Dana Air – Nigeria
3 June 2012
Boeing/ McDonnell Douglas MD-83
On 3 June 2012, a Dana Air Boeing / McDonnell Douglas MD-83 aircraft flying from Abuja, Nigeria, crashed in a suburb five miles north of Murtala Muhammed International Airport in the capital city of Lagos. All 153 people on-board, and a number of people on the ground, were tragically killed.
The Nigeria Leadership Initiative (NLI), a not for profit organization in Lagos, provided assistance to families who had lost loved ones. With a lack of information from the accident investigation and the complex legal issues surrounding the case, the NLI invited Jim Morris to advise families at two Families’ Forums organised by the NLI. The first Forum was in Lagos in July 2012 and the second was in Abuja in August 2012.
With his expertise as a RAF pilot on large Boeing aircraft, Jim provided the families with specialist advice on the accident, the investigation, the law and the best litigation strategies. Subsequently, in December 2012, Jim was invited by the Nigeria Bar Association to present a lecture on international aviation litigation to Nigerian lawyers in Lagos.
Unfortunately, during the 4 years following the accident, the Nigerian Aviation Authorities only published very brief preliminary reports that indicated a loss of power in both engines. Despite indications that a comprehensive final report would be published in 2016, the authorities did not publish a report that year.
Notwithstanding the absence of a final accident report, proceedings were filed in Florida, USA, and the High Court of Lagos, Nigeria. The actions alleged that Dana Air and the pilot of the aircraft were the primary parties liable for the accident.
Finally, in March 2017 the Nigeria Accident Investigation Bureau (AIB) published the Final Report. The Report confirms that the Flight Data Recorder (FDR) suffered extensive heat damage in the fire which meant that no data on the aircraft engines or systems could be recovered. Although the data on the Cockpit Voice Recorder (CVR) was recovered, the lack of FDR data meant that the information to determine what exactly happened was limited.
The report concludes that the probable causal factors were:
- Engine number 1 lost power seventeen minutes into the flight, and thereafter on final approach, Engine number 2 lost power and failed to respond to throttle movement on demand for increased power to sustain the aircraft in its flight configuration.
- The inappropriate omission of the use of the Checklist, and the crew’s inability to appreciate the severity of the power-related problem, and their subsequent failure to land at the nearest suitable airfield.
- Lack of situation awareness, inappropriate decision making, and poor airmanship.
The report also identified that the Captain had been suspended in 2009 by the US FAA, he had recently joined Dana and that his line training with Dana was hurried. Using his professional pilot experience, Jim analysed the timeline of events and the report contents.
The aircraft was registered with Dana Air on 3 February 2009, so had been with the airline for over 3 years. The left (No1) engine was overhauled at a facility in Florida on 3 August 2011 and the right (No2) engine was overhauled at the same facility on 2 December 2011. It is not clear from the report when these engines were fitted to the aircraft, but the accident happened 11 and 6 months after the respective overhauls. An important point that is highlighted by the report is that the left engine had not been upgraded in accordance with Special Bulletin (SB) 6452. SB6452 was required to prevent a problem with the engine (including shutdowns, air returns and fires) which was caused by assembly fractures in the secondary fuel manifold. The report focuses on this issue as there was a loss of power in another Dana Air MD-83 aircraft on 6 October 2013. The engine on this incident was overhauled by the same facility in Florida and also was not upgraded to SB6452 and the cause of the loss of power was secondary fuel manifold tube fractures.
However, metallurgical tests by the US air accident investigators (who assisted the Nigerian AIB) couldn’t find any evidence that the manifold tubes were fractured pre impact. As such, in the absence of reliable physical evidence and FDR data on the engines, the Nigerian AIB was unable to confirm that crushed or fractured fuel feed lines caused leaks/ restrictions to both engines prior to impact.
To lose power in both engines in an airliner is an extreme situation and should never occur. Looking in detail at the events on the day of the accident gives an insight into a difficult in-flight emergency (that could have been recoverable) that became an in-flight catastrophe.
1458:00hrs - the aircraft was airborne reporting a fuel endurance of 3 hours.
1513:44hrs - This was the start of the recording on the Cockpit Voice Recorder (CVR), which retained about 30 minutes 53 seconds of the flight to impact. The transcript began as the aircraft was in cruise (at 26,000 feet) heading toward Lagos. When the recording started the Captain and First Officer (F/O) were in a discussion of a non-normal condition regarding the correlation between the engine throttle setting and an engine power indication, with particular reference to the ‘left’ which AIB believed was a reference to the left engine.
