Bill, Tom and Dog — The Missing Tank

Bill had bought a new pump.

It was, according to Bill, an exceptionally good pump.

“It’ll move 600 litres a minute,” he told Tom, standing beside it with the satisfaction of a man who had recently acquired considerably more pump than circumstances required.

Tom looked along the poly pipe running towards the cattle yards.

“And what’s it feeding?”

“The troughs.”

Tom looked at the troughs. Each had a little float valve in it.

“How much water do they use?”

Bill shrugged. “Depends how thirsty the cattle are.”

Dog sat down.

Bill started the pump.

For a few seconds everything looked magnificent. Water roared up from the dam, the pipe went hard and Bill smiled.

Then the trough float valve shut.

The pump did not care for this development.

The pressure climbed. The pipe twitched. Somewhere near the shed a fitting developed an expression of deep concern.

Bill shut the pump down.

“Float valve must be faulty.”

Tom inspected it.

“No. Float valve’s working perfectly.”

They started the pump again.

The trough filled.

The float valve shut.

The pump once again attempted to send 600 litres a minute through a hole that was now approximately no litres a minute in size.

Bill shut it down.

“Pump problem.”

Tom shook his head. “Pump’s working perfectly too.”

Bill looked at the pump.

Then at the trough.

Then back at the pump.

“Well they can’t both be working perfectly.”

Dog wandered over to the old header tank beside the shed and sat underneath it.

Tom pointed.

“There’s your problem.”

“The tank?”

“The missing tank.”

Bill frowned.

Tom explained that the pump wanted to run for a decent period and shift a lot of water. The cattle wanted small amounts of water whenever they happened to drink. There was nothing wrong with either end of the system. The problem was expecting one to follow the other directly.

“So the pump fills the tank,” Bill said slowly.

“And the tank feeds the troughs.”

“And if the cattle stop drinking?”

“The pump can still finish its run.”

“And if the pump stops?”

“The cattle still have water.”

Bill looked at Dog.

Dog looked at Bill.

It was difficult to tell whether Dog understood process engineering, but he had been sitting beside the answer for the last ten minutes.

Bill eventually nodded.

“So what you're saying is that I bought two perfectly good bits of equipment and built a stupid system.”

Tom considered this.

“Yes.”

Dog lay down.

There was really nothing further to add.

The Missing Buffer

There is a basic engineering question buried in the Glan Devon proposal that has received surprisingly little attention: what happens when the Xetrov and the biosolids dryer do not want to operate at the same rate?

The proposed process couples two quite different pieces of equipment. The Xetrov burns non-recyclable waste to provide heat. The amended May 2026 material identifies the dryer as an SMS thin-film dryer, transferring that heat into wet biosolids through a hot-oil circuit. The same design also appears to return dryer off-gas to the Xetrov for thermal treatment.

That creates a fairly tight process loop. The dryer depends on the Xetrov for heat, while the Xetrov may depend on the operating dryer as a source of gas requiring treatment. Neither piece of equipment, however, has any reason to behave exactly like the other.

Thin-film dryers hold relatively little material and can respond quickly to changing feed conditions. The Xetrov also appears to be a relatively low-inventory thermal system, with prepared fuel screw-fed into a high-temperature vortex rather than sitting in a large combustion bed. That potentially gives both systems relatively rapid response characteristics. Without more engineering detail and equipment specifications I cannot be sure. So, let's just assume this is how these bits of gear behave.

The interesting question is what happens when you combine the two, and their responses are different.

Something has to absorb the difference

A biosolids sludge dryer does not consume exactly the same amount of heat every minute of every day. Feed moisture varies. Feed rate varies. Downstream equipment can constrain throughput. The dryer can slow, trip or stop altogether.

The Xetrov has a different set of constraints. Fuel feed, combustion air, temperature, residence time and emissions all have to remain within an acceptable operating envelope. If the dryer suddenly reduces its heat demand, stopping the NRW feeder does not instantly remove the energy already in the system. Fuel already inside the unit continues burning. The refractory remains hot. The heat exchanger remains hot. The thermal oil continues carrying that energy around the circuit.

So where does the heat go?

That is the missing buffer problem.

Any coupled process needs some means of absorbing short-term differences between one part of the plant and another. In mining we solve the same problem all the time. A stockpile separates the rhythm of the mine from the rhythm of the mill. Surge bins, and intermediate tanks exist because real plants do not operate as a perfectly synchronised chain.

The Glan Devon equivalent could involve several things: thermal-oil inventory, Xetrov turndown, controlled fuel feed, biosolids storage, NRW storage and blending, dried-product surge capacity, bypass arrangements, cooling capacity or some form of heat rejection. The control system may coordinate all of those, but the control system cannot create buffer capacity that does not physically exist.

At megawatt-scale heat flows this matters. Thermal oil can absorb excess energy for a period, but it is not an infinite heat sink. Once the available temperature margin has been consumed, something else has to happen. Either heat production falls quickly enough, heat demand rises, heat is rejected elsewhere, or the system moves towards a trip.

The same problem exists in the opposite direction. If the Xetrov output drops while wet biosolids continue arriving, the dryer cannot simply manufacture the missing heat. Material has to wait somewhere, throughput has to fall, or another heat source has to take over.

The engineering question is therefore not merely whether the Xetrov can heat the dryer. It is how the plant behaves when heat production, heat demand and material flow temporarily stop matching.

The buffer is physical

This becomes important to the development application because process buffering is not an abstract control-system issue. It occupies space, contains material, consumes power and changes the way the site operates. It may increase the on-site inventory or biosolids and non-recyclable waste.

