Water Quality / Water Filter Primer

Joined
Apr 30, 2000
Messages
847
I was prompted by another current thread here on W&SS to share something that I have. About 12 years ago I participated in an online discussion (on another forum) with the water filter developer from MSR. We had some offline communication as well. I was persistent enough ion my questions that he shared some things with me that were not part of the online discussion. One was a full water filter test report. Another was the article I will post below, which I subsequently found on the Marathon Ceramics web site, the company who was making the ceramic elements for MSR.

Although some parts of this are clearly marketing, and this is dated (it does not discuss chlorine dioxide or UV lamps), it contains a lot of good discussion about water quality in general.

I'll post this in several parts since it is fairly long.

Portable Water Filters: a Designer’s Perspective
Dan Vorhis
version 7-30-97A

Most people know by now the advantages of carrying a portable water filter. Filters usually weigh less than carrying water and deliver drinking water quickly, with relatively little effort. And even if available surface water is cloudy, filtered water is usually pretty clear. Until about six years ago, you would find one or two "old-faithful" portable water filters to choose from on the shelf under the bike pumps. Now you go in a store and can’t miss a huge display with a dozen likely-looking devices sitting in a prominent place beside an aquarium from which to pump. The backpacker has benefited from all these new product introductions - filters are easier to use and less expensive than ever before. But the added selection and marketing pressure has left a lot of people bewildered or, worse yet, has created product gurus who would answer your questions based on a casually-informed article or on persistent marketing hype.

One thing different about water filters compared to, say, stoves or snowshoes - you can’t really tell if they work. Yes, you can see if they pump water easily. You can see if they seem to be well designed. You can read the reports about clogging rate, or even find out for yourself how fast they clog. But you can’t really tell if they are protecting your health adequately. It is so intimidating an issue that even the backpacking journals have avoided questioning the effectiveness of devices, choosing instead to parrot manufacturer’s claims on well laid-out charts.

This paper describes my perspective on recent issues related to portable water filters, based on my experience as a designer of water treatment devices for Mountain Safety Research, Inc. These observations are probably not entirely objective, or at least a competitor of MSR’s might make that case! You will probably notice a definite lean toward microfiltration here, a position which I think is justifiable.

What can make you sick?

Microorganisms.

It is estimated that about 1/3 of human illnesses in the world are caused by drinking contaminated water or eating contaminated food. According to recent reports, in the U.S., microorganisms (protozoa, bacteria, viruses) account for about 40% of waterborne illness, non-living or "chemical" contaminants account for about 10%, and the cause of the remaining 50% have not been identified. These statistics are gathered by the Center for Disease Control and Prevention (CDC), and others, from outbreaks mostly associated with community water treatment systems. Nobody does a very good job of tracking what makes backcountry travelers sick.

The "50% unknown" is interesting. Many of what are now known to be major waterborne pathogens (microorganisms that make you sick), such as Giardia, Cryptosporidia, and Campylobacter jejuni, were discovered to be human pathogens within the past 20 or 30 years. There are sure to be more "discovered" in the future.

Your friend may drink improperly-treated water and never show symptoms, while you drink the same water and are nauseous for weeks. Or the scenario might be reversed after your friend hasn’t gotten enough sleep for a week, and is under the gun at work. Our ability to resist infection from bad water varies with age, stress, what we’ve eaten for lunch. Children are often more susceptible than healthy adults, as are older people and people with weaker immune systems.

For example: In a Baltimore, Maryland, study, healthy young adult volunteers were fed various doses and strains of a bacterium called Campylobacter jejuni. Campy is a bacterium found worldwide, including in the backcountry. When volunteers were fed 800 bacteria, 10% became ill with diarrhea and fever, 50% showed evidence of infection with no symptoms. When the dose was increased to 90,000 organisms, 46% became ill, and 85% showed evidence of infection.

