[Noah]

I was looking online for an Oxygen regulator online when I came across this picture, which looks a lot like the regulator I occasionally use during high flights. Certainly got my attention! The spontaneous combustion resulting SIMPLY FROM TURNING ON THE TANK VALVE killed a firefighter, and the NIOSH report analyzing the failure had my jaw on the floor several times.

For those (like me) who are using similar aluminum medical regulators in flight - you need to read and understand this report:

http://www.cdc.gov/niosh/fire/reports/face9823.html

Some excerpts below:

As he opened the cylinder post valve, the cylinder emitted a loud popping sound and then flashed, releasing two 4-foot flames...his clothes ignited from the waist up.

The most probable ignition mechanism was particle impact on the filter during the initial flow transient after the fire fighter opened the cylinder valve. This ignition led to the burning of the regulator?s aluminum body which caused the flash. Particle impact has been shown to be one of the most efficient mechanisms for directly igniting metallic components in a high-pressure oxygen environment. Particle impact ignitions occur when a metallic particle contained in the oxygen flow contacts a rigid surface and ignites. The ignition of the particle then promotes ignition of the target material.

With the exception of aluminum, testing indicates that the particle itself must ignite during the impact event for an ignition of the target material to occur. For aluminum, however, even inert particles, such as grains of sand, have been shown to cause ignition of the target material. Testing also indicates that aluminum particles, such as would be produced from the aluminum cylinder, are susceptible to ignition when they are in an oxygen flow stream. Aluminum has been proven to ignite by particle impact at low temperatures and at sonic flow rates similar to the conditions that exist in the valve and regulator assembly. The aluminum is flammable at pressures as low as 35 pounds per square inch gauge (psig) and has been shown to ignite by particle impact at the flow and temperature conditions present in the valve at the time of the incident.
In this incident, tests show that the particle traveled with the oxygen downstream (into the regulator) and ignited as it made contact with the bronze sintered-inlet filter. Testing indicates that while bronze has been shown to resist ignition and sustained combustion at these pressures, the thin cross-section of the filter and the very close proximity to the aluminum body provides for a kindling path to the aluminum body for particle impact ignitions on the surface of the filter. While bronze is resistant to ignition, experience has shown that sintered filters can melt and break apart when exposed to a strong ignition mechanism like particle impact.

The regulator involved in the incident was an aluminum body regulator with a bronze sintered-inlet filter housed inside an aluminum downstream flow path. The design of the high-pressure section provides minimal protection of the highly flammable aluminum to promoted ignition mechanisms. Further, the significant amount of aluminum in this regulator, directly exposed to the high pressure environment and oxygen flow, produced a design that is susceptible to an ignition. The design also allows for combustion in the high-pressure port to punch through the main seat in the regulator directly and progress into the piston barrel leading to rapid involvement of the low-pressure components and venting of combustion by-products outward (through the vent ports), potentially towards the operator... Bronze or brass, which are both non flammable at the pressures in the regulator and do not ignite by particle impact, act as a shield between the particle that ignites and the aluminum body. In this type of design, a particle ignition usually will burn itself out before kindling the ignition and combustion of surrounding materials.

This was a real eye opening report and very educational, especially with regard to the recommendations at the end (I'm only copying the most pertinent ones but they are all worth reading):

Recommendation #1: Fire departments should use oxygen regulators constructed of materials having an oxygen compatibility equivalent to brass.

DISCUSSION: Aluminum alloys are attractive candidate materials for pressure vessels because of their high strength-to-weight ratios. High pressure oxygen system components for portable or flight use must be lightweight, so it may appear to be desirable to build their housings from such lightweight metals as aluminum. The use of aluminum alloys in lines, valves, and other components should be avoided whenever possible because they easily ignite in high-pressure oxygen, burn rapidly, and have very high heats of combustion. Aluminum is ignited exceptionally easily by friction because the wear destroys its protective oxide layer.

Aluminum is easily ignited by particle impact, and aluminum particulate is a far more effective ignition source than many other metal particulate tested to date (titanium particulate has not been tested). High-pressure oxygen systems fabricated from aluminum must be designed with extreme care to eliminate particulate. Testing has shown that aluminum is substantially more flammable in oxygen than brass or other high copper or high nickel alloys.7
Sources indicate that commonly used aluminum alloys can easily burn in the presence of high-pressure pure oxygen once an ignition is present. Thus, aluminum will burn in pure oxygen at a pressure of 35 pounds per square inch (psi) (this is about twice the normal atmospheric pressure) whereas some brass alloys require over 5000 psi of pure oxygen to burn. Aluminum will also produce approximately 10 times the amount of heat provided by copper alloys when burning.

