Fifteen Hundred Seats Is a Fire Code Number. It Is Not Fifteen Hundred Diners.

A client sent us his architect's drawings and asked for tank dimensions to fit the volume he could allocate. He had done the sensible thing and had numbers ready before he called: the project's own plumbing sheet carried a design flow, and the architect's plan carried a seat count for the restaurant. Both numbers were real. Both had been produced by a professional. Neither of them measured what I needed to size a plant against.

It is worth separating this from two things I have already written about. Three ways a flow number usually arrives covers the headcount, the fixture count and the owner's own estimate, and the water bill that beat all three — that post keeps the arithmetic and I am not repeating it here. How much a finished box can treat depends on which tenants get signed covers the case where the concrete is already poured and the tenancy mix moves the answer. This note is a fourth case, and it is the nastiest of the four, because nothing about it looks like a guess. The figure is printed. It is authoritative. It is arithmetically correct. It is simply counting something other than people producing sewage.

Three quarters of the plant sat in one line of a schedule

The owner told me what he could give me: he could allocate for a hundred and fifty cubic metres a day. That is a useful way for a client to open, and I say so, because it tells me the constraint before I have spent a week discovering it. So I built the demand up from his building to see whether a hundred and fifty was generous, tight or wrong.

When I added the pieces, the restaurant came out at roughly three quarters of the whole figure. Everything else in the development together — retail, the offices, the staff, the common areas, the housekeeping water — made up the remaining quarter. One line in one schedule was carrying the plant.

That is not unusual in a mixed-use building and it is not a mistake on its own. Food and beverage is the heavy tenant in almost every development we look at, for flow and for strength both. What made me stop was not the share. It was the number the share had been calculated from.

The restaurant was described as fifteen hundred seats.

What a seat count on an architect's drawing actually measures

Fifteen hundred is not a forecast of how many people will eat there. It is an occupant load: the figure a fire code calculation produces by dividing floor area by an area-per-person factor for that use, and it exists to size exits, doors, stair widths and egress routes. It answers one question, which is how many human beings might lawfully be inside that room at the worst moment, and it answers it deliberately high, because the consequence of getting it low is people who cannot get out.

A sewage design asks a completely different question. Not how many could be in there, but how many will be in there, with water running, on an ordinary day — and then what happens on the busiest day of the week. Those two numbers are related only in the loosest way. A dining room laid out with tables, service aisles, a bar, a waiting area and a kitchen behind it seats a fraction of its egress occupant load. The egress figure does not care about furniture. That is the whole point of it.

So the number was not wrong. It was being asked to do a job it was never computed for. The honest basis for that restaurant is the seating layout — tables and covers as the designer intends to arrange them — and if a kitchen is involved, the kitchen water as well, which is a load in its own right and does not scale with diners at all. We have written separately about why what arrives from a kitchen needs catching before the plant, and about what commercial kitchen water does to equipment upstream.

The second number on the same drawing was doing the same thing

While I was looking at the seats, the plumbing sheet gave me the other one. It carried a figure labelled as a daily peak flow, a little over twice the daily volume anybody was talking about.

A figure like that is almost always a peaking factor applied to a daily volume. Somebody takes the average day, multiplies it by two or two and a half to represent the busiest hours, and prints the result. That is legitimate engineering and it has a purpose: you size a pipe or a pump for the moment of highest flow, not for the average, because a pipe that passes the average and surcharges at lunchtime is a pipe that has failed.

What you cannot do is take that number into a tank schedule. A treatment plant is sized on the mass of load arriving over a day and on the time the biology needs to work on it. Feed a peaked figure into that calculation and you have bought roughly twice the concrete, twice the air and twice the running cost to treat a volume that does not exist — and you have done it with a number that will survive every review, because it is printed on a consultant's sheet with a professional title next to it.

The right place for a peak is in the equalisation tank, which is the one chamber whose entire job is to accept a peak and release a flat flow into the process behind it. Peaks are a hydraulic problem solved by buffer volume. They are not a reason to double the plant.

Two numbers, one failure mode

I did not think much of it at the time, but writing it down now, the two figures fail in exactly the same way, and that is worth naming because it is the useful part of this note.

Both are real measurements of real things. An occupant load genuinely measures the safe capacity of a room. A peak flow genuinely measures the highest rate a pipe must pass. Both were produced correctly by people who were doing their own jobs properly. And both were about to be used as the design basis for a sewage treatment plant, which is neither of those jobs.

The failure is not in the number. It is in the handover. Nobody checks the units of a figure that arrives with a seal and a sheet number on it, and the further a number travels from the person who calculated it, the more authoritative it looks and the less anybody remembers what question it was answering. By the time it reaches a tank schedule it is just a quantity. I have written the same thing about a regulation clause copied into a specification from somewhere else, which had also travelled a long way from whoever first wrote it, and about a BOD figure on a drawing that nobody could trace. It is the same disease in a different organ.

Which is why I would rather a client checked our own numbers than took them on trust. Every figure in a design should be able to answer the question of what it is measuring and who measured it.

What it does to the plant if the seating layout comes back smaller

Here is the arithmetic, because this is not a semantic point. It is a plant.

Against a hundred and fifty cubic metres a day, the restaurant at fifteen hundred seats is about a hundred and twelve of them and everything else about thirty-eight. If the seating layout comes back at a third of the egress figure — which is an ordinary outcome once you draw in tables, aisles, a bar and a kitchen — the restaurant falls to about thirty-seven, and the total to about seventy-five.

