Installation and Commissioning Are Not the Same Job
Installation is a logistics exercise. The equipment is moved to the position on the layout, bolted down, connected and switched on. Commissioning is a proof exercise: it demonstrates that the machine, the utilities, the media, the formulation and the operator behave as one system, and it produces the evidence that they do.
The distinction matters because the two are usually scheduled as if they were one item. Installation finishes when the cable is terminated. Commissioning finishes when a batch has been measured, not when the motor turns. Everything in between is the part that consumes the calendar.
In practice commissioning runs as four phases, and they are sequential for a reason. Each one produces a condition the next one depends on:
- Site readiness, the plant can supply power, cooling, drainage, media and material at the point of use, at the quantity and quality the machine was specified for
- Mechanical checks, the machine is level, grouted, aligned and rotating in the right direction, with every interlock proven before anything goes into the chamber
- The water run, the seal, the cooling circuit and the control loop are proven with a harmless charge, and grinding media is added in stages only after that
- The first charge, a deliberately slow batch is run and sampled to build a fineness curve against specific energy, which becomes the operating window for production
Skipping ahead is the single most common cause of a start-up that drags. Charging media into a machine whose seal has not been proven means that any leaking found later has to be fixed with the chamber full.
What Site Readiness Actually Requires
Site readiness is unglamorous and it is where the schedule is won or lost. The list below is the one that matters most often, in the order it usually bites. Note that almost none of it is equipment supply, which is why it sits in the customer’s scope and why it slides.
| Requirement | What ready looks like | Why it stops the line |
|---|---|---|
| Floor and anchor positions | Level, cured, anchor holes set out to the certified drawing, grout available on site. | Correcting level after grouting means cutting anchors and re-setting the baseplate. |
| Power and protection | Supply sized for the installed motor rating, with protection and isolation matched to the drive. | A supply that trips on inrush cannot complete a water run, so nothing downstream can be proven. |
| Cooling water | Flow and inlet temperature at the specified values, with a return path that can be checked visually. | Without cooling there is no way to measure the temperature rise, which is half of the acceptance criteria. |
| Compressed air | Pressure and dryness adequate for any pneumatic lift, seal or valve on the machine. | Hydraulic and pneumatic lift machines cannot even be positioned for the mechanical checks. |
| Drainage and washdown | A floor drain or sump within reach of the discharge and the sample points. | Water runs and cleaning produce volume, and there is nowhere for it to go. |
| Grinding media | Correct material, size and quantity, on site and dry, before the water run finishes. | The chamber cannot be charged, so the first charge cannot start. This is the most frequent single delay. |
| First-batch material | Enough resin, solvent, pigment or filler to run a full batch plus a second one for comparison. | A half batch cannot be sampled properly, and a single batch proves nothing about repeatability. |
| Access route | The largest item can physically reach its position, at the width and height it will arrive in. | Machines that will not go through the door get partially dismantled, which resets the alignment work. |
Two of these deserve emphasis. Grinding media selection is not a purchasing afterthought; bead diameter, density and filling ratio set the energy density in the chamber, so the media specification is part of the process design rather than a consumable. Our guide to grinding media selection covers the sizing logic in detail.
The second is the first-batch material. Ask for two batches worth, not one. A single batch shows that the machine runs. Two batches, run identically, show whether it repeats — and repeatability is the property the plant is actually buying.
Mechanical Checks That Cannot Be Deferred
These are the checks that become expensive the moment anything is put inside the chamber, because repeating them afterwards means emptying it first.
- Level and shim, the baseplate is set on shims and checked in both axes before grouting. A mill that leans loads its seal unevenly, and an unevenly loaded seal is the first thing to weep
- Grout, poured only after level is confirmed, and cured before the machine is run. Grout is not a packing material; it transfers load, and it cannot do that while it is wet
- Coupling alignment, checked with a dial gauge or laser in both angular and parallel planes, at the operating temperature if the drive runs hot. Alignment done cold on a warm drive is alignment done twice
- Rotation direction, confirmed by bumping the motor before the chamber is filled. A machine run backward can unseat a mechanical seal in seconds
- Seal and jacket integrity, proven with a pressure test and a short water run, watching for a pressure drop on the cooling circuit and for any trace at the seal face
- Interlocks, every one proven to stop the drive, including the cooling flow switch, the chamber pressure limit and the lift or lid position sensor where fitted
The platform arrangement in the photograph above is common on coating and adhesive plants, and it concentrates risk: several vessels share one deck and often one utility manifold, so a leveling error at one end propagates. It is worth spending a morning on the deck before a single vessel is anchored to it.
The Water Run, Then Media in Stages
The water run is the least expensive test in the whole sequence and the one most often shortened. Clean water goes in, the machine runs at low speed, and four things are recorded: whether the seal stays dry, whether the cooling circuit holds pressure, whether the no-load current matches the expected figure for that speed, and how much the water temperature climbs over a fixed interval.
That last number is the useful one, because it establishes the baseline heat input before any product is in the machine. Once the batch is in, the temperature rise is compared against that baseline, which separates heat from the grinding work from heat from a restricted cooling path. Without the water baseline, the two are indistinguishable.
Media charging happens after the water run passes, and it happens in stages with the machine stopped between them. Charging in one pour risks a wedged shaft and a stalled drive, and it gives no opportunity to observe the current step between fill levels. Watching the current rise in steps is a direct check that the filling ratio is what the process design intended — a chamber that draws noticeably more than expected at the target ratio is usually overfilled.
