Exclusive: MH370 Was Crippled by Sudden Electrical Failure
New data reveals a runaway power outage doomed the 777, supporting the theory a fire in the cargo hold turned the jet into a flying zombie.
The pilots of Malaysia Airlines Flight MH370 were suddenly confronted by a cascading loss of electrical power in which many of the airplane’s vital systems shut down, placing an urgent demand on the crew to understand and deal with the failures.
Before this loss of power occurred the crew had been able to make regular contact with air traffic controllers and the airplane was able to automatically transmit its position.
After it, no word was ever heard again from the pilots. Its two automatic reporting systems, the transponder continually sending the airplane’s position and a separate system reporting the condition of its critical systems at half-hourly intervals, both stopped working.
This new revelation of a serious technical problem and its immediate effects is buried in the arcane detail of a lengthy report (PDF) issued last week by the Australian Transport Bureau, which is directing the search for the Boeing 777. It is the first official acknowledgement of what had previously been only speculation—that there was a sudden loss of electrical power capable of disabling vital systems.
As well as portraying a sudden crisis of control in the cockpit, the report greatly undercuts theories that the pilots themselves went rogue—far from harming the airplane it is much more likely that they were struggling to save it in a situation that most pilots would find hard to master.
The purpose of the report was to reinforce confidence that the undersea search for the airplane is being carried out in the right part of the Indian Ocean and has a high chance of success.
Flight 370 took off from Kuala Lumpur, Malaysia, at 12:42 a.m. (Malaysia time) on March 8, 2014, bound for Beijing. Normally that flight would take around five and a half hours. In fact, it ended seven hours and 38 minutes later somewhere over the southern Indian Ocean, creating the greatest mystery in the history of modern aviation.
The last voice contact with the flight came 37 minutes after takeoff, with the captain signing off with the air traffic controllers in Kuala Lumpur, saying “Good night. Malaysian three seven zero.” The airplane was then on course heading out over the South China Sea.
Two minutes later the blip indicating the airplane’s position on the Kuala Lumpur controllers’ radar screens disappeared—indicating that the transponder was no longer working. At around the same time (as revealed later by military radar that had picked up the flight) the airplane made a sharp left turn, taking it back over Malaysia toward the Strait of Malacca.
The new report is not precise about when the airplane suffered its loss of electrical power: It places the blackout inside a 56-minute window between the final scheduled transmission from the system monitoring the airplane’s critical functions, the Aircraft Communications Addressing and Reporting System, ACARS, and an unsuccessful attempt by the airline’s dispatchers to contact the crew.
But that window can actually be narrowed: The power loss must have occurred in the time between the attempt from the ground to contact the airplane and the last normal contact between the controllers and the captain, some 44 minutes, and very likely it happened very rapidly after the captain signed off—when the transponder failed.
However, whatever the extent of the power loss, the report makes clear that, remarkably, at least one system was able on its own to recover power and continue functioning.
Twenty-one minutes after the airline’s dispatchers tried to contact the flight the airplane was able to transmit a scheduled electronic “handshake” to a satellite.
Tracking the flight path of the Malaysian jet has always rested on one slender thread of data that was detected by the London-based satellite operator Inmarsat.
An Inmarsat ground station in Australia recorded seven electronic “handshakes” transmitted automatically from the 777 beginning with one before takeoff. From those brief and impersonal pulses and after many hours of calculations the searchers were directed to an area deep in the southern Indian Ocean, called the seventh arc, between latitudes 40 and 50 and more than 1,500 miles from the nearest land mass, southwestern Australia.
The handshakes, more commonly called pings, were sent at hourly intervals.
Amazingly, though, the system used to transmit the hourly pings, the Satellite Data Unit, SDU, was able to reboot itself within 60 seconds of the power failure and was able to send the subsequent hourly pings for the rest of the flight, while the ACARS remained silent, as did the transponder.
What caused the power loss?
The Australian report gives four possible causes:
One, a sudden failure that caused the airplane’s Auxiliary Power Unit, APU, to kick in to restore emergency power.
Two, an action carried out in the cockpit using overhead switches.
Three, someone accessing the Main Equipment Center below the flight deck, pulling out circuit breakers and, later, resetting them.
Four, intermittent technical failures.
Clearly, these possibilities suggest a choice between actions that required deliberate human intervention (using the overhead switches in the cockpit or someone gaining access to the Main Equipment Center, pulling out the circuit breakers and then later resetting them) or the sudden onset of technical failures that the airplane’s backup systems were able to restore, at least in part.
