9.2 Leak Testing, Pressurization & Purge Operation

9.2 Leak Testing, Pressurization & Purge Operation

Type I and Type II equipment operate at positive pressure, so a leak pushes refrigerant outward and can often be found with electronic detectors or soap bubbles at normal operating pressure. A low-pressure chiller flips this problem around: because it runs at or below atmospheric pressure, a leak pulls air and moisture inward rather than pushing refrigerant out. Finding that leak means temporarily bringing the system to a positive pressure - and doing so safely, within tight limits, is one of the most heavily tested Type III skills.

Staying Below the Rupture Disc

Every low-pressure chiller is fitted with a rupture disc - a one-time-use pressure-relief device built into the shell that is designed to burst and vent the refrigerant charge if internal pressure climbs too high, protecting the vessel from a catastrophic shell failure. The disc is rated at the chiller's maximum working pressure; Trane, for example, rates the disc on its R-123 CenTraVac chillers at 15 psig ±2 psig.

EPA lists the maximum leak-test pressure for low-pressure centrifugal chillers as a Type III test topic, and the controlling number is the manufacturer's test-pressure limit, which sits below the rupture-disc rating. A technician who pressurizes past it risks triggering the rupture disc mid-test, which vents the entire refrigerant charge to the atmosphere in the middle of what was supposed to be a controlled leak check - turning a diagnostic step into an accidental, large-scale refrigerant release.

The test limit always sits below the rupture-disc rating; check the nameplate and the manufacturer's instructions before pressurizing.

Preferring Controlled Heat Over Raw Nitrogen Pressure

To raise a low-pressure chiller from vacuum toward its test pressure, EPA's order of preference is controlled heat - most commonly, circulating warm water through the chiller's tube bundle, or applying other controlled heating methods specified by the manufacturer - over simply pumping in high-pressure nitrogen from a cylinder.

The logic follows directly from refrigerant pressure-temperature behavior: raising the refrigerant's temperature raises its saturation pressure in a predictable, gradual, and self-limiting way. A technician using warm water can reference a pressure-temperature chart for the refrigerant in the system, target a specific water temperature that corresponds to a safe pressure below the manufacturer's limit, and let the system rise to that pressure gently rather than forcing it there with an external gas source. This method is inherently more controllable and repeatable than valve-metering nitrogen from a cylinder that may be charged to well over 2,000 psig on the supply side.

Dry nitrogen remains an available and legitimate tool to supplement a leak test - for example, to nudge pressure up a bit further once heat alone has raised it partway, or to add a trace amount of refrigerant vapor for electronic leak detection. But nitrogen must always be added carefully and incrementally, watching the gauge continuously, precisely because excessive nitrogen pressure can blow the rupture disc just as easily as excessive heat can. A technician should never simply open a nitrogen regulator to some convenient setting and walk away; the manufacturer's limit has to be actively monitored and respected throughout the pressurization, regardless of which method - heat, nitrogen, or a combination - is being used to get there.

Purge Operation as a Leak-Detection Signal

As introduced in the previous section, a low-pressure chiller's purge unit automatically removes accumulated non-condensable gas - almost always air that has been drawn in through a leak while the system sits under vacuum. This makes purge behavior a genuinely useful diagnostic signal even before a dedicated leak test is performed:

• Normal, occasional purge activity is expected on most low-pressure chillers, since a small amount of air infiltration through shaft seals and fittings over time is common even on a tight system.
• Excessive or steadily increasing purge runtime, especially compared to that chiller's own recent history, signals that a leak is pulling in more atmosphere than normal and should prompt a full leak-testing sequence rather than being dismissed as routine.

Tracking purge runtime over time - not just glancing at a single reading - is what turns this into a reliable early-warning tool. A sudden jump in how often or how long the purge unit cycles, compared to that same machine's baseline, is one of the clearest field indicators that a chiller has developed a new leak somewhere in its shell, piping, or fittings.

Leak-Rate Repair Thresholds Still Apply to ODS Chillers

Many low-pressure chillers still in service use ozone-depleting substance (ODS) refrigerants, most commonly R-11 or R-123 (a hydrochlorofluorocarbon, and therefore still an ODS). For appliances holding a full charge of 50 lbs or more of an ODS refrigerant, EPA's leak-rate repair thresholds continue to apply by appliance category:

Most low-pressure chillers used to cool buildings fall under the comfort cooling category, which carries the strictest, lowest threshold of the three: 10%. If a chiller's calculated annual leak rate exceeds 10%, the owner or operator is obligated to repair the leak (generally within a set timeframe, with follow-up verification), rather than simply topping off refrigerant indefinitely. This 10/20/30% structure is the same leak-rate framework introduced for ODS appliances generally in the Core chapters of this guide - it is not a separate rule invented just for chillers, but it applies squarely to comfort-cooling low-pressure chillers whenever their full charge meets or exceeds the 50 lb threshold and their refrigerant remains classified as an ODS.