Pressure Pulsation: The Simple Scientific Explanation

A reciprocating pump ran at a nominal 850 psi, and the pipe cracked anyway — twice, the second failure arriving within 18 months. The gauge never showed an overload. What it could not show was the pressure wobbling 48 psi either side of the set point, thirty times a second, with check-valve spikes pushing past 120 psi. The problem was never the pressure. The problem was the pressure that would not sit still.

That wobble has a name: pressure pulsation. It is a periodic fluctuation in fluid pressure around a mean value, and it is one of the most quietly expensive phenomena in engineering. It shakes pipelines, cracks welds, wears out pumps, corrupts flow measurements, and makes healthy machines fail on a schedule nobody can explain. It also runs through your own arteries every second.

Most writing on the subject is either dense acoustic theory or a vendor brochure. This is the plain version: what pulsation is, where it comes from, why it breaks things, and which fixes actually work.

The pressure the gauge doesn't show

Every pressure reading you have ever trusted is an average. A static gauge or a slow transmitter samples the line and reports one number — say 850 psi — while the true pressure is dancing above and below it. Pulsation is that dance.

Two numbers describe it. Amplitude is the size of the swing, usually quoted peak-to-peak: the distance from the lowest trough to the highest crest. Frequency is how often the swing repeats, in hertz. A triplex piston pump spinning at 600 rpm produces its main ripple at 30 Hz, thirty pressure cycles a second, with further energy at 60 Hz and 90 Hz.

There is a second way to size a ripple, and it matters for energy. RMS, or root mean square, squares the instantaneous values and averages them before taking the square root — the result is the ripple's energy content rather than its crest. An engineer asks for peak-to-peak when worrying about a spike, and RMS when worrying about the shaking force.

Slow gauges miss all of it. That is why a system can look healthy on paper and be destroying itself in practice.

Where the ripple comes from

Start with the pump. A positive displacement machine — piston, plunger, gear or vane — does not push fluid continuously. It traps a fixed volume and squeezes it, then releases it, over and over. The flow leaving a plunger pump is therefore lumpy: a series of slugs rather than a stream, and each slug sends a compression wave through the fluid.

Reciprocating compressors do the same thing to gas, and the shape of the resulting ripple depends on cylinder count and crank phasing. A detailed walkthrough in Processing Magazine traces how more cylinders raise the dominant frequency but lower the amplitude — in theory, at least, before real-world interactions complicate the pattern.

Turbomachinery is tidier but not innocent. A centrifugal pump or fan produces a pulsation at its blade-passing frequency — the rate at which each impeller vane sweeps past the casing cutwater — and at multiples of it. The ripple is smaller than a piston pump's, but it excites piping the same way.

Then there is the violent cousin: water hammer. When a valve closes or a pump trips, a moving column of liquid is forced to stop. Its kinetic energy converts into pressure, and a wave races back through the pipe at the speed of sound in the liquid. The Joukowsky equation — Δp = ρ × a × Δv — sizes the spike: density times wave speed times the change in velocity. Stopping a 1 m/s flow in a steel pipe can add roughly 1.3 MPa, close to 195 psi, almost instantly. Close the valve slowly and the spike shrinks: a 100 ft line carrying water at 6 ft/s, shut in 0.1 seconds, can spike about 420 psi, while a one-second close holds it near 42 psi.

Pulsation is the regular, repeating ripple. Water hammer is the irregular, one-off slam. They share wave physics, but a fix for one does not fix the other.

Photo by Lucas Santos on Unsplash

Here's the catch

Amplitude alone rarely breaks a pipe. Frequency decides whether the amplitude gets multiplied.

Every length of pipe is an acoustic system with natural frequencies of its own, a little like an organ pipe. When a pulsation frequency lands on one of those natural frequencies, the pressure wave reflects back and forth and stacks on itself, and a ripple of a few percent can grow many times over. This is acoustic resonance, and it is the most common reason a modest pulsation becomes a crack.

Even the contents of the fluid matter. Entrained or dissolved air changes how fast pressure waves travel through a liquid. As little as 0.1% air by volume can cut the acoustic velocity of a water line roughly in half, dragging its resonant frequencies somewhere the designer never expected them.

The 850 psi pump is the textbook case. Its discharge line had a natural frequency at 92 Hz. The pump's third harmonic sat at 91 Hz. That one-hertz near-miss amplified the ripple locally until cyclic stress in a threaded joint reached about 65% of the metal's yield strength — enough to predict fatigue failure in three to four years, which matched the observed 18-month cracks almost exactly.

