Technical guide · Separators
Gravity disc selection, and what the wrong disc costs
A gravity disc is a flat ring with a hole in it, one of the cheapest parts inside a separator, and the part that decides whether the machine cleans the oil or quietly throws some of it away with the water. Nothing on the disc says which oil it suits, and neither does the nameplate. The right size follows from the oil in the tank today.
What the disc is actually doing
Inside a purifier bowl two liquids spin at the same speed, both pressed outward. Cleaned oil leaves near the axis over a ring on the light-phase side, separated water further out over the bore of the gravity disc. Each stands as a column, pressing inward on the boundary where they meet according to its density and its depth, and the boundary settles where the pressures balance. That is the interface: an equilibrium, not a feature machined into the bowl, and the disc is the only thing that moves it.
Open the bore and water spills at a larger radius, so its column is shallower and presses less; close it and the water is held deeper and presses harder. The maker's manuals put it in a line: a larger hole moves the interface towards the bowl periphery, a smaller hole places it closer to the bowl centre.
It belongs in a narrow band: outside the disc stack edge, inside the outer edge of the top disc. Inside the stack edge, water intrudes into the channels the oil rises through; past the top disc, the seal keeping oil out of the water side is gone.
The disc is also half of a pair. That light-phase ring fixes the other radius in the same balance, and on many bowls it changes when the machine is converted between purifier and clarifier duty. Rebuild a bowl with the wrong one and the disc takes the blame.
The oil decides, not the machine
The right disc therefore follows from how far apart oil and water are at the temperature the oil is treated at, not from what the machine is called. Two identical separators on one bedplate, same duty, can need different discs because one is still on the last bunker. What the ISO 8217 grade on the delivery note stands for is set out in the marine fuel grade reader.
Density on a bunker delivery note is quoted at a fixed reference temperature; the separator does not run there. The same reference temperature is what the bunker quantity calculator uses to turn a volume at tank temperature into the volume at 15 °C and tonnes. Both liquids thin as they heat, but not at the same rate: in the upper part of the range used for residual fuel, water loses density faster than the fuel does, so the gap the separator works on is narrowest exactly where heavy fuel is treated. That is why a published upper fuel density exists beyond which no disc will hold an interface at all: at treatment temperature the fuel and the water have become, to a centrifuge, one liquid. That figure is in your maker's documentation, not in a supplier's catalogue.
The charts also assume fresh water as the heavy phase. Oil carrying sea water works against a heavier heavy phase than the chart expects, and the maker's note says that may call for a larger hole.
Two ways to be wrong, and only one of them shouts
Too small a bore holds the interface too far in. The water layer thickens until it reaches the rising channels in the disc stack and water starts leaving with the cleaned oil. That failure announces itself: a water alarm, water in the service tank, emulsion at the outlet.
Too large a bore holds the interface too far out. Past the outer edge of the top disc the seal breaks and oil leaves the bowl with the water. From outside this looks like nothing: the cleaned oil is clean, cleaner than usual even, and no alarm is wired to the water side. The evidence sits where nobody looks: a sludge tank filling early, a bunker running out ahead of the figures. That is the failure worth fearing, and the one least likely to be reported.
The maker's fault list shows the asymmetry. Poor separation points to a hole too small; oil in the outgoing water, and a broken water seal, point to a hole too large. Both entries carry a second cause unrelated to size — the seal ring under the disc, which when perished gives the same picture as a disc a size too big. Order the ring with the disc.
On a self-cleaning machine there is a timing clue. The water seal is remade after every sludge discharge, so a disc marginally too large can hold for hours and let go in the minute after one. A loss that appears only around discharges is the machine showing where its margin went.
Fuel oil and lubricating oil are not the same problem
On heavy fuel the oil and the water are close in density and the exercise runs near its limit: small changes matter, and a bunker change is a reason to revisit the size. Lubricating oil duty is roomier but sharper. The oil is far lighter than water, so the choice is less delicate; what changes is where the water comes from — a cooler, a stuffing box or condensation, not a bunker barge — and what is at stake. A lube oil purifier losing oil to the sludge side is draining a sump the engine depends on.
