This article was originally published in THE WEEK issue dated November 24, 1996
He came in wishing the tower good evening on very high frequency. He said he was KZA-1407, flying at 23,000 feet (FL-230 in air traffic parlance) and descending to 18,000 feet (FL-180). He was 74 nautical miles from DPN, meaning Delhi Palam.
The descent meant many things to many people. To the 30-odd passengers on Ilyushin-76, it meant they were close to Delhi where they were to spend a few inexpensive nights before flying back to Kazakhstan with tonnes of garments. To the pilots, it was an ordinary instruction preceding landing.
On ground, responsibilities changed in split seconds. As the plane crossed FL-200, its monitoring passed from the area control office to approach control office.
As instructed, the pilot, or his English interpreter, informed the tower on reaching 15,000 feet. It was
then that Saudi Airlines Boeing-747 (flight SV-763) came in, making the conversation a trialogue. Having taken off from Delhi, with around 300 passengers for Dhahran and Jeddah, the Saudi pilot told the tower that he was approaching 10,000 feet (FL-100). The trialogue was a routine one. Two flights were on the same path, one taking off and the other landing. The tower quickly told the Saudi plane to climb to 140. "Approaching Level 140 for higher," responded the Saudi pilot, meaning he wanted to climb further. The tower response was quick: "Maintain Level 140; stand by for higher."
In came the voice of the Kazakh pilot: "Reach 46 miles DPN; radial 270." He was 46 nautical miles from Delhi. The tower told him to remain at 150 and said: "Identify traffic; 12 o'clock reciprocal; Saudi Boeing-747; 14 miles; report when in sight."
The instruction meant: look out for traffic; there is one coming straight in your lane in the opposite direction.
Fliers describe angles by the clock position; 12o'clock means two planes one above the other like the two hands of a clock and reciprocal means in opposite direction. The Kazakh, well-alerted, asked for a repeat of the miles, the horizontal distance between him and the Saudi plane. Tower repeated "14 miles". "Roger," replied the Kazakh, meaning he has got it. "Traffic is 13 miles," tower alerted him and asked him to maintain his Level 140. The Kazakh acknowledged it by repeating his flight number of 1407. And those were his last words.
The radar blips of both planes disappeared at 18:40 hours, exactly seven minutes after the Saudi Boeing had taken off. The two planes are believed to have run into each other at 18:41 hours over Kheri Sanwal village in Bhiwani district of Haryana.
Fifteen years after a Pushpak trainer crashed into a Chetak helicopter, the first major mid-air collision took place in Indian skies. There is still no definite clue to how it happened, but Indian air traffic experts, ruling out an error from the tower, point to three possibilities:
· The Kazakh pilot heard and acknowledged the instructions, but did not obey them. He might have come down lower than 14,000 feet. The theory that he could not have understood the English instructions has been ruled out after the revelation from the tape of the trialogue. Yet another theory, that he might not have been conversant with the feet measurement as most ex-Soviet planes and pilots use the metric system, too, has been ruled out. The tape reveals that he clearly carried the conversation without the slight hint of a doubt or confusion. That leaves only the possibility of him having flouted the instruction in his hurry to reach the ground, a tendency shown by many sky-weary pilots.
· The Saudi pilot either did not hear the instruction to remain at 140 or flouted the instruction. Strengthening this suspicion is the fact that he did not acknowledge when told to "stand by for higher" than 140. Already, he had expressed a desire to climb higher than 140. Many pilots show a dangerous haste to climb high as soon as possible to save time and fuel and to reach a comfortable cruising speed and level.
· Both flouted the instructions.
The decoding of the black boxes and the digital flight data recorders would reveal the reasons, but the fact remains that a collision was inevitable in the overcrowded Indian skies. The authorities, interestingly, have been quick to come to the rescue of the air traffic control (ATC) personnel who had been on a strike till hours earlier. They denied that the ATC had put the planes too close to each other.
"A separation of 1,000 feet is the standard distance and both were aware of each other's presence," pointed out Civil Aviation Director-General H.S. Khola. "That is well within the stipulations of the International Civil Aviation Organisation (ICAO)."
But what about the horizontal distance?
Though described as 12 o' clock position, the planes were not to pass exactly one over the other. The
last transmitted information is that the Saudi plane was 13 nautical miles ahead of the Kazakh plane. But what was the distance sideways? Civil aviation authorities are tight-lipped. Certainly the radar would have been fed with data from DME (distance-measuring equipment), a system which only a few airports in India including Delhi can boast of. According to aviation experts, Annex-14 of ICAO regulations says that the normal practice when two planes are about to collide is to blindly swerve right. Western experts talk of devices that help the cockpit computers to sense action ahead and turn the plane vertically upwards, but apparently ICAO relies on the right-turn method which is considered safer. In fact, Indian Air Force officers talk of many incidents where the right-turns both on runways and in the air have saved planes and lives.
