Thai Airways GE90 Incident Opens Door For Data Analytics

A Thai Airways GE90 failure opened the door for GE Aviation to leverage its data analytics to identify risk factors.

Using deep data dives to flag reliability issues is a promising tool for predictive analytics, but the ongoing probe into a GE90 engine failure shows how precise analysis can help isolate engines for checks stemming from airworthiness issues.

The FAA on Jan. 17 issued an emergency airworthiness directive (EAD) calling for removal of high-pressure turbine (HPT) interstage seals from 16 GE90-115Bs identified by serial number. The EAD and a related GE service bulletin issued days earlier stem from the ongoing probe into the October 2019 failure of a GE Aviation GE90-115B on a Thai Airways Boeing 777-300ER. The flight was departing Bangkok for Zurich when the crew rejected the takeoff while still at low speed. Debris from the damaged engine impacted the 777’s fuselage and other engine.

Soon after the incident, GE targeted eight engines operated by five airlines for HPT interstage seal checks, which the regulator mandated in October. The January action narrowed the risk population to two airlines and eight aircraft that operated similar mission profiles. A source with knowledge of the situation tells Inside MRO the common links in these incidents included operating shorter routes, using shorter runways and turning the aircraft more quickly than a typical 777 long-haul mission.

GE Aviation, citing the ongoing investigation, declined to provide details beyond what was in the FAA directives. But the specificity of each mandate points to GE using its vast vault of engine data to analyze the issue and identify common trends between the failed engine and others in the fleet.

GE used a similar approach to target CFM Leap 1Bs for inspections following a March 2019 failure of a Southwest Airlines 737 MAX being ferried from Orlando, Florida, to Victorville, California.

GE quickly linked the failure to coking—deposits of evaporated fuel and other material on fuel nozzles that lead to uneven temperature flow regions within the combustion chamber and hot spots within the high-pressure turbine. These hot spots can cause premature wear. Within hours of the failure, the company analyzed the engine’s operating history and compared it against data from each of the other 1,560 Leaps in service. 

Aware of coking, a common issue with gas turbine engines, the manufacturer had rotable pools of fuel nozzles ready for use at certain thresholds. Following the Southwest failure, GE revised its analytics and reduced those thresholds. Engines exceeding the revised limits were recommended for inspections. 

The GE examples underscore how the combination of quality data and experts who can generate actionable insights from them is proving valuable. Operators are spared unnecessary work that broader fleet mandates would require, and the manufacturers and regulators can more quickly pinpoint airworthiness issues while getting clear pictures of their scope.

Sean Broderick

Senior Air Transport & Safety Editor Sean Broderick covers aviation safety, MRO, and the airline business from Aviation Week Network's Washington, D.C. office.