HISTORY OF FLIGHT
On June 26, 2008, about 1756 eastern daylight time, a Cessna 310R, N4912A, registered to a private individual and operated by Marco Aviation, Inc., experienced total loss of engine power from both engines and was ditched in the Gulf of Mexico near Goodland, Florida. Visual meteorological conditions prevailed and a visual flight rules (VFR) flight plan was filed for the 14 Code of Federal Regulations (CFR) Part 135 non-scheduled, domestic, passenger flight. The flight departed from Key West International Airport (EYW), Key West, Florida, and was destined to Marco Island Airport, (MKY), Marco Island, Florida. The airplane sustained substantial damage and the certificated airline transport pilot (ATP) plus one of five passengers were not injured. Three passengers sustained minor injuries and one passenger sustained serous injury. The flight originated about 1722, from EYW.
The pilot stated that before takeoff, he did not check the fuel tanks and he did not brief the passengers on the emergency procedures. He reported the reason he did not brief them was because he had flown them earlier that day from MKY to EYW, and had briefed them before that flight. Before departure of the accident flight, the weight and balance form reflected that the fuel load was 280 pounds, which agreed with the reading on the fuel quantity indicator. The right fuel quantity was indicating a slightly higher amount than the left. He thought he had an adequate fuel supply for the flight, and operates the engines with both mixture controls in the full rich position. After takeoff, the flight climbed briefly to 3,000 feet mean sea level (msl), then descended to and remained at 2,500 feet msl, and proceeded to the destination airport. When the flight was near the destination airport, he reduced both throttles to 21 inches manifold pressure, and began descending at less than 500 feet-per-minute. Approximately 15 miles from MKY, while flying at 1,500 feet msl, the right engine "coughed a couple times" with corresponding right yaw. He noted the right fuel flow was "down" and described a split between the left and right fuel flow readings. He turned the right auxiliary fuel pump to the low position, and moved the right fuel selector to the "left main" position in an effort to restore engine power. The engine kept surging, so he feathered the right propeller. At that point he advised the passengers to don their life vests, which were located on a shelf behind the 3rd row of seats.
The pilot announced on the MKY common traffic advisory frequency (CTAF) that he was operating single-engine, and to clear the traffic pattern. Personnel of a fixed base operator (FBO) at MKY, who heard the communication immediately called 911 to request assistance, and remained in contact with the pilot. The pilot further stated that while flying single-engine, he maintained 120 to 130 knots, and stopped descending at 800 feet msl. At approximately 1758, a pilot with Collier County Sheriff Department Aviation Unit (Collier County SD) was alerted to the need for assistance, and within 4 minutes departed from Naples Municipal Airport (APF) and proceeded to the area where the airplane was reported to be located.
The accident pilot further stated that approximately 4 minutes after the right engine began losing power, and when the flight was 9 miles from MKY, the left engine began indications similar to the right engine, and then quit. The pilot announced on MKY CTAF that he would be ditching and to send help. He switched the left fuel selector to the right main tank position in an effort to restore engine power, which was unsuccessful. He recalled the left propeller did not feather, and in preparation for ditching, reviewed in his mind the ditching procedures, which were full flaps extended and gear up. He slowed to 93 knots, and just before ditching, he placed his arm in front of the 10 year old passenger seated in the copilot's seat. Also, just before ditching, the left and right fuel quantity were indicating 70 and 100 pounds, respectively.
The airplane first contacted the water with the curved portion of the bottom of the fuselage, and lunged forward, then rebounded. After coming to rest, the pilot attempted to open the cabin door but felt resistance. He then opened the emergency exit window, and was able to fully open the cabin door. All occupants exited the airplane from the cabin door, with one passenger reporting he was the last person out of the airplane because he was looking, but could not find a life vest for himself. All occupants then stayed momentarily on one of the wings, but the airplane began to submerge. They then entered the water and the pilot tried to keep everybody together, but it was difficult due to the waves. The pilot helped one passenger inflate the second chamber of her life vest, and noted that another passenger that did not have a life vest was clinging to two of his daughters.