1515:56hrs - the Captain stated that “we just want to get # home” which AIB believed was the aircraft.
1515:56hrs - 1522:00hrs - the crew continuously discussed the problem with the left engine with vague references to EPR, non response from throttle inputs and left engine operating at idle power or inoperative. There was no evidence from the CVR that the crew called for non-normal/ emergency checklist to ascertain non response of the left engine to throttle movement.
1518:00hrs - The aircraft made contact with Lagos Area Control Centre and provided an ETA at Lagos at 1545:00hrs
1522:00hrs - The aircraft started the initial descent from cruise altitude (26,000 feet). At this point the crew were becoming increasingly concerned at the condition of the left engine.
1530:00hrs - the aircraft reported descending through 18,100 feet.
1531:12hrs - the crew confirmed that there was no throttle response from the left engine.
1531:27hrs - the captain took control as flying pilot.
1531:27hrs - 1534:33hrs - at some point during this period there was confirmation of throttle response on the right engine and the engine anti ice, bleed air and ignition were all switched off.
1532:05hrs - the crew observed the loss of thrust of the left engine.
1534:33hrs - the Captain confirmed that “okay this one is good for us so far”.
1537:00hrs - the crew started performed pre landing tasks.
1537:07hrs - extension of slats.
1537:10hrs - speed brakes were deployed.
1537:53hrs - the Captain requested Flap 15.
1538:35hrs - the crew indicated that they were 21 miles out on the localizer (extended centreline) for the runway.
1539:10hrs - the Captain asked for landing gear down.
1540:00hrs - The aircraft reported descending through 7,700 feet.
1540:42hrs - the F/O confirmed that they were at 5000 feet at 15 miles out.
1541:46hrs - F/O inquired, “both engines coming up?”
1541:48hrs - the Captain replied “negative”.
1541:56hrs - the F/O asked the Captain if he could declare an emergency and the Captain agreed.
1542:10hrs - The crew radioed an emergency indicating dual engine failure - “… negative response from throttle”.
1542:35hrs - the Captain requested Flap 28 and this was selected.
1543:45hrs - the Captain reported runway in sight.
1543:49hrs - the Captain instructed the F/O to retract the flaps.
1543:50hrs - the Captain instructed the F/O to retract the landing gear.
1543:50hrs- 1544:04hrs - during this period the Captain informed the F/O that he had lost both engines.
1544:04hrs - 1544:29hrs - the flight crew attempted to recover engine power without reference to any checklist.
1544:29hrs - the last event captured on the CVR “too low gear”.
1544:33hrs - the aircraft crashed on the extended centerline of runway 18R around 5.8 miles short of the runway.
The report states that there was never any call for the checklist both normal and non-normal/ emergency throughout the entire flight and use of the Quick Reference Handbook (QRH) was never detected.
Given that at the start of the CVR recording (1513:44hrs) the crew were discussing an issue with the left engine, at this point they should have been going through the relevant checklist process to identify/ rectify the problem. Also, having identified a problem with an engine, they should have been considering the nearest suitable airfield to divert to. At this point two suitable airfields were Abuja and Ilorin international airport (which was the officially filed alternate airport for the flight that is approximately half way between Abuja and Lagos and around 75 miles to the northwest of the direct track from Abuja - Lagos).
Unfortunately the Final Report does not set out the exact track of the flight nor does it provide a clear timeline analysis of the time left to total loss of power and the time it would take for the aircraft to turn back to Abuja or divert from the planned track to Ilorin or any other suitable airfield. As the CVR transcript indicates that between 1531:27hrs - 1534:33hrs there was confirmation of throttle response on the right engine, at the start of the CVR recording (1513:44hrs) it appears that there were possibly around 17.5 minutes of power remaining on the right engine. As such had they had diverted to Abuja or Ilorin at that point and had the power lasted for 17.5 minutes, they may have been able to touch down safely with power from the right engine.
At 1522:00hrs the crew were approximately half way between Lagos and Abuja. At this point the crew were becoming increasingly concerned about the left engine yet they still did not carry out any of the checklists. The report does not confirm the nearest suitable airfields at this point, but it is likely that Akure and Ilorin airfields were closest, yet the crew did not divert but continued to Lagos and started to descend.
