Each Trap Holds About Eighty Litres. At Its Rated Flow, That Is Ninety Seconds.

The drawings for a new restaurant fit-out came through and they were tidy work. Four grease traps, each rated at fifteen gallons a minute, four-chamber, stainless, set flush in the finished floor. One at the dishwash, two along the cook line, one at the coffee and bar counter. Vents off every one of them. I had been asked to quote the dosing for it, so strictly speaking the traps were not my business.

I spent the afternoon on them anyway, because of a division nobody on the drawing had done. Fifteen gallons a minute is about fifty-seven litres a minute. A unit of that class holds somewhere between sixty and ninety litres. Divide one by the other and the trap gives the wastewater about a minute and a half of quiet at the flow it is rated for. That number is not a fault. But it decides almost everything about what the kitchen will be dealing with for the next ten years, and it was set on a plumbing drawing months before anybody there had thought about grease at all.

This is a different note from the three others on this blog about grease. The one on the grey crust that is soap rather than grease is about the chemistry of what forms in a trap and the hard water behind it. The one on why our conditioner has no enzyme in it is about what is in the bottle. The one on a dose of four hundred millilitres and the smallest pump we could buy is about the hardware that puts it in. This one is about the trap itself, and the choice made before any of those three questions exists. I am deliberately not re-arguing the pump here.

A flow rating is not a size

A grease interceptor is a residence-time device. It does not treat anything. It gives fat enough seconds of stillness to rise and solids enough seconds to fall, and it has a physical barrier that keeps the risen layer away from the outlet. That is the whole mechanism, and everything a trap does or fails to do comes back to how many seconds it can offer.

So a rating in gallons per minute is a statement about the worst minute, not a statement about capacity. It says this unit will still be separating when the dishwash sink is dumped at the end of service. Underneath it there is a volume, and the volume is what buys the time. Sixty to ninety litres against fifty-seven litres a minute is a minute and a half. If two fixtures discharge together into the same trap, it is less.

Set that against the other kind of installation, the one I spend most of my time with: a buried interceptor of several cubic metres in a yard or a driveway, serving the whole kitchen through one line. At the same instantaneous flow that device offers well over an hour. Two things sold as the same item, doing nominally the same job, with retention times that differ by a factor of fifty or more. Nobody writes that comparison down because the two arrive through different trades — one comes off a plumbing schedule, the other off a civil drawing. On our own new-build jobs we form the interceptor in reinforced concrete alongside the rest of the inlet chambers we draw for screening, grit and grease rather than buying a unit in, and we price it on its own line.

The small trap separates worse. It also never rots.

This is the trade, and it is the most useful thing I can tell an owner who is about to sign a plumbing drawing.

More retention time means better separation. It also means the contents sit still for hours or days. Still contents with fat in them go anaerobic, generate hydrogen sulphide, and in water carrying a few hundred milligrams a litre of hardness they build the metallic soap crust described in the note above. The crust needs time. It is a slow reaction between liberated fatty acids and the calcium that is already in the wash water, and a vessel that holds its contents for a day is a far better reactor for it than a vessel that holds them for ninety seconds.

A small trap emptied by a person every few days never gets the time to do either job well. It separates less, and it also never rots. The big device gives you good separation and a crust. The small device gives you weaker separation and fresh contents. Neither of them is the better engineering answer in the abstract. They fail differently, and which failure you are signing up for is mostly a function of that retention number.

I find this the same shape as something I have argued about plants: a five cubic metre a day plant is harder to run than a five hundred, because small vessels have no buffer in them and depend on attention instead. A point-of-fixture trap is the same bargain in miniature. It trades volume for attendance.

Four traps means four tasks, and the labour has moved indoors

The buried interceptor is one task on a truck calendar. A hauler arrives, drums are filled, somebody opens a cover and scrapes, and in the bad version of it a person leans into an unvented pit to reach the surface. That is confined-space work and I will not soften it; I set out what hydrogen sulphide does to the people who go in there, including the fact that a particulate mask does nothing against a gas, in a note of its own. Provision for it is not a premium item either — we carry gas detection, forced ventilation and a second person at the hatch in our proven rates.

Four traps on a kitchen floor is a different arrangement entirely. Four tasks, inside the kitchen, done by people who are already standing in the room, with a bucket, every few days. Nobody enters anything. No truck, no drums, no hauler on a quarterly cycle. On this job that deleted a whole block of work I would normally be pricing.