The passenger without the life vest reported that he was clinging to his daughters for support, and contemplated pushing away from them because he was concerned about their safety trying to keep him afloat, but about that time he spotted a helicopter.
The flightcrew of the Collier County SD helicopter that proceeded to the area first spotted the occupants at 1820, or approximately 24 minutes after the ditching, and immediately flew to a nearby boater and motioned for the boater to follow the helicopter. The boater proceeded to the area and all occupants were placed in the boat. Additionally, emergency medical service personnel boarded the boat and remained with the pilot and passengers until their transfer to land, where they were then transported to a hospital. The pilot reported to a Collier County SD individual and also to one passenger, while on the boat, that he ran out of fuel. He also told the passenger that the fuel gauges were incorrect.
The airplane was ditched during daylight hours, approximately 11 nautical miles south-southeast from the center of MKY.
PERSONNEL INFORMATION
The pilot, age 65, held an ATP certificate with a rating airplane multiengine land. Numerous type ratings for transport category airplanes are listed on his ATP certificate. He also held a commercial pilot certificate with an airplane single-engine land rating, and was issued a first-class medical certificate on December 3, 2007. The medical certificate listed a limitation that the holder must have available glasses for near vision and was not valid after December 31, 2008.
The pilot verbally reported to NTSB personnel that he is a retired airline pilot and has flown general aviation type airplanes, having accrued approximately 1,500 hours in various makes and models. The pilot reported on the NTSB “Pilot/Operator Aircraft Accident/Incident Report” having a total time of over 18,000 hours, and approximately 200 hours in the accident make and model airplane.
The pilot began his initial ground training with Marco Aviation, Inc., on December 1, 2007, and completed it on January 20, 2008. He began his initial flight training on January 7, 2008, and completed it on January 16, 2008; accruing 15.9 hours. His airman competency/proficiency check in accordance with 14 CFR Part 135.293 and 135.299 occurred on February 4, 2008. He was qualified to act as pilot-in-command (PIC) in Cessna 310R airplanes, and prior to being hired had not previously flown a Cessna 310 type airplane.
Since employment, he reported accruing approximately 200 hours in the accident make and model airplane. For the month of May he logged approximately 40 hours during the course of 52 flights in the accident airplane. For the month of June, excluding the accident date, he logged approximately 29 hours during the course of 38 flights in the accident airplane. On the accident date, excluding the accident flight, he logged 3 flights in the accident airplane.
AIRCRAFT INFORMATION
The airplane was manufactured in 1978, by Cessna Aircraft Company as model 310R, and was designated serial number 310R1399. It was certificated in accordance with Civil Air Regulation (CAR) 3 and sections of 14 CFR Part 23. It was powered by two 285-horsepower Teledyne Continental Motors IO-520 engines and equipped with two McCauley constant-speed, full-manual-feathering propellers. It was also equipped with three rows of two forward facing seats in each row.
Safety Board review of CAR 3 pertaining to unusable fuel revealed that amount is the amount at which the first sign of a malfunction occurs. With respect to the fuel quantity indicator, the regulation required that it be calibrated to read zero during level flight when the quantity of fuel remaining in the tank is equal to the unusable fuel supply.
The airplane’s fuel system consists of a 51.0 gallon total capacity aluminum tank attached at each wingtip, and also a 32.0 gallon total capacity bladder tank installed in each wing. The total usable capacity of each main and auxiliary fuel tank is 50.0 and 31.5 gallons, respectively. A dual fuel quantity indicator installed in the co-pilot’s instrument panel reflects the quantity of fuel in the tank selected at the fuel selector. The fuel quantity indicating system consists of a capacitance type probe installed in each tank, associated wiring to a signal conditioner installed in each wing, and associated wiring connected to the fuel quantity indicator.
Review of the maintenance records revealed the airplane was last inspected in accordance with a 100-Hour inspection on April 24, 2008; the airplane total time at that time was 8,835.6 hours. The airplane had accumulated 78.3 hours since the last 100-Hour inspection.
