Dry block or liquid bath? Compare accuracy, stability, speed, safety and NATA traceability so you can pick the right temperature calibration source for your lab.
Choose a dry block when you need speed, portability and a clean, low-maintenance setup for probes that fit the insert. Choose a liquid bath when you need the best stability and uniformity, or when your sensors are short, odd-shaped or glass. Most Australian sites with mixed sensors end up needing both, or a supplier who has both. This guide sets out how the two compare on the things that decide an audit: uncertainty, immersion, safety, running cost and traceability under NATA requirements.
Both devices do the same job. They create a stable, known temperature so you can compare a sensor against a reference thermometer. NATA’s technical guidance treats calibration as comparison: the reference and the sensor under test must experience the same temperature at the moment you read them. How each device achieves that is where the differences start.
A dry block (also called a dry-well calibrator) heats or cools a metal block with drilled wells. You insert the sensor into a well, often through a removable insert sized to the probe. There is no liquid.
A liquid bath holds a fluid such as water, silicone oil or alcohol, depending on the temperature range. The fluid is stirred so the temperature is even throughout. You immerse the sensor and the reference directly in the fluid.
There are also hybrids with separate dry inserts and liquid inserts, as Beamex notes. They suit labs that cannot decide, but they carry the compromises of each side.
| Criterion | Dry block | Liquid bath |
|---|---|---|
| Stability and uniformity | Good, but generally lower than a bath | Generally better homogeneity and stability |
| Sensor shape | Needs a matching insert; poor for short or odd shapes | Handles short, odd-shaped and glass sensors |
| Speed | Fast heating and cooling | Slower to reach and settle at temperature |
| Portability | Usually light enough for site work | Heavier, fluid has to be transported and managed |
| Maintenance | No fluid to change or clean | Fluid care, cleaning, monitoring |
| Main error sources | Sensor loading, stem conduction, insert fit | Immersion depth, fluid choice, stirring |
| Safety | Hot surfaces | Hot oil, fumes, steam if water contacts hot oil |
| Best for | Field checks, routine probes of standard diameter | Low-uncertainty work, mixed or unusual sensors |
The general direction in this table matches what manufacturers publish. Beamex lists lower accuracy and stability, and difficulty with very short or odd-shaped sensors, as the drawbacks of dry blocks. It lists better homogeneity and stability, and suitability for odd-shaped sensors, as the main benefits of a bath. Your exact numbers depend on the model, so always read the specification and the uniformity data for the unit you intend to use.
A dry block reaches setpoint quickly and cools down quickly. For a technician working through a plant with thermometers, probes and process sensors on a schedule, that matters. Less time waiting means more instruments checked in a shift, and less time a process sensor is out of service.
With no oil to select, replace, filter or dispose of, a dry block removes a whole category of housekeeping. It also removes the contamination risk: a dry block cannot leave oil residue on a sensor that later goes into a food or pharmaceutical process. For food and pharmaceutical sites that is a real advantage, because every extra cleaning step is a step someone can skip.
Surfaces get hot, but there is no open fluid at high temperature. The steam-explosion hazard of water meeting hot oil does not exist. Training an operator on a dry block is generally faster than training them on a bath.
Stirred fluid surrounds the sensor on all sides, so the temperature around the probe is more even than in a drilled well. When the uncertainty you need is tight, or when you are calibrating reference-grade thermometers, this is the reason to use a bath.
Short probes, bent probes, thermocouples with unusual tips and liquid-in-glass thermometers are all easier in a bath. Fluke points out that liquid-in-glass thermometers can stick in a dry-well insert or break during calibration, which is why baths are recommended for them.
A dry block depends on the gap between the probe and the insert wall. A loose fit lets the probe’s temperature drift from the block’s. In a bath, the fluid fills that space for you.
Choosing the device is half the job. The other half is using it in a way that survives an audit. The NATA temperature metrology annex to ISO/IEC 17025 is the document Australian laboratories work from, and it is specific about both.
Heat travels along the probe stem, in or out of the block. This stem conduction is often a dominant uncertainty source in dry block work. NATA’s guidance describes a pull-up test for probes that are not immersed to at least 200 to 250 mm. You withdraw the probe 20 mm and watch the reading change, which lets you account for the error. For dry blocks, NATA also expects axial uniformity to be assessed over a 20 mm distance.
Fluke makes a related point: dry blocks are more sensitive to sensor loading errors, caused by heat moving in and out of the block through the inserted sensor. Loading a block with several large probes at once makes this worse.
For baths, the NATA annex expects documented assessment of sample loading effects and thermal losses. It also calls for accessories such as sensor guides. Without one, a probe can touch the bottom of the tank, obstruct the stirring and ruin uniformity.
This is the most common failure in the field. Whether you use a block or a bath, the reference thermometer must be in the same medium as the device under test and at a comparable immersion. Any difference in depth or contact between the two breaks the comparison and, with it, traceability.
The guidance also says to record the actual value of the medium, measured by the reference, alongside the readings from the devices under test. Do not rely only on the calibrator’s setpoint. A display that says 100.0 °C is a request, not a measurement.
