N729MSBeech B100 · 1976

On October 26, 2009, approximately 1143 central daylight time, a Beech B100 King Air, N729MS, registered to Mazak Properties, Incorporated, and operated by a private pilot, impacted terrain after encountering severe weather near Benavides, Texas. The personal flight was operated under the provisions of Title 14 Code of Federal Regulations Part 91 and an instrument flight rules (IFR) flight plan was filed. Instrument meteorological conditions prevailed. The private pilot and three passengers received fatal injuries and the airplane was destroyed. The flight departed Garner Field Airport (KUVA) Uvalde, Texas, and was destined for Leesburg International Airport (KLEE), Leesburg, Florida.

Prior to departure, the pilot received three separate weather briefings from an Automated Flight Service Station (AFSS). A review of these communications revealed that the pilot was provided information regarding the severe weather conditions along his planned route of flight. The pilot expressed his concern about these conditions and altered his route of flight further south so he could maneuver around and through “holes” in the weather.

A review of air traffic control (ATC) communications revealed that the accident airplane departed from Uvalde, Texas, about 1058. The pilot contacted Houston Air Route Traffic Control Center (ARTCC) sector 59 at 1105:08, and reported that he was at an altitude of 10,200 feet and climbing to 23,000 feet. He also continued to say, "All right if you don't mind helping us we're looking at the radar – it might be better for us to go down toward Laredo <unintelligible> route…looks like a squall line." The controller responded, "Niner mike sierra yes there is a very significant squall line between you and your destination. Not sure how you'll get through, but we'll work on it somehow." The pilot responded, "All right – I sure appreciate the help."

At 1107:44, the pilot requested to proceed direct to Laredo. The controller responded, "…direct Laredo, direct Corpus, direct Leeville, rest of route unchanged." The pilot read back the route, and the controller cleared the accident airplane to climb to 25,000 feet, which was the pilot's requested final altitude.

At 1122:23, the pilot stated that he had a request, and the controller acknowledged. The pilot continued, "uh, we're looking at a hole um going towards Corpus is it possible we could get about a one five zero degree heading try and work through that way <unintelligible>. The controller replied, "November niner mike sierra fly heading of one two zero when able proceed direct Corpus Christi rest of route unchanged." The pilot responded, "…sounds great one two zero and when able direct Corpus for mike sierra thank you."

At 1140:43, the controller transmitted, "November niner mike sierra verify you're level flight level two five zero." The pilot replied, "No sir, we had dropped down to – we'd gotten into some pretty good turbulence we're at two four zero." The controller responded, "November niner mike sierra you've got to tell me these things…I've got a MOA below you at flight level two three zero and below so november niner mike sierra when able maintain flight level two five zero." The pilot transmitted, "Yes sir um I'm sorry sir.'" The controller then asked,"…and uh any injuries or uh damage?" The pilot did not reply and there were no further communications with him.

At 1142:07, an expletive and propeller noise were heard on the same frequency used by the pilot, and for the next 35 seconds there was sound similar to that made by a stuck microphone conflicting with other transmissions. In response, the controller transmitted, "All right now everybody use caution – we had a stuck mic there for a while, that's very bad now."

A review of recorded radar data revealed that after departing Uvalde, the airplane traveled on a southeasterly heading before it made a turn to the south at 1108. The airplane then traveled on this southerly heading for approximately 15 minutes before it turned back to a southeasterly heading toward Benavides, Texas. At 1140, when the airplane was approximately 6 ½ miles north of Benavides at an altitude of 25,000 feet, it entered a right turn toward the northwest and started to descend. The last radar return was received at 1141:46 at an altitude of 22,200 feet. The radar track went into "coast" status at 1142:55, indicating that the radar data processing system had lost track of the aircraft.

A review of the National Weather Service (NWS) Corpus Christi, Texas, Weather Surveillance Radar-1988, Doppler (WSR-88D) weather and radar track data revealed that the airplane entered a line of echoes in the range of 40 to 54 dBZ (very heavy to intense) before it began to lose altitude and reverse course. The Federal Aviation Administration (FAA) defines reflectivity of 50 dBZ and greater as “extreme” intensity with severe turbulence, lightning, hail likely, and organized surface wind gusts. Stronger echoes with maximum reflectivies to 63 dBZ or “extreme” intensity were located in the immediate vicinity along the intended route of flight.

The pilot held a private pilot certificate for airplane single and multi-engine land, and instrument airplane. His last Federal Aviation Administration (FAA) second class medical was issued on January 10, 2008. At that time, he reported a total of 550 flight hours. The pilot’s logbook was never located.

