Cold loops that sanitize themselves
An ozonated purified water loop carries 0.02 to 0.2 ppm of dissolved ozone from the tank through the piping, and a UV destruct unit at 254 nm strips it out ahead of each use point. Every UV lamp puts most of its input into the water as heat: a 5 kW destruct unit on a 100 gpm loop raises the water about 0.2 C per pass, and with the pump added the loop climbs a degree or two an hour until something removes it.
That something is a loop cooler, and it lives in a harder place than a plain trim cooler because the water on one side of it is an oxidizer.
The cooler duty is everything that warms the loop. UV lamps convert 85 to 90 percent of their input to heat: a 3 kW TOC unit and two 5 kW destruct units add about 40,000 Btu/h. A 7.5 hp pump adds 19,000 Btu/h. A 3,000 gallon tank and 500 ft of pipe in a 24 C plant pick up another 10,000 to 20,000 Btu/h when the water is held at 20 C, because now the insulation works against you. Call it 85,000 to 95,000 Btu/h with 25 percent margin for fouling and summer chilled water: about a 1 C rise per pass on a 100 gpm loop, which a small double tubesheet unit removes on 7 C chilled water.
Dissolved ozone at loop concentration does nothing to 316L. The damage is to organics: nitrile gaskets crack in weeks, EPDM lasts months and sheds particles as it goes, silicone the same. Every clamp gasket, valve diaphragm and pump seal in the ozonated section is PTFE, PVDF or an ozone-rated fluoroelastomer, and the exchanger ships with those fitted, not the EPDM set that ships by default. The shell side is chilled water and needs nothing special. The one thing ozone asks of the exchanger body is full drainability, because standing ozonated water off-gasses into an ozone-rich pocket that finishes any gasket near it.
The destruct units sit just ahead of the use points; the cooler goes upstream of them, in the ozonated section, on the return leg or just after the pump. There it removes the heat from every lamp on the loop and it sees ozone, which is why its materials matter. Placed downstream of the last destruct unit it sees clean water but cannot cool the loop between the tank and the use points. The common mistake is exactly that: an engineer specifies EPDM gaskets, the reviewer moves the cooler past the UV to protect them, and the suite complains all summer that the water is 28 C.
Ozone sanitizes the wetted surfaces it reaches, but not the tank above the spray pattern, the vent filter, or a dead leg behind a closed valve. Many sites keep an annual hot water sanitization at 80 C for those, and then the loop cooler either survives 80 C on the tube side, which a double tubesheet shell-and-tube does without comment, or is isolated and bypassed. HeatX asks at quote time so the cooler is rated for the hottest thing the loop will ever do.
Send the loop flow, lamp ratings, pump horsepower, tank and pipe details, chilled water temperature and the sanitization method, and the cooler ships with the right gaskets fitted.
Call and talk it through with an engineer: 1-805-484-2992
Ozone limits what can touch the water, so these rows pair the duty with the materials allowed in the path. Compare them with your loop and send your ozone level and sanitization practice with the quote request.
| Condition | Detail |
| Product | Ozonated USP Purified Water, 0.02 to 0.2 ppm O3 upstream of UV destruct |
| Flow range | 20 to 250 gpm loop flow; the cooler sees the full loop |
| Temperatures in / out | 26 to 30 C in, 20 to 23 C out; continuous |
| Hold | None; ozone contact time is the loop transit, typicaly 5 to 15 minutes |
| Utility | Chilled water 7 to 10 C or a glycol chiller loop on the shell side |
| Approach | 3 to 5 C; a small unit sized on a modest duty |
| Construction | Double tubesheet straight tube 316L, or double-wall plate for loops that are never steamed |
| Finish / class | 316L, 20 Ra electropolished on the water side; ASME BPE |
| Ozone materials | 316L, PTFE, PVDF and ozone-grade fluoroelastomer in the water path; no nitrile, EPDM, silicone or brass |
| CIP / SIP | Ozone sanitizes continuously; periodic hot water sanitization at 80 C where the SOP keeps it |
With a dissolved ozone analyzer after the last destruct unit, alarmed near zero and interlocked to the draw valves. Qualification backs the analyzer with grab samples by a colorimetric method at the use points, and the destruct units are sized for the ozone dose, roughly three times the UV dose that disinfection needs. The cooler's part is indirect but real: destruct efficiency and the analyzer's reading both drift with water temperature, so a loop held at 20 to 23 C gives a steadier trend and a clearer margin than one that swings between 20 C in winter and 28 C in summer.
Not while it runs full and closed; the concern is wherever ozonated water is opened to the room. The tank vent, drains and sample points release ozone gas, and the OSHA exposure limit is 0.1 ppm averaged over a shift, so the tank vent carries a destruct unit and the room an ambient monitor. The exchanger's contribution is to drain completely to a vented drain before service, so maintenance staff never open a head full of ozonated water, and to have no pocket where standing water off-gasses into an ozone-rich space that finishes any gasket nearby. Both are drawing details, and we add them when the inquiry says the loop is ozonated.
Replacing lamps changes little; adding a unit adds heat the cooler was never sized for. A lamp's electrical input sets the heat it leaves in the water, and that input stays close to rating as the lamp ages, so a like-for-like change is invisible to the cooler. Adding one 5 kW destruct unit for a new use point adds about 15,000 Btu/h at 85 to 90 percent conversion, which is 15 to 18 percent of the 85,000 to 95,000 Btu/h duty on this page. That modification goes through change control with a new heat balance, and the cooler is rechecked against it before the lamp is energized.
Chilled water at 7 to 10 C on most loops; glycol only where a dedicated chiller serves the water system or the plant chilled water runs too warm in summer. A room at 24 C and 50 percent humidity has a dew point near 13 C, so either fluid makes the shell sweat, and glycol at -5 C sweats harder and frosts the connections. The shell, heads and headers are insulated with a sealed jacket whichever fluid is used, because condensation dripping onto a clean-area floor is a finding by itself.
The exchanger a WFI system is validated around: 80 C, double tubesheet, 15 to 20 Ra, and nothing added.
+ Learn More
Ambient loops run cheap and sanitize on a schedule; the exchanger has to do both jobs on the same tubes.
+ Learn More
Tank, loop and sub-loops held at 80 C by one heater sized on losses, cooled only where the suite needs it.
+ Learn More