Suppose commissioning shows that reliable operation requires a larger wet-biosolids inventory so the dryer can be fed at a steadier rate. That means more biosolids stored on site for longer periods, with more handling, mixing and potentially more odorous material inside the building.

Suppose instead that the Xetrov needs a more consistent fuel supply. The answer may be more NRW storage, better segregation or blending of fuel before it reaches the feeder. That means a larger combustible inventory.

A dried-product buffer creates another inventory. A larger thermal buffer means more hot oil or additional equipment. A heat-rejection system means cooling equipment, pipework, pumps and another operating state that must be considered in noise, heat and possibly water assessments.

None of these things is simply a change to a software setting.

They potentially alter the environmental design of the facility.

Now look at the negative-pressure building

The proposal relies heavily on enclosing odorous activities within a building maintained under negative pressure. That sounds straightforward until the actual operating inventory and operating pattern are considered.

If more wet biosolids have to be stored to decouple receival from drying, where are they stored? For how long? How are they turned over or mixed? Are doors opened more frequently? Are additional bins, bunkers or transfer points required? Does the required extraction rate change?

Negative pressure is not a magic property of a building. It depends on fans moving enough air to overcome leakage and deliberate openings while capturing contaminated air from the places where it is generated. The treatment system then has to handle both that airflow and its contaminant load.

The ventilation requirement therefore depends partly on how the plant is actually operated.

If the real operating plant requires materially greater buffering than the application assumes, then the odour-control duty may also be different from the one currently being assessed. A larger biosolids inventory may mean a larger or differently distributed odour source. More handling may create additional release points. Different storage arrangements can change the air volumes that need to be captured.

That relationship between process design and odour control is important because the buffering question cannot be separated neatly from the environmental assessment.

Then there is fire

The same applies to NRW.

If reliable Xetrov operation requires a larger or better blended fuel inventory, the site contains more combustible material. Storage geometry, pile size, separation distances, detection systems, firefighting access and water requirements then become part of the operating design.

A larger buffer may make the process easier to control while increasing the consequence of a fire.

The same principle applies elsewhere. Wet biosolids, dried pellets, residues, thermal oil and process liquids all have to remain somewhere when equipment stops. A plant designed around smooth continuous flow can look very different once the credible operating states include eight hours, a day or several days without one of its major process units.

That is why storage capacity should not be treated simply as a maximum tonnage number on a drawing. It is part of the plant's response to abnormal operation.

And what happens when the power goes off?

This is where the coupling becomes particularly interesting.

A power failure does not necessarily make a hot process safe. It may remove the very equipment required to bring that process safely to rest.

After a trip, the NRW feeder may need to stop immediately, but hot-oil circulation may need to continue. Residual material may still be burning. Fans may still be required to move combustion gases through the system. Air-pollution-control equipment may need to remain operational. The building extraction system may still be required to maintain negative pressure. Cooling equipment may need to remove residual heat.

All of those systems use power.

For a rural South Burnett site exposed to regular storm activity and power interruptions, that raises an obvious set of design questions. Which loads are essential during shutdown? Which controls and actuators are UPS-backed? Do the hot-oil pumps have emergency power? What happens to the combustion air and induced-draft systems? Can negative pressure be maintained? Is there standby generation? How quickly does it start? How long can it run? What happens if it does not start?

Most importantly: what is the safe state of the plant if mains power disappears and the backup system also fails?

These are not obscure questions. They follow directly from putting a combustion system, a thermal-oil circuit, a biosolids dryer, air-pollution-control equipment and an enclosed negative-pressure building together on the same site.

Yet the application provides remarkably little visibility of how that shutdown architecture is supposed to work.

Stage 1 does not answer the approval question

The proposal repeatedly places considerable weight on Stage 1 operation to optimise or validate the process. Some optimisation during commissioning is inevitable. The important distinction is between tuning a plant whose fundamental design has already been established and using commissioning to discover what the plant needs in order to work.

There is a large difference between adjusting controller settings and discovering that the plant requires another 500 tonnes of storage.

There is a difference between tuning Xetrov fuel feed and discovering that the fuel has to be stockpiled and blended in a way not presently shown.

There is a difference between adjusting thermal-oil temperatures and discovering that the circuit requires a substantial heat-rejection system.

There is a difference between balancing ventilation dampers and discovering that the real storage and handling configuration requires larger extraction fans or additional odour treatment.

And there is a difference between testing emergency procedures and discovering that critical pumps, fans or pollution-control systems require emergency generation that was never part of the assessed design.

Those are not commissioning adjustments. They are design consequences.

They also flow into matters that Council and SARA are supposed to assess: odour, air emissions, noise, fire, water management, storage quantities, traffic, abnormal operation and potentially the physical layout of the facility itself.

That is why the lack of visible buffering and shutdown detail matters now rather than later.

What plant is actually being assessed?

The central question is not whether competent engineers could eventually find a way to make a Xetrov and an SMS thin-film dryer operate together. Presumably they could design some arrangement to do it.

The relevant question is much narrower and much more important:

What arrangement is actually required?

How much wet biosolids storage does stable operation require? How much NRW must be held and blended? How much thermal capacity sits between heat production and heat consumption? Where does excess heat go? What happens when the dryer trips? What happens when the Xetrov trips? What continues running during a power failure? What inventories accumulate during a prolonged outage? What ventilation, fire protection and emergency power are required for those operating states?

Those answers define the plant.

If the answers are already known, they should be visible in the design being assessed. If they are not yet known because Stage 1 is intended to establish them, then there is a more fundamental problem: the development application may be seeking approval before the operating configuration needed to make the process work has actually been determined.

That is the issue hidden inside the missing buffer.

Commissioning should show that the approved plant works. It should not be the stage at which the proponents discover what plant they actually need.

© 2026. All rights reserved.