Lubinpla

Coolant sampling from a machine sump: where to take the sample

Summary

Sampling coolant from a machine sump takes four steps: run the pump, sample the circulating flow, avoid surface and bottom, and label the bottle. Draw from the return flow or mid-depth of the circulating zone, at the same point and time every round, so concentration, pH and microbial results describe the coolant the tool actually receives.

  • HSE recommends measuring water-mix coolant concentration with a refractometer and pH every week, and taking dipslides once a week (HSE MW5, 2021).

  • Dipslide results consistently at or above 10,000 CFU/ml indicate bacterial growth; at or above 1,000,000 CFU/ml immediate action is required (UKLA, 2025).

  • Pall's sampling protocol for hydraulic and lubricating oil recommends at least 30 minutes of operation before sampling and a tank sample taken about halfway under the surface (Pall Corporation, accessed 2026).

  • ASTM E2694-26 measures ATP in water-miscible metalworking fluids in about 5 min, against 36 h to 48 h for culture colonies to become visible (ASTM, 2026).

What do you need before taking a coolant sample from a machine sump?

Before taking a coolant sample from a machine sump you need a clean, labeled bottle for each test type, a sampling tube or dipper long enough to reach mid-depth, gloves, and a written point and time for that machine. The OSHA metalworking fluids manual recommends a written procedure for fluid testing that states where and when to collect samples, how to treat them after collection, which tests to run, the protocol for each test, and who performs and records them (US OSHA, accessed 2026). ASTM E1497-23, the practice for selection and safe use of metal removal fluids, covers product selection, storage, dispensing and maintenance (ASTM E1497-23, 2023).

Settle the procedure with the coolant supplier and the laboratory before the first round, since the bottle and the time limit depend on the test. For concentration and pH, a clean, dry bottle is enough. A microbial sample goes into the sterile container or onto the dipslide (an agar-coated slide dipped in the coolant, then incubated) that the laboratory or kit maker specifies.

The UKLA and HSE guide recommends single-use gloves when handling a dipslide and the coolant sample (UKLA, 2025). Gloves bring a hazard of their own, though: HSE warns that gloves increase the risk of entanglement at moving machinery (HSE MW5, 2021). Take samples with the spindle stopped and guards closed, and follow the machine maker's instructions and the coolant SDS for skin protection.

Where in the machine sump should the coolant sample come from?

The coolant sample should come from the circulating flow, with the pump running: the return to the sump, a sampling valve on the supply line, or mid-depth in the zone where the return stirs the sump. Coolant guidance does not fix a sampling point. The closest published rules are Pall's, written for hydraulic and lubricating oil.

Pall recommends sampling from a turbulent section where contaminants are uniformly suspended, not using a tank drain port directly, and, in a tank, inserting the tube about halfway under the fluid surface (Pall Corporation, accessed 2026). Pall also states that using the same sample location each time is important for trending, so that one round compares with the next.

That logic carries over to a coolant sump with one change. Coolant carries tramp oil floating on top and chips and sludge at the bottom, so both edges of the sump misrepresent the fluid. Use the table below to decide whether a coolant sample can be kept or must be retaken.

Check

Pass

Fail and what to do

Sample point

Return flow, supply-line valve, or mid-depth near the return

Surface, bottom or a stagnant corner: retake

Pump running before sampling

Running for a set, recorded time

Pump just restarted after a stop: run it, then retake

Tramp oil layer in the bottle

None visible after standing

Oil layer from the surface: retake below it

Bottom sludge or chips in the bottle

None, or fines only

Sludge or chips: retake higher in the flow

Container for microbial sample

Sterile, as the lab or kit specifies

Reused or open bottle: discard and retake

Label

Machine, time, point, sampler

Any field missing: complete it before testing

The first row, sample point, decides the rest. A dip from the top of a quiet sump picks up the tramp oil layer, and no label or container can make that sample represent the circulating coolant.

Why does a coolant sample from a dead zone of the sump mislead?

A coolant sample from a dead zone, a part of the sump where the coolant barely moves, misleads because stagnant coolant separates and grows microbes differently from the circulating fluid the tool sees. The OSHA metalworking fluids manual notes that quiet time lets tramp oil float and solids settle, and that extended inactivity of more than 8 to 10 hours promotes anaerobic bacteria in fluids that are heavily contaminated or lack the correct concentration of the right biocide (US OSHA, accessed 2026). HSE guidance names points of low or no flow, such as dead-ends and dead-legs, as places where biofilm grows, and recommends keeping the fluid circulating with pumps (HSE MW5, 2021).