In making this range of possibilities clear the report demonstrates that there is no data that could make a persuasive argument for either scenario. That can only be settled when—or if—the remains of the airplane are found and recovered.
However, this new information seriously undermines one of the most persistent conspiracy theories: that the pilots did it.
First, the theory widely advanced in the early days of the disaster that as a first step to make the airplane “vanish” the pilots switched off the transponder. Nobody switched off anything at that moment—it now appears that a power interruption or failure could have disabled the transponder. (A transponder only works for ground tracking within radar range, otherwise its signals can be picked up only by other airplanes that are nearby.)
Second, that one of the pilots left his seat, opened a hatch in the floor, went down into the Main Equipment Center, pulled out the circuit breakers and later reset them.
I asked an expert on the 777 and its systems to comment.
He said that the idea that a pilot went below to pull one or more circuit breakers was extremely unlikely, even bordering on the absurd. He added: “Few airline pilots would even know how to get down to the lower deck while in flight.
“And even if they tried, few would be familiar with the locations of avionics components, or be able to find the relevant circuit breakers to pull. That kind of information is not even contained in the typical pilot training or operating manuals.”
He also explained that the pilots would most likely need to be following “non-normal” procedures to use the overhead switches that control electrical power generation as part of coping with failure messages flashing on their instrument displays.
Indeed, rather than this being an attempt to harm the airplane, the expert said, the pilots could very well have been implementing “a well-defined non-normal procedure” to respond to what was a “very complex failure”—and that those actions were exactly what the pilots should have done.
However, he added, if it was a failure that went beyond anything anticipated in their training—“like a severe uncontained fire”—the crew may not have fully understood the severity of what was happening. “They would simply have no way of knowing.”
Simultaneously, he said, “they would have been trying to decide whether to divert and get on the ground as fast as possible.”
The captain, Zaharie Ahmad Shah, was very experienced, with more than 18,000 hours flying time and 8,659 flying 777s. Fariqu Abdul Hamid, the co-pilot, had only 2,800 flying time experience and—this could well have been significant in a crisis, only 39 hours on the 777, no more than a few flights.
Most of the power to run all the 777’s systems and avionics comes from generators attached to each of the two engines. It is distributed throughout the airplane through multiple connections, many with backup systems and controlled by computers. The main concentration of computers, including those controlling the airplane’s communications systems, is in the Main Equipment Center.
In a Daily Beast special report, I examined a scenario in which a fire in the forward cargo hold of the 777, originating in a consignment of lithium-ion batteries that were being shipped on the airplane, could have breached a wall and reached the Main Equipment Center, seriously degrading the airplane’s avionics and leading to the incapacitation of the crew and passengers.
However, the avionics for the Satellite Data Unit, sending the pings, was located not in the Main Equipment Center but well clear of it, in the roof of the cabin behind the wings, because that is where the antenna to access the satellite is best positioned.
The picture in the Australian report of an airplane stricken by a sudden and extensive loss of electrical power, while in no way definitive, is entirely consistent with this scenario.
Indeed, the report gives dramatic new clarity to the “zombie flight” version of events in which the airplane, by then fatally crippled, makes one final change of course and then flies into the vast emptiness of the southern Indian Ocean without any sign of human direction or control. There is also much more detail about the airplane’s final moments in the air.
The report’s account draws on a scenario followed by Boeing in an engineering simulator (first reported by The Daily Beast) that shows Flight 370 cruising at a constant altitude of 35,000 feet for more than 5 hours at which point the airplane begins to run out of fuel.
The assumption is made that once Flight 370 made a left turn over the Straits of Malacca it was then being flown on autopilot. (The new report cautions: “The specific settings input into the autopilot are unknown. Furthermore, it is also unknown what changes (if any) were made to those settings throughout the accident flight.”)
Considering how little is known of what happened to turn the airplane “dark” the reconstruction of the flight and its conclusion is surprisingly graphic. As the 777 runs out of fuel the right engine flames out first, followed by the left engine 15 minutes later. The airplane then descends in a circling glide, covering as many as 100 nautical miles, hitting the water “uncontrolled but stable.”
As luck would have it, the final—seventh—ping sent from the airplane and intercepted by the Inmarsat satellite ground station was sent about 10 minutes before the airplane hit the water. Within those 10 minutes the SDU had lost power from the engines, the APU had automatically started (taking about a minute to restore power) and the SDU, because power had been interrupted, began automatically to log on again with the Inmarsat satellite and completed that process within seconds of the airplane crashing—thereby providing the Inmarsat analysts with one more essential clue to the final position of the airplane.