Further review of the airframe maintenance records that begin with the first entry dated August 21, 1978, to the last entry dated April 24, 2008, revealed there were two entries indicating calibration of the fuel indicating system. The first entry was dated February 19, 1988, at an airplane total time of 2,707.4 hours and the second entry was dated August 21, 1995, at an airplane total time of 3,920.6 hours. There was no record of removal, repair, or replacement of either main tank probes, either signal conditioner, or fuel quantity indicator. Calibration of the fuel quantity indicating system is not periodically required, but is required when components of the system such as the probes or signal conditioners are removed and replaced. Special test equipment is required to perform the test.
METEOROLOGICAL INFORMATION
A surface observation weather report taken at Naples Municipal Airport (APF), Naples, Florida, at 1753, or approximately 3 minutes before the accident indicates the wind was from 070 degrees at 10 knots, the visibility was 10 statute miles, and few clouds existed at 3,500 feet. The temperature and dew point were 30 and 23 degrees Celsius respectively, and the altimeter setting was 30.09 inches of Mercury.
COMMUNICATIONS
While proceeding towards MKY, the pilot established contact with the MKY airport on the CTAF and advised that he was operating single-engine and requested the traffic pattern be cleared of other airplanes. An individual at an FBO at MKY broadcast on the CTAF asking the accident pilot for clarification, but the pilot did not reply. Another FBO employee broadcast on the CTAF asking the pilot what was the nature of the problem, then asked him for the number of people on-board, the amount of fuel on-board, the location, and the nature of the problem. The pilot advised the number of people on-board, and the fact that he was operating on one engine, but did not report the on-board fuel load. The FBO employee advised the pilot that he would be contacting emergency services and called 911.
The FBO employee further advised the 911 operator that the pilot of a twin-engine airplane was inbound to MKY with one engine inoperative, and requested emergency services respond to MKY. While on the phone with the 911 operator, the pilot announced on the CTAF that the other engine had quit, he was going to ditch the airplane, and the flight was located 6 miles south of MKY. The 911 operator was updated as to the location and the ditching intention.
The CTAF at MKY is not recorded.
WRECKAGE AND IMPACT INFORMATION
The airplane came to rest in approximately 7 feet of water and was located on a magnetic heading of 080 degrees. The airplane was raised using airbags and towed to shore where lifting straps were placed just outboard of each engine nacelle. Damage to both wings in the area of both straps was noted.
Examination of the airplane following recovery revealed damage to the main spar of the right wing near the wing root area. Control continuity for the primary and secondary flight controls were confirmed. All three landing gears were retracted and the flaps were extended, but the flap actuator was not measured to determine flap extension.
Examination of the airframe following recovery revealed no evidence of fuel stains at any fuel caps or fuel drains. Each main and auxiliary fuel tank were drained using the sump drain; all contained salt water. Only the left main fuel tank was found to contain any fuel; a total of 14 ounces of blue colored fuel consistent with 100 low lead aviation gasoline (100LL) were drained from that tank. Due to the fact that the airplane was submerged for several days, no attempt was made to apply power to the airplane’s electrical system. The left and right signal conditioner and each probe from the left and right main fuel tanks were removed for further examination.
Examination of the dual fuel quantity indicator in the cockpit revealed water was trapped inside the instrument. The left pointer was indicating approximately 210 pounds, while the right pointer was off scale low. No obvious impact damage was noted to the instrument or to the surrounding area of the instrument panel. The fuel quantity indicator was retained for further examination. Further examination of the cockpit revealed a portable global positioning system (GPS) receiver was mounted to the pilot’s control yoke. The GPS receiver was retained for further examination.
Examination of the cockpit and cabin revealed all seats remained attached to their respective attach points, with no observed deformation to any seat. No shoulder harnesses were installed at any seat position; all lapbelts were found released and were noted to latch and release with no discrepancies noted. No life vests were found in the pilot’s seat pocket, the cockpit, or cabin.
