Significance of Bacteriologic Plate Counts

Significance of Bacteriologic Plate Counts

Microbiologists have used aerobic plate counts to demonstrate the presence and number of bacteria in foods and beverages, including water, for more than a century. The procedure is reliable and accurate when applied to nutrient-rich substrates such as foods and medical specimens, but organisms living in potable water are usually starved or actually dormant and are unable to respond to the wealth of nutrients suddenly provided.

Less than 1% of the bacteria known to be present and viable in water samples are able to grow into countable colonies within the 5-7 days of incubation provided by the Heterotrophic Plate Count (HPC) procedure. Furthermore, very few of those that do grow are able to grow a second time in the secondary cultures designed to identify them. That means the number of colonies counted is always at least 99% low, and their identities are totally unknown—the procedure is so terribly inaccurate and uninformative that it is almost not worth doing.

However, experience over the years has shown that HPC values remain about the same in well-run water systems, so that similar numbers (within a power of ten) obtained from regular samplings indicate a healthy status quo. This is the true and only value of HPC data for drinking water: as a “general quality control” indicator of a well-run system.

Other plate count procedures are designed to imitate the internal body environment and support the growth of sewage organisms or specific pathogens preferentially. Such bacteria are usually quite delicate and die off quickly after leaving the rich, warm environment of our intestines, and detection of even one or two survivors per 100 mL sample has great sanitary significance. Every positive Total Coliform plate count must be confirmed and retested for Fecal Coliforms or E. coli bacteria; repeated positive results trigger public notification of an unsafe condition advising citizens to boil their water before drinking it.

Although ordinary aerobic plate count procedures such as the HPC are not intended to indicate health hazards and have no sanitary or health significance, they are unfortunately often misused in that way. Over the years, two specific aerobic plate count or HPC values have acquired inappropriate health significance in different regions. In Europe, an aerobic plate count of 100 colony-forming-units (CFU) per mL was associated with the effluent from slow sand filters that had been optimized for hydraulic factors.

That number has no intrinsic health significance, but it was officially recommended (never required) as the maximum plate count for drinking water for many years in several European nations, the European Union, and the World Health Organization. That recommendation has recently been recognized as defective and has been officially removed from the EU and WHO regulations, but many people in positions of authority still cling to the old rules.

In addition, several nations, including many around the “Pacific Rim,” originally misinterpreted the intent of the old rules and made the 100 CFU/mL maximum a requirement instead of just a recommendation. In the U.S. and Canada, an HPC (formerly SPC) of 500 CFU/mL or less was associated with water samples in which the detection of coliform bacteria was most sensitive. Like 100 CFU/mL, that number also has no intrinsic health significance, but it never achieved the status of an officially recommended maximum plate count in North America. Nevertheless, both 100 and 500 CFU/mL have long been improperly used to imply or suggest that filtered ingredient waters with plate counts exceeding those values are unsanitary or unsafe.

That is just not true. In fact, there is considerable evidence showing exactly the opposite: the bacteria comprising the aerobic plate count or HPC in disinfected drinking water or coming from filters producing ingredient water are entirely harmless. Actually, they are even better than harmless: they are actively beneficial in reducing, even eliminating the health hazard from any pathogens that may occasionally escape disinfection. The more HPC organisms, the better!

There are three lines of evidence supporting that contention. The first is simply the complete absence of any evidence to the contrary: there is not even a single confirmed case of an outbreak of human enteric disease caused by bacteria growing in supplemental water treatment products, anywhere in the world, since such products first appeared about a century ago. Of course, that is negative data, which is scientifically inadmissible, but it is very significant and impressive, nonetheless.

A second line of evidence comes from laboratory studies in which scientists have intentionally inoculated water filters with pathogens and coliforms. Those studies showed, in every case, that the number of bacteria from the plume of added contamination was reduced by a factor of 100 or greater by the filters, to below infectious levels. The naturally-occurring water system bacteria are much more vigorous and hardy than enteric pathogens from sewage, and such organisms are readily overwhelmed by their growth.

This has been demonstrated with scanning electron microscopy. Depending on how new and fresh the filter cartridges were at the time of the contamination, the media might or might not sustain a long-term colonization: fresh, virgin, uncolonized media permitted the inoculated organisms to establish a tentative foot-hold, but the colony always failed within a few weeks or months. Filters that were already colonized
by the harmless, naturally-occurring water system organisms resisted colonization by the test bacteria.

