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

P+R+O+O=F of Performance

December 1, 2006 By    

A two-year study of propane vapor regulators has determined that age has little effect on the performance of first-and second-stage regulators, but it plays a significant role in how well single-stage devices work.

It also found a higher percentage of failures in warm, dry regions, suggesting that drying or hardening of the rubber components may be more significant in causing failures than metal corrosion.

The research was directed and funded by the Propane Education & Research Council to determine if there was scientific or engineering evidence to substantiate the service life recommendations of regulators manufacturers, who for decades recommended replacing the devices every 15 years. Some U.S. manufacturers have extended the service life of their products from 15 to 20 or 25 years since 2003.

PERC’s study found that the common two-stage regulator systems show no significant degradation during that period, suggesting that marketers can safely continue to use those products beyond the suggested time frames.

The report drew rebukes from the leading domestic regulator manufacturers. Representatives from both RegO and Fisher Gas Products, the two largest U.S. regulator manufacturers that account for about 85 percent of domestic sales, say they are still reviewing the study and will report their specific concerns to PERC.

Regulator failures by age
Regulator failures by age

They chastised PERC for spending more than $300,000 on a study they say duplicates existing research and paints an incomplete picture of regulator performance to the propane industry and the public.

“What benefit does this report provide to the consumer and to the industry? Does it put a cap on service life recommendations? We don’t know. At the end of the day, they are basically agreeing with what RegO did three years ago when it changed its recommendation to 25 years,” says Brian Davidson, director of sales and marketing for RegO.

“I almost think the report opens up many more questions than it answers. You know this report will show up in litigation somewhere down the road, and that will cost both the manufacturers and marketers.”

They also criticized the study’s author, Battelle Applied Energy Systems, for not involving them in the project. Battelle was paid $262,000 for the research.

“The report says the reason for doing the study was because there is no basis for manufacturers recommending a 15-year-serviced life. There was, but nobody ever asked us. Nor did anyone ever ask us why we switched it to 25 years,” Davidson says.

“Our biggest problem was that Battelle had a good source of information sitting right in front of them that they chose not to tap. And we can’t tell you why. ”

Regulator failures by environmental condition
Regulator failures by environmental condition

PERC officials say representatives from RegO, Fisher and Sherwood were consulted and, in fact, provided valuable input about the test protocol in late 2004 and early 2005. None was provided a copy of the report before its release, however.

Ed Varney, director of sales for Fisher Gas Products, likened the study to a public challenge of a car maker’s recommendations for an oil change. He says his company would have shared findings from its years of product testing had it been asked.

“As a company we are very serious about safety and would never do anything to compromise that,” Varney says. “We are not challenging the accuracy of the study; it’s more the whole impression of what it is trying to say. We have concerns that we wanted to make industry aware of and we raised them. We are still reviewing the report and will submit those issues to PERC.”

PERC President Roy Willis admits the process can be improved with a pre-publication briefing among interested parties that allows for feedback. That step will be added to future studies, he says.

He maintains that an independent study of regulator performance was needed and its findings vital to propane marketers.

“From my perspective, the question that generated the study was answered in a reasonable way. Some have expressed concerns about the study, but nobody has presented us with substantive information that differs from the findings of the study,” he says.

“Is there a safety concern about using this product? If it is prone to failure at a higher rate beyond 15 years, that’s information that users of this product need to have. If they have a perfectly functioning, closed-gas system, the question is whether it should be opened to replace a part on a timed interval in order to enhance safety – especially since opening the system invites human error. Our study didn’t find a basis for doing that.”

Regulator failures by environmental condition
Regulator failures by environmental condition

In addition to collecting and testing regulators, PERC required researchers to review U.S., European, Australian and Japanese propane regulator manufacturer’s catalogs, technical bulletins, instructions and safety literature to document service life standards. It found none.

The Gas Technology Institute – formerly known as the Gas Research Institute – was paid $50,000 to do the literature search and serve as advisors for the regulator testing. It cites a 1999 Korean study that raised the issue of the long-term effects a propane operating environment has on elastomer and spring performance.

But Davidson says the literature search ignored a 1981 study of more than 177,000 in-service regulators by now-defunct propane retailer Cal Gas. That study’s findings were the basis of RegO’s 15-year service life recommendation, Davidson says.

Battelle’s Rodney Osborne, project manager of the PERC report, says the Cal Gas study did not turn up in the literature search. Although the project has been talked about in the industry for years, nobody connected with PERC’s research project could produce it. A 1983 propane industry trade journal article about the Cal Gas study was reviewed after the PERC report was issued, but Osborne and Willis say it has no bearing on the outcome.

“We didn’t hear about it until after our report went public. It could have been a very thorough and complete study, but you can’t tell from the article. What was the protocol? What data does it contain? We don’t know because the article was more observational,” Osborne says.

“But after reading the article, it doesn’t change our findings at all because ours isn’t opinion. We believe it is an objective study and that the conclusions we drew were from objective data. That’s really the key.”

Willis says he is open to revisit the issue if new, scientific information is produced.

“I think everyone recognizes that those involved in this debate over the service life of a product have some vested interest. That’s why we try to focus on an independent, third-party analysis while the debate continues between the commercial interests and the scientific findings,” he says.

“The bottom line to all of this is that our work found no scientific basis for a 15-year service life. The rest of it is just noise.”

The study

Battelle collected 773 regulators no longer in service from 49 propane marketers in 27 states, with more than half coming from South Dakota, Mississippi and Iowa. The collection represented 13 brands and various types of regulators of all ages and service conditions. The reason the equipment was removed from service was not used as a selection criterion.

