How Can I Fail Thee? Let Me Count the Ways ...

  • E-Mail this Article
  • View Printable Article
  • Text size:

    • A
    • A
    • A

Recent high-profile accidents have demonstrated that even when redundant or back-up systems are present, they don't guarantee a successful outcome to a vacuum system failure. Indeed, the vacuum pump and its associated bits and pieces are some of the most failure-prone components with which pilots of piston-powered aircraft surround themselves. AVweb's Scott Puddy takes a close look at the vacuum system and its many weaknesses while offering some insights on what you can do about them.

Airborne, the originator and dominant manufacturer of dry vacuum pumps, now attaches this disclaimer to each pneumatic pump that it ships:

WARNING: FAILURE TO FOLLOW THE FOLLOWING INSTRUCTIONS MAY RESULT IN DEATH, BODILY INJURY, OR PROPERTY DAMAGE:

1. A BACK-UP PNEUMATIC POWER SOURCE FOR THE AIR DRIVEN GYROS, OR A BACK-UP ELECTRIC ATTITUDE GYRO INSTRUMENT, MUST BE INSTALLED IN ALL AIRCRAFT WHICH FLY IFR.

2. ANY INOPERATIVE AIR PUMP OR OTHER COMPONENT OF THE GYRO SYSTEM, AND ANY INOPERATIVE BACK-UP SYSTEM OR COMPONENT, MUST BE REPLACED PRIOR TO THE NEXT FLIGHT.

3. THIS PILOT SAFETY WARNING MUST BE PERMANENTLY RETAINED IN THE PILOT'S OPERATING HANDBOOK FOR THE AIRCRAFT INTO WHICH THIS AIR PUMP IS INSTALLED.

Now wait a minute! Section 91.205(d) doesn't list a backup vacuum source as required equipment for IFR operations. Why is Airborne attempting to rewrite the FARs?

For years your instructors admonished you to "trust your instruments." Why is Airborne now saying, "For God's sake, don't trust your instruments"?

The reason? Dry vacuum pumps are designed to fail abruptly and without warning. Evidently, some people (or their survivors ... or their lawyers) have determined that a statistically intentional and precipitous failure of our principal instrument system is unsporting.

Meanwhile, the National Transportation Safety Board (NTSB) has reported pneumatic system failures as a factor in an average of two fatal accidents per year over the past decade. Some of those accidents, including the 1999 crash that killed well-respected pilot and instructor Itzhak Jacoby, his wife and his daughter, or the recent one that resulted in the deaths of Missouri Governor Mel Carnahan, his son and his aide, were widely reported by AVweb and other media. Numerous other pneumatic system failures occur each year without causing accidents, incidents, or sensational headlines. Those events may have escaped your attention but are communicated to the FAA and your maintenance technician through Service Difficulty Reports (SDRs) and manufacturer's Service Bulletins.

So what's going on? How frequent and how significant are these systems breakdowns? Must you have a back-up vacuum system to fly IFR? Should you? Do you need attitude instruments for VFR? Is a standby vacuum source sufficient protection for you or is it merely an adequate means of insulating Airborne from legal liability?

A System Under Siege

Several years ago, AVweb Editor-In-Chief Mike Busch wrote two excellent articles about attitude indicators and the vacuum pumps that drive them that are well worth reading and reading again. Since then, the plot has thickened considerably in the wake of three significant Airworthiness Directives (ADs), multiple Service Bulletins, and a mountain of SDRs relating to failures of various components of air-driven instrument systems. One of these ADs was directed against ... you guessed it ... Airborne.

Defective Pumps...

The Airborne AD resulted from a series of premature (between 0-400 hours in service) "normal-mode" failures of Airborne pumps. Upon investigation, it was determined that batches of flexible couplers bearing manufacture dates of "12/97" or "5/6/98" were defective. Priority Letter AD 98-23-01, issued October 29, 1998, mandated immediate replacement of the targeted parts and restricted operations of affected aircraft to daytime VFR pending compliance. More recently, Airborne issued Service Letter 53 on July 12, 2000, relating to excessive wear of the teeth on the drive spline of its clutch-operated dry air pumps which precluded the clutch assemblies from properly engaging and disengaging.