The worst performance by a water filter in a test conducted at the EPA labs in Cincinnati was the reduction of millions of Pseudomonas aeruginosa bacteria inoculated into the influent to only 8 or 9 CFU/mL, immediately after the contamination. That filter continued to produce P. aeruginosa counts of 1 or 2 CFU/100 mL for some months afterward, but those numbers do not constitute a health hazard.

The third line of evidence that the growth of HPC bacteria in water treatment products is actively beneficial to the health of people using them comes from long-term, whole-community epidemiological studies. Like the laboratory studies mentioned above, they are unanimous: in every case the health of families using supplemental treatment was equal to or better than that of the control families that used regular tap water, in spite of the fact that, in every case, the level of HPC bacteria in the specially-treated water was 10 to 100 times higher than in the regular tap water. There is a clear and convincing correlation between the highest plate counts in the specially-filtered water and the lowest incidence of gastrointestinal illness.

Plate count samples have value for assessing the overall quality of the treatment and the physical integrity of the system only when taken as part of a test program, in which the same trained people collect samples regularly from the same sites, in the same way, at the same time of day. But individual “grab samples” taken here and there are so variable that the results have almost no meaning at all. This is especially true of water treatment unit influent-effluent sample pairs, of which only the influent was thoroughly flushed before being sampled. Naturally, the plate count will be higher in water coming from a treatment unit or water-using appliance than in well-flushed samples from the main, but the actual number has very little meaning, and no sanitary significance at all.

Since tap waters known to be safe regularly show Standard Plate Counts (SPCs) of 30,000-40,000 CFU/mL, which are equivalent to more than 500,000 CFU/mL by the newer HPC procedure, it is evident that the “magic numbers” of 100 or 500 CFU/mL have no value at all in identifying “safe” conditions. It is standard practice in the well-informed national laboratories to consider HPC sample averages in controlled tests statistically equivalent if they are within one log of each other. That is, a group of filter influent plate counts averaging 300 CFU/mL and their corresponding effluents averaging 2990 CFU/mL would not be considered different.

If the flushed water samples from the mains in a particular system have always fit into a 12-month running geometric mean of, say, 30 CFU/mL, except for one sample with an HPC of 15,000 CFU/mL, then an investigation should be launched to find an explanation for the deviation. The surprise count of 15,000 does not signify an unsanitary or unsafe condition—only a positive Fecal Coliform or E. coli test would do that—just an unexpected result that demands an explanation. If investigation showed that the sample was not a properly flushed sample from the main but a carbon filter or water softener effluent, then the deviation would be explained, and that should be the end of it.

The same is true of random positive Total Coliform samples, even if confirmed, but not associated with a positive Fecal Coliform or E. coli test result in the same sample or from the same site. Environmental coliforms that have nothing whatever to do with sewage are very common and hardy, and they have confused the health issue for generations. The only purpose of the Total Coliform test is as a preliminary screen for possible fecal contamination.

It used to take extra time, effort, and money to run both a Total Coliform and Fecal Coliform analysis, but no longer. The 100 mL Presence-Absence test using “defined substrate media” containing the metabolic additives “MMO” and “MUG” gives both results simultaneously in 24 hours; it’s fully approved by all public health agencies; it’s totally insensitive to HPC overgrowth; and it actually costs less than doing it the old way.

So there is no reason to settle for confusion and incomplete information any more. Bacteriological samples just for Total Coliform analysis should no longer even be collected, and any results from them should not be given any health significance. Any positive results should be investigated, just like an unusual HPC level, but then ignored if no pattern is evident or there are no fecal coliform organisms.

This is not an appeal to develop “bacteria-breeder” products for their beneficial health effects. Rather, it is an appeal, both to our customers and to public health officials everywhere, to recognize and accept the normal regrowth of survivors of disinfection in supplemental water treatment equipment as a beneficial, protective force of Nature. The biofilms flourishing on the media and other surfaces in water treatment products are entirely analogous to the schmutzdecke (that developed on those early sand filters that produced cholera-free water with aerobic plate counts around 100 CFU/mL.

Of course, most of today’s tap water is well treated and entirely free of pathogens, susceptible only to the occasional influx of untreated water through cracks in the mains when the pressure fluctuates. But that does happen, and often, in many systems. Usually it involves only very small numbers of pathogens diluted by great volumes of water, which is no health hazard. But once in a while a bigger plume of contamination may come along, so that a few of the most susceptible individuals receive an infectious dose of something.

