A rotary vane vacuum pump uses an offset rotor whose sliding vanes sweep gas from inlet to exhaust, sealed and lubricated by oil. It reaches deeper vacuum than water-based pumps and holds it steadily. The oil is also its weakness: solvent vapour dissolves into it, and contaminated oil loses vacuum fast.
One thing up front, because it shapes everything below: we do not build rotary vane pumps. We build water-ring circulating water vacuum pumps, and we sell them to labs deciding between the two. So we have no reason to oversell rotary vane technology, and none to pretend our own pumps do things they cannot.
Table of contents
- How does a rotary vane vacuum pump actually work?
- Why is the oil both the trick and the trap?
- How does a water-ring pump move gas differently?
- Rotary vane or water-ring: which survives solvent vapour?
- How much vacuum do you actually need?
- What do the pumps we build actually pull?
- How do you keep whichever pump you buy alive?
- FAQ
How does a rotary vane vacuum pump actually work?
Picture a cylindrical housing with a rotor mounted off-centre inside it, so the rotor almost touches the wall at one point and leaves a crescent-shaped gap everywhere else. Slots in the rotor hold flat vanes that slide in and out. As the rotor spins, centrifugal force throws the vanes outward, and they stay in contact with the wall as the gap widens and narrows.
Each pair of adjacent vanes traps a pocket of gas. On the inlet side the pocket is expanding, so gas rushes in from your system to fill it. As the rotor carries that pocket toward the narrow point it shrinks, squeezing the gas until its pressure exceeds atmospheric and pushes open the exhaust valve. Then the vane pair comes back around and does it again.
That is the whole idea: positive displacement. The pump does not suck. It removes discrete volumes of gas and refuses to let them back in. Vacuum is what is left when you keep doing that.
The reason this design goes deep is the sealing. The clearance between vane tip and housing is tiny and filled with oil. Oil is nearly incompressible and flows into every gap, so almost nothing leaks backward from the high-pressure exhaust side to the low-pressure inlet. Backflow limits every vacuum pump. Kill the backflow and you go deeper. Some designs stack a second pumping stage so the first exhausts into the second rather than to atmosphere, cutting the pressure the seal must hold against.
Why is the oil both the trick and the trap?
The oil does three jobs at once: it seals the clearances, lubricates the sliding vanes, and carries heat away from the compression zone. Take it out and the pump does not just get weaker — it destroys itself in short order.
Now think about what a lab actually pumps. Pulling vacuum on a rotary evaporator distilling dichloromethane, a filtration flask of ethanol-wet cake, or a reactor stripping methanol, you are not pumping air. You are pumping air plus a large load of organic vapour. That vapour hits the compression zone, gets squeezed, and a good fraction condenses into liquid right there in the pump — compressing a vapour is one of the two classic ways to condense it.
Where does that condensate go? Into the oil. Solvent dissolves in and thins it. Thinner oil seals worse, so ultimate vacuum degrades. It lubricates worse, so the vanes wear. And the dissolved solvent has its own vapour pressure inside the pump, setting a floor on how low the pump can pull no matter how good the mechanism is. The datasheet was not lying — your application changed the machine.
Water is worse than most solvents, because water and hydrocarbon oil do not mix. Pumped water vapour condenses and emulsifies the oil into a milky mess that seals poorly and corrodes internals. That vacuum pump water problem is the most common complaint we hear from labs running oil pumps on aqueous work.
There are answers. A gas ballast valve bleeds controlled air into the compression stage so vapour never reaches its condensing pressure and leaves with the exhaust. A cold trap freezes vapour out before it arrives. Both work — and both are things someone must remember to do, correctly, every run, forever. Gas ballast costs ultimate vacuum while open; a cold trap needs coolant and emptying. Neither survives a busy lab with rotating staff for long.
How does a water-ring pump move gas differently?
A water-ring pump — the type we build — starts from the same off-centre rotor idea and removes the oil entirely. The impeller spins inside a housing partly filled with water, and centrifugal force flings that water into a ring against the wall. Because the impeller is offset, the ring is not concentric with it, so the space between the blades and the water surface grows on one side and shrinks on the other. The water ring becomes the seal: a liquid piston. Gas is drawn into the widening cells, carried around, compressed by the closing water surface, then pushed out.
Same positive-displacement logic — but the sealing fluid is water, which gets circulated, cooled, and simply changed when dirty. That swap cuts both ways.
