Several years ago, I was called to a facility being constructed for the Ohio Air National Guard. Over the weekend, the boiler had exploded, which damaged the flue and blew the side covers off the boiler and across the mechanical room.

The contractor said there had been several overpressurizations of the gas system over the previous weeks. Plastic shavings had been found in the utilities gas-service regulator. Each time the installing contractor and the local utility company cleaned, flushed, and reinstalled the regulator, they assured the contractor that the problem had been solved.

The gas service supplied 5 psig from the street to the meter/regulator assembly located outside the mechanical room. The regulator was set to deliver 14 in. of water column to the interior gas-distribution system. The regulator had an internal safety that would vent to atmosphere if the pressure reached 28 in. of water column (1 psig). The regulator functioned as designed and vented gas to atmosphere each time the regulator jammed with plastic shavings.

With the boiler having been specified with its own regulators as part of the gas train, could the elevated pressure have caused the explosion?

In reviewing submittal, installation, and maintenance data, a small item jumped out at me. Contained in the small print was a statement that the main regulator was “non-locking.” This type of regulator is limited in production to a few manufacturers. It regulates downstream pressure to the set point and design flow rate. Before that point is reached, the inlet and outlet pressures are the same. This could explain the explosion within the boiler that caused the damage.

I believe the following sequence of events happened: Though the contractor felt the problems associated with the plastic shavings had been corrected, this was not the case. The regulator overpressurized because of fouling by the plastic shavings. This meant the boiler's gas train was seeing 28 in. of water column, not the 14 in. it was designed to receive. Upon a call for heat, the boiler's controls initiated the firing sequence. The burner ran its pre-purge and called for pilot. The pilot was equipped with an industry- standard, locking-type regulator. That regulator could handle the 28-in. overpressurization. The pilot started its ignition sequence. The pilot flame ignited and was proved. Now, the controller knows it has pilot and calls for main burner. The main gas valve opens, allowing gas into the chamber at 28 in. of pressure.

This pressure, along with the air for combustion, was outside the combustible fuel/air-ratio mixture for ignition. As the gas flow developed and pushed into the combustion chamber, the non-locking regulator started to function and regulated the pressure to the correct setting of 14 in. Once the pressure dropped and the fuel/air mixture reached the limit of combustion, the main burner ignited. However, the combustion chamber already was packed with excess gas. The flame front from the main-burner ignition quickly expanded across the combustion chamber, resulting in a contained explosion and external damage. After much thought and discussion, this appeared to be the only logical and reasonable explanation of the situation.

The non-locking-type regulator was replaced. The utility regulator was replaced, and both the newly installed street main and gas-service lateral were flushed fully to prevent any further problems with the system. The boiler system continues to work without any known problems. Yes, the flue was repaired, the boiler factory-tested, and the damaged parts replaced.
David D. Dexter, PE, CPD, CPI, LEED
Korda/Nemeth Engineering Inc.
Columbus, Ohio

Send “war stories” of 400 or fewer words to the attention of Scott Arnold, managing editor, at sarnold@penton.com. Authors are paid $50 per published war story.