Swiftair MD83 Loss Of Control In flight final report

Approved on April 22, 2016

The aeroplane took off at night from Ouagadougou airport at about 1 h 15 bound for Algiers. During the climb, the crew made several heading changes to avoid a stormy area before reaching cruise level FL 310. A few minutes later, the aeroplane’s speed, piloted by the autothrottle, decreased due to the obstruction of the pressure sensors on the engine nose cones, likely by ice crystals. The autopilot then progressively increased the aeroplane’s pitch attitude to maintain the altitude, until the aeroplane stalled. The aeroplane’s stall was not recovered.

The aeroplane maintained its nose-up attitude and left bank while the control surfaces remained mainly deflected in a pitch-down attitude and with a right bank. The aeroplane struck the ground at high speed.

MD 83 LOC accident Malí

Image from BEA Final Report

FACTUAL INFORMATION

  • Aircraft: MD-831 registered EC-LTV (Although the official designation is McDonnell Douglas DC-9-83, for reasons of brevity the term MD-83 will be used throughout this report.)
  • Date and time: 24 July 2014 at 1 h 472 (Except where otherwise indicated, the time given in this report are expressed in Universal Time Coordinated (UTC), which was also the official time in Mali on the day of the accident)
  • Operator: Swiftair S.A.
  • Place of accident: 80 km south-east of Gossi, Mali
  • Type of flight: International public transport of passengers flight AH5017 Ouagadougou (Burkina Faso) – Algiers (Algeria)
  • Persons on board: Captain; copilot; 4 cabin crew; 110 passengers
  • Consequences and damage: 116 fatalities; aircraft destroyed

History of Flight

Note: the following elements are based on data recorded on the FDR, radio communications and witness statements. No useable data was recovered from the CVR

On 24 July 2014, the MD-83 registered EC-LTV was programmed to operate flight AH 5017 from Ouagadougou bound for Algiers. One hundred and ten passengers and six crew members were on board.

The flight plan filed planned a departure via Niamey (NY), then ROFER via route UM608.

At 1 h 02 min 20, the crew was cleared to start up for a departure from active runway 22.

At 1 h 10 min 14, the crew was cleared to taxi to runway 22 and said that they wanted flight level FL 330 for cruise, then changed their minds and requested FL 310 initially, because the aeroplane’s weight was too high for FL 330.

At 1 h 13 min 05, the controller cleared the crew to perform the departure via EPEPO, towards FL 310, with a turn to the right after takeoff. The controller had prepared for a departure of this flight via GAO through the EPEPO point, through which the aeroplane had passed at the time of its arrival at Ouagadougou from Algiers.

At 1 h 15 min the crew took off, then turned to the right and flew 023° heading. At an altitude of about 10,500 ft, the left side autopilot was engaged (This means that the Captain was likely the Pilot Flying (PF)), the autothrottle having been active since takeoff.

Nine minutes after takeoff, the crew said that they were passing through FL 145 and that they estimated EPEPO point at 01 h 38, and Algiers at 05 h 06.

At 1 h 28 min 09, climbing through FL 215, the aeroplane was transferred to the Ouagadougou ACC to which the crew said that they were turning to the left on heading 356° due to an avoidance manoeuvre (The precise message from the pilot was as follows: « we are turning left heading 356 to avoid »). The total air temperature was then 9°C5.

During the climb towards FL 310, the crew made three heading alterations to the left (of 28°, of 4° and then of 8°), then an alteration to the right of 36° to return to heading 019°, close to the initial heading. The TAT reached 6°C at 01 h 31 min 11

At 1 h 37 min 28, the aeroplane levelled off at FL 310 at Mach 0.740. The autopilot then maintained the aeroplane’s altitude and heading, while the speed was controlled by the autothrottle. At the same time, the aeroplane was transferred to the Niamey ACC. The Total Air Temperature was then -5°C.

In the two minutes following level-off, the aeroplane’s speed increased.

From 1 h 38 min 34, and for about 30 seconds, the autothrottle was in MACH ATL mode. The engines’ EPR (Ratio between total pressure at the engine outlet and that at the inlet) stabilised around 1.92 and the Mach changed from 0.758 to 0.762. The autothrottle then returned to MACH mode and the aeroplane continued to accelerate up to Mach 0.775.