Although the reasons for loss of power in both engines could not be determined, if the left engine lost power for the same reason as the later flight in October 2013 (fractures in manifold fuel feed), it would be very unusual for the right engine to fail for exactly the same reason, especially as the right engine had been upgraded with SB6542. The cause of the right engine failure is a mystery, but if the extraordinary situation existed where both left and right engines suffered the same fuel feed fracture problems, then the loss of power in both engines may not have been preventable. In this situation the failure of the crew to carry out the checklists and divert to the nearest airfield when they first noticed the problem may have meant that they placed themselves in a situation where the aircraft would lose power on both engines before they could touch down on a suitable airfield.
On the facts available for the accident flight, the Captain was indicating that he was not getting power on both engines at 1541:48hrs. At this point, the Captain should have asked the F/O to commence the double engine failure checks and the Captain should have selected a gear/flap/ speed configuration that would maximise the glide range to give the greatest chance of reaching the runway. Unfortunately, the report is not clear on the aircraft height, speed and its distance from the runway at that point and whether it could have glided to the runway.
There are some glide simulator tests at page 54 of the report. These tests indicate that if the aircraft had been at 5000 feet and 15 miles from the runway with gear down and Flap 28 or 15, it could have reached the runway. If it was at 2200 feet at 15 miles with gear and Flap 15, it would not have had enough height to reach the runway. The next part of the simulator test is not clear in the report, but it appears to indicate that at 2200 feet at with gear and flap up, the aircraft did have enough height to reach the runway. It is disappointing that there is not more detail/ simulator reconstruction on the height and distance from the runway when the aircraft lost power and whether it could have glided to the runway and what configuration it would need to be in.
When an aircraft loses total engine power, the priority for the pilot is to is to maximise its gliding range, which maximises the chance of reaching a suitable runway. Glide range is maximised by minimising the aircraft drag (gear and flaps retracted) and flying at the glide speed for the aircraft (by setting a nose down pitch angle that achieves the glide speed). The Captain should have known the gliding performance and speed for the aircraft.
When the power didn’t come up at 1541:48, the Captain should have immediately determined the distance to the runway, the height and speed of the aircraft, and put it in the best configuration to reach the runway and avoid an aerodynamic stall (flying at too low an air speed to provide lift for a given wing flap selection). As a minimum, the landing gear should have been retracted immediately as this causes drag and does not contribute to lift. However, the Captain did not call for gear retraction until 1543:50 - 2 minutes after loss of power. In fact, contrary to minimising drag, at 1542:35 (47 seconds after total loss of power) the Captain requested Flap 28, which is a higher drag setting for the flaps. Had he selected gear up, maintained Flap 15 and pitched the nose to maintain the Flap 15 glide speed there would have been less drag and more glide range. Furthermore, if at the time of losing engine power the aircraft was above the stall speed for a clean wing (flaps up) glide, retracting the gear and flap and pitching the nose to maintain clean wing glide speed should have maximised glide range. From the report, it appears that the Captain delayed the decision to retract gear and flaps until he stated the runway was in sight at 1543:45hrs.
This 2 minute delay to minimise drag combined with the decision to increase drag by setting Flap 28 will have adversely impacted on glide range. It is possible that if the Captain had correctly maximised glide range at the point that he realised that he had lost the right engine, the aircraft may have reached the runway. In fact, if the simulator reconstruction at 2,200 feet with gear and flap retracted was a replication of the actual height, speed and range of the accident aircraft and the simulator reached the runway, this would indicate that promptly setting the correct glide configuration could have enabled the accident aircraft to reach the runway and reduce the severity of the losses.
In addition to correctly maximising the glide range, had the double engine failure checks been completed it may have been possible to re-start one of the engines and land on the runway with power.
This Final Accident Report identifies a devastating series of events that contributed to the tragedy. It is very unfortunate that no data could be recovered from the FDR as this may have enabled the investigators to identify what went wrong with the engines. Notwithstanding the difficult technical problem(s) with the aircraft, the poor airmanship and failure to use the required checklists made matters far worse, so these human factors were a major contribution to the catastrophic loss of the aircraft. Clearly the Nigerian aviation authorities and the aviation industry need to learn lessons from this accident and implement all necessary measures to prevent a similar future accident and improve flight safety in Nigeria.