How accurate should airworthiness directives be? Before you answer that question, let me give you an example of an actual AD applicable to the Airbus A318/319/320/321 and then you can decide whether the information provided is sufficient for a mechanic to perform the required maintenance properly. I know you’re not all mechanics, but I don’t think you need to be one to see the problem. And if you fly or maintain GA aircraft, and think this issue doesn’t apply to you, think again. If the FAA doesn’t scrutinize ADs for large aircraft that carry millions of people around the world, just how much scrutiny do you think it’s giving ADs applicable to GA aircraft?
As you all know, airworthiness directives are issued only when a known safety problem exists and that problem is likely to exist or develop in other similar products. To quote the regulation exactly, FAR 39.3 states, “The FAA issues an airworthiness directive addressing a product when we find that: (a) an unsafe condition exists in the product; and (b) the condition is likely to exist or develop in other products of the same type design.”
This means that the safety problem isn’t a speculative one. It’s not a potential safety problem. A safety problem actually exists that the FAA determines must be addressed by a rule, which becomes mandatory for anyone operating the product to which the AD applies. ADs are federal aviation regulations, and mechanics performing maintenance are required to comply with the requirements of the ADs, usually by performing required inspections or repairs.
The AD for the Airbus models, 2004-13-06, requires repetitive inspections to detect cracks in the keel beam side panels and make repairs if necessary. Keel beams support the structural integrity of the fuselage. The AD was issued “to detect and correct fatigue cracks on the side panels of the keel beams, which could result in reduced structural integrity of the airplane. This action is intended to address the identified unsafe condition.” The point of all this is that ADs in general, and this one in particular, are issued to address serious unsafe conditions. Reduced structural integrity of an airplane is clearly a serious safety issue.
Since ADs address serious, unsafe conditions, it seems obvious that the maintenance procedures developed to communicate what a mechanic has to do need to be as clear and as accurate as possible so that a serious unsafe condition can be found and addressed before an accident or incident results, with injuries or deaths. There should be no argument on this, right?
Ambiguous and Incorrect Directions
The recent emergency revocation of two mechanics’ certificates (which they won on appeal to the full NTSB on timeliness grounds) involves their alleged failure to comply with this very AD. In brief, the FAA discovered the alleged violations in January 2010. Notwithstanding how serious the FAA claimed the violations to be in January 2013, the agency allowed the mechanics to continue to perform maintenance on dozens of airliners carrying hundreds of thousands of passengers for three years without recalling or re-inspecting a single aircraft. It’s almost incredible to think that, if the FAA believed the violations were so serious, it would have waited three years to bring an emergency revocation, knowing that the mechanics were continuing to work on airliners that whole time. And, what’s more, how could the FAA determine that the mechanics could no longer be trusted to perform maintenance properly in 2013 and yet never check a single one of the dozens of aircraft the mechanics worked on in those intervening three years between when the violations were allegedly discovered and when the emergency order was issued? I digress. Makes the IRS look competent by comparison.
Ultimately on appeal to the full NTSB, the mechanics prevailed; the full Board determined that the FAA’s action was untimely. The NTSB found that the mechanics had been prejudiced by the inordinate amount of time that had passed and were unable to properly defend themselves because of that lapse of time. (Note: I was an expert witness for the mechanics.) Because the NTSB dismissed the cases on timeliness grounds (“laches” in legalese), the full Board never addressed the issues regarding the validity of the AD on appeal. So here are the technical issues related to the accuracy of the AD the mechanics were supposed to comply with.
The AD incorporates by reference Airbus’s Service Bulletin, which requires, among other steps, removal of a pump and accumulator and a “detailed visual inspection of the area around the rivets” for cracks. The mechanics are alleged not to have removed the pump and accumulator and the FAA thereby claimed the required inspection could not be done without these components being removed.
So here is the first accuracy issue with the AD: it requires–via the Airbus Service Bulletin that’s incorporated by reference–that the pump be removed first and then the accumulator. Once the inspection is completed, it requires the pump to be re-installed followed by the accumulator. It doesn’t take a mechanic to figure out that there’s something wrong with these directions: when removal of parts is done sequentially, their re-installation is also done sequentially but in the reverse order. So if the pump were removed first, for example, it would normally be re-installed last. In this case, if the steps were followed as required by the AD, they could be accomplished but it would take significantly longer than removing the accumulator first and then the pump, and then reinstalling the pump and then the accumulator. According to the FAA, sequential steps in an AD must be followed sequentially even if the outcome, as in this case, is absurd.
