Could a $10-million system have saved lives in Kozhikode plane crash?

While most plane crash investigations blame dead pilots, Kozhikode investigators may have to dig much deeper

On a rainy August night last year, the Air India Express flight IX-1344, on a mission for repatriation of stranded Indian nationals in West Asia, overshot the tabletop runway at Kozhikode airport in Kerala. The plane fell a few feet below in the gorge and broke into two, killing 20 people, including both pilots. In many ways, it was an uncomplicated investigation for the Aircraft Accident Investigation Bureau (AAIB), as its black box containing the data flight recorder was salvaged almost immediately without any damage.
 
The delayed investigation
 
The preliminary report should have been out within 30 days, according to the International Civil Aviation Organisation (ICAO) guidelines. But little is known publicly about it as of date. The January 2021 deadline for submitting the final report, as also the two-month extension given by the civil aviation ministry to the five-member investigation team, is over. Now, the investigators are racing against time to piece together a report before August 7, the first anniversary of the horrific crash. Even the Mangalore air crash of 2010, one the deadliest in India’s aviation history and similar in nature to Kozhikode, was investigated within months of the disaster. One of the members of the investigating team told Business Standard that the delays were “due to Covid-related restrictions and the final report will be out soon”.

 

Meanwhile, a public interest litigation (PIL) taken up by the Supreme Court in January this year could make the matters more complicated as investigators seek an answer to the most crucial question in any air crash probe: what caused the plane to crash? While most investigations in the past have blamed dead pilots for errors in judgement, Kozhikode investigators may have to examine past bureaucratic developments that could have potentially saved not just the lives but also the aircraft whose fuselage broke into two on impact at Kozhikode. The Ministry of Civil Aviation (MoCA), Directorate General of Civil Aviation (DGCA) and Airports Authority of India (AAI) did not respond to emailed questions on the matter sent on June 22-23. Their responses and inputs will be published when received. 

 
A 2007 plan shelved
 
In 2007, the AAI was looking for a solution to prevent aircraft from overshooting runways at Kozhikode and zeroed in on a system that was first installed at New York’s JFK airport in 1996. While similar plans were afoot for Mangaluru, too, it was decided that the system should be installed at Kozhikode first. A contract for this was drafted in 2008. But the project was subsequently put on the back burner. “I had several meetings with the AAI and also met its then chairman. They barely had the money to pay staff salaries. They were in financial trouble. The global financial crisis of 2009 made matters only worse,” said the petitioner in the Supreme Court case, Rajen Mehta, now 85. He was the India representative of the American company that was manufacturing the system at the time. His PIL was taken up by the SC earlier this year with the court asking the MoCA to consult with Mehta on the issue. A team of ministry officials subsequently met Mehta and took his inputs. The ministry will be submitting its post-consultation report on the matter to the court.

 
However, with the 2010 Mangalore crash – in which 158 of the 166 passengers and crew members died as the Air India Express plane overshot the runway and slid into the gorge below – civil aviation ministry bureaucrats’ interest in the system piqued again. Within three months of the Mangalore crash, the arrestor system’s manufacturers were again asked to send a draft proposal for the Kozhikode airport.
 
The arrestor system in question
 
Officials were interested in installing an Engineered Materials Arresting System (EMAS), where cement blocks were to be placed at the end of runways (called the runway end safety area, or RESA, in aviation parlance) to prevent overshooting planes from plunging down the tabletop runway at Kozhikode. These specially-engineered blocks were to be made of cellulite cement and crush under the weight of a speeding aircraft to maximise its deceleration. Such a system is usually deployed at airports where runway space is limited and the required RESA cannot be maintained in line with ICAO norms. The tyres of a speeding aircraft that has suffered a braking failure upon landing or is at the risk of overshooting due to an unusual landing would sink into the cellulite cement material, radically slowing down the plane to minimise injuries and damage to the plane. In the US, these installations are wider than the runway and provide for steps and a road on either side to allow the exit of passengers after an overshooting plane has been stopped.

 
The system was a result of research initiated by the US Federal Aviation Administration (FAA) in the 1990s to improve safety at airports where runway safety areas were constrained. The FAA worked alongside the Port Authority of New York and New Jersey, University of Dayton and a private US corporation to develop EMAS. In 2010, the year proposals were invited in India, 50 such systems were installed at airports across the world, 46 of them in the US alone. In 2020, 66 US airports had such systems. According to the FAA, between 1999 and 2019, EMAS led to 15 such arrestments of overshooting aircraft at US airports. These were carrying over 400 passengers. The system is believed to have never failed in preventing an arrestment. While ICAO recommends a RESA of 240 metres, Kozhikode had just 90 metres due to land constraints.

 
The $10-million proposal and a tragic prophecy
 
After a request by the AAI, the US manufacturer of this system again sent a proposal in September 2010. The proposal made an observation that would turn tragically prophetic a decade later. It noted, “Each end of the Kozhikode runway has a 90-metre runway end safety area, with dramatic changes in elevation beyond making it extremely difficult to obtain the ICAO-recommended length of 240 metres. This system (EMAS) offers an alternative to Kozhikode, where topographic features can cause catastrophic damage and loss of life if an overrunning aircraft is not contained within the runway end safety area. This system can provide that improved safety in far less space.”

