Lubinpla

Coolant Bacteria Testing: Why One Reading Is Not a Trend

Summary

HSE recommends weekly coolant dipslide counts in CFU/ml, but one count shows only that day's control band, not whether bacteria are rising. A trend needs a run of same-method counts at a fixed interval, read with pH and concentration from the same dates. This holds for water-mix metalworking fluids under HSE MW5 (2021) and HSE web guidance as of 2026-10-23.

  • HSE bands for dipslide counts: below 10,000 CFU/ml is good control, 10,000 to below 1,000,000 CFU/ml is reasonable control, and 1,000,000 CFU/ml or above is poor control, where draining and cleaning should normally take place (HSE, 2024).

  • HSE MW5 asks for dipslides once a week and records of all tests kept for at least five years in a table or graph (HSE, 2021).

  • In 44 in-use metalworking fluids, qPCR bacterial concentrations were up to five orders of magnitude higher than culture counts (Gilbert et al., 2010).

  • ASTM E2694-26, active as of 2026-10-23, states that culturable colonies need 36 h to 48 h or longer to become visible (ASTM International, 2026).

This reading applies to water-mix metalworking fluids checked with dipslides or laboratory plate counts. Neat oils, fungal counts and specific organisms such as non-tuberculous mycobacteria need their own methods. No bacteria count replaces the workplace risk assessment or a health assessment of exposed staff.

What does coolant bacteria testing with dipslides measure, and how?

Coolant bacteria testing with dipslides measures the culturable bacteria in a sample of water-mix metalworking fluid, meaning the ones that grow into colonies on the slide. The result is read after incubation and reported as colony-forming units per millilitre (CFU/ml). The HSE describes the device: "A dip slide consists of a plastic carrier coated with a sterile culture medium, which is dipped into the liquid to be tested." The slide is then incubated so colonies grow, and their density is compared with a chart to estimate the count (HSE, 2024).

HSE MW5 (2021) sets three handling rules: stir the sump before testing so tramp oil does not coat the slide, use a dipslide incubator so every test develops at the same temperature, and follow the dipslide supplier's incubation instructions. The sources give slightly different incubation windows. The UKLA good practice guide says bacterial colonies can generally be counted after around 24 to 48 hours of incubation (UKLA, 2025). ASTM E2694-26 gives 36 h to 48 h or longer for culturable colonies to become visible (ASTM International, 2026).

The count is a culture result: it covers only the bacteria in that sample that grew on that medium, at that temperature, in that time. ASTM E2694-26 notes: "It has been estimated that less than 1 % of the species present in an environmental sample will form colonies under any given set of growth conditions" (ASTM International, 2026). A dipslide count therefore works as an index of control, to be compared with earlier counts from the same method. It is not a census of every organism in the sump.

Which coolant bacteria count bands call for action?

The HSE sorts coolant bacteria counts from dipslides into three bands, and the upper two call for action (HSE, 2024). Below 10,000 CFU/ml is good control, and regular checks should continue. From 10,000 to below 1,000,000 CFU/ml is reasonable control: review the fluid, adjust its parameters to those the supplier recommends, and if bacterial growth continues despite these adjustments, add biocide at the dose the supplier recommends (HSE, 2024). 1,000,000 CFU/ml or above is poor control. There the HSE says immediate action should be taken in line with the risk assessment, and draining and cleaning should normally take place.

The UKLA guide frames the middle band as a pattern, not a single slide: "Bacteria growth is indicated when dipslide results are consistently at or above 10,000 CFU/ml" (UKLA, 2025). Two slides in a row in that band tell you more than one.

MW5 also fixes the order of steps. Do not add biocide until the good practice measures have been followed: concentration, pH, tramp oil content, metal contamination, operating temperature, agitation and flow (HSE, 2021). When biocide does go in, the UKLA guide asks that every addition is noted and records kept, and that the supplier's SDS is checked for the biocide's hazards and the protective equipment it needs (UKLA, 2025).

These bands are British guidance from the HSE and UKLA. A fluid supplier or another jurisdiction may set different action levels.

How do you read coolant bacteria results over consecutive samples?

Coolant bacteria results over consecutive samples are read by band and direction together. Band is where the latest count sits against the HSE thresholds. Direction is whether same-method counts at the same interval are rising, flat or falling. MW5 asks for records "in a table/graph (see included example) to make it easy to spot trends and act on the results to maintain fluid quality" (HSE, 2021). The OSHA best practices manual makes the same point about all fluid data, which "should be tabulated so that relationships and trends in the data can be spotted" (OSHA, 1999 committee basis).