Examination of both engines revealed corrosion attributed to salt water submersion; no mechanical or impact damage was noted. The lower spark plugs of all cylinders were removed and the crankcases of both engines were flushed with fresh water. Fresh water was also sprayed on the exterior of both engines and engine accessories, and both magnetos of both engines were flushed inside and out with fresh water. A corrosion inhibiting spray was applied to all cylinder walls of all cylinders of both engines. Each engine-driven fuel pump was removed for visual inspection; each drive coupling was not fractured. The flexible fuel hoses in each engine compartment were all secured and intact. No fuel was found in any fuel line; however, a smell of fuel was noted when the fuel lines were removed. The fuel lines were then reinstalled. Visual inspection of the left propeller revealed all blades appeared to be in the low pitch position and exhibited varying degrees of aft bending. Visual inspection of the right propeller revealed all blades were in the feathered position. Both propellers were removed from the engines, which were then removed from the airframe.
Both engines were taken to a Federal Aviation Administration (FAA) certified repair station for operational testing with NTSB oversight. A test club propeller, appropriate for the engine make and model, was utilized for the test runs. With respect to the left engine, the spark plugs were removed and cleaned to remove corrosion, and the point of each magneto was lightly cleaned with sandpaper to remove corrosion. The engine was started and operated; however, the engine ran rough when the left magneto was grounded. The right magneto was removed and a serviceable unit was installed. The engine was restarted and operated to approximately 2,250 rpm (static rpm specification with the test club is 2,200 to 2,300). The right magneto was retained for further examination. No evidence of mechanical failure or malfunction was noted to the left engine. With respect to the right engine, the spark plugs were removed and cleaned to remove corrosion, and the point of each magneto was lightly cleaned with sandpaper to remove corrosion. The engine was rotated and produced compression, but did not start. The right magneto was removed and a serviceable unit was installed. The engine was started but failed to continue to run when the right magneto was grounded. The left magneto was also removed and a serviceable unit was installed. The engine was started and operated to approximately 2,150 rpm. Both magnetos were retained for further examination. No evidence of mechanical failure or malfunction was noted to the left engine.
SURVIVAL ASPECTS
The passenger (adult male) who remained in the airplane looking for his life vest reported that he could not find it. While the operator reported there were six life vests aboard the airplane at the time of the accident, the missing life vest was not located when the airplane was recovered. Additionally, the passenger briefing card located in the airplane depicts that flotation equipment will be stored in the pouch behind each seat. Personnel of the operator reported that they repositioned the flotation equipment to a shelf behind the rear seats because passengers were opening the sealed bags. There was no provision to secure the flotation equipment to the shelf.
TESTS AND RESEARCH
With permission of the operator and NTSB oversight, the operator’s other Cessna 310R airplane (N310BV), was utilized for testing of the accuracy of the fuel quantity indicating system. Each main fuel tank was drained using the sump drain until no fuel remained. The aircraft’s battery power was applied and the left main fuel tank quantity indicated slightly less than 10 pounds of fuel, while the right main fuel tank fuel quantity indicated 0 pounds of fuel. A total of 1.1 gallons of fuel (.1 gallon greater than the unusable amount) were added to the left main fuel tank and the fuel quantity gauge still indicated slightly less than 10 pounds remaining. Fuel was then incrementally added to the left main fuel tank bringing the total quantity in the tank to 11, 21, 31, 41, and 51.0 gallons (full). During the fueling process the fueler stopped filling the tank when it contained 50.5 gallons (.5 gallon less than full) as the level of fuel in the tank was where he would normally stop fueling. At 11, 21, 31, 41, and 51 gallons total fuel quantity, the indicator read 9, 20, 29, 37.5 and 46.6 gallons respectively. With respect to the right main fuel tank, 1 gallon of fuel was added to the tank and the fuel quantity indicated 0. Fuel was then incrementally added to the right main fuel tank bringing the total quantity in the tank to 11, 21, 31, 41, and 51.0 gallons (full). At 11, 21, 31, 41, and 51.0 gallons total fuel quantity, the indicator read approximately 10, 20, 28, 37, and 46 gallons respectively.
The right magneto from the left engine and both magnetos from the right engine were examined with NTSB oversight at an FAA certified repair station. All magnetos were placed on a test bench with no repairs and found to produce spark at all leads when operated up to 300 magneto rpm. The measured gaps of all points, and the internal timing of all magnetos were within specification. No evidence of carbon tracking was noted on the distributor block of any magneto. The capacitor of the right magneto from the left engine tested .33 microfarads (specification is .34 to .41 microfarads), while the capacitor of the left magneto from the right engine tested .30 microfarads. The capacitor of the right magneto from the right engine tested .34 microfarads (within limits).
