I Built This Plant Twice. The Client Only Paid Once.

This is the project I think about most, and it is the one I got wrong.

It was early in the life of our company. A food processing plant — hams, hotdogs, beef jerky — needed a sewage treatment plant built and running. I will not name them. What matters is what happened, and what it taught me, because I have never repeated any of it since.

We walked into a job someone else had already abandoned

Before us, another contractor had taken the project. They collected about ninety percent of the contract value and then stopped coming. No plant, no handover, no explanation. By the time we arrived, the client had paid for something that did not exist and was being asked to trust another contractor with the same promise.

You cannot argue a client out of that feeling. I did not try. I told them we would build it, and I stopped invoicing.

That is not a sales tactic. It was the only honest position available. If the last people took the money and left, then the only thing that separates you from them is that you have not taken money yet.

The first mistake: I read the fats wrong

Meat processing wastewater carries fats, oil and grease. Everybody knows this. What I did not fully respect at the time is that the FOG coming out of a hotdog and jerky line is not the FOG in a restaurant grease trap.

In a kitchen, oil floats. Give it time and a quiet chamber and it will separate on its own. That is why a grease interceptor works there.

In a meat processing line, the oil has been through emulsifiers, hot water, high-pressure washdown, and mechanical shear. Protein and detergent hold it in suspension. It does not float. It stays dispersed in the water as fine droplets, and it will pass straight through a grease interceptor as if the chamber were not there.

I specified a grease interceptor. It was the wrong unit for that water. What was needed was dissolved air flotation, which does not wait for oil to rise but attaches microbubbles to the emulsified droplets and lifts them out. The difference between those two choices is the difference between a plant that works and a plant that chokes.

The second mistake: I let the two waste streams become one

The site had domestic sewage from toilets and canteen, and it had process wastewater from production. I allowed them to be combined into a single line into the plant.

At the time this looked like simplification. One line, one plant, less pipework, less cost. It is the kind of decision that feels efficient on a drawing.

What it actually does is change the legal identity of the discharge. Once process water from a food plant mixes with domestic sewage, the whole combined flow is no longer treated as domestic wastewater by the regulator. The stricter parameters now apply to everything, including the toilet water that would have been easy to treat on its own. I did not just make the load bigger. I made the compliance target harder for every litre passing through.

Since that project, I segregate. High-strength process water gets its own pretreatment before it is allowed anywhere near the biological plant, and where it is possible to keep clean streams clean, we keep them clean. It costs a little more in pipe. It saves an enormous amount in treatment. It is also the decision that decides whether you can ever reuse the water.

The third problem: the plant was on the second floor

The treatment plant sat on the second level of the building. There was no ground area available, so that is where it went.

A plant at ground level that smells is a problem for the people standing near it. A plant on the second floor that smells is a problem for the entire neighbourhood, because you have handed the odour to the wind at elevation.

And it did smell. Undigested fat coats everything. It blankets the surface of the aeration tank, blocks oxygen transfer, smothers the biology underneath, and then the tank turns septic. Septic means sulphide. Sulphide is the rotten-egg smell that people can detect at concentrations far below anything that would show on a routine lab report.

The complaints started with the neighbours. They went to the Barangay. From the Barangay it went to the City Health Office. From the City Health Office it reached the Department of Health.

I want to be clear about what that means for a client. They were not only dealing with a plant that did not work. They were dealing with officials arriving at their production facility, in a food business, where a health finding is not a fine but a shutdown. And the contractor responsible for that situation was me.

What we actually did about it

We stopped billing entirely. Not reduced, not deferred — stopped. We would not invoice again until the plant passed and stayed passing. That decision was made once and never revisited, and it removed every conversation about money from the site for months.

Then we corrected the design.

The flotation stage went in ahead of the biology, so the emulsified fat was removed as fat instead of being fed to bacteria that could not cope with it. And into the existing aeration tank we retrofitted FBBR-style fixed-film media.

That second decision is the one worth explaining. In a conventional activated sludge tank, your entire treatment capacity is the bacteria floating in the water. If those bacteria get shocked — by a fat slug, by a caustic washdown, by a weekend with no flow and then a Monday with all of it — you lose your biomass and you are starting over.

Fixed-film media gives the bacteria a surface to live on. You are holding a much larger population in the same tank volume, and a large part of that population is anchored in a biofilm rather than suspended in the flow. It does not wash out. It survives shocks that would have wiped out a suspended-growth system, and it recovers in days instead of weeks. For a facility with irregular production, holidays, cleaning cycles and variable load, this is not a refinement. It is the reason the plant stays alive. It is also what keeps the slow bacteria that remove ammonia in the tank when everything else is being washed out.

We worked that site at every hour it needed. Nights when the load pattern only showed itself at night. Holidays, because bacteria do not observe holidays and a plant recovering from a shock cannot be left alone for four days. That period was not comfortable and it was not profitable.

The plant stabilised. It passed. It kept passing.

What I took from it

Look at the actual water, not the label on the industry. "Food processing" tells you almost nothing. Emulsified fat and free-floating fat need completely different equipment, and the only way to know which one you have is to look at the process that made it.

Keep streams apart until you have a reason to join them. Mixing is cheap on the drawing and expensive for the next twenty years of operation.

Design for the bad week, not the average one. Averages do not damage plants. Shocks do.

And when the design is wrong, the fastest way through is to say so and fix it at your own cost. I did not win that client with a proposal. I won them by refusing to leave, which is exactly what the previous contractor had not done. I have only ever needed to do that once, because the mistakes behind it were never repeated.

I share this because I meet plant owners regularly who are living inside some version of it — a plant that will not hold, a smell nobody can explain, a contractor who has stopped answering. Usually the problem is not the operator and not the bacteria. It is a decision made on a drawing years earlier, by someone who did not look closely enough at the water. The plants we have built since are here.

If that sounds like your plant, tell me what is happening and I will tell you honestly what I think is causing it. You can book a free 15-minute consultation with us. We are one of the wastewater companies in the Philippines that would rather talk you out of the wrong system than sell you one.