Half the plant. Not ten per cent, not a refinement. Half.

Half the plant is a different box, a different civil works quantity, which we price three separate ways, a different excavation, a smaller blower, a smaller clarifier, and a running cost per cubic metre that the owner lives with for the life of the building. It is also a different declared flow on a discharge permit, which is a ceiling he agrees to and is then held to, so it is not a figure to inflate casually either. And it changes what the job costs before anybody pours anything, which is where almost all of the cost of a plant is actually decided.

What we did, which was not to reduce it on our own authority

I want to be careful here, because the tempting move is the wrong one. Having worked out that fifteen hundred is an egress figure, the clever thing to do is divide it by three, design to that, and look decisive.

We did not do that, and we would not. I do not know how many covers that dining room will have. I know the fire code number is not it. Those are two different states of knowledge and only one of them is mine to act on. Dividing somebody else's number by a factor I invented would replace a traceable wrong figure with an untraceable one, which is worse, because at least the first one could be checked.

So we did three things. We asked for the architect's seating layout, naming exactly what we wanted from it: covers as laid out, kitchen type, and service hours. We flagged in writing that the seat figure appeared to be an occupant load and said why, so that if anybody downstream wanted to keep using it they were doing so knowingly rather than by accident. And we carried on with the layout work against the volume the owner had allocated, because he had given us a constraint and the tank geometry inside that footprint was work we could do while the question was open.

That last part matters more than it sounds. The seat question changes the plant's capacity, but it changes almost nothing about the shape of the box, the depth available, the cell order or the structural case — and on that job the structural case was the interesting one, because a buried tank on a high water table tries to float when it is empty. Work that does not depend on the open number should not wait for it.

Designing while a number is still open

The way we handle an open figure is to design at more than one case and show them side by side rather than picking one and presenting it as fact. It is the same instinct as sizing to a band instead of a single number: the owner sees what the plant becomes at the fire code figure, at a plausible real seating, and somewhere in between, and he can see which decisions move with it and which do not.

Some of them do not move at all. Screening, grit and grease removal are set by what arrives rather than how much. The choice of treatment technology rarely turns on a factor of two at this scale. The media we use in the fixed-film stages does not change. The things that do move are volumes, air, and the awkward fact that a small plant is harder to run than a large one, so halving a plant is not purely a saving.

And one thing moves in a direction people do not expect. A plant sized for twice its real load does not run twice as cleanly. It runs badly, because the biology is underfed relative to the volume it is sitting in, aeration is oversized for the oxygen demand, and the operator is managing a tank that never reaches the condition it was designed for — and the operator will not be the same person in two years. Oversizing on an inflated number is not a safety margin. It is a different plant with its own faults, and the owner pays for it monthly.

Five questions worth asking about any number in your own design basis

1. What does this figure measure, in words, without using its label? If the answer is the label again, nobody knows.
2. Who calculated it, for what purpose, and on which sheet? A number with no author is a number nobody will defend when it turns out to be wrong.
3. Is it a maximum, an average or a peak? Plants are sized on daily mass and on retention time. Pipes are sized on peaks. Exits are sized on maxima. They are not interchangeable.
4. Which single line in the basis carries the largest share of the demand, and how confident are you in that one line? If three quarters of your plant sits on one assumption, that assumption deserves three quarters of the scrutiny.
5. What happens to the plant if that line is half what it says? If the answer is a materially different plant, the number is worth a week's delay to establish properly.

None of those five require an engineer to answer. They require somebody to be willing to ask where a figure came from, which is a different skill and a rarer one. If you are holding a design or a quotation and want to put those questions to it, that is the reading we do when somebody sends us a finished drawing and asks for a price, and it is what we offer as a consultation. It is also what separates two quotations that look the same — we have written about reading a quotation scope by scope rather than on its total.

Where this does not transfer

An occupant load is not always the wrong basis, and I should say so. On a building where the population genuinely is the occupancy — a dormitory, a school at full attendance, a staff canteen serving a known shift — the headcount and the design figure converge, and using the printed number is reasonable. The gap only opens where the use involves furniture, service area and turnover, which is why restaurants, function rooms and events spaces are where it bites hardest.

Nor is every peaked flow figure a trap. On a plant where most of the day's water genuinely arrives in a few hours, the peak is the design case for a great deal of the works, and treating it as an average would be the error running the other way. The question is never whether a peak matters. It is which part of the plant the peak is sizing.

And this was an existing design basis handed to us before anybody had measured anything. A building already operating has a water bill and a history, which is the method in that earlier note; in that situation none of this arises, because you are reading measurements rather than inheriting assumptions. A retrofit and a new build are two different jobs for exactly that reason. A lab result on the real influent beats every method on this page, which is why we ask for one and almost nobody has it.

Finally, every site differs. The share a restaurant carries, the gap between an egress figure and a real seating, the peaking factor somebody chose, the standard the discharge has to meet and the class of the water it goes to all move the arithmetic on this page around. What travels is not the figures. It is the habit of asking what a number is counting before you build a plant on it. If you want to see roughly where your own volume lands, there is a price estimator and a design generator on the site, and the parameters that matter by industry are set out separately. Our broader approach to wastewater treatment and the plants we have finished are there too, as is what we do across Cebu and as a design-build contractor nationally.

If you are holding a design basis you did not produce, and one line of it is carrying most of your plant, send it over and we will read it. You can book a free 15-minute consultation, or start a new plant enquiry with us if you are earlier than that.