They Asked Us to Halve the One Tank That Treats Nothing.

A client sent our drawing back with one tank marked up. Not the aeration tank, not the clarifier, not the plant room. The equalisation tank, with a note asking whether it could be halved. His reasoning was perfectly sensible and I have heard it a dozen times: it was the largest single chamber on the drawing, there was no equipment in it, and it treated nothing.

All three of those things are true. The equalisation tank does not treat anything. It holds water and lets it out slowly. That is the whole of its job, and it is the reason every machine downstream of it can be smaller than it would otherwise have to be. It is also the tank that is cheapest to build on the day the slab is poured and close to impossible to add afterwards. So it gets cut, and then a year later somebody is looking at a lab sheet wondering why the plant fails on some days and passes on others.

This is what we are actually deciding when we size it, and where we have watched shrinking it come back.

It is a flow number, not a strength number

When we draw a plant against more than one wastewater strength, about half the concrete barely moves between the columns. Screening, grit, the clarifier surface, the disinfection contact volume, any treated-water storage — and equalisation — are all sized off flow and the shape of the day, not off BOD. The aeration volume and the air are what move when the strength moves.

That has a practical consequence. An owner who has no influent lab result, which is most owners, can still size his balance tank properly, because the information it needs is not a laboratory number. It is the operating pattern of the building, and he already knows that better than we do. What time does the kitchen close. When does the wash-down start. How many rooms, and do they all shower between six and eight.

The buildings that need it most rarely look like they do

The clearest case we have worked on recently was a dressing plant that made sixty to seventy per cent of its daily wastewater inside a three-hour wash-down window. On paper it was a modest daily volume. In practice, for three hours a day it behaved like a plant three or four times the size, and then for twenty-one hours it behaved like almost nothing at all. Sizing anything on the daily average there would have produced a plant that was wrong for every hour of the day.

Commercial kitchens do the same thing on a smaller scale, and they do it worst on the days nobody expects. We once traced a failed BOD three weeks before a permit sampling to a quarterly deep-clean the night before, with the grease interceptor pumped the day before that. The plant was fine. The day was not. Equalisation is what turns a day like that into a slightly stronger week instead of one bad sample.

Hotels and dormitories have a shape too, just a gentler one — breakfast, check-out, laundry, evening. And the smallest plants have the sharpest peaks of all, because there is no crowd to average anybody out. A 500 m³/day plant smooths itself. On a 5 m³/day plant, one function room booking is the entire day's design flow arriving in an evening.

What the peak does if you let it through

Nothing dramatic, which is exactly why it goes unnoticed for so long. Four things happen quietly.

The clarifier stops clarifying. When we see solids coming over the weir, the thirty-minute jar test sorts it: good compact sludge with cloudy water above it is not a biology problem at all, it is water moving through the tank faster than the solids can fall. People buy chemicals for that. What it needs is the peak held back upstream.

The disinfection contact time collapses when you need it most. Contact time is volume divided by flow, so doubling the flow for two hours halves the contact you paid for, in the two hours the flow is dirtiest.

The chemical dosing goes out of proportion. Phosphate removal is set against a ratio, and a dose calculated on average flow is simply the wrong dose during the peak — under-dosed on the way up, over-dosed on the way down. The same goes for alkalinity correction.

And the nitrifiers take a shock. The ammonia-oxidising population is the slow, fussy one: it doubles in a day or two while the carbon-eating bacteria double in hours. A slug of ammonia arriving in three hours does not give it time to respond, so the parameter that fails is the one with the least headroom — 4 mg/L as NH₃-N for Class C, 3 for Class B. We have written up what those three nutrients actually look like on site if you want the readings rather than the theory.

How we size it, and it is not a percentage

There is a rule of thumb in circulation that the balance tank should be some fraction of daily flow. We do not use it, because two buildings with the same daily flow can need tanks that differ by a factor of three. What we do instead is put the day on paper hour by hour — as well as the owner can describe it, which is usually well enough — take the cumulative inflow against a constant outflow, and read off the largest gap between the two. That gap is the volume. Then we add for the fact that the tank cannot usefully be drawn down to nothing and cannot be filled to the brim.

The outflow rate is the other half of the decision and it is the half people forget. Equalisation only works if what leaves is genuinely constant, which means a transfer pump on level control rather than gravity, and it means one duty and one standby rather than a single pump, because the one tank whose whole purpose is to never overflow is a poor place to have a single point of failure.

The tank also has to be kept moving. Raw sewage held for several hours with no air goes septic, and then what reaches the aeration tank is a sulphide-smelling feed that has already lost some of the carbon the denitrification stage was counting on. So the balance tank gets coarse air. On a small plant that means an air pump rather than a blower — the balance tank is often the chamber nearest the occupied part of the building, and a machine somebody switches off at night is worse than no machine at all.

