In-House Water Testing for Sterilizers: Conductivity, Hardness and Chloride Methods a Clinic Can Run

Four self-sealing sterilization pouches in increasing sizes, 90 x 230 mm on the largest

Quick facts

  • Three of the Canadian water limits for reprocessing — conductivity, hardness and chloride — can be measured at a clinic bench with a meter, a titration kit and a wash bottle; four others cannot be measured there at all.
  • The Canadian single-value table comes from provincial guidance adapted from the CSA reprocessing standards: pH 6.5–8, evaporation residue ≤15 mg/L, conductivity ≤50 µS/cm, hardness ≤0.1 mmol/L, chloride ≤3 mg/L.
  • Conductivity is measured against a potassium chloride standard of about 1,413 µS/cm at 25 °C, and every reading has to be temperature-compensated to 25 °C before it is compared with a limit.
  • Hardness at 0.1 mmol/L is the same as 10 mg/L expressed as calcium carbonate, and a kit that reports in grains per gallon or in drops without a conversion has not given you a result you can file.
  • Steam purity is not a bench measurement at all: non-condensable gases, dryness value and superheat need a steam quality survey with dedicated sampling apparatus.

Most clinics discover their water problem through a symptom rather than a reading: instruments come out spotted, pouches come out damp, or a biological indicator fails in a cycle that looked normal on the printout. By then the question is already expensive. The three parameters that cause most of those symptoms, though, are cheap to measure in-house — a conductivity pen, a hardness titration kit and a chloride titration kit cost less than one cancelled list, and they produce a number you can plot month to month.

This article is about the measurement side of the problem: which parameter can be tested where, with what instrument, in what units, and where the boundary sits between a bench test and a laboratory result. The packaging is a reminder of what is at stake downstream — a box of Class 4 self-sealing pouches holds a sterile barrier claim that rests on a rinse water quality nobody inspects unless a residue test is run — but the tests below are about the water, not about the pack.

Which water parameters can a clinic measure on its own bench?

The Canadian limit table is short and unusually practical, because it was written for facilities rather than for laboratories. Each row below is paired with the method that can produce it, and with an honest statement of what that method can and cannot establish.

Parameter Canadian limit Bench method What the bench method establishes Sent out?
pH 6.5–8 pH meter with two-point buffer calibration Acidity or alkalinity of the final rinse; a proxy for detergent carry-over No
Conductivity ≤50 µS/cm Conductivity meter or pen, temperature-compensated to 25 °C Total dissolved ionic content — the fastest screen for a softener or reverse-osmosis failure No
Hardness ≤0.1 mmol/L EDTA titration kit (drop count or burette) Calcium and magnesium, the scale-forming pair No
Chloride ≤3 mg/L Argentometric titration kit Chloride, the pitting-corrosion driver for stainless instruments No
Evaporation residue ≤15 mg/L Gravimetric: evaporate a measured volume, dry and weigh Total non-volatile residue, including species the three tests above miss Yes
Iron ≤0.2 mg/L Colour-comparison kits are semi-quantitative only Dissolved iron, a corrosion and staining marker Yes
Phosphate ≤0.5 mg/L Not practical at the bench Detergent residue carry-over Yes
Silicate ≤2 mg/L Not practical at the bench Dissolved silica from glassware, cement or certain water sources Yes

The right-hand column is the part clinics skip. Conductivity, hardness and chloride are the three parameters where a bench result is defensible; the other rows require an accredited laboratory because the methods are instrumental, the limits are low, and the result has to survive a challenge.

How do you run a conductivity reading that holds up?

Conductivity is the highest-value bench test because it is the earliest indicator that a treatment stage has failed, and because the meter is the least costly instrument in the kit. The method is standardised — United States EPA method 120.1 covers the determination of specific conductance in water — and the practical steps are the ones that get skipped.

