Choosing a Laboratory Vacuum Pump: Pressure, Gas Load and Duty

A pump can evacuate an empty vessel quickly and still struggle to hold the pressure once a sample starts releasing vapor. Those two tasks call for different numbers: how much gas the pump can remove at the working pressure, and how low the pressure can go when there is little or no incoming gas. Start with the job at the vessel, then read the pump specification sheet.

Learning objectives: By the end of this guide, you should be able to compare diaphragm and oil-sealed rotary vane pumps, estimate the pumping speed a known gas load requires, identify the ratings that need confirmation, and plan a dry checkout before introducing a sample.

Define the pressure and gas load

Write down the required absolute pressure at the vessel, an acceptable band, and the time allowed to enter that band. Absolute pressure is measured from zero pressure. Gauge pressure is referenced to the surrounding atmosphere, which varies with weather and elevation. At a local atmospheric pressure of 101.3 kPa, a vessel at 20.0 kPa absolute reads about −81.3 kPa gauge. For a pump curve or a boiling calculation, use the absolute value. For a vessel exposed to room air, the approximately 81 kPa difference is the load across the wall. That distinction matters when choosing both the gauge and the vessel.

Separate pump-down from holding under load. Pump-down removes the gas initially in the volume. Holding under load requires continually removing gas from sources such as a controlled air bleed, a drying sample, or dissolved gas escaping from a liquid. Write down the expected gases, whether any can condense, and how long and how often the pump will run. For a vapor-producing process, estimate the peak load downstream of any intended condenser and identify whether liquid droplets or particles can carry over. A dry checkout has little power to predict that wet load.

A free-air displacement figure such as 3.5 cubic feet per minute (CFM) is a nominal inlet-volume figure. It cannot substitute for a pumping-speed curve across the pressures you actually need. At a given temperature, a liter of gas at 200 mbar contains only about one fifth as many molecules as a liter at 1,000 mbar. This is why gas throughput is expressed as pressure times volumetric pumping speed, for example in mbar·L/s. Ultimate pressure, by contrast, describes a near-zero-load limit under defined test conditions. A catalog may instead report a base pressure measured after a specified time. Look up the conditions and distinguish either value from what a loaded vessel will reach. Pfeiffer’s vacuum technology guide explains these definitions; Chen’s throughput study illustrates why speed is measured at multiple pressures.

Compare pump mechanisms

In a diaphragm pump, a flexing membrane changes the volume of a chamber. One-way valves admit gas as the chamber expands and discharge it as the chamber contracts. The pumping chamber does not need sealing oil. That is useful when oil contamination would compromise the work or when process vapors would degrade oil. General-purpose diaphragm pumps and chemistry-rated models may use different diaphragms, valve materials and head materials. For an approved condensable vapor application, a chemistry model may also specify a ballast setting or pair with a separator and condenser. The VACUUBRAND MD 4C NT datasheet shows the sort of model-specific information to look for; its ratings do not apply to other diaphragm pumps.

In an oil-sealed rotary vane pump, sliding vanes on an off-center rotor partition the chamber. Rotation increases the volume on the inlet side and compresses the gas before discharge. Oil lubricates and seals internal clearances. Putting two compression stages in series can lower the attainable inlet pressure; stage count alone does not predict speed at the working point. This design is worth evaluating when an oil-free candidate cannot meet the required pressure under load and when the process can tolerate the oil and exhaust requirements. Check whether the pump has an inlet isolation or anti-suckback valve, what happens on power loss, and whether your process needs separate protection. A valve is not a substitute for following the shutdown procedure.

During compression, a condensable vapor can reach its saturation pressure and liquefy inside the pump. Gas ballast introduces a specified noncondensable gas into the compression stage so the discharge valve can open at a lower vapor partial pressure. That can help expel vapor within the model’s documented limits, usually at the cost of a higher lowest attainable inlet pressure. Use the manufacturer’s limits and ballast settings; adding air is not a remedy for flammable or reactive mixtures. An inlet trap protects against carryover; an exhaust mist separator removes oil droplets from the exhaust stream. Neither makes hazardous exhaust safe to vent into the room. See Pfeiffer’s rotary vane manual for these mechanisms and their limits.

