Vibration Fault Table — Air Separation Units

Analyse a reading attach a DCS trend, FFT or waveform and get candidate faults

Attach or photograph a screenshot
DCS trend, FFT spectrum or time waveform
PNG or JPG

Say what each line colour is, which points and probes, units, the running speed (1×), alarm and danger values, the time window, and what changed and when. The more of this you give, the better the answer.

Start here

A vibration plot is a row of bars. Each bar sits at a frequency (how often per second something happens) and its height is how hard the machine shakes there, in mils. The most important frequency is 1×: once for every turn of the shaft. Most faults show up as which bars are tall and how they change over time.

Find the machine's running speed. That is 1×. Everything else is measured as a multiple of it: 2× is twice per turn, ½× is every other turn.
Ask where the tall bars are: at 1× and 2×, below 1×, way up high, or just a noisy hump with no bar at all. That points you to one or two colour families below.
Before blaming the machine, check the grey family. A surprising number of alarms are the sensor, the cable or the setup.
Open the cell, read How to be sure, and only then act. Anything marked P1 goes to a specialist before the next start.

Most common faults on ASU rotating equipment — ranked

In order of how often each one actually turns out to be the answer on air separation plants. Start at the top. The dangerous ones sit low on the list because they are rare, not because they do not matter. Tap a row to open the fault.

Top findings by machine

Main air compressor	Dirt building up (fouling), surge events, gear wear, duct and pipe shaking, pad problems after an outage, thrust position
Booster air compressor	Dirt and dust erosion, gear wear, gas-force instability at high pressure, surge, thrust position
Feed N2 compressor	Surge and stall from cold box upsets, gear mesh wear, misalignment on the train, pinion instability at low load
Recycle N2 compressor	Surge on liquefier trip, load-dependent instability, dry gas seal distress, pinion thrust collar wear and thrust drift
Combined FNC/RNC	Cross-service interaction after a recycle unload, gear mesh wear on the shared bull gear, thrust reversals on pinions
Turboexpander	Deposit and ice unbalance, thermal transients on start, thrust position drift, sub-synchronous instability, seal gas problems
Main drive motors	Bearing lubrication and wear, 2×line frequency (air gap, stator), misalignment, soft foot, VSD bearing currents
Cooling water pumps	Cavitation and low-flow running, bearing wear, misalignment, pipe strain, structural looseness
Cooling tower fans	Blade pitch unbalance, gearbox wear, drive shaft joints, deck resonance, belt faults on smaller units
Cryogenic pumps	Vapour lock and cavitation, bearing wear, cooldown thermal effects, seal problems
Any machine	Measurement artefacts — probably the most frequent single reason for an unnecessary callout

This ranking reflects typical field experience across ASU fleets. Replace it with the client's own CMMS and condition-monitoring history once twelve months of data are available — the order for a specific plant can differ, and that difference is itself informative.

What these faults look like on a DCS trend

The DCS does not show you a spectrum. What reaches the operator screen from an API 670 machinery protection rack is a handful of scalar values — overall shaft vibration in mils peak-to-peak from each X and Y probe, axial position, speed, and alongside them the bearing metal temperatures and process tags. Diagnosis needs the condition monitoring system. But the shape of a one-second trend tells you a great deal about what kind of fault you are dealing with, and how fast you need to move.

Each sample below is a one-second trend of the kind an operator or manager can pull up without any specialist software. Learn the shapes — and learn the normal ones first.

Every coast-down: what to look for

Rundown time. Note the seconds from breaker open or trip to slow roll and compare with the last five coast-downs. A shorter rundown means something is dragging — a rub, a seal, a bearing, a brake. A longer one usually means a coupling or load has been removed.
Shape of the fall. Unbalance vibration falls with the square of speed, so the curve should be smooth and steep at the top. A flat or rising region on the way down is abnormal.
The critical speed bump. It should appear at the same speed and roughly the same height as on the run-up and on previous coast-downs. A shift in speed means the supports or bearings have changed; a taller peak means damping has been lost or unbalance has increased.
Slow-roll reading. Below about ten percent speed the probe reads mechanical and electrical runout, not vibration. It should return to the same small value every time. A change means a bent shaft, a damaged journal surface, or a probe problem.
Bearing temperatures and oil pressure. Auxiliary or emergency lube oil should carry the machine to a stop. Temperatures should fall, not rise, during rundown.
Axial position at rest. It should come back to the same resting value. A different value means the thrust bearing or coupling has moved.

Words you will meet, in plain English

Word	What it means
1×	Once for every turn of the shaft. 2× is twice per turn, ½× is every second turn. Nearly everything is measured against this.
mils	The unit of shake for shaft probes: thousandths of an inch, peak to peak. 1 mil is about the thickness of a sheet of paper.
in/s	Inches per second: the unit for sensors bolted to the casing (accelerometers). Used on motors, pumps and fans.
spectrum / FFT	The row-of-bars plot: how much shake at each frequency.
waveform	The raw wiggly line over time. Good for spotting knocks and rubs that the bars hide.
below 1×	Also called sub-synchronous. Bars slower than the shaft turns. On compressors these are the ones to take seriously.
side bars	Small bars evenly spaced either side of a big one. Their spacing names which shaft is causing it.
timing (phase)	When in the turn the shake happens, measured against a mark on the shaft. Comparing timing at two points tells apart faults that look the same.
blade pass	Number of blades times 1×. The pulse each blade makes going past a fixed part.
gear mesh	Number of teeth times that shaft's speed.
BPFO, BPFI, BSF, FTF	The four frequencies a ball or roller bearing makes when its outer ring, inner ring, balls or cage are damaged. Worked out from the bearing part number.
twice line frequency	100 Hz on a 50 Hz supply, 120 Hz on 60 Hz. The magnetic signature of a motor.
tilt pad bearing	The bearing type on every compressor here: several separate pads, each rocking on a pivot, floating the shaft on oil. Very stable, but it depends on the pads being fitted exactly right.
axial position	Where the rotor sits along its own length. Watched by the thrust probes. Drift here matters even when vibration is low.
alert / danger	Alert means investigate. Danger means the protection system will trip the machine.
rundown time	Seconds from breaker open to standstill. A shortened rundown means something is dragging.

