Although the public will not know all the details of the TAM 3054 crash until the official accident report is released, the basic facts are clear and suggestive. Unfortunately, what they suggest is that airport technology itself will not prevent accidents.
ANATOMY OF AN ACCIDENT
TAM 3054 was a short-haul flight from Porto Alegre to Congonhas Airport at Sao Paulo. Two experienced pilots in their mid-50s were in the cockpit of a twin-engine Airbus 320 aircraft, which is standard equipment in the fleets of many airlines.
On this particular airplane, though, the thrust reverser on the right-side engine was inoperative and had been deactivated before the flight, a fact known to the pilots. However, a single reverser is not regarded as a mission-critical component, and TAM therefore cleared the airplane for flight under ‘minimum equipment list’ (MEL) status.
The flight was trying to land on runway 35L at night, in rainy conditions. Indeed, Congonhas had experienced heavy rain for much of that afternoon, and the runway areas could easily have experienced drainage problems (which may or may not have been justifiable, depending on the condition of Congonhas’s drainage system). Again, air traffic control (ATC) had relayed weather reports to the pilots, who expressed some concern.
THE CONGONHAS CHALLENGE
Taking off and landing at Congonhas in the modern aviation era has always been challenging for two reasons. Because it opened in 1936, before the advent of jet propulsion, Congonhas has short runways: 35L is only 6,300ft long, and lacks the overrun area that is standard at modern airports.
Moreover, Congonhas is one of the few remaining airports located in the centre of a major city. Consequently, Congonhas lacks room for runway expansion, and surrounded by development, the airport’s runways offer no safety-valve flight path to aircraft in trouble. Indeed, the terrain just past 35L drops off sharply toward an area of public roads and buildings. Nevertheless, Congonhas was still the busiest airport in Brazil.
On top of their intrinsic external limitations, the runways presented additional challenges at the time of the accident. Responding to longstanding concerns about runway conditions, the runway had recently been resurfaced, but the new pavement had to cure (like the grout between shower tiles) before construction crews could groove it.
Lacking transverse grooves, the runway could have collected a thin sheet of standing water that could have caused tires to hydroplane; as a result, carriage braking can either become wholly ineffective or lead to a yawing skid.
In any event, the airplane’s flight recorder subsequently showed that the airplane touched down properly in all respects except one – the pilots had not set the right-side engine thrust lever to idle.
Under other circumstances, this would not have been a critical error, but on this occasion, the mistake caused a failure cascade:
- Upon touchdown, the right engine actually added forward power while the pilots were putting the left engine into reverse
- This caused the spoilers, which are automatically linked to engine configuration, not to deploy
- Without spoilers, which apply downward pressure on the airplane, insufficient ‘weight on wheels’ makes hydraulic braking through the tires difficult – even under good conditions
- With the lack of deceleration from spoiler failure and the yaw produced by asymmetric thrust, the plane, which was rapidly running out of runway, started to veer off 35L at high speed
- The pilots tried to execute a ‘go-around’ or re-take-off – which Airbus strongly warns pilots not to do in this situation because the mechanics of re-reversing the fully functional engine could cause a stall
- Even so, the airplane cleared the field and is at least airborne, but because of the surrounding congestion, the pilots don’t have enough space to regain altitude
- The airplane crashed into a building, with predictable results
Even this recitation might seem to be enough to determine cause. However, causality in breakdowns of complex systems is rarely as simple as it appears, for a variety of reasons.
For example, the lack of runway grooving would probably not have made a difference given the malfunctions of the vehicle’s systems. The pilot error was facilitated by sub-optimal human factors engineering on the A320, which has been involved in several similar incidents around the globe. Finally, TAM may have been a little too eager to avoid further delays by certifying an airplane under MEL allowances.
The definition of ‘mission critical’ depends on the mission, and some pilots have compared flying into Congonhas at night in the rain with landing on an aircraft carrier in rough seas. In short, everyone bears some of the responsibility, which implies that nobody should have to take all the blame.
THE AIR SAFETY RECORD
The TAM 3054 crash received extra scrutiny as it was bracketed by two other aviation accidents in Brazil within the past year.
