Three situations are common on site: the log book holds nothing but dates and signatures, and nobody can say when the element was last changed or which oils the machine has processed; filter elements pile up in the store and go mouldy while the pump oil is out of stock; the maintenance plan is set on a fixed cycle that has nothing to do with the actual condition of the machine. On the surface this is record keeping that has not been done properly, but in substance the content of the log cannot support any decision, so recording it is no better than not recording it.
Whether a log is useful can be checked on the spot with four questions: when was the element last changed; how much oil has this machine processed in total; which oils has this machine processed; what was the cause of the last fault and how was it resolved. If all four can be answered on the spot, the log is alive; if they cannot, it is only a registration book.
A log usually fails for three reasons
- It records the action but not the result: it notes that the filter element was changed, but not which model was fitted, how much oil had been processed when it was changed, or what condition the removed element was in.
- It records the symptom but not the conclusion: it notes that the pump was noisy, but not which component the fault was finally traced to, how it was dealt with, or whether it recurred.
- The records do not stay with the machine: a paper book is indexed by work position, and once the machine has been moved the records are broken and the accumulated experience can no longer be found.
The minimum usable set of fields: seven
These seven fields are not a case of the more the better, but of exactly enough to answer the decision questions:
- Equipment number: identical to the nameplate, with the model and the main specifications recorded as well. It answers which machine the record refers to, and because the number travels with the machine when it is moved, the records stay connected.
- Installation position: down to the room and the work position. It answers where the spare parts have to be delivered and which line has to be stopped for maintenance.
- Oil type and grade: insulating oil, hydraulic oil and gear oil must be recorded separately, with the grade copied from the label on the drum. It answers what oil the machine processed last time; mixing the records makes a contamination problem impossible to trace.
- Volume processed each time: with the unit fixed to litres or tonnes, one unit used throughout. It answers how much work the machine actually did and how quickly the elements are consumed.
- Total running hours: added up after each run. It answers when the pump oil is due for a change and when the pump is due for dismantling and inspection.
- Element replacement record: the date, the model and specification and the reason for replacement, with the reason split into end of life, differential pressure over limit, damage and accuracy not met. It answers whether this replacement was normal consumption or a premature failure.
- Fault and conclusion: the symptom, the operating conditions at the time, the action taken, the final conclusion and whether it recurred. It answers whether the same problem keeps coming back.
How the log drives spare parts purchasing
Use the volume processed between two element changes to work out the actual capacity of a single element, multiply it by the number of machines in service, and that gives the annual demand for elements, a figure more reliable than ordering by guesswork. Compare the total running hours with the pump oil change records and the consumption rate of the pump oil can be seen to be abnormal or not. Put the recurring problems from the fault records side by side: those are the priority for spare parts and for modifications. Count the share of premature failures in the replacement reasons on its own; if that share rises, the problem lies in an earlier stage of the process, and replacing more elements will not help.
How the log drives the maintenance plan
Scheduling a pump overhaul when the total running hours are reached fits the actual situation better than a calendar plan. Where one position reports the same fault repeatedly, put the modification or the specific re-check into the next maintenance window. Shorten the inspection interval for machines that process a lot of oil and lower the frequency for machines that have been out of use for a long time; idle machines should not keep taking up spare parts and inspection hours. The conclusion recorded for each fault can be rewritten as a reusable inspection step and added to the inspection sheet, so that the next time the same symptom appears the check can simply be followed.
| Symptom | Possible cause | How to confirm | Action |
|---|---|---|---|
| The duty staff cannot say when the element was last changed | The log holds only dates and signatures | Pick three machines at random and check their last three records | Rebuild the record format around the seven minimum fields |
| Element stock is piled up while the pump oil is out of stock | Spares purchasing is not linked to the actual consumption | Back-calculate the demand from the volume processed between two element changes | Order against the demand and count the share of premature failures |
| The same fault keeps recurring on the same machine | Faults are recorded by symptom only, without the conclusion | Summarise the fault records by equipment number | Add the missing conclusions and list the machine for a specific re-check in the next maintenance window |
| The history of a machine cannot be followed after it has been moved | The log is indexed by position instead of by equipment number | Compare the nameplate number with the number in the records | Use the nameplate number as the primary key and the position as a variable field |
| The maintenance plan is unrelated to the actual condition of the machine | The plan is set on the calendar without using the total running hours | Compare the planned maintenance items with the fault records | Re-sort using the total running hours and the fault frequency |
The value of the log lies not in how much has been written down, but in whether three questions can be answered on the spot: how many spare parts to order, which machine to repair first, and what parts to take to site next time.