Examination of the fuel quantity indicator, both main fuel tank probes, and both signal conditioners were performed at the manufacturer’s facility with FAA oversight. Following removal from the airplane, the components were submerged in fresh water in an effort to minimize further corrosion. Visual inspection of each signal conditioner revealed each environmental seal appeared intact. The seal screw of each was removed and the units were placed into a chamber and dried internally. The indicator was also placed in a chamber and dried internally. All components were tested in accordance with the Acceptance Test Procedure (ATP) for that respective component. The left and right tank probes tested 34.17 and 34.26 picofarads (pF) respectively. The specification is 34.5 +/- 0.50 pF. Both signal conditioners exhibited external corrosion while the left signal conditioner exhibited a bent connector. No bent pins or FOD bridging was noted on any of the pins of either signal conditioner. Testing of the signal conditioners was done without adjustment, which revealed both were within specification for supply current, the empty and full settings. Testing of the indicator revealed the left coil was in an open condition and the right coil was within limits. No determination was made as to the reason for the open condition of the left coil. Operational testing of the indicator was initiated but the unit did not function. Disassembly of the indicator revealed extensive corrosion on the inside of the indicator and inside components of the indicator. Limited range of motion of both meter movements was noted. The left signal conditioner and probe and also the right signal conditioner and probe were separately paired with an exemplar indicator for testing purposes. The testing was performed in order to determine the empty tank readings with a dry probe revealing both indicated negative 5 pounds. The manufacturer’s personnel reported that the findings of being less than zero are normal because the empty settings on the signal conditioner are adjusted on airplane and compensate for aircraft harness stray capacitance and installation effects. Since the laboratory set-up does not have those affects, a down scale shift of the reported magnitude is normal. Personnel from the manufacturer also reported that the Cessna Indicator is designed so that the pointer will indicate a "slightly below zero" reading when unpowered; therefore, if there is an open wire in left coil, the left pointer will indicate a "slightly below zero" reading regardless of the fuel quantity in the tank.
The check pilot and co-owner of Marco Aviation, Inc., reported that when the airplane is at MKY, they routinely fill each main tank, which are located at each wingtip. Personnel from the MKY Airport reported that they only have one AVGAS truck which carries 100LL fuel. Their records reflect that they last fueled the accident airplane with 100LL on June 24th. At that time a total of 39.2 gallons of fuel were added, which filled each main tank. The airplane was not fueled at MKY after June 24th because the truck was removed from service at 1700 hours on June 24th for scheduled preventative maintenance, and was not returned to MKY until 1735 hours on June 26th. The truck was originally scheduled to be returned at 1700 hours on June 25th; however, additional work was performed.
Fuel records from an FBO at EYW indicate that on the day of the accident, a total of 30.0 gallons of 100LL fuel were added. The pilot reported that the fuel was added after landing, following his first flight that day from MKY to EYW. Other than the accident date, there was no record that the accident airplane fueled at EYW in May or June 2008. Postaccident, a fuel sample was taken for testing from the same source at EYW Airport that fueled the accident airplane on the accident date. The results revealed the sample met specification for 100LL fuel.
Fuel receipts from Marco Island Airport for the months of May and June of 2008, were correlated with the Weight and Balance Sheets for the same months associated with the accident airplane. Fuel purchased at MKY for flights made between May 1, 2008, and June 22, 2008, total 2,323.5 gallons, while the airplane was operated 66.2 hours between these same dates. The calculated average fuel consumption for the last 2 month period based on the fuel purchase amount and reported flight time was approximately 35.09 gallons-per-hour (gph). While the operator reported that the fuel consumption is 35.0 gph on their typical legs, the pilot reported that he was trained the fuel consumption is 32 to 35 gph, and he flight plans 16 gallons total fuel consumption for each 35 minute flight that does not have any weather or air traffic control delay.