The closest weather reporting facility to the accident was from Jim Hogg County Airport (KHBV), Hebbronville, Texas, located approximately 24 miles southwest of the accident site at an elevation of 663 feet. The airport had an Automated Weather Observation System (AWOS-3) and reported the following conditions surrounding the time of the accident:

At 1103, the weather was reported as wind from 210 degrees at 8 knots, visibility 7 miles, scattered clouds at 1,600 feet, ceilings broken at 2,200 feet, overcast at 7,000 feet, temperature 26 degrees Celsius, dew point 22 degrees Celsius, and a barometric pressure setting of 29.91 inches of Mercury. Remarks: automated observation system, lightning distant north, northeast, and southwest, temperature 25.9 degrees Celsius, dew point 21.5 degrees Celsius.

At 1125, the weather was reported as wind from 310 degrees at 18 knots gusting to 26 knots, visibility 5 miles in moderate rain, scattered clouds at 1,000 feet, ceiling broken at 1,500 feet, overcast at 2,800 feet, temperature 20 degrees Celsius, dew point 18 degrees Celsius, and a barometric pressure setting of 29.95 inches of Mercury. Remarks: automated observation system, lightning distant northeast, hourly precipitation 0.19 inches, temperature 19.7 degrees Celsius, dew point 17.6 degrees Celsius.

At 1143, the weather was reported as wind from 330 degrees at 21 knots gusting to 31 knots, visibility 4 miles in thunderstorms and heavy rain, scattered clouds at 800 feet, ceiling broken at 2,300 feet, overcast at 3,000 feet, temperature 18 degrees Celsius, dew point 17 degrees Celsius,

and a barometric pressure setting of 29.94 inches of Mercury. Remarks: automated observation system, lightning distant north east and east, hourly precipitation 0.24 inches, temperature 17.7 degrees Celsius, dew point 16.5 degrees Celsius.

At 1203, the weather was reported as wind from 350 degrees at 22 knots gusting to 32 knots, visibility 10 miles, scattered clouds at 800 feet, scattered at 2,300 feet, ceiling overcast at 3,000 feet, temperature 18 degrees Celsius, dew point 16 degrees Celsius, and a barometric pressure setting of 29.95 inches of Mercury. Remarks: automated observation system, lightning distant northeast through southeast,temperature 17.6 degrees Celsius, dew point 16.4 degrees Celsius.

The NWS Corpus Chisti (KCRP) upper air sounding for 0700 indicated an unstable atmosphere with a Lifted Index of -6.8 and a K-Index of 31.6, which supported thunderstorm development across the region. The degree of instability and vertical wind profile also supported a moderate to strong risk of multicellular type thunderstorms in lines and clusters, with the potential for strong updrafts and downdrafts. The tropopause was identified at 54,000 feet.

At the airplane's cruising altitude, the WSR-88D 2.4° elevation scan documented echoes of 40 to 49 dBZ along the flight track when the pilot lost altitude and reported encountering turbulence, and turned back into an area of 40 dBZ when the upset occurred. Echo tops of 18 dBZ or higher

were detected between 30,000 to 34,000 feet over the accident site, with tops to 45,000 feet approximately 10 miles south-southeast of the accident site.

The National Lightning Detection Network (NLDN) detected 140 cloud-to-ground lightning strikes between 1122 and 1202 within 15 miles of the accident site, and confirmed that active thunderstorms were occurring in the vicinity of the accident site.

The NWS had Convective SIGMET 38C current over the area for an area of thunderstorms moving southeast at 20 knots, with tops to 37,000 feet. The advisory implied severe to extreme turbulence, severe icing, and low-level wind shear associated with the thunderstorms.

The NWS also had AIRMET Zulu current for moderate icing conditions from the freezing level identified at 13,500 feet to 25,000 feet over the area.

The National Transportation Safety Board’s on-scene investigation was conducted on October 27-29, 2009.

A hand held global position receiver (GPS) was used to identify the position of the main wreckage as 27.6663 degrees north latitude, 98.4613 degrees west longitude. The main wreckage was located approximately 6 1/2-miles north of Benavides, Texas, on a remote cattle ranch. The associated debris was located in two general areas, situated about 0.8 nautical miles from each other. The overall wreckage distribution was consistent with an in-flight break-up.

The main wreckage was located in a shallow impact crater that measured approximately 30-feet-wide. Damage to the dense overhead foliage surrounding the wreckage was limited, consistent with a near vertical descent path. The main wreckage was inverted and consisted of the cockpit, both engines, inboard sections of each wing, both propeller systems, the fuselage, all three landing gear, and the upper portion of the right outboard wing. The main wreckage was consumed by post-impact fire. The second area of wreckage was scattered in a centralized area south of where the main wreckage came to rest. It consisted of the outboard section of left wing, sections of the upper portion of the right wing, sections of the vertical stabilizer, both horizontal stabilizers, both elevators, and the rudder.