Three settings that are really one setting
Disc, separating temperature and feed rate cannot be chosen independently, and the maker's chart says so by taking all three at once. But the chart is a calculation from densities, and the makers are candid that it is theoretical: it gives the disc to try first, not the disc to fit and forget.
What it leaves out is that the oil has to be pushed through the stack and out over its ring. Anything that makes that harder thickens the oil layer and shoves the interface outward, toward the seal-breaking side. Cold oil, thick oil and too much feed all do it. That is why the same fault list gives a temperature too low and a throughput too high as causes of a broken water seal, in the same column as a disc too large. The maker's rule follows: the heavier or more viscous the oil and the larger the feed, the smaller the bore — and fit the largest disc that still holds the seal.
So a disc is correct only against a stated temperature and a stated feed rate. A machine driven hard to catch up on tank levels, or with a heater that cannot keep up in the cold, has moved two of the three variables it was chosen against. Where the outlet is pumped, back pressure on the clean oil side belongs in the same balance. For a fuel, the fuel viscosity-temperature calculator shows how thick it still is at the temperature the heater actually reaches, from two measured points.
The machines that no longer have one
Once fuel densities passed what a fixed dam could hold, it had to go. Alfa Laval ALCAP machines have no gravity disc: a transducer watches the water content of the cleaned oil and the control system discharges when it rises, so the unit follows the fuel instead of being set for it. GEA's unitrol and the water detection on Mitsubishi SelfJector machines answer the same question by measurement rather than geometry.
That did not retire the disc. Conventional purifiers stay in service across an enormous installed fleet, and on lubricating oil duty the margin never demanded anything cleverer. Hence the split behind every separator enquiry: on one machine the disc size is part of the order, on the other the part does not exist. Asking which, first, separates an order from a week of correspondence.
Frequently asked questions
What happens if the gravity disc bore is too small?
The interface sits too close to the bowl centre. The water layer thickens until it reaches the rising channels in the disc stack and water starts leaving with the cleaned oil. This failure is loud: a water alarm, water in the service tank, emulsion at the outlet.
What happens if the gravity disc bore is too large?
The interface sits too far towards the bowl periphery and the water seal at the top disc breaks, so oil leaves with the water. From outside this looks like nothing — the cleaned oil is clean — and the evidence only shows up later, as a sludge tank filling early or a bunker running out ahead of the figures.
Does every marine separator have a gravity disc to size?
No. Alfa Laval ALCAP machines have none — a transducer watches water content in the cleaned oil and the control system discharges instead. GEA's unitrol and the water detection on Mitsubishi SelfJector machines answer the same question by measurement rather than geometry. Conventional purifiers, and most lubricating-oil duty, still use one.
What to send when ordering a disc
The designation alone will not do — it names a bowl size and a generation, not the oil.
- The marking on the fitted disc, with a caliper reading if the stamp is worn. The fitted disc is evidence, not an answer.
- The oil described by what it is — residual, distillate or the lubricating grade — with its density and the temperature that figure is quoted at.
- The temperature the oil is actually treated at, read off the running machine rather than the plan, and the feed rate it is set to.
- Make, model and serial number from the nameplate, with a photograph.
- Duty and configuration — purifier or clarifier, and which ring is on the light-phase side if the bowl has been converted.
- What the machine is doing, if the request came from a fault: water in the oil, oil in the water, a seal lost after discharges.
- Whether you want the set rather than one size.
What Vessel Core will not do before an enquiry is reviewed
Vessel Core will not tell you which disc your machine needs. Nobody outside your engine room can: the answer comes from your oil, your treatment temperature and your feed rate, read against your maker's chart and proved by running the machine and watching both outlets. A supplier can quote the size you identified, the set around it and the seal ring with it, and say plainly when the machine has no disc.
Send the enquiry through the RFQ form with a nameplate photograph and, if you have it, a photograph of the fitted disc. If the model is unclear, the no-part-number route explains what to send instead; wider context sits on the separator spare parts page and in the separator sourcing guide.
Maker names identify equipment only. Vessel Core has no dealership, distribution, representation or partnership agreement with Alfa Laval, GEA Westfalia, Mitsubishi or any other manufacturer, and works as an independent spare parts supplier.
