A hydraulic cylinder that creeps under load is the most common complaint on any hydraulic machine, and it is one of the few faults where replacing the cylinder itself is regularly the wrong answer. Drift means oil is bypassing somewhere between the pressurised chamber and the low-pressure return. That bypass may be inside the cylinder, across the directional control valve, or through a load-holding valve that has stopped holding — and all three produce an identical symptom.
This guide sets out the three causes that account for almost all drift, the isolation test that tells them apart in minutes, and what each result means for the repair.
What cylinder drift actually is
In a healthy cylinder, oil trapped in the cap-end or rod-end chamber holds the piston stationary against the load. Drift is that oil moving from the high-pressure side to the low-pressure side, allowing the load to push the piston. The load does not move because the oil is being lost from the system in most cases — it moves because oil is being transferred past a seal or a valve spool.
Every cylinder drifts to some degree. Piston seals are elastomer components under continuous differential pressure; a microscopic amount of leakage is inherent. What matters is whether the rate falls within what the application can tolerate.
Acceptable drift rates by application
There is no universal acceptable rate, because a scissor lift and a bucket on a loader have nothing in common. The figures below are the working ranges generally accepted for industrial and mobile duties.
| Application | Typical acceptable drift | Notes |
|---|---|---|
| Presses and clamping | 0.1 – 0.5 mm/min | Position tolerance is the limit, not the cylinder |
| Machine tool positioning | Below 0.1 mm/min | Anything more shows as position error |
| General industrial load holding | 1 – 3 mm/min | Often tolerated but should be trended |
| Mobile loader arms and tippers | 3 – 5 mm/min | Duty cycle and safety margins allow more |
| Overhead and personnel lifting | Effectively zero | Requires load-holding valves, not cylinder seals alone |
Measure drift against these figures over a fixed interval rather than by eye. Mark the rod position with a marker or tape, load the cylinder, place the control valve in neutral, and time the movement over ten minutes.
How much oil a drift rate represents
Leakage flow can be calculated from the drift rate and the piston area, which is useful because it converts an apparently trivial movement into a measurable number.
Q = A × v
Where A is the affected piston area in mm² and v is drift in mm/min, giving Q in mm³/min. Use the full piston area when the cylinder drifts on extension, and the annular area when it drifts on retraction.
For a 100 mm bore cylinder with a 50 mm rod drifting at 2 mm/min on retraction, the annular area is 5,890 mm², so the bypass flow is about 11.8 cm³/min — a small-looking figure that is still enough to ruin press tolerance.
Bypass flow at common drift rates
| Drift rate (mm/min) | 100 / 50 cylinder, retract drift | 63 / 45 cylinder, retract drift |
|---|---|---|
| 0.5 | 2.9 cm³/min | 0.8 cm³/min |
| 1.0 | 5.9 cm³/min | 1.5 cm³/min |
| 2.0 | 11.8 cm³/min | 3.1 cm³/min |
| 3.0 | 17.7 cm³/min | 4.6 cm³/min |
Note that leakage is not linear with drift on a given cylinder — it scales with area. A large cylinder drifting slowly can move more oil than a small cylinder drifting quickly. When specifying a replacement seal, this is the number that matters.
Cause 1: piston seal bypass
This is the most common cause, and the one most often misdiagnosed. The piston seal separates the cap-end and rod-end chambers. When it degrades — through abrasive contamination, an over-pressure event, thermal ageing, or chemical attack on the seal compound — oil crosses the piston under the differential pressure of the suspended load. The high-pressure chamber empties slowly and the load pushes the rod in the direction it is already pushing.
Typical triggers are contaminated oil abrading the seal lip, a spike in system pressure extruding the seal into the clearance gap, or a seal compound that is unsuited to the fluid or the temperature range. A scored bore will also produce bypass, and this is the case that causes repeat failures: new seals fitted to a damaged bore fail again within weeks.
Diagnosis is confirmed by the port-block test below. The repair is a cylinder rebuild with a complete seal kit, and the bore must be inspected and honed before the new seals go in.
Cause 2: directional control valve leakage
Every directional valve has a finite internal spool leakage, specified by its manufacturer in cm³/min at rated pressure and temperature. As the valve ages, the spool-to-bore clearance increases and leakage rises above the threshold the seals were sized for. Oil on the load-holding side drains to tank, the trapped volume falls, and the cylinder moves.
The distinguishing feature of valve leakage is that the cylinder holds perfectly when the lines are disconnected and capped. The cylinder is healthy; the circuit is not.
The fix is valve repair or cartridge replacement, and checking the spool for wear and contamination. On machines where drift has been an ongoing nuisance, the more robust answer is to fit a pilot-operated check valve directly at the cylinder port. That arrangement traps oil in the cylinder regardless of what the directional valve does, and removes this failure mode altogether.
Cause 3: load-holding valve failure
Cylinders on cranes, aerial platforms, tippers and other load-holding applications should not rely on cylinder seals to hold a suspended load — a seal is a seal, not a safety device. They use one of three arrangements.
| Valve type | Function | When it fits |
|---|---|---|
| Pilot-operated check valve (POCV) | Blocks oil leaving the load chamber until pilot pressure opens it | Standard on cranes and aerial platforms; true zero leak |
| Counterbalance valve | Maintains back-pressure on the load side | Where the load would otherwise accelerate away from pump control, such as boom lowering |
| Load-holding cartridge | POCV function mounted directly in the cylinder port | Eliminates the external line, the tightest possible arrangement |
When these valves are present and the cylinder still drifts, the valve seat is usually contaminated with debris from a failed seal or piston, or the spring has fatigued below its cracking pressure. Flush the circuit and replace the cartridge — a seat contaminated with metallic debris should not be cleaned and reused.