The good: solvent vapour condensing inside the pump is a non-event. It goes into the circulating water, which you drain and refill. Nothing emulsifies because there is no oil. The water also acts as a direct-contact condenser, so the pump handles vapour loads that would flood an oil pump.
The bad: water has a vapour pressure of its own, and it is inside your pump on purpose. That sets a hard floor on the vacuum a water-ring pump can reach, and no engineering gets past it. It is thermodynamics, and it applies to our pumps exactly as it applies to everyone else’s.
Rotary vane or water-ring: which survives solvent vapour?
Here is the comparison you rarely see published, because most pages about vacuum pumps are written by companies that only make one kind.
| Question | Rotary vane (oil-sealed) | Water-ring (what we build) |
|---|---|---|
| Sealing fluid | Oil | Circulating water |
| Depth of vacuum | Deeper — oil seals better than water and adds no vapour pressure of its own | Limited by the vapour pressure of its own water; SHZ-95B is rated to 0.098 MPa maximum vacuum |
| Solvent vapour in the stream | Contaminates and thins the oil; ultimate vacuum degrades run by run | Vapour condenses into the circulating water, which you drain and refill |
| Water vapour in the stream | Emulsifies the oil unless gas ballast or a cold trap is used every run | Irrelevant — the sealing fluid is already water |
| Routine maintenance | Oil changes, oil level checks, gas ballast discipline, mist filter | Change the tank water; keep it cool |
| Consequence of neglect | Vacuum quietly degrades; the pump can seize | Vacuum quietly degrades as the water warms and loads up; refill restores it |
| Serving several stations | One inlet; needs a manifold you build yourself | SHZ-95B ships with 5 taps, 10 L/min per tap |
Read that as a fork, not a scoreboard. If your process is dry — freeze drying downstream of a good trap, gloveboxes, vacuum ovens on already-dry material, leak testing — a rotary vane pump is the correct machine and we will not pretend otherwise. It goes lower, holds steadier, and the oil never sees anything that hurts it.
If your process is wet in any sense — rotary evaporation, aqueous filtration, solvent stripping, degassing anything containing water — the water-ring pump is usually the one still working in year three. Not because it is a better pump in the abstract, but because your process is trying to destroy the other one’s sealing fluid, and it cannot destroy water with water.
The failure nobody warns you about is the middle case: a lab buys a rotary vane pump for its vacuum rating, runs it on a rotary evaporator without a trap, and concludes months later that the pump is defective. It is not defective. It is full of dichloromethane. Everything in the vacuum pump range we build is aimed at that middle case, because that is where labs keep getting burned.
How much vacuum do you actually need?
This is where most selection goes wrong. People buy on the ultimate vacuum number, as if deeper is always better. It is not. It is a number that has to clear your requirement, after which it stops mattering. The honest framework is short:
- Start with the boiling point you need, not the pressure. Your process defines a temperature you must not exceed — thermal degradation, a bath limit, a solvent you want off gently. That temperature plus the solvent’s vapour pressure curve defines the pressure you need. Work backwards from chemistry, not from a spec sheet.
- Check that requirement against the water floor. If the pressure you need sits above the vapour pressure of the water inside a water-ring pump at your ambient temperature, a water-ring pump does the job. If it sits below, the pump physically cannot, and no vendor claim changes that. Routine solvent evaporation is usually above the floor; deep drying and high-vacuum work is below it.
- Then check flow, separately. Ultimate vacuum tells you the floor. Pumping rate tells you how fast you get there and whether you hold it against your vapour load. A pump with a great floor and a small throat will crawl on a system actively boiling solvent at it.
- Then count stations. One evaporator is one inlet. Four sharing a pump is a manifold problem, and the pump needs total capacity — not per-station capacity — to serve them at once.
- Then decide what your sealing fluid will be eating. Everyone skips this step. It decides whether the pump is still good in year three.
Notice what is not on that list: brand, and a headline vacuum rating divorced from context. For that reasoning applied to a specific rig, the way we match a vacuum pump and chiller to a rotary evaporator walks the same logic through a real configuration, and our equipment selection guide covers these questions across the rest of the line.
What do the pumps we build actually pull?
