Air Safety Week: On Becoming Electrically Inert

FindArticles > Air Safety Week > July 2, 2007 > Article > Print friendly

On Becoming Electrically Inert

Switches that Kill

After takeoff and in instrument conditions, at a point six minutes after takeoff, the pilot of a Beech 200 KingAir progressively lost his EFIS instrumentation (Proline 21 avionics) and all communications.

He assumed that he had a major avionics failure and concentrated on the standby flight instrument display indications until the aircraft had climbed clear of cloud and was level at FL150.

Scottish ATC became aware of the aircraft's predicament when they saw its tracking change and lost its transponder signal and all comms. They dispatched a Tornado fighter to intercept. After some confusion, the KingAir followed the fighter but then lost control upon entering cloud.

Reason? The backup standby flight instrumentation was also EFIS and it too had ultimately failed.

The Tornado made a Mayday Relay call and vectored back onto the aircraft, a short while later finding it flying between cloud layers. A passenger aboard the KingAir contacted ATC by cellphone and ascertained that the plan was to recover them to Leuchars.

The KingAir was guided clear of cloud to a successful visual approach and landing at RAF Leuchars on the East coast of Scotland. The aircraft had been airborne for almost two hours and had been without electrical power for at least 90 minutes.

Technicians at Leuchars applied ground power and could not fault the aircraft. Investigators later analyzed the timings of the progressive failure and factually deduced the likely cause. The pilot was the Chief Pilot and he normally flew the Beech 200 fitted with electromechanical instrumentation.

There was no requirement for him to "convert" to the ProLine 21 version (see displays at tinyurl.com/367fzp). However, he had right-seated about 10 times for another pilot. Climbing into the LH seat he had actually been entering an alien and unfamiliar environment with respect to the switchology layout and displays.

Once the emergency commenced, the pilot considered that the workload involved in maintaining the standby instrument scan for controlled flight had made fault finding "almost impossible". This is a classically common observation for single-pilot IFR flight emergencies.

After the flight, he stated that he had seen no warning or caution lights illuminate while airborne and he could not recall whether he had checked the voltage/loadmeter gages or the battery ammeter gage during the flight. Once on the ground, the commander checked the battery voltage and noted that it was very low.

There was no data recorder and the CVR had been overwritten, so investigators relied upon the recorded Radar times and R/T transmissions. There'd been only six minutes between takeoff and the loss of the aircraft's secondary radar return. The KingAir has a 28VDC starter-generator on each engine, two main batteries and a single NICAD battery for backup instrument power.

The Generator field switches are situated to the left of the control column and sit beneath a guard labeled "master switch". If a GEN trips, the switch must be moved up to a spring-loaded reset position and released. Unguarded engine auto ign, engine anti-ice and ignition and engine start switches are clustered near the generator and battery master switches.

These switches are not visible to a pilot who's in (and previously only flown in) the right seat. Battery Ammeter and Load and Voltage meters are located in the overhead panel. A sole pilot experiencing flight instrument woes would not want to look up and back at them, because that would be vertigo inducing. The pilot did recall looking down at the battery and generator switches and seeing that they appeared to be on.

As there were no indications of a GEN trip, he did not attempt any GEN resets until much later, just before landing when the gear wouldn't extend. A failure of either/both generators should illuminate the master caution light together with an associated L GEN and/or R GEN amber caution caption(s).

The only other advisory, besides the meters, that there is an electrical failure underway would be the amber caption on the caution annunciator panel and there are two traps connected to actually "seeing" that. Some caution panels have a caption bright/dim switch but many other annunciator panels automatically go to DIM if a night-lighting rheostat is turned on (or is left ON from a prior night flight).

In some aircraft, even if a single rotary rheostat is just "cracked", the caution panel lights will dim and be virtually invisible for daytime flight. The other trap is that the caution panel's captions dim progressively as it's also inevitably affected by any electrical failure and a consequent flat battery, i.e. it may display the caption for five minutes maximum before "light's out".

