Water Quality Failures in Microbiology

By Dr. Zubair Khalid, DVM, MS, PhD · 2026-11-10

A water quality failure in a microbiology laboratory almost always starts upstream of the bench. In most investigations the first real finding is a degraded pretreatment pack, a fouled reverse osmosis membrane, or an exhausted polishing cartridge that has been passing ions, chlorine, or organics into a distribution loop that cannot compensate. The first thing to check is the point-of-use resistivity or conductivity reading against its own trend line, together with the maintenance log for the last cartridge, filter, and UV lamp change. A displayed 18.2 MΩ·cm proves very little if the conductivity cell has drifted, if the temperature compensation is misconfigured, or if the water has been sitting stagnant in a loop for three weeks.

Effective microbiology water quality failure troubleshooting is a sequence, not a guess. It moves from the written specification, to the measurement, to the point of use, to the treatment train, and finally to the feed water. Each step removes a class of causes rather than a single suspect. A laboratory that follows that order will resolve most failures within a single working day and, more importantly, will be able to document why the failure happened. That documentation is what keeps a quality management system credible when an auditor, a client, or a clinician asks whether a result produced during the incident is trustworthy.

This guide covers the practical microbiology of laboratory water: what the grades mean, which failures are common, which are rare, how each failure produces a bench-visible effect, and how to verify a repair before patient or client samples are run again.

What Counts as a Water Quality Failure in Microbiology

Purified water is a reagent in almost every microbiology workflow. It reconstitutes dehydrated media, dilutes stains, carries antibiotics into susceptibility plates, feeds molecular master mixes, rinses glassware, and supplies automated instruments. A water quality failure is any deviation between the water actually delivered at the point of use and the specification the laboratory has committed to, whether or not a test result has yet been affected. A failure can be chemical, microbiological, physical, or operational.

• Chemical failure. Dissolved ions, total organic carbon (TOC, the carbon content of organic molecules in the water, reported in parts per billion), endotoxin, dissolved gases, or residual disinfectants exceed limits.
• Microbiological failure. Planktonic bacteria, biofilm, or the lipopolysaccharide fragments called endotoxin exceed limits. Endotoxin is measured in endotoxin units per milliliter (EU/mL).
• Physical failure. Particulates, resin fines, or gas bubbles enter the product water.
• Operational failure. The correct water is available but the wrong grade is used for a task, or water is stored and dispensed outside its validated conditions.

Resistivity is the resistance of water to electrical current, expressed in megohm-centimeters (MΩ·cm) at 25 °C. It rises as ions are removed, so it is the most common continuous indicator of ionic purity. Conductivity is its reciprocal, expressed in microsiemens per centimeter (µS/cm), and some systems display it instead. A bacterial count is reported as colony-forming units per milliliter (CFU/mL) after incubation on a low-nutrient medium.

The grade definitions used across clinical and veterinary microbiology laboratories trace to CLSI document GP40 and to ISO 3696. Laboratories should always follow their own standard operating procedure, but the commonly used targets are summarized below. These are specifications, not universal legal limits.

Two points about this table cause most of the confusion seen in laboratory investigations. First, resistivity and TOC measure different things. Water can hold 18.2 MΩ·cm and still carry a biofilm downstream of the sensor, because the biofilm sits on the tubing wall and the cells are only intermittently shed into the sample. Second, Type II water stored in an open or vented container absorbs atmospheric carbon dioxide, which forms carbonic acid and drives pH toward roughly 5.5 to 6.0 while resistivity stays acceptable. Neither reading alone is a complete quality check.

The Causes, From Most Common to Rarest

1. Feed Water and Pretreatment Problems

Feed water is the raw municipal supply entering the purification system. Pretreatment, usually a sediment filter, carbon block, and water softener, protects the reverse osmosis (RO) membrane. The RO membrane is a semipermeable barrier that rejects dissolved salts and most organics under pressure.

The single most common cascade is this: the carbon block exhausts, chlorine or chloramine passes through, and the thin-film composite RO membrane oxidizes. Oxidized membranes lose salt rejection while continuing to produce water. The polishing cartridge downstream then works harder, exhausts early, and the loop begins to show falling resistivity and rising TOC. The laboratory replaces the cartridge, sees a temporary recovery, and the cycle repeats.