At 1 h 39 min 36, the aeroplane’s speed started to decrease.

At 1 h 40 min 10, the autothrottle changed to MACH ATL and for about thirty seconds this mode alternated with MACH mode. At 1 h 40 min 46 the autothrottle changed back to MACH ATL mode while the Mach was 0.752. Following that and until 1 h 45, the altitude remained stable, pitch and EPR increased progressively, while the engines’ N1 (The N1 parameter represents the rotation speed of the engine low pressure rotor, expressed as a percentage of a reference speed) remained stable and the speed continued to decrease.

Between 1 h 41 min 38 and 1 h 44 min 29, the Niamey ACC and flight AH5017 tried to get in contact but did not manage to do so. Flight RAM543K offered to act as the intermediary. The crew of flight AH5017 announced, at 1 h 44 min 29, that they were at FL 310 on an avoidance manoeuvre. The Niamey ACC heard this radio exchange and then asked them to squawk the 3235 transponder code. He also asked them to call back passing GAO and to transmit estimates for MOKAT point.

No answer or any other messages from flight AH5017, reached the Niamey ACC and the transponder code used by flight AH5017did not change.

At 1 h 44, EPR and N1 fluctuations on both engines appeared for about 45 seconds.

Then, for about twenty seconds, the EPR increased then decreased on two occasions from 1.6 to about 2.5. The N1s increased up to 91% during the first oscillation and remained between 83 and 87% during the second. Some roll oscillations between 4° to the left and right appeared. The autothrottle was disengaged between 1 h 45 min 02 and 1 h 45 min 06 (Autothrottle modes are only recorded every 4 seconds).

This disengagement occurred between the first and second EPR variations.

At 1 h 45 min 06, the calibrated airspeed was 203 kt, the Mach 0.561, the angle of attack was 9° and the aeroplane started to descend. Pitch increased until it reached 10° at 1 h 45 min 17, and then decreased slightly while the deflection of the elevators and the position of the trimmable horizontal stabiliser (THS) continued pitching up. The EPR and the engines’ RPM started to decrease towards values corresponding to idle. The roll oscillations continued and the speed continued to decrease.

At 1 h 45 min 35, the autopilot disengaged. The altitude had fallen by about 1,150 ft in relation to cruise flight level, the calibrated airspeed was 162 kt, the Mach 0.439, the angle of attack was 25° and both engines were almost at idle. The aeroplane’s pitch began to decrease and the bank was increasing to the left.

The aeroplane’s pitch and bank were then subject to significant changes. They reached, respectively, 80° nose-down and 140° bank to the left. The aeroplane was pitched nose down and banked to the left until it struck the ground. The flight control surfaces remained mainly deflected pitch-up and in the direction of a bank to the right. Around twenty seconds before the impact, the flight control surfaces pitch-up deflection decreased, then the engine speed increased again and reached values close to maximum thrust.

The last values were recorded at 1 h 47 min 15:

 Pressure altitude: 1,601 ft, (in relation to the 1013 hPa isobar)

 Calibrated airspeed: 384 kt

 Pitch: 58° nose-down

 Bank: 10° to the left

 Magnetic heading: 099°

Swiftair MD83

EC-LEY similar to the crashed aircraft. Photo (C) Ray McFadyen http://www.airteamimages.com/mcdonnell-douglas-md-80_EC-LEY_swiftair-28spain29_125995.html

EC-LTV-3

Wreckage from the EC-LTV Photo: http://www.baaa-acro.com/2014/archives/md-83-is-missing-over-mali-with-116-people-on-board/

CONCLUSION

Findings

 The aeroplane took off from Ouagadougou bound for Algiers with 116 people on board with no known technical problems.

 The crew held the licences and ratings required to carry out the flight.

 The aeroplane had a valid airworthiness certificate.

 The aeroplane weight and balance was within operating limits.

 The meteorological situation was what could be expected at that time of the year in the intertropical convergence zone.

MD 83 LOC accident Malí (2)

Image from BEA Final Report

 A CVR malfunction that preceded the event flight prevented it from producing a usable recording of the noises and conversations in the cockpit. The crew could not detect this malfunction during the pre-flight test. The malfunction occurred after the last maintenance action where it could have been detected.