Battelle noted that the collection process was slow and difficult because many propane marketers had reservations about participating in the study.

“Particularly, some felt the study was flawed since regulators are removed from service based on their condition; not because of an arbitrary age designation. This very fact would affect the results of the study since many of the regulators tested would be defective and/or deteriorating,” the report acknowledges.

The collection was pared to 266 for external and internal inspections to evaluate corrosion, damage, or missing components; lock-up testing at three different inlet pressures and four different flow rates; and pressure relief testing. All but two of the tested devices were Fisher and RegO products, and all were within a month of having been removed from service, according to Osborne.

To test performance, Battelle used the flow/lock-up test and pressure relief/relief capacity tests in the Underwriters Laboratory standards for LP-Gas regulators. The U.L. standards are used to establish the operating parameters of newly manufactured regulators.

The report emphasizes that the test results do not project field failure rates among the millions of low-pressure propane vapor regulators in service nationwide.

“These failure rates are the result of an extremely rigorous test protocol that stresses the regulators to conditions not seen frequently in the field. Indeed, the particular combination of tests that were prescribed in the protocol may never be experienced in the field. The intent of the newly developed testing criteria was to generate failures. These failure rates could then be compared between independent parameters of manufacturer, environment, and others,” the report says.

Visual inspections

Before the regulators were tested, external and internal visual inspections were performed to identify and document any significant corrosion, damage, or missing components that might correlate with performance issues, including:

  • Corroded regulator body
  • Missing parts (screws, vent screens, bonnet caps)
  • Excessive paint
  • Physical damage to the regulator (holes or cracks in the regulator body)
  • Removal of fittings (corroded; difficult to remove; could not remove)
  • Excessive dirt

Regulators with seized fittings were not tested. Those that were corroded or had an accumulation of dirt or debris on the exterior were still tested. Those with cracks or holes in the regulator body could not be tested, but the damage was documented.

Issues identified from the internal visual inspection included:

  • Cross-threaded adjusting screw
  • Damage to adjusting screw
  • Adjusting spring seized in place or stiff
  • Burred threads
  • Internal corrosion and/or contaminants (dirt, oil)

Lock-up test results

Regulator lock-up was measured at three different inlet pressures and recorded in three successive trials for each inlet pressure. For first-stage, integral two-stage, and single-stage regulators lock-up was measured at 100 psig, 25 psig, and 250 psig inlet pressures. Lock-up for second-stage regulators was recorded at 10 psig, 5 psig, and 15 psig inlet pressures.

Test results show that the lock-up test performance of the first-stage regulators remains fairly consistent – and for the most part remains within U.L. criteria – regardless of age. In fact, the report says there does not appear to be a significant correlation between regulator performance and age, environment, or manufacturer.

The performance of second-stage regulators was more scattered, with the number of units not meeting test criteria increasing after 35 years of age.

Single-stage regulators tested poorly. The percentage of units that did not meet lock-up test criteria far exceeded that for the other regulator types. The report notes an apparent trend in the deterioration of regulator performance by age, although single-stage regulators of all ages did not perform well. No significant trend was found based on environment or manufacturer.

Although single-stage regulators are still in service, they are no longer permitted for new installations per NFPA 58 regulations.

Pressure relief test results

The pressure relief device start-to-discharge and reseat pressures were measured and recorded in three successive trials for each test.

The test results for first-stage regulators were fairly consistent with only five out of 50 not meeting the pressure relief test criteria. Each failed unit had start-to-discharge and/or reseating pressures that were too low. All regulators that did not meet the test criteria were from the same manufacturer, which was not identified in the report. Four out of the five were under 15 years of age. Regulators in this test were either from a cool, dry or warm, dry environment. None was older than 25 years.

Results for second-stage regulators were widely scattered for reseating pressures, regardless of regulator age. However, there did appear to be a slight age affect for the start-to-discharge pressures. More regulators older than 15 years tended to exceed the pressure relief test criteria than regulators less than 15 years in age. Twenty-seven of the 78 second-stage regulators did not meet the test criteria. Most had start-to-discharge pressures that were too high, ranging from 33.2 inches of water to 60.1 inches of water. One regulator still did not relieve after reaching 65 inches of water.

Integral two-stage regulators also produced widely scattered results regardless of age. Of the 11 integral two-stage regulators tested, four did not meet the test criteria. Most had start-to-discharge pressures that were too high.

All regulators that did not meet the pressure relief test criteria were from the same unidentified manufacturer and came from two locations. The report notes that such a small sample size makes it difficult to determine any trends between age, environmental conditions, and manufacturers and regulator performance.

Test results of single-stage regulators were widely scattered regardless of regulator age. Nearly half of the 17 regulators tested did not meet the test criteria. Most had start-to-discharge pressures that were too high.

Causes of failure

The report says the main causes of regulator failure identified in the research include:

  • Regulator chatters and leaks through the pressure relief device (PRD)
  • Leak in regulator body
  • PRD start-to discharge and/or reseat pressure too low
  • PRD start-to discharge pressure above the Underwriter Laboratory specification
  • Regulator discharge pressure will not stabilize
  • PRD did not relieve

Several of the regulators identified as failures were selected for detailed analysis. Findings from the analysis indicate a few possible trends as to why some regulators did not meet the test criteria.

For the regulators that were disassembled and analyzed, debris within the regulator body was the single most common potential cause for high regulator lock-up and/or leaks through the PRD. Some of the debris appeared to be corrosion from piping or containers, but others appear to be related to equipment manufacturing.

Others showed some damage to the regulator seat disc, which could have led to high lock-up pressures.

A free copy of the report is available from the PERC website at

Comments are currently closed.