One concern of using aluminum in the regulator flow path is the possibility of particle impact and the aluminum not being able to contain the ignition. Particles can be introduced into oxygen resuscitators in many different ways. Experts suggest the presence of a particle or particles in the cylinders is not as problematic as the design of the oxygen flow path and the materials used.

The cylinder has a post valve that closes off the oxygen opening and allows the regulator to attach (Diagram 2). When the post valve is screwed into the aluminum cylinder, there is a possibility that the two metals rubbing together (galling) could create metal particles that would remain enclosed in the cylinder body. Galling is a condition involving smearing and transfer of material from one surface to the other and particles could be introduced by metal-to-metal of seals rubbing which occurs when the post valve is opened and closed. The frictional heat of the galling could lead to ignition of the valve; or the particles generated by the galling could cause malfunction or ignition of another component within the regulator.

Therefore, the design of the regulator?s flow path should be resistant to ignition if particles should occur. Experts suggest the regulator flow path should be lined with brass, bronze, or a similar material which would resist particle ignition, and that using such a material would shield the particle ignition and provide the opportunity to burn out. Particulate migration from the cylinder can be minimized by the installation of a standoff tube (bayonette) at the inlet of the post valve.

Experts suggest that the design of the regulator involved in this incident allowed for combustion in the high-pressure port to punch through the main seat directly and progress into the piston barrel, leading to rapid involvement of the low-pressure components and venting of combustion byproducts outward (through the vent ports).

In this incident, the regulator (with the aluminum flow path) could not contain the ignition of the particle impact. The significant amount of aluminum in this regulator, directly exposed to the high-pressure environment and oxygen flow, produced a design that was susceptible to an ignition mechanism of this nature. The testing laboratory recommended that the flow path should be constructed of a metal such as brass or bronze to reduce the risk of a flash.

Recommendation #3: Fire departments should ensure that when opening a cylinder post valve with the regulator attached, it should be opened slowly and positioned away from the operator and other people

 

[Low Pass]

Interesting and thanks for sharing, but with the millions of portable CGA-870 style aluminum oxygen bottles and regulators that are and have been in use for decades with minimal failures, I'm not seeing a substantial risk. I didn't read in this report about hydrocarbon contamination. That is a source for spontaneous combustion in the presence of pure oxygen. Maybe this was a factor.

 

[BillL]

I was just reading (transfill research) about bomb tests that NASA ran on pure oxygen at high pressure and sensitivities thereof. Brass is very good, actually stainless was less than brass. The initial opening of the valve results in rapid pressurization of the regulator cavity and instantaneous local high temperatures. It is a good reminder not to jamb open the valve.

As a corollary, gassy mines (MHSA part 31) certification used to disqualify aluminum engine components due to spark hazard unless coated.

Thanks for the reminder.

 

[Chkaharyer99]

Are Aluminum Oxygen Regulators Safe?

ECRI Institute article:

http://www.mdsr.ecri.org/summary/detail.aspx?doc_id=8313

Mitigation:


"Particle ignition can typically be avoided by following basic compressed-gas safety practices. These include clearing the cylinder valve stem (by "cracking" the cylinder valve to release a short burst of gas) before attaching a regulator and ensuring that the regulator is in a no-flow state before attaching it to the cylinder. In the aluminum regulator fires that ECRI has investigated, such basic safety practices were violated. Also, while not a sure preventive measure, the use of a sintered inlet filter disk rather than an easily displaced mesh screen inlet filter can help reduce the fire risk. (One major regulator manufacturer has made this change to its regulators.)"

"The risk of adiabatic compression ignition can be minimized both by keeping all regulators free of oils, dirt, and other combustible substances and by opening the cylinder valve slowly, rather than rapidly."

Thank you for the article. Its a great reminder to use good quality (proven) clean equipment and to follow specific procedures.

