The Checkbook Is Not Balancing

What led to the China Airlines accident of a Boeing 737-800 last August? Bart Crotty speculates that as with any accident, there were many factors including a loose bolt, a service letter and a possible error in the re-assembling of some parts after inspection.

The disastrous China Airlines Boeing B737-800 hull-loss ground fire on August 20, 2007, where all 165 occupants amazingly evacuated without serious injury, strongly appears a case of operator maintenance and manufacturer personnel or organizational error. And this mishap also brings to light a long existing need to revitalize the management and performance of RII (required inspection item(s)).

The irony of some human error caused aviation accidents/disasters is that subsequent investigations reveal the causal actions were in fact well-meaning positive efforts being carried out with the sole intention of increasing safety. The China Airlines B737-800 hull-loss ground fire at Naha Airport, Okinawa, may be a prime example of just such a bizarre catastrophic happening. However, at this early stage in the official investigation the writer’s armchair analysis is just a "what if shot in the dark;" well, may not be too far off the bull’s-eye mark.

The Japanese investigating authority and several media sources have reported that most likely loose parts, in this case a "liberated" bolt from the right wing main slat track downstop assembly, pierced the surrounding wing fuel cell metal vertical wall (the "can"), a mere 2 mm (0.079 in) thick. The resulting fuel cell hole was reported as 23 mm (0.9 inch) wide and 41 mm (1.6 inches) long.

After a mid-morning uneventful landing and normal wing slat retraction and taxi into its gate position, the airplane was about to park at the terminal when airport service personnel saw fuel gushing out from the right wing leading edge and onto the apron. It then spread under the fuselage and ignited under the left engine and the fire quickly spread across the B737-800’s wings and engulfed the airplane. A full evacuation followed. Just before the airplane exploded, all 165 passengers and crew members had successfully evacuated.

The airplane, a relative young B737-800 (Taiwan registered B-18616) was delivered new to China Airlines in July 2002 and had accumulated about 13,500 hours of service time.

FAA’s ADs

Five days after the incident, the U.S. Federal Aviation Administration (FAA) issued an emergency airworthiness directive, AD-2007-18-51, dated August 25, 2007. This required an initial detailed visual inspection of the wing main slat track downstop assembly for missing parts, FOD (foreign object damage) and the correct bolt/nut torque, all due within 24 days and a repeating inspection every 3,000 flight hours. However, this AD in itself did not require or mention disassembly of the downstop’s eight parts to perform the inspection. It went on to caution about having sufficient light and possible use of supplement visual inspections aids. This AD did refer to Boeing Service Letter (BSL), B737 SL-57-084-B, revised July 10, 2007 sent to operators concerning two instances of similar slat loose parts problems but the AD said no more about the BSL. The BSL was originally issued in December 15, 2005. Soon after the fire incident, the contents and instructions of the BSL (not mandatory to accomplish) were not released to the general public. At the time of issue of this first AD, my suspicion was that the BSL may have advised operators or given them an option during the inspection to disassemble the downstop parts buildup.

Early reports said China Airlines had accomplished some sort of related inspection of the accident airplane and the downstop assembly bolt had been removed and reassembled. At this point in time of my armchair analysis I began to think of "excessive maintenance," doing more work than was necessary, i.e. as a possible case of not letting well enough alone — classic maintenance human or organizational error.

Three days later, the FAA issued AD-2007-18-52, which superseded AD-2007-18-51. This was issued after the FAA became aware of some additional similar B737NG (models -600 thru -900) incidents of similar FOD and/or loose parts. The new AD now required inspection no later than 10 days, not 24 days as before, but the checking of the correct torque remained 24 days. The 3,000 flight hour re-inspections also remained the same. Interestingly enough, optional borescope inspections were now mentioned. Additionally, this second AD went to lengths saying that disassembly of the downstop assembly is not required and doing so might cause a maintenance error. Missing from the second AD is any mention of the BSL originally mentioned in the first AD. A call to the FAA Seattle, Wash., Large Aircraft Certification Office revealed the BSL did contain advice on conducting an inspection but the FAA office did not disclose specifics of the matter. A few weeks later, the BSL revealed that it did not, repeat did not, direct or suggest disassembly of the downstop assembly as originally suspected and this would further indicate that China Airlines did so on their own initiative.