Confirm the cause by measuring feed conductivity and testing for free chlorine with a DPD colorimetric test at the carbon outlet, not just at the tap. Record the RO percent rejection (the difference between feed and permeate conductivity divided by feed conductivity, multiplied by 100). A rejection value that has fallen from the high 90s to the low 90s points to membrane compromise. Also review the pretreatment change log. Carbon blocks are rated on chlorine throughput, not calendar time alone, so a facility that increased water use can exhaust one early.

2. Exhausted Ion-Exchange Cartridges

Mixed-bed ion-exchange resin swaps cations and anions in the water for hydrogen and hydroxide ions. When capacity is spent, ions break through into the product stream. Divalent ions, silica, and carbon dioxide tend to break through before the resistivity reading collapses, so resistivity is a late indicator. Confirm exhaustion by comparing inlet and outlet conductivity across the cartridge and by reviewing the rate of change. A cartridge that falls from 18.2 to 12 MΩ·cm in a week is exhausted, whereas a cartridge that drops only after a long holiday shutdown may simply need flushing.

3. Biofilm in the Loop, Tank, and Point-of-Use Filters

Biofilm is a structured community of microorganisms attached to a surface and embedded in a self-produced matrix of extracellular polymeric substances. It is the classic hidden cause of a microbiological failure with perfect resistivity. Stagnation, dead legs (pipe runs with no flow at the end), oversized storage tanks, and warm room-temperature loops all favor attachment. Once established, biofilm sheds planktonic cells and endotoxin into the water at a rate that no UV lamp downstream can fully control.

Confirmation requires sampling at more than one point. Take a sample from the dispensing valve, a swab from the inside of the valve spout, a swab from the point-of-use filter housing, and a sample from the tank if one exists. Plate each on R2A agar, a low-nutrient medium designed for heterotrophic water bacteria, and incubate at 20 to 28 °C for five to seven days. Growth that is present at the valve but absent at the tank inlet localizes the problem to the loop. A simultaneous rise in TOC with stable resistivity is a strong hint that organic material from the biofilm is present.

4. UV Lamp and Ultrafilter Aging

UV lamps at 254 nm injure and inactivate microorganisms, and lamps at 185 nm generate hydroxyl radicals that oxidize organics and lower TOC. Output falls gradually long before total failure. Ultrafilters remove particles, pyrogens, and microorganisms by size exclusion and can become colonized on the upstream side if sanitization is neglected. Confirm both by reviewing hours-in-service logs rather than waiting for an alarm. Manufacturer intervals are usually expressed in hours of operation or in months, and the shorter of the two governs.

5. Storage and Dispensing Errors

Water held overnight in a carboy, a graduated cylinder, or an uncovered flask is no longer the water that left the system. Carbon dioxide dissolves, bacteria multiply, and plasticizers can leach from new tubing. Common specific mistakes include refilling a carboy without sanitizing it, storing Type I water in a container previously used for tap water, dispensing into a beaker that was rinsed but not dried, and drawing water from a still that has been idle and has no recirculation.

6. RO Membrane Scaling and Temperature Effects

Hardness scale, silica, and high feed total dissolved solids reduce permeate flow and raise feed pressure. Water temperature matters too, because RO membranes are rated at 25 °C and produce substantially less permeate at 10 °C. A system that suddenly fails in January in a cold municipal supply may be suffering from temperature, not from a broken component. Confirm by checking normalized permeate flow and pressure drop against baseline, not by absolute gauge readings alone.

7. Sensor Calibration and Monitoring Failures

A meaningful fraction of reported failures are instrument artifacts. Conductivity cells foul with biofilm or scale, temperature compensation coefficients get set incorrectly, TOC analyzers drift when their UV reactor lamp weakens, and handheld meters go out of calibration. Confirm by measuring the same sample on an independent calibrated meter and by checking the cell against a certified reference solution, for example a 100 µS/cm potassium chloride standard. Pay attention to temperature compensation, because resistivity specifications only apply at 25 °C.

