COVID-19 Resources for Contractors (PDF)

Trapped refrigerant: Excessive liquid can lead to system failure


This hydrostatic expansion resulted from isolating an evaporator. (Technical Safety B.C.)

By Greg Scrivener

Trapping or isolating refrigerant can be dangerous if it is done with too much liquid present. Most of us are familiar with the rules about filling refrigerant recovery tanks and propane tanks – usually to 80 percent maximum liquid by volume.

Those who work on refrigeration systems may also be familiar with the regulations surrounding liquid storage in high-pressure receivers – a maximum 90 percent liquid by volume, according to the CSA B52 Mechanical Refrigeration Code. Fewer of us are familiar with the possible consequences of trapping refrigerant in condensers, evaporators and piping.  The goal of this article is to shed some light on why these regulations exist and what situations require caution.

Excessive pressure

The first thing to understand is that any time you trap a high-volume percentage of refrigerant as a liquid, you are creating a scenario where excessive pressures could develop.  While the different refrigerants all have slightly different coefficients of expansion, they all cause a similar pressure increase.  Fig. 1 shows the calculated pressure increase of six different refrigerants trapped with 80 percent liquid volume at a temperature of 20°F in 20 feet of one-inch pipe and subject to an energy input of 1,200 Btu’s.

Trapped liquid caused this weld fracture in an evaporator. (Technical Safety B.C.)

An important observation is that there need not be 100 percent liquid to cause a rapid pressure increase. It is also poignant how little energy is required to increase the pressure; you can see in Figure 1 that for the small volume in 20 feet of one-inch pipe, it only took 1/20 of a ton to hit 5000 psi for several refrigerants!

Fig. 2 shows the results from a similar calculation, except that it is for a single refrigerant at several different initial temperatures. In this case, it is a one foot square vessel of R134A. You can see that at 80°F, the rapid pressure increase begins at around 2,000 Btu and at 0°F it begins at 3000 Btu.  That is a 50 percent increase in the energy required to cause an expansion when the refrigerant started out cold.

This discovery is somewhat misleading. As you add energy you increase the temperature and as you increase the temperature, you decrease the amount of energy available.  In most cases we have an energy source heating our refrigerant piping and vessels that has a limited temperature (i.e. the ambient temperature in your truck, for example).

The curved red line on Fig. 2 shows the pressure at 150°F. It may indeed require more energy to cause the refrigerant with the lower temperature to have a pressure increase above the design pressure, but it can all happen at low heat source temperatures (like the back of a work truck parked in the sun).

If you only have a 150°F heat source available, it would be impossible to over pressurize the refrigerant vessel that started at between 40°F and 80°F and was 80 percent full.  This explains why the transportation regulations require that a cylinder is no more than 80 percent full.  As we increase this volume of liquid, the temperature at which the over-pressure happens is lower.  Figure 3, which uses an initial temperature of 60°F and R134A, shows that if the liquid volume percentage increases from 80 percent to 90 percent we can see excessive pressures happen more rapidly.

I hope we’ve developed an appreciation for the dramatic pressure increases that can occur when we trap refrigerant with high volumes of liquid.  This is important because there are several ways this situation can develop in our systems and we should be vigilant during design, operation and service to avoid them.  There are also several situations that can develop where the heat source is at a much greater temperature than the 150°F we’ve been working with so far.

Figure 1: The calculated pressure increase of six refrigerants that were trapped at 20F in 20 feet of one-inch piping.

Cold liquid

Cold liquid is always more susceptible to an over-pressure in the event it is trapped. Industrial refrigeration mechanics and operators usually understand this because liquid overfeed systems are so common.

In typical air conditioning systems, it is rare to have liquid that is very cold. However, in certain remote condenser applications that run in the winter (i.e. server rooms), there is a chance that cold liquid is entering the evaporator from the condenser.

Figure 2: The calculated pressure rise caused by adding energy to one-foot square of R134A at different initial temperatures. The red curve indicates the pressure at 150°F for all temperatures.

It is a similar case for most commercial refrigeration systems where sub-cooling circuits on the condenser are fairly common. The liquid coming in from outside in the winter is cold and often, the line needs to be insulated to prevent condensation. In grocery store applications, we frequently use the high or medium temperature rack to provide sub-cooling for the liquid on the low temperature rack. This is a very efficient option, but it means that we are using cold liquid throughout the store.

There are several things you should consider when designing and operating a refrigeration system when liquid can be trapped:

  • If liquid can be trapped (even warm liquid) between automatically closing valves, you must provide a pressure relief valve for the piping in between. This is often an internal relief valve that simply bleeds back into the system. Most refrigeration solenoids will not hold pressure on the outlet side so, if the outlet pressure increases above the inlet pressure, they will open and prevent the over-pressure situation. However, not all solenoids do this and almost no motorized valves do either. Use caution selecting or replacing valves if they could potentially trap liquid.
  • If the liquid can be trapped between manual valves, then it is advisable to car-seal one of the valves open. At the very least, this makes anyone interested in closing the valve ponder why someone would have it car-sealed open. In reality, sites with valves like this should have a valve control program that outlines who is allowed to use these particular valves and have procedures in place for such use.

Figure 3: The calculated pressure rise at different initial liquid volumes for one-foot square of R134A with an initial temperature of 60F. The redline is an isotherm at 150F.

Heat sources

So far in this article, we really only considered 150°F heat sources, but there are several higher temperature sources we should consider.

  • There are code requirements for pressure relief valves on evaporators where there is a heat source within 18 inches. This would seem to apply to defrost heaters. In the U.S., exemptions are made for typical small UL listed evaporators. However, on larger equipment, relief valves should be considered and may be required.

    Use caution when assuming solenoid valves will prevent over-pressure by allowing reverse flow. Check valves installed right after solenoids often negate this feature.

  • Confined spaces, such as shipping containers in the sun, can become extremely hot.
  • Heating overfilled refrigeration vessels with torches can be very dangerous.
  • Condensers don’t usually require relief valves, but consideration should be given to condensers that are isolated in the winter (split condensers with no bleed valve and systems where condensers are shut down). In the winter, the refrigerant is attracted to the low pressure caused by the cold outside. There have been several reported cases of the isolation valves leaking enough that refrigerant vapour leaked into the condenser; once there, the refrigerant condensed and the condenser became full of liquid. Once the condensers warmed up in the spring they ruptured.

All of this information helps us understand how important it is not to create situations were liquid is isolated in refrigeration systems.  We’ve seen that even at 80 percent volume, catastrophic pressures can develop with relatively low temperatures. We have certainly not covered all of the relevant information on this topic, but I hope that what we have covered has been useful and helps you prevent future failures.


Comments are closed.