2airishuman
Contributor
Last week, the burst disc in one of my LP72s failed prematurely while the cylinder was flat in the bed of my pickup truck. The truck was parked on a warm, sunny day. The force of the air being discharged was sufficient to lift the end of the cylinder from the truck bed, causing the cylinder to strike the front of the bed with considerable force, as well as striking the rear window of the truck. The bed will require repair to allow the tonneau cover to seal, and the glass will have to be replaced. Estimated repair costs are $2150.
There were no injuries, since there was no one in the immediate vicinity of the cylinder when this happened. Witnesses heard the noise but there was nothing to see but the aftermath by the time they arrived to investigate.
I had loaded a number of cylinders in the truck in anticipation of a day of diving for 3 people in the Brainerd/Crosby area over Memorial Day weekend. Due to bad weather, we never went, and I left the cylinders in the truck in anticipation that we would dive them somewhere else over the coming week.
The cylinders were not overfilled. In fact, the affected cylinder had been used briefly on a dive that was called within the first few minutes and only had about 2200-2300 PSI remaining.
Disassembly of the valve showed that the burst disc had sheared completely:
Analysis
This cylinder was one of 12 LP72s that I use for shore dives. I had purchased these cylinders from various sources over the last two years. I have replaced most of the valves with Thermo Pro valves, but a handful of the nicer original valves remain in service. LP72s have a service pressure of 2250 PSI and require a burst disc of 3750 PSI. This size of burst disc has been recently discontinued, but some old stock is still available from dealers, all of it with the newer 1-piece construction.
I did not replace the burst disc in this valve when reconditioning the cylinder because the newer 1-piece burst discs, the only ones available in the proper pressure rating, would protrude from the valve body making the cylinder awkward to lift and carry by the valve. However, the valve had been disassembled for cleaning, with the burst disc removed and re-installed. I believe that the removal and reinstallation of the burst disc, combined with the unknown prior service history of the burst disc assembly, were the main cause of the accident.
While the cylinders had been left in the sun on a warm day (85 degrees), I do not believe that the cylinder temperature would have exceeded about 120 degrees, which would lead to only a 10% increase in pressure over that at 72 degrees, not enough to contribute meaningfully to a burst disc failure.
The failure of the burst disc may have been uneventful had the newer style been used. A known limitation of the flush style burst disc assembly on this valve is that it discharges air in a single high-pressure stream, unlike newer assemblies that have 3 or 6 discharge holes arranged radially. The single discharge resulted in a large enough amount of force to cause the cylinder to lift upright and cause damage.
DOT and SCUBA industry guidelines
The valve was in place when the cylinder was hydro tested earlier this year. The hydro facility passed the cylinder and also issued a VIP sticker. While DOT requirements mandate that requalifiers confirm that the correct size PRD (pressure relief device) be installed, if the valve is installed in the cylinder at the time of requalifiaction, replacement of the PRD is not required.
PSI procedures require that a SCUBA shop providing a hydro test service (even if the actual hydro is subbed out as is commonly the case) replace the burst disc at time of hydro if a new burst disc is available. PSI shops differ in their interpretation of this rule.
Conclusions
Though known, the hazards posed by burst disc failure, and the safety problems specific to older single-outlet burst disc assemblies, are not considered serious and are not widely communicated to divers or dive shops. Accidents usually result in property damage (only) or minor injuries, and are not treated as "diving accidents," leading to underreporting and a lack of appreciation for the risks.
The use of burst discs in the USA and in jurisdictions that have copied its regulatory framework is largely an accident of history, because the wording of legislation intended to cover steam boiler safety devices in the late 1800s was broad enough to apply to compressed gas cylinders. Pressure relief devices are not used on high-pressure cylinders in Europe, which instead relies on tighter controls of the filling process for protection against accidents resulting from overfilling. It is unclear whether a balance-of-harms analysis would come out in favor of the U.S. approach.
There were no injuries, since there was no one in the immediate vicinity of the cylinder when this happened. Witnesses heard the noise but there was nothing to see but the aftermath by the time they arrived to investigate.
I had loaded a number of cylinders in the truck in anticipation of a day of diving for 3 people in the Brainerd/Crosby area over Memorial Day weekend. Due to bad weather, we never went, and I left the cylinders in the truck in anticipation that we would dive them somewhere else over the coming week.
The cylinders were not overfilled. In fact, the affected cylinder had been used briefly on a dive that was called within the first few minutes and only had about 2200-2300 PSI remaining.
Disassembly of the valve showed that the burst disc had sheared completely:
Analysis
This cylinder was one of 12 LP72s that I use for shore dives. I had purchased these cylinders from various sources over the last two years. I have replaced most of the valves with Thermo Pro valves, but a handful of the nicer original valves remain in service. LP72s have a service pressure of 2250 PSI and require a burst disc of 3750 PSI. This size of burst disc has been recently discontinued, but some old stock is still available from dealers, all of it with the newer 1-piece construction.
I did not replace the burst disc in this valve when reconditioning the cylinder because the newer 1-piece burst discs, the only ones available in the proper pressure rating, would protrude from the valve body making the cylinder awkward to lift and carry by the valve. However, the valve had been disassembled for cleaning, with the burst disc removed and re-installed. I believe that the removal and reinstallation of the burst disc, combined with the unknown prior service history of the burst disc assembly, were the main cause of the accident.
While the cylinders had been left in the sun on a warm day (85 degrees), I do not believe that the cylinder temperature would have exceeded about 120 degrees, which would lead to only a 10% increase in pressure over that at 72 degrees, not enough to contribute meaningfully to a burst disc failure.
The failure of the burst disc may have been uneventful had the newer style been used. A known limitation of the flush style burst disc assembly on this valve is that it discharges air in a single high-pressure stream, unlike newer assemblies that have 3 or 6 discharge holes arranged radially. The single discharge resulted in a large enough amount of force to cause the cylinder to lift upright and cause damage.
DOT and SCUBA industry guidelines
The valve was in place when the cylinder was hydro tested earlier this year. The hydro facility passed the cylinder and also issued a VIP sticker. While DOT requirements mandate that requalifiers confirm that the correct size PRD (pressure relief device) be installed, if the valve is installed in the cylinder at the time of requalifiaction, replacement of the PRD is not required.
PSI procedures require that a SCUBA shop providing a hydro test service (even if the actual hydro is subbed out as is commonly the case) replace the burst disc at time of hydro if a new burst disc is available. PSI shops differ in their interpretation of this rule.
Conclusions
Though known, the hazards posed by burst disc failure, and the safety problems specific to older single-outlet burst disc assemblies, are not considered serious and are not widely communicated to divers or dive shops. Accidents usually result in property damage (only) or minor injuries, and are not treated as "diving accidents," leading to underreporting and a lack of appreciation for the risks.
The use of burst discs in the USA and in jurisdictions that have copied its regulatory framework is largely an accident of history, because the wording of legislation intended to cover steam boiler safety devices in the late 1800s was broad enough to apply to compressed gas cylinders. Pressure relief devices are not used on high-pressure cylinders in Europe, which instead relies on tighter controls of the filling process for protection against accidents resulting from overfilling. It is unclear whether a balance-of-harms analysis would come out in favor of the U.S. approach.