Bacteria. Bacterial symptoms, if they appear, come about 6 hours to 3 days after exposure. Symptoms include "explosive" diarrhea, vomiting, fever, headache, dizziness, weakness, sometimes bloody stools. For most healthy adults, bacterial infections are self-limiting, and symptoms disappear after 3 or 4 days, although a small percentage of people will remain ill for weeks. Often, immunity against re-infection is maintained as a sort of compensation for all the fun. For example, in the Campy study mentioned above, seven volunteers who had experienced illness, and twelve who hadn’t, were re-challenged one month later with 100,000,000 Campy bacteria. (Yummy. Actually, you wouldn’t see or taste this quantity of bacteria in the water.) None of the veterans showed symptoms, while six of the "controls" paid their dues. It can be seen why a grim harvest of young children (death by dehydration) takes place each year in places where high quality drinking water is unknown - the kids haven’t had a chance to build up any immunity.

Protozoa. Protozoan infections usually take longer to show symptoms - one week to several months. Giardia cysts "hatch" after a period of time in the small intestine and reproduce, interfering with adsorption of nutrients and, in some people, causing symptoms such as diarrhea, gas, cramps, weight loss, weakness, nausea. Symptoms of Giardiasis usually appear 1 to 8 weeks after ingestion of the cyst. Without treatment, Giardiasis infection will continue for two or three months in most people, and can permanently scar intestines.

The protozoa Cryptosporodia parvum may cause copious watery diarrhea, cramps, nausea, vomiting, low-grade fever usually 1 to 2 weeks after exposure. The illness lasts for 7 to 20 days in "immunologically healthy" adults. "Crypto" has been getting a lot of press lately following about a dozen recent worldwide outbreaks. The largest outbreak so far occurred in Milwaukee, Wisconsin in 1993 - an estimated 403,000 ill, around 100 deaths. In its cyst form, Crypto is incredibly resistance to chemicals like chlorine and iodine. In one study, Crypto oocysts (the environmentally resistant form of Crypto) were able to infect mice after soaking 24 hours in straight household bleach! This organism’s resistance to iodine is one of the reasons why the U.S. Marines have stopped relying on iodine tablets for water treatment. Crypto oocysts are smaller than Giardia cysts (see Figure 1). They can break through sand filtration at water treatment plants and make people sick even when the plants are working as they should. Infectious dose (how many organisms it takes to make you sick) is thought to be quite low, some say as low as one oocyst, for this organism. There is no known cure for Cryptosporidiosis - once you’ve got it, you have to suffer through it.

In a sampling of 257 water samples from 17 states (Rose, Gerba, Jakubowski 1991), Crypto oocysts were detected in 55% of surface water samples, with an average concentration of 43 oocysts/100 liters. Giardia cysts were found in 16% of the same samples, with an average concentration of 3 cysts/100 liters. Pristine waters seem to be less affected than lowland waters exposed to urban and agricultural run-off. Young calves are a major reservoir - the CDC estimates that 90% of dairy farms are infected with the cyst.

The USEPA is currently overseeing an even more thorough investigation of the microbiological state of surface water sources used by communities for water treatment. Out of this study will come more reliable information about the concentration of viruses, bacteria and protozoa in these waters. There have been questions about the reliability of previous studies because of examination methods used. The current exhaustive effort should deliver the most accurate picture yet.

Viruses. Waterborne viruses that make people sick usually originate from human fecal material. This differs from bacterial and protozoan pathogens, many of which cross species freely, and are therefore carried in the backcountry not only by humans, but by animals including small rodents and deer. Viruses do not reproduce in surface water like some bacteria. Most virus particles "die" (there is some question about whether they were actually alive in the first place) within days of exposure to the environment. Nevertheless, an individual may become infected after ingesting only a few virus particles. Anywhere there are infected humans and the chance of human fecal contamination, there is the chance of viral infection.

Illness from waterborne viruses takes many forms. Hepatitis A, spread through fecally contaminated food or water, is a highly contagious liver disease with symptoms that include yellow skin and eyes, dark-colored urine, flu-like symptoms (fever, chills, weakness), stomach pains. Young children often show no symptoms, and about half of adults who become infected show no symptoms. Illness from Hepatitis A, if it occurs, appears two to six weeks after exposure, and may require 6 months for complete recovery. Note that there are other Hepatitis (inflaming the liver) viruses like Hepatitis B and C, spread through contact with bodily fluids, not through drinking water.

