Paul Industries designs cleaning, descaling and turnaround systems for Wyoming mineral and chemical plants. A continuous soda ash operation cleans on a completely different logic from a food or pharmaceutical plant. It does not clean to a hygiene standard between batches; it cleans because heat transfer has decayed to the point where the economics no longer work, and the whole question is when to stop and how long it takes.
Request a quote or call 201-450-8280
The cleaning decision is an economic calculation, and most plants do not have the data for it
Scale on a heat transfer surface reduces its duty. The plant compensates, by raising steam pressure, increasing temperature difference, or accepting reduced throughput, and each of those has a cost that accrues continuously. Against that sits the cost of stopping: lost production, the cleaning itself, and the risk that comes with any shutdown and restart.
The optimum is the point where the accumulating loss overtakes the cost of the intervention, and it is a genuinely calculable number. Most plants do not calculate it. They clean on a calendar, or when something trips, or when a shutdown happens for another reason and the opportunity is taken.
Making it calculable needs three things, none of them exotic.
Measure the fouling, not the symptom. Track the actual heat transfer coefficient over time rather than the steam pressure the operators have crept up to. Inlet and outlet temperatures on both sides plus flow gives the duty, and duty against clean-condition duty gives the fouling curve. That curve is what turns cleaning from a habit into a decision.
Know what the loss costs. Extra fuel, extra electricity, and lost production if the unit is a bottleneck. That last one dominates when it applies and is often not attributed to fouling at all.
Know what the clean costs and how long it really takes. Including cool-down, isolation, the clean itself, inspection, and return to stable operation, which on a crystallizing plant can be a substantial tail.
With those three, the cleaning interval stops being a matter of opinion. Plants frequently find they have been cleaning either considerably too late, having absorbed months of degraded performance, or somewhat too early, having given up production that was still profitable.
Cleaning without damaging what you clean
| Method | Suits | Cost to the metal |
|---|---|---|
| Hot water or condensate wash | Soluble scale, caught early | None; always try this first |
| High-pressure water jetting | Adherent deposit, accessible surfaces | Low, if pressure is controlled |
| Mechanical cleaning | Tubes and defined geometries | Low to moderate; can score surfaces |
| Chemical descaling | Inaccessible or heavily fouled surfaces | Real; attacks the substrate too |
| Doing nothing longer | Nothing | Caustic concentrates under the deposit |
Two rows deserve expansion.
Chemical descaling damages the substrate. Conditions aggressive enough to dissolve tenacious scale act on the metal as well, and a surface that has been repeatedly descaled is rougher than it started. A rougher surface nucleates deposit more readily, so the next interval is shorter, which means more frequent descaling and further degradation. That feedback loop is real, it takes years to become obvious, and it presents as equipment that fouls faster than it used to, which reads as aging.
The lever that costs the metal nothing is circulation velocity. More flow and more turbulence during a clean contributes genuine removal capability at no cost to the substrate, and it is consistently the last thing plants reach for because concentration and temperature are the knobs on the dosing skid.
Tolerating a deposit is not neutral. As covered on our Wyoming caustic cracking page, liquid trapped under scale on a heated surface concentrates well beyond the bulk, and caustic gouging attacks the metal underneath. A deposit that is only costing heat transfer today may be quietly costing wall thickness as well, which is why a deposit found on a heated caustic surface should put the metal beneath it on the inspection list rather than simply being cleaned off and forgotten.
| Heating load | Per clean | Per 12 cleans |
|---|---|---|
| 60 kW | $57.31 | $688 |
| 120 kW | $114.62 | $1,375 |
| 240 kW | $229.25 | $2,751 |
The energy in the cleaning itself is trivial at Wyoming’s 7.96 cents per kilowatt-hour, just under the 8.13 cent national average (EIA, 2024). The cost that matters is the production the clean occupies, which is why shortening the cleaning window is worth far more than making the clean cheaper, and why cleaning access designed into the equipment repays itself repeatedly.
Designing so the clean is short
Most of the duration of a turnaround clean is not cleaning. It is getting to the surface, and getting back.
Isolation that allows a unit out without taking the plant down. Where the process permits parallel trains or bypass, a clean becomes a maintenance activity rather than an outage. This is a capital decision made at design and it is the single largest determinant of how often a plant is willing to clean.
Access that does not require dismantling. Removable heads, inspection ports, cleaning connections plumbed in rather than improvised, and enough space around the equipment to actually work.
Drainage and fill provision. A vessel that drains completely and refills quickly saves hours at both ends of every clean, and low points that hold liquor extend both the clean and the subsequent inspection.
Somewhere for the spent solution to go. Cleaning produces a waste stream with its own handling and disposal requirement, and in a water-constrained basin with limited disposal routes that is a design question rather than a detail. A clean that cannot be drained anywhere is a clean that does not happen on schedule.
