Before a single pass of steam touches an industrial pump, someone has to hand the cleaning crew a state they can work in. That is staff preparation, and it is the part of a cleaning trial that most often gets improvised. A pump exterior is a small, awkward object: a casing with a paint system, cooling fins that hold a baked film, a terminal box, a shaft end, a baseplate sitting in its own run-off. A trial on a pump measures three things at once, whether anyone planned for them or not: whether the pump survives the method, whether the method removes the film, and how long the job actually takes.
The order below is the preparation staff can do without a decision from engineering, plus the short list of items they cannot decide for themselves. Run it on one machine before you generalise to a fleet.
Write down what the trial has to prove before staff touch the pump
A trial without a written question cannot fail, which is why it usually teaches nothing. Fix one machine and one limited set of surfaces, and write what counts as a result before anyone prepares the surfaces: a coating that looks the same after the pass as it did before; a film that comes off in one pass rather than three; a cycle you can repeat on the next pump in the row.
Write the acceptance as a pair of photographs. The first set is taken while the surfaces are still as they were found, framed so the buildup reads in the picture, before any liquid goes near them. The second is taken after the pass and after the surfaces are dry. Put the two side by side and let that be the acceptance standard. A trial with no before photographs is a discussion about memory, not a result.
Staff need the stop conditions written down before the first pass, not after the first surprise: painted areas where lifting or crazed paint spreads, a label or safety marking whose edge lifts, a surface that comes up dulled or hazed, liquid reaching a row that was meant to stay dry. Any of those and the crew halts work on that row, and the row goes onto the exclusion list with an alternative method written beside it.
Give the crew a machine they can actually work around
Preparation fails most often because of the state of the pump, not the cleaning method. Three conditions have to be true before a single surface is wiped, and staff should be told which of them they own.
The pump is isolated, and the isolation belongs to the site. Power is off and locked or tagged out, driven equipment on the same shaft line is secured, and valves are set and recorded. Staff preparing the exterior confirm that the isolation exists and who signed it; they do not create it and they do not release it. The lubricant stays inside the machine. A pump that is drained and flushed for a separate mechanical job is a different, worse cleaning problem, because the internal surfaces are now open and every joint on the outside has become a route inward.
The pump is cold, and the temperature state is recorded. A warm casing changes how a deposit behaves and how a surface dries, and it changes what a cycle time means. Staff should be told which state to prepare in and should write the state down alongside the trial conditions, so that the second run is the same job as the first.
The work area can hold water and survive the run-off. That means an understood route for the liquid, a standing point, and lighting that shows what is actually on the surfaces rather than what the overhead lamp suggests. Staff walk the area before starting and say out loud where the run-off will go and where it will stop. If the answer is nowhere, the preparation cannot be signed off.
Break the pump into rows instead of treating it as one object
Staff preparing a pump exterior should write a list, one row per surface, on one sheet. A pump family that behaves consistently gives roughly this set:
The main casing and its paint system; the cooling fins and any finned motor housing; the motor exterior and its fan cowl and vents; the terminal box and its cable entries; the coupling guard and the shaft end behind it; the bearing housings; the drain plugs, sight glasses and the joints where casing halves meet; nameplates and rotation-direction arrows and safety markings; the baseplate, its feet and the shims under it; the pipework flanges and the pump-to-floor interface; and the apron of floor that belongs to the pump.
Each row gets three columns. The first says what is actually on it, in the words the crew uses: a wet oil film, a dried and baked varnish-like film where mist has settled on warm fins, grease worked into a corner with grit in it, a dry dust cake, an emulsified smear where a leak has met wash water. The second says what the surface is and how it is today: painted steel, bare machined metal, stainless, anodised aluminium, a decal, paint that is already crazed or patched or re-coated, corrosion already starting at an edge. The third says whether it can be reached: an open face, an internal corner, a fin stack, the underside of the baseplate, the gap between the guard and the casing that needs a hand removed first.
Those three columns are where the staff preparation earns its keep. A row holding a lubricant that is still wet is not the same job as a row whose residue has been cooked onto warm fins, and both rows sit on the same machine. Without that distinction, the trial reports one result for what was really several separate tasks on the same machine. A row carrying a fresh film is not a row carrying a film that has been cooked onto metal.
Mark the exclusions before the pass, and decide what replaces them
Damage during pump cleaning nearly always comes from liquid arriving somewhere it was never meant to go. The preparation that stops it is cheap and is done with tape and a marker, by staff, before the machine is started.
The boundary on a pump usually sits at the seal faces and gland area, the bearing housing and its lubrication points, the terminal box and its cable glands, the motor fan cowl and cooling openings, the coupling guard and the shaft end, instrument tails and small-bore connections, any weep hole or vent, and nameplates and safety markings with an edge already lifting. Painted surfaces whose coating is crazed, patched or re-coated belong on the same sheet, and so do bonded joints and adhesive trim.
Tape the boundary where it can be taped and mark it where it cannot. Then write what happens instead on every excluded row: a dry wipe, a cloth dampened with the solvent the site already approves, a brush, a hand clean, or a visual check only, with the trade-off the site accepts for that choice. A sheet naming only the rows to clean gets abandoned the first time the pump will not restart. A sheet that also names what to leave alone is the sheet that survives contact with a real maintenance week.
Choose the mode on the row, not on habit
Cleaning machinery around vehicle work, the page for Fixed Steam Cleaning Machine PowerDe HW02 publishes the choice between dry steam and wet steam on a fixed unit, along with its supply and connection arrangement and a protection list. It publishes no rate for cleaning pumps or production machinery, and the figures on that page describe that machine and its recorded use, not your pump and not your shop. Ask for the rows you need as rows, for the exact unit quoted.