 
The AAI had been given three different price options after it had previously conveyed its “budgetary constraints” to the American company. The most effective and expensive one for Kozhikode was priced at $10 million. It involved placing 5,520 such cellulite cement blocks to make an arrestor bed of 85.2 metres in length and 49.4 metres in width. Each of these crushable blocks was four-by-four feet in dimension. The second option, priced at $9.4 million, involved installing 400 fewer blocks in a reduced length of 79 metres. The third option, priced at $9 million, involved placing of even fewer blocks in a length of 75.3 metres. The company claimed these systems were capable of stopping a typical Boeing 737 aircraft landing at 70 knots with brake failure and no reverse thrust, and prevented it from falling into the gorge at the end of the Kozhikode runway.

 
DGCA’s rejection vs corporation’s counter
 
In September 2010, within days of receiving the proposal, the AAI forwarded it to the DGCA for approval. However, there seemed to have been little movement on the Kozhikode proposal until 2015, when a committee of DGCA officials met to discuss installing the same system at the Mangaluru airport. The DGCA rejected the system on many counts. It noted that the system was designed for aircraft touching down at 70 knots and could not arrest planes that landed at higher speeds. It also quoted a FAA report to state that the system needed to have a length of 120 metres while the proposals were for much smaller lengths. Among other things, it rejected the system on the ground that it would take 48 days to replace the concrete blocks, so the maintenance time would be high. It also stated that the life of the system was just 10 years, so replacing it every decade would be a costly affair.

 
The American company sent its response, saying that the life of the system was 20 years, not 10 as mentioned by the DGCA, and that the replacement time was 48 hours, not the 48 days recorded by the DGCA. It rejected the length observation by stating that the proposed system on a shorter area was consistent with the paucity of land at Mangaluru and Kozhikode. Systems on much smaller areas had been installed globally. The US FAA had conducted a data analysis of aircraft over a 12-year period that showed that 90 per cent of all runway overruns happened at a speed of 70 knots or less, with most aircraft stopping within 305 metres from the runway end. An FAA advisory circular dating back to 2005 during the early days of the installation of these systems in the US had calculated the length of the required EMAS system at 120 meters by presuming the aircraft was landing at 70 knots with “no reverse thrust and poor braking”.

 
Again on the back burner, and another Kozhikode prophecy
 
With the DGCA’s rejection, the project to install these systems at Kozhikode and Managluru was again put on the back burner. The Civil Aviation Security Advisory Council (CASAC), at a meeting in February 2013, had agreed to install EMAS at both Kozhikode and Mangaluru airports. Officials from both the DGCA and AAI were present at the meeting. Business Standard spoke to an official who was part of the deliberations at the time. Requesting not to be named, the official explained that even though the court of inquiry into the Mangaluru crash had recommended installing EMAS at tabletop airports, the DGCA was more concerned about its long-term feasibility and operational costs.

 
The official said: “Funding and maintenance costs were a problem area. There were other unspoken concerns. There was only one company in the world that was capable of manufacturing these systems. In 2012-13, everyone was afraid to take decisions because there were scams all around. It was better to wait and watch than act. There was nothing wrong in buying something from a vendor if that was the only enterprise supplying that particular product. But that was a tricky time to take any decision, leave alone a decision like this one.” Rajen Mehta, meanwhile, wrote another letter in 2015 to the new civil aviation minister, Ashok Gajapati Raju, with a copy forwarded to his deputy, Mahesh Sharma. The letter urged the installation of EMAS and ended with a warning that “accidents are just waiting to happen at Managluru and Kozhikode airports.”

 
The Mangaluru redux at Kozhikode?
 
In many ways, the conclusions of the Mangaluru crash probe of 2010 could echo in the Kozhikode crash investigation. Like Mangaluru and most other fatal air crashes, the pilots’ role could be primarily called into question. The court of inquiry headed by former vice-chief of air staff Air Marshall BN Gokhale was scathing in its remarks on the Serbian commander’s role in the crash of IX-812 Dubai-to-Mangaluru flight in May 2010. The report had primarily blamed Zlatko Glusica for the crash. The probe found that despite an abnormal descent Glusica deployed both manual braking and thrust reversers while landing. But just seconds from touchdown, he stowed the thrust reversers and again went full throttle in a bid to make another landing after a turnaround. The probe said that with maximum braking and thrust reversers “the aircraft could have stopped in the overshoot area and the accident might have been averted”.

 
What could set the Kozhikode probe apart is that it would also have to examine what DGCA and civil aviation ministry officials missed in the Mangaluru report. A key recommendation of Air Marshal Gokhale’s report that remained buried as a footnote for a decade read: “In order to help in retarding an aircraft in the overshoot area, ideally, a system such as EMAS, be installed at the tabletop airports. However, in case it is not cost-effective, then at least a soft ground arrestor (SGA) should be available as part of runway end safety area.” While the role of the dead pilots in Kozhikode would certainly be in focus, the reality of that fateful August evening could have been different if warnings since 2007 had been heeded.