficacy, either higher performance or longer duration, the disinfecting medium volume must be increased, post treatment time extended or prefiltration methods optimized.

Flow rate
Flow rate and use pattern are critically important for how the device will be tested and how it will actually be used. Beware of accelerated testing in order to obtain an answer more quickly, as the answer may not be meaningful in comparison to real-world use patterns. Most accelerated tests are designed to be followed by real-world use conditions.

Consumers do not typically just open the faucet and filter 4,000 liters (1,057 gallons) all at once, nor do they drink 50 to 100 liters (13.2 to 26.4 gallons) per day in their home. Will the device sit idle with long periods of dormancy such as in a vacation home? Know the NSF, WQA or local government test criteria for certification requirements but of equal importance, spend time in the field understanding use patterns that may affect design considerations.

Many treatment processes are time dependent, either in terms of contact time with the media or in terms of dwell time downstream of halogen disinfection, in order to allow adequate time for completion of the oxidation process for disinfecting the water.
Obviously, higher flow rates shorten the available time for treatment processes and a system certainly should not be designed and tested at a slower flow rate than it is likely to experience in the hands of actual users.

Post treatment time
Dwell time downstream of treatment media is particularly important with disinfection systems like halogenated resins and ozone. The clock starts ticking once the water comes into contact with the disinfecting media and continues to run until either the water is consumed or the water passes through scavenging or neutralizing media to halt the disinfection process. Bacteria being attacked by chlorine, bromine or iodine can literally be rescued and remain viable if rushed too quickly into a lifesaving bed of GAC.

So, how much dwell time is needed? It depends. Probably the biggest factor is dosage. Higher concentrations or dosages of disinfecting media require less time for micro organism destruct, while lower dosages require more time.

The classic example of this factor is when an iodine resin manufacturer or supplier says just a few cubic inches of their media will treat 37,854.11 liters (10,000 gallons) of water. However, that is provided post treatment storage of sufficient size is engineered to allow for a greatly extended dwell time, as the leach rate from the resin declines to lower and lower dosage over its lifetime.

By comparison, some organisms like polio virus are notoriously more resistant to halogens than E. coli. Water temperature can be a huge factor as well. The level of microbial contamination expected to be present and the required performance efficacy must be factored into the equation as well.

Remember, everything is interrelated.

Post treatment
What is the required taste or odor profile of the final product water? Does your newly designed, compact and affordable water treatment device have to replicate the bottled water quality results generated by equipment costing tens of thousands of dollars?

Does your treatment process cause bad taste or odor in the product water, perhaps with excessive residual halogen that must be scavenged? Are problematic levels of heterotrophic bacteria present? Has something potentially harmful, such as odorless, tasteless iodide residual, been created during treatment?

Purifying drinking water for health and safety is only half the battle if the water has to be stored for later use. Protecting it from recontamination by bacteria colonization can be just as important. This necessitates at least a low-level residual chemical presence for lasting benefit, preferably below offensive taste and odor thresholds if not below human detectability.

User interface/acceptability
About now, you may be thinking that you have worked through the preceding list of considerations and have come up with an ideal combination of treatment technologies for your new ‘killer’ drinking water disinfection device (pun intended). But wait a minute; let’s not forget the consumer or end-user.

Is the device going to be user-friendly and not overly complex to operate and maintain? Did it become over-engineered and too costly for the target market, though it can perform flawlessly? Costs need to be considered from both acquisition and operating viewpoints. Does the device meet or exceed the consumers’ minimum required output?

Even the over-simplified chart in Figure 1 lists 20 variables that are not simply present or absent but come in gradations. Thus, there is a near infinite array of possible combinations, making it exceedingly difficult, if not impossible, to generate an algorithm to automatically provide reliable answers for all your water treatment device design questions. Such design decisions must therefore be ultimately resolved via prototyping, validation and optimization testing in properly structured experiments to generate reliable empirical data.

The next time you are asked, “How much of your media will be required to treat water?” remember your first answer should always be, “It depends.” Then help your customer to understand that you are not being evasive.

You simply want to be accurate and avoid being too quick to provide what may turn out to be an inaccurate answer (high or low), once the rest of the design considerations are fully understood and tested.

A final reminder: everything is interrelated.

About the authors
Duane Dunk is Global Business Unit Director for HaloPure drinking water disinfection markets at HaloSource, Inc. in Bothell, WA. He has over 15 years of POU industry experience. Dunk may be contacted at Gregg McCarthy is a product development engineer at HaloSource, Inc. A degreed chemical engineer, he has been working in water and wastewater treatment for seven years. He currently focuses on joint development with partner companies of a wide variety of HaloPure contact biocide POU disinfection systems, where he constantly has to respond, “It depends.” He may be contacted at


Comments are closed.