Hydrogen testing in an accumulation chamber can be used to test multiple components for leaks simultaneously.

In any process cooling system, quality is as crucial as the price and design of the equipment, and the most important quality factor is the system tightness. A miniscule leak as small as a bacteria cell can cause the loss of 1 g of refrigerant per year, rendering the system inefficient or inoperable in a few years.

For equipment manufacturers, such defects can significantly increase warranty costs and degrade brand image. Organizations and facilities that use process cooling equipment also pay a heavy price for leakage; the U.S. Navy, for example, estimates that it spends $2.7 million each year replacing leaked refrigerants. Refrigerant leakage also has been linked to ozone depletion and global warming.

Table 1. All leak testing methods have advantages and disadvantages.
For these reasons, all components that contain a refrigerant must be thoroughly and systematically tested to ensure that they meet rigorous tightness standards. Several methods of leak detection can be used to achieve this goal. For many years, companies have employed wet methods in which an operator watches for bubbles to form at any leaks. Dry methods use gas detectors, either to detect refrigerant leakage after charging the system, or during leakage testing with a tracer gas to ensure tightness prior to charging the system with refrigerant. Each method has advantages and disadvantages (table 1).

Wet Testing Methods

Two types of wet or “bubble” testing have been used since the earliest days of process cooling quality assurance testing.
  • In water dunk testing, the component is filled to a specific positive air pressure and submerged in a well-lit tank filled with clear water.
  • In soap bubble testing, an operator observes a pressurized component that has been sprayed or brushed with a soapy solution.
Theoretically, with both methods, bubbles form at the source of a leak as a result of air pressure, and the amount of bubbles per minute can signify the size of the leak. The process is able to detect very small leaks if the operator patiently and carefully observes the component during the entire waiting time. Typically 10 to 20 min elapse before a bubble appears at leaks the size of 1 g/yr (a typical maximum allowable leakage standard for the HVAC/R industry).

The main advantage of bubble-testing methods is that they require little capital investment. Water is readily available, easily disposable and essentially cost-free when used in the small quantities required for water-dunk testing. Water dunking also has the advantage of simultaneously providing integral testing (checking tightness of the component as a whole) and location testing (pinpointing a leak).

However, bubble testing provides minimal quality assurance. Both the water dunk and soap bubble methods are highly dependent on the operator, who must be meticulous and patient. The operator's perspective can be limited, and small leaks might remain hidden on the reverse side of the component or in a recess. With water dunking, the operator must be careful not to pull down air bubbles that mask bubbles from a small leak or result in false rejection. With soap-bubble testing, larger leaks sometimes do not cause the formation of bubbles; instead, the compressed air just blows in a manner that can be difficult to observe.

The water or soapy solution used in wet processes also introduces problems. Capillary force can be extremely strong in the types of holes that need to be detected. The smaller the hole, the stronger the capillary force. The result is that liquid that has been sucked into a microscopic leak by capillary action cannot be forced out with compressed air, and no bubble will appear. In addition, the water can have a corrosive effect on the component being tested. Oxidization of surfaces can make it difficult to braze, solder and weld repairs. Moreover, soapy solutions are messy and can be difficult to clean off test objects.

Dry Testing Methods

Dry methods to detect leaks in process cooling equipment and systems use specially designed detectors or mass spectrometers to spot the presence of tracer gas or refrigerant passing through a leak, or to sense the total accumulation of gas that has escaped from the test object into an enclosed chamber.

Dry methods are faster and less prone to operator error than bubble testing, and they can be automated on production lines. They are also noncorrosive and do not require cleaning before rejected items are repaired. Dry testing can be used to detect the leakage of refrigerant already introduced into the system, or to ensure tightness using a tracer gas prior to introducing refrigerant.

Detecting refrigerant leakage with halogen detectors is a common procedure for service technicians repairing process cooling systems in the field. A similar method is used during the manufacturing process. After charging the system with refrigerant, manufacturers will “sniff” valves and joints with mass spectrometers or probe with halogen detectors. The selectivity of halogen detectors has improved over the past few years, allowing them to be used as an alternative to mass spectrometers.