Why it is the cheapest tank to build and the dearest to add

On the drawings we produce now, chambers are blocked into one monolith with shared cross walls rather than drawn as separate boxes with gaps between them. In that arrangement, extra volume on the balance tank is mostly one more bay along the row — a longer base slab, a longer top slab, and one additional cross wall. Our own tank rate has been running at about ₱14,000 per cubic metre of capacity on a 3,000 psi mix, and a client's own contractor working to our drawing came in around ₱10,500. For a sense of scale, one three-chamber tank of 70,000 litres with a 4,000 psi top slab came to about ₱1.2 million. Those are figures from specific jobs we actually sent out, not a rate card — excavation depth, water table, access and whether you are building beside an existing structure all move them, and your site is not that site.

Now price the same volume eighteen months later. It is a separate structure, so it is four walls instead of one. It needs its own excavation, beside a plant that is running and a building that is occupied. It needs new inlet and transfer pipework cut into a finished tank, a pump, a level control and power run to it. And it is very often the reason a plant has to be shut down for a week in a building that cannot spare the week. The volume costs a few hundred thousand pesos at design stage. The same volume as a retrofit is a different order of decision — which is the whole argument in design being where the money on a plant is actually decided. If the tanks are above ground rather than in the ground it is a much easier retrofit, and that is one of the quieter arguments for the above-ground option on a site that may grow.

What the tank buys back

This is the part that never appears on the same line as the concrete, and it is why we argue for the tank rather than just defend it.

Every machine downstream can be sized on the equalised rate instead of the peak. On the dressing plant, the choice was a small dissolved air flotation unit sitting behind the balance tank against a much larger one sized to swallow the three-hour peak. Same result, considerably less machine. The aeration turn-down is gentler, the clarifier surface can be smaller, the disinfection volume can be smaller, and the dosing pumps run at a steady rate rather than chasing a curve.

So the trade is concrete against machinery. Concrete is bought once, does not wear out, makes no noise and cannot be switched off. Machines cost money every month you own them, need spares, and are the things that fail. Given a choice between a bigger balance tank and a bigger machine to do the same work, we take the tank almost every time, and that preference is a cost argument rather than an engineering nicety.

It also buys operating room for the plant. When a blower trips over a weekend, holding flow back is the first thing we do while the air comes back. When a plant is new and the nitrifiers have not established, the same tank is what lets us feed it gently for a fortnight. Neither of those moves exists on a plant with no buffer — an operator can carry a plant, but not a plant that gave him nothing to work with.

Where we will not use it

Equalisation buffers a peak. It does not create capacity. If the average daily load is above what the plant can treat, a larger balance tank only moves the failure a few hours later in the day, and by the end of the week it is not moving it at all. When the arithmetic says that, the honest answer is that the plant is undersized or the flow is above what was declared on the permit, and no amount of concrete upstream fixes either.

We also do not let it become a holding tank by default. A balance tank that is kept full because somebody likes the margin is a septic tank with a pump on it. And it is not a substitute for taking strong process flows out of the domestic line in the first place — combining them is a drawing decision that creates work nobody needed, and no balance tank undoes it. Nor does it remove fat: that is a grease interceptor question, upstream, and a balance tank full of floating grease is its own problem.

What travels and what does not

The hour-by-hour method travels. The numbers do not. A building with a genuinely flat pattern across the working day — some light industrial sites are close to it — earns very little from a large balance tank, and there the money is better spent on the nutrient stage. A site that already has an oversized existing sump may effectively have equalisation and not know it, which is worth checking on any retrofit before drawing anything new. On a retrofit you also have the luxury that the flow pattern can simply be measured for a week instead of estimated.

And the profile we assume for a Cebu hotel or a Cebu dressing plant is a Cebu profile. It comes out of the buildings we have worked in and the operating patterns around here, not out of a textbook, and it will be wrong somewhere with a different working day. Our own cone readings and habits sit on the same footing: they have held for us over and over again, which is the strongest thing anybody can honestly say about them. The reason I would still argue for the tank on a building I have never seen is not that our numbers transfer. It is that the cost of being wrong is so asymmetric — a balance tank that turns out larger than it needed to be is a small amount of concrete doing nothing, and one that turns out too small is a machine, an excavation and a shutdown.

If you want a shape for your own volume before talking to anyone, the price estimator and the design generator will both give you one, and it is worth running them once on your average day and once on your peak to see how far apart the two answers sit. Our wastewater page sets out what we do on the plant side, the parameters your industry is held to are listed separately, and the plants we have finished are mostly within a couple of hours of Cordova. If reuse is part of the plan, the treated-water storage that needs is a second volume decision and belongs in the same drawing.

If you have a drawing in front of you with a balance tank on it and you are not sure whether the volume is right, or somebody has suggested you take it out, book a free 15-minute consultation and we will walk through your day with you. For a plant already running and failing on some days but not others, operation and maintenance support is the better starting point, and a design review or an ocular visit can be arranged from there.