 

[Noah]

After some more research on this topic, I found the following ASTM article:
http://www.astm.org/SNEWS/ND_2008/chiffoleaunewton_nd08.html

Some excerpts copied here:

Developing an Effective ASTM International Standard to Prevent Further Devastating Medical Oxygen Fires

The U.S. FDA received 16 reports of other similar fires between 1993 and 1999 involving aluminum regulators attached to cylinder valves of portable oxygen cylinders. Considering the number of devices in clinical use, oxygen regulator fires are relatively rare; however, the consequences of these fires were quite serious. In total, these incidents caused severe burns to 11 health care workers, EMTs and patients, and resulted in the loss of two ambulances and significant damage to a fire station.3,4 What makes this more tragic is that most of the fire victims were first responders, EMTs, firefighters ? people who save lives ? and the cause of their injuries was life-saving equipment. In February 1999, the FDA and the U.S. National Institute for Occupational Safety and Health released a joint Public Health Advisory titled ?Explosions and Fires in Aluminum Oxygen Regulators? to alert routine users of the possible hazards associated with the equipment.4
Because of the severity and increasing frequency of these types of incidents, FDA and NIOSH attempted to identify the causes of these fires by consulting the expertise of organizations such as the National Aeronautics and Space Administration and ASTM International. Members of ASTM Committee G04 on Compatibility and Sensitivity of Materials in Oxygen Enriched Atmospheres had already been involved with the investigation of the medical oxygen regulator fires and were familiar with the cause of the fires. The consensus of the committee was to form a task group to develop a standard for evaluating ignition sensitivity and fault tolerance of medical oxygen regulators with the goal of preventing further fires.

This paper follows the development of a successful standard ? from the need for a new standard to establishing a provisional standard and performing round-robin testing on the current version of ASTM standard G175, Test Method for Evaluating the Ignition Sensitivity and Fault Tolerance of Oxygen Regulators Used for Medical and Emergency Applications. The benefit and impact of this standard is clear and simple yet extremely remarkable: Since the inception of ASTM G175, no fires of oxygen medical regulators that have successfully met the requirements of this standard have been recorded.

The Need for a New Standard
By the year 2000, oxygen fire forensic experts had investigated or inspected evidence from 11 separate incidents, including the first two presented in this paper. To identify specific design or usage problems that were potentially contributing to the fires, they performed an evaluation of adverse event information and forensic analyses of burned regulators. The forensic analyses of failures employ a root cause and origin analysis approach. For oxygen fires, this involves determining the point of ignition and the method of ignition, which is otherwise known as an ignition mechanism.

In order for a fire to occur, three elements (fuel, oxidizer and ignition) are required, as depicted by the fire triangle. Two elements are always present within medical oxygen equipment: The equipment materials are considered the fuel and the pressurized oxygen is considered the oxidizer. Therefore, only an ignition mechanism strong enough to ignite the fuel is required for medical oxygen equipment to catch fire. The strength of an ignition mechanism required for the ignition of materials within a high pressure oxygen environment is much less than the strength required in an air environment because of the relatively low ignition energy of materials in oxygen.

In all of the investigated fires, the ignition originated within the oxygen-wetted areas of either the regulator or cylinder valve. Of the 11 fires examined by WHA, a variety of four ignition mechanisms contributed to the fires: heat of compression ignition, contaminant ignition, particle impact ignition and promoted ignition.

Other than igniting the regulator materials, heat of compression is capable of igniting any contaminants present. Contaminants (i.e., flammable foreign matter not intended to be present in oxygen components) such as hydrocarbon oils are easily ignited by heat of compression compared to the solid materials. Once ignited, contaminants can release sufficient energy from their heat of combustion to kindle the materials of the regulator. This is known as contaminant ignition.

When the cylinder valve is first opened, the gas flow is extremely fast, with velocities approaching the speed of sound across the seat of the valve. Accelerated by this high velocity flow, small metallic particles generated during assembly and operation can shoot out of the valve into the regulator. The impact of these particles against the regulator materials causes a transfer of kinetic energy to thermal energy, potentially igniting the particle and the impacted material. This is known as particle impact ignition. Aluminum is extremely susceptible to this ignition mechanism.
The last ignition mechanism of concern for the type of oxygen equipment shown in Figure 7 is known as promoted ignition. This requires another ignition mechanism like the three described above, however, the ignition occurs upstream of the gas flow, and the fire propagates with the flow to kindle flammable materials in its path. If ignition occurred within the cylinder valve, such as ignition of the valve seat, then the gas flow would force the fire toward the regulator and kindle the flammable materials of the regulator.

Each of the above ignition mechanisms leave different fire patterns and clues that help identify the root cause of the fire. The major problem identified was the use of aluminum in critical areas of some regulators. Aluminum is used because it is lightweight, however, it is also flammable and highly susceptible to most types of relevant ignition mechanisms. Every incident listed in Table 1 involved a regulator constructed mainly of aluminum.

