Aircraft Accidents and Aviation Lessons Unlearned LIV: Southwest Airlines flight 812

Southwest Airlines flight 812, aircraft N632SW

On April 1, 2011, around 15:58 (3:58 PM) Mountain Standard Time, Southwest Airlines flight 812 (SWA812) experienced a rapid decompression during climb out at a flight level of 34,000 feet. The flight diverted to Yuma International Airport (NYL) in Yuma, Arizona. The aircraft, registration number N632SW, a Boeing 737-3H4 (-300 series), serial number (S/N) 27707, landed safely. The National Transportation Safety Board (NTSB) assigned the event accident number DCA11MA039; the Accident Brief – NOTE: not Report – AAB-13/02, was adopted on September 24, 2013.

The NTSB determined, “… the Probable Cause of this accident was the improper installation of the fuselage crown skin at the S-4L lap joint during the manufacturing process, which resulted in multiple site damage fatigue cracking and eventual failure of the lower skin panel.” The cause of the accident was attributed to a rapid decompression because fasteners and a skin panel at the crown were improperly installed by the manufacturer, resulting in cracking and eventual failure. The NTSB was correct in stating that a structural problem may have been built into the panel. However, clearly the blame for preventing the panel anomaly was Southwest’s inspection abilities and its maintenance program – specifically the structural inspection task cards – not a Boeing manufacturing error.

There were two problems with SWA812: For one, there were no fatalities. As indelicate (cynical?) as that sounds, nobody would notice an aviation event unless there were more horrifying consequences. This is human nature; after all, why be concerned if there was no threat of death. The NTSB decided an Aircraft Accident Brief (AAB) was adequate, not an Aircraft Accident Report (AAR); they did not feel that the SWA812 deserved anything more. This NTSB error led to the second problem, which was ignoring the fact this was not the first time this happened, albeit not on the scale of the previous event/accident.

On April 28, 1988, Aloha Airlines flight 243 (Aloha243), a Boeing 737-297 (-200 series), S/N 20209, suffered rapid decompression and a catastrophic structural failure of the crown from the forward bulkhead at Body station (BS) 360 to the manufacturer’s splice at BS 540, just forward of the wings; from the left-side floor to the right-side floor. S/N 20209, became infamous when the disfigured airliner’s pictures were splashed all over the media for weeks. Everybody knew. Aloha243 also suffered a fatality; a flight attendant was killed when the crown separated; she was pulled out of the aircraft and lost at sea near the Hawaiian Islands.

Why was the Aloha243 accident significant to the SWA812 event? NOTE: Serial numbers will be used as opposed to registration numbers, which can change when an aircraft is sold from operator to operator. First point: S/N 20209 was a 737-200 series; S/N 27707 was the next series, the -300 series, designed, approved and flying before Aloha243 occurred. Both aircraft were also Maintenance Steering Group (MSG) -2 certificated.

What made the dismissal of SWA812’s importance more egregious were the number of flight hours and flight cycles accumulated. A flight hour is measured from engine start to engine shut down or how many hours the aircraft is operated. A flight cycle measures how many flights an aircraft takes, from wheels off to wheels on the ground. An aircraft may fly five flight hours in one flight, but that is only considered one cycle. A cycle records how many times the aircraft is pressurized and depressurized, which places stresses on the aircraft skin and structural members, how often it expands and contracts. The Boeing 737 was originally designed as a short-range aircraft, resulting in a closer number of flight cycles to flight hours and that was an important issue that led to the Aloha243 accident.

At the time of its accident, S/N 20209 had 35,496 flight hours and 89,680 flight cycles for an average of 2.5 hours per flight; it had been operated in salt air conditions (Hawaii), which contributed to the accident. By contrast, S/N 27707 had 48,748 flight hours and 39,786 flight cycles, for an average of 1.2 hours per flight; it was owned exclusively by Southwest Airlines for fifteen years before the event. What both Aloha Airlines and Southwest Airlines experienced were failures in the maintenance inspection program; the NTSB caught this with Aloha243, but missed it completely with SWA812, twenty-three years later.

The NTSB website was referenced successfully for docket information, specifically the Maintenance investigation notes and ten attachments, including the NTSB Maintenance Group Chairman (MGC) Factual Report. A review of these documents confirmed that SWA812 was not a Structures accident, as the NTSB determined, but a Maintenance accident. The Structures Group could have been playing a supporting role in damage analysis, but this accident was not due to structural engineering and had less to do with manufacturer culpability. To record this event in a Brief demonstrated that the NTSB did not just miss the target, they missed the broad side of the barn the target was hanging on. It must be asked: Did the NTSB Board Members actually read the Accident Brief before adopting it?

AAB-13/02 used five of the Brief’s fifteen pages to elaborate on the structural testing accomplished on the failed panel, surrounding structure and fasteners; ‘good-to-know’ information that failed to address the accident’s Root Cause. This raised the question: At what point, e.g., flight hours, flight cycles, years, was an operator culpable for failing at its Inspection program? The Brief diverted attention from actual root cause to irrelevant issues. This made the industry less safe; nothing was learned.

Those five pages glorifying structural testing brings to mind one scene from the 1992 movie, My Cousin Vinny; the Prosecutor’s expert witness bragged about his tire testing equipment, “I have a dual-column gas chromatograph, Hewlett-Packard model 5710a with flame analyzing detectors.” In the scene, the illusion worked; the jury, unfamiliar with automotive jargon, were dazzled. However, the expert witness never answered the simple question: Did all the impressive testing information prove the case? No, it did not; it was a distraction … just like with the SWA812 Brief.

