Do vacuum pump failures cause accidents? The AOPA Air Safety Foundation found 40 accidents from 1983 through 1997 involving vacuum pumps. Thirteen listed a broken vacuum pump as the cause of the accident, while most listed the failed pump as one factor contributing to the accident, but not the only one.
Were the accidents serious? Four of the 40 accidents caused no injuries, three resulted in minor injuries, and one led to serious injuries, but 32 of the accidents resulted in fatalities.
Even VFR pilots need to consider the loss of a vacuum pump as a potentially serious emergency, depending on flight conditions. The National Transportation Safety Board reports (see the Web site, www.ntsb.gov/Aviation/FTW/97A358.htm) that on September 20, 1997, a Mooney pilot noticed low vacuum-pressure readings, followed by slow failure of the attitude and heading indicators. The wing leveler system failed as well, since it, too, was a pneumatic (air-driven) system. The pilot elected to continue into deteriorating weather that became so bad he was forced below VOR reception altitudes and ended up landing in an open field. There were no injuries, although the aircraft was damaged.
Another pilot taking off at night in California in 1990 (see "Proficient Pilot: Vacuum Failure," June 1995 Pilot) wasn't so lucky. The Cessna 182 departed in VFR weather beneath an overcast sky, crossed the Pacific shoreline--resulting in loss of a visual horizon--and entered a descending spiral at 500 feet, crashing into the ocean at full power. The vacuum pump shaft was found sheared, as it is designed to do in the event of pump failure; it had probably sheared as takeoff power was applied. There was one fatality.
A look at actual incidents shows that experienced pilots are not immune to loss of control after a vacuum system failure. An ATP flying a Piper Seneca II over Oregon crashed in a steep, nose-down attitude during a no-gyro approach (see "Safety Corner," September 1985 Pilot). He had 6,500 hours total time and 500 in type.
A flight instructor with 1,270 hours total time and 256 hours of instrument time died along with his passengers when he was unable to control a Beech B55 Baron following vacuum system failure. Just before the crash, the pilot told ATC he was "in trouble, in a dive." Flight conditions included low ceilings, poor visibility, snow, and turbulence.
Unless using a simulator, most of us have never been trained in the actual way vacuum pumps fail, so the slow-death routine of most attitude indicators is at first difficult to identify. Attitude indicators die like bad actors in early Western movies, taking forever in order to draw out the scene. When simulated during training, the instructor simply reaches over and slaps a suction cup on the instrument face. Electrically powered horizontal situation indicators (HSI) and AIs, on the other hand, die rapidly. HSI failure is often accompanied by a red flag.
From the accidents reviewed here, it is obvious that cross-checking a dying attitude indicator is more difficult than it sounds, especially if the vacuum pump failure occurs in the high-workload environment associated with instrument weather, or shortly after takeoff. Clues come from the other instruments. The turn-and-bank indicator can confirm or deny what the attitude indicator or heading indicator appear to be saying about a turn. The altimeter, airspeed indicator, and vertical speed indicator--not to mention engine and wind sounds and the change in air pressure in your ears--can help confirm or deny indications of a climb or descent. It all sounds so simple when comfortably seated and reading this magazine. Without partial-panel flying practice, however, such advice is meaningless.
The pilot of a Cessna P210 and his friends were killed near Albuquerque in January after the aircraft's two vacuum pumps failed. He had told controllers at one point that he thought he had just been in a spin, and later that he had "just done a roll." He had a backup electrically driven attitude indicator, but he reported that it did not seem to be in agreement with a remotely driven compass. A witness saw a wing separate from the aircraft after it emerged from the base of a high overcast.
Maryland flight instructor Mike King suggests that when flying partial panel, "As soon as you make an adjustment, back it up." For example, if you bank, confirm that a turn has begun on the turn-and-bank indicator and check the altimeter to make sure the turn is level. When the turn is stopped, assure that it is stopped by cross-checking with other instruments, and that the plane is not changing altitude.
Help may be available from other sources in your aircraft. At Aero-Tech in Lexington, Kentucky, instructors teach students to consider the resources at hand, such as an autopilot that can control the aircraft in the Nav mode and take its information from the VOR or GPS receiver. Autopilots using turn coordinators as a reference are unaffected by a vacuum failure. An ADF with a compass card that is slaved to a magnetic heading sensor can replace the heading indicator. A GPS can supply almost instantaneous updates of the aircraft's heading during a turn.
Professional Instrument Courses teaches students the configurations and power settings needed for all phases of flight, from climb and cruise to precision approach and nonprecision approach descent, to ease the workload during partial-panel flying.
Airline and military pilots will tell you that they practice partial-panel flight and approaches infrequently, if ever. Redundant systems and backup instruments make instrument failure highly unlikely. It's possible to have backup systems even in the single-engine piston world as well, with standby vacuum systems going for less than $500. That's a reasonable price to pay to reduce the chance of a potential emergency.
