Troubleshooting Diaphragm Pump Vibration in Rigid Piping

The motor starts and the entire skid moves. Pressure gauges blur so badly you cannot read them, and the shaking is worse at the pump inlet than at the discharge. Stainless tubing runs hard-piped from the pump with no hose, no dampener, and the pump sits on vibration mounts bolted to a base plate. Kerosene is the fluid.

That symptom set is not one fault. It is three possible faults that look identical from two metres away: unrestrained reciprocating inertia forces, suction-side acceleration head and cavitation, and an acoustic or structural resonance in the tubing. Work them in order. Guessing costs you a rebuild you did not need.

Read What the Skid Is Telling You

First, settle what kind of pump you have. A motor-driven diaphragm pump is a reciprocating pump. It behaves like a piston or plunger pump: discrete stroke volume, discrete displacement pulse, and unbalanced inertia forces along the plunger axis. The "it's not a piston pump so it doesn't need a dampener" reasoning is wrong and it is the single most common reason these skids ship without pulsation control. Only an air-operated double-diaphragm pump escapes that logic, and even then only partly.

Second, note where the motion is worst. Inlet worse than outlet points at the suction side. Suction lines are usually longer, larger in swept volume relative to available pressure, and carry the least margin above vapour pressure. Kerosene is low viscosity and low density, so it accelerates readily and gives up very little friction damping.

Third, note what is actually moving. Gauges vibrating hard while the main run moves less means the gauges are on unsupported stubs acting as cantilevers, not that the fluid pulsation is worst at the gauge. Fix that separately, and do not use the gauges as your vibration indicator.

Check 1: Did It Shake on Day One?

Start here. This one question splits the tree and costs nothing.

• It has vibrated since commissioning. This is a design problem: pulsation, acceleration head, mounting stiffness, or piping acoustics. Go to Check 2.
• It ran smoothly and started vibrating recently. This is a mechanical or wear problem. Pulsation dampeners will not fix it and buying them wastes time and money.

For the recent-onset branch, look at loose hold-down bolts and cracked grout, a failed or stretched diaphragm, a broken or fouled check valve spring in the suction or discharge valve, worn crosshead or connecting rod bearings, motor-to-pump coupling misalignment, and gas or vapour accumulation in the pump head. A single failed check valve in a duplex or triplex pump doubles the pulsation amplitude at once and is a classic "it was fine last month" cause.

What you see	Most likely cause	First reading to take
Whole skid moves in one axis, in step with strokes	Unbalanced reciprocating inertia forces reacting into soft mounts	Overall vibration at pump feet vs base plate vs floor
Inlet worse than outlet, rumble or crackle in the pump head	Acceleration head exceeding available NPSH; cavitation	Suction pressure at the pump inlet, running, close-coupled
Steady hammer at stroke frequency, worst on one pipe span	Pressure pulsation with acoustic or structural resonance	Frequency spectrum: compare stroke frequency and harmonics
Only gauges and small-bore stubs shake	Cantilevered branch resonance	Hand-check stub stiffness with pump stopped
Onset was sudden after months of clean running	Failed check valve, torn diaphragm, loose bolts, worn bearings	Discharge pressure ripple and bolt torque check

Check 2: Pull the Vibration Mounts and Bolt It Down

Vibration mounts under a reciprocating pump are usually the problem, not the cure. Isolation mounts work by lowering the mounted natural frequency well below the exciting frequency. A slow-speed reciprocating pump excites at stroke frequency and its low harmonics, which is exactly the range where soft mounts amplify instead of isolate. You have given the unbalanced plunger inertia force something to push against that moves.

The check: with the pump running, compare motion at the pump feet, at the base plate, and at the structure below it.

• Pump feet move a lot, base plate and structure move little. The mounts are the amplifier. Remove them and bolt the pump solid.
• Base plate and pump move together, structure is quiet. The base plate is too light or too flexible. Anchor it to a heavier foundation or stiffen it.
• Everything including the floor or frame moves. The exciting force is larger than the foundation can absorb. Go to Checks 3 and 4 before you add steel.

The rule of thumb that holds up in the field: if you cannot balance the forces, bolt the machine to something big enough to swallow them. Use real hold-down bolts, flat and shimmed mounting pads, and full contact under the feet. Vibration mounts and flexible pipe belong on a pump only when the foundation is what is shaking the pump, which is not the case here.

Things that vibrate come loose and break. Every soft element you add is another item that needs re-torquing.

Check 3: Work the Suction Side First

Inlet vibration worse than outlet vibration is your strongest clue and it points to acceleration head. On a reciprocating pump the entire suction column has to be accelerated and decelerated once per stroke. The pressure the pump must have available to do that is on top of friction and static losses, and it grows with suction line length, flow velocity, and pump speed.

The standard form is:

h_a = (L * V * n * C) / (K * g)

L = actual length of suction line (not equivalent length)
V = mean velocity in the suction line
n = pump speed, strokes per minute
C = constant for pump type (simplex / duplex / triplex, single or double acting)
K = constant for fluid compressibility
g = gravitational constant

Read C and K from the table in Cameron Hydraulic Data for your pump configuration and for a light hydrocarbon; do not carry over values used for cold water. Then compare:

NPSH_available = h_static + h_atm - h_vapour - h_friction - h_a

Against the pump's NPSH required on its curve, at the running speed. If h_a alone eats the margin, you have found it. Kerosene near ambient temperature has meaningful vapour pressure and no cushioning viscosity, so the margin disappears faster than water intuition suggests.