This is why API 618, the American Petroleum Institute standard for reciprocating compressors, treats pulsation as a frequency problem first. It caps unfiltered pulsation at the cylinder flange at 7% of mean absolute line pressure, or three times the stage pressure ratio, whichever is lower, and applies a frequency- and diameter-dependent limit through the piping. The standard exists because resonance turns percentages into failures.

Your arteries run the same experiment

The human circulatory system is a pulsation problem with a clever damper built in. Each heartbeat ejects blood in a burst, so arterial pressure rises to a peak — systolic — and falls to a trough — diastolic. The gap between them is pulse pressure; for a healthy 120/80 mmHg reading, that gap is 40 mmHg.

A normal pulse pressure sits around 40 mmHg, with 40 to 60 mmHg generally considered the healthy band. Below 40 is low, above 60 is high, and a pulse pressure above 100 is described as truly widened. A widening pulse pressure is not a curiosity; it signals stiffer arteries that no longer cushion the heart's pulsatile output.

That cushioning is real engineering. Arterial walls are elastic, with a compliance of roughly 2 mL per mmHg, so they swell during systole and recoil during diastole. The storage-and-release smooths the flow by the time it reaches the narrow capillaries, which is why blood moves through tissues nearly continuously even though the heart delivers it in slugs. The time-averaged value — mean arterial pressure, roughly diastolic plus one-third of pulse pressure — is what organs actually feel.

The body does exactly what a pulsation dampener does. It just builds the dampener out of living tissue.

What the gauge averages away

You cannot manage what you cannot see, and most instrumentation is built not to see pulsation. A standard pressure gauge is deliberately slow, averaging the ripple away so the needle stays readable. Convenient for the operator, dangerous for the pipe.

Serious measurement uses a fast dynamic pressure transducer mounted flush with the pipe wall, so a connecting tube cannot add its own resonance. The raw signal is then broken into frequencies with a fast Fourier transform, or FFT, which shows exactly which harmonic is carrying the energy. A peak-to-peak number tells you the worst crest; the spectrum tells you why it is there.

There is one practical trap: location. In a standing wave, some points along a pipe barely move while others move violently. A transducer sitting at a pressure node can read near zero while the pipe a few metres away is shaking itself apart.

Photo by hidde schalm on Unsplash

Taming the ripple

For a piston or plunger pump, the standard fix is a pulsation dampener: a small pressure vessel, usually split by a bladder or diaphragm, with a compressible gas such as nitrogen on one side. Liquid above the mean flow rushes in and compresses the gas; on the trough, the gas expands and pushes it back out. The gas side is precharged to about 80% of minimum system pressure so there is always liquid in the vessel, and a well-sized unit can remove more than 95% of the flow and pressure variation.

Volume matters. A rough starting point for an accumulator is 5 to 15% of the pump's displacement per stroke. Too small and it barely breathes; too large and you have bought an expensive pipe fitting.

Other tools attack the system rather than the source. Orifice plates burn pulsation energy as pressure drop. Extra supports and shorter spans push a pipe's natural frequency up and away from the pump's harmonics. Slower valve closure tames water hammer, and surge suppressors — gas vessels precharged to about half of operating pressure — catch the one-off spikes a continuous-ripple dampener was never built for.

There is steady innovation, too. A 2022 study of a compact component built around an elastic tube, with no gas bladder at all, cut downstream pulsation by about 43% compared with the upstream side at a mean pressure of 3.5 MPa.

Hype versus reality

Every fix on the market has a brochure, and every brochure has a claim. A few deserve a closer look.

• Claim: a dampener eliminates pulsation. Reality: no. A good one removes more than 95% of the variation, but residual ripple always remains, and it should be checked, not assumed away.
• Claim: a small ripple is safe. Reality: the API 618 limits are frequency-dependent precisely because a small ripple at a resonant frequency is the dangerous one. Amplitude without a spectrum is half the story.
• Claim: slow the valve and water hammer disappears. Reality: true for transients, but it does nothing for the continuous ripple from a running piston pump. The two problems need different tools.
• Claim: stiffen the pipe and resonance goes away. Reality: it can help, but a piston pump excites many harmonics, and no real installation can push every natural frequency above every harmonic. Damping usually beats detuning.

The pattern is consistent. The physics of pressure pulsation is simple; the discipline is not. Watch the ripple and know its frequency, and a system can run for decades. Ignore the gauge's quiet average, and the metal files its own report.

Researchers studying pressure fluctuations in marine high-pressure common-rail fuel systems — lines that run up to 1,800 bar — put the payoff plainly: keep fluctuations inside the design range and “the fatigue life of the system is effectively infinite.” Let resonance push them past it, and the alternating loads become the root cause of fatigue failure.