RAS design tricks in water disinfection

I have been working for the past few months on the design of a number of laboratory facilities for the upcoming IRCAO building at the University of Liège. Among these, there will be facilities for disease and vaccine testing involving some important aquaculture viruses. That means the building needs to meet stringent biosecurity requirements, including for the treatment of water leaving the laboratories.

The team at the university, led by Dr. Alain Vanderplasschen and Dr. Owen Donohoe, had a very good idea of what they needed from such a system. We needed to deliver several times the required inactivation dose for the most resistant virus we were targeting, IPN, in a single pass. We needed to be able to treat effluents in batches. And all of this had to fit into a small system that could pass through normal doors, was simple to operate and maintain, and didn't break the bank.

The system also had to be automatic, but in a very particular way. No programming. No fiddling with electrical wires or electronics. If something breaks, someone with a limited technical background should be able to understand what is happening and repair it.

The first challenge was deciding on the disinfection technology: in aquaculture, this essentially comes down to UV irradiation or a powerful oxidation process, most commonly ozone. There are other possibilities, but these are the two obvious candidates for this application.

Ozone is a very powerful oxidant and disinfectant, but that is also part of the problem. It reacts with everything in the water, not just the pathogen we are trying to remove. A proper ozone installation also brings additional safety considerations, requires dedicated equipment and, depending on the ozone generator, a supply of pure oxygen and cooling.

Then there is the question of controlling the actual ozone dose. If the system is going to operate automatically and the treatment performance needs to be demonstrated, controlling and monitoring the process becomes considerably more complicated.

So, UV it was.

The basic principle is simple enough: the required UV dose depends on the UV output and how the water moves through the reactor. But implementing that principle in this particular application was not so straightforward. The water arrives to be treated in batches. That means the UV system would need to be switched on and off for every treatment cycle.

If you keep the UV lamps running without water circulating, the reactor can overheat. If you switch them off, you have to wait for them to power up again before treatment can begin. Frequent cycling also has implications for lamp life.

When you need to treat batches of water every day, this quickly becomes an undesirable way of operating the system. So, instead of switching the UV on and off, we keep the water moving. Here is where the RAS box of tricks starts to be useful.

In RAS, water treatment processes are almost never 100% efficient. If you need to remove enough of something from the water — let's say CO₂ — one of the simplest ways of doing it is to pass the water through the treatment device repeatedly. There is another useful trick: prevent the treated water from immediately mixing back with the untreated water.

This allows a treatment process to operate as a side-stream loop, continuously treating water at a relatively high flow rate while gradually accumulating treated water somewhere else, ready to be discharged or returned to the system. I used these principles to design the IRCAO system.

The infected water first enters a collection sump (Tank 1), which also acts as a settler. From the clarified end of the sump, a pump draws water and sends it through fine solids filtration and then through the UV reactors. The UV-treated water is discharged into a baffled buffer tank (Tank 2).

The buffer tank is connected to the pump that eventually discharges the treated effluent from the building. The geometry of the tank is designed so that the discharge pump inlet is as far away as possible from the point where treated water enters the tank. Internal baffles reduce the possibility of short-circuiting.

There is also another outlet located close to the UV discharge point inside the buffer tank. Most of the water leaves the tank through this outlet, back to sump, for another UV irradiation pass. The system receives roughly 0.5–1 m³ of polluted water per day. With a relatively modest 2 m³/h circulation pump, the UV system can process that volume approximately 40–90 times over the same period.

The rest is essentially a matter of engineering the UV system to deliver the required inactivation dose with a healthy safety margin on a single pass, and adding enough redundancy that a single component failure doesn't compromise the whole process.

The final system therefore has very little automation. Level controllers monitor the process and trigger the discharge of treated water from the buffer tank when the collection sump reaches its maximum level. No PLC is required to make the basic process work. For long term monitoring, UV sensors on the reactors can signal a system shutdown if the UV irradiation doses are not being met.

The system is being built as two skids that can fit through normal doors and are operated at approximately chest height. At the moment, everything is being fabricated, and we expect to have a functioning system in a few weeks.

There were, of course, another myriad of design challenges around hydraulics, failsafes, component selection, maintenance and making sure that the system behaves predictably under abnormal conditions. Those took a good part of the design effort, but I won't bore you with those for now.

If you have an idea, a problem or a system that needs to become a real piece of equipment, don't hesitate to get in touch. I like co-designing these things and getting them to fruition.

ChimanaTech

Chimana Management BV

Hemelrijk 2A

5281PS

Boxtel

The Netherlands

carlos@chimana.tech

+31 612 769 754

© 2025. All rights reserved.