 During climb to FL310, the crew made some heading changes to avoid cloudy areas, which led it to fly on the edge of a convective cloud system.

 The route chosen led to them flying in an area where the presence of ice crystals was likely.

 The anti-icing systems had not been activated during aeroplane climb and the cruise phases during which the autothrottle was in MACH ATL mode.

 When levelling off, the autopilot went into altitude and heading hold mode while the speed was piloted by the autothrottle in MACH mode.

 The EPR values of the right engine then the left engine became erroneous probably following the obstruction of the Pt2 pressure sensors by ice crystals.

 These erroneous EPR values caused the autothrottle to limit the thrust produced by the engines to a level lower than the thrust required to maintain FL310.

 About two minutes after levelling off, the aeroplane speed started to decrease.

 The autopilot offsets the speed decrease by increasing the aeroplane flight attitude to maintain altitude.

 After the transfer of flight AH5017 to the Niamey ACC, radio contact with the aeroplane was not immediately established. Flight RAM543K, flying in the same sector, acted as a relay between flight AH5017 and Niamey ACC.

 Between the transfer of the crew to the Niamey ACC and the end of the flight, 15 attempts at radio contact were recorded on the FDR. Only two messages were received by the Niamey ACC.

 The Niamey ACC heard the crew of flight AH5017 saying that they were at FL310 and undertaking an avoidance manoeuvre.

 The Niamey ACC received no other message from the crew after that.

 Two EPR fluctuations of strong amplitude caused by crew input on the throttle levers were observed. The autothrottle disengaged between these two variations, and the aeroplane started to descend. At the same time, the last activation of the VHF, 4 seconds long, was recorded on the FDR.

 When the aeroplane speed reached 203 kt, vibrations attributed to buffet appeared.

 Four seconds later, when the aeroplane’s speed was 200 kt, the stick shaker activated, followed three seconds later by the stall warning. At that time, the aeroplane reached the 12° stall angle of attack.

 Both engines suffered a surge due to the aeroplane’s high angle of attack.

 The autopilot disengaged about 22 seconds after the triggering of the stall warning. The aeroplane angle of attack had then reached approximately 25°. There was no apparent crew action between the stall warning activation and the autopilot disconnection.

 During the aeroplane’s fall, its pitch attitude and bank changed significantly. The aeroplane continued to pitch down with a left bank angle down to the ground. During this flight phase, the control surfaces remained mainly deflected pitch up and in the direction of a bank to the right. When the elevators were commanded close to the neutral position, around 27 seconds before the impact, the stick shaker and stall warning system intermittently deactivated and the engines resumed normal operations.

 No problems were mentioned by the crew during their contacts with the Ouagadougou and Niamey air traffic controllers.

 No distress message was received by the control centres.

 The last recorded values were a 58° pitch-down attitude, a 10° bank to the left and a calibrated airspeed of 384 kt.

 The emergency DETRESFA phase was triggered at 04 h 38.

 The wreckage was recovered at 18h23.

 The crew had not received any training relating to approach to stall and recovery from it since joining Swiftair. The investigation could not determine how far back their last training on these points took place.

 During simulator training, the approach to stall is performed manually by the crew after disengagement of the autopilot.

 The stick shaker and stall warning initiation speeds on the airline’s simulator are values lower than those calculated for the accident flight.

 The FCOM procedures relating to engine or airframe icing do not mention the characteristics of icing by ice crystals, and are thus not adapted to the specifics of this type of icing.

 The phenomenon of obstruction of the Pt2 pressure sensors due to icing is only described in the FCOM in the climb phase

 The stall warning devices triggered at high altitude with a speed margin below the 7% mentioned in the certification regulations. These specific features were not brought to the attention of operators.

 The autopilot remained engaged beyond the triggering of the stall warning devices and the stall. This specific feature, and its consequences on the detection of and recovery from a stall, is not explicitly detailed in the manufacturer’s FOB, the only documentation supplied to operators on this point.

Causes

The aeroplane speed, piloted by the autothrottle, decreased due to the obstruction of the pressure sensors located on the engine nose cones, probably caused by ice crystals. The autopilot then gradually increased the angle of attack to maintain altitude until the aeroplane stalled. The stall was not recovered. The aeroplane retained a pitch-down attitude and left bank angle down to the ground, while the control surfaces remained mainly deflected pitch up and in the direction of a bank to the right. The aeroplane hit the ground at high speed.