The NTSB has used distraction before. In the April 2013 National Air Cargo B747 accident where a military all-terrain vehicle (M-ATV) moved aft during takeoff, the NTSB’s Structures investigator’s report showed numerous color pictures of the M-ATV’s pallet, particularly the underside, where red paint scrapings scored the M-ATV’s pallet. Everybody knew the M-ATV pallet moved aft; the paint scrapes proved nothing. Worse, the investigator never answered the basic question: Why did the M-ATV pallet move aft? Ironically, Boeing answered in the report that the M-ATV pallet’s weight exceeded the floor’s structural strength. The pallet’s weight, exaggerated upon landing in Bagram Air Base, broke the floor; there was nothing to anchor the pallet in place; the pallet was free to slide aft. That was the answer: clear, simple, factual. The cargo floor failed on its previous landing – period! The M-ATV pallet slid aft when the aircraft rotated, nose up. Paint scrape pictures were irrelevant.

Just like National Air Cargo, the SWA812 Brief needed common sense and attention to analysis, not worthless technical jargon. The report should have focused on Maintenance and Inspection. The MGC did come close to Southwest’s Inspection problems in his Factual Report, but either his data was dismissed, or he did not understand it. In AAB-13/02, page five, was this throw-away statement, “… the Southwest Airlines maintenance records for the accident airplane were examined and contained no evidence of any major repairs or alterations performed on the accident crown skin or side skin panels.” Major Repairs? Alterations? That’s it? What about Inspections? Was the Inspection schedule given more than a hurried glance? The MGC identified himself as an Aerospace Engineer, which meant he possessed ZERO skills in various inspection techniques and how to follow the dictates of an Inspection program. The MGC did identify inspections conducted on S/N 27707 over fifteen years. On pages four through seven of the MGC’s 18-page Factual Report, the MGC referred to Southwest Airline’s Maintenance Inspection Program, recorded its heavy structural inspection intervals and dates they were accomplished, but the SWA812 Inspector-in-Charge failed to include any of this information in the Accident Brief.

Per AAB-13/02, “The fracture extended between BS 666 and BS 725 and through the lower row of rivets of the lap joint, intersecting 58 consecutive rivet holes at approximately 1-inch intervals.” The crown section, per Figure 3, was between BS 360 and BS 908; the left and right limits were between Stringer 14-Left over the top to Stringer 14-Right. As per the MGC’s Factual, this area had received several general visual inspections, which were limited by paint and primer not being removed.

The Southwest maintenance program for the B737-300 series required heavy inspection checks during S/N 27707’s lifetime leading up to the event. The original Maintenance program had upgraded from MSG-2 to MSG-3 in 2004, so the Maintenance program was improved for this model B737. The MGC’s Factual showed S/N 27707 had undergone several ‘C’ Check phase inspections and ten ‘Y’ inspections since 2004. During this time, there were more involved inspections marked ‘INSP’; the MGC did not document how detailed the INSP inspections were nor how much access to the failed panel area was exposed. This was crucial information to understanding the integrity of the Southwest Structural Inspection Program. Why did the NTSB not know about inspections or why dismiss this information?

It was relevant to these points that S/N 27707 underwent a Non-Destructive Inspection (NDI), most likely Eddy Current. The NDI was performed at the location of the failed skin panel on February 2, 2011, fifty-eight days prior to the SWA812 inflight event. The MGC did not specify what type of NDT was used nor did he investigate the NDT’s findings, which is why AEROSPACE … ENGINEERS … SHOULD … NOT … BE … LOOKING … AT … MAINTENANCE … ISSUES!

The ongoing foolishness, where unqualified NTSB engineers keep missing maintenance issues has been documented in almost every maintenance-related NTSB accident report reviewed on this website. The NTSB continuously avoids employing industry-experienced airframe and powerplant FAA-certificated technicians as Lead Maintenance Investigators; this guarantees that major investigation mistakes will persist and maintenance issues will not be corrected.

The whole purpose of an aircraft accident investigation, no matter how involved, is improved aviation safety; the industry benefits, lives are preserved. When I was the sole NTSB Maintenance Major Accident investigator, I would talk frequently with the Federal Aviation Administration (FAA) investigators I worked with. If the NTSB dragged their feet on maintenance issues, the FAA investigators would raise the issue(s) to FAA’s upper management – safety was improved. Former Member John Goglia, the only mechanic Board Member, was often frustrated by the NTSB’s inaction on maintenance issues; he would walk across the street and speak with FAA management himself – safety was improved. Industry knows investigations are error-filled; they also take the initiative. Boeing would have addressed National Air Cargo’s floor collapse to guarantee it did not repeat – safety was improved. Southwest and its FAA certificate office would have corrected the B737-300 inspection program – safety was improved.

It would be small consolation if the SWA812 Brief’s Probable Cause was close, but it was not. It could be argued that the structures investigator was right about the metal analysis; that might be true. The truth is, in accident investigation, there is a large difference between “being right” and “getting it right”; Accident Brief AAB-13/02 for SWA812 was neither.  

SWA812 faded from a lack of attention; no one saw it as a big problem; no one analyzed Root Cause; no one related it to Aloha243; no one at the NTSB felt – still feels – that Maintenance issues deserve the careful attention they deserve. In thirty-four years of existence – minus the time John Goglia and I were there – the NTSB still ignores the fact that over half the FAA workforce deals strictly with Maintenance; it is that important. Until the NTSB hires qualified mechanics, my Aviation Lessons Unlearned website will – unfortunately – have plenty of monthly accident reviews.

Leave a Reply

Your email address will not be published. Required fields are marked *