Rotavirus is another common waterborne virus, with symptoms more similar to bacterial infections; vomiting, watery diarrhea, low-grade fever. Rotavirus infection is one of the most frequent causes of severe dehydrating diarrhea in children. Spread of the disease occurs not only through drinking fecally-contaminated water, but also through contact with utensils and objects that come into contact with feces.

Note that few water treatment facilities in the world routinely check for viruses in their water supply. Some cities - along the Ohio and Missouri rivers, for example - rely on water that has been "previously used" by citizens upstream. During low water times, a glass of water might quench more than one thirst on its way to the ocean. Viruses have been detected in community water systems along these routes. Another "home use" water source, well water, has also come under scrutiny. Apparently, household wells are sometimes located too close to septic systems, and viruses can find their way into drinking water more or less easily, depending on soil conditions.

"Chemical" contaminants.

This class of contaminant is often conveniently ignored in the portable device industry because most waterborne illness is caused by microorganisms, and illness caused by chemicals rarely causes acute symptoms, making tracking difficult. This class of contaminants can also be a problem for portable devices to deal with - a few grams of granular carbon doesn’t do it. What kinds of chemicals can make you sick?

Pesticides are usually associated with urban or agricultural run-off. This can mean aerial sprays in backcountry timber stands. Concentrations of various pesticides in surface water go up in the early growing season as a result of higher application rates and more rains.

Bacterial Toxins are produced by what used to be called blue-green algae, now called cyano-bacteria, which bloom in warm water and form greenish soup. Other toxin-forming organisms are thought to occur on snow ("watermelon snow"). Filtering out the algae does not remove the toxin in the water. Such toxins have been known to kill large animals within minutes. Adequate contact with quality activated carbon has been shown to greatly reduce concentrations of bacterial toxins. Inorganic contaminants like mercury and arsenic don’t just enter surface water through industrial run-off. In the big, dry interior basins in western U.S., for example, low rainfall allows accumulation of metals and minerals in valley surface water. Drainage from mine tailings can also contribute. Some of these contaminants are tough to remove even with a block of high quality activated carbon. Halogens such as iodine and chlorine, used in water treatment, are toxins in sufficient concentration, or create toxic by-products when they react with natural organic materials in surface water.

Some common misconceptions about the causes of waterborne illness.

"Giardia is mostly what I have to worry about in the backcountry, right?"

It surprises me how many people who should know better - microbiologists at the health department, for example - continue to spread this tale. Cryptosporidia is obviously another backcountry threat. The tiny bacteria Campylobacter jejuni is also a concern. A study in the early 1980s found 23% of patients showing up at a clinic near Grand Teton National Park were infected with Campy (compared to 8% infected with Giardia). Last summer (1996), another such clinic reported seeing about two Campy-caused illnesses for every one illness caused by Giardia. There are certainly other bacteria and protozoa in the backcountry, carried by animals and humans and transported into surface water from feces, that cause human illness.

"Viruses are the major health concern outside the U.S."

Food and water borne disease are the number one cause of illness in travelers, according to the Center For Disease Control and Prevention (CDC), and "traveler’s diarrhea" is the most frequent health problem. According to the CDC, "…Although viruses are commonly acquired by travelers, they do not appear to be frequent causes of Traveler’s Diarrhea in adults." Bacterial infections from contaminated food or water, especially from toxic strains of Eschericia coli, Salmonella species, Shigella species, and Campylobacter jejuni, cause most of the diarrhea.
 
"Viruses are not a concern in the U.S."

Because waterborne viruses that cause illness in humans are spread only in human feces, not in animal feces (unlike bacteria and protozoa, as mentioned above), countries where sewage treatment is commonplace have less worry about viruses in surface water. However, viruses have been detected in run-off from urban areas, especially after periods of high rainfall. "Pristine" waters in backcountry camping areas, where no sewage treatment facilities exist, are at risk more or less, depending on the human pressure. The trouble is, you can’t tell by looking at the water what happens to be hidden under a pile of leaves upstream.

Portable Treatment Options

Boiling. The most effective way to kill microorganisms in water is to boil the water. Boiling doesn’t make muddy water clear, of course, and metals or other non-volatile contaminants in water will remain after boiling. To kill all known waterborne pathogens, the CDC recommends bringing water to a rolling boil for one minute or, at altitude above about 6000 feet, bring to a rolling boil for 3 minutes. The kills occur as a result of the temperature and the time at that temperature, so don’t cheat. Let it boil and cool before drinking.

Iodine tablets will protect against bacteria and viruses if manufacturer’s instructions regarding water temperature and turbidity (cloudiness) are followed. Two iodine tablets in one liter of water were shown to kill almost 3 log (99.8%) of Giardia cysts in a lab brew of cold, somewhat turbid water. However, Cryptosporidia oocysts are not killed under realistic contact times and exposures to the iodine.

Iodinated Resins Iodinated resins look like tiny black beads, about the size of fine sand. Microorganisms in the water are killed in two ways - organisms collide with beads, and iodine comes off the beads into water and kills the bugs as they soak in the iodine solution. Given adequate contact time, iodinated resins are very effective against bacteria and viruses. Because protozoan cysts are resistant to iodine, resins must be used in combination with a filter to insure protection against cysts like Cryptosporidia.

A water filter that pushes water through the iodine beads too fast, especially in cold water, or a device that removes the iodine too soon after treatment (with activated carbon, for example), can allow pathogens through. Stuff in water like silt or organic molecules are an additional load on the iodine - the iodine can’t tell the difference between a soil mineral particle and an organism. Organic materials and silt in water can also "blind" the beads over time if they are inadequately protected by filtration and activated carbon upstream.

Microorganisms in surface water tend to clump together, or clump with the fecal material from whence they came. Organisms in the center of the clump are shielded from the iodine. Clumping (also called "agglomeration") is an enemy of chemical treatments, but a friend of filtration technologies, because filters have an easier time removing these clumps.

The EPA recommends that devices which add iodine to water be used for short periods of time only - 2 to 3 weeks max., according to the late Ruth Douglas at the USEPA. Pregnant women, hyperthyroid women (about 1% of women in the U.S.) and others with thyroid problems should avoid drinking iodinated water. These recommendations are based, in part, on tests which indicate that certain individuals can develop sensitivity to iodine after prolonged exposure even to relatively dilute concentrations.

Filtration. Looking through a microscope, most water filter media resembles a sponge more than a window screen. Clogging particles - mostly soil mineral particles - are either stopped near the surface of the filter or migrate deep into the media and stop there.

If particles like bacteria are stopped simply because they are too big to fit through a hole, we call that "sieving". When a particle or molecule sticks to the filter media like a ball bearing sticks to a magnet, that is called "adsorption". All filter media rely on some combination of sieving and adsorption to remove stuff from water (see Figure 3). Some filters (ceramic filters, for example) are heavy on the sieving, others (carbon block filters) are heavy on the adsorption.

Ceramic filters.

Ceramic filter media is made from diatomaceous earth that has been heated to a point where it just begins to melt ("sintering"). Quality cartridges made from sintered diatomite catch most particles within about .005 inch of the surface. When the ceramic filter clogs, the clogged pores are abraded and flushed away, and the cycle begins anew. Flow rate recovers almost 100% after each abrasive cleaning. The entire depth of the filter media may be effectively and systematically used, 5 thousandths of an inch at a time. A gauge supplied with the ceramic cartridge lets the user know when it is time to replace the cartridge.

Ceramic cartridges last longer than other filter media of similar overall size, although they usually require more frequent cleaning. For example, one pleated fiber cartridge has a surface area of about 130 square inches, a filter wall thickness of less than .015 inch, and an average pore size of 3-5 microns or so. This filter will produce high flow rates and require very little maintenance for many liters, and then it will clog. A ceramic cartridge, in comparison, begins with only 19 square inches of surface area, a filter wall thickness of .250 inch, and an average pore size of about 1.8 microns. The ceramic will clog more frequently, but is capable of being cleaned at least 30 times, so effective surface area is

30 X 19 square inches = over 500 square inches.

Ceramic filters don’t reach the end of life without warning - a gauge clues the user when to replace the cartridge. When the ceramic in the example above is used up, it still has about .150 inch of filter media left - at least 10 times more than the pleated filter media. According to our tests, these factors of wall thickness and pore size, coupled with pressure, relate directly to microorganism retention capability.

Ceramic cartridges are more fragile than the plastic or glass fiber pleat packs. Care should be taken especially when the ceramic cartridge is out of its housing during cleaning or drying.

Micron ratings. Remember the old filter controversy… "nominal" vs. "absolute? Certain companies tended to blur the distinction. "Nominal" means about 98% (1 log) of a certain size particle are retained. "Absolute means "100%" of a certain size particle is retained. The difference between "nominal" and "absolute" is significant in some surface waters, where 500,000 to 5,000,000 or more pathogens per liter are possible. "98% retention" of 500,000 still leaves 10,000 bacteria per liter. In any case, "nominal vs. Absolute" is yesterday’s controversy. Today, manufacturers avoid use of the word "nominal". Now, you have to ask for a definition of "absolute"!

Occasionally I get a call from somebody making a "water filter comparison" chart for a store or magazine. One question that eventually gets asked is, "What is the micron rating on your filter?" To answer this, I painfully ask the following questions:

• Are you referring to nominal or absolute retention?
• How do you define absolute? Is it 2 log (99%), 4 log (99.99%), 7 log (as per Health Industry Manufacturer’s Association (HIMA) definition for pharmaceutical filter media?
• Are you testing at maximum flow rate? Minimum wall thickness? Bacterial challenge or solid particle challenge? How are you feeding the bacteria? Has the media been exposed to all the gunk in surface water to use up adsorptive capacity?

Of course, the person on the phone trying to put together a simple water filter comparison chart is usually put off with this line of questioning. When deciphered, it implies that an accurate comparison is impossible unless all the devices are tested the same way. "Can’t you please just give me a number?", I will hear. At least one questioner resorted to phone hopping around MSR until he finally located somebody who would give a "simple" (albeit meaningless) micron rating on our device.

Here is one definition of "absolute". Let’s say we want to show that a window screen keeps out 100% of house flies - that is, that the window screen is an "absolute filter" for house flies. In order to prove this, we would have to test with an infinitely large number of house flies which is, of course, messy. Organizations have come up with tests to deal with the "infinitely large" problem. In the pharmaceutical industry, when they want to insure that absolutely no bacteria get into something, a filter media is tested according to a HIMA (Health Industry Manufacturer’s Association) standard. The standard specifies that a filter must be challenged with a high concentration of a certain .3 micron bacterium (Brevundimonas diminuta). Even what these bacteria are fed is specified in the HIMA test method, because people had figured out a way to fatten them up and make a filter look better than it was! If the filter media retains greater than 99.99999% (seven log) of the organism, it is called a "sterilizing", or ".2 micron absolute" filter. "Seven log" removes 1000 times more organisms than "4 log" (99.99%), and 10,000 times more organisms than "2 log" (99%). A "micron rating" comparison is useless unless all devices are tested identically.

Some common misconceptions about filters.

"A .2 micron filter protects my health better than a .3 micron filter."

Based on the discussion above, you might see why a device advertised as ".3 micron" might protect your health better than one advertised as ".2 micron":

• Some manufacturers base their claims on testing done with test dust or latex beads. Others use actual bacteria to rate their filter. Bacteria can squeeze through a hole half their diameter, so the test using bacteria is tougher.
• Some manufacturers do their tests when the filter is brand new, others do it after the filter has been used a lot in the field. Factors that come into play here: wall thickness and adsorptive capacity. A filter that has been abraded away is less effective. A filter whose adsorptive capacity has been exhausted is not as effective.
• Some manufacturers run water through their device at low pressures, so bacteria don’t blow through. Others test at maximum pressure to mimic a realistic, worst case.

In our last field comparison of portable filters, for example, one filter listed as a ".3 micron" filter on the magazine charts removed about 170 times more bacteria than a filter claiming to be a ".2 micron" device. Such a discrepancy has been more the rule than the exception in our comparison tests.

"A filter should have a pre-filter".

Pre-filters seem to make sense at first glance. Here is a closer look:

• Most clogging occurs when particles in the water that are very close to, or smaller than, the pore size of a filter media. Usually, particles that are much larger than the media don’t restrict flow much. So, for example, a bowling ball would not clog a window screen. Many pre-filters currently sold are of relatively large pore size. These pre-filters will clog, but they are clogging with particles that are larger than the particles that would clog the main filter media. So, now you have two filters that are clogging, and the first filter isn’t necessarily making the second one last longer.
• There is one pre-filter currently offered which does have a pore size that is fine enough to make a difference. If you find yourself with a device that clogs quickly, this accessory is one answer. However, as with any set of "in-line" filters, you now have two clogging filters to deal with. And replacing pre-filters that clog quickly can be a nuisance, and more expensive than alternative technologies.
• The concept of a pre-filter has some other weaknesses. When a filter is placed downstream from the pump, as in most portable devices, water is pushed through the filter. A pre-filter is typically placed upstream from the pump, which means water is sucked through the filter. A phenomenon called "cavitation" (gases getting sucked out of the water in the form of bubbles that
interfere with pumping) occurs at a vacuum of only about 8-10 psi. Not much pressure to work with.
• In a partial vacuum, hoses have to be beefy or they collapse as the pre-filter clogs.

Adsorption is a phenomenon of small particles or molecules wherein the particle sticks to the surface of a filter medium, like our example of a ball bearing sticking to a magnet. Bacteria, viruses, chemicals, organic molecules that add color or taste to water if they come into contact with the surface of the filter media and the water doesn’t hold tight, they take up residence in the filter.
 
Activated carbon is very good at grabbing things (like hydrocarbons, pesticides, taste/odor compounds, some metals, iodine) out of water. This is not because carbon has any special surface chemistry, but more because it has a lot of surface area - 850 to 1400 square meters per gram, or about one football field per teaspoonful! If the pesticide molecule, viral particle, or taste/ odor compound gets close to the carbon surface, it jumps out of the water and grabs onto the carbon. Other, non-carbon filters have a lot less surface area. Bacteria will adsorb onto a non-carbon filter (see Figure 3), but there is very little removal of chemicals like taste and odor compounds and pesticides from water. Carbon or no carbon, when the surface of the "magnet" is filled, incoming "ball bearings" don’t have any place to rest, and they go on through the media.

"Granular" carbon (little loose grains of carbon) is the inexpensive form, and permits much of the water to flow around the carbon, through the channels between the grains. "Block" carbon (looks like a black porous brick) eliminates the large channels, forcing the water into closer proximity to the carbon. As a result, "block" carbon is much more effective than granular carbon at removing the things carbon removes.

Activated Carbon that has seen a lot of use one day may recover a great deal of its adsorptive capacity if given a chance to "rest" for a few hours. The adsorbed molecules will, over time, vibrate deep into the recesses of the carbon structure and re-expose adsorptive surface to passing water.

Colonization. Bacteria will grow in any damp media, including activated carbon, iodinated resin beds and silver-impregnated media. While most studies have failed to find disease-causing bacteria colonizing damp filter cartridges, this is certainly a possibility. Drying the cartridge thoroughly between uses will reduce bacterial growth. Where that is impossible (on the trail for a week or more), a device may give you the option of boiling the cartridge to kill the bacteria growing within. Most cannot be boiled.

How long before it clogs?