It also moves the work into the kitchen, which has its own consequences and they are not small. Cleaning a trap now competes with service. It is done by staff whose job is food, and it will be done well or badly depending on supervision rather than on a contract. The honest version of this is the same point I keep making about plants: a good operator can carry a plant and cannot carry a bad design, and when people change, the knowledge leaves with them unless something written stays behind. Four small traps need a habit. One pit and a hauler needs a purchase order. Habits are cheaper and they are also more fragile, which is the same reason we treat how a plant is looked after as part of designing it rather than as something that starts after handover.

And the material still has to go somewhere, which is the part that never disappears. Whatever comes out of those four buckets leaves the building as waste, exactly as everything you take out of a plant has to have a destination. Moving the labour indoors does not change the mass of fat the kitchen produces. It only changes who handles it and how often.

These were vented. The pit I had been looking at was not.

Every trap on this drawing had a fifty millimetre vent on it. That single line is the reason a small trap set in a finished floor is not the hazard an unvented pit is. The headspace has a path out, the volume of the headspace is tiny, and nobody has any reason to open the lid and lean in.

A vent is only a vent if it goes somewhere acceptable, and inside a tenanted building that is a real question rather than a formality. The drawing said the draw-off and vent connections were to be verified on site before purchase, which is the right instruction and is also exactly the kind of instruction that gets overtaken by a delivery date. I have written about how a smell complaint turned out to be an air path rather than a treatment failure, and the lesson transfers directly: the air in these systems always leaves somewhere, and if the drawing does not decide where, the building decides for you.

Worth saying plainly, since it is the opposite of what people expect: the vent on a grease trap is not there to help the grease. It is there for the gas that the stagnant water underneath it generates, and for the air the water displaces when a sink empties into it. It is a safety and odour fitting, not a process one.

Where the dose has to go is decided by the layout, not by the pump

Back to the job I was actually quoting. Four traps in parallel, each taking its own fixtures, all of them joining a common kitchen waste line downstream. The obvious place to put a dosing point is that common line, because it is one point instead of four. It is also useless. Anything injected there is downstream of all four traps, and the only thing it will condition is the sewer.

So the topology is set by the trap layout, not by the equipment. Four traps is four dosing points, which is what drove the whole hardware question in the note on the dose and the pump. And the traps hold only sixty to ninety litres, with flow passing through them the whole of service, so the dose has to go in at the end of the day when the kitchen is closed and the contents are sitting still. A dose added during service is washed through within the minute and a half the arithmetic at the top of this note describes.

One point on that job is still dosed by hand, and I would rather state it than let it be discovered. The bar trap sits too far from the cabinet for a tube run I would trust, so on the automatic option three traps are dosed by the pump and the fourth is poured by a person. That is not elegant, and I did not find a way to make it elegant. Pouring from a marked bottle has nothing in it to fail, so the fourth trap is arguably the most dependable one in the set — but it is an exception in a system sold as automatic, and an exception has to be written into the handover or it becomes a missed trap in month four.

Four chambers inside a box the size of a drawer

These units are four-chamber. That matters more than it sounds, because compartments are what makes ninety seconds work at all. Most of the interceptors I open around Cebu are single open boxes with no internal baffle, which means nothing slows the water down and nothing holds the floating layer back from the outlet. Take the baffles out and you have a septic tank that happens to be near a kitchen.

The compartments cut both ways in a unit this small. They are what buys the separation, and they are also what makes the thing awkward to clean properly. The first chamber does most of the work and holds most of the solids, and in a four-chamber unit set in a floor it is the chamber that is hardest to get a scraper into. I expect the first chamber to be the one that gets skipped, for the same reason we no longer draw a weir plate into our clarifiers: an adjustment nobody can reach is not an adjustment, and a chamber nobody can reach is not a chamber. Before any of these are bought I would want to see somebody put a hand and a tool into the first compartment with the lid off, the same way we measured a manhole before believing a partition could go through it.

What I could not do on this job, and said so

The decision was already made. The traps were specified and not yet purchased, which sounds like a window and mostly is not one. There is no ground to dig in a fit-out on a retail floor, the drain runs are already set in a slab that is not the tenant's, and a tenant does not get to re-plumb a landlord's building. So I quoted the dosing for the traps that were going in, and I did not propose a different device. What I could do was write down in the proposed method of treatment what those traps will and will not do, so that it is on record before the first service rather than discovered during one.