This is where the health value of point-of-use/point-of-entry products and their biofilms of harmless, naturally-occurring organisms is greatest. When there has been a significant incident of contamination, often it is two days or more before the condition is recognized, confirmed by testing, and publicized. During that time everybody will have been consuming contaminated water, but those businesses and homes using supplemental water treatment equipment will benefit from some unknown (and probably unquantifiable) level of protection against waterborne enteric illness due to the presence of an active biofilm on the filter media.

This protective effect of HPC growths is not something special achieved only by special products with unusual designs and exotic media, but by even the simplest granular activated carbon bed filters, sand filters, and water softeners. The overall level of protection is magnified if the water treatment product is also a fine-filter capable of removing colloidal particles such as parasitic cysts and/or contains an effective
adsorbent for virus particles and other contaminants.


Regarding the plate count procedure:

Reasoner, D. J. and E. E. Geldreich. 1985. A New Medium for the Enumeration and Subculture of Bacteria from Potable Water. Appl. Environ. Microbiol. 49:1, pp. 1-7.

Staley, James T. 1985. Enumeration and Identification of Heterotrophic Bacteria from Drinking Water. EPA/600/S2-85/061.

Byrd, Jeffrey J., et al. 1991. Viable but Nonculturable Bacteria in Drinking Water. Appl. Environ. Microbiol. 57:3, pp. 875-878.

Regarding filter inoculation studies:

Rollinger, Y. and W. Dott.1987. Survival of Selected Bacterial Species in Sterilized Activated Carbon Filters and Biological Activated Carbon Filters. Appl. Environ. Microbiol. 53:4: pp 778-781.

Camper, A. K., et al. 1985. Growth and Persistence of Pathogens on Granular Activated Carbon Filters. Appl. Environ. Microbiol. 50:6, pp 1378-82.

Geldreich, E.E., et al. 1985. Bacterial Colonization of Point-Of-Use Water Treatment Devices. J. AWWA 77:2, pp. 72-80.Reasoner, D. J., et al. 1987. Microbial Characteristics of Third-Faucet, Point-of-Use Devices. J. AWWA 79:10, pp. 60-66.

Regarding epidemiological studies:

Calderon, R. G. and E. W. Mood. 1988. Bacteria Colonizing Point-Of-Use, Granular Activated Carbon Filters and Their Relationship to Human Health. Final Report, CR-811904-01-0. U.S. EPA.

Calderon, R. G. 1991. Bacteria Colonizing Point-Of-Entry, Granular Activated Carbon Filters and Their Relationship to Human Health. Final Report, CR-813978-01-0.

Payment, P., et al. 1991. A Randomized Trial to Evaluate the Risk of Gastro-intestinal Disease Due to Consumption of Drinking Water meeting Current Microbiological Standards. Am. J. Pub. Health 81:6, pp. 703-708.

Payment, P., et al. 1997. A Prospective Epidemiological Study of Gastro-intestinal Health Effects Due to the Consumption of Drinking Water. Internat. J. Environ. Health Res. 7:1, pp. 5-31.

Regarding European Union and World Health Organization Standards:

Lightfoot, Nigel. 1998. Domestic Softened Water and Health. British Water Review Report, 23 November, 1998

World Health Organization. 1999. Guidelines for Drinking Water Quality, ISBN 92 4 154480 5, Vol. 2.

Regarding high plate counts in the distribution system:

Becker, R. J. 1975. Bacterial Regrowth Within the Distribution System. Proceedings, AWWAWQTC, Session 2B-4.

Perry, D. L., et al. 1981. Development of Basic Data and Knowledge Regarding Organic Removal Capabilities of Commercially Available Home Water Treatment Units Utilizing Activated Carbon. Phase 3/Final Report. EPA No. 68-01-4766.

Regarding the Total Coliform test:

Payment, P. 2000. Coliforms as Indicators: After a Century of Good Service, Is It Time To Retire?
Proceedings, ASM 100th General Meeting, Special Session: “Coliforms to Fluorochromes: A Century of Water Quality Assessment.”

Percival, S. L., et al. 2000. Microbiological Aspects of Biofilms and Drinking Water. CRC Press, Boca Raton, pp. 115-116.

This Technical Bulletin is courtesy of Pentair Everpure, Inc.