...Defective Filters...

Airborne competitor RAPCO was likewise the subject of an AD. The FAA issued Priority Letter AD 97-16-10 on July 31, 1997, because of several reports of cracked filter housings on RAPCO's in-line pressure filters. Used principally in twin-engine aircraft that employ a pneumatic pressure system rather than a vacuum system, those filters were reported to be failing within the first two to six hours of service because the plastic was excessively brittle and subject to cracking when exposed to a low humidity/high temperature environment. More recently, Airborne issued Service Letter 56 on April 19, 2000, which required inspection of certain of its filters to remove loose particles of black elastomeric compound that may have been left behind during the manufacturing process.

...Defective Vacuum Lines...

Also during this period, Cessna issued Service Bulletin SEB 96-10 on June 7, 1996 and Beechcraft issued a safety communiqué in August 1996 relating to Imperial Eastman Hytron Redi-seal hose. Those vacuum lines failed internally without giving any visible external indications. The red inside liner could crumble, undetected, and contaminate the other vacuum system components.

...And Defective Valves

Meanwhile, even back-up vacuum systems drew FAA fire. Effective January 14, 2000, the FAA issued AD 99-24-10 relating to the very popular Precise Flight Model SVS III Standby Vacuum System which was installed on 51 models of Beechcraft airplanes, 170 models of Cessna airplanes, 12 models of Mooney airplanes, and 74 models of Piper airplanes (amongst numerous others). This is a repetitive inspection AD resulting from shuttle valve failures which prevent the standby vacuum system from operating after the primary system fails.

A Wake-up Call For General Aviation

These events are a wake-up call to general aviation. Any vacuum-driven instrument system consists of the instruments themselves, a pump, filters, valves and air lines to connect all of the above. As the events of the past few years aptly demonstrate, each of those components is life-limited and susceptible to failure.

The Advances To Failure...

Prior to recent "advances" in equipment design, GA pilots relied on external venturi tubes to power the flight control instruments. Venturi tubes were pretty much maintenance-free, devoid of moving parts and worked well, unless and until visible moisture was encountered at below-freezing temperatures. In that event, the venturi tubes were the first place to look for warnings of accumulations of ice. Fair-weather friends to be sure, but you weren't supposed to be operating your Bugsmasher II in icing conditions, anyway.

The next power source was the "wet" vacuum pump. Those engine-driven devices were functional, but inelegant and inappropriate for some applications. Wet pumps have metal vanes and were lubricated by engine oil. The pumps were long-lasting and could be overhauled once they demonstrated signs of wear. However, the wet pump's discharge air contains an oil mist requiring an air-oil separator to return most of the oil to the engine. Even so, they tended to lubricate the belly of the airplane as though in constant preparation for a gear-up landing. The oil mist also tended to deteriorate rubber deicing boots in applications where the pump powered such a system. Finally, wet pumps could not be used in positive pressure systems because the oil would contaminate the system filters and the instruments themselves. The move to dry pumps was on.

In the early 1970s, general aviation switched to dry vacuum pumps manufactured principally by Airborne and relative newcomer Sigma-Tek. Although there are differences between the two, Airborne and Sigma-Tek pumps use self-lubricating graphite vanes which are both the solution (no engine oil required for lubrication) and the problem (the pumps tend to produce spec vacuum throughout their lives and then disintegrate internally in a cloud of carbon dust). Compounding the sudden failure mode is a frangible coupling that is designed to shear abruptly in the event of overstress or sudden stoppage. In most cases, the aircraft operator will see no sign of any problem until the vacuum gage reading falls precipitously to zero.

...Warranties/Frequencies...