  • Calibrate against a potassium chloride standard, not against the previous reading. A 0.01 M potassium chloride solution reads about 1,413 µS/cm at 25 °C, which is the anchor point every meter expects. A pen that has drifted 20 % still gives a stable, plausible number.
  • Compensate to 25 °C, and check what the meter means by compensation. Conductivity rises roughly 2 % per degree Celsius, so a sample at 30 °C reads about 10 % high if the meter does not correct it. A limit of 50 µS/cm is meaningless without a reference temperature.
  • Check the unit before comparing. The Canadian table is written in microsiemens per centimetre. Some published summaries of the American water standard print the same parameter in millisiemens per centimetre, which is a factor of one thousand; a comparison made across those units produces a confident and completely wrong conclusion.
  • Sample at the point of use, not at the treatment plant. The final rinse outlet, the rinse arm or the basin where instruments are rinsed is where the number has to hold. A reading taken at the reverse-osmosis outlet proves the membrane works, not that the water arriving at the instrument does.
  • Record the reading with its temperature, instrument identity and calibration date. A conductivity value without those three fields cannot be compared with last quarter's value, which is the only thing a trend is made of.

Detergent chemistry is the other half of the residue question. A phosphate result above the limit is usually a rinse that did not clear the cleaning step rather than a water supply fault, which is why the cleaning chemistry and the water have to be reviewed together: a concentrated multi-enzyme ultrasonic cleaning solution is formulated to be rinsed clear, and a rinse stage that is too short or too cool leaves the residue that later reads as a water problem.

The same instrument will not measure steam purity. Condensate conductivity is a different sample, taken from a cooled steam sample, and it sits in the steam survey rather than in the daily log.

How do you run a hardness titration without a laboratory?

Hardness is measured by titrating the sample with a chelating agent — EDTA — until an indicator changes colour, and the volume consumed gives the calcium and magnesium content. Drop-count kits do the same arithmetic with a fixed drop volume, which makes them fast and makes their resolution the thing you have to check.

The number that trips people is the unit. A limit of 0.1 mmol/L is stated in millimoles per litre; kits frequently report in milligrams per litre as calcium carbonate, in grains per gallon, or in German degrees of hardness. One millimole per litre of calcium carbonate is about 100 mg/L, so 0.1 mmol/L equals about 10 mg/L as calcium carbonate, which is about 0.56 grains per gallon and about 0.56 German degrees. A kit that reports "hard" or "soft" without a conversion cannot be filed against a numeric limit.

Two practical rules follow. First, use a kit whose smallest increment is small relative to the limit: if a drop is worth 17 mg/L as calcium carbonate, the kit cannot resolve a result near 10 mg/L, and the reading is effectively a pass or fail with no trend. Second, take the sample after the softener and after any blending, because a blended supply can pass on one day and fail on another as the mix ratio drifts.

Why does chloride at 3 mg/L matter more than the number suggests?

Three milligrams per litre is a very low limit, and the reason is metallurgical. Chloride in the presence of moisture and heat drives pitting and stress-corrosion cracking in stainless steel instrument surfaces, particularly where passivation has been worn by repeated handling. The damage is not visible in the rinse water and not visible on the instrument until the pit is deep enough to trap soil.

That has two consequences for testing. First, a general-purpose chloride strip is not adequate: many strips step at 10, 25 or 50 mg/L, so the entire compliant range sits inside the first colour block. A titration kit for chloride, run argentometrically, is what produces a number at this concentration. Second, sample handling matters more than at higher concentrations — a rinsed container, a full bottle and a prompt reading avoid the contamination and evaporation errors that dominate at the milligram level.

Practically, chloride is the parameter most often skipped and the one most often implicated when instruments develop pits or discolouration that cleaning cannot remove. It is also the parameter most sensitive to a change of water source, which is exactly the kind of change that happens without anybody writing it down.

Where is the boundary between the bench and the laboratory?

The boundary is defined by the limit and by the method's resolution, not by convenience. Four parameters have limits so low, or methods so instrumental, that a bench result would be an estimate dressed as a measurement.

