CEN14FA230

The pilot's decision to operate the banner-tow flight below the minimum safe altitude specified by federal regulation, which likely limited his ability to recover from a potential loss of airspeed due to a gusting wind condition. Also causal was the pilot’s delayed decision to release the banner, which allowed the airplane to exceed its critical angle-of-attack and resulted in an aerodynamic stall/spin from which he was unable to recover.

The pilot reported that the purpose of the banner-tow flight was to tow an advertisement billboard over a ballpark. He stated that, despite the engine operating normally, the airplane had little to no climb performance as he flew southbound toward the ballpark; therefore, he decided to make a 180-degree turn and return to the departure airport. He released the banner after the airplane developed an excessive descent rate during the turn. The pilot reported that he was unable to recover sufficient airspeed after releasing the banner and that the airplane encountered an aerodynamic stall/spin at a low altitude. The airplane then descended, inverted, into a house where a postimpact fire ensued. A postaccident examination of the airplane revealed no preimpact mechanical failures or malfunctions that would have precluded normal operation. At the time of the accident, the low-level surface winds were from the south between 5 and 10 knots with gusts reaching 20 to 30 knots. A review of radar track data confirmed that the accident airplane flew over a congested area at altitudes below 1,000 ft above ground level (agl). The final portion of the accident flight included an S-turn maneuver below 500 ft agl. According to radar data, during the S-turn, the airplane flew within 1,000 ft laterally and 200 ft vertically of the pilot's personal residence. Additionally, radar data revealed that the pilot made a tight 360-degree turn near his residence, between 600 and 700 ft agl, during another recent banner-tow flight. Based on the available radar evidence, the pilot, on at least two occasions, flew below the minimum safe altitude required by federal regulation 91.119(b), which prohibited operating an airplane, over a congested area, at an altitude below 1,000 ft above the highest obstacle within a horizontal radius of 2,000 ft. The pilot's intentional flight below the required minimum safe altitude likely limited his ability to recover from a potential loss of airspeed due to a gusting wind condition during the banner-tow flight. Additionally, his delayed decision to release the banner allowed the airplane to exceed its critical angle-of-attack, which resulted in an aerodynamic stall/spin from which he was unable to recover.