Aviation experts believe that either one or both the aircraft did steer right on seeing each other. "Anyway, it was not a head-on collision," said civil aviation secretary Yogesh Chandra. "The windscreen and the fuselage of the Ilyushin were found intact. The passengers' bodies too were rather unmutilated, unlike in the case of the Boeing. In fact, the police told us that they had found $35,000 in currency in a passenger's pocket."
The Saudi aircraft which had just taken off had its tanks full, the fire from which enveloped the plane instantly. "One can surmise, from the nature of the debris that one collided on the other's wing or tail while peeling off," said an aviation expert.
The first question that arose from the debris was why planes taking off from and landing in Delhi have to take the same northwesterly route. Brijendra Shekhar, general secretary of the Air Traffic Controllers' Guild, blamed it on the Air Force which runs four stations in and around Delhi. But the Air Force clarified that it had granted permission to have separate arrival and departure routes in June 1996.
It was a permission sought half-heartedly and granted grudgingly. The fact is that the civil authorities are not 'equipped' to run separate routes. "You would need double-directional runways for that," pointed out Yogesh Chandra. "We cannot afford them right now." Added Khola: "Airports in Switzerland give the same corridor for arrival and departure. There are uniform guidelines and specifications on how much should be the vertical and longitudinal separation between aircraft. We operate under them."
The unnerving truth, however, is that not only the airport approaches, but most of the flight-permitted Indian airspace is too crowded. There are too many planes flying through very small corridors in the Indian airspace. The Delhi flight information region (FIR) handles more than a thousand aircraft every day, though many of them may not be landing at all anywhere in India. The whole country is divided into five FIRs around Delhi, Mumbai, Calcutta, Chennai and Guwahati. Calcutta was the most crowded till recently and Guwahati FIR was created last year to relieve the pressure on Calcutta.
The International Air Travel Association (IATA) had virtually warned air traffic controllers in the entire south and southeast Asian region that the airspaces were getting too crowded for them to handle with their existing equipment. India has the largest airspace in the region, which naturally should allow
the overflying aircraft plenty of room to cruise, but security considerations force Indian authorities to limit their routes to limited corridors. Naturally these corridors get overcrowded like narrow streets in Indian towns after the automobile boom in the last one decade.
Of course, much of it is beyond the control of Indian authorities. The air transport is growing at a rate of eight per cent every year in the Asia-Pacific region, which is higher than most other parts of the world. In fact, 35 per cent of people all over the world who travelled out of their countries in 1993 were in this region and IATA estimates that by 2010, more than half the international travellers would be from here. And in Asia the Indian sky is the most crowded because of the narrowness of the corridors.
Most of the flights between European and southeast Asian destinations do not land in India, yet they
have to be handled by Indian ATCs. Direct flights from southeast Asian cities to Europe have increased from 364 a week in 1989 to 672 this year. Of them 594 are from just four airports—Hong Kong, Bangkok, Kuala Lumpur and Singapore. And 530 of these flights have to cross the India airspace, with 462 reporting first to Calcutta and subsequently to Delhi.
The rush across the Calcutta and Delhi FIRs is easy to explain: the planes just want to avoid the Himalayas. Only 54 of the 594 take the northerly route across China and Russia for two reasons. One, the route is longer and two, Chinese ATCs are ill-equipped to handle heavy traffic.
The pressure thus on the Indian air corridors has been mounting. In addition to these overfliers are the hundreds of flights landing in and taking off from the country's airports. The approach controllers would have to take care of only the landing and take-off of the planes, but the area and zonal controllers have to watch them as well as the overfliers. "Every day, we have to track and control 700 to 1000 aircraft in the Delhi FIR alone," said an official.
Airlines have been demanding that India widen its corridors or open more of them. Their argument is simple: most of the direct flights between Europe and southeast Asia are carried out by the wide-bodied aircraft like the Boeing-747. The optimum cruising altitude for a Boeing-747-400 two hours after take-off is FL-300 (30,000 feet). Two hours is approximately the time taken for a flight to enter Indian airspace from southeast Asia. But the lanes across India at that height are so crowded that Delhi and Calcutta order many of them to descend to such low levels as 24,000 feet. The airspace being subcontinental in expanse, flights on this route often have to fly for nearly five hours at such low altitudes across India. A recent study by Boeing found that about a third of the total flying hours is spent at such low altitudes which is expensive in terms of both fuel and time.
Essentially, what the airlines are demanding is a larger or an alternative corridor, but the Air Force which is primarily responsible for security of the airspace is not willing to concede. The Air Force has a chain of radars running from Kashmir to Kanyakumari which can detect every plane entering the airspace. But detection is not its only job. Along with civil air traffic controllers, it has to monitor every flight across the airspace lest Purulias should be repeated. It would therefore like to limit civil flights to corridors that it can manage. Of the many such corridors that have been opened across the Indian airspace, the one across Delhi is the most congested. Mumbai handles more landing and take-off traffic but has much less overflying traffic. Adding to the worries of the night-shifters in ATCs is that most southeast Asian fliers like to leave in the evening and reach Europe early morning.