Reading Why it has to be a laboratory result What the bench can do instead
Evaporation residue ≤15 mg/L Gravimetric: a measured volume is evaporated and weighed on an analytical balance; the result depends on drying temperature and desiccation Track conductivity as a screen and send a sample when conductivity drifts
Iron ≤0.2 mg/L Colorimetric or spectrometric at low concentration; visual kits are semi-quantitative Note staining on instruments and send a sample when it appears
Silicate ≤2 mg/L Requires colorimetric instrumentation and careful blank handling; glass containers contaminate the sample Use plastic containers for every sample, and never store a sample in glass
Steam purity tests Non-condensable gases ≤3.5 % (v/v), dryness value ≥0.95, superheat ≤25 °C are measured with a steam sampling apparatus at the sterilizer inlet Commission a steam quality survey and file the report

The steam row is the one that is most often mistaken for a water test. It is not: the three limits are properties of the steam arriving at the chamber, and they are measured with a condenser, a gas collection arrangement and a throttling calorimeter rather than with a conductivity pen. Those tests belong to a survey performed at installation, after major repair and on the facility's own schedule, and their results are what a sterilizer's certificate is being tested against in practice.

What does the Whitehorse case show about self-testing?

In 2025 the Yukon's largest hospital ran a sterilization failure that postponed 165 surgeries, including 135 cataract cases, after instruments emerged from sterilizers with residue and staining. The root cause, confirmed by an independent review, was the use of utility water rather than critical water during the final washing of instruments, compounded by a naturally hard water supply. The hospital's own account of the recovery, including the installation of deionization and ultrafiltration systems, is published in its quality performance reports, and the operational sequence is described in the published case review.

Two lessons transfer directly to a clinic bench. The first is that the failure mode was a stage mismatch, not a catastrophic water supply: the water was acceptable for one purpose and not for another, which is exactly the distinction the American water standard draws when it separates utility water from critical water and requires the critical column to be used for the final rinse of critical devices. The second is that residue and staining on instruments are visible evidence that the water chemistry is out of range, and they appear before a biological indicator fails — which is why a monthly conductivity and hardness record has diagnostic value beyond the number itself.

A third lesson is about escalation speed. Once residue was recognised as a water problem, the response was to stop relying on the affected supply: instruments were sterilized off site and a bottled reverse-osmosis water protocol was used for limited services. A clinic with a bench log can make that call in a day rather than in a season, because the log shows whether the reading moved before the symptom appeared.

What fields belong in the water record?

A record is only useful if a person who was not present can reconstruct the measurement and decide whether it stands. The fields below are the minimum for a bench result; add the laboratory report number for any sent-out parameter.

Field Why it is there Example format
Date and time of sampling Conductivity and hardness move with demand and with regeneration cycles 2026-10-30 08:15
Sampling point The point of use is the only location a limit applies to Final rinse outlet, decontamination room
Parameter and method A number without a method cannot be reproduced or challenged Conductivity, EPA 120.1, meter SN 4471
Reading with its reference temperature 50 µS/cm at 25 °C is not the same claim as 50 µS/cm at 30 °C 38 µS/cm at 25 °C
Calibration evidence Standard value, date, and who performed it 1,413 µS/cm standard, 2026-10-30, technician initials
Limit and pass or fail The limit belongs in the record, not in a policy binder Limit 50 µS/cm — pass
Action taken Nothing is more useful at audit than a documented response Re-sampled after softener regeneration
Reviewer and date Separation between the person measuring and the person accepting Reviewed 2026-10-30 by the IPAC lead

The same fields carry across the three bench parameters, and the record is what makes a monthly pattern visible. Three consecutive conductivity readings at 40–45 µS/cm against a 50 µS/cm limit is a softener that is about to fail, and a record is the only place that pattern can be seen.

What do you do when a reading is out of limit?

Work in a fixed order, because the order is what makes the response auditable. Confirm the reading first — re-calibrate, re-sample from the same point, and check the sample temperature and container. Then look at the treatment train: softener regeneration, reverse-osmosis performance, filter changes, and any recent change of water source or plumbing. Then decide about loads: instruments processed since the last in-limit reading are the affected population, and the response for them follows the same logic as a failed load — hold, re-process, and record why.

The wider sampling plan, the three water classes and the frequency expectations are set out in three water classes, one sampling plan: reprocessing water under ANSI/AAMI ST108, and the Canadian limits are compared with the American standard in autoclave water quality in Canada vs the US. The wider frame — which standard owns which limit, and where the provincial expectations sit — is mapped in the sterilization compliance hub. Clinical engineering teams that maintain the sterilizers themselves usually want the chemistry record alongside the cycle record, and the compliance log centre carries the four record groups used here — including the water row — in a form that can be printed as a blank sheet or generated in the browser without an account.