Further review of the Weight and Balance Sheets for May and June 2008 revealed that with the exception of the accident date, the fuel consumption for all flights from MKY to EYW was always listed as being 120 pounds, which equates to 20.0 gallons. The sheets also reflect the round trip flight duration from MKY to EYW as varying in length from 1.1 hours on May 10, 2008, 1.6 hours on May 14, 2008, 1.2 hours on May 18, 2008, 1.1 hours on May 22, 2008, 1.2 hours on June 1, 2008, 1.3 hours on June 2, 2008, 1.1 hours on June 3, 2008, 1.2 hours on June 4, 2008, 1.1 hours on June 11, 2008, 1.2 hours on June 18, 2008, and 1.4 hours on June 25, 2008.
Postaccident fuel consumption calculations were performed from the date the main tanks were last topped off (June 24th), to the accident flight. The first flight after fueling was on June 25th (the pilot was off June 23rd and June 24th), and was a round trip flight from MKY to EYW. The flight departed MKY with nearly full main fuel tanks and the total round trip flight duration was recorded to be 1.4 hours. While the Weight and Balance sheet reflects a total of 210 pounds consumed, the actual fuel consumption based on the historical amount for the previous 2 months would have been approximately 295 pounds, resulting in a landing fuel load of 305 pounds. On June 26th, the Weight and Balance sheet reflects that the fuel load was 390 pounds, which was a carry-over from the previous day, instead of 305 pounds based on the historical fuel consumption. The flight proceeded from MKY to EYW, and based on the pilot’s account lasted approximately 35 minutes. The fuel consumption for that leg based on the historical rate would be approximately 122 pounds, resulting in a landing fuel load of approximately 183 pounds. As previously mentioned, a total of 180 pounds of fuel were added and the flight departed to return to MKY with an estimated fuel load of approximately 363 pounds. The flight duration was reported to be approximately 35 minutes, and the estimated fuel load after landing based on the historical fuel burn would have been 241 pounds, though the Weight and Balance Sheet reflects the fuel load was 380 pounds. No fuel was purchased, and the flight proceeded to EYW; the flight duration was again reported to be approximately 35 minutes. The estimated fuel load after landing based on the historical rate would have been 119 pounds, though the Weight and Balance Sheet reflected the fuel load to be 280 pounds. No fuel was purchased, and the flight departed to return to MKY. Based on the historical fuel rate, 119 pounds of fuel would have allowed approximately 34 minutes of engine operation/flight time at the normal operating power settings.
Readout of the portable global positioning system (GPS) receiver was performed by NTSB personnel from the Vehicle Recorder Division. The unit which stores flight information is triggered when groundspeed exceeds 30 knots and altitude exceeds 500 feet, and ends when groundspeed drops below 30 knots for 10 minutes or more. A tracklog consisting of latitude, longitude, date, time and groundspeed is stored within the unit whenever the receiver has a lock on the GPS navigation signal. Visual inspection of the receiver revealed minor external damage. Internal inspection revealed minor damage to one ferrite coil. Continuity was confirmed for the coil. The main PC board of the receiver was placed into a surrogate case, liquid crystal display (LCD), and soft-key assembly. The unit was powered and started up normally; a total of 41 tracklogs between June 11 and June 26 were recorded. With respect to the accident date, 3 tracklogs were recorded. The first tracklog for the accident date originated near Key West, Florida, and ended near Marco Island, Florida; the duration was approximately 39 minutes. The second tracklog originated near Marco Island, Florida, and ended near Key West, Florida; the duration was approximately 39 minutes. The third tracklog associated with the accident flight began near Key West, Florida, and ended at a point located at 25 degrees 49.038 minutes North latitude and 081 degrees 38.703 minutes West longitude, or approximately 172 degrees and 10.7 nautical miles from the center of the destination airport (MKY). The flight duration was approximately 34 minutes.
Safety Board review of FAA Service Difficult Reports for the period from 1995 to July 23, 2008, associated with the airplane by make and model pertaining to the fuel system, revealed no records pertaining to the signal conditioner or fuel quantity indicator. One entry associated with another airplane (N4919A) reported chafing of the electrical wires for the left and right auxiliary fuel tank probes.