Work a sizing example

Suppose a dry-air teaching setup must hold 200 mbar absolute while a metered air bleed introduces 60 mbar·L/s. This is a hypothetical requirement, not a result measured on a product. Assume steady, approximately isothermal operation, negligible additional gas load, and that the pump’s no-load pressure is small compared with the target. The required effective speed at the vessel is then approximately 60 mbar·L/s ÷ 200 mbar = 0.30 L/s, or 18 L/min. If a known bleed is given in standard liters per minute rather than pressure-volume units, confirm its reference pressure and temperature before converting.

Now consider a candidate that delivers 0.60 L/s at the actual inlet operating pressure. Assume the connecting assembly can be represented by a conductance of 0.60 L/s over the pressure drop and for the gas in this example. In a one-path, steady-flow approximation, 1/S_eff = 1/S_pump + 1/C. This gives 0.30 L/s at the vessel. Equivalently, the same throughput would give about 100 mbar at the pump inlet and a 100 mbar drop across the line. So the speed to verify on the pump curve is at approximately 100 mbar in this example, not 200 mbar. Real conductance may depend on pressure, gas, and the fittings; use an appropriate model or measured flow data rather than assuming a hose has constant conductance across the entire pump-down. The example meets the target only on paper and leaves no allowance for added gas load or uncertainty.

Use the calculation to ask a concrete question of a manufacturer: “With my inlet pressure, gas composition, operating settings and exhaust arrangement, can this model continuously handle the required throughput?” For intermittent pump-down, separately check how long it takes the proposed assembly to reach the target. Comparing those answers often gives a better basis for choosing between two models than ranking their headline flow figures.

Read the datasheet and check compatibility

Read the speed curve with the ballast state, electrical frequency and test gas in mind. Check maximum permitted inlet pressure for sustained operation, allowable exhaust backpressure, cooling clearance, and whether the published duty applies to the full working range or only part of it. For long runs, ask about unattended operation, restart behavior, overtemperature protection, and planned service. Motor thermal protection does not establish a continuous-duty rating for a particular inlet load.

Then trace every surface the gas or liquid could reach, including the inlet fitting, hose, seals, trap, pump head or oil, and exhaust accessories. Record the exact connector standard: a matching nominal diameter can hide differences between a flare fitting, ACME thread, hose barb and flange. Verify the vacuum, temperature and chemical limits for each component. Select a gauge with a useful range at the vessel; if the sensor depends on thermal conductivity, check the gas for which its indication is calibrated. See the Büchner filtration guide if the task is filtration: the filter and receiver have their own operating limits, and additional vacuum can change cake behavior without improving the result. The filtration equipment listings can help identify component types; check the documentation for the exact assembly before applying vacuum.

The TA350 single-stage rotary vane pump listed by Viking Lab Supply is an example of how to read a catalog entry. The catalog and Kozyard’s TA350 page agree on 3.5 CFM nominal free-air displacement (about 99 L/min) and an oil reservoir with a sight glass. They disagree on the ultimate pressure, electrical and motor ratings, oil capacity, and details of the inlet fitting. Those conflicting figures cannot identify what the supplied unit supports. Confirm the nameplate, inlet hardware and revision-specific manual before connection or use. The shared 3.5 CFM figure alone does not establish vessel performance, vapor compatibility or duty rating.

Run a dry checkout

Use the following checkout as a supervised teaching exercise with clean, dry air and an empty, vacuum-rated vessel. Use the institution’s approved procedure and the equipment manuals. It does not establish suitability for a hazardous sample, nor is it a validated regulatory or clinical method.