We have spent this page describing someone else’s pump. Here are our own numbers, so you can hold us to the same standard.
| Parameter | SHZ-D(III) | SHZ-95B |
|---|---|---|
| Pump type | Water-ring circulating vacuum pump | 5-tap water-ring vacuum pump |
| Power | 180 W | 550 W |
| Flow rate | 60 L/min | 80 L/min |
| Suction capacity | 10 L/min (single tap) | 10 L/min per tap; 50 L/min total |
| Maximum vacuum | — | 0.098 MPa |
| Tank size | — | 50 L |
| Voltage | 220 V | 220 V or 110 V |
The two dashes are deliberate. We publish the maximum vacuum and tank volume we measured on the SHZ-95B, and leave blank what we have not measured the same way on the SHZ-D(III). A blank is more useful than a number we invented.
The benchtop SHZ-D(III), drawing 180 W and circulating 60 L/min through a single 10 L/min tap, is built for one station — one evaporator, one filtration setup, one reactor. Buying more than that is buying bench space you do not get back.
When several stations run at once, total capacity takes over from per-tap capacity. The vertical SHZ-95B, with its 50 L tank, 550 W motor, 5 taps at 10 L/min each and 50 L/min total suction, exists for that case, and it is what we point at 5 L and larger equipment. The 50 L tank does more work than it looks like: a bigger volume of circulating water absorbs more condensed vapour and more heat before warming enough to lift the vacuum floor. Tank size is a vacuum-stability spec disguised as a plumbing spec.
Skip the SHZ-95B if you have one evaporator and no plans for a second. Skip the SHZ-D(III) if three people already queue for the same pump.
How do you keep whichever pump you buy alive?
The maintenance logic follows from what the sealing fluid is — the thread running through this whole page.
Running a rotary vane pump: put a trap in front of it, open the gas ballast when pumping vapour instead of leaving it shut because it costs vacuum, and change the oil on condition rather than on a calendar. Cloudy oil means water. Discoloured or thin oil means solvent. Either way the oil has stopped being a seal and is now just liquid taking up space. Checking the sight glass takes three seconds and predicts nearly every rotary vane failure we hear about.
Running one of ours: change the tank water when it stops being clear, and watch its temperature. Warm circulating water has a higher vapour pressure, which raises the floor — so vacuum gets worse on a hot afternoon in a room with no air conditioning, then recovers after a refill. Users report this as intermittent pump failure. It is the pump behaving as physics requires.
For both: fix your leaks before buying a bigger pump. Oversizing to compensate is the most expensive way to avoid finding a bad joint. Check tubing internal diameter too — a long narrow line between pump and vessel throttles flow whatever the pump is rated at. Our equipment FAQ covers the tubing and connection questions that come up most.
FAQ
Is a rotary vane vacuum pump better than a water-ring pump?
For dry applications needing deep vacuum, yes — the oil seal reaches lower and holds steadier than any water-sealed design can. For streams carrying solvent or water vapour, usually no: that vapour condenses into the oil and degrades the vacuum the pump was bought for. It is a fit question, not a quality ranking.
Can I use a rotary vane vacuum pump on a rotary evaporator?
Yes, but only with a cold trap in front of it and disciplined gas ballast use, because rotary evaporation sends a continuous solvent vapour load straight into the pump. Without that protection the oil takes the solvent in and ultimate vacuum falls off within months. Labs that will not maintain the trap are better served by a water-ring pump.
Why does my vacuum pump get worse when it is pumping water?
In an oil pump, water vapour condenses under compression and emulsifies the oil, which then seals poorly and corrodes internals. In a water-ring pump, pumped water is harmless, but warming circulating water raises its vapour pressure and lifts the vacuum floor. Changing the tank water restores an oil-free pump; contaminated oil must be drained and replaced.
How many pieces of equipment can one vacuum pump serve?
Count total demand, not station count. Our SHZ-95B provides 10 L/min per tap across 5 taps for 50 L/min total, so five stations each needing 10 L/min are within spec. The single-tap SHZ-D(III) at 10 L/min serves one. Splitting a single-tap pump across a manifold divides that capacity, it does not multiply it.
Where this leaves you
If your vacuum stream is dry and you need to go deep, buy a rotary vane pump, budget for the oil, and use the gas ballast. We do not build one and will not sell you a compromise instead.
If your stream carries solvent or water — most rotary evaporation, filtration and reactor stripping in a working lab — start from what your sealing fluid will be eating, then size on total flow, then count stations. One station at a time means the 180 W, 60 L/min SHZ-D(III). Several at once, or 5 L and larger vessels, means the 550 W SHZ-95B with its 50 L tank and 5 taps. Tell us the solvent, the temperature ceiling and how many stations share the pump, and we will tell you which of the two fits — or that it is a rotary vane job and you should buy one elsewhere.