You may be trying to access this site from a secured browser on the server. Please enable scripts and reload this page.
Turn on more accessible mode
Turn off more accessible mode
Skip Ribbon Commands
Skip to main content
Top Link Bar
NEWS & EVENTS
Speeches & Testimony
Most Wanted List
The Investigative Process
Data & Stats
General Aviation Safety
Assistance to Families & Victims
Operations & Policy
Administrative Law Judges
Strategic Plans & Reports
Safety Recommendation Details
The Investigative Process
Data & Stats
General Aviation Safety
On February 7, 2006, about 2359 eastern standard time, United Parcel Service Company (UPS) flight 1307, a McDonnell Douglas DC-8-71F, N748UP, landed at its destination airport, Philadelphia International Airport (PHL), Philadelphia, Pennsylvania, after a cargo smoke indication in the cockpit. The captain, first officer, and flight engineer evacuated the airplane after landing. The flight crewmembers sustained minor injuries, and the airplane and most of the cargo were destroyed by fire after landing. The scheduled cargo flight was operating under the provisions of 14 Code of Federal Regulations (CFR) Part 121 on an instrument flight rules flight plan. Night visual conditions prevailed at the time of the accident.
TO THE FEDERAL AVIATION ADMINISTRATION: Ensure that the performance requirements for smoke and fire detection systems account for the effects of cargo and cargo containers on airflow around the detection sensors and on the containment of smoke from a fire inside a container and should establish standardized methods of demonstrating compliance with those requirements. (Superseded by safety recommendation A-12-68)
Original recommendation transmittal letter:
Closed - Acceptable Action/Superseded
Philadelphia, PA, United States
In-Flight Cargo Fire, United Parcel Service Company Flight 1307, McDonnell Douglas DC-8-71F, N748UP
Addressee(s) and Addressee Status:
FAA (Closed - Acceptable Action/Superseded)
Safety Recommendation History
From the transmittal letter (greensheet) issuing safety recommendations A-12-68 through A-12-70: The NTSB used two types of containers to evaluate the time it takes to detect fires that originate from within those containers. The NTSB tested each type of container twice. The tests revealed that the elapsed time between fire initiation and detection ranged from 2 minutes 30 seconds to 18 minutes 30 seconds, all exceeding the 1 minute detection time required under current regulations. From the tests conducted in the study,11 the fires grew very large, capable of causing significant damage to an aircraft, shortly after becoming a detectable fire (see figure 1). Based on the test data, the NTSB is concerned that when fires inside containers become detectable to the aircraft’s smoke detection system, there is little time until the fires reach levels that can compromise the integrity of the cargo compartment and then threaten the structure and systems of the aircraft. The presence of containers and pallets prevent fires that start within them from being detected within 1 minute per 14 CFR 25.858, reducing the time to react. On December 17, 2007, as a result of the UPS flight 1307 accident, the NTSB issued Safety Recommendation A-07-98, which asked the FAA to do the following: Ensure that the performance requirements for smoke and fire detection systems account for the effects of cargo and cargo containers on airflow around the detection sensors and on the containment of smoke from a fire inside a container and should establish standardized methods of demonstrating compliance with those requirements. On September 6, 2010, the FAA responded that a test program had been completed12 using a Boeing model 727 main deck cargo compartment to evaluate the effect of cargo containers on airflow around the smoke detectors. The FAA reported that although there was some variability in the test results, in general, smoke detectors sounded quicker in loaded compartments than in empty compartments. However, the FAA study addressed the effect of cargo containers on the airflow around the smoke detectors for only one type of aircraft configuration and did not address the issue of overall detection time of the fire, specifically with the effect of smoke confinement within the containers. NTSB’s safety recommendation A-07-98 is currently classified “Open—Acceptable Response.” Based on the circumstances of the UPS flight 6 accident and the NTSB cargo container fire study, the focus of the delayed detection issue has shifted from the effect of airflow on smoke to reach detectors to the effect of smoke confined within a container or pallet to reach detectors. The NTSB study established that there could be a long time during which a fire originating inside a container produces smoke that does not exit the container. The NTSB study also found that due to this delay caused by the container concealing the smoke, once a fire becomes detectable to the aircraft’s smoke detectors, it is not long until it burns through the container and becomes a substantial threat. The NTSB concludes that if the fire were to be detected while generating smoke inside the container, valuable time would be gained for alerting flight crews and mitigating the effects of the fire. Accordingly, the NTSB reclassifies Safety Recommendation A-07-98 CLOSED—ACCEPTABLE ACTION/ SUPERSEDED, and recommends that the FAA develop fire detection system performance requirements for the early detection of fires originating within cargo containers and pallets and, once developed, implement the new requirements.