NATA criteria expect the reference thermometer’s uncertainty to be no more than one fifth of the uncertainty required for the working thermometer being calibrated, according to the same NATA annex. That applies to dry blocks and baths alike. A very good bath with a poor reference still gives you a poor calibration.
A dry block is mostly a burn risk. A bath adds fire, fumes and steam.
All three points come from the NATA annex linked above. Follow the manufacturer’s manual and your site’s safety procedures as well. If your team has not run a hot oil bath before, have an experienced technician supervise the first runs.
A dry block has a lower ongoing cost in consumables. You buy inserts for the probe sizes you use, and that is about all. A bath needs fluid, and fluid degrades, collects dirt and has to be matched to the temperature range. Calright describes liquid bath systems as delivering strong stability but needing more care, cleaning and monitoring. That labour is the real cost, and it is easy to overlook when comparing purchase prices.
There is also the cost of the uncertainty budget. A dry block that is poorly matched to your probes can force you into a larger stated uncertainty, which can in turn mean more of your instruments fail a tolerance check on paper. Cheap hardware that widens your uncertainty is not cheap.
| Situation | Likely choice |
|---|---|
| Food production line probes checked on site | Dry block, with matched inserts |
| Pharmaceutical lab, reference thermometers | Liquid bath, with a good reference |
| Mixed fleet with glass thermometers and short probes | Liquid bath |
| Routine verification of handheld probe thermometers | Dry block |
| Calibration of data logger probes in bulk | Either, depending on probe size and the uncertainty needed |
These are starting points, not rules. Your own uncertainty requirement should override any table.
We calibrate instruments for manufacturers, laboratories, food and pharmaceutical sites, and the temperature problems we meet are rarely about which box someone bought. They are about how it was used.
The usual failure is a reference thermometer placed differently from the sensor being tested: a different depth, a different well, a different medium. The paperwork looks fine, the certificate exists, and the comparison is still not valid. That is the trap, and it catches teams with expensive equipment as easily as teams with cheap equipment.
Our advice is to settle three things before you spend anything. First, write down the tolerance your process needs. Second, check that your sensors physically suit the device. Third, decide who owns the fluid, the inserts and the uniformity records. If any of those is unclear, a calibration done by an accredited provider will often cost less than the confusion.
We are NATA accredited and ISO 9001 certified, we service multiple equipment brands regardless of where the equipment was purchased, and we calibrate both on site and in our workshop. That means we can work with whichever approach your site has, and tell you straight when the approach is the problem.
— Kaz
Whichever device you use, the audit question is the same: can you show that every reading traces back to a reference with a known uncertainty, and can you show what you did when something was out of tolerance? Both parts need a process, not just a certificate.
An out-of-tolerance finding is where many teams stall. The question is not only whether the sensor is adjusted, but which production or test results relied on it since the last good calibration. This out-of-tolerance guide for calibration teams sets out a structured way to handle those events so the evidence is ready when an auditor asks.
If you want to understand how accreditation and certificates fit together, read our guides on NATA calibration standards in Australia and NIST vs NATA certificates. If an audit is close and an instrument has just failed, our emergency calibration plan lays out the steps.
Owning a dry block or bath only makes sense if you calibrate often enough to justify the equipment, the reference thermometer, the training and the uniformity records. For many sites, checking a handful of probes a few times a year, buying in accredited calibration is the more defensible route. The certificate arrives with an uncertainty statement already worked out under accreditation, and you do not carry the fluid, the hot surfaces or the documentation burden.
There is a cost angle here too. On-site work usually costs more than workshop work because of travel and set-up, and we explain why in our NATA calibration cost guide. Sending probes to the workshop in batches is often the cheapest way to stay compliant.
We calibrate temperature instruments under NATA-accredited and traceable calibration, on site or in our workshop, and we also offer hire equipment, preventative maintenance and repairs from the same page. For day-to-day checks, our range of thermometers and measuring equipment includes probe thermometers such as the Testo 905-T1 and the Testo 175 T3 dual thermocouple logger. Tell us what you measure, the range and the tolerance you need, and we will quote for your site. Prices are tailored to the job, so we do not publish a fixed rate.
Usually not quite. Dry blocks are generally less stable and uniform than a stirred bath, and they are more affected by sensor loading and stem conduction. For many routine checks the difference is small enough not to matter. For low-uncertainty work, a bath is the safer choice.
A liquid bath. The fluid surrounds the sensor regardless of its shape, whereas a dry block needs an insert that fits. Glass thermometers are also better in a bath, because they can stick or break in a dry-well insert.
A calibration bath is a tank of stirred fluid, such as water or silicone oil, held at a controlled temperature. You immerse the sensor under test and a reference thermometer in the fluid and compare their readings. The fluid has to suit the temperature range and be kept clean.
Yes. For traceable work, record the actual temperature of the block or bath as measured by the reference thermometer, alongside the readings of the devices under test. NATA’s guidance for accredited work does not accept the calibrator’s setpoint alone.
It has to be in the same medium, at a comparable immersion, so both experience the same temperature. A difference in depth or contact weakens the comparison and can invalidate traceability.