But, regardless of the order in which these components are to be removed, the FAA contended that the detailed visual inspection of the “area around the rivets” for cracks required by the AD could not have been accomplished without removal of those components. Without going into whether an inspection of the area under those components could be done without their removal, the issue here is that there are no rivets (as the term is defined by Airbus in its manuals, which are the manuals that apply to this AD) under these components. If there are no rivets, then mechanics can’t inspect for cracks around them–with or without the components removed.
So there’s the second accuracy issue: while there are no rivets under these components, there are fasteners–Hi-Lok pin fasteners to be specific–and it could be that Airbus in its Service Bulletin meant for the mechanics to inspect for cracks in the area around the fasteners. Maybe that’s what the FAA meant when it issued the AD. Maybe not. Maybe the rivets are somewhere else on the keel beam. Maybe they’re under different components. But are mechanics supposed to guess? Are mechanics’ licenses subject to revocation if they fail to conclude that Airbus meant “fastener” when what was written was “rivet” even when Airbus’s own manual differentiates between fasteners and rivets? According to Airbus’s manual, all rivets are fasteners but not all fasteners are rivets.
It seems that if Airworthiness Directives are issued to address known unsafe conditions, mechanics are entitled to the most accurate paperwork possible. And they certainly shouldn’t be subject to revocation when the paperwork is inaccurate. According to the FAA’s deposition of an airline mechanic not involved in these enforcement actions, this mechanic submitted a request to the airline years before these violations allegedly occurred, requesting a correction of the job cards used to comply with the AD to change the sequence in which components were removed because the order was clearly incorrect. To date no change in the sequence has been made by the airline, Airbus or the FAA.
So while I’m not aware that the rivet versus fastener issue was raised to the FAA or the airline before these alleged violations occurred, it has been raised at least since the NTSB hearing in February this year. As I write, no change to the AD has been made.
System Failure at the FAA
If you’re thinking this is one case, with one inspector or one FAA office not doing their jobs, it’s not. The problem is all the way up the chain of command at the FAA. (I won’t begin to speculate on how this Service Bulletin escaped Airbus’s quality control.) First, you have the FAA incorporating by reference an Airbus Service Bulletin into an AD without apparently ever checking if the Service Bulletin makes sense. Then you have emergency orders based on this AD used to revoke airmen certificates without anyone checking the AD for accuracy. And those emergency orders are cleared through Flight Standards management in the field, Flight Standards and legal management in the region and the highest levels of Flight Standards and legal in Washington, D.C. So you really have two broken systems at the FAA: the issuance of airworthiness directives and the issuance of emergency orders.
If you believe the mechanics should have refused to sign off the work until they were handed correct maintenance procedures, I totally agree. But in the real world, that’s not happening. If mechanics did that, aviation would grind to a halt. That’s how common paperwork problems are in maintenance. (In fact, one union official told me recently that when mechanics insist on accurate paperwork to sign off maintenance, the airlines treat it as a job action.) This particular AD was issued in 2004 and there are hundreds of these aircraft in the U.S. fleet that have been subject to these repetitive inspections. I know other mechanics have noted the inaccuracies but just created their own workarounds. This clearly puts their certificates at risk if the FAA finds out. But the bigger question is: do the workarounds all address the unsafe condition the AD was meant to detect?
The FAA knows it’s a problem; the FAAst Team issued a bulletin in January 2011 titled Incorrect Maintenance Manual Procedures. The first line is: “How many times have you done a job and realized the maintenance manual was incomplete or incorrect?” Mechanics don’t typically report paperwork problems at least in part because so many maintenance procedures are incorrect and mechanics are used to developing their own workarounds to get the job done. The airlines know incorrect procedures are common and tacitly, if not explicitly, condone the workarounds in the name of on-time departures. The manufacturers know there are problems because the maintenance procedures are not fully vetted by mechanics before they’re distributed.
This AD is but one example, the tip of an iceberg. But fixing the problem of incorrect maintenance paperwork has to start somewhere. FAA? Airbus?