A representative sample, at a temperature a technician can hold
Every WFI loop is sampled at the return, the tank and a set of use points on a schedule, and every pure steam generator is sampled as condensate. The sample has to be the loop: same water, same bioburden, same endotoxin, at a temperature that fits in a bottle without flashing or scalding the technician. A sample cooler is a very small double tubesheet exchanger, 1 to 5 sq ft, that takes 0.5 to 2 gpm through a sanitary sample valve and drops it from 80 C to 20 C.
The engineering is not the heat transfer. It is making sure the cooler adds nothing, holds nothing between samples and can be proven clean.
A sample cooler is on the sampling SOP, so it is validated equipment. It can defeat a sample three ways: by holding stagnant water between samples that grows organisms, by a rolled tube joint that lets tower water into the sample, or by an internal surface that sheds particles or rouge. The design answer is a double tubesheet unit with the sample in a single pass of electropolished 316L tube, tower or chilled water on the shell across an open gap, vertical so the tube drains through the outlet valve, and no elastomer in the sample path but the clamp gasket at the loop valve. Small units are often a coil rather than a bundle; the coil is still double tubesheet at both ends.
The first water out of a sample cooler is whatever sat in it since the last sample, and it is never the loop. The SOP flushes the cooler for a set time before the bottle opens; at 1 gpm through a cooler holding 0.3 liters, thirty seconds clears it about six times over, and most SOPs call for a minute or more with the loop-side valve fully open. The outlet is a sanitary valve, not a hose barb, because a hose is a dead leg with a bacterial population of its own and cannot be steamed. Between samples the loop-side valve is closed, the outlet stays open, and the cooler drains dry by gravity. A cooler that holds water between samples is the one the environmental monitoring trend catches first.
Pure steam is sampled as condensate, and the condensate is tested against the WFI monograph. The sample cooler on a pure steam header takes steam at 30 to 45 psig, condenses it and sub-cools the condensate to 20 to 30 C in one small unit. It sees 135 C on every sample, which sterilizes it, and the concern shifts from bioburden to non-condensable gases and superheat: the cooler is vented so air in the header does not blanket the tube, and the sample valve is a sanitary steam-rated valve with a slow opening so the first slug of steam does not hammer the coil. The shell is rated for the header pressure in case the tube fails; a domestic sample cooler with a 75 psig shell is not the right part.
Buying a boiler-house sample cooler. The coil-in-a-can units sold for power plant water are single tubesheet, 304 or carbon steel shell, rolled joints and a mill-finish coil, at a tenth of the price. On a WFI loop the rolled joint weeps tower water into the sample, the mill finish grows a film, the unit cannot be steamed, and QC spends a quarter chasing a bioburden trend that lives in the cooler. HeatX builds the sanitary version with the same five-minute installation and the right metallurgy.
Send the loop or header conditions, the number of sample points and the cooling water, and the coolers ship as a set.
Call and talk it through with an engineer: 1-805-484-2992
A sample cooler runs briefly and must leave nothing behind. The rows below show what it is built to handle at a loop or steam sampling point; send your sample source and valve arrangement with the quote request.
| Condition | Detail |
| Product | WFI or PW from a loop; pure steam condensate at a generator or header sample point |
| Flow range | 0.5 to 2 gpm during the sample; zero between samples |
| Temperatures in / out | 80 C to 20 to 25 C for loop water; 121 to 135 C steam to 20 to 30 C condensate |
| Hold | None; the flush before the sample clears the cooler volume at least three times |
| Utility | Chilled water 7 C or tower water 30 C on the shell, 2 to 5 gpm |
| Approach | 5 to 10 C; not critical, the outlet is a comfort temperature |
| Construction | Double tubesheet coil or straight-tube 316L, vertical, drains through the sample valve |
| Finish / class | 316L, 15 to 20 Ra electropolished, ASME BPE; steam version rated 150 psig shell and tube |
| Sample path | Sanitary diaphragm valve on the loop, cooler, sanitary outlet valve, no hose |
| CIP / SIP | Steamed from the loop side; pure steam version sterilizes itself every sample |
Only if the sample path adds something, and the design on this page is built so it cannot. Elastomers leach organics that read as TOC against a 500 ppb limit, an unpassivated tube can lift conductivity toward the 1.3 microsiemens per cm stage 1 limit at 25 C, and water held in the cooler between samples adds both. An all-welded electropolished 316L path with the single clamp gasket at the loop valve, flushed for the SOP time, gives grab results that track the online analyzer. The cooler also suits the conductivity test, because the stage 1 table is read at the measured temperature and 25 C is where its 1.3 figure applies. A persistent offset between grab and online readings is the first sign the cooler needs attention.
On a hot WFI loop the flush does most of the work; on an ambient PW loop the cooler joins every loop sanitization. Holding the cooling water off for the first part of the flush lets 80 C water sweep the product path before the cooler drains dry, which on a hot loop is a sanitization every sample. An ambient loop gives no such help, so the cooler is left open to the loop during hot water or ozone sanitization and steamed where clean steam reaches the sample point. The interval belongs in the sampling SOP, because a cooler that is not on the sanitization checklist is the one that gets forgotten when the loop is done.
Cooling helps the count rather than hurting it, because 80 C water would keep killing organisms in the bottle. What moves the result is time and the hardware: samples are plated within about two hours at room temperature, or refrigerated and plated within the hold the method validates, and any growth that occurred inside the cooler is counted as the loop. That is why the cooler drains dry between samples and why the flush clears its volume many times over before the bottle opens. A count that rises at one sample point and nowhere else usually lives in that point's cooler or hose, not in the loop.
A sanitary diaphragm valve with a short self-draining outlet, never a hose barb or a ball valve. Ball and needle valves hold water in the body cavity and can not be sanitized or drained. The valve is opened by hand or by a timed actuator that runs the flush, and its spout sits at a fixed height so the technician fills the bottle without touching it. On the pure steam version the outlet valve sees condensate at header pressure, so it is a steam-rated diaphragm valve; the slow opening belongs to the inlet valve on the header, which is what keeps a slug of steam off the coil.
Small enough to lift, built like the production unit, and sized so the data scales.
+ Learn More
Salts dissolve hot, buffers are used cold, media foams if you let it: the exchanger on the transfer line.
+ Learn More
The last exchanger before the needle: product at fill temperature, sterile boundary kept closed.
+ Learn More