The FAA indicated (1) that the smoke detection time of an empty cargo compartment is generally longer than that of a loaded compartment and (2) that these findings are consistent with the current methodology used for certification testing. The NTSB believes that the FAA’s research and analysis of the effects of cargo and cargo containers on airflow around smoke and fire detection systems address the first portion of this recommendation. However, we remain concerned that the performance requirements for smoke and fire detection systems still do not take into account the effects of cargo and cargo containers on the containment of smoke from a fire inside a container. We would like to point out that our ongoing investigation of the UPS accident in Dubai suggests that, in that accident, a delay in smoke detection occurred due to smoke containment in the collapsible shipping containers. We will discuss this recommendation further in our final accident report; in the interim, Safety Recommendation A-07-98 remains classified OPEN – ACCEPTABLE RESPONSE.
CC# 201000359: - From J. Randolph Babbitt, Administrator: The Federal Aviation Administration recently conducted tests to evaluate whether smoke detection times are affected when the same quantity of smoke is released in an empty and a fully-loaded cargo compartment. Current smoke detection testing uses a smoke generator to simulate smoke from a small fire. Per existing FAA guidance, applicants typically perform this test in the most critical condition, with the cargo compartment smoke source located farthest from any smoke detector, to show that the smoke detection system meets the required response time. We conducted these tests in the main deck cargo compartment of a Boeing Model 727 series airplane and in the aft, below-floor cargo compartment of a Boeing Model 747SP series airplane. Although there was some variability in the test results, in general, the smoke detectors sounded quicker in loaded compartments than in empty compartments. We reported the smoke detection test results in DOT/FANAR-09/52, Effects of Cargo Loading and Active Containers on Aircraft Cargo Compartment Smoke Detection Times, published December 2009 (enclosure 1).
The FAA replied that it would investigate the effects of cargo and cargo containers on airflow around smoke and fire detection sensors by conducting tests at the FAA’s William J. Hughes Technical Center and by using a computer model of computational fluid dynamics. In addition, the FAA stated that it would review the technical standards order (TSO) for cargo containers with regard to containment of smoke from a fire inside a cargo container. Pending completion of the research and analysis of the effects of cargo and cargo containers on airflow around smoke and fire detection systems and incorporation of appropriate revisions to the TSO based on this research and analysis, Safety Recommendation A-97-98 is classified OPEN -- ACCEPTABLE RESPONSE.
Letter Mail Controlled 9/18/2008 12:11:57 PM MC# 2080573: - From Robert A. Sturgell, Acting Administrator: FAA regulations §§25.855, 25.857, and 25.858 require the presence of smoke detectors in class B, C, and E cargo compartments. The FAA currently has no performance criteria addressing concerns about loading of the cargo and its effects on airflow around the smoke detectors. Historically, the FAA permitted the smoke detection testing to be carried out in empty cargo compartments with normal ventilation conditions because an empty cargo compartment would represent the maximum amount of mixing and dilution of the smoke; hence a worst case scenario. We had no data indicating that the presence of cargo would represent a more challenging smoke detection scenario. We plan to investigate the effects of cargo and cargo containers on airflow around the detection sensors. We intend to conduct tests at our Technical Center and to utilize a Smoke Transport Computational Fluid Dynamic code to assess the effects of airflow on smoke detection. We will also review the current Technical Standard Order requirements for cargo containers to determine what design changes may be practical with regard to containment of smoke from a fire inside a cargo container. I will provide a status update by January 2009. I will keep the Board informed of the FAA’s progress on this safety recommendation.
Strategic Plan, Performance & Accountability Reports & More
Directions to Conference Center
Web Policies & Notices
Annual Review of Aircraft