8. Autoclave and Water Bath Cross-Contamination

Water baths are well-known reservoirs of Pseudomonas species, Aeromonas species, and nontuberculous mycobacteria. Splashes and aerosols reach open plates, slants, and staining trays. Autoclaves fed with hard water scale their chambers and jackets, and contaminated house water can carry organisms into the chamber, particularly on gravity-displacement cycles with poor drying. Confirm by culturing bath water and by swabbing bath walls and gaskets.

9. Rarer Causes

Chlorine or chloramine breakthrough into the product water is uncommon but serious, because it will inactivate reagents and injure cell cultures before any plate count rises. Organic leachates from new tubing or from a new plastic reservoir can raise TOC for weeks after installation. Particulate shedding from a failing filter housing, silica breakthrough past an exhausted anion bed, and endotoxin spikes from a gram-negative biofilm are all seen in practice but are far less frequent than the causes listed above. Gas supersaturation and microbubble formation after a pump or pressure change can interfere with optical instruments.

Troubleshooting Table

Use this table as the core of the investigation. Start at the top and work down, because the upper rows account for most incidents.

A Diagnostic Sequence for Microbiology Water Quality Failure Troubleshooting

Follow these steps in order and write down each result. The written record is the difference between a repair and a recurring problem.

1. Read the specification. Confirm what grade the affected procedure actually requires by checking the standard operating procedure, not by memory. Molecular methods, cell culture, and susceptibility media usually demand a higher grade than glassware rinsing.
2. Verify the measurement before you verify the water. Measure the same sample on a second, calibrated instrument. Check the reference solution and the temperature compensation setting. Eliminate the sensor as a cause first.
3. Sample at multiple points. Draw from the point of use, from the loop return, and from the tank or system outlet. Compare. A failure confined to the valve points to the valve or the point-of-use filter. A failure present at the system outlet points upstream.
4. Run the full panel. Resistivity or conductivity, TOC, heterotrophic plate count on R2A incubated at 20 to 28 °C for five to seven days, and endotoxin if the application is endotoxin-sensitive. Do not release the system on resistivity alone.
5. Review the maintenance log. Extract the dates and hours for the last cartridge, sediment and carbon filter, UV lamp, ultrafilter, and RO membrane changes, and the last sanitization. Compare against the manufacturer's interval.
6. Interrogate the feed water. Record feed conductivity, feed temperature, and free chlorine at the carbon outlet. Ask the facility whether the municipal supply changed, whether plumbing work occurred, or whether water use increased.
7. Calculate RO performance. Compute percent rejection and normalized permeate flow. A falling rejection with stable feed quality is membrane compromise until proven otherwise.
8. Inspect physically. Open the point-of-use filter housing and look for discoloration or slime. Swab the valve spout and the inside of the tank. Smell the water. These cheap checks resolve a large share of cases.
9. Identify the mechanism. Decide whether the failure is chemical (ions, TOC, endotoxin), microbiological (planktonic cells, biofilm), physical (particulates), or operational (wrong grade, poor storage). The category dictates the fix.
10. Correct, then verify. Repair the cause, sanitize or replace as needed, flush to the manufacturer's specification, and repeat the full panel before returning the water to service.
11. Assess impact on results. Identify every test performed with out-of-specification water and decide which results need review, repeat, or a qualified comment.

How Water Failures Surface at the Bench

The bench effects differ by discipline, and recognizing the pattern shortens the investigation.

In media preparation, ion contamination can precipitate phosphates and calcium during autoclaving, producing a cloudy or granular appearance. High TOC can depress growth of fastidious organisms. Dissolved carbon dioxide lowers pH, and a medium that is 0.3 pH units off can shift the performance of selective agars appreciably.

In staining, hard rinse water leaves mineral deposits that appear as crystals under the oil objective, and residual ions can interfere with decolorization. A Gram stain that looks washed out across multiple operators is worth checking against the rinse water before blaming technique.

In antimicrobial susceptibility testing, the divalent cation content of the water used to prepare cation-adjusted Mueller-Hinton broth matters directly, because calcium and magnesium affect aminoglycoside and tetracycline activity. A slow drift in MICs for quality control strains is a classic early signal of a water problem, not a resistance problem.