Fixes on this branch, in order of effect: shorten the suction line, increase suction line bore to drop V, raise the supply level, slow the pump down (h_a falls linearly with n), and fit a suction stabiliser close-coupled to the pump inlet. Increasing bore and shortening the run attack L and V simultaneously and are almost always the cheapest wins on a sample skid.

Check 4: Size Dampeners From Pump Data, Not From Pump Type

Before you call anyone, collect the data that decides whether a dampener helps and how big it is: pump make and model, simplex, duplex or triplex, single or double acting, stroke volume, strokes per minute, discharge pressure, suction pressure, suction and discharge line length and bore, and fluid temperature. "Diaphragm pump" on its own tells a dampener supplier nothing.

What a dampener does and does not do:

• It absorbs the flow variation between strokes, cutting residual pulsation to a fraction of the bare-pump value. That is the fluid-borne part of your problem.
• It does nothing about unbalanced mechanical inertia forces. If Check 2 said mounting, a dampener will not save you.
• It must be close-coupled. Every extra length of tubing between pump port and dampener leaves an unprotected column that still pulses.
• Fit one on the suction side too when acceleration head is the driver. Discharge-only installations are the usual reason "we fitted a dampener and it still shakes."
• Set the gas pre-charge to the fraction of mean line pressure the dampener manufacturer specifies, with the line depressurised, and re-check it after commissioning. A dampener at the wrong charge is a dead volume.

Verify bladder and diaphragm elastomer compatibility with kerosene when you specify. Give the supplier the full data set and let them run the sizing.

Check 5: Find the Resonance in the Tubing

Rigid stainless tubing hard-piped from a reciprocating pump is a tuned acoustic pipe with a pulsating source at one end. If a pipe run length happens to match a quarter or half wavelength of the stroke frequency or one of its harmonics, you get a standing wave and a local pressure amplitude far above what the pump generates. The rest of the system looks fine and one span shakes.

Calculate the excitation frequencies first:

f_stroke (Hz) = strokes-per-minute / 60
f_plunger (Hz) = f_stroke * number-of-plungers
Check harmonics: 1x, 2x, 3x of both

Then take a spectrum at the worst point and see which line dominates. A peak at running speed alone points back to mechanical imbalance or coupling. A peak at plunger frequency or its harmonics is pulsation-driven, and the cure is either the dampener from Check 4 or a change in the piping.

Piping cures, in order: add clamps to break long unsupported spans and shift the structural natural frequency, change the run length or add a volume bottle to detune the acoustic length, and re-route to remove dead legs. For the shaking gauges specifically, clamp the gauge stubs to a rigid bracket, keep the stubs short and stiff, and fit gauge snubbers or liquid-filled gauges so you can read them. Shaking gauges are a symptom of a cantilever, not a measure of system severity.

Decide the Flexible Hose Question

Flexible hose is the reflex answer and it is the one most likely to disappoint. A hose does not remove the exciting force. It decouples the piping from the pump, which moves the reaction somewhere else, and braided hose is stiff axially, so it transmits pressure-pulse thrust straight through unless both ends are anchored. On the suction side an unrestrained hose can also breathe with each stroke, adding compliance that changes the acoustic tuning in ways you did not design for.

Use hose when the foundation is the vibration source and you need to isolate the pump from it, or as a deliberate isolator with hard anchors on both sides. Do not use it as a substitute for bolting the pump down or for pulsation control. If you do test with hose, treat it as a diagnostic: if the skid quietens down, you have confirmed force transmission through the rigid tube, and the permanent fix is still anchoring plus dampening. Add elastomer compatibility with kerosene and hose fatigue life to the maintenance burden before you make it permanent.

Fix Procedure and Verification

1. Record a baseline. Overall vibration velocity at the pump feet, base plate, suction line, discharge line, and the structure, plus a spectrum at the worst point. Photograph the whole installation including supports and mounts.
2. Answer the commissioning question. Day-one vibration goes down the design branch; recent onset goes to mechanical inspection first.
3. Remove the vibration mounts. Bolt the pump feet solid to a flat, shimmed, full-contact base and torque the hold-downs. Re-measure. This alone often halves the motion.
4. Anchor the base plate to a foundation with adequate mass, or stiffen it. Re-measure.
5. Clamp the piping. Break long unsupported spans, brace small-bore stubs and gauge take-offs to a rigid bracket. Re-measure.
6. Run the acceleration-head and NPSH calculation with your real suction geometry and pump speed. If the margin is short, shorten and enlarge the suction line, raise the supply, or slow the pump.
7. Send the full pump and piping data set to a pulsation dampener manufacturer. Fit suction and discharge units close-coupled to the pump ports, set the pre-charge per their instructions, and confirm elastomer compatibility with kerosene.
8. Re-measure after each change, not after all of them. One change at a time is the only way to know what worked.

Verification targets: gauges readable at running speed, no crackle or rumble audible at the pump head, suction pressure at the inlet staying above vapour pressure through the whole stroke, and the pulsation harmonics dropped out of the spectrum. Loose fasteners are how a solved vibration problem comes back.

If the skid still moves after the pump is solidly anchored, the piping is properly clamped, and correctly sized close-coupled dampeners are fitted at the specified pre-charge, stop changing hardware. At that point you need the pump OEM to confirm the unbalanced force and permitted foundation loading for your model, and a dampener manufacturer or piping specialist to run a pulsation and acoustic study on the actual system geometry. Send them the baseline data, the spectra, and the photographs you took in step 1.

Frequently Asked Questions