This is the most frequently asked question by backpackers looking for a filter, for a very good reason - people need to know if they can afford to use the thing. Unfortunately, most filters will not deliver as much water as people have been led to expect. MSR has paid for over a dozen comprehensive field water clogging trials over the past six years, filtering from a variety of water sources. Most devices have never come close to delivering as many liters as advertised - usually, they have been off by a factor of about ten or more. Total filter element life (life to replacement, following manufacturer’s instructions for cleaning, etc.) for many popular devices is 20-100 liters in water with a visibility of 1 to 4 feet. This means that many people will be buying a $20 to $40 replacement cartridge after filtering less than 15 gallons of water. There are a few notable exceptions. Ceramic filters provide the longest life, according to our tests. For details why, see the section under "Filtration".

Water filters are an interesting category. I think a many folks use this gear less than they would like. It may take two or three years for the vast majority of filter owners to put 50 liters through their device. So, in the presence of inspired marketing and in the absence of good field trials, a mediocre product may develop quite a following in two or three years. After that, results start to come in. By contrast, a camping stove which fails to deliver doesn’t make it far out of the starting gate.

Gradually, MSR’s "field water comparisons" have become more sophisticated. For example, several years ago we began challenging portable devices with a microorganism after the clogging test. The purpose of the organism challenge is not so much to determine if a device gives adequate protection against a pathogen - you would need to use the actual pathogen for that. But it is a way to compare the devices relative to each other. The organism we chose is a bacterium called Brevundimonas diminuta, the HIMA bacterium mentioned. B. diminuta is a tough challenge for a filter because it is so tiny - we don’t know how resistant to iodine it is compared to other bacteria. The results of both the clogging comparison, what we term "cost per liter", and the bacterial challenge have been startling.

Figure 4 is a summary of the results of our most recent comparison, performed early 1997. The water we chose this year was fairly cloudy - visibility of about one foot. It wouldn’t have been the greatest water to filter from if clearer water were available, but certainly wasn’t as bad as most mid-western streams and rivers. I have taken the liberty of distinguishing with a gray-colored block those devices that are currently being marketed as purifiers to accentuate a point - "purifier" status does not necessarily guarantee superior performance.

Figure 4. "Use Cost" and Relative Bacterial Retention Capability of 10 different portable water treatment devices in one test conducted by MSR, Inc. (January 1997)


Cost per liter filtered, based on replacement element cost.
(U.S. currency) Lab challenge results: Average concentration of living, non-pathogenic bacteria that got through a device.
(challenge water: B. diminuta @ 4.8  108/ Liter, average)
(red indicates device is marketed as a "purifier")
Device No. 1 $0.37 about 11 bacteria/liter
Device No. 2 $0.60 about 1,600,000 bacteria/liter
Device No. 3 $0.60 about 7,000,000 bacteria/liter
Device No. 4 $3.09 about 1,300 bacteria/liter
Device No. 5 $3.28 about 11,800 bacteria/liter
Device No. 6 $0.75 about 10,800,000 bacteria/liter
MSR WaterWorks II $0.65 about 12 bacteria/liter
MSR MiniWorks $0.34 about 90,000 bacteria/liter
Device No. 9 $0.45 about 15,600,000 bacteria/liter
Device No. 10 $1.92 about 73 bacteria/liter

There are many details left out of Figure 4. For example, one device broke during the microbiological challenge, which, in a perverse sort of way, had the effect of making that device look better in the organism challenge (less challenge water was run through it). A couple of devices were not tested at the end of their useful life, making them, perhaps, look better than they would have otherwise. All devices were pumped at normal flow rate - no double passes, etc. Notice that two of the top three devices in terms of bacteria removal/kill were not purifiers, yet all three devices which cost more than $1.00 per liter were purifiers.

"Purifiers"

The U.S. Environmental Protection Agency (USEPA) proposed a "purifier" standard in the mid-1980’s to help regulate portable water treatment devices. The standard has not yet been officially adopted due, in part, to variability in the test method. Usually, a standard is written so any lab can follow the testing guidelines and come up with similar results. If the standard is unclear, results can come in all over the place. Such has been the case with the "purifier" protocol. But that isn’t the only problem. Lack of rigor in the standard - loopholes - has permitted devices to enter the market as "purifiers" that would not pass a more stringent standard.