After departure, at 1514:30, the pilot established radio communications with Denver Terminal Radar Approach Control and requested visual flight rules flight following for a several hour banner-tow flight over Coors Field. At 15:14:54, the approach controller issued a transponder identification code to the pilot. At 15:17:06, the approach controller acknowledged radar contact with the accident airplane. At that time, the accident flight was located about 15 nm northwest of Denver International Airport (DEN) at 5,600 ft msl. There were no additional communications with the accident flight. A Garmin GPSMAP 96C, a JPI Engine Data Monitor, an Apple iPad, and a SatLoc device were located at the accident site and submitted to the National Transportation Safety Board (NTSB) Vehicle Recorders Laboratory in Washington D.C. for potential non-volatile memory (NVM) data recovery. An internal examination of the JPI Engine Monitor, Apple iPad, and SatLoc devices revealed extensive fire-related damage that precluded the recovery of any data. An external examination of the Garmin GPSMAP 96, s/n 98831422, revealed the postaccident fire had damaged the molded exterior case; however, the internal circuit board containing the NVM memory chip was relatively undamaged. The NVM data was extracted from the memory chip using forensic equipment. The extracted data included two flights, one completed on May 3, 2014, and the other, the accident flight, on May 5, 2014. The GPS device recorded position data every 3 minutes; including the date, time, latitude, longitude, and GPS altitude. Course and groundspeed parameters were derived from the recorded parameters. The extracted position data did not yield useful aircraft performance data because of the limited data-sampling rate (once every three minutes). A postaccident review of the pilot's flight logbook established that he had flown a similar banner-tow flight over Coors Field on May 3, 2014. A review of radar track data for the May 3, 2014, flight established that the airplane paralleled Interstate I-25 while en route to/from downtown Denver. According to plotted radar data, while returning northbound, the airplane completed a single 360-degree left turn about 0.7 nm east of where the airplane crashed on May 5, 2014. The 360-degree turn was centered over a residential area located immediately west of Interstate I-25 and between Kennedy Drive and Community Center Drive. The radius of the turn was about 0.2 nm. The 360-degree turn took about 1 minute 23 seconds to complete and the airplane's altitude varied between 6,000-6,100 ft msl (600-700 ft agl) during the turn. The pilot's personal residence was located about 1/2 nm east-southeast of where the airplane crashed on May 5, 2014. A radar plot for the May 3, 2014, flight confirmed that the pilot flew over his residence at 600 ft agl during the 360-degree left turn. Further review of radar data for the May 5, 2014, accident flight established that the pilot flew within 1,000 ft laterally and 200 ft vertically of his residence during the second half of the S-turn. Federal regulation 14 CFR Part 91.119(b), stipulates that no person may operate an aircraft over any congested area of a city, town, or settlement, or over any open air assembly of persons, at an altitude below 1,000 ft above the highest obstacle within a horizontal radius of 2,000 ft of the aircraft. On May 18, 2012, the FAA Denver Flight Standards District Office (FSDO) issued the operator, Drag 'N' Fly Banners, a Certificate of Waiver to conduct airplane banner-tow operations. One of the waiver's provisions was that the operator must maintain compliance with the conditions of Federal regulation 14 CFR Part 91.119, including the minimum safe altitudes and distances from obstacles. On August 5, 2013, the operator added the accident pilot to the waiver's list of approved pilots. On February 14, 2014, the operator added the accident airplane (N4519Y) to the waiver's list of approved airplanes. On March 24, 2014, a FAA inspection identified no discrepancies with the banner-tow equipment. The main wreckage consisted of the entire airplane, which was found inverted and nose down in a house. The engine, firewall, and propeller were located within the house. The remaining steel-tube fuselage structure, both wings, and the empennage were located outside the residence. The postimpact fire destroyed a majority of the airplane, including the cockpit, instrument panel, interior, pilot seat and fuselage/wing fabric covering. The empennage surfaces were relatively undamaged. A postaccident examination of the airplane confirmed flight control cable continuity from the cockpit controls to their respective flight control surfaces. The flight control trim position could not be determined due to fire damage. The flap control handle and its ratchet mechanism were in an extended position; however, the actual position of the flaps at the time of impact could not be determined due to impact damage. The postaccident examination revealed no evidence of a preimpact mechanical malfunction or failure that would have precluded normal airplane operation. The engine remained attached to the firewall by its mounts. The engine exterior exhibited damage from the postimpact fire. An engine control continuity check was not possible due to the extent of damage; however, all observed separations were consistent with impact related damage. Internal engine and valve train continuity was confirmed as the engine crankshaft was rotated. Compression and suction were noted on all cylinders in conjunction with crankshaft rotation. No anomalies were identified with the accessory drive gears. Both magnetos exhibited extensive thermal damage and would not rotate by hand. Further examination revealed the internal components of both magnetos had melted during the postimpact fire. The spark plugs exhibited features consistent with normal engine operation. A borescope inspection revealed no anomalies with the cylinders, valves, or pistons. A disassembly of the carburetor did not reveal any anomalies with the metal floats, needle valve, or inlet fuel screen. The two-blade propeller remained attached to the engine crankshaft flange. One blade appeared straight with leading edge damage. The other blade exhibited an aft bend about midspan, leading edge damage, and chordwise scratching. The postaccident examination revealed no evidence of a preimpact mechanical malfunction or failure that would have precluded normal engine operation. At 1536 MDT, the automated surface observing system (ASOS) located at Erie Municipal Airport (EIK), about 8 miles north-northwest of the accident site, reported: wind from 170 degrees at 10 knots, gusting 20 knots, visibility 10 miles, clear sky, temperature 27 degrees Celsius, dew point -10 degrees Celsius, and an altimeter setting of 29.78 inches of mercury. The National Weather Service (NWS) Surface Analysis Chart for 1500 MDT, depicted a low pressure system at 1000-hectopascals (hPa) over Wyoming along a frontal wave with a stationary front extending south-southeast across eastern Colorado into the Oklahoma and Texas panhandles, where a second low pressure system was located at 1001-hPa. Another high pressure system at 1011-hPa was over southwest Colorado. The resultant pressure systems resulted in a general southerly wind flow over eastern Colorado. The NWS regional radar mosaic indicated no significant weather echoes over the Denver metropolitan area. The Denver upper air sounding for 1800 MDT had a wind profile that supported mountain wave activity near 10,000 ft msl with the potential for strong up/downdrafts and moderate and greater turbulence. According to a survey of nearby weather observation stations, the low-level surface winds were from the south between 5 and 10 knots with gusts reaching 20 to 30 knots. The density altitude was about 8,500 ft, based on the weather conditions reported near the accident site. Acc