ERA22FA114 · Cessna 310 R

The pilot was performing an aerial survey flight, and after completing a preflight inspection, he taxied toward the runway for engine run-up and surveying computer start-up. During taxi and the subsequent run-up, the airplane was positioned for about 8-10 minutes with a quartering tailwind. Track data revealed that shortly after takeoff, the airplane’s climb rate decreased, and its acceleration stopped. Shortly thereafter, the airplane began a 10°-bank-angle left turn at an airspeed of about 136 knots, followed by a rapidly descending right turn and impact with terrain. Postaccident examination of the wreckage revealed that the left fuel tank selector handle was in the OFF position, the left throttle was near idle, the left propeller control was near the feather position, and the rudder was trimmed to the right. These control positions were consistent with the left engine being partially secured, which would result in a lack of power and the loss of climb rate noted shortly after takeoff. Additionally, the right fuel tank selector handle was found in the left main fuel tank position. The examination of both engines revealed no evidence of any preimpact anomalies or malfunctions that would have precluded normal operation, and no reason for why the pilot might have partially secured the left engine. In the event of an engine failure during takeoff, the airplane manufacturer’s Pilot’s Operating Handbook (POH) assumes that the inoperative propeller is feathered and that 5° of bank toward the operating engine is used to balance the side force generated by a full rudder input. If these conditions do not exist, the airplane can quickly become uncontrollable at airspeeds much higher than the published single-engine minimum controllable airspeed (Vmc). The physical evidence, along with a performance analysis of the airplane’s flight track, showed that the left engine was not fully secured, the right engine fuel selector was set to the left tank, and the airplane banked 10° into the inoperative engine at an airspeed of about 136 kt shortly before the airplane entered a steep, descending right turn. This turn toward the inoperative engine would have dramatically increased the airplane’s minimum controllable airspeed above that assumed by the POH (80 knots), and the pilot's ability to maintain control of the airplane would have been significantly reduced. It is likely that during this left turn, the pilot allowed the airplane's airspeed to decrease below a speed for which the airplane would have been controllable, which resulted in a loss of control and led to the airplane's roll to the right and rapid descent toward the terrain.

Postaccident toxicological testing performed by a state office of forensic science revealed that the pilot’s carboxyhemoglobin, a marker of carbon monoxide (CO) exposure, was elevated at 31%. Although the Federal Aviation Administration Forensic Sciences Laboratory toxicology results did not show elevated carboxyhemoglobin, these test results might have been misleadingly low if there was an actual postmortem decrease of carboxyhemoglobin in the tested blood. This may have occurred if the specimens were obtained from a collection site where blood intermixed with gastric acid. The carboxyhemoglobin percentage measured in the blood specimen tested by the state forensic science office was confirmed by a second distinct technique, and the probability is small that the elevated result was attributable to postmortem changes. Examination of the airplane’s combustion heater assembly revealed no defects that could have allowed the combustion biproducts to intermix with the ventilation air, and examination of the wreckage revealed no evidence of inflight or post-impact fire. A postaccident test with an exemplar airplane (the same make/model as the accident airplane) that was equipped with an electronic CO detector revealed that when taxiing and performing an engine run-up with a quartering tailwind, the exhaust from the left engine was able to penetrate the cockpit. Based on the observations from this test, it is possible that engine exhaust gasses containing CO could have entered the cockpit while the pilot was conducting his pre-takeoff tasks. Given that the airplane was equipped only with a disposable “spot” CO detector, the pilot would not have been alerted to increasing CO levels unless he had looked at the device and observed a color change. Given that the temperature on the day of the accident was 33° F, it is likely that the airplane’s heater was operating. It is possible that its fan could have drawn additional air containing engine exhaust gasses and CO into the cabin heater air intake, and then into the cockpit, which would have increased the pilot’s the level of CO exposure. No other source of abnormal CO was identified. Based on available operational and physical evidence, it is likely that the pilot was impaired due to CO exposure. It is possible that this impairment could have resulted in his perception of a left engine problem, and resulted in him partially securing it, as demonstrated by the postaccident positions of the engine controls. Ultimately, the turn into the partially secured engine resulted in a loss of control and impact with terrain.

Verbatim NTSB analysis and probable cause from the NTSB dataset