The accident was the result of a combination of the following events:

 the non-activation of the engine anti-icing systems

 the obstruction of the Pt2 pressure sensors, probably by ice crystals, generating erroneous EPR values that caused the autothrottle to limit the thrust produced by the engines to a level below that required to maintain the aeroplane at FL310.

 the crew’s late reaction to the decrease in speed and to the erroneous EPR values, possibly linked to the work load associated with avoiding the convective zone and communication difficulties with air traffic control.

 the crew’s lack of reaction to the appearance of buffet, the stick shaker and the stall warning.

 the lack of appropriate inputs on the flight controls to recover from a stall situation.

These events could be explained by a combination of the following factors:

 The FCOM procedure relating to the activation of the anti-icing systems that was not adapted to Pt2 pressure sensor obstruction by ice crystals

 Insufficient information for operators on the consequences of a blockage of the Pt2 pressure sensor by icing

 The stick shaker and the stall warning triggering logic that led these devices to be triggered belatedly in relation to the aeroplane stall in cruise;

 the autopilot logic that enables it to continue to give pitch-up commands beyond the stall angle, thereby aggravating the stall situation and increasing the crew’s difficulties in recovery.

The absence of a usable CVR recording limited the possibility of analysing the crew’s behaviour during the flight.

Specifically, it was not possible to study CRM aspects or to evaluate the possible contribution of the employment context and the experience of the crew members.

Swiftair MD83 (2)

EC-LVT Photo (C) David Tobarra http://www.airliners.net/photo/Swiftair/McDonnell-Douglas-MD-83/2455162/&sid=867b5878c1eac34b6f201f895ccd8b3c

SAFETY RECOMMENDATIONS

1. Recommendations issued during previous investigations and relevant to the analysis of the accident to flight AH5017

1.1 Accident on 16 August 2005 to the DC-9-82 (MD-82) operated by West Caribbean Airways

The investigation recalls that the Venezuelan authorities, at the end of the safety investigation, had recommended that:

 058/2005-AA2: the aviation authorities require the inclusion in flight crew training of recovery from high-altitude stalls. This is justified on the grounds that in simulator training, low-altitude stalls are induced, from which the aircraft can recover more quickly, because it can increase power in order to maintain altitude, whereas at high altitudes, the behaviour of aircraft is different, calling for more precise manoeuvres in terms of time of execution.

 058/2005-AA5: In those countries in which there are air operators certified to operate MD-80 series aircraft, they increase and optimise the requirements in the flight crew programmes in the presence of buffeting at high altitudes, and also those with regard to the various configuration modes for the autothrottle system (ATS), the anti-ice system, and the monitoring of altitude and speed and their relation to aircraft power status.

 058/2005-ODF1: Boeing study the possibility of designing a new algorithm or of reviewing (with a view to improving) the existing algorithms in these aircraft or systems which trigger alerts or warnings so that they provide sufficient time for flight crew, firstly to recognise in good time aural and/or visual indications of any abnormal or hazardous situation, and secondly also to react in time to such signals or warnings and then carry out a rapid and appropriate analysis and decision-making process. In this particular case, the present Civil Aviation Accident Investigation Board suggests that an additional warning, both aural and visual (e.g. lights and a voice saying “Warning: Performance”, “Warning: Performance Conflict”, etc.), in what is determined to be sufficient time, could alert the crew and put it in a state of situational awareness in a more appropriate manner, and initiate more timely corrective action in order to avert this type of accident. We, therefore, recommend analysis of the inclusion of an additional audio-visual warning appropriate to the situation detailed here and to the causal factors of this accident.

1.2 Serious incident on 4 June 2002 to the DC-9-82 (MD-82) operated by Spirit Airlines. The investigation recalls that the NTSB, at the end of the safety investigation, recommended that:

 A-04-034: the FAA issue a flight standards information bulletin to principal operations inspectors to alert all affected air carrier flight crews about the icing situation encountered by Spirit Airlines flight 970 and to emphasize the need to maintain vigilance for the signs of high-altitude icing conditions, the effect these conditions can have on airplane and engine performance, and the need for the appropriate use of the engine anti-ice system.