The table below pairs a reading's band and direction with the pH and concentration logged on the same dates, and gives the action that follows. The patterns are illustrative; the band values and actions come from the cited sources.

Result

Meaning

Action

One count at 10,000 to below 1,000,000 CFU/ml after weeks below 10,000

Reasonable control that day (HSE, 2024); one point is not yet a trend

Check concentration, pH, tramp oil and temperature; repeat the dipslide at the next weekly test with the same incubator

Count band rising over three weekly dipslides, same method (illustrative)

Growth is outpacing fluid management

Correct fluid parameters first; if growth continues, dose biocide at the supplier's rate and log it (HSE, 2021)

Count flat, pH falling, concentration below the supplier range

A sharp drop in pH may indicate high bacteria levels (UKLA, 2025); the count may lag

Restore concentration, recheck pH, and test more often until the readings settle

Dipslide and laboratory plate count disagree

Different medium, temperature, volume and transit; both can be valid (Chemaqua, 2018)

Keep trending one method; compare methods side by side over several samples

Low count after biocide, but odour returns

Biofilm may be reseeding the fluid (Xi and Franzblau, study 2009-2011)

Inspect sump walls, pipes and chip beds; plan a full system clean

Count at or above 1,000,000 CFU/ml

Poor control (HSE, 2024)

Act immediately per the risk assessment; normally drain and clean

The row that most often changes a decision is the flat count with falling pH and low concentration. The dipslide still reads in band, but the fluid is losing the conditions that hold bacteria down. Waiting for the count to rise only delays the correction (inference from UKLA, 2025).

Why do dipslide and lab plate counts give different coolant bacteria numbers?

Dipslide and laboratory plate counts are different culture tests, so they give different coolant bacteria numbers. Medium, incubation temperature, sample volume, reading method and transit time all differ, and neither number is the true count. A water treatment bulletin on cooling systems lists these causes and names the culture medium as the most significant, since dipslides usually use a high-nutrient tryptic soy agar while laboratories may use a high-nutrient medium or a low-nutrient medium such as R2A agar (Chemaqua, 2018).

The same bulletin notes that dipslides are often incubated at room temperature while laboratory plates are incubated at 95 °F (35 °C), that a dipslide designed to absorb a 0.1 mL sample may take up more or less, that dipslide reading is a visual estimate, and that bacteria can grow or die in transit to the laboratory (Chemaqua, 2018). The bulletin is about cooling water, but the same mechanisms apply to coolant samples by inference. A metalworking fluid also adds tramp oil, which MW5 warns can coat the slide (HSE, 2021). The Chemaqua bulletin concludes that differences between dipslide and laboratory results are to be expected.

Both methods also share one limit: they count only what grows in culture. Gilbert et al. (2010) tested 44 in-use metalworking fluids and reported: "Concentrations obtained by qPCR were up to five orders of magnitude higher than by culture, suggesting that MWF contamination is generally underestimated." A University of Michigan study funded by NIOSH found that less than 10% of bacterial cells could be cultured (Xi and Franzblau, study 2009-2011).

Why read coolant bacteria counts with the same method and interval?

Coolant bacteria counts need the same method and interval because a change in slide type, incubation temperature, incubation time, sampling point or test spacing can move the count on its own. The trend then shows the method rather than the fluid. MW5 asks for three things: a dipslide incubator that keeps the development temperature the same for every test, the supplier's incubation time, and dipslides once a week (HSE, 2021). The UKLA guide adds a dipslide on the supply water as a control test and a check that slides are within the manufacturer's shelf life (UKLA, 2025).

The interval matters as much as the method. A weekly count sees the fluid once every seven days, and a warm sump can change between tests: the UKLA guide notes that operating a water-mix fluid above 30 °C creates optimal conditions for microbial growth (UKLA, 2025). The HSE allows a lower test frequency only when records show that fluid quality management is effective, "eg dipslide, fluid concentration and pH test records" (HSE, 2024).

A practical rule follows (inference). On every reading, note the method, incubation time and temperature, sampling point, and whether biocide or fresh concentrate went in since the last test. A jump in the count can then be traced to the fluid or to the method. Keep those records: MW5 asks that records of all tests be kept for at least five years (HSE, 2021).

