What model selection actually decides before a trial
A steam-cleaner model decision normally has to settle four things, and each one changes what the trial must test. First, the mounting: a fixed unit stays where the services are and brings the lance to the work, while a mobile unit follows the work and carries its own tank. Second, the cleaning mode: cold pressure, heated water, or steam. Third, the duty arrangement: single operator, two operators, or a station that several shifts share. Fourth, the consumable and utility supply: water source, drainage route, power point, and the drying material the finish standard depends on.
The cleaning machinery category lists hand-pushed machines, walk-behind and wheeled units, and fixed steam machines as different families, and it carries entries for the hand-pushed, walk-behind, fixed and petrol-heated arrangements. The site also offers built-to-order steam cleaning equipment, which means the boundary between catalogue models is not always where the buyer expects it. A model comparison that starts from catalogue families rather than from these four decisions tends to compare brochures instead of comparing duties.
For the fixed side of the range, this site documents one Fixed Steam Cleaning Machine PowerDe HW02 entry carries a heater rating row, a stainless tank capacity row, a single-phase supply row, a stated footprint, a one-half-inch water inlet and a set of protections, and the same entry publishes the application scope it was written for: vehicle exterior, interior and engine-compartment cleaning, plus cleaning around parking and garage areas. Treat those lines as the description of that one catalogue entry, not as a description of every fixed steam machine, and not as a claim about kitchen, production-line or machinery duties.
Read the published application scope before it shapes the shortlist
The published scope matters for model selection in a practical way. On the mobile side, the walk-behind model is listed with two battery arrangements and is positioned around indoor garage bays, tight outdoor parking areas, and work where a hose run or an extension lead would get in the way. On the fixed side, the documented entry is positioned around vehicle exterior, interior and engine-compartment work. The product pages for the hand-pushed and tanked models describe construction and features and do not publish application text at all, so for those models the buyer has no published scope to lean on and the trial carries the whole weight of the decision.
Write the scope down as two columns before the trial: what the catalogue page says the machine is positioned for, and what your job actually is. Where the two match, the trial can concentrate on finish and pace. Where they do not match, the trial has to test both feasibility and acceptability, and the result must not be reported as a catalogue claim. Where the page is silent, the trial result is the only evidence anyone has, and it belongs to your surfaces rather than to the model.
Turn your requirement into a sheet the models are scored against
A model decision is only as fair as the sheet it is scored against. Build that sheet from the job, and write it before the trial rather than during it.
Work area and access: how far the machine has to travel, how many doors and corners it passes, whether it has to leave the building, whether a hose run crosses a walkway. Deposit type: grease, oil, food or biological residue, dust, coolant film, salt, and how long each deposit has been on the surface. Surface list: painted panels, stainless frames, anodised parts, seals and gaskets, bearings and shaft ends, electrical enclosures, instrument openings, bonded joints, and any surface with a coating rating.
Throughput: how many items or square metres per shift, and how many people are available. Services: water source, drainage route, permitted wash area, power point, ventilation. Tolerance for wetness: how long a surface may stay wet, whether a dry finish is required before the item returns to service, and what drying material is acceptable. Finish standard: a written definition of clean, with photographs, so the trial result is not argued afterwards.
A specification column belongs on this sheet too, but as questions in the exact unit each page uses: supply phase; heater rating; tank capacity and tank material; water inlet connection and whether the arrangement supports continuous running; footprint and mass; and the protection list with the conditions and the reset route. Ask for the exact unit per row rather than translating values, because the site's own notation for values is not always consistent.
Compare model families with trial questions, not brochure values
Fixed versus mobile is the most consequential model choice, and it rarely turns on the heater. It turns on where the work happens. A fixed unit assumes the work comes to it and that the site can run a hose to the job; a mobile unit assumes the machine follows the work and that its own tank and battery arrangement are enough for the task length. If the job is one position and a defined shift, the fixed arrangement is usually the simpler answer, and the question becomes whether its published scope covers your duty. If the job moves between bays, or if a hose run or an extension lead would be unsafe, the question is whether the mobile arrangement sustains the task length. If the work is genuinely outside both published scopes, the honest route is customized steam cleaning equipment and a trial against the built specification, not a forced fit to the nearest catalogue entry.
Variant comparison is the second decision, and on this site the same base model can be sold in more than one arrangement. The walk-behind model is documented with a lead-acid option and a maintenance-free battery option. The published pages say what the model offers; they do not publish a run time, so the trial is what tells you whether either arrangement covers your task length without a mid-task stop. Compare variants on the questions your own sheet raises: task length, refill frequency, recharge or swap time, what happens when the task is interrupted, and where the machine waits between shifts.