 A-04-035: the FAA actively pursue research with airplane and engine manufacturers and other industry personnel to develop an ice detector that would alert pilots of air inlet pressure sensor icing.

1.3 Accident on 1 June 2009 to the Airbus A330-203 operated by Air France The investigation recalls that the BEA, at the end of the safety investigation, recommended that:

FRAN-2012-032: ICAO ensures the implementation of SAR coordination plans or regional protocols covering all of the maritime or remote areas for which international coordination would be required in the application of SAR procedures, including in the South Atlantic area.

2. Interim Recommendations

The investigation confirmed that at the time of writing, documentation such as the AFM (Aircraft Flight Manual) does not contain specific procedures to enable crews, from the indicated engine parameters, to rapidly highlight a situation of EPR inconsistency resulting from an obstruction of the nose cone pressure sensor. Detection of this inconsistency would enable flight crews to react before the aeroplane came dangerously close to a stall situation.

Thus the Commission d’Enquête sur les Accidents et Incidents d’Aviation Civile du Mali and the BEA recommended that:

– the FAA, as primary certification authority, or if not, EASA, require a modification MD80 type flight manual to :

– draw crews’ attention to the risks linked to possible icing of the PT2 pressure sensor at cruise altitude including where there are no visible signs of icing, specifically, when the engine anti-icing system is not activated.

[Recommandation MLI-2015-002] [Recommandation MLI-2015-003]

[Recommandation FRAN-2015-014] [Recommandation FRAN-2015-015]

– provide them with the means to rapidly detect an erroneous EPR indication and to remedy it. [Recommandation MLI-2015-004]

[Recommandation MLI- 2015-005]

[Recommandation FRAN-2015-016]

[Recommandation FRAN-2015-017]

– that the FAA and EASA study the need for similar action for other aeroplanes equipped with engines using the same thrust management principles.

[Recommandation MLI-2015-006] [Recommandation MLI-2015-007]

[Recommandation FRAN-2015-018] [Recommandation FRAN-2015-019]

3 New Safety Recommendations

3.1 Anti-icing protection devices

In accordance with the current FCOM procedures, crews, though made aware of the risks of classic icing, can be unaware of the precursor signs (or their absence) associated with icing through ice crystals, and be unaware of the possibility of obstruction of the PT2 pressure sensors and their related consequences in cruise. Due to the impact of the engine anti-icing system on the performance of the MD80 in cruise, crews may be tempted only to activate the anti-icing system when they expect to encounter icing conditions. Icing of the PT2 sensor can occur, when the anti-icing devices are not activated, in the absence of any other sign of icing on the airframe or on the windshield wipers or the engines, even in situations of low concentration of ice crystals which may not be readily visible, and may not be detectable on the weather radar

Consequently the Commission d’Enquête sur les Accidents et Incidents d’Aviation Civile du Mali and the BEA recommend that:

 the FAA ask the manufacturer to study the feasibility of installing a permanent anti-icing system for the PT2 sensors, independent of any activation by the crew of existing anti-icing systems for the engines or the airframe.

[Recommandation MLI-2016-001] [Recommandation FRAN-2016-022]

 pending the introduction of such a system, the FAA require that the manufacturer’s FCOM procedures related to “Engine Anti-Ice on Ground and in Flight” mention the difficulties of ice crystals detection in particular at night and define clearly the associated criteria for the anti-icing systems activation.

This may require the systematic activation of these systems in flight as soon as the total temperature is below 6°C, without taking into account the criteria of visible humidity or signs of ice accretion on the windshield wipers.

[Recommandation MLI-2016-002] [Recommandation FRAN-2016-023]

3.2 Stall in cruise on MD80 type aeroplanes and associated training

On the MD 80 buffet may appear late and may not be clearly noticeable. In addition, the stall warning systems such as the stick shaker and the stall warning in cruise are triggered in flight with a warning period that leaves little or no time for the crew to react before stall actually occurs. Further, the autopilot can remain engaged until a true stall and even beyond, which then aggravates the stall situation and increases the difficulty of recovery by the crew.

Test on a training simulator suggests that simulators may not be sufficiently representative as regards to the triggering of approach to stall warnings.

