Hydraulic system tracing

Hydraulic System Fault Tracing Procedure

Specially for Loesche Roller Mill — A Field Guide

Anyone who's chased a hydraulic fault on a Loesche mill knows it's rarely as simple as “check the pump.” The hydraulic system on these mills is doing a lot — holding grinding pressure on the rollers, allowing them to swing away during high vibration or tramp material events, and doing it all under high pressure with a fair amount of heat and contamination working against it. Here's how I actually walk through it when something's not right.

Step 1: Understand What the System Is Telling You First

Before touching a single valve, pull up the trend for hydraulic pressure, tank level, oil temperature, and roller position over the last few hours. Nine times out of ten, the trend already tells you whether you're chasing a slow-developing problem (seal wear, filter clogging, cooler fouling) or a sudden one (a burst hose, a stuck valve, a sensor failure). Don't skip this step just because the alarm feels urgent — five minutes here saves an hour of guessing later.

Step 2: Confirm It's Actually Hydraulic, Not Instrumentation

A surprising number of “hydraulic faults” are really a transmitter or pressure switch giving bad readings. Cross-check the DCS/PLC pressure value against a local gauge at the accumulator or cylinder manifold. If the local gauge reads normal and the DCS doesn't, you've just saved yourself from tearing into the hydraulic power unit for no reason — go chase the transmitter, wiring, or loop instead.

Step 3: Check the Basics on the Power Unit

• Oil level in the tank — low level is one of the most common root causes, and it's the fastest to check.

• Oil temperature — too hot and viscosity drops, causing internal leakage and pressure loss; too cold on a startup and you'll get sluggish response.

• Return line filters — a clogged filter can starve the pump or trip a bypass indicator, and it's often overlooked because it “was fine last week.”

• Pump running current — compare against normal baseline; a pump working harder than usual for the same pressure output often means internal wear.

Step 4: Trace the Pressure Path Rather Than Guessing

Work from the pump outlet toward the rollers, checking pressure at each stage:

1. Pump discharge pressure — confirms the pump itself is making pressure.

2. Pressure at the accumulator bank — confirms the accumulator is charged and holding (a low nitrogen pre-charge is a very common cause of pressure fluctuation under load).

3. Pressure at each roller cylinder — tells you whether the loss is upstream (pump/accumulator/valve) or isolated to one cylinder (seal or hose leak on that specific roller).

If pressure is fine at the accumulator but drops at one particular cylinder, you've narrowed it to that roller's hose, seal, or directional valve — no need to open up the whole system.

Step 5: Check the Directional and Relief Valves

Sticking directional valves are a classic Loesche hydraulic issue, especially when contamination has worked its way into the system. A valve that won't fully shift can cause a roller to hang up in the wrong position, or cause erratic swing-out behavior during a trip. Relief valves that have drifted out of setpoint will either bleed off pressure early (weak grinding pressure) or fail to protect the system (pressure spikes). Both are worth checking against OEM setpoints if the fault is intermittent.

Step 6: Inspect for External Leaks and Hose Condition

Walk the hose runs to each roller cylinder, particularly at the swivel joints and crimped fittings — these flex constantly and are the most common source of slow leaks that eventually show up as “can't hold pressure” complaints. A hose that's been in service several years and shows any surface cracking or weeping at the fitting should be flagged for replacement before it becomes an unplanned trip.

Step 7: Don't Forget Contamination

Hydraulic oil contamination — whether from a failed seal letting in dust, water ingress through a breather, or just accumulated wear particles — is behind a lot of “mystery” faults: sticky valves, erratic pressure, premature pump wear. If you're seeing repeat faults with no obvious mechanical cause, it's worth pulling an oil sample and checking particle count and moisture content rather than just replacing parts one at a time.

Step 8: Common Instrumentation Faults on the Hydraulic System — and How to Fix Them

Since we already talked about ruling out instrumentation in Step 2, it's worth going deeper here, because a good chunk of “hydraulic” faults I've chased over the years turned out to be instrumentation problems wearing a hydraulic costume.

Fault	Likely Cause & Solution
Pressure transmitter reading zero or dropping out intermittently	Usually traces back to a loose or corroded connector at the transmitter head, especially in an area exposed to oil mist and dust. Open the terminal box, check for oil ingress into the electrical connection, clean and re-terminate, and make sure the cable gland is properly sealed. If it's intermittent under vibration, check the cable for chafing along its run — a nicked conductor shorting intermittently is a classic culprit.
Pressure transmitter reading a fixed, “stuck” value	If the reading never moves even when you know pressure is changing (confirmed on the local gauge), suspect a blocked impulse line or a failed sensing diaphragm rather than the electronics. On hydraulic applications, a small amount of varnish or sludge buildup in the sensing port is common if the oil hasn't been changed on schedule. Isolate, bleed, and flush the impulse line before condemning the transmitter.
Erratic or noisy pressure signal	This is often electrical, not mechanical — check for cable routing too close to VFD or motor power cables, and confirm shield grounding is only at one end (usually the PLC/marshalling end) to avoid ground loops. If the noise correlates with pump motor starts, it's almost certainly EMI coupling rather than an actual pressure fluctuation.
Roller position transmitter (LVDT or similar) giving inconsistent feedback	These take a beating from vibration and thermal cycling. Check mechanical mounting first — a loosened bracket lets the sensor shift relative to the rod it's tracking, throwing off the calibration even though the sensor itself is fine. Re-verify zero and span against known mechanical positions before assuming the transmitter has failed.
Level switch or level transmitter in the hydraulic tank giving false low-level alarms	Foaming oil (often from air ingress on the pump suction side, or from a return line dumping oil above the fluid surface) can fool a float-type switch into reading intermittently low. If alarms come and go with pump cycling, check the return line submergence and look for suction-side air leaks before replacing the switch.
Temperature transmitter drifting or reading high compared to a handheld check	RTDs in hydraulic tanks are prone to giving false-high readings if the thermowell isn't properly seated or if there's an air gap reducing thermal contact. Pull the RTD, check the thermowell for oil fill (some are meant to be filled with a conductive compound), and verify against a calibrated handheld probe before adjusting anything in the control system.
Solenoid valve not responding to PLC command	Confirm 24V (or applicable voltage) is actually reaching the coil first, with a meter at the valve terminals — don't just trust the PLC output indication. If voltage is present but the valve doesn't shift, it's mechanical (stuck spool, contamination) rather than instrumentation, which sends you back to Step 5.

Bottom Line

The pattern that works is: trend first, confirm instrumentation second, then trace pressure stage by stage from the pump to the roller rather than jumping straight to the component you suspect. Most “hydraulic mysteries” on a Loesche mill turn out to be one of a handful of usual suspects — low oil level, a weak accumulator charge, a sticking valve, a slow leak at a hose fitting, or a transmitter/sensor issue masquerading as a process problem. Working through them in order — and not ruling instrumentation in or out until you've actually checked it — gets you to the real cause a lot faster than chasing symptoms.