How do you read coolant bacteria counts together with pH and concentration?

Coolant bacteria counts are read together with pH and concentration for two reasons. The HSE says direct bacterial measurement should be used with those checks, and pH and concentration often move before a culture count does. The HSE page states: "Direct means of measuring bacterial contamination should be used in conjunction with other checks on fluid quality, eg fluid concentration and pH." MW5 asks for refractometer concentration and pH every week against the supplier's recommended levels (HSE, 2021).

The UKLA guide spells out how the three are linked. Diluting a fluid below its recommended concentration "may increase the risk of bacterial contamination and corrosion," microbial growth can occur when pH drops below the recommended operating range, and "a sharp drop in pH may indicate high bacteria levels" (UKLA, 2025).

Read the three series on the same dates. A count that stays flat while pH drifts down and concentration sits below range suggests the fluid is losing its defences, and the culture count may follow (inference).

A rising count with steady pH and in-range concentration points instead to a source such as tramp oil, stagnant zones or contaminated make-up water; the UKLA guide says tramp oil above 2 percent encourages microbial growth, links stagnant fluid to bacterial odour and advises a control dipslide on the water supply (UKLA, 2025). A record layout that keeps these readings per machine and per site is described in how to keep a coolant concentration and pH history per site.

What can coolant bacteria testing miss in biofilm?

Coolant bacteria testing on a sump sample can miss much of what lives in biofilm: the organisms attached to sump walls, pipes and chip beds rather than suspended in the fluid. A dipslide or plate counts only the suspended ones. The cooling water bulletin states that bulk water bacteria counts "don't reflect bacteria growing inside biofilms" (Chemaqua, 2018). The OSHA manual describes chip beds as a nesting area for bacteria where the biomass can be too large for simple biocide treatment to be effective (OSHA, 1999 committee basis).

Metalworking fluid shows the same gap: a University of Michigan study sampled bulk fluid and biofilm from seven systems in two automobile plants monthly for a year and found that "the relative abundance of M. immunogenum was significantly higher in biofilms than in bulk fluid samples" (Xi and Franzblau, study 2009-2011). The authors found that biofilm communities resembled the fluid of the same system, "suggesting biofilms may act as a seeding source of fluids for bacterial regrowth," and named biofilms as a primary focus of management.

Odour can also come from organisms a routine count leaves out: anaerobic sulfate reducers cause most of the odour, routine laboratory sampling usually tests only aerobic bacteria, and accurate anaerobic results take 120 to 160 hours (Master Fluid Solutions, 2022). Mycobacteria need their own method: ASTM E2563-23 is a plate count practice for non-tuberculous mycobacteria in aqueous metalworking fluids with a detection limit of one CFU/mL (ASTM International, 2023).

Common mistakes in reading coolant bacteria counts

Three mistakes are common in reading coolant bacteria counts: treating one slide as a verdict, comparing counts from different methods, and adding biocide before fixing the fluid. One slide in the 10,000 to below 1,000,000 CFU/ml band is a prompt to check the fluid (HSE, 2024). The UKLA guide ties bacteria growth to results that are consistently at or above 10,000 CFU/ml (UKLA, 2025).

Mixing methods hides trends. A laboratory plate count, a dipslide from another supplier and a rapid adenosine triphosphate (ATP) result under ASTM E2694-26 each measure something different. ATP reflects total bioburden, and the standard notes that "different factors affect ATP concentration than those that affect culturability" (ASTM International, 2026). Trend each method against itself.

Biocide first is the third mistake. MW5 says not to add biocide until the good practice measures have been followed (HSE, 2021), and the UKLA guide warns that overuse of biocides may increase the risk of developing biocide resistant organisms (UKLA, 2025). Log every dose with its date and amount so the next count can be read against it. Before handling the biocide, read its SDS and follow COSHH, or your local hazardous substance rules.

In Lubinpla, same-method dipslide, pH and concentration readings can be kept as dated measurement history in a project's conditions, with each biocide addition logged as a product use record. A new count then reads as part of a trend rather than as one value.

  • One dipslide count places the fluid in an HSE control band for that day. It does not show direction.

  • Trend counts from one method, one incubator setting and one sampling point, at a fixed interval.

  • Read every count with the pH and concentration measured on the same date.

  • Dipslide and laboratory plate counts differ by design, so do not mix them in one trend.

  • A low count with returning odour points to biofilm or to anaerobic bacteria the count does not see.

Related posts