View Full Version : Histories of Oxygen Generators

Data Guy
30th Mar 2011, 08:03
Source; NTSB report and FAA Tech Center Tests -- Aug 1986 to April 2003.
NTSB ValuJet Report NTSB/AAR-97-06 adopted Aug 19, 1997 regarding the loss of Valujet flight 592 on May 11,1996. Briefs Only; see full NTSB report, at Link > http://www.ntsb.gov/publictn/1997/aar9706.pdf
7 Incidents Involving Chemical Oxygen Generators. Ref pgs 93-96 (Ed. Note; Only 1and 2 were fires).
Incident 1. 8/10/86. Aircraft destroyed at the gate. NTSB cause; the fire was initiated by improper handling and that the generator had not been packaged properly. Generator was a Scott P/N 801836-06C - ref Appendix P, Hazardous Material Factual Report, Exhibit # 15-Q, pg 000444. (Ed. Note; the same model shipped on ValuJet 592 (ref NTSB Final Report-ValuJet; page 14, “Scott P/N 801386-06, three mask”.
Incident 2. 2/19/88. In-flight fire by a rupture (3 inch hole) of a Duo-Pac Avoix generator caused by blockage of the output tube. NTSB recommended relocation of the relief valve directly to the generator body. Ref to NTSB Safety Letter and Link > http://www.ntsb.gov/recs/letters/1989/a89_106_107.pdf
Incident 3. 11/7/92.Cargo facility fire. Improper labeling. No claim generator initiated fire.
Incident 4. 9/24/93. Cargo facility fire. Improper labeling. No claim generator initiated fire.
Incident 5. 10/21/94. Cargo facility fire. Improper labeling. No claim generator initiated fire.
Incident 6. 1/26/96. Cargo facility. Improper labeling. No claim to a fire.
Incident 7. 4/25/96. Repair station. Improper labeling. No claim to a fire.
> Ref pg 54, Five Fire Tests, “Test Conditions and Results”. No fire with tests 1 and 3. Fire with tests 2, 4 and 5 on the fourth try.. (Ed. Note; not known if outlets were plugged as in the May 1999 DOT/FAA Tech Center tests below.
> Ref pg 133, “Conclusions; # 6, The activation of one or more chemical oxygen generators in the forward cargo compartment of the airplane initiated the fire on ValuJet flight 592. Conclusion # 21; Because of the rapid propagation of the oxygen fed fire …..”.
> Ref Appendix P, Hazardous Material Factual Report, Exhibit # 15-Q, pg 000446; “test data indicate that the outside surface of Scott oxygen generator canisters may reach 420 to 443 degrees F during production of oxygen”.
May 1999. DOT/FAA/AR-TN99/9. By O’Connor and Hagen. Activation of Oxygen Generators in Proximity to Combustible Materials. Link > http://www.fire.tc.faa.gov/pdf/tn99-9.pdf
Abstract; This report presents the results of a series of tests performed on oxygen generators contained in cardboard shipping containers and packing materials to witness the probability of ignition in the event one of the generators was activated. Test results indicated that in the presence of an activated generator, combustible materials will produce a fire. Since it is conceivable that the sodium chlorate in a inactivated generator could melt and plug all outlets, the outlet and pressure relief valves were plugged before activating the generator to simulate this melting situation. The activation of an aircraft oxygen generator canister produced temperatures of 410 F). No fire in 3 of 7 tests.
April 2003. DOT/FAA/AR-TN03/35. By David Blake. The Response of Aircraft Oxygen Generators Exposed to Elevated Temperatures. Link > http://www.fire.tc.faa.gov/pdf/TN03-35.pdf
Abstract; The purpose of this testing was to determine the temperatures that would cause self-activation of sodium chlorate oxygen generators. The data will be used to establish the degree of thermal protection that would be required to prevent the activation of chemical oxygen generators should they be exposed to heat from a cargo compartment fire involving other materials. The minimum temperature that caused the activation of one of the generators was 600°F. Due to uncertainties with other designs not tested and the physical properties of sodium chlorate, it is recommended that the generators not be exposed to temperatures above 400°F.
End -----------

Data Guy
31st Mar 2011, 11:11
One other historical document not mentioned in my prior post was a (Danish) AAIB DENM Report 2/96. I had left it out because, unlike the other references with full URL source links, this one is not available. Google searches and a query to the Danish web site and to the Danes have failed to find a URL. And although this 1993 MD-87 incident does have passing references in two FAA Tech Reports and in the TSB’s Swissair flight 111 report, those references only address the newly realized flammability issues of Mylar acoustic insulations which saw AD action 12 years later for removal of all metalized Mylar (MPET) by 2005.

TSB reference Link > > Transportation Safety Board of Canada - AVIATION REPORTS - 1998 - A98H0003 (http://www.tsb.gc.ca/eng/rapports-reports/aviation/1998/a98h0003/01report/01factual/rep1_14_12.asp)

AD reference Link > McDonnell Douglas Model DC-9-80 and MD-90-30 Series (http://www.airweb.faa.gov/Regulatory_and_Guidance_Library/rgAD.nsf/0/7157A2905D2A3E7286256A0800694EA9?OpenDocument&Highlight=blankets)

On Nov 24, 1993, SAS MD-87, Reg # SE-DIB incurred a severe fire over the right rear lavatory. Because O2 generators were consumed, the AAIB asked the FAA Tech Center to test these for any possible effects on that fire. From that May 4, 1994 Tech Center response, the AAIB Report ( concluded that “the generators had only a minor effect on the fire in SE-DIB”. Of the tests, the Tech Center said, “Two tests were conducted using solid oxygen generators. The first test utilized a generator mounted above a pan of jet fuel fire. Over-heating from the fuel fire caused the generator to activate. Intensification of the fuel fire was limited to the immediate area of the discharging oxygen.

In the second test the oxygen generator was placed in a cardboard box containing shredded newspaper. After the box was ignited and burned for a period of time, the generator activated due to the surrounding fire. Again, intensification of the fire was contained to a small local hot spot”.

“Based on these tests, it was concluded that the involvement of a discharged oxygen generator had minimal effect in the SAS MD-87 fire”.

Data Guy
31st Mar 2011, 11:17
To my "More History" post above the TSB and the AD Links didn't come through.

Perhaps the moderator can patch them back in