Initial actions upon receiving a RED MASTER WARNING or amber MASTER CAUTION is to cancel it and then look down to identify the malfunction's caption. However, in the circumstances described above, a lit caption will simply not be visible. The pilot will naturally assume a transient self- correcting fault condition had caused the MASTER to flash.

In the event of complete DC generation failure, the aircraft battery was certified to provide power for 30 minutes, with this duration depending upon the pilot recognizing the problem and shedding non-essential electrical loads (which would include non-essential functionalities of the Proline21).

However, if the Proline stayed "up" and the gear and flap was actuated, the 30 minutes would decrease to less than 10 minutes, but the separate NICAD would still power the STBY instruments for up to 30 minutes. From the timeline, this is exactly what happened.

The purpose of UK CAP 747, generic reqmt #4 under 'Mandatory Requirements for Airworthiness' was to ensure that 'certain aircraft' under 5,700kg maximum authorized weight would provide the pilot with a clear and unmistakable warning in the event of a loss of electrical generation. EASA Certification Standard CS 23.1322 defines a warning indication as 'red and non-dimmable'.

The avionics manufacturer confirmed that if the aircraft had suffered a progressive failure of its electrical supply, this should have been recorded on both the Maintenance Diagnostic Computer (MDC) and Flight Management Computer (FMC). Both were removed and their non-volatile memories were downloaded.

They contained no data relevant to the flight, confirming that the double generator failure had occurred prior to 80 knots and activation of the weight- off-wheels switch on takeoff. Neither device will work just off the main batteries.

As mentioned earlier, the cluster of switches just left of and below the LH pilot's yoke (tinyurl.com/34vwfl) contains two similar groupings of switches. If, during the pre-takeoff checks, a pilot was to inadvertently select the ignition and engine start switches to 'on' rather than the almost identical and adjacent eng auto ign switches, per the checklist, the result would have been that the generators would have been tripped off-line.

This action would have resulted in the battery being unable to support all the aircraft systems within about six to seven minutes after takeoff (which is what happened). A MASTER CAUTION wouldn't have revealed any captions on the night-dimmed annunciator panel and so the pilot would have "pressed on", canceling and disregarding the apparently spurious master caution.

Even if he'd had reason to look up at the ammeter, investigators found that a discharge deflection on the needle was easily mistaken for a charging indication.

The pilot stated that he attempted to reset the generators just prior to landing at Leuchars. If the problem was caused by having the start switches in the ON position, then he would have been unable to reset the generators until he noticed his mistake and selected the start switches to the OFF position.

EASA (Europe's FAA equivalent) is considering an Airworthiness Directive to ensure that the aircraft type complies with the requirements of EASA CS 23.13 09(b)(3) and 23.13 53(h) by providing red warning annunciations when both generators are off-line, and a "low volts" warning when the aircraft battery is supporting any part of the aircraft's electrical load.

If the pilot is aware of reversion to battery power, 30 minutes should usually be sufficient time in which to get on the ground. However, if the pilot is UNaware that both generators are off-line, in this aircraft variant both the main and standby instruments could fail in quick succession, as happened here.

On January 27, 2001, N81PF, a Beech 200 carrying 10 members of the Oklahoma State University basketball team back to Stillwater from a game, crashed 18 minutes after takeoff from Jefferson County Airport Broomfield, Colorado.

In the absence of any evidence to the contrary, the accident was attributed to an AC inverter failure and the failure of the pilot to select the alternate AC inverter. After reading the above analysis of this UK incident (report just released), you may well harbor some doubts about that OSU accident finding.

For a non Proline21 equipped KingAir aircraft, the OSU accident's timing (until loss of DC-powered inverter-driven AC electrics including attitude instruments and mode C), leading to a loss of control, is about right (tinyurl.com/2xrela).

A DC-powered AC inverter will drop offline while there is still power available to drive DC powered equipment, and in doing so, will free up remaining battery capacity. This double-generator failure scenario appears not to have been considered for the OSU accident.