An advantage of charging the equipment or system with refrigerant prior to testing is that leak testing and charging are reduced to a single step. If the object passes the leak test, it is already charged with refrigerant and ready to move to the next step in the manufacturing process.

However, many manufacturers cannot introduce refrigerant into an untested object due to environmental concerns and regulations. An additional problem is that rejected objects must be cleared of refrigerant prior to repair, and the gas must be recovered according to government regulations and industry standards. This is not an easy process; refrigerants are heavy gases, and their slow diffusion rate makes them difficult to clear.

For these reasons, many facilities are opting to perform leak testing prior to charging systems with refrigerant.

Hydrogen testing also is effective when used to probe specific joints and valves without an accumulation chamber.

Leak Detection Using Tracer Gas

Two tracer gases commonly used in dry leak detection are helium and hydrogen. (A third tracer gas -- refrigerant diluted in nitrogen -- is seldom used because of the cost of the gas and the environmental limitations on refrigerant emissions.)

Helium. Helium testing is used for the integral testing of components such as coils and tubing. It requires the use of specially designed mass spectrometers that are extremely sensitive to trace amounts of this gas. The method is capable of detecting very small leaks.

In a helium test, a helium-filled object is placed in a vacuum chamber. After a series of pumps achieves a vacuum in the chamber, a mass spectrometer is used to analyze the amount of helium in the chamber. Unfortunately, creating the vacuum lengthens the test cycle time.

Leak detection with helium can be expensive. A mass spectrometer is a complex instrument that requires regular maintenance. The vacuum chamber must be well constructed to withstand the vacuum, and the seals must be well maintained. The two or three stages of vacuum pumps also require maintenance and periodic replacement.

Additionally, helium gas is a natural resource that is in limited supply and can be expensive to obtain. For these reasons, in some cases, continuous test operations with helium can cost more than $100,000 per year.

Hydrogen. Hydrogen is a plentiful natural resource that costs one-third or one-fourth as much as helium gas. The recommended tracer gas mix (5 percent hydrogen/95 percent nitrogen) is safe and nonflammable.

The lightest element in the universe, hydrogen fills test objects quicker than helium, spreads easily throughout the test object, penetrates leaks readily and vents away much faster than other tracer gases. Hydrogen also has the advantages of not sticking to surfaces of the testing apparatus.

With hydrogen, integral testing does not require the creation of a vacuum. Hydrogen tests can be performed at atmospheric pressure in accumulation chambers. A fan simply clears the accumulation chamber between tests. Instead of mass spectrometry, the hydrogen method uses a microelectronic probe that is 100 percent selective to hydrogen.

Hydrogen testing also is effective when used to probe specific joints and valves without an accumulation chamber. For example, a U.S. tractor manufacturer performs air-conditioning leak tests by positioning a hydrogen probe at seven joints. A touchscreen programmable logic controller (PLC) displays graphical images and auditory signals that lead an operator sequentially through a 10-sec test of each joint.

Alternatively, specific joints or valves can be tested by enveloping these test points inside a small local enclosure, with a hydrogen probe performing an accumulation test of the limited air space (containing a dead volume of only a few cubic centimeters). This method is effective in detecting extremely small leaks. Hydrogen is well-suited for this application because it does not linger in the local enclosure and cause false rejections.

Manufacturers that require a fast test cycle time use another local enclosure method, called a “flush test.” Air is continually drawn through the local enclosure past the test point into a sampling probe. Because the airflow is constant, the concentration of trace gas in the airflow is a direct measure of the gas leakage rate. The flush method is very fast because it does not involve vacuum pumping or accumulation; however, it is slightly less sensitive than the accumulation method and detects leaks down to 5 g/yr.

Users of process cooling equipment have historically relied on bubble-testing methods because they considered helium mass spectrometry to be cost-prohibitive. Due to environmental and safety concerns, many facilities cannot charge their systems with refrigerant without pre-testing to ensure tightness. Hydrogen leak testing is gaining acceptance as a safe and environmentally friendly approach that provides high sensitivity and reliability without the high cost and complexity of helium testing.