That is the real finding and it is not a technical one. The trap is chosen by a plumbing consultant at design stage, thinking correctly about pipe sizes and fixture units, months or years before anybody has a grease problem. The person who will live with the consequence is not in that conversation. This is the same pattern as a seat count taken off a fire-code drawing and used as a population, or a clause copied into a specification from a regulation that had been dead for ten years: a number or an item that is real, printed and authoritative, carried forward by people who had no reason to question it. Where we are brought in early enough, reading a finished drawing before pricing it is most of the value we add, and the boundary between what we will look at for nothing and what we charge for is written up in the note on where our free design help stops.

Five things worth settling before the traps are bought

  1. Work out what each fixture actually delivers, and check the rating sits above it. The rating is a statement about the worst minute, so the question is which minute you are protecting against — one sink, or the dishwash dumped while the pot wash is running.
  2. Ask how many litres the unit holds, and do the division. Volume divided by flow is the retention time, and if it comes out under about a minute the trap is depending entirely on being emptied often.
  3. Confirm every trap is vented and settle where each vent terminates, verified on site before the units are ordered rather than after.
  4. Name who empties them and how often, and have somebody physically reach into the first chamber with the lid off before you buy four of them.
  5. Locate any dosing or cleaning point relative to the traps. Anything downstream of all of them does nothing at all.

None of the five is about the make of the trap, and none of them needs a laboratory. They are the kind of thing I would want settled before signing anything, in the same spirit as the four questions I would ask before signing for a plant. They are all arithmetic or a tape measure, which is the same argument I make about a water bill that had been measuring the flow for eleven years while everybody waited for a flow meter, and about reading a settling cone instead of buying an instrument.

What the trap decides downstream, and what it does not

A trap intercepts what it catches and nothing more. Fat that passes it carries on to whoever is treating the building, and in a multi-tenant development that is usually a central plant the tenant never sees. Fat arriving at an aeration tank shows up as brown greasy foam and as solids going over the clarifier weir, which is a problem the landlord then owns and the kitchen never hears about. Four tenant traps working perfectly do not fix a central plant that was built short, and I have watched that argument run both ways in a development more than once. If the plant behind the kitchen is the real question, the price estimator and the design generator will give you its shape, though neither of them sizes a grease trap.

It does not make anybody compliant either. The parameters a discharge is held to come from the receiving water, not from the kitchen — our pages on discharge compliance and the parameters that matter by industry set that out, the nutrient field guide covers what fat and the nitrogen alongside it do to a biological plant, and the pollution control officer help centre is the place to start if the question is a permitting one rather than a plumbing one. What an inspection actually asks for and what a declared flow commits you to are both separate notes.

And if the crust in an existing trap is already grey and hard, none of this is the fix. That is a water chemistry problem, and the answer is softening the kitchen hot water before the calcium ever meets a fatty acid, which is a sizing exercise with its own trap for the unwary and raises a question about pipework that is usually answered by the pipe schedule. Dosing a culture into a metallic soap crust is one of the clearest cases of when I tell people not to buy the product we sell. That the answer sits on the water treatment side of the business rather than the wastewater side is most of why it gets missed.

Where this stops transferring

Several of the conclusions above belong to this building and would be somebody else's differently.

A ground-floor restaurant with a yard behind it is a different problem, and there the external interceptor is usually still the better device — volume is the thing that buys separation, and a properly baffled, vented, regularly serviced pit is not a problem anybody needs to solve. Nothing here is an argument for ripping one out. A working installation and a failing one are two different jobs, and most of what I have written applies to a drawing, not to an existing kitchen.

Hard water pushes the sums the other way, in favour of the small trap. The crust needs residence time to form, so frequent emptying of a small volume is a genuine advantage against it rather than a compromise. In soft water that advantage mostly disappears and the case goes back to being about separation alone.

Volume of fat changes it again. A restaurant and a commissary are not the same site. A kitchen laminating pastry or frying at volume produces fat by the kilogram an hour and will overrun ninety seconds of retention whatever the rating on the lid says — I have built a plant twice for a food processing site because of what the fat did, and a dressing plant making most of its wastewater in three hours is the same lesson about peaks. Those need volume, which means an interceptor sized honestly and probably a tank whose job is to smooth the flow behind it.

The retention figures in this note come off one drawing and one class of unit. Rated flows and holding volumes vary between manufacturers by more than you would expect, and the only numbers worth designing against are the ones on the submittal for the unit actually being bought. Do the division yourself.

If you are planning a kitchen fit-out or working through a grease problem in Cebu, you can see the wastewater work we do across the region and what we have built, read who we are or how we approach wastewater, send the site details through the request form, look at how a plant gets built or how we work with clients, and browse the rest of these notes. Or just book a free 15-minute consultation and tell me what the trap is doing.