The wreckage was recovered and moved to a secured facility where it was examined under the supervision of the Investigator-in-Charge on November 4-5, 2009. A Safety Board Structural Engineer, along with a representative of Hawker Beechcraft and Honeywell also participated in the examination/reconstruction.

Examination of the recovered sections of the outboard section of wings, vertical stabilizer, rudder, horizontal stabilizer, and elevators revealed that all of the fractures examined exhibited signs consistent with overstress failure. There was no evidence of pre-existing cracking on any of the fracture surfaces examined.

The left engine was still contained in the nacelle hardware and exhibited heavy impact damage to the upper portion of the nacelle. The nacelle sheet metal was formed around the upper portion of the engine and mount structure. The upper cowl sheet metal was removed to access the engine for detailed examination. The engine hardware aft of the gearbox accessory face was blackened with soot. The upper mount was bent down toward the top of the power section. The starter was in place on the aft side of the gearbox but the starter mount flange was fractured.

The engine was not removed from the lower nacelle and mount hardware.

The gearbox housing had impact damage to the left side of the top face and was cracked. The gearbox was also fractured in the inlet flow path area. The fuel pump housing was fractured in the fuel filter area and opened at the center split line between the low pressure and high pressure stages. The fuel control was impact damaged on top of the unit. The propeller pitch control and propeller governor were in place and their housings were intact. The fuel shutoff valve had impact damage and the lever arm was separated. Inspecting up the inlet of the engine, the leading edges of the impeller vanes appeared to be intact and undamaged. The power section rotating group was not free to rotate, as well as, the propeller and propeller shaft.

The starter spline drive shaft was intact and was not sheared.

The oil filter bypass indicator was not extended.

The plenum exhibited impact damage on the top side of the part. The igniters were bent toward the bottom of the part and the plenum skin was torn around them. Several of the fuel manifold hoses were damaged by fire. The engine exhaust was not removed to inspect the turbine end of the engine. Several of the fuel manifold hoses were damaged by fire.

Examination of the left propeller revealed that the forward housing was separated from the threads on the forward side of the hub. The three blades still attached to the hub were in a coarse blade angle position. The damage to all four blades is as follows:

1 blade was separated at the grip area. The separated blade was bent approximately 1 foot from the grip area.

1 blade had a slight bend approximately 1 foot from the tip.

1 blade was bent toward the front side of the blade.

1 blade was bent toward the aft side of the blade.

The right engine was still contained in the nacelle hardware and there was heavy impact damage to the upper portion of the nacelle. The nacelle sheet metal was formed around the upper portion of the engine and mount structure. The upper cowl sheet metal was removed to access the engine for detailed examination. The engine hardware aft of the gearbox accessory face was blackened with what appeared to be soot. The amount of soot increased further aft on the top of the engine. The upper mount was separated at the forward end and was bent down against the top of the engine. The starter was separated at the mounting flange and was lying against the upper left portion of the power section. Several of the fuel manifold hoses were damaged by fire.

The engine was not removed from the lower nacelle and mount hardware.

The starter drive-spline shaft was intact and was not sheared, but was bent.

The oil filter bypass indicator was not extended.

The gearbox housing had impact damage to the left side of the top face and was cracked. The fuel pump and fuel control had been pushed down toward the engine and had pulled the fuel pump mounting studs from the back side of the gearbox housing. The gearbox to fuel pump drive shaft was intact and in place but was slightly bent down toward the power section of the engine. The fuel pump to fuel control flanges were broken, exposing the fuel control drive splines. The fuel control drive-spline shaft appeared to be intact. The prop pitch housing appeared to be intact and the mounting flange did not appear broken. The prop governor mounting flange was fractured. The fuel shutoff valve was impact damaged and fire damaged. Inspecting up the inlet of the engine, the leading edges of the impeller vanes appeared to be intact and undamaged. The power section rotating group was not free to rotate as well as the propeller and propeller shaft.

The second stage compressor housing had wrinkles in the housing between the flanges. The engine plenum housing exhibited impact damage on most of the top side of the part. Several of the fuel manifold hoses were damaged by fire.

Examination of the right propeller revealed that the spinner was impact damaged on one side and formed around one side of the propeller hub. One blade counterweight had an impact location through the spinner that indicates a coarse blade angle at the time of impact. After removal of the spinner housing, the right propeller hub was intact. The pushrods were intact and connected to the blades and piston. All four blades were in a coarse pitch position. The damage to all four blades is as follows:

1 blade appeared to be undamaged.

