The audible and visual overpressure alarm that stops the machine is a safety protection. Once the alarm sounds, the order in which the site responds matters more than the response itself. Two instinctive actions are the ones most likely to cause trouble: resetting and restarting as soon as the alarm sounds, and switching everything on the machine off. Both wipe out the basis for the judgement; the first overwrites the state of the site, the second shuts off the running information that may still be useful.
Step one: preserve the state, do not reset in a hurry
After the protection has acted, the first thing to do is to record, not to act. Write down three things: the moment the alarm sounded, the pressure reading shown at that moment, and what the machine was doing before the alarm, whether it was feeding oil, heating, or had just finished an element change. These three set the starting point for every judgement that follows. Write them down first and only then decide how far to shut down; do not let the operating action run ahead of the record.
Overpressure protection is part of the protection chain of the machine, one of several lines of defence alongside phase-failure and phase-sequence protection. Its operation shows that the protection condition was genuinely reached; that is not direct evidence of a spurious operation, but it is not evidence that the machine is faulty either. The protection has left the machine in a safe state, and that has to be acknowledged.
Step two: decide which section the pressure comes from
A rising pressure has only a few possible sources, and they can be ranked by how easy they are to confirm on site. The first is increased resistance in the filtration section: if the element has taken up dirt or is blocked, the upstream pressure keeps climbing, and the remedy is to clean or replace the element after shutdown as the model requires. The second is an obstructed downstream path: if the oil outlet valve or the bypass valve is not fully open, the pressure builds up near the outlet of the filtration section. The third is the pump itself: a screw oil pump delivering less, or air in the pipework preventing the pump from drawing, which shows as an abnormal pressure reading with a flow rate that also fails to rise. The fourth is a temperature control setting that is too high, which raises the viscosity, worsens the flow behaviour and lifts the whole pressure reading.
The way to tell them apart is to look at the combination of pressure and flow rather than at the pressure figure alone. High pressure together with falling flow usually points to resistance in the filtration section or an obstruction downstream; high pressure with normal flow sends you back to the temperature control and the running state of the pump. None of these judgements requires dismantling the machine, but they do require the pressure gauge and the flow indication to be at hand at the same time.
| Pressure reading | Flow behaviour | The more likely direction | What to do first |
|---|---|---|---|
| Keeps rising | Clearly falling | Increased resistance in the filtration section or an obstruction downstream | Stop the machine and check the elements and the position of the outlet valve |
| High but steady | Close to normal | Temperature control setting or a change in viscosity | Check the actual oil temperature against the set value |
| The reading fluctuates | Fluctuating | Air trapped in the pipework or a problem at the tapping point | Check the suction conditions of the pump |
| Rises slowly | Falls slowly | The element is gradually blocking | Arrange the element replacement ahead of the replacement interval |
| Jumps up suddenly | Drops sharply or is interrupted | The pump has stopped turning or a valve has been closed in error | Confirm the running state of the pump first |
Step three: act on the judgement and leave a record
In terms of order, first restore the observable state, then restore the runnable state. In practice: if the problem is in the filtration section, shut down, clean or replace the elements and restart; if it is the position of a valve, reset the valve to its specified position and restart; if it is the temperature control setting, set the value back into the range the process requires and restart. After each action, observe for a period rather than treating the action as the solution.
The record matters just as much. Look at the number of protection operations, the time of each one and the operating condition at the time together; if the same protection operates repeatedly, what it points to is a fixed weak point rather than an occasional event. A single operation may come from the operating sequence, while repeated operations usually point to a structural or a maintenance problem.
- The moment the alarm sounded, the pressure reading and what was happening at the time: write down all three before touching anything.
- Do not reset and restart repeatedly before the pressure has been released and the cause understood.
- Look at the pressure and the flow together while investigating; the pressure on its own confuses the two directions.
- Count the protection operations per batch, and investigate repeated operations of the same kind separately.
- Only after the problem is confirmed as dealt with, write the actions of this event into the handover notes, so that the next shift can pick it up.
A protection operation is a signal from the machine. When the signal appears, record first, judge next, act last; if the order is wrong, nothing that is investigated afterwards adds up.