If China Airlines, as reported in the media, did carry out the BSL subject inspection in the weeks before the accident, then let’s suppose the airline or an individual maintenance worker might have found it easier to accomplish the visual inspection task by disassembling the downstop assembly of eight parts. Then, during the reassembly, the error chance of misassembling (sequence of parts or loosing some part) of the downstop assembly of parts buildup would be an added risk on top of doing a proper initial inspection. The irony here is that quite probably, during manufacture of new B737NGs, this same error-prone situation of assembling and installing of the eight parts was originally committed at random times by Boeing production workers and now the situation is repeating itself in the field by airline maintenance personnel; a human factors case of déjà vu.

Neither AD said at how many main slat track stop assembly there are per airplane and if there is more than one downstop assembly per wing, which one required the inspection. It could well be that the working space in the subject slat area is very limited, and further considering the relatively small parts size, both factors would make for a tedious reassembly of the eight parts, not a job to be easily and quickly accomplished. The idiom comes to mind of one being or a job being "all thumbs" meaning a clumsy person or a task challenging a person’s fine hand skill abilities.

What May Have Happened

My guess is that the China Airlines maintenance worker was doing the BSL inspection job for the first time on the accident aircraft, had correctly or incorrectly disassembled the assembly of eight parts stack up and made an error of some sort in the parts re-assembling. Further, if China Airlines required and performed a double inspection to check the satisfactory work of the first worker, then it too was apparently not carried out correctly. If so, as I continue to armchair theorize, we have in play here a chain of two successive maintenance errors, a double irony, two failed maintenance actions that were acutely intended and directed to solely enhance safety.

The only upside it seems at this point to this disaster is there were no fatalities or serious injuries of the 157 passengers or eight crew members on board the airplane.

Three remaining begging questions are: (1) were there a number of new delivered B737NGs, i.e. -600 thru -900 series, airplanes assembled by Boeing with latent parts errors of the slat track downstop assembly, and (2) if and when operators performed routine inspections (not the BSL or AD inspections) or maintenance did they commit errors in re-assembling the downstop assembly parts buildup?

The answer to question No. 1 is: Further media reports have indicated that some operators of applicable B737NGs have reported nearly 21 new reports of loose or missing parts and particularly so after the issue of the ADs. Since the BSL was first issued in December 15, 2005, it may be a case of several airlines not having accomplished the BSL suggested inspection or now via the new ADs, going back and more intently re-inspecting the downstop and finding earlier missed problems.

The answer to question No. 2, did airlines induce problems during regular inspection or maintenance, may never be known as most airlines would not want to publicly admit to mistakes.

A final nagging No. 3 question, possibly not applicable here depending on China Airlines’s inspection policy/standards, but very disturbing industry-wide, is why relatively so many RII/double/duplicate inspections (know under different names) are botched and fail to catch possible maintenance/inspection errors they are specifically called on to uncover.

In this regard, during the past 10 years I can think of several resulting U.S. fatal accidents and foreign operator fatal accidents. For those readers not knowledgeable, certain maintenance safety critical job/tasks are self-determined and listed by individual airlines in their manuals to require a double/duplicate inspection, the U.S. industry used term is required inspection item (RII.) These second inspections (the first inspection being inherent with the worker doing the subject job/task requiring a RII) are performed by certain airline authorized and qualified maintenance personnel who cannot be involved in doing or being part of the original job/task. As infrequent random errors will occur in maintenance, the RII inspections of highly safety critical jobs/task (e.g. replacing engines or control surfaces or flight control components, major repairs/alterations, performing AD related work, etc.), are to ensure that any faulty or poor work is uncovered and therefore does not cause or lead to an accident or serious incident.

What Has Become of the RII?

In my much younger airline employment days as an FAA certificate mechanic, the on-job occasion or call for an RII inspection definitely raised the level of attention and job performance of the work force. And those few inspectors or foremen given RII inspector authority were looked upon with much respect. Nowadays, the whole situation of the purpose of RIIs, its management and effectiveness in my opinion has deteriorated considerably. U.S. air carriers, by regulation, must have a written policy, training and procedures to manage their RII program but few airlines do a really good job of it and FAA inspectors don’t challenge or encourage the airlines to improve this situation.

It seems that the FAR 121 and 135 RII regulation requirements, which actually cover more specific regulation print space details than any other maintenance/inspection areas/items therein, have lost their teeth in the recent years; how can this be? I will continue to lobby the FAA and industry and hopefully help correct the deteriorated RII situation.

About the writer: