Your behavior appears to be a little unusual. Please verify that you are not a bot.


Methods for troubleshooting leaks

February 1, 2016 By    

Editor’s note: At the request of our readers, LP Gas magazine is launching this new Service and Maintenance column. Feel free to share topics that you’d like for us to address in future issues.

Many of us have been there, alone at 11 o’clock at night, in the dead of winter, with a concerned customer smelling propane throughout the kitchen.

On one particular evening, I showed up to find the customer had already shut off the ball valve before the second-stage regulator and on the 500-gallon underground tank, as instructed during a Gas Check a few years earlier. After confirming these valves were off, I investigated the odor using a combustible gas detector and my sense of smell. I confirmed there was a slight odor but only in the kitchen around the four-burner cooktop. The customer told me the gas had been shut off for about 40 minutes.

Troubleshooting the leak

Now let’s get into the different methods used to troubleshoot the leak.

Our most accurate testing method on a standard, interior working pressure system (9 to 11 inches of water column), used in this house, is a manometer-type gauge. The industry’s preferred low-pressure testing instrument is a water or spring manometer.

Remember it was the middle of the night, and I did not know the severity of the leak. I did know the first- and second-stage systems were shut down. If this was a simple interruption of service, I would test the entire system using a high-pressure gauge at the tank. The tank was 100 feet from the house.

Leaving the second-stage valve off, I tested the 100 feet of underground gas piping to eliminate the first stage of leaks. I did this by installing a 0-30 pounds-per-square-inch gauge (PSIG) on the first-stage regulator. The first-stage system was found to be leak free. The reason for testing this way was to minimize piping lengths and ultimately get me closer to the leak that was suspected inside. Splitting the leak-testing method into two systems may take a bit longer but in the long run can be very effective.

Now, getting back to the manometer. With the ball before the second stage still in the off position, I turned on the tank, pressurizing the first-stage line that I knew was gas tight. I installed a manometer into the tap of the second-stage regulator and then slowly opened my valve below the second-stage regulator to pressurize the system. Once the pressure was bled below the regulator’s lockup, I waited. Within 30 seconds, I noticed my gauge had dropped from 9.5 inches of water column to 5 inches of water column – and realized I had a good-size leak on my hands.

I could not use a leak detector on the fittings under the cooktop while keeping the gas piping live due to a number of reasons: how far my second-stage regulator was from the kitchen, the time of night, me being alone and the customer already in a panic.

Quick reaction

So I decided to quickly pull the second-stage regulator off the piping, install a Rogers gauge (air test), shut off the ball valves at all of the gas appliances and complete a pressure test by pumping the system full of air to 3 PSIG.

There was no point in watching the gauge drop; I just needed the air so I could look for the leak inside using a leak detector.

Using a soap-style leak detector, I lathered the fittings directly beneath the cooktop, where I suspected the leak. Within seconds, I noticed large, consistent bubbles at the copper tubing flare connection located below the appliance shutoff. I disassembled the fittings to inspect the flare. Right away, I noticed small burs on the female side of the manual flare. I then cut the tubing and reflared and reassembled the fitting.

After my repair, the system was tested again, first using the Rogers gauge, followed by a low-pressure leak test. The leak was found; the system was made tight; and, most importantly, the customer was safe and happy.

There are many factors that play into leak detection, and often we only use one method to repair a leak. However, we must be prepared to troubleshoot in different environments and situations.


Ryan Card is the service manager for D.F. Richard Energy in Dover, N.H. He can be reached at rcard@dfrichard.com or 603-516-3225.

This article is tagged with , and posted in Current Issue

Comments are currently closed.