There can be no precise picture of how the airplane broke up on hitting the water. The only physical remnant from the crash appeared four months ago, washed up on the island of La Réunion near Madagascar in the western Indian Ocean.
That piece of wreckage was a flaperon, a part of the airplane’s flight controls. There is one flaperon on the rear of each wing close to the fuselage. Although it is relatively small, the flaperon is very busy throughout the whole flight. It is part flap, the control surface that is lowered in a series of phases to increase lift for takeoffs and landings, and part aileron, a separate surface that moves up or down to control “roll”—to keep the wings laterally level at all speeds and altitudes or to control the degree of banking in a turn.
Because of its hyperactive role in the airplane’s flight controls the value of the flaperon to investigators is far greater than its size would suggest. Given the final minutes of the flight as simulated by Boeing, its actions would have been essential to maintaining stability in the glide. For that reason its discovery could add some better understanding of how the airplane hit the water.
The flaperon was in remarkably good condition, given that it had spent nearly 17 months in the water. In photographs the only visible sign of damage is that its thinnest part, the trailing edge, is badly shredded. The forward part, where it is hinged to the wing, appears to have made a clean break.
Estimating the forces that produced that break would be an important part of what investigators would do in order to try assess what role the flaperon was performing right up to the moment of impact. And, by looking at that, the investigators could get clues to how violent—or otherwise—the final seconds of the flight were.
The flaperon was taken from La Réunion to France, where it remains in the hands of the Bureau d’Enquetes et d’Analyses, BEA, having been examined there by experts who confirmed that it came from the Malaysian Boeing 777. (The BEA did not respond to a request from The Daily Beast for information on the examination of the flaperon.)
Meanwhile, in Australia the investigators seized on the discovery of the flaperon as a chance to confirm that their search was being conducted in the right place. Was landfall on the island consistent with the path that any floating wreckage would have taken if it originated in the area being searched?
A team at the Commonwealth Scientific and Industrial Research Organisation, CSIRO, including oceanographers and weather experts, had been working for 16 months using a technology called drift modeling, to predict where, if any floating wreckage survived, it would wash up. Now they reverse-engineered the flaperon’s path from La Réunion back to the search area at the other end of the Indian Ocean, based on the elapsed time, distance, and oceanic conditions from July 2015 back to March 8, 2014, the day that the airplane disappeared.
The result, however, was rather less than assured. Indeed, in describing the findings the CSIRO team leader, Dr. David Griffin, was careful to hedge the bets: The arrival of the piece of wreckage on La Reunion Island “does not cast doubt on the validity of the present MH370 search area” he said, but then added, “it is impossible to use the La Réunion finding to refine or shift the search area.”
It was wise of the scientists to be as careful as this because they had made an embarrassing error in a previous drift model. They originally predicted that the first wreckage would wash up on the west coast of Sumatra, Indonesia, by July 2014—some 4,000 miles northeast of La Réunion.
When this didn’t happen they went back to the numbers and discovered that the data had been corrupted by a significant miscalculation of the effects of wind on the ocean.
It’s fair to say, then, that drift modeling, no matter how conscientiously conducted, is as yet far from being an exact science.
However, the absence of any further floating wreckage since the flaperon was discovered in July lends credence to the idea that perhaps major parts of the 777 did remain intact after impact and then sank, possibly through wave action forcing water into the engines and empty fuel tanks.
I discussed this possibility with the expert on the 777.
He advised caution on reaching any firm conclusions on the basis of a single piece of physical evidence—particularly when the flaperon is visible only in photographs and not by way of a physical inspection.
Nonetheless, he told me, “Even a pilotless jet could possibly get lucky and enter the water at a shallow angle and minimum sink rate that minimizes the impact.
“Most of the structure could have remained intact, or at least separated into only a few big pieces. Not a lot of extraneous debris may have exited the fuselage, particularly if there was no attempt at opening doors or deploying rafts in the water evacuation.”
That would be encouraging for the undersea search because the larger the pieces of wreckage the more likely they are to be detected.
Last week, when the new report was released, the Australian Deputy Prime Minister Warren Truss said that he was “hopeful, indeed optimistic, that we will still locate the aircraft.”