In molecular diagnostics, residual organics, metal ions, and nucleases inhibit amplification. The first visible sign is often a failing no-template control or a rising cycle threshold across all samples in a run. Autoclaved water that has sat in a bottle is not equivalent to freshly dispensed Type I water for this purpose.

In cell culture and virus isolation, endotoxin and trace organics reduce plating efficiency and can produce morphologic change that mimics cytopathic effect. In parasitology, the water used for floatation solutions and wet mounts can introduce organisms or alter specific gravity if the grade is wrong.

Corrective Actions That Hold

Replacing a cartridge fixes a symptom. These actions fix the cause.

Sanitize the whole wetted path, not just the tank. Use a disinfectant compatible with the system materials and with the manufacturer's instructions, and follow the contact time. Peracetic acid and dilute hypochlorite are common choices. Rinse thoroughly afterward and verify the absence of residual disinfectant before returning the loop to service. Residual disinfectant will damage media and cell cultures more reliably than the biofilm did.

Eliminate dead legs and stagnation. Reconfigure the loop so that every branch has flow. Remove unused valve branches and capped tees. If a tank is oversized for the daily draw, either add recirculation or reduce tank size.

Control temperature and storage time. Keep stored water cold if the system is designed for it, keep containers closed, and set a maximum holding time in the standard operating procedure. Draw fresh water for critical work rather than using what has been sitting.

Replace consumables on condition and on interval. Track hours for UV lamps and liters processed for cartridges and pretreatment. Trend the data so replacement happens before failure rather than after.

Fix feed water before replacing polish. If rejection is poor or chlorine is breaking through, a new polishing cartridge will fail in weeks. Treat the upstream problem first.

Qualify replacement components. New tubing, new tanks, and new filter housings can leach organics. Flush them per the manufacturer's instructions and check TOC before use.

Safety note. Peracetic acid, hypochlorite, and other sanitizers are corrosive and can release irritating vapors. Wear splash goggles, gloves, and a lab coat, work in a ventilated area, and never mix hypochlorite with acids. UV lamps used for water treatment are a serious eye and skin hazard if the housing is opened while energized. RO systems operate at high pressure and should be depressurized before any fitting is disturbed.

Verifying the Repair Before Sample Testing Resumes

Do not release the system on a single good resistivity reading. A defensible verification includes all of the following.

1. Resistivity or conductivity at or above the specification for the grade, measured at the point of use with a calibrated instrument at 25 °C.
2. TOC below the specification for the grade.
3. Heterotrophic plate count on R2A incubated at 20 to 28 °C for five to seven days, below the laboratory's limit. Include a sample from the valve and a swab from the valve spout.
4. Endotoxin below the specification when the application requires it.
5. A negative control in the actual application. For media, prepare a small lot with the repaired water and check sterility and growth promotion with the appropriate control organisms. For molecular work, run a full extraction and amplification negative control. For cell culture, run a plating check.
6. Absence of residual sanitizer, verified by the method the manufacturer specifies.

Record all six. A verification record that lists only resistivity is the finding most likely to be cited in a laboratory accreditation assessment.

Prevention Through Laboratory Quality Management

Microbiology quality control for water is a monitoring program, not a single annual test. A workable program has five elements.

A written specification per use. Define which grade each procedure requires and record it in the standard operating procedure. This prevents ad hoc substitution when the Type I dispenser is down.

Scheduled monitoring with trend review. Check resistivity or conductivity daily if the system displays it, and record it. Measure TOC, plate counts, and endotoxin on a defined interval tied to the grade and the application. Plot the values. A trend line catches a failing UV lamp or a developing biofilm weeks before an alarm does.

Change control. Any change to the system, the plumbing, the feed water source, or a consumable supplier triggers a verification. The most common preventable failure in well-run laboratories is an unrecorded plumbing repair that introduced debris or a change in municipal disinfection practice.

Documented corrective and preventive action. When a failure occurs, record the root cause, the corrective action, the verification results, and the impact assessment on affected test results. This is the core of laboratory quality management and the part auditors examine most closely.