On a pump, the two modes differ in the amount of water they put near the machine, and that is the whole difference. They are not a weaker and a stronger setting. Low-moisture work suits warm-adjacent surfaces, closeness to electrical neighbourhoods, and fin stacks where you want heat and dwell on a baked film without sending liquid down a casing joint or into a lap between two halves. Higher-moisture work suits thick crust on an open flat face, where the volume of water breaks the bond and where the liquid that comes off can be directed somewhere it can be collected. Staff should not be making that choice at the machine with the lance in hand; the mode belongs on the row.
Two loose habits to keep out of the trial record. The temperature stated for a generator is not the temperature arriving at the surface, and a tank capacity is not the time a crew can work before refilling. Both get quoted carelessly, and both change what a cycle looks like. Neither figure has to appear in your record at all, because the numbers that matter to your trial are the minutes your crew spends on the surfaces you are actually standing in front of.
Have the supplies, the route and the timing in place before the first pass
The unglamorous part of staff preparation is what most often stops a trial in its first hour. Before starting, the crew confirms and records the following:
Where the unit will stand and how it reaches that point, including kerbs, thresholds, channels, expansion joints and any doorway on the route, and whether that standing point lets one position cover the whole row set or forces a move.
What supply exists at the point of work rather than across the shop, how much cable and hose has to be pulled across a wet floor to get there, and how those lines are kept clear of the run-off route.
Where the water goes in and where it goes out: which drain, channel, tray or collection point each row discharges to, away from the baseplate feet and the pump-to-floor joint, and how the room or bay re-opens afterwards.
Consumables and refill limits inside one shift, and whether the trial can be finished without a refill in the middle of a timed run. A refill that lands inside a pass changes the measurement.
A dry finishing condition that closes each wet row: no standing liquid in a joint or at a foot, no wet label, a visual check of the internal corners and of the top shoulder where mist settles. The condition closes the row, not the clock.
Time blocks named in advance so the trial produces a usable figure: isolation confirmed, shielding and taping, the pass itself, drying and re-check, removing the protection and restoring the machine, and handing it back. Run the same row set twice from the same starting condition, take the number from the second run, and leave the blocks visible in the record so a slow one stands out and can be attacked separately.
Run the trial as a bounded job and let it answer the question
Choose a small row set on one pump: a stretch of flat painted panel, one corner where crust has built up, one finned housing with a history of running warm, and one pump-to-floor junction. Work only those rows, in the order written, with the exclusions already taped and marked. Staff photograph before and after, time the named blocks, and stop on any of the written stop conditions.
Inspect the same rows the next working day, after the pump has run a normal shift. A surface that looks clean and then re-soils by the end of the day shows that the interval was the wrong choice, not the cleaning. A second pump of the same family is the next test, and it belongs after the first result is written down, not during it.
Ask the crew doing the preparation what they found, while it is still in front of them. They are the people who know which row could not be reached, where the run-off paused, and which surface they were unsure about. That information becomes the standard if someone writes it down.
What staff cannot decide, and who has to answer them
A trial should not invent site authority. Staff preparing a pump for a steam trial need these settled in writing by others before they start, and they should be able to point at the answer when asked:
The isolation and its release, named and signed, including who is allowed to hand the pump back.
Whether this pump is in scope at all. Any pump with a seal or joint already weeping, an enclosure already open, or a coating already failing on the row you intend to clean should be settled before the trial, not discovered during it.
The run-off route and who signs it. On some sites that route carries oil and is a waste-handling question rather than a housekeeping one, and the crew cannot decide it at the machine.
The replacement method and its trade-off for every excluded row, so that an exclusion never becomes a blank space nobody owns.
Whether the trial may proceed on painted, lagged or marked surfaces the site treats as protective or contractual, and who signs that decision.
The rows to have answered in writing for the machine on the quotation
Once the surface list, the exclusions and the run-off route exist, the questions to a supplier of industrial steam cleaning equipment stop being general and become comparable answers. Ask for these as rows, for the particular unit on the quotation, and treat a brochure answer as no answer:
What supply the unit requires and what the site has to make available where the work is done; the inlet fitting for the water feed matched to the site tap; the footprint and the travel route, including kerbs, thresholds, channels and doorway clearances; whether hose and wand reach cover the reachability column on your sheet; the protective features present on that unit, listed by name rather than counted; the dry and wet switching your rows depend on; refill and consumable limits inside one shift; and how the trial is arranged, covering what the supplier brings to a taped patch test and what the site supplies.
Where a pump family turns out to need something the catalogue does not offer, that becomes a specification built from your own sheet, which is the sort of request handled as customized steam cleaning equipment rather than a catalogue purchase.
Signs the preparation is not finished
A few checks staff can run against themselves before the trial starts. If the row set has no photographs, the acceptance is a mood. If a row has no named mode, the operator will decide at the machine. If nobody on the crew can rattle off three excluded rows without looking, the exclusions have been written rather than prepared. If the run-off answer is vague, the trial will make a mess that the site then treats as the cost of the method. If the temperature state is not recorded, the two runs are not the same job. If the isolation cannot be pointed to, nothing else on this list matters.
Prepared in this order, staff preparation turns a cleaning trial on a pump into a bounded measurement: a written question, a machine in a defined state, a sheet of rows, taped boundaries, and two timed runs. What comes back is a cleaning standard your own crew can repeat, and a set of supplier answers that can be compared against each other.