We design and install cleaning circuits and connections, isolation and bypass arrangements, spent solution handling, access modifications on existing equipment, and the instrumentation that makes fouling measurable rather than inferred. Piping follows ASME B31.3 with the fluid service category documented, and on the product-contact side we build to ASME BPE with orbital welding to AWS D18.1 and passivation to ASTM A967 after cleaning per ASTM A380.
Standards referenced: EIA electricity price data · ASME BPE · ASTM A967 · ASTM A380 · ASME B31.3
Frequently asked questions
Do you design cleaning and turnaround systems for Wyoming plants?
Yes, across Green River, Rock Springs, Casper, Cheyenne and statewide: cleaning circuits and connections, isolation and bypass arrangements, spent solution handling, access modifications on existing equipment, and fouling instrumentation.
How should we decide when to clean?
By calculating where the accumulating loss from fouling overtakes the cost of stopping. That needs the actual heat transfer coefficient tracked over time, the cost of the loss including lost production if the unit is a bottleneck, and the true duration of a clean including cool-down and return to stable operation.
What do most plants do instead?
Clean on a calendar, when something trips, or when a shutdown for another reason creates an opportunity. Plants that start measuring usually find they have been cleaning considerably too late, having absorbed months of degraded performance, or somewhat too early, giving up production that was still profitable.
Why track heat transfer rather than steam pressure?
Because steam pressure is the compensation, not the measurement. Operators creep it up as duty decays, which masks the fouling. Inlet and outlet temperatures on both sides plus flow gives the actual duty, and duty against clean-condition duty gives the curve you can make a decision from.
Does chemical descaling damage equipment?
Yes. Conditions aggressive enough to dissolve tenacious scale act on the metal too, and a repeatedly descaled surface is rougher than it started. A rougher surface nucleates deposit more readily, so intervals shorten, which means more descaling and further degradation. It reads as aging and it is not.
What is the safest way to improve a clean?
Circulation velocity. More flow and turbulence contributes genuine removal at no cost to the substrate, and it is consistently the last lever plants reach for because concentration and temperature are the knobs on the dosing skid.
Is leaving a deposit in place harmless?
No. Liquid trapped under scale on a heated surface concentrates well beyond the bulk, and caustic gouging attacks the metal underneath. A deposit costing heat transfer today may be costing wall thickness too, so a deposit on a heated caustic surface should put that metal on the inspection list.
Where does the time in a clean actually go?
Not into cleaning. Into getting to the surface and getting back: isolation, cool-down, dismantling, drainage, refill and return to stable operation. That is why access and isolation design determine cleaning duration far more than the chemistry does.
What about the spent cleaning solution?
It is a waste stream with its own handling and disposal requirement, and in a water-constrained basin with limited disposal routes that is a design question rather than a detail. A clean with nowhere to drain to is a clean that slips off schedule.
How do I get a quote for a Wyoming cleaning project?
Use the form on this page or call 201-450-8280. Useful inputs are which equipment fouls and how fast, current cleaning frequency and duration, whether fouling is measured or inferred, what isolation exists, and where spent solution currently goes.
How is heat exchanger fouling distinguished from other performance losses?
By comparing the overall heat transfer coefficient, corrected for flow and temperature, with its clean baseline, so that a drop from fouling is separated from drops caused by low steam pressure, air in the shell or a flow restriction. Without the correction, a plant cleans exchangers that were never dirty.
What data does a heat transfer trend need from the plant's historian?
Inlet and outlet temperatures on both sides of the exchanger, the flow on at least one side, and the steam pressure or heating medium temperature, all logged at the same interval so that a corrected coefficient can be calculated. Most plants already log these; the calculation is what is missing.
What chemistry removes soda ash and trona scale?
Acid cleaning for carbonate and bicarbonate scale, with inhibitors to protect the steel, and mechanical or hydroblasting methods where deposits are thick or hard. The chemistry is chosen against the deposit analysis, and the equipment metallurgy sets the limits on acid strength and temperature.
Can descaling be done in place?
Often, by circulating cleaning solution through the equipment with temperature control and monitoring of the reaction. In-place descaling avoids opening the equipment and shortens the outage, provided the circuit can be isolated and the spent solution handled.
Which parts of an evaporator are most at risk during a chemical clean?
Gasket and seal materials, tube-to-tubesheet joints, instrument connections and any carbon steel component in an otherwise stainless system, because the acid attacks them faster than it dissolves the scale. The clean is designed around those weak points with inhibitors, temperature limits and time limits.
Can spent descaling solution be neutralised on site?
Yes, by caustic addition in a tank sized for the volume, with the resulting salt solution and dissolved metals checked against the plant's discharge permit before release or sent to the plant's own effluent system. Soda ash plants generally have the caustic and the tankage; the check against the permit is the step that is skipped.
How does design reduce scaling in the first place?
