Five Parameter Mistakes Pump OEMs Make When Specifying Mechanical Seals

Every pump an OEM ships leaves the floor with a seal already mounted, pulled from a spec sheet somebody filled in. Drawings mostly get the size correct. The duty kills the seal. Five parameter gaps show up on those sheets over and over, and they reach the field before anyone bolts the baseplate down.

1. Sizing the seal from shaft diameter alone

Shaft diameter gets the catalog to build the right bore. It never says how much axial room a cartridge actually burns inside the stuffing box. That seal needs length the box has to give. A shallow box, or a bore that runs tight, leaves it with nowhere to sit. Spec the seal off the shaft line, skip the box depth, and the result is a part that passes the drawing and fights the pump at install. Read the box dimensions off the pump model before that seal lands on the bill of materials.

2. Writing the rated pressure and forgetting the transients

Operating pressure is the one number the data sheet always carries. The seal sees more than that. The startup kick arrives first, then the deadhead when a valve stays shut, then the backpressure nobody designed the process to hold. Rate a single seal to the steady figure and it can crack open on a transient nobody asked about. The box pressure at start and stop goes on the spec. It should not land first in the failure report.

3. Quoting the fluid temperature instead of the seal chamber temperature

Line fluid carries a temperature the engineer reads off the P&ID. The seal chamber runs hotter, usually by enough to push the elastomer past its ceiling while the fluid sheet still reads safe. Heat works back from the bearing. Discharge fluid carries its own heat, and a sealed chamber holds both. A boiler feed pump sits near steam temperature while the suction line still reads cold, and the seal sees the hotter one. Build the spec on line temperature and the secondary seal gets handed a temperature it never agreed to. The chamber reading, or a margin somebody wrote down, is the figure the seal desk needs.

4. Relying on nameplate speed without checking face velocity

Nameplate speed is the number the motor carries. The seal lives or dies at the face velocity. Face velocity rises with shaft speed, and it climbs a second time with seal size. A VFD past nameplate, or a shaft fattened for horsepower, drives the face past what the design can throw off as heat. A small shaft turned fast beats a big one turned slow. A spec that lists only nameplate speed misses the number the faces actually watch.

5. Naming the main fluid and leaving out what rides with it

Process fluid is the one the spec names. It leaves out the trace abrasive in water that was supposed to be clean, the second chemical that shows up during clean-in-place, and the flush water that looked harmless on paper and carries its own grit. A seal meets every one of those, not the headline on the datasheet. Abrasion, a stray solvent that passed the compatibility check on paper, and flush water nobody graded take more seals than the primary fluid. The materials call rests on the full wetted list, flush medium included.

None of the five needs a different seal model. What each one needs is the right number on the spec. FBU builds cartridge seals across the API 682 arrangements for pump OEM programs, and the data sheet for that exact model holds the final box, pressure, temperature, speed, and fluid limits. Fill those five in and a drawing match becomes a duty match.

Quick reference

Box geometryShaft size picks the bore, not the fit. Box depth and bore decide whether the cartridge actually seats. Both come off the pump drawing, not the shaft line alone.

PressureSteady differential is only part of it. Startup surge, deadhead, and backpressure all hit the faces. Put the box pressure at start and stop on the spec.

TemperatureMost jobs run the line cooler than the chamber. The faces eat the hotter figure. A chamber reading or a written margin keeps the rubber alive.

SpeedNameplate speed is not the face speed. Peripheral velocity climbs with speed and size, and a VFD or fat shaft blows past the limit. Spec the highest speed the seal will see.

FluidEverything that enters the box reaches the seal. Trace grit, a second chemical, flush solids, all of it. The full wetted list, flush medium included, goes on the spec.

Frequently asked

What is the most overlooked dimension when an OEM specifies a seal?
The stuffing box depth and bore. Shaft diameter sets the seal size, yet a cartridge needs axial room the box must give. Size it off the shaft line and the drawing clears a part the pump rejects.

Why does rated pressure fall short for seal selection?
The seal sees more than the steady differential. Startup surge, deadhead against a shut valve, and process backpressure all reach the faces. Rate a single seal to operating pressure only and it can open on a transient it was never qualified for.

Is line fluid temperature enough to pick the elastomer?
No. The seal chamber runs hotter than the line fluid in most services, and that gap can push the elastomer past its ceiling while the fluid sheet still reads safe. The chamber temperature, or a margin somebody documented, is the figure the seal desk needs.

Does nameplate speed decide the seal?
Not alone. The faces care about peripheral velocity, which grows with both shaft speed and seal size. A VFD ramp past nameplate or a shaft fattened for horsepower can drive the face past what the design sheds as heat. The highest speed the seal will actually see goes on the spec.

Why list only the main fluid on a seal spec?
The seal meets everything that reaches the box. Trace abrasives in clean water, a second chemical from clean-in-place, and grit in the flush water take more seals than the primary fluid ever does. The material call rests on the full wetted list, flush medium included.