Consequently the Commission d’Enquête sur les Accidents et Incidents d’Aviation Civile du Mali and the BEA recommend that;

 the FAA require that the manufacturer integrates into the documentation provided to operators the specific features of a stall in cruise on MD 80 type aeroplanes, linked to the late appearance of buffet, of the stick shaker and of the stall warning and with the non-automatic disengagement of the autopilot after the stall warning.

[Recommandation MLI-2016-003] [Recommandation FRAN-2016-024]

 the FAA and EASA require that these specific features of MD 80 type aeroplanes be taught during type rating and recurrent crew training.

[Recommandation MLI-2016-004] [Recommandation MLI-2016-005]

[Recommandation FRAN-2016-025] [Recommandation FRAN-2016-026]

 the FAA ensure that the data made available by the manufacturer Boeing for the design of MD80 simulators be representative as regards the triggering of approach to stall warnings and the non-disengagement of the autopilot after the stall, at low altitude and at cruise level.

[Recommandation MLI-2016-006]

[Recommandation FRAN-2016-027]

 the FAA and EASA ensure the simulators used for training MD80 crews, be representative of the triggering of approach to stall warnings, and the non-disengagement of the autopilot after the stall, at low altitude and at cruise level.

[Recommandation MLI-2016-007] [Recommandation MLI-2016-008]

[Recommandation FRAN-2016-028] [Recommandation FRAN-2016-029]

 the FAA require that the manufacturer Boeing studies the feasibility of a modification of the autopilot’s engagement logic to allow automatic disengagement when the approach to stall is detected by the MD80 systems.

[Recommandation MLI-2016-009] [Recommandation FRAN-2016-030]

3.3 CVR

The pre-flight test performed by crews does not make it possible to detect the two major malfunctions of the CVR in this accident: the failure of the erase function and the failure to record cockpit sounds on the CAM. These two malfunctions deprived the safety investigation of essential information to understand the event and to make recommendations to improve flight safety.

A CVR maintenance check is performed at each type C check, every 3,600 flying hours or every 15 months at the latest. This procedure allows a malfunction of the erase function to be detected but does not check a failure to record cockpit noises on the CAM track.

In addition, ICAO Annex 6 relating to the technical operation of aircraft states that magnetic tape CVRs shall no longer be used after 1 January 2016 for international commercial air transport activities.

Nevertheless, pending the enforcement of these provisions in national or European regulations, the Commission d’Enquête sur les Accidents et Incidents d’Aviation Civile du Mali and the BEA recommend that;

 the FAA ensure that the maintenance check procedure for CVRs on MD80 type aeroplanes is modified by the manufacturer in order to ensure that all of the recording tracks are tested, including the CAM track.

[Recommandation MLI-2016-010]

[Recommandation FRAN-2016-031]

 the FAA and EASA ensure that this modification is implemented by the operators concerned. [Recommandation MLI-2016-011]

[Recommandation MLI-2016-012]

[Recommandation FRAN-2016-032]

[Recommandation FRAN-2016-033]

3.4 SAR

The investigation showed that there was no coordination between the Niamey RCC and the Ouagadougou and Bamako RSC, whose areas of responsibility bordered on the accident area.

Inadequate resources and a lack of practice at the Niamey RCC meant that rapid integration of available information was not possible.

Consequently the Commission d’Enquête sur les Accidents et Incidents d’Aviation Civile du Mali and the BEA recommend that;

 the national authorities of Niger, Burkina-Faso and Mali put in place coordination plans and ensure that their effectiveness is validated by regular exercises.

[Recommandation MLI-2016-013]

[Recommandation MLI-2016-014]

[Recommandation MLI-2016-015]

[Recommandation FRAN-2016-034]

[Recommandation FRAN-2016-035]

[Recommandation FRAN-2016-036]

Excerpted from BEA full final report: https://www.bea.aero/en/investigation-reports/notified-events/detail/event/accident-of-a-mcdonnell-douglas-dc-9-83-md-83-registered-ec-ltv-operated-by-swiftair-sa-flight-ah-5017-on-july-24-2014-in-the-gossi-area-mali/

**********************

Posts in this blog:

Human Factors in Aviation

Accidents/Incidents investigation

Don’t forget your email address registration to receive instant notifications. Follow me on facebook Living Safely with Human Error  and twitter@dralaurita. Lots of Human Factors information updated almost every day.

14 thoughts on “Swiftair MD83 Loss Of Control In flight final report

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s