On basket and immersion designs the geometry check comes first: the basket has to sit at the intended depth in the vessel, with clearance at the bottom and enough travel left for the lift to work. Getting that wrong is not a process problem, it is a positioning problem, and it is much easier to fix before the vessel is full. The same logic of matching machine geometry to the material is what separates wet grinding from dry grinding as a process choice, not just a machine choice.
The First Charge Is a Measurement, Not a Production Run
The instinct on a new line is to run the material at the rate the plant will actually use, because that is what the customer is waiting for. It is the wrong instinct. The first charge exists to find the shape of the fineness curve for that formulation with that media charge in that machine, and a curve needs points.
So the first batch runs slowly and is sampled after each pass. Each sample gets a fineness reading, and each reading is paired with the energy consumed to reach it. Plotted together, they show where additional passes stop buying fineness and start buying heat. That point, not the mill’s rated capacity, is the operating window. Once it is known, throughput is raised to just below it — and the throughput relationship can be set from evidence instead of from the nameplate.
The temporary loop in the photograph is deliberate. A short circuit through a tote lets the operator take a sample every few minutes without disturbing the plant piping, and it lets the media charge be adjusted between runs without a full drain-down. Permanent piping goes in once the window is known, not before.
Temperature is the constraint that usually decides when to stop. The work going into the chamber comes back out as heat, so the rise across the chamber tracks the specific energy being applied. A properly sized cooling circuit, a correctly charged chamber and a sensible flow rate will hold that rise to around ten degrees Celsius; if it climbs well past that at modest throughput, something is wrong, and continuing will bake product onto the internals rather than make it finer.
Reading the First Batch
Three-roll refining and bead milling fail in different ways, and the symptoms below are what the first batch tends to show when the setup is off. The value of running a deliberately slow first charge is that these appear one at a time instead of all together.
| Symptom on the first batch | Most likely cause | What to change |
|---|---|---|
| Fineness stops improving while power keeps climbing | The energy is going into heat rather than into new surface; the chamber is past its window. | Reduce the number of passes and raise the flow rate so product leaves the chamber sooner. |
| Temperature rises further on each successive batch | Cooling flow is below the specified value, or the jacket has an air lock. | Measure flow at the return, vent the circuit, and compare the rise against the water-run baseline. |
| No-load current is higher than expected | Media charge is above the design filling ratio, or media has compacted at one end. | Stop, drain, and recharge in stages while watching the current step. |
| Separator or screen blinds within a few hours | Flow is too low to carry product out, so solids pack against the separation element. | Raise the flow rate first; check media size against the slot or screen opening second. |
| Seal weeps on the first warm run | Alignment moved after the cold check, or the seal was run dry at any point. | Re-check alignment hot, and confirm the seal is pressure-fed before the drive starts. |
| Color strength reads low against the retained reference | Dispersion is incomplete, not grinding: the pigment is under-wetted before it reaches the mill. | Fix the pre-dispersion stage before touching the mill settings. |
On a three-roll refiner, the apron film is the in-process control that costs nothing. Paste that runs down in a smooth, uniformly colored sheet is dispersed; paste that shows streaks, specks or a grainy texture is not, and no amount of additional passes at the same gap will fix it. The working principle behind the three-roll refiner explains why the gap sequence, not the pressure, does the work.
What Only Shows Up After Two Weeks
A start-up that passes on day one can still be wrong by week three, because several failure modes are slow. These are the ones worth watching for in the first month, and they are the reason a start-up report is not the same document as a performance review.
- Media wear outrunning the budget, the charge is too aggressive for the target fineness, or the media material is wrong for the product’s hardness. Measure a sample of media before and after, not by eye
- Progressive separator blinding, slots or screen openings are marginal against the media size. It looks fine cold and clogs warm
- Seal face wear, usually a cooling or flush deficiency rather than a seal defect. Check the flush is present at every start, including short ones
- Coupling and baseplate movement, grout that was never fully cured, or a drive that was aligned cold and runs hot. Re-check alignment after the first two weeks of production
- Control drift, temperature or pressure sensors reading consistently high or low. Compare against a handheld instrument before trusting the trend
All five are cheaper to catch in a scheduled two-week review than in an unplanned stop. The practical format is a short log: media weight sampled, separator differential, seal flush confirmed, alignment re-checked, and sensor readings against a reference instrument. Our maintenance and troubleshooting guide covers the diagnostic side in more depth.
Planning a Start-Up? Send Us the Layout
Tell us the product, the target fineness, the batch size and what utilities the plant can supply. Our engineers will confirm the machine sizing, the media specification and the commissioning sequence, and quote directly from the factory.
Send your process detailsChat on WhatsApp
WhatsApp +86 138 1677 4633 export@polycmill.com
Where This Equipment Is Built
Bead mill chambers, dispersion discs, roller assemblies and vessel shells are machined and assembled in the same plant, which is why commissioning questions usually reach the people who built the machine rather than a service sub-contractor.
Staging the machines in the plant before dispatch is worth the floor space. Every one of them is run through its full travel, its control sequence and its rotation direction before it is loaded, so the checks that happen on site are confirming a known state rather than establishing one for the first time. A line specification for a full paint and coating production line is built on exactly that assumption.