The area being searched totals more than 46,000 square miles of which around 29,000 square miles have so far been covered. As a result of the new analysis of the flight path, priority has been given to the southern sector—the total search area is as long as the distance between New York City and Charleston, North Carolina, and about as wide as the I-95 corridor, little more than 60 miles. Using the new calculations, the length may be shortened as the width is expanded.
And, as the area remaining to cover diminishes—according to the math—the chances of finding the Boeing 777 should increase exponentially.
“We are anticipating that the search will take to around mid-2016 to be completed,” the official spokesman for the Australian Transport Safety Bureau, Dan O’Malley, told The Daily Beast.
The search has continued, operating 24 hours a day, during the southern hemisphere winter, even though the conditions were often appalling.
“There have been times when the vessels were obliged to break off searching because of rough weather,” said O’Malley. “The highest waves were 50 feet in a tropical cyclone. When the weather is really poor work becomes very difficult and obtaining adequate rest is difficult too, so it’s also very fatiguing.”
On two occasions crewmembers fell ill and their ships had to break off and return to their home port of Fremantle, 1,700 miles away.
“There is no helicopter with the range to fly out and recover a patient, and it’s too risky to winch a person from a ship in rough conditions. It’s at least 10 days sail for the round trip, so this delays progress on the search,” said O’Malley.
Before the flight disappeared this was one of the most remote stretches of ocean in the world and its floor had never been mapped. Some of the ocean is as much as 20,000 feet deep, with extremes of terrain. Now, after a bathymetric survey using state-of-the-art equipment, the Australians believe that they have an accurate and detailed map of every piece of the seabed.
These extreme depths and challenging terrain call for the most advanced search equipment, an autonomous underwater vehicle, AUV. Since last May rough weather made it impossible to use this system.
This week, with the financial help of the Chinese (153 of the passengers were Chinese), a third ship equipped with an AUV will join the search.
For months the search had been limited to two ships deploying torpedo-like towfish that scan the ocean bed with sonar. “The deep tow equipment is the most efficient method to search large swathes where the seafloor is relatively flat,” explained O’Malley. “However some of the seafloor features have very steep gradients and maneuvering the towfish over them can leave ‘terrain avoidance’ gaps in the data. These are the areas we will search with the AUV.”
One thing is for sure among many that are not: Should the searchers find the remains of an airplane that took 239 people to their deaths in such baffling circumstances it will be an unmatched achievement in the history of air crash investigations, and the only thing that can finally explain what really happened.
Update, 12/11/15: Three days after the publication of this story the Australian Transport Safety Bureau amended the section of their report that referred to the power loss.
The original text, “power loss occurred” was changed to read “power loss to the SDU.”
In an exchange of emails with the Daily Beast, spokesmen for the ATSB said that the data used in their report related only to the power outage suffered by the Satellite Data Unit, and added that “it is not known if any other systems stopped working during this time”—or whether other systems were affected by the same loss of power that affected the SDU.
There is, however, an explicit clue that not only were other systems affected but that the power problems were related to avionics systems located in the Main Equipment Center below the flight deck.
As the story makes clear, the failure of the jet to send regular half-hourly reports monitoring its vital systems via the ACARS, Aircraft Communications Addressing and Reporting System, is at the heart of the mystery about what made the flight suddenly go “dark.”
It is important to realize that the ACARS messages are sent to a satellite via the SDU, as were the “pings”. After the SDU had suffered a power loss and re-booted it resumed sending the pings – but no further ACARS messages were sent. (The transponder, which also stopped working, used a different antenna to communicate directly with radar, not with a satellite.)
The ACARS data is collected and processed in the Main Equipment Center and then sent to the SDU through dedicated data-carrying lines called busses. Something happened that caused that connection between the MEC and the SDU to be broken, and it remained so for the rest of the flight.
At the same time, the SDU, located well away from the airplane’s lower deck in the roof of the cabin behind the wings, was the only part of the airplane’s communications left working, providing the sole clue to the course of the final hours of Flight MH370. In effect, the SDU had successfully isolated itself from whatever systems were failing elsewhere.
The new amendment to the ATSB document says, flatly, “…the transponder loss of comms [communications] and other aircraft systems are not discussed.”
Well, so be it, that may not be “discussed.” But left unchanged in the report are the four theories put forward to explain the loss of power, as explained in the story above, and they include inferences of foul play by either the crew or others. No evidence is offered to support these theories—nor, obviously, is there any intention of discussing how extensive the loss of power really was.