Indeed, the night traffic into, from and across India is chaotic. Communication with and separation between flights is a nightmare with the available equipment. Reported instances of near-misses have been galore. According to information available with the civil aviation authorities, there were 50 near-collisions between 1991 and 1994. Among the narrowest misses were those between flights of Cathay Pacific and British Airways over Delhi, Indian Airlines and Aeroflot near Farrukhabad (both in 1989), Royal Nepal Airlines and Scandinavian over Lucknow, Indian Airlines and Bangladesh Biman near Calcutta.
Most near-misses have been while changing altitudes, and so too happened with the ill-fated Saudi and Kazakh planes. In 1990, Malaysia Airlines flight from Delhi to Kuala Lumpur narrowly missed a KLM aircraft while climbing to cruising level over Farrukhabad.
Two years later, a KLM nearly collided with a Cathay Pacific near Calcutta. Days later, a Swiss Air almost flew into a Japan Airlines flight near Lucknow. Most such near-collisions have been reported from Delhi FIR. In the four years from 1991, about which information has been released, there were 50 such near-misses in the Indian airspace.
Officials say that much of the problem could be solved with the opening of the Future Air Navigation System (FANS-1). The system involves phased induction of a whole new family of technologies, the first being communication, navigation, surveillance/air traffic management (CNS/ATM) which involves installation of controller-to-pilot data link communications (CPDLC) and automatic dependent surveillances (ADS). According to Yogesh Chandra, "Delhi and Mumbai would have state-of-the art systems installed by Raytheon by early next year."
The biggest advantage of the CNS/ ATM would be that separation between aircraft can be reduced even to eight kilometres. As an air traffic controller explained, "It would be still like bumper to bumper traffic both ways on a narrow road, but if everyone is made to stick to their lanes there would be a reasonably good flow of traffic without touching each other."
An interim CNS/ATM in the Calcutta region is expected to permit reduced separation between aircraft over the Bay of Bengal—more aircraft, flying closer to one another, would be accommodated in the corridors.
An upgrade of the system is expected by 1999 and will add ground-to-ground data links. Aviation experts say that a decreased separation under CNS/ATM need not be of much concern. The automatic dependence surveillance (ADS), tested in the South Pacific in 1988-89, uses data links to transmit accurate aircraft positions to controllers which they see on their displays.
When the system is in place, controllers will have access to the aircraft's flight management system (FMS) computer. Through that, even if the pilot errs in his communication to the controller, the cockpit computer would directly give the controller error-free automatic position reports wherever in airspace.
Auckland (New Zealand), Singapore and Brisbane already have the system installed. Bangkok and the four Indian traffic systems are about to have it. Another prospect of relief for the Indian traffic controllers is that China is planning to join the FANS club which would enable it to open another corridor in the Far East Asia route, taking away some of the traffic across India.
However, as M.S.G.K. Warrier, former director of modernisation in National Airports Authority, observed, "FANS has little to do with the present situation as it is meant for en route control of flight path." But controllers say that it would still relieve the high pressure in the FIRs. Moreover, many of the near-collisions of the last few years have been of overflying aircraft.
But then, as the recent collision has shown, the crowded approach too should be of as much concern as the narrow corridor. Presently, there is only one westerly route (G-52) for arrival to and departure from Delhi. There is only a primary radar system which sends out signals and receives them back from the aircraft. A secondary radar, on the other hand, has a transponder which directly receives signals from the aircraft and displays on the screen the call sign, level, speed and such details.
The controller presently has to depend on the information given orally by the pilot. If the pilot says he is at 14,000 feet, the controller has to trust him and put another flight 1,000 feet higher. On the other hand a secondary radar would give the controller what is known as the 'synthetic height', that is information given automatically by the system. Even if the pilot is wrong, the controller would still get the right information.
But Yogesh Chandra observed: "Even with the best technology there could be factors beyond human control". In fact, many aviation experts do not think that the failure in this case was of technology. "More accidents take place because of human errors," they point out. "Of the 24 accidents (none of them collisions) that occurred to civil-registered aircraft in India between mid-1993 and mid-1995, none had been due to equipment defect."
Many have therefore wondered whether the recent accident was caused by one such human error. "It is possible that the Kazakh crew failed to grasp the phraseology used in communication from the ATC," said Air Vice Marshal H.M. Shahul, former chairman of the Airports Authority of India. But the Kazakh authorities rule out the theory that their pilot could not have understood English or that his metric system could not quickly convert the altitude instructions into feet. They maintain that the Ilyushin's altimeter had feet readings and that the co-pilot had undergone courses in aeronautical English.
also read
But there is one thing on which almost everyone agrees: The Indian air corridors are too crowded for the present range of control equipment to cope with. But the security authorities are not willing to expand the corridors for fear of Purulias. The only option then is to upgrade the technologies.