The Blower Stopped for Three Days. The BOD Came Back First.

The call came on a Monday morning. The blower had tripped on Friday afternoon, nobody was on site over the weekend, and by the time the operator walked in on Monday the aeration tank had been sitting still for close to three days. The question on the phone was the one we always get in that situation: is the plant dead, and do we have to start it again from zero?

The answer, most of the time, is no. But the recovery does not come back in one piece. It comes back in two, and the second piece takes far longer than the first. That is the part that catches people out, so it is worth writing down what we actually saw and what we did about it.

What we found when we opened the tank

The liquor had gone grey and flat instead of chocolate brown. There was a smell — not the earthy smell of a working tank, but the sulphur smell that tells you the tank went anaerobic somewhere in those three days. We took a litre and put it in the cone. The sludge was still there. It settled, and it settled to something inside the range we normally work in, roughly 50 to 200 mL/L, which is the number we watch on a day-to-day basis instead of sending samples away for MLSS. We read a cone rather than track MLSS for exactly this reason: when you need to know something on a Monday morning, a lab result that arrives in seven to ten days is not a decision-making tool.

So the biomass had not washed out. It had been starved of oxygen, not removed. That distinction decided everything that followed.

The two populations do not recover at the same speed

There are two workforces in an aeration tank. The one that eats the organic load — the BOD — is fast. Under the conditions in a working plant those bacteria double in a matter of hours. The one that oxidises ammonia is slow; nitrifiers take on the order of a day or two to double, and they slow down further as the temperature drops. Standard wastewater engineering texts (Metcalf & Eddy) give the growth rates behind that, and the practical consequence is simple arithmetic: a population that doubles in hours climbs back inside a week, and a population that doubles in a day or two needs weeks to rebuild the same numbers.

That is what happened here. The colour came back first, then the settling, then the BOD. The ammonia number sat there and refused to move for much longer than the client expected. Nothing was wrong. We were simply watching the slower workforce do what it does. We have written before about why ammonia is the parameter that fails first in the results that reach us, and a blower outage is one of the cleanest demonstrations of it.

What we did, in order

First, air. Before any dosing decision, before any sampling, get the blower running or get a temporary one in. Nothing else you do matters while the tank is still anaerobic.

Second, we stopped wasting sludge. When the population you need is depleted, removing more of it is the one thing that guarantees a longer recovery. Wasting resumes when the cone reading climbs back past about 200 mL/L, not before.

Third, we held the flow back. Whatever equalisation the plant had, we used it, and we asked the building to spread its washdown rather than dump it. A depleted plant can handle its normal load spread evenly and cannot handle it in a two-hour peak. Where the plant has no equalisation volume, this is where its absence gets expensive — the same weakness we described in the post on what you declared on the permit and what the plant actually receives.

Fourth, we left the chemicals alone. There is a strong pull, in that first week, to buy something — a bacterial dose, a chemical, anything that feels like action. We sell probiotics ourselves and we still tell clients when not to buy them. A tank whose biomass is intact and merely starved does not need to be re-seeded; it needs air and time. The one thing we do watch is alkalinity, because nitrification is acid-forming and consumes it as the ammonia oxidation restarts. If the pH starts sliding as the nitrifiers come back, that is a real intervention, and lime or soda ash handles it. We would rather correct alkalinity than let a recovery stall on pH — and we design so that compliance never depends on a chemical being available locally that week.

Losing them and failing to grow them are different problems

The three-day outage above was a starvation event. The biomass stayed in the tank. A different plant taught us the harder version of the same lesson: a small system where the transfer pump kept burning out on low water, and every time it failed the tank overflowed and the nitrifiers left with the water. That plant was not failing to grow nitrifiers. It was losing them, over and over, and no amount of aeration was going to fix an overload relay that was not there.

So when a plant comes back from an air outage and the ammonia still will not settle after several weeks, the question we ask is not "how do we grow more" but "where are they going". Usually the answer is physical — an overflow, a carry-over, a clarifier handing solids forward. That is a different post: the clarifier is usually not the problem, it is where the problem shows up.

Where fixed-film media changed the arithmetic for us

On plants where we have installed FBBR-style fixed-film media — polypropylene rope bundles run the full depth of the tank, spaced so dissolved oxygen can move between them — the recovery after an outage has been noticeably less dramatic in our experience. Our reading of it is that a nitrifier attached to media is not carried out with an overflow the way a suspended one is, so the plant starts its recovery with a base population instead of from near zero. We hold that as our own experience rather than as a proven mechanism; we have not run the controlled comparison that would let us claim it as fact.

It is also why we default to media on most designs. Most operators cannot culture a suspended activated sludge properly, and a plant that forgives an imperfect week is worth more than a plant that only performs when everything goes right. That is not an excuse for poor operation. It is a design choice made with real operators in mind — and it connects to something we see repeatedly, which is that turnover in the operator's chair shows up as a plant problem.

The design question sitting underneath all of this

A three-day outage is not really a biology story. It is a story about a plant with one blower and nobody on site at the weekend.

We put a standby blower and an automatic changeover into designs where the numbers justify it, and we say plainly what it buys. The duty-standby arrangement costs what it costs, once. Against that, an outage of a few days buys you several weeks of ammonia that will not pass, and if a sampling date falls inside those weeks you are looking at a failed result, a re-test at your own cost, and a conversation with the regulator that you did not need to have. That is arithmetic, not alarm. Sometimes the arithmetic says the standby is not worth it — a small plant with an operator on site daily and a spare unit available in town can reasonably carry the risk. We say that too, when it is true. Every component we put in a quotation has to be justified against a financial benefit, and one that cannot be justified should not be there.

The other half of the answer is duller. That blower did not fail at random. Air filters that nobody cleans, a check valve nobody looks at, a belt nobody tensions — that is where most blower failures we attend actually start, and it is the kind of thing a maintenance routine exists to catch. If you are sizing or budgeting a plant and want to see what the compliance side actually demands of it, the DENR discharge standards page lays out DAO 2016-08 and DAO 2021-19 by water body class, and the STP price estimator will give you a range for a system of your size before you talk to anybody.

What transfers, and what does not

The order of recovery transfers: BOD returns before ammonia, always, because of the growth rates involved. The instinct to stop wasting sludge and hold the flow back transfers too.

The numbers do not. Our own plants sit at very different dissolved oxygen levels — we have left a large plant running at 1 mg/L because the liquor was brown, the sludge settled and there was no septic smell, while on other plants anything below about 4 goes anaerobic and starts to smell. A recovery time from one site tells you nothing reliable about another. Temperature, load, how much media is in the tank, how long the outage ran, and whether the biomass stayed in the tank or left with an overflow all move the answer.

So take the sequence, not the schedule. Look at your own plant, in a cone, on the day. And if you have had an outage and the ammonia is not coming back, the useful thing is to find out where the bacteria are going before buying anything at all. You can see the kind of work this comes from on our completed projects page, or read more about how we approach wastewater treatment generally.

If you would like to talk one through, book a free 15-minute consultation and we will look at it with you.