7.1 General

7.1.1 Leaks and leak detection

In non-destructive testing, a leak is defined as a hole, a porous area, a permeable area for gases or a different structure in the wall of a test specimen through which a gas can escape from one side of the wall to the other due to a difference in pressure or concentration [33]. Expressed in simpler terms, leaks are small holes through which gases or liquids flow from the side of higher pressure to the side of lower pressure. The geometry of the holes is not known. This means that the tester does not know whether the leak is a smooth-walled round pipe or occurs in the form of a crack or gap, for instance. Assumptions and calculations can only be made for ideal geometries. Since the real geometry of a leak channel is usually unknown, only calculated values can be assumed as an upper limit for a leakage rate. NOTE: The European standard DIN 1330-8 referred to previously uses the term "leakage rate". In the interests of readability we will continue to use the more common term "leak rate" in this book.

A leak can be a harmless leak such as a dripping water faucet. Leaks involving the escape of aggressive media or toxic substances can have more serious consequences. The accident suffered by the US space shuttle Challenger in 1986 was also due to the failure of an O-ring on the solid fuel rocket and the leakage of hot combustion gases.

Any number of technical products will not function, or will not function for an adequate period of time, if they have leaks.
Examples include:

The refrigerant circulation system in refrigerators
Air conditioning systems in cars
Automobile tires
Automotive fuel tanks or heating oil tanks
Processing systems in the chemical or pharmaceutical industries.

In many cases, the leak-tightness of machines and systems in the production process is an indispensable prerequisite for the quality of the manufactured products.

Returning to the original definition of a leak, we thus find that it is impossible to completely prevent substances from flowing through a wall. The term ”tight“ therefore refers to the requirements of the respective machine, plant or vessel, and must be quantified accordingly.

7.1.2 Leakage rate

Let us consider a bicycle tube having a volume of 4 liters. It has been inflated to a pressure of three bar (3,000 hPa), and without any additional inflation should have a maximum pressure loss of 1 bar (1,000 hPa) after a period of 30 days. The leakage rate has already been defined in Chapter 1.3.3: (Formula 1-35).

Q_{L}=\frac{\Delta p\cdot V}{\Delta t}

Q_LLeakage rate [Pa m³ s^-1]
Δp Pressure change during measurement period[Pa]
V Volume[m³]
Δt Measurement period[s]

Or to illustrate: The leakage rate of a vessel having a volume of 1 cubic meter is 1 Pa m³ s^-1, if the interior pressure increases or decreases by 1 Pa within 1 second. Please refer to Table 1-8 or to our app for conversion to other customary units.

Inserting the values for our bicycle tube then yields the permissible leakage rate

Q_{L}=\frac{1\cdot10^5\mathrm{\ Pa}\cdot4\cdot10^{-3}\mathrm{m}^3}{30\cdot24\cdot3,600\mbox{s}}=1.5\cdot10^{-4}\mathrm{\ Pa\ m}^3\mathrm{s}^{-1}

and we find that the bicycle tube with this leakage rate is sufficiently tight. These kinds of leakage rates can be found by means of the well-known bubble test method (Figure 7.1).

Figure 7.1: Bubble leak test on a bicycle tube

test

Now let us consider a refrigerator in which a loss of 10 g of refrigerant is allowable over a ten-year period. The refrigerant we use is R134a (1,1,1,2-Tetrafluoroethane) with a molecular weight of 102 g mol-1. The permissible loss is therefore about 224 Pa m³. This results in a permissible leakage rate of

Q_{L}=\frac{224\mathrm{\ Pa\ m}^3}{10\cdot365\cdot24\cdot3,600\mathrm{\ s}}=7.1\cdot10^{-7}\mathrm{\ Pa\ m}^3\mathrm{s}^{-1}

These kinds of leakage rates can only be localized and quantified by means of extremely sensitive measuring methods, for example with mass spectrometry and test gases that are not present in the atmosphere.

7.1.3 Tracer gases

The test gases that are used for leak detection (also called tracer gases) should satisfy the following conditions:

They should

Be non-toxic for humans, animals and the environment
Not displace air, as hazardous situations, such as suffocation, could otherwise occur
Be inert, i. e. slow to react, and should neither react chemically nor be flammable
If possible not be present in air. Only with a gas that is present in the smallest possible concentration in the ambient air is it possible to detect even the smallest leaks
Not be mistakable for other gases
Be quantifiable through test leaks.

The tracer gas helium satisfies all of these requirements. As a noble gas, it is not capable of chemically reacting. Only 5 ppm of it is present in atmospheric air, thus enabling even the smallest leakage to be detected. Since it is lighter than air, it does not pose a health hazard. Specific detection is possible using mass spectrometry, a highly sensitive and very selective analytical process (see chapters 6.1 and 7.2). There are many commercially available test leaks that are designed either as a diffusion leak or a flow leak.

The criteria described above are met by hardly any other test gas, an exception being forming gas 95/5 which is a mixture of 95 % nitrogen and 5 % hydrogen. The combustible hydrogen which is explosive in a wide mix range with air, is diluted to a degree where the mixture is neither explosive or combustible and is therefore safe for use as a test gas. The same mass spectrometry detectors can also be used as a sensitive test for hydrogen. Due to the higher background signal of hydrogen in the analytical technology used, it does not attain the same detection sensitivity as with the test gas helium, but it by far exceeds the detection sensitivity of the pressure decay method.