When worn vanes should be replaced: back-calculating effective pumping speed from pump-down time

Three statements come up whenever vane wear is discussed on site: the sound changed, the oil went black, the body runs hot. All three are hints; none of them answers “how much is left”. Only one number does that — the effective pumping speed the pump still delivers on a fixed chamber — and it can be back-calculated with a stopwatch, a gauge and one pump-down run, without opening the pump. The five-signal checklist is covered in five measurable signals and the replacement point; this article turns those signals into threshold numbers.

1. The rated figure is not the on-site figure

The DPL-V300B datasheet carries two columns: nominal pumping speed 280 m³/h and pumping speed 240 m³/h at 50 Hz. The nominal value is swept volume times rotational speed; 240 m³/h is what the pump measures under defined conditions. Add pipework, valves and an inlet filter, and what reaches the chamber is a third figure again.

So wear is never judged against 280 or 240 directly. It is judged against the pump-down time on the same system, the same chamber and the same target pressure — the one quantity that wraps pump, piping and chamber together and repeats.

2. Step one: turn pump-down time into effective pumping speed

For an isothermal pump-down:

t = (V / S) × ln(p₀ / p₁)

with V the chamber volume in m³, S the effective pumping speed in m³/h, and p₀, p₁ the start and target pressures in mbar absolute. Solving for speed:

S = (V / t) × ln(p₀ / p₁) × 3600

Take a 3 m³ foreline chamber on a coating line, pumped from atmosphere (1013 mbar) to 10 mbar each cycle.

Commissioning run, 215 s:

S = (3 / 215) × ln(1013 / 10) × 3600 = 0.01395 × 4.618 × 3600 = 232 m³/h

That is 96.7% of the 240 m³/h datasheet figure; the missing 3.3% sits in pipe, valve and flange resistance and is normal. This run becomes the baseline.

Two years later, same chamber, same target, 268 s:

S = (3 / 268) × 4.618 × 3600 = 186 m³/h

77.5% of 240 m³/h — a 22.5% loss, past the point of “worth a look”.

In production terms: 53 s extra per cycle. At 30 cycles a day that is 26.5 minutes, and over 250 working days about 110 hours. The pump never stopped and never alarmed; the cost appeared in cycle time.

3. Step two: discount the piping before blaming the vanes

Pump speed and chamber speed are separated by pipe, bends, valves and filters, in series:

1 / S_eff = 1 / S_pump + 1 / C

with C the conductance of the line. For a 240 m³/h pump:

Line conductance C	Effective speed at the chamber
240 m³/h	120 m³/h
120 m³/h	80 m³/h
60 m³/h	48 m³/h

When C is of the same order as the pump speed, half of it stays in the pipe. Before opening the pump, confirm three things: the pipework has not changed since the baseline, the inlet filter is not blocked, and the gas ballast position matches the baseline. Ballast is a large effect — from V0040A to V0100 the ultimate pressure is 0.1 mbar closed and 0.5 mbar open, so comparing one run with ballast open against one with it closed is meaningless.

If piping, filter and ballast all check out and the speed is still down by a fifth, the gap is inside the pump: vane tips widen, vanes shorten radially, compressed gas slips back to the suction side, and volumetric efficiency falls with it. The gap-side geometry is set out in rotor and vane clearance: what a wider gap costs in speed and life.

4. Step three: separate pump backflow from system leakage

A speed loss still leaves one question open: is it the pump or the chamber? The rate-of-rise test answers it in two minutes.

Close the inlet valve, stop the pump, log the pressure rise:

Q = V × Δp / Δt

A 0.5 m³ chamber (500 L) rising from 0.5 mbar to 3.5 mbar in 60 s:

Q = 500 × 3 / 60 = 25 mbar·L/s

That is a leak or a large outgassing source to be found — flanges, bellows, viewport seals, moisture carried in on workpieces.

Now the pump-side figure: speed fell from 240 to 186 m³/h, a difference of 54 m³/h, or 15 L/s. At a 10 mbar working point that equals 150 mbar·L/s of equivalent backflow. No flange leak delivers that; the number points at internal clearance.