1 blade had a slight dent in the trailing edge.

1 blade was bent forward.

1 blade was bent aft.

A Safety Board Air Traffic Control Specialist convened an ATC group on November 3, 2009, at the Houston Air Route Traffic Control Center. The group met with facility staff, conducted interviews, and reviewed recorded radio and radar data and training materials.

In an interview, the controller-in-charge of the accident flight reported that he had just returned back to work after having two days off and was not dealing with any personnel issues. In addition, there was no equipment or operational issues apparent except for the severe weather affecting the area. The controller was asked to review the Satori replay of the accident sequence.

The controller was assigned to work sector 59, which he noted was handling oceanic traffic that would not normally be routed through the area. He did receive a position-relief-briefing when he took over, but it did not include information about the unusual traffic flow. His recollection was that the routing from Laredo to Corpus Christi was largely clear of precipitation, with most of the weather being located to the north of that route.

In regards to his handling of the accident airplane, the controller recalled when the pilot requested a 150 heading toward a hole in the weather. The controller responded by issuing a 120 heading and cleared the pilot to proceed direct Corpus Christi when able. The controller explained that at the time the pilot requested the heading change, there was a large hole in the line of weather that he believed the airplane could pass through safely and would have no further weather issues because it was a clear route. The pilot gave a positive response and the controller took that to indicate that the pilot also saw the same hole in the weather.

Around this time there was also a Global Hawk unmanned aerial vehicle (UAV) in the sector with an unusual route in its flight plan that the controller believed was incorrect. He was confused about the UAV’s purpose of flight so it took him additional time and effort to correct the problem.

It was also noted that on a few occasions, the controller issued pilots clearances direct to fixes that were not part of their filed route of flight. He explained that this was because he misunderstood what clearances had been issued to those aircraft by preceding sectors in response to the turbulence reports in sector 59.

In regard to advising pilots about weather deviations, the controller was asked if they are approved to be equivalent to providing detailed weather information to the pilot as required by the controller handbook. The controller said his objective is to get the aircraft safely past the weather. He will also provide a complete description of the weather if he believed that a pilot was having a hard time getting around it, or appeared to be turning toward the weather instead of away from it. If a pilot appeared to be aware of the weather ahead and the controller believed that he was proceeding appropriately, the controller would not necessarily spend the 10 or 15 seconds per aircraft necessary to provide the detailed weather information specified in the controller handbook. The controller was asked to explain his understanding of the weather services requirements of the controller handbook. He stated that it was one of the services that controllers provide after separation and safety alerts. Asked when he felt obliged to use the specified phraseology and descriptive terms in the controller handbook, the controller stated that he does so when it was to his advantage, or if the pilot appeared to be deviating toward weather that may be a hazard. When asked about his understanding of the term "pertinent weather," the controller stated that it applied to weather that would have an effect on the aircraft's flight…weather that the pilot would want to know about.

The controller was queried about his initial discussion with the accident pilot in which he described a squall line ahead of the flight and said that he did not know how the pilot would be getting through it. The controller said that at that point he did not recognize the

identifier for the aircraft's destination and was uncertain whether the flight would best be routed north or south of the line of weather. When the pilot requested direct Laredo, the controller took that as an indication that it would be best for the pilot to proceed south of the

squall line.

The controller was then shown the archived displayed weather data provided by the FAA Technical Center, which did not depict a hole through the line of weather at the point that he initially described. He reiterated that he had seen a gap in the line, and that he believed that the 120 heading he issued would have put the accident airplane through a hole. The controller stated that the recorded weather information showing no hole in the line was incorrect. He confirmed that he had selected NEXRAD intensities 1, 2, and 3, with altitude filters 000B600. He stated that he configured his scope to have a blue background, which would have caused the displayed weather to appear darker than the background. Moderate weather would have appeared to him as solid black on the display. The controller vectored the airplane toward the hole that he saw and received a read-back from the pilot sounding like the pilot agreed with the routing. At that point, he focused most of his attention on other aircraft, but continued to periodically monitor the accident airplane. The controller recognized that it was about 20 minutes from the time that he issued the vector until the airplane disappeared from radar, and he did not recall any changes in the weather during that period.