Four self-sealing sterilization pouches in increasing sizes, 90 x 230 mm on the largest
Four pouch sizes, 90 × 230 mm on the largest. A pouch shows the result of a water problem before a biological indicator does: damp paper, water marks and staining on the paper side are the visible end of a rinse that was out of range.
Three 24-hour steam biological indicators in clear tubes with red caps for steam sterilization monitoring
A steam biological indicator is the end of the line: by the time one fails, the water and steam readings that predicted it were already out of range somewhere upstream. The bench log exists to catch the problem before this point.

Ordering for a clinic, a lab or a care home? Wholesale and multi-site ordering covers case pricing and account setup, and the B2B wholesale collection lists the lines stocked for institutional buyers. Reselling? become a distributor.

Have a question about your own facility? Send it in and you will get a written answer specific to your setup, with the regulation or standard it is based on cited. Ask your compliance question.

Related reading

Frequently Asked Questions

Can I use pool test strips for reprocessing water?

Not for a filed result. Pool strips are built for ranges two to three orders of magnitude above the limits that apply here: a chloride strip that steps at 10 or 25 mg/L cannot resolve a 3 mg/L limit, and a hardness strip with a 17 mg/L increment cannot resolve a 10 mg/L limit. A strip is useful as a rough screen on a suspect sample; it is not a measurement you can put in a log against a numeric limit.

Does conductivity replace the hardness test?

No, although the two track each other when a softener fails. Conductivity measures total dissolved ionic content across all species, so it rises when hardness rises but also when sodium, chloride or detergent residues rise. A softener that is exhausted releases sodium in exchange for calcium, which can keep conductivity almost flat while hardness climbs, so a softener failure can be invisible on a conductivity trend alone.

What temperature should the sample be at when I measure conductivity?

Report the reading at 25 °C, and note the actual sample temperature if you measured before compensation. Most meters compensate automatically to 25 °C, but the compensation is a model rather than a measurement, so a sample taken hot from a rinse line should be cooled rather than corrected, particularly near the limit. The Canadian table's 50 µS/cm is a 25 °C figure.

How do I convert 0.1 mmol/L hardness into the units my kit reports?

Multiply by 100 to get milligrams per litre as calcium carbonate, which gives about 10 mg/L, then convert to the kit's unit. Ten mg/L as calcium carbonate is about 0.56 grains per gallon and about 0.56 German degrees. Ask the kit supplier for the increment in the same unit as the limit before you start, because a conversion made after the fact is where filing errors come from.

Is the water used for the final rinse the same as the water fed to the sterilizer?

No, and that is the stage distinction the Whitehorse case turned on. The final rinse of instruments is where the critical water requirement applies, because residue left on the instrument is residue that enters the sterilizer. Feed water is a separate limit set, and steam purity is a third, measured at the sterilizer inlet as condensate. Three stages, three sets of numbers, and a pass at one does not establish a pass at the next.

How often should the bench tests be run?

Set a frequency in the facility's own procedure and tie it to the treatment train: a clinic on a softener with a monthly regeneration schedule has a different risk profile from one on a reverse-osmosis system with continuous monitoring. What matters for audit is that the frequency is written, that it is actually followed, and that an out-of-limit result triggers the documented response rather than a second sample until one passes.

Can I send a sample to any laboratory?

Use a laboratory accredited for the specific parameter, and ask for the method with the result. A result reported without a method and a detection limit cannot be compared with a limit of 0.2 mg/L or 2 mg/L, and it will not survive a challenge. The sample container also matters: plastic for silica and for low-level metals, filled without headspace, delivered within the holding time the laboratory states.

Every CliniEco Medical product is manufactured under an ISO 13485 quality management system and distributed under Health Canada MDEL #35334. If your question is how a specific reading maps onto the water limits that apply to your facility, our specialists will work through the numbers with you rather than quote a range.

Sources

Already working with clinics? Become a regional representative.