5. Three thresholds

Criterion	Value	Action
Effective speed vs baseline	down 15% (240 → 204 m³/h)	Check oil level, oil grade, ballast, filter, piping; only then suspect clearance
Effective speed vs baseline	down 25% (240 → 180 m³/h)	Schedule the vane change
Blanked-off ultimate pressure	0.08 mbar class only reaching about 0.8 mbar	Strip and inspect, no need to wait for a speed loss

Three hard signals also trigger replacement on their own: silver-white or black fines in the drained oil; metallic knocking as vanes rattle in their slots; oil temperature 10 ℃ or more above baseline at the same duty. Oil charge is 8.5 / 11.5 L on the DPL-V300B and 20 / 23 L on the DPL-V630B — record it in the same table as the pump-down time.

6. Four things to do in the same maintenance window

Replace all vanes on a rotor as a set. A few tenths of a millimetre of height difference loads the new vane harder and pushes wear onto the remaining vanes and the bore.
Clean carbon from the slots and inspect the bore. A scored bore will wear a new set straight back to where the old one was.
Change oil, mist separator, inlet filter and end-cover seals together. Splitting them means paying the labour three times.
Repeat the same pump-down curve and write the new figure into the baseline table. Without a baseline, the next judgement is back to listening for sounds.

Oil-free vane pumps follow the same logic. The DPL-V25 runs a dry chamber at 25 m³/h with 850 mbar gauge (about 150 mbar absolute); wear shows as shorter carbon vanes, slower pump-down and a higher blanked-off pressure. The criteria are still pump-down time and ultimate pressure — not running hours, which say how long the pump ran, not how much material it lost.

7. Keeping a usable baseline

Date	Chamber	From → to	Time	S_eff	Ambient	Oil temp	Note
2026-01	3 m³	1013 → 10 mbar	215 s	232 m³/h	24 ℃	62 ℃	new pump, new filter
2026-06	3 m³	1013 → 10 mbar	236 s	211 m³/h	27 ℃	66 ℃	after filter change
2026-11	3 m³	1013 → 10 mbar	268 s	186 m³/h	25 ℃	71 ℃	window to be scheduled

One run per quarter, five minutes each. Two consecutive drops on the sheet and the vanes should already be on order. Theoretical times can be checked first with the site selection calculator and compared with what the pump actually does.

One note for export projects: at 50 Hz and 60 Hz the same model runs at different speed with a different motor configuration, and some models are not suitable for 60 Hz regions. Confirm the configuration before ordering, and never judge a 60 Hz pump against a 50 Hz baseline.

Frequently asked questions

How many running hours should a vane set last?

There is no fixed hour count. Use the pump-down baseline on the same chamber: look into it at a 15% effective-speed loss and schedule the change at 25%. On a 3 m³ chamber the DPL-V300B going from 215 s to 268 s is exactly that 25% case.

Can I replace only the vanes that look worn?

No. A few tenths of a millimetre of height difference loads the new vane harder and moves wear onto the remaining vanes and the bore, so the pump comes apart again sooner.

Does slower pump-down always mean worn vanes?

Not always. Rule out three things first: changed pipework, a blocked inlet filter, and a different ballast position from the baseline. If those are clear, the rate-of-rise test decides — a rise around 25 mbar·L/s belongs to the system, not the pump.

Do oil-free carbon vanes follow the same rule?

Yes. Oil-free vane pumps such as the DPL-V25 are judged on pump-down time and blanked-off ultimate pressure just like oil-sealed ones; shorter carbon vanes show up as slower pumping and a higher ultimate pressure, with no oil level or oil grade to confuse the picture.

What else should be replaced at the same time?

Oil, mist separator, inlet filter and end-cover seals, plus cleaning the rotor slots and checking the bore. Oil charge is 8.5 / 11.5 L for the DPL-V300B and 20 / 23 L for the DPL-V630B.