Syndicate content

How to test water quality? Low-cost, low-tech options for microbial testing

Pratibha Mistry's picture
A multi-compartment bag with a colorimetric reagent for E. coli testing
using a most probable number approach.

Photo credit: National Snow and Ice Data Center

Water quality is central to the challenge of ensuring safe water for all. Here we present the third entry in a three-part blog series on low-cost, low-tech water quality testing. In previous posts, we discussed options for measuring physical and chemical aspects of water quality. In this final post, we explore low-cost, low-tech options for microbial testing.

Microbial testing is arguably the most important factor when monitoring the quality of drinking water. Without noticing anything strange, a single glass of cool, refreshing water could contain billions of microorganisms. Just a few dozen of these organisms could be enough to make you very sick.[1] According to estimates from the World Health Organization, diarrheal disease from contaminated drinking water causes more than half a million deaths annually.[2]

The holy grail

In a recent presentation, Dr. Mark Sobsey, of the University of North Carolina at Chapel Hill, outlined the ideal characteristics of a microbial test for water quality monitoring on a limited budget:

It should be portable, low-skill, self-contained, lab-free, and electricity-free. It should be available globally at a cost of less than $0.10 (USD) per test, and it should be easy to interface with data reporting and communications technologies. It should also be integrated into education programs to mobilize stakeholders. To this we might add that (semi)quantitative results should be available quickly, without a 12 (24, 48, etc.) hour incubation period.

This “holy grail” test does not yet exist.

Waterborne viruses, protozoa, and helminths (worms) can all be harmful to human health. However, testing for a large number of different pathogens is infeasible. A more practical solution is to test for an “indicator organism,” which acts as a sign of fecal contamination. According to the WHO Guidelines for Drinking Water Quality, Escherichia coli (commonly abbreviated as E. coli) is the preferred indicator organism.[3] The target concentration is 0 organisms per 100 mL sample.

The substitutes

There are a large number of microbial testing options available today, but they are not perfect. Selective media and enzyme-based assays that detect E. coli’s beta-glucuronidase enzyme are often well-suited for low-resource settings.[4] Overall, these microbial water quality tests take one of three approaches:
  • Presence-absence (P-A): P-A tests don’t provide quantitative information about microbial water quality. Instead, they change color to tell you whether or not microbial contamination has been detected. Test kits are comparatively inexpensive, but often involve adding a powdered nutrient mixture and allowing a 24-hour incubation period for organisms to grow. P-A tests are suitable for screening in situations where microbial contamination is not expected (e.g., deep groundwater).
  • Most probable number (MPN): MPN tests are semi-quantitative. Several samples of the same water are tested in tubes, plastic bags, or small plastic plates with multiple “wells.” The user adds a nutrient solution (“culture media”) and waits 12-48 hours for organisms to grow before counting the number of positive samples, indicated by a color change. The user then converts that number of positives to a statistical estimate of bacterial concentration, as per the instructions for the particular test.
  • Membrane filtration: Membrane-based tests are the most quantitatively accurate. In general, a 100 mL water sample is forced or vacuumed through a small, round filter paper (the membrane) using a little hand pump. All the bacteria in the sample are caught on the filter as the water passes through. The filter is then incubated with some sort of culture media. Each bacterium caught on the filter will multiply into a little colony. After the incubation, the user counts the colonies – possibly with the aid of a magnifying glass – to determine how many “colony-forming-units” were present in the original 100 mL sample. Due to the filtration step, membrane-based tests are more difficult when water samples contain a lot of suspended material, and they can take a bit of time.
When selecting a test, it is helpful to consider not only the quantification needs and the cost, but also factors like the sample volume, the format and stability of the culture media (e.g., powder, liquid, agar, films, absorbent pads), the incubation time, and the ease of reading results. Incubation temperature is also an important practical consideration: Some tests require a warm incubator, while others can handle ambient (warm climate) temperatures. A few small-volume tests are even designed to be incubated with the user’s own body heat (e.g., in your pocket)!

As previously mentioned, the “holy grail” of microbial testing does not yet exist, and all of the currently-available testing options have drawbacks. Fortunately, researchers are working on the problem from a number of angles, including immunogenic, chemical, and molecular approaches. These efforts to advance low-cost, low-tech microbial testing are essential. The harmful – even deadly – effects of waterborne disease are felt across the developing world, particularly by the most vulnerable people. Advances in microbial testing will make it easier to meet the challenge of safe water for all.

Thank you for following along in our water quality testing series!

Additional resources

[1] The WHO defines highly-infective microorganisms as those for which an “infective dose” of just 1-100 organisms are enough to make you sick. Viruses, protozoa, and helminths are usually highly infective. Bacteria are usually less infective, which is to say that you have to ingest more of them (thousands, perhaps millions) before you get sick.
[2] In contrast, naturally-occurring chemical contaminants are usually present at levels that become problematic only after chronic exposure. And water affected by acutely dangerous levels of (human-caused) chemical contamination is likely to be unappealing.
[3]The topic of indicator organisms is an area of active research. More discussion is available in the WHO Guidelines document, and in the CAWST manual in the links section. Older testing methods have used thermotolerant coliforms (TTC) or total coliforms (TC) as indicators of fecal contamination.
[4] For detailed discussion and comparison of specific microbial testing products, please refer to the resources links at the end of this post.


Submitted by Mariame on

appreciate that experiences/tools from other institutions were included. grounded comments! thanks.

Submitted by Joep Appels on

Please send comparison of specific microbial testing products, maybe we can help as we develop these kind of tests.

Submitted by Lisa Hirsh on

However one defines the "holy grail" of microbial water quality tests, we respectfully submit the Aquagenx Compartment Bag Test hit the marks described in this blog post. Our CBT E. coli Kit was developed by Dt. Mark Sobsey. It is widely used throughout the world for the following reasons:

+ MPN test results scored by easy color-match or smart phone apps, Akvo Caddisfly and mWater Surveyor (both apps automatically calculate CBT test results , create and geocode surveys, and share test results in real-time)
+ Easy for anyone to use with little training
+ Easy to carry and transport in the field, compact and lightweight
+ Ambient temperature incubation at 25° Celsius and above
+ No electricity, labs, cold chain, expensive equipment, specialized technicians required
+ Works at variable temperatures, constant temperature control in incubator not required
+ Test results in 24-48 hours depending on ambient temperature
+ Meets World Health Organization (WHO) Guidelines for Drinking Water Quality
+ Available globally, under $10 per test pricing available

Please visit our website and contact us for more information!

Add new comment