Through velocity that keeps solids in suspension, temperature control that avoids local overheating, surface finishes that resist adhesion, and materials chosen for the service. Scale is never eliminated in this process, but the rate can be reduced.
What is the safest way to shorten a cleaning outage?
Prepare everything before stopping: circuits made up, chemicals on site, permits ready, and the sequence rehearsed. The outage itself is then cleaning and inspection, not setup. Prefabricated cleaning connections on the equipment make this possible.
How does restart differ from normal operation?
Equipment is cold, empty and often air-filled, and the process has to be brought to steady state without thermal shock, without scaling the freshly cleaned surfaces, and without process upsets. A restart procedure with defined ramps is part of the cleaning plan.
Does cleaning affect the crystallizer's product quality?
Yes. A freshly cleaned crystallizer operates differently until it reaches steady state, and the crystal size distribution can be off-specification for a period. Planning for that transition avoids treating it as a failure.
How is confined space entry managed during cleaning?
Entry is a permit activity under OSHA rules, and cleaning designed to avoid entry is safer and faster. Where entry is unavoidable for inspection, atmosphere testing, ventilation and rescue provisions are part of the plan.
What about cleaning the brine and process water side?
Brine systems scale and corrode, and cooling water systems foul. They are cleaned on their own schedule with chemistry matched to the deposits, and their condition affects the main process heat transfer as much as the product side does.
Can heat transfer monitoring be automated?
Yes. A calculation running continuously on the plant's control system gives the operators a live heat transfer coefficient and alarms when it crosses the threshold that the economic model set. It is inexpensive relative to the value of cleaning at the right time.
Does the same approach apply to bicarbonate and other Wyoming mineral products?
Yes. Crystallizers, evaporators, dryers and heat exchangers in any continuous mineral process foul and are cleaned on the same economic logic, with chemistry matched to the specific deposit.
What is the commonest mistake in Wyoming mineral plant cleaning?
Cleaning on a fixed calendar interval that was set years ago and has no relationship to the current fouling rate. Either the plant cleans too often and loses production, or too rarely and runs inefficiently for months. Measuring fixes both.
The restart is the expensive half, and nobody budgets for it
When a plant estimates the cost of a clean it usually counts the clean. The larger number is often on the other side: getting a crystallizing plant back to saleable product after it has been stopped, drained and refilled.
A clean strips the system of exactly the thing a crystallizer needs to work. Crystal population is gone, so there is no surface for dissolved material to grow on. The vessel walls are clean and, for a while, unusually smooth. Temperatures are wrong throughout. And the first liquor in is not at steady-state composition.
What follows is predictable. Supersaturation builds because there is nothing to consume it. When it finally breaks, it breaks by spontaneous nucleation rather than by growth, and the plant produces a large population of fine crystals. Those fines then dominate the size distribution for a long time afterwards, because the same dissolved material is now spread across far more particles. Product that is off-specification for size goes somewhere, whether that is rework, blending, or a lower-grade sale.
Three things shorten that tail and they are all worth designing for rather than improvising.
Seed deliberately. Starting with a controlled population of good crystals gives the supersaturation somewhere to go, so growth wins the race against nucleation from the first hour rather than after the system has already nucleated once. A plant with a reliable seed source and a defined seeding procedure restarts far better than one that waits for the crystallizer to seed itself, and the difference shows up directly in how quickly product comes back on specification. That means somewhere to store seed, a way to introduce it cleanly, and a specification for the seed itself.
Bring temperature up before composition. Reaching thermal steady state on clean liquor, then moving composition toward operating conditions, avoids the supersaturation spike that comes from doing both at once. It takes longer on paper and it is usually faster to saleable product.
Do not chase rate. The instinct after a shutdown is to recover lost tonnes, and pushing feed rate raises supersaturation precisely when there is least crystal surface available to absorb it. That produces more fines, which extends the off-specification period, which loses more tonnes than the push gained. Restarting gently is faster in output terms and it is the opposite of what shutdown pressure encourages.
There is also a subtlety about the clean surfaces themselves. A freshly descaled vessel is smoother than a seasoned one and initially offers fewer nucleation sites, which sounds helpful and is not. It means supersaturation can climb higher before anything happens, and when it does happen it happens abruptly. Plants sometimes observe that the first days after a particularly thorough clean produce worse product than the days after a lighter one, and that is the reason.
The practical conclusion is that the cleaning decision described above should count the restart. A plant that cleans a week earlier than optimal and then spends three days off-specification has not saved anything. We would rather help a client measure the restart tail than quote a faster clean, because on a crystallizing plant the tail is frequently the larger number and it is almost never on the work order.
We design seeding and seed storage arrangements, restart-oriented instrumentation, and the isolation and fill provisions that let a unit come back on temperature before it comes back on composition.
Cleaning too often, or not often enough, at a Wyoming plant?
Tell us whether you measure fouling or infer it from steam pressure. If it is the second, that is where to start. Call 201-450-8280 or use the form below.