Model and model are the third comparison, and it is usually about reach and duty rather than about power. The catalogue distinguishes hand-pushed machines with a tank from walk-behind and wheeled machines, and from fixed steam machines. Ask what each family has to reach, what it has to carry, and what it has to survive being dragged through, and score those questions on your own surfaces.
Design the trial so it can decide between models
The trial is only a decision tool if it is set up to discriminate. Run the candidates over the same surfaces, prepared the same way, with the same operator, and score them against the sheet rather than against each other.
Start on sacrificial or removable items rather than on the asset itself: spare panels, a removed guard, a detached cover, a spare hose reel or bracket. Where the real asset has to be used, cover an area with removable tape or film and work only on the covered patch, photographing before and after so the change is visible rather than argued. Keep at least one untreated reference area. Where your job is outside a model's published scope, begin with the shortest possible run and record what happens, rather than running a full cycle and explaining afterwards.
Write the stop conditions before the trial starts, not after an incident: any sign of water ingress into a bearing, motor, switch, connector or instrument; any lifted, softened or blistered coating; any displaced lubricant, label or seal; any loose or cracked insulation; any visible substrate damage. A trial that can be stopped is a trial that produces evidence; one that runs to failure produces a repair.
Keep the trial bounded so its result does not overreach: one shift or a few hours, two or three surfaces, the deposit you see most often, and the access route you actually use. A bounded trial answers whether a model suits your job; an unbounded trial generates anecdotes.
Score each candidate against your sheet and keep the record
Score the trial on the dimensions your sheet carries, and let the same scores decide between models. The useful dimensions are few and observable.
Effectiveness against the finish standard: did the patch reach the written standard, and on which surfaces did it not. Time and labor: elapsed time for the same area with the same operator, including setup, hose handling, refills and drying - this dimension is where fixed and mobile arrangements usually separate, because a machine that is easier to push can still lose time on refill and recharge. Consumables and services: water drawn, water source used, drainage actually needed, how often the tank ran out. Access and handling: whether the machine passed the doors, corners and floor condition you recorded. Quality and risk: the stop conditions, and anything they touched. Operability: how quickly a second operator reached an acceptable result, and what the written instruction had to contain.
Keep the record per candidate and per patch: date and shift, operator, surface and deposit, mode and distance and dwell, before and after photographs, the score against each sheet row, and any exception. That record is the evidence that the model decision was established rather than assumed. A trial without a record establishes nothing; a trial with a record can be re-read by whoever signs the order.
Keep the site's own documentation in the record as well. The PowerDe cleaning equipment category is where the families, variants and application scope live; the trial record is where your duty and your result live. The two should be read together, and neither should be quoted as the other.
Turn trial results into a model decision without over-claiming
Three results are possible, and all three are useful. A candidate meets your standard on your surfaces within your services and your task length: the decision is that model, and the record says which surfaces and which deposit it holds for. A candidate meets the standard but not the arrangement: the mode or the mode plus consumables is right, the mounting is wrong, and the decision moves to the other family rather than to another supplier. Neither candidate meets the standard: that is a requirement problem, not a catalogue problem, and the next step is an engineered specification with the trial conditions attached.
Keep the wording honest when the result travels into a purchase document. Write it as: trial on named surfaces, named deposit, bounded duration, result against a named standard, model and variant named. Do not write it as a cleaning rate, a cleaning temperature, a pressure, or an assurance for other surfaces. Do not carry a figure from one page onto another model: the site's own model pages publish overlapping specifications, and a fixed machine's published tank and heater values describe that entry alone.
Where the trial shows the duty sits outside every published scope, ask for the built-to-order route and repeat the trial against the engineered specification, including the utility arrangement and the protections, before the order. Where the trial shows the duty is inside a published scope, still record the surfaces, because the scope describes positioning, not a warranty over your plant.
Questions to put to the supplier after the trial, not before
The most productive supplier conversation comes after the trial, because you can now quote your own conditions. Ask which model or variant your trial record maps to, and which parts of your job the supplier says are outside the published scope; what the machine's protection list is and what each protection interrupts; how the water connection is expected to work over a full task and what the refill or drain routine costs in operator time; how the tank material and the machine construction are specified for your deposit type; whether spare lance, hose, seal and consumable availability matches your shift pattern; and what happens when the trial result is used for a second site.
Ask for specification answers in the exact unit each page uses, and refuse merged values. Ask for the application scope of the exact entry you are buying, not of the category. Ask which parts of your trial record you may reuse if you replicate the job elsewhere, and which parts you may not.
A trial establishes a model decision when both sides can read the same record. Your sheet defines the requirement, the trial defines the conditions, and the record defines the result - and the decision inherits exactly those limits, no more.