The controller said that he immediately noticed when the accident airplane left its assigned altitude and contacted the pilot. The pilot responded that he had encountered turbulence. The controller said that this was the first time he noticed the airplane was operating in weather. The airplane was descending through 22,000 feet in an area where he should have been able to maintain radar contact down to about 2,500 feet. He knew there was something wrong and noticed that the airplane appeared to be in precipitation. The controller said that he missed the expletive apparently transmitted by the pilot after the upset and did not associate the stuck microphone on frequency with the accident airplane at the time that it occurred. He just recognized that it was a bad time to have a stuck microphone on frequency.

Additional investigation revealed that the controller was repeatedly transposing call signs of various aircraft under his control. He stated that he has experienced that problem at various points throughout his career and was not aware of any vision issues or other difficulties that may have caused him to transpose call signs. The controller was required to wear corrective lenses for distant vision; however, he was not wearing his glasses at the time he was handling N729MS.

Upon completion of this investigation, it was concluded that the ATC services provided to N729MS were not in compliance with FAA requirements per FAA order 7110.65, paragraph 2-6-4, “Weather and Chaff Services.” Although the pilot requested assistance in avoiding the “squall line” noted by the controller as being ahead of the airplane, the controller did not provide the pilot with the information as required. Review of recorded precipitation data showed that there was substantial information available to the controller about moderate to extreme weather along the airplane’s route of flight. While the controller stated that he saw a hole or clear area ahead of the aircraft, this is contradicted by both the recorded data and the statement of a second controller working the D-position at the time of the accident.

An autopsy on the pilot was conducted on October 28, 2009, by the Nueces County Medical Examiner, in Corpus Christie, Texas. The cause of death was determined to be "multiple blunt trauma.”

Toxicological testing was conducted on the pilot by the FAA’s Medical Laboratory in Oklahoma City, Oklahoma. The findings were negative for drugs and alcohol.

The airplane was self-insured and the wreckage was released to the pilot’s attorney on December 21, 2009.

The aircraft was equipped with an Enhanced Ground Proximity Warning System (EGPWS). The unit was found in the nose avionics area of the aircraft. The housing was impact damaged and there was a small amount of what appeared to be heat damage to the housing.

Examination of the EGPWS was conducted at the facility of Honeywell Aerospace, Redmond, Washington, on February 3, 2010, under the supervision of a Safety Board investigator. The EGPWS was removed and exhibited impact related damage to the case structure, exposing the internal printed circuit boards (PCB). Heat damage and soot was also observed on the casing. The three internal PCB’s were damaged and appeared distorted. The EGPWS was disassembled and the PCB’s were removed. The flash memory chip that contained non-volatile memory was observed intact and undamaged. Due to the damage to the PCB, the flash memory chip was removed from the PCB and placed on a memory chip reader. A binary file was successfully downloaded from the flash memory chip. The downloaded binary file data was decoded using a company software program. This information was sent to the Safety Board's Flight Data Recorder (FDR) Laboratory, Washington, DC, for review.

According to a FDR Specialist, the EGPWS's non-volatile memory (NVM) does not continuously record, but rather stores data only when certain criteria are met. The readout process at the manufacturer’s facility produced several files of flight history data which encompassed operational, documentary, fault and warning information. The flight history data warning file outputs performance data as related to the operation of the aircraft. These data do not continuously record but, rather, if an alert or warning related to the EGPWS function activates, the unit retains data points for 20 seconds prior to the activation of the warning and 10 seconds afterwards. The EGPWS parameters are only sampled 1 time per second but the actual time of occurrence can be anywhere within the second. The downloaded files contain data logged based on hours of operation (operational time) of the individual EGPWS unit and have no reference to any other time base. In the data

files, each power cycle is tagged with a sequential flight leg number. The accident flight was recorded as flight leg 205 and did record a warning that was most likely triggered by a “Terrain Caution” and “Terrain Pull-up” alert.

Performance data was then calculated from radar and EGPWS returns, with some interpolation for raw data claims, which are the 40 second gaps between the end of the radar range and the EGPWS and the EGPWS and the wreckage location respectively.

At 1140:43, when the pilot made his last transmission, the airplane had an approximate groundspeed of 278.67 knots, and was descending at a rate of 1,871.5 feet per minute.

At 1142:07, when the expletive and stuck microphone was heard, the airplane had an approximate groundspeed of 139.61 knots, and was now descending at a rate of 3,794.7 feet per minute.

Over the next 26 seconds, the airplane increased its descent rate to 24,111 feet per minute before it reached a peak descent rate of 40,398.8 feet per minute before the data ended at 1143:31. At that time, the airplane’s approximate groundspeed had slowed to 102 knots and the descent rate had decreased to 13,100.4 feet per minute. The last data point was received at 27.6597 degrees north latitude and 98.4612 west longitude.</unintelligible></unintelligible>