1. Prepare the apparatus. Verify the vessel, trap, gauge and lines are rated for the intended pressure differential and temperature. Inspect glass for chips, scratches and cracks; use the required shield and eye protection. Support the vessel and trap independently. Place the pump on a stable surface, check the correct supply and cooling clearance, and verify its approved lubricant and fill level if applicable. Arrange the exhaust as required by your laboratory; do not block it or add a restrictive filter without verifying backpressure.
2. Make and verify the connections. Fit a gauge on the vessel side, using the right range and units. Install the approved inlet trap in the correct orientation and arrange the isolation and vent controls according to the equipment instructions. Confirm the start-up valve positions and use the manufacturer’s start-up sequence. Record the actual configuration, including the gauge, hose, trap and ballast position, so a later result can be compared on equal terms.
3. Observe the pressure and time. With the system shielded, evacuate according to the approved procedure. Record time to reach the specified band and the pressure there. If the setup is already designed for a clean-air metered bleed, apply its approved load and check whether the vessel pressure stabilizes in the band. Observe pump temperature or alarms according to the manual. An empty-vessel result without a bleed confirms only dry evacuation; it does not test the intended process load.
4. Shut down and handle waste. Follow the model’s isolation, vent and shutdown order so oil or trap contents cannot be drawn into the vessel. Confirm the vessel is back at atmospheric pressure before opening it. Let hot surfaces cool before servicing. In routine sample work, label and collect trap condensate and used pump oil according to their actual contaminants and your laboratory’s waste procedure. Do not discharge contaminated oil or condensate into a drain. Consult Princeton EHS and Cornell EHS for vacuum shielding, trapping and exhaust guidance.

Troubleshoot selection problems

The empty vessel reaches target, but the approved test load pushes it above the band.

Compare the specified load with the curve and the estimated effective speed. A partly closed valve or restrictive trap may explain a low vessel speed even with a suitable pump. If the assembly behaves normally, revisit the selection or reduce the approved gas load rather than changing the target mid-test.

Oil becomes cloudy after vapor service.

Stop and use the manual and local procedure to assess contamination. Verify the intended trapping and gas ballast arrangement and the model’s vapor-load limits. Service and dispose of contaminated oil as required before returning the pump to use.

The pump gets hot or repeatedly trips during a long run.

Stop and isolate it as directed. Check cooling, supply, inlet load, exhaust restrictions and permitted duty. Do not reset a protective trip repeatedly. If dry performance is poor as well, use the separate weak vacuum troubleshooting guide for leak and restriction checks before replacing the pump.

Frequently asked questions

How do microns on a vacuum gauge relate to pressure?

In vacuum specifications, a micron commonly means a micrometer of mercury: 1 micron Hg = 0.001 Torr, approximately 0.133 Pa. A hypothetical 75 micron reading therefore corresponds to about 10 Pa or 0.10 mbar absolute. Confirm the units on the instrument before recording a result.

Can a small diaphragm pump be a better choice than a larger rotary vane pump?

Yes, if the diaphragm model reaches the required pressure at the expected load and is documented for the process. It can avoid oil maintenance and reduce the chance of oil reaching the apparatus. A diaphragm pump still has wear parts, and an oil-free chamber does not imply chemical resistance.

Does a dry checkout qualify the pump for a new sample?

No. Wet loads can change pressure, contaminate oil, or exceed a trap or pump’s capacity. Review the sample’s hazards and the expected gas and vapor load separately, and use the laboratory’s approval and operating procedure before changing service.

References

1. Pfeiffer Vacuum, Vacuum generation: pumping speed, throughput and pressure limits; vacuum calculations.
2. Sheng-Jui Chen, System for measuring vacuum-pump performance using the standard throughput method, Acta IMEKO 7(1), 2018.
3. VACUUBRAND, MD 4C NT chemistry diaphragm pump technical data.
4. Pfeiffer Vacuum, DuoVane 1.6 operating instructions, pump mechanism and gas ballast.
5. Kozyard, Kozyvacu TA350 product specifications.
6. Princeton University EHS, Pressure and Vacuum Systems.
7. Cornell University EHS, Laboratory Safety Manual: Physical Hazards.