HPAC Magazine

The Quest For The Smoking Gun

May 1, 2014 | By DAVID DEMMA

Performing a complete and thorough leak check of all potential leak paths is an art.

In my near 25 years employed as a sales/application engineer at a valve manufacturer I had many opportunities to visit the manufacturing plant. The manufacturing process is amazing to behold. It is a constant flow of product, which means there is a ongoing stream of processes, repeated over and over with great accuracy by all who are part of the functioning team at the plant.

It’s that fact that always amazed me…that you could have a workforce of hundreds, with each individual doing mostly the same thing for the entirety of their eight-hour (or longer) workday without losing momentum or focus, all the while maintaining the highest standards of efficiency and accuracy.

Now, what does this have to do with anything? During the time when I was a refrigeration technician, there was one task that gave me a small feeling for what it might be like to work in a manufacturing plant performing the same task all day long. That task is leak checking a large multi-evaporator system, such as a supermarket refrigeration system.

There are many equipment designs that might be employed in a supermarket, as such there is no typical system. The store design could call for any of the following designs:

(1) Individual condensing units piped to individual evaporator systems (for example: one compressor for four to 12 ft. meat display cases, etc.)

(2) Small multi-compressor racks of small/medium capacity, each piped to four to eight evaporator systems operating in a similar temperature range.

(3) Larger multi-compressor racks piped to 10 to 12 evaporator systems operating in a similar temperature range.

(4) Large split suction multi-compressor racks piped to 15 to 20 evaporator systems operating in two different temperature ranges.

(5)  A single central plant, operating in several different temperature ranges, with every evaporator system in the store piped to it.

Typically, the larger the system is, the larger the refrigerant charge is. Given the price of refrigerants, along with the environmental impact of refrigerant leaks, there has been a movement towards smaller systems with shorter piping runs to reduce the total overall refrigerant charge. However, there are still many larger multi-compressor rack systems in use. The fact is, research has shown that the average supermarket will experience an annual leak rate of approximately 20 per cent. Even if it is a smaller multi-compressor rack system with a charge in the 400 lb to 500 lb range, 20 per cent of that amount per year along with the leak check and repairs is a fairly large expense to incur.


The same research project that found the average supermarket annual leak rate in the 20 per cent range, also found that the common culprits for leaks were loose/missing Schrader valve caps and loose/missing TEV adjustment caps. These are easily preventable.

Now, as to my previous contention that leak checking a multi-compressor rack is similar to the repetitive tasks performed in a manufacturing facility, why do I make this comparison?

Let me illustrate using the following example: Several years ago I was asked to visit a few “trouble” jobsites, at a major supermarket chain. The problem was thought to be leaking gaskets on some valves. As it was reported, these leaks were amounting to refrigerant losses in the hundreds of pounds per month and they were occurring at multiple store locations.

I arrived at the store and the technician got his gear out and we went straight to the alleged leaking valve. Now, the technician was using his state of the art electronic leak detector, which was capable of detecting a leak in the range of 0.5 oz/year. Within a matter of moments the detector was making its alarming sound indicating the detection of a leak. I like to use the electronic detector to locate leaks and then verify those leaks with soap bubbles, so I asked the technician to get his bubbles. This alleged several hundred lbs/month leaking valve generated about one bubble the size of a pin head approximately every minute.

While the technician had already dutifully replaced the gaskets in this valve twice, this was definitely a false alarm.

Just for perspective, with a leak rate of 0.5 oz/year it would take over three years to leak one pound of refrigerant. Today’s leak detectors can locate miniscule leaks, but those leaks are not likely responsible for large repetitive refrigerant loss issues. Not to say that they shouldn’t be repaired (smaller leaks can always become larger leaks) but this valve was clearly not the smoking gun we were looking for.

It was at this point if I asked the technician if he had done a complete leak check of every fixture, component, and visible piping on this system. Sheepishly, he said “No.”

Now, you have to realize that leak checking the multi-compressor rack in the machine is a task in and of itself. A sample rack legend is shown in Table 1. This is a split suction rack, so there are medium temperature and low temperature fixtures piped to the rack. At the rack, there are five or more compressors with over 20 evaporator systems piped to it. Some of those systems are comprised of multiple evaporators.