Some requirements in the current version of the "purifier" standard are tough. For example, 4 log (99.99%) reduction of viruses is required - not an easy task for most technologies. However, other areas are less well defined. For example, health-related filter test standards typically require that a media be tested at its maximum flow rate. Multiple passes of the challenge water through the filter would be totally unacceptable, and the device which failed on the first pass would fail the test. The current USEPA "purifier" standard does not specifically state that multiple passes of challenge water are not permitted. Another example: Certain iodinated-resin "purifiers" are sold with a carbon accessory to remove the iodine and make the water more palatable. But this accessory reduces the device’s ability to kill bacteria and viruses. One manufacturer has decided to test their devices with the carbon accessory removed, and then sell them with the carbon accessory. The EPA has not stepped in.

Some common misconceptions about "purifiers":

"A filter is different from a purifier."

This is a false distinction. According to the section 1.3.1 of the "Guide Standard and Protocol for Testing Microbiological Water Purifiers", revision 1987, "…A review of treatment units that might be considered as microbiological purifiers discloses a number of different types covering treatment technologies ranging from filtration and chemical disinfection to ultraviolet light radiation." There is no requirement for chemical treatment in the standard.

"A purifier kills viruses, a filter doesn’t remove viruses."

Not necessarily true. High quality microfiltration has been shown to remove 2 to 3 log (99 to 99.9%) of viruses on a single pass. This is not enough for the USEPA "purifier" - 99.99% of specified viruses is necessary. However, no iodinated resin-based portable water treatment device passes the USEPA "purifier" standard without pumping the water through the device twice and/or reducing the flow rate to a fraction of the normal flow rate of the device. The jury is out over which technology is more effective against viruses in field water.

"A purifier protects my health better than a non-purifier."

It would make things simpler, but it isn’t necessarily so. Some companies, as an ethical stance, refuse to test their devices using multiple passes of a challenge water and refuse to test using slower flow rates than used in the field. And so their filters face a self-imposed hurdle that is very much higher than is required by the current version of the "purifier" standard. In addition, the "purifier" standard addresses strictly microbiological risks - it doesn’t deal with pesticides, industrial chemicals, mercury, bacterial toxins. Most "purifiers" currently sold come ill-equipped to deal with this class of contaminant.

Care and feeding of one "purifier"

A quick read, recently, of the instructions in one device marketed as a "purifier" turned up some noteworthy tidbits. Toward the back of the instruction booklet was a little note describing exactly what steps had been necessary for this particular device to pass the "purifier" microbiological reduction requirements.

• Remove the carbon cartridge that is provided with the device to remove iodine taste.
• Pump water at a rate of no faster than about 6 minutes per liter into a separate container.
• Re-filter that liter of water again, at the same flow rate (total time per liter: over 12 minutes).
• Wait 20 minutes before drinking the water. (total time: 32 minutes)
• Don’t filter more than 2 liters at a time. Wait 2 hours between pumping sessions so the resin beads can re-charge.

Does anyone do this? These instructions are couched in such a way to make it seem that such remarkable advice need rarely, if ever, be followed. Meanwhile, the consumer may be tempted to believe they are getting superior health protection while pumping at the approximately 40 seconds per liter that the device can deliver, rather than 32+ minutes per liter necessary to get the protection defined under the conditions of the standard.

Standards-writing process could use your help
Water treatment can be complicated. It is not fair to expect people, who may be expert in their own field, to also be an expert in the field of water treatment before making an informed buying decision. That is what good standards are for.

ASTM (American Society for Testing and Materials) is an (almost) 100 year-old, not-for-profit organization that helps manufacturers, buyers and sellers write good standards such as standard test methods. By requiring input from everybody involved in, for example, the portable water treatment industry - manufacturers, sellers and users of the equipment, it is believed that all needs will be met and a solid standard will result. It is very difficult, through the ASTM process, for one powerful interest to exhibit undue influence over the outcome.

The negative side of a good standard is that it takes a long time to create. I do not expect to see an ASTM standard covering health and health-related aspects of portable water treatment devices before the year 2000.
 
Back
Top