What is the expected lifespan of a dry vacuum pump? Although times vary from installation to installation (and everyone seems to have "horror" stories telling of dramatically shorter lifespans), some guidance is found in the warranties the pump manufacturers offer. Airborne has a two-year/1,000-hour warranty on its 215/216 series pumps (used on most non-deiced single engine planes) and a one-year/400-hour warranty on its 240-, 400- and 800-series pumps (used on deiced and larger aircraft). Sigma-Tek offers a two-year/1,000-hour warranty on its pumps. Airborne also specifies that its pumps should be routinely replaced at intervals varying from 300 to 1,200 aircraft hours depending on the unit. For non-deiced singles, the applicable replacement interval is 1,200 hours. Airborne schedules filter replacement on much shorter intervals — every year or 100 hours (whichever comes first) for the vacuum regulator filter and every year or 500 hours (whichever comes first) for the central air filter element.

...Back-Up Vacuum Sources...

As do other manufacturers, Airborne recommends (and sells) a standby electric auxiliary vacuum pump for systems having only one primary pump. The auxiliary pump connects to the primary pneumatic system through a manifold downstream of the vacuum regulator. During normal operations, the auxiliary pump is separated from the primary system by a check valve. In the event the primary pump fails, the auxiliary pump is activated and the check valve separates the failed primary pump from the vacuum system.

A more rudimentary but common standby vacuum source is a cockpit-controlled connection between the vacuum system and the engine intake manifold. So long as the pilot selects an engine power setting that is low enough to maintain a differential between ambient air pressure and the intake manifold pressure, there will be vacuum to power the instruments.

The main weakness of the electrical standby pump installation and the intake manifold standby connection is that they provide redundancy for only the vacuum pump itself. When the system is operating in the auxiliary mode, it uses the same instruments, filters, vacuum lines, and regulator as it does when operating in the normal mode. If any of those shared components have failed, the standby vacuum source may not restore system operation. In essence, installing a standby vacuum source places single-engine airplanes on a par with light twins that have two pumps powering a single system and a single set of instruments. In other words, the standby vacuum system does not provide total systems redundancy.

Failures, Failures And More Failures

If, as Airborne suggests, you were to install a standby pump and routinely replace your vacuum pump prior to its failure, you could minimize the risks inherent in the design of dry pumps. However, you would be addressing only that specific risk. Murphy's Law states that any component that can fail will fail — and at the least-opportune moment. Service Difficulty Reports filed during the last five years provide ample evidence that Murphy's law applies to general aviation. Here s a sampling of what can go wrong with your vacuum system.

Vacuum Pump SDRs...

Sometimes vacuum pumps fail in ways you might not expect. Vacuum pump-related failures caused a loss of engine oil pressure in several cases. In four cases, the vacuum pump housing literally split in two. In three instances, the vacuum pump adapter cracked. In two other incidents the vacuum pump gasket was defective.

Failures result from installation errors as well as from product defects. A new Grob 115C suffered camshaft drive gear damage because the factory had mis-installed the vacuum pump. In two other cases, vacuum pumps were incorrectly installed in the field resulting in a losses of engine oil.

Operator error is also a factor. One pilot attempted to operate his C-172 without having the vacuum pump installed, with predictable adverse results.

...Dual Vacuum Pump SDRs...

Dual pneumatic pumps yield redundancy only so long as the pilot is able to detect the failure of the first pump.

  • A rare partial failure of a dry pump in a twin-engined Piper evaded detection because the minimal pressure (one inch) produced by the failed pump, while insufficient to power the instruments, was sufficient to avoid triggering the warning indicator.

  • In another case, because there was no annunciator installed on this system, the pilot was unaware that the primary pump had failed and flew for hours with the standby pump as his only available source of pneumatic air.

  • In a third instance, the contamination resulting from the failure of the first pump obscured the system warning indicators. The pilot did not notice the initial breakdown until the contamination disabled the second pump resulting in a complete system malfunction.

...Vacuum Filter SDRs...

There were numerous failures of filters resulting from product defects, installation errors and inadequate maintenance.