In September 2006, Gol 1907, a Boeing 737, collided in midair with a private executive jet operated by its manufacturer, Embraer. According to the Aviation Safety Network accident report, a combination of pilot deviation in flight plan, delayed ATC response, radio communications difficulties, and an emergency transponder failure caused the head-on collision, which occurred over a remote section of the Amazon.
On 22 July, only five days after the TAM crash, a power failure shut down the ATC radar for over the hours at the Manaus centre, which controls airspace over the Amazon. The resulting disruption forced over 50 aircraft to divert, and produced massive aviation delays throughout Brazil. The radar blacked out when backup generators failed to operate after the main power feed failed.
Official inquiries into both incidents have been as disturbing as the incidents themselves. In the Gol crash, authorities took pains to blame the pilots, whose American citizenship was emphasised, while downplaying the contributing factors.
Regarding the radar blackout, various entities within the ATC institutional umbrella have produced a range of allegations, including (a) a power outage caused by the local utility, which denied the claim, (b) bungled routine maintenance on the backup generators, which wouldn’t explain the presumed primary failure, (c) a lack of batteries to start up the backup generators themselves, and (d) sabotage by disgruntled air controllers, who of course vehemently denied the charge.
The common element here has been finger pointing, as the Air Force, the controllers, and other entities have sought to offload responsibility onto others.
ORIGINAL SIN: CRACKS IN THE SYSTEM
Clearly, the overarching problem is not one of technology, nor even infrastructure, but of management. Institutionally, responsibility for civil aviation is divided between Infraero, which operates the major airports, ANAC, the new aviation agency created by President Lula da Silva last year and the Air Force, which still oversees air traffic control.
Notably, the Air Force has an antagonistic relationship with the air traffic controllers, who are underpaid and understaffed, and have complained publicly about underinvestment in ATC infrastructure.
Investment, however, may not be lacking so much as misdirected.
Brazil has long struggled with infrastructure development in general, and commercial transportation infrastructure in particular has been a primary driver of the notorious ‘Brazil cost’ (the country’s premium in operating expenses relative to those of other nations).
In its zeal to build tourism and commerce, moreover, Infraero has prioritised modernising terminals – in contrast, ANAC, which does not interact directly with passengers, is working with a budget one-third less than it requested.
Nevertheless, blaming infrastructure problems solely on management would be as simplistic as blaming the TAM accident solely on pilot error. Brazil’s aviation breakdowns also reflect national growth rates that would challenge the ability of any government to provide sufficient infrastructure in a timely manner.
Air traffic in Brazil has grown by at least 15% annually over the last three years; if it continues, this growth implies that traffic in 2010 will be double that of 2004.
To keep pace with surging demand, airport operators in particular face a difficult juggling act featuring many tradeoffs along a number of dimensions. For example:
- Urban, surburban and even exurban sprawl always catches up with airports, which often have to limit runway and airway usage because of neighbouring high-rise construction and residential complaints about noise
- Shutting down runways for renovation exacerbates congestion in the short term, and the resulting public irritation can induce airport operators to cut corners (such as not waiting for runway concrete to cure)
- When budgets are fixed, old infrastructure renovation competes for budget dollars with new infrastructure construction
- Authorities can stretch budgets by avoiding more expensive material, such as more absorptive asphalt that does not require grooving, but such tradeoffs are often penny wise and pound foolish
Certainly, technology can remedy almost every problem, but fixes are constrained in many cases by insuperable external barriers.
Regarding landing incidents, for example, crushable concrete (EMAS) and similar materials can reduce overruns the same way that gravel inclines corral trucks with brake failure, but this fix assumes extra land for runway extension. Aircraft tires can be ‘stickier’, but at the cost of more susceptibility to blowouts, which operators and fabricators regard as more dangerous.
Ultimately, the best way to prevent accidents is to focus less on minimising equipment error and more on minimising human error. At the operational level, this usually requires more people and more training, both of which are relatively simple, although not always financially easy.
At the management level, the problem is not so much cognitive error as conflicts of interest, self-protection and ego gratification, and corroded ethics. Unfortunately, there is no technological fix for any of these.