But materials selection alone cannot guarantee oxygen regulator safety. It must be coupled with good design practice and proper use. Proper filter design and material selection is essential to mitigating the risk of ignition mechanisms like particle impact ignition, contaminant ignition and promoted ignition. Other design factors contributed to the fires, for example, the lack of protection around flammable components such as valve seals and springs. Finally, user error contributed to some fires, such as the use of multiple gaskets, hydrocarbon contamination, improper maintenance and other practices. However, good components should be designed and tested to safely tolerate reasonably foreseeable user error.

At the time of these fires, an ISO test method was available for assessing regulator vulnerability to heat of compression ignition, however, the majority of the ignition mechanisms that caused the fires in Table 1 were not heat of compression. The other ignition mechanisms were not addressed in the ISO standard. In fact, regulators that pass the ISO test were involved in the fires in Table 1, demonstrating that the existing standard was not adequate to ensure oxygen regulator fire safety.

Therefore, a standard was needed that accounted for all potential types of ignition mechanisms present under normal and reasonably foreseeable abnormal conditions (including user error). The objective was to have a test standard that ensured that regulators were both ignition-resistant and fault-tolerant, having a low probability of ignition and, should it occur, a low consequence of ignition.

Development of the Standard
The task group formed within ASTM Committee G04 involved industry, technical experts, users and regulatory agencies. The committee incorporated the existing ISO test for assessing ignition resistance to heat of compression (Phase 1) and undertook development of a new promoted (forced) ignition test for assessing fault tolerance (Phase 2).
The Phase 2 test subjects regulators to a clinically realistic and reproducible ignition event, simulating real-world conditions such as particle impact ignition, contamination and/or promoted ignition. The regulator is subjected to a single oxygen pressure shock similar to that used in the Phase 1 test that creates compression heating. The major difference is that in the Phase 2 test, an ignition pill is positioned at the regulator inlet where it ignites by heat of compression and promotes ignition of the regulator if the regulator is not fault tolerant. Fault tolerant regulators swallow the ignited pill (dissipate the heat without burning out), posing less of a hazard to users.

Because the ASTM standard closely represented the ignition mechanisms of particle impact, contaminants and promoted ignition, manufacturers began to redesign their regulators to defend against these mechanisms. The result was the new range of safe, ignition and fault tolerant regulators that are currently on the market.

ASTM standard G175 has gained wide acceptance throughout the United States in a relatively short period since its promulgation in 2003. In February 2007, the FDA proposed a new rule in the Federal Register that involves the development of a special controls guidance for medical oxygen regulators.

Conclusion
The effectiveness of ASTM standard G175 has clearly prevented any further devastating fires with conforming medical oxygen regulators. The use and applicability of the standard is increasing internationally and for other types of oxygen components. The standard has and will continue to ensure that rescue personnel and the people they serve are benefiting from safe medical oxygen equipment.

 

[Noah]

So I think the lesson here is, if you are using an aluminum medical regulator, make sure it conforms to ASTM G175. Older aluminum regulators, especially those manufactured before about 2007, should be especially scrutinized.

I know I bought my regulator on EBay many years ago without much thought about any safety risks - and it was unbelievably inexpensive. This makes me wonder if it was an old regulator being dumped because it did not conform to this new safety standard. You can bet I am going to verify that it conforms to ASTM G175 or it is going in the trash!

 

[agirard7a]

Thanks

Thanks Noah for the education. Your good at that!

 

[bret]

I thought the Respirators Firefighters use is filled with air (80% nitrogen-20% O2) like SCUBA? not pure Oxygen?

 

[rmartingt]

I thought the Respirators Firefighters use is filled with air (80% nitrogen-20% O2) like SCUBA? not pure Oxygen?

Firefighting SCBAs (that are used when fighting fires) simply use regular compressed air. But pure medical oxygen cylinders are carried for medical applications.

 

[Doggtyred]

In-Hospital use has been using vendor filled tanks with integral regulators for over a decade, after the powers that be required that medical oxygen be handled differently (serialized tanks, tracking, taking a tank to vacuum prior to refilling, etc) that it previously had been.

I would have expected that the separate tank and separate regulator assembly pictured in the first post would have been phased out of service with fire and EMS too, along with the station based transfilling equipment likely used to service it.

Horrible accident. I always knew we were to purge a cylinder before putting the regulator on it. Always thought it was to keep trash out of the regulator, not to keep it from torching me.