  • An in-line filter was installed backwards causing a vacuum pump to break down after seven hours in service.

  • Another in-line filter disintegrated upon removal after having never been replaced during the plane's 2,170 total hours in service.

  • Yet a third filter failed in flight after 2,898 hours in service.

  • A fourth in-line filter deteriorated and crumbled after two years in service.

  • The main vacuum filter was waterlogged in two cases — the first because of a leaking windshield seal and the second because it was exposed to torrential rain while baggage was loaded into the nose cargo bay.

  • Several plastic in-line filters melted because of engine or exhaust manifold heat.

Other filters cracked and leaked, leading to vacuum pump failures in some cases and low system pressure in others. In one of those instances, the mechanic had not hard-mounted the filter. It shifted and was cracked because it interfered with movement of the control yoke.

...Vacuum Valve SDRs...

A number of valve failures made the list.

  • There were multiple cases of broken check valves that prevented systems on twin-engine planes from developing pressure from operable pumps on one side following a failure of the opposite side pump.

  • A Piper twin experienced low vacuum pressure because both relief valves had stuck in a partially closed position.

  • A Piper Malibu suffered multiple failures of its standby vacuum pump because a mis-set regulator caused the main pump to build up excessive pressure that over-worked the standby pump. In another instance, a pneumatic regulator leak resulted in inadequate system pressure.

...Vacuum Hose SDRs...

In addition to the problems with the Imperial Eastman Hytron Redi-seal hose experienced primarily in single-engine Cessnas, the latex surgical tubing type vacuum lines used in late-model Mooneys drew fire. There were several reports suggesting that those lines tend to fold over and kink after a few years, resulting in system blockage. Other reports reflected that retracting the nose wheel in some Bonanzas was crimping air lines. There were also a few miscellaneous instances of lines failing because of chaffing from air conditioner ducts and the like.

...Vacuum Gage SDRs

Even the vacuum gages themselves can fail.

  • There was a report of a gage that was leaking internally, yielding abnormally high indications which resulted in the system pressure being set too low to power the instruments.

  • In another case, there was a system leak at the coupling to the gage which resulted in an pressure indication that was lower than the actual system pressure. Adjustment of the regulator to yield a normal reading commanded excessive system pressure, causing a premature vacuum pump failure.

What's A Pilot To Do?

Federal Aviation Regulation Part 91 allows you to fly IFR with no back-up for any component of your air driven instrument system other than needle, ball and airspeed. Airborne recommends and the FAA's Safety Article, P-Pamphlet #8740-52 strongly suggests that at least a back-up vacuum source should be installed in any plane that is flown regularly under IFR. However, based on the above, you may determine that a mere standby vacuum source does not provide sufficient protection for all the potential failures that are inherent in the system. What's a pilot to do?

Why not consider what the regulations call for when true systems redundancy is required? Turbojet airplanes operated under Part 135 or Part 121 are required to have a third gyroscopic bank-and-pitch indicator (artificial horizon) that:

  1. Is powered from a source independent of the electrical generating system;

  2. Continues reliable operation for a minimum of 30 minutes after total failure of the electrical generating system;

  3. Operates independently of any other attitude indicating system;

  4. Is operative without selection after total failure of the electrical generating system;

  5. Is located on the instrument panel in a position acceptable to the Administrator that will make it plainly visible to and usable by each pilot at his or her station; and

  6. Is appropriately lighted during all phases of operation.

These devices, sometimes known as "peanut gyros," are the latest rage among many owners of high-performance piston-powered airplanes with one effect being that the price for secondary attitude indicators has increased and supply has dropped.

Are they the silver bullet? Are they STC'd for your plane? What do they cost? What is the price relative to that of a standby vacuum pump? Where can you mount one? (Hint — you probably have a redundant piece of equipment mounted in your panel that starts with "c" and rhymes with "block.")

Sorry, we're out of room but those are all good questions for next time. Until then, brushing up on your partial panel skills would be a good idea.