What to measure where

Machine	Primary measurement
Main air compressor (integrally geared)	Shaft proximity probes on each pinion & bull gear, plus casing acceleration for mesh
Booster air compressor	As above; add axial position on thrust ends
Feed N2 compressor (FNC), integrally geared	X-Y proximity probes on each pinion, axial position on each pinion, bull gear thrust bearing temperature, casing acceleration for mesh
Recycle N2 compressor (RNC), integrally geared	As FNC, plus dry gas seal vent flow and seal gas dP on the high-pressure pinions where fitted
Combined FNC/RNC unit	As above on every pinion of both services — plus process trends from both services on the same screen
Turboexpander / booster	Proximity probes, axial position, bearing temperatures — casing readings are of limited value
Motors, cooling water pumps, blowers	Casing velocity (in/s) plus acceleration enveloping for bearings
Cooling tower fan drives	Low-frequency velocity at the gearbox and on the fan deck structure
Cryogenic (LOX/LIN) pumps	Casing acceleration at the warm end; trend with flow, suction pressure and sub-cooling

Rolling-element faults live in acceleration and envelope data. Balance, alignment and looseness live in velocity. Fluid-film and rotordynamic behaviour lives in shaft displacement.

Screening severity — casing velocity (in/s)

Zone	Broad meaning
A	Newly commissioned condition.
B	Acceptable for unrestricted long-term running.
C	Unsatisfactory for long-term running. Plan an intervention.
D	Severe enough to cause damage. Investigate now.

Zone boundaries depend on machine class, mounting and support stiffness — take them from ISO 20816 and the OEM manual for the specific machine, not from a single number. A rising trend against a machine's own established baseline is a stronger signal than any absolute limit.

Before you call it a fault

Was the reading taken at the same point, direction and mounting as the baseline?
Was the machine at steady load, speed and temperature?
Does a repeat reading reproduce it?
Has anything changed in the process — flow, suction temperature, adsorber switching, cooling water?
Is it on the machine, or on the pipework and structure attached to it?

The grey family at the end of the table covers measurement artefacts that look alarming and are not faults. Rule those out first.

Nitrogen compressors: what is different

Configuration	What to expect
Independent FNC	Takes warm gaseous nitrogen from the cold box and raises it to liquefier or pipeline pressure. Clean, dry gas, so fouling is rare; suction conditions swing with the cold box, and it sees every column upset first.
Independent RNC	Circulates nitrogen around the liquefier refrigeration loop at high pressure ratio. Usually the highest-pressure machine on site, with the hardest-worked pinions. Gas-force instability, thrust collar and dry gas seal problems live here. A liquefier trip removes its load in seconds.
Combined FNC/RNC	Feed and recycle stages on one bull gear, one casing set, one lube system. Anything that happens on one service reaches the other through the gear and the thrust balance. Diagnose it as one machine, never as two.

On all three the drive is large and often variable speed. The ASU-family cells for icing, adsorber carryover and liquid ingestion do not apply — nitrogen service is dry and clean. Surge, instability, gear mesh, dry gas seals and thrust are the ones that matter.

Tilt pad bearings: what to expect

Every compressor pinion, bull gear and expander here runs on tilt pad bearings. That changes what to look for:

Classic oil whirl and whip, the half-speed instability of old sleeve bearings, is designed out. If a compressor shows a bar below 1×, the likely causes are gas forces at high pressure (Ac), pads fluttering (Pf), or pads damaged or fitted wrongly (Lp, Pi), not the oil.
The bearing's stiffness comes from the pads. Worn pivots or wrong pads make the machine softer: 1× creeps up, the rotor sits lower, and the critical speed on a coast-down moves. Compare every coast-down with the last one.
The first start after any bearing job is the most likely time to find a problem. Compare it with the last run before the outage, not with the alarm limits.
Thrust is taken on tilt pad thrust bearings too. Watch axial position and thrust pad temperature; vibration will not warn you.
Pad metal temperature is often the earliest and most reliable warning on these bearings. Trend it alongside vibration, never instead of it.

Setting the trends up so they are worth having

A one-second scan rate is only real if the historian keeps it. Swinging-door or deadband compression will quietly flatten a surge event into a single sample — check the compression deviation on every vibration and axial position tag, and keep the raw data for these points.
Build the trend groups before you need them. For each machine: both X and Y probes at each bearing, axial position, speed, bearing metal temperatures, lube oil supply temperature and pressure, and the process tags that load the machine.
Vibration and process on the same chart. Almost every question an inexperienced manager asks about a trend is answered by whether load, flow or temperature moved at the same moment.
Alert and danger are different things. Alert is a call to investigate; danger is a protective function with a short voted time delay. Neither is a number to argue with while the machine is running.
Never defeat, bypass or raise a danger setpoint to keep a machine running. If a bypass is genuinely needed, it goes through management of change with a time limit on it.
Record every surge event, every trip and every alarm excursion with its timestamp. The trend is only evidence if someone can find it three months later.

Shaft vibration limits for these machines come from the OEM and the site protection philosophy. Where a general benchmark is needed, the usual starting point for fluid-film machines is the ISO 20816 shaft vibration guidance, which scales with running speed — high-speed pinions and expanders have far tighter limits in mils than a 1800 rpm motor.