Each compressor has several pressure control connections, an oil level control connected to the compressor body, but also with a sight glass (both neoprene seals), a Schrader valve or two, suction and discharge service valves, and maybe a suction filter-drier shell at the compressor inlet. Then there is a liquid solenoid valve, liquid ball valve, suction regulator, suction ball valve, possible hot gas solenoid valve, more Schrader valves, liquid filter-drier shell, receiver level gauge, receiver valves, possible subcooler circuit with TEVs and solenoid valves, heat reclaim valve, possible discharge/liquid pressure regulating valve for defrost; oh, and more Schrader valves, oil separator, oil reservoir, oil filter, access valves on the oil reservoir, oil piping to each compressor, multiple pressure transducers, relief valves, and a multitude of locations where piping is sitting on strut with clamps that should be tight, but may not be. And that is just the machine room.

There is a condenser, along with the piping lines going to and from the condenser. It is best to turn the system off and shut the condenser fans off, so a leak detector can properly pick up any hint of refrigerant in the tube bundle, the return bends, access valves, etc. While it might not seem to be a high priority to check the piping, it is not unheard of for a pipe clamp to loosen and vibrate against the strut, wearing a hole in the pipe. I will come back to this in a moment.

Then there is a heat reclaim coil, with the piping going to and from it. The same procedure applies as stated for the condenser.

Now, move out to the sales floor and various W/I boxes to check the low side of the system. If it is a gas defrost system, running the system through a defrost cycle will elevate the pressure and make leak detection easier. If not, then shut the entire system down and allow the low side pressure to rise to higher than operational levels. For display cases, running the electronic leak detector probe in the discharge air will sense any refrigerant that might be present. If this proves positive, it is necessary to remove any p
roduct from the display case and inspect the TEV, check valves if it is a gas defrost system, solenoid valves if present, and yes, Schrader valve for checking pressure to determine superheat and use soap bubbles to precisely locate the leak. This procedure needs to be repeated for each display case.

A similar procedure needs to be repeated for each W/I box evaporator.

If the store piping is routed underground, then it is recommended to locate the point of penetration and “sniff” with a detector there too. While underground leaks are not a common location of refrigerant leaks, they are always a possibility. Given that the piping is inside a “blanket” of insulation, any refrigerant leak in an underground line will travel through the insulation and will typically allow for detection above ground.

Back to the example: The technician and I had a short, but important conversation. The essence of the conversation was this: Locating one leak in a system that has hundreds of potential leak paths does not mean squat. Unless the technician is willing to perform a 100 per cent complete leak check on the system, including the machine room, high side, low side and all accessible piping, then he is just wasting his time and his customer’s money. Really, what good is it to find and repair one leak, only to return the following month and try to sell your customer on why you did not get it right the previous month.

Yes, it is time consuming. Yes, it is absolute drudgery akin to a worker in a manufacturing plant performing the same operation over and over all day long. You cannot lose determination or focus. You cannot afford to assume that “this one leak” is all there is. And you certainly cannot perform leak detection with any less urgency and exactness the closer you get to the end of the job. If you have not done a complete and thorough leak check of all potential leak paths, then you have not done the job right.  <> 

Dave Demma holds a degree in refrigeration engineering and worked as a journeyman refrigeration technician before moving into the manufacturing sector where he regularly trains contractor and engineering groups. He can be reached at ddemma@uri.com.


I called the technician a day later to learn what he had found. He reported one expansion valve leaking, one solenoid valve leaking at a W/I box, a few leaking Schrader valves and a leaking pipe (caused from a lose clamp that had vibrated and rubbed on the strut that it was sitting on). This was actually the largest of the leaks that were found.

After repairing these major leaks, and more importantly, knowing that he had performed a thorough and complete leak check of the system, the technician left the jobsite with the confidence that can only come from doing the job right.

I checked with this technician a few months later to see if there had been any additional leak issues at this store. He reported that there had not been any refrigerant loss issues.

The grocery chain in question used this issue to change the policy regarding leak detection, and mysteriously the leaking valve issue became a non-issue.



Stories continue below