The port-block test: which one is it
This is the single most time-saving diagnostic available, because it separates the three causes without disassembly. Counterbalance valve seat leakage and piston seal bypass produce identical visible symptoms; only this test tells them apart.
- Extend the cylinder and apply the working load
- Place the control valve in neutral, with the spool centred
- Mark the rod position and start a ten-minute timer
- Disconnect both hydraulic lines at the cylinder ports and cap both ports
- Support the load mechanically — the cylinder must not be the only thing holding it
- Measure the rod movement over the next ten minutes
Reading the result
| Result | Conclusion | Action |
|---|---|---|
| Still drifting with ports capped | Fault is inside the cylinder | Rebuild with a full seal kit; inspect and hone the bore |
| Holds when ports are capped | Fault is upstream in the circuit | Repair or replace the directional valve, check load-holding valve cartridges |
| Drift only after the system warms up | Thermal effect on leakage | Check oil temperature and viscosity grade; consider cooling |
| Drift started after a recent rebuild | Wrong seal profile or damaged bore | Verify seal profile and groove dimensions before refitting |
Interpret the ten-minute measurement
| Drift over 10 minutes | Implied rate | Assessment |
|---|---|---|
| 0 – 1 mm | Below 0.1 mm/min | Within press and clamping tolerance |
| 1 – 5 mm | 0.1 – 0.5 mm/min | Normal; record as a baseline for trending |
| 10 – 30 mm | 1 – 3 mm/min | Acceptable on general industrial duty; plan seal replacement |
| 60 mm | 6 mm/min | Marginal; internal bypass developing, schedule service |
| 250 mm and above | 25 mm/min | Severe bypass; remove from service before the bore scores further |
The last row matters more than it looks. A cylinder left running with a bypassing piston seal keeps abrading its own bore, and the repair that would have been a seal kit becomes a new barrel. Severe drift is a stop-work condition, not a monitoring condition.
Drift in safety-critical applications
In overhead lifting, aerial work platforms and any personnel-carrying application, treat a drifting cylinder as a falling load. Fit rated mechanical locks, blocks or cribbing before working on it — a cylinder that has already drifted has proven it cannot hold.
Lock out every energy source, not only electrical: hydraulic pressure, gravity, and stored spring or chain tension. Accumulators hold energy long after the pump stops, so bleed stored pressure and verify at a gauge. Hot oil burns and injects under pressure; never search for a leak by hand, and pass a card along the line instead.
Preventing drift rather than fixing it
- Control contamination. Keep oil at or better than ISO 4406 20/18/15 in service, 18/16/13 for new filtered oil. Sample on a schedule rather than filtering blindly.
- Use the correct seal profile and material for the pressure and temperature. Backup rings on high-pressure applications prevent extrusion.
- Protect the rod from side load and corrosion, and replace wiper seals as part of every rebuild — a failed wiper is the fastest route to rod and gland damage.
- Maintain surface finish. Honed bore and chrome rod should be Ra 0.2 – 0.4 µm for most elastomer and PTFE seals.
- Torque the tie rods correctly on tie rod cylinders, in a diagonal sequence in several stages.
- Log every drift event with a specific cause code rather than recording it as a generic hydraulic fault. A cylinder that drifts every quarter is an engineering problem, not a maintenance one.
Send us the measurements
If you have measured drift, send the bore, rod diameter, working pressure, the drift rate, and the result of the port-block test. We will tell you whether the fault is in the cylinder or the circuit, and supply the piston seals, rod seals, wipers or complete seal kit appropriate to the bore and pressure, together with any dimensional confirmation needed before a replacement barrel is built.
Frequently asked questions
Is some cylinder drift normal?
A small amount is inherent, because piston seals are elastomer components under continuous differential pressure. One to two millimetres per minute is often treated as acceptable on general industrial machinery. Anything faster, or any drift in a safety-critical load-holding application, needs investigation.
How do I know whether the leak is in the cylinder or the valve?
Perform the port-block test. Cap both cylinder ports. If it still drifts, the leak is internal to the cylinder. If it holds, the directional valve or a load-holding valve is leaking in neutral. This is the only method that distinguishes the two without disassembly.
Why do my new seals keep failing?
Usually the bore is scored and was not honed before fitting. New seals in a damaged bore fail fast because the same abrasive channel that caused the first failure is still there. Inspect the bore and rod during every rebuild, and fit the complete seal kit rather than replacing one component.
Can a counterbalance valve cause cylinder drift?
Yes, through a leaking seat. A counterbalance valve that is misadjusted, fatigued, or contaminated by debris can open prematurely under static load and allow the load to creep. The symptom is identical to piston seal bypass, which is why the port-block test comes before ordering any parts.
How much oil is lost when a cylinder drifts?
Multiply the drift rate by the affected piston area. For retraction drift use the annular area, for extension drift the full piston area. A 100 mm bore, 50 mm rod cylinder drifting at 2 mm/min is passing roughly 11.8 cm³/min across the piston — small in volume, but decisive for press and clamping applications.
Does oil temperature affect drift?
Yes. Higher oil temperature lowers viscosity, which increases leakage across both valve spools and piston seals. A system that holds perfectly at cold start may drift noticeably after two or three hours of running. If drift appears only after the oil is hot, temperature is part of the problem and a higher viscosity grade or cooling may be part of the answer.