Training and competency. Everyone who draws water for critical work should know which grade applies, how to dispense without introducing contamination, and how to report an alarm. Untrained relief staff drawing water from the wrong tap is a recurring cause of failures that no maintenance program can prevent.

Common Mistakes and Limitations

Trusting 18.2 MΩ·cm as a complete quality statement. Resistivity is an ionic measurement. It says nothing about bacteria, endotoxin, or TOC. A loop with a healthy biofilm and a functioning ion-exchange bed will display a perfect number while dispensing contaminated water.

Sampling only at the dispenser. Biofilm is a surface phenomenon. A grab sample from a flowing valve under-represents the sessile population. Swabs of the valve spout, filter housing, and tank wall are often more informative than the water sample itself.

Replacing the cartridge without fixing the feed. This is the most expensive recurring error in water system maintenance. If the RO rejection is poor or chlorine is passing the carbon block, the new cartridge will fail again.

Ignoring temperature compensation. Resistivity specifications are stated at 25 °C. A meter compensating to a different reference, or a sample measured cold, will not match the specification.

Holding water in open containers. Carbon dioxide uptake changes pH, and airborne organisms enter. Water drawn for critical work should be used promptly.

Assuming autoclaving sterilizes and purifies. Autoclaving kills organisms but does not remove endotoxin, ions, or organics. Autoclaved water stored in a bottle can grow organisms after the cycle if the container is not protected.

Skipping the application-level check. A water sample can pass every chemical and microbiological specification and still fail in a specific assay. Growth promotion testing, negative controls, and QC strains are the final arbiters.

Limitations. Water system design, sanitization chemistry, and replacement intervals are manufacturer-specific and must follow the system documentation and the laboratory's accreditation requirements. Individual incidents need local root cause analysis and, where patient or client results may be affected, review by the laboratory director or the attending veterinarian. This guide describes general mechanisms and standard practice, not a substitute for the written procedures that govern a specific laboratory.

Frequently Asked Questions

How often should microbiology laboratory water be tested?

Follow the interval in your laboratory's standard operating procedure, which is usually set from the grade required and the manufacturer's recommendations. Resistivity or conductivity is checked daily when the system displays it, while TOC, heterotrophic plate counts, and endotoxin are checked on a longer defined cycle and after any repair or change.

Why can resistivity read 18.2 MΩ·cm when bacterial counts are high?

Because resistivity measures ions, not organisms. Bacteria and biofilm contribute little to conductivity, so an ion-exchange bed can keep the display at specification while a biofilm in the loop, valve, or point-of-use filter sheds cells into the water. Only a plate count and a TOC trend will reveal this pattern.

Can deionized water be used instead of Type I water for media?

Only if the procedure and the grade specification allow it. Deionization removes ions but does not reliably remove organics, bacteria, or endotoxin, and a service deionized supply can vary day to day. Critical media, molecular reagents, and cell culture work require the grade named in the procedure.

What causes a sudden drop in resistivity?

The usual causes are an exhausted ion-exchange cartridge, chlorine breakthrough that has damaged the RO membrane, a change in feed water quality, or a sensor fault. Confirm by comparing inlet and outlet conductivity, calculating RO rejection, and checking the reading on an independent calibrated meter.

Why does water from a stored carboy fail when the dispenser sample passes?

Because storage changes the water. Carbon dioxide dissolves and lowers pH, organisms multiply in a closed but non-sterile container, and plasticizers can leach from the walls. Draw fresh water for critical work and set a maximum holding time in the procedure.

Does boiling water make it suitable for microbiology use?

No. Boiling kills many vegetative organisms but does not remove ions, organics, endotoxin, or bacterial debris, and the water can be recontaminated as it cools. Purified water for microbiology must come from a qualified purification system.

How long can purified water sit in a carboy?

There is no single safe number, because the answer depends on the grade, the container, and the storage temperature. Set a holding time in your procedure, keep the container closed, and treat any water held beyond that time as out of specification.

What should be done with samples tested while the water system was out of specification?

Identify every affected test, review the results against expected performance and controls, and repeat those that cannot be defended. Record the impact assessment with the corrective action. Results that may have been compromised should be reviewed by the laboratory director before release.