So a dip from a dead corner can show a higher bacteria count, a lower pH or more tramp oil than the working fluid. A dip from the top can blur the refractometer reading: the UKLA and HSE guide shows a hazy boundary line when tramp oil is present in old coolant (UKLA, 2025). Either way, the result describes a corner of the sump, not the machine.

A dead-zone sample still has a use when you are looking for a biofilm source. Take it as a separate, labeled sample and never average it into the trend for the circulating fluid.

When should you take the coolant sample?

You should take the coolant sample at the same point in the shift each round, after the pump has circulated the sump and before any top-up of water or concentrate. A sample drawn right after a top-up measures the addition, not the sump. One drawn at start-up after a weekend or a shutdown measures coolant that has been standing, and the OSHA metalworking fluids manual links more than 8 to 10 hours of inactivity with anaerobic growth in poorly controlled fluids (US OSHA, accessed 2026).

Pall's protocol for hydraulic and lubricating oil recommends at least 30 minutes of operation before sampling, and if sampling during operation is not possible, sampling as soon as safely possible after shutdown to prevent settling (Pall Corporation, accessed 2026). Coolant guidance gives no minimum circulation time. Set one for each machine type, write it into the procedure and keep it constant.

On frequency, HSE recommends weekly refractometer concentration and pH checks and weekly dipslides for water-mix coolant, and allows a lower dipslide frequency only when records show fluid quality management is effective (HSE MW5, 2021). HSE also recommends checking fluid appearance and odor every day between sampling rounds (HSE MW5, 2021).

How is a coolant microbial sample taken differently from a concentration sample?

A coolant microbial sample goes into a sterile container or onto a dipslide and is tested quickly, while a concentration and pH sample needs only a clean bottle and a representative draw. HSE recommends stirring the sump before a dipslide test so that tramp oil does not coat the slide (HSE MW5, 2021). The UKLA and HSE guide recommends submerging the dipslide agar for about 10 to 15 seconds, running a second dipslide on the supply water as a control, and reading bacterial colonies after around 24 to 48 hours of incubation (UKLA, 2025).

For a laboratory count, use the container the laboratory supplies and keep it closed. Ask the laboratory for its time and temperature limits between sampling and testing, because microbes keep growing in the bottle.

ASTM E2694-26 is a faster route: it measures ATP, a constituent of all living cells, as an indicator of total microbial contamination, with results in about 5 min instead of 36 h to 48 h for culture (ASTM, 2026). The standard also notes that the method does not tell bacteria from fungi. A concentration sample is read by refractometer instead, and the UKLA and HSE guide notes that the reading must be multiplied by the product's adjustment factor (UKLA, 2025).

What should you record with each coolant sample?

Each coolant sample should be recorded with the machine and sump, the exact sample point, the date and time, how long the pump had run, the last top-up, the sampler's name and the tests requested. Add what the sampler saw: odor, color, tramp oil on the surface, foam and chips. Without the point and the pump time, two results from the same machine cannot be compared, because a surface dip and a return-flow sample are different samples.

HSE recommends keeping records of all fluid tests for at least five years, in a table or graph that makes trends easy to spot (HSE MW5, 2021). The OSHA metalworking fluids manual recommends tabulating physical observations, laboratory analyses and additions to the system so that trends and relationships can be seen (US OSHA, accessed 2026).

When sampling is part of a customer visit, the same fields belong in the visit record. A guide on writing a technical service report the customer acts on shows where they fit. Lubinpla's site inspection treats a sample the same way: it uses a result for a decision only when the record ties it to the sampler, time, location and the range of fluid it represents.

Frequently asked questions

Can you take a coolant sample from the sump drain valve?

Avoid it. A drain valve draws from the bottom, where chips and sludge collect. Pall's protocol advises against using a tank drain port directly and describes a tube through the port reaching into the fluid when no other point exists (Pall Corporation, accessed 2026).

Should a coolant sample be filtered before testing?

Not before microbial testing, because filtering removes organisms. For a refractometer reading, let the sample stand until chips settle and read from the clear fluid. Note any tramp oil layer on the label rather than discarding it unrecorded.

How much coolant should you sample?

Take the volume the laboratory or test kit asks for, plus enough to repeat a test. A refractometer needs only one or two drops (UKLA, 2025), but a laboratory panel needs more.

Related posts