Experimental Test Pilot 28 November 2007
The Boeing Company
By Greg Lausin
How about that fuel sample???
Is a phrase that every aviator has spoken when querying his crew chief before pre-flight.
As an Army aviator with twenty years of flight experience, I spent the early part of my
career as an AH-64A Maintenance Test Pilot (MTP), then as an AH-64D Longbow MTP,
and then as an Experimental Test Pilot (XP). I was fortunate enough after retiring from
the US Army, to continue my aviation career as an XP for Boeing Helicopters in Mesa. I
must admit that a cursory look at the fuel sample was all that was required and
sometimes, it was nothing more than asking the crew chief if he took one or not. I rarely
gave it more thought than that. Until now…
One afternoon my boss approached me with a question regarding engine oscillations.
What could cause them, and what to do to fix them. During the week of 8 October, 2007,
a non-U.S. AH-64D operator had experienced uncommanded engine oscillations on three
AH-64D’s, while operating in the vicinity of Frosinone, Italy. One AH-64D experienced
two separate episodes of uncommanded engine oscillations, and two other AH-64D crews
experienced one episode each. In the most serious incident one of the crews responding
to severe engine oscillations had retarded one power lever to idle, when the remaining
engine, also oscillating, flamed out, requiring the crew to execute an autorotation to an
emergency landing. Two aircraft experienced engine oscillations nearly simultaneously
(within minutes of each other) and approximately 12 km apart. In responding to my
boss’s question, the only semi-intelligent thing I could think of saying at the time was
The non-U.S. AH-64D operator requested Boeing on-site participation in the ensuing
investigation, beginning a very strange journey. We started out by looking at the
Maintenance Data Recorder (MDR) data from each of the aircraft, and noticed some very
unusual occurrences, that did not appear to make sense. The uncommanded engine
oscillations were not consistent with any previously observed behavior pattern of the GE-
T-700-701 engine series. They exhibited unusual engine indications, i.e. low torque
simultaneous with a high gas generator (Ng) and high Turbine Gas Temperature (TGT)
readings. The engines were surging well into the engine stall region, with large torque
splits, and divergent high / low oscillating power turbine speed (Np) and rotor speed (Nr),
to engine over and underspeed limits. An initial challenge was sorting out whether one or
both engines were misbehaving in these events. We thought that the data had to be wrong
– it contradicted our pre-conceived notions about how twin-engine helicopter propulsion
malfunctions should appear.
One of the things I learned during XP training at the United States Naval Test Pilot
School was; “the data is always right, and if it doesn’t look right, you just don’t
understand it”. So how could this be? Every time we thought we figured out a cause of
the phenomena, we realized that it had occurred to multiple aircraft. And because
inconsistent engine behavior was the only substantial evidence presented, multiple
hypotheses were considered to explain the behavior. Since the occurrences happened to
multiple aircraft over a fairly large geographic area, could Electromagnetic Interference
(EMI) from a nearby source be the culprit? Was a new secret directed energy weapon
being used by terrorists? What about the possibility of sabotage? How about fuel
contamination? But contaminated fuel would just kill engines outright, and generally
would affect both engines at the same time, wouldn’t it? But in our deliberations we
found ourselves back to: “Yeah, but it happened to multiple aircraft and two of them at
the same time” and was not experienced by any other aircraft operating in the area at the
time (Frosinone is an Italian flight training facility, as well as an Augusta manufacturing
facility). Fuel contamination had been discounted by everyone involved, because the
initial fuel samples taken for analysis all came back good. We were stumped. Some of the
smartest people in the industry were sitting around a table scratching their heads.
As we continued to apply analytical thought to these events, we knew we had to figure
out what conditions, systems or components on the aircraft could cause the engines to
behave in the way they had. We were not worrying about the why at that point.
Experience told us that during EMI testing, very high power levels could affect aircraft
systems but on the AH-64D it usually affected cockpit displayed indications, and not the
actual function of systems. There were always some other visual indications observed by
the pilot e.g., scrambled images on the cockpit displays.
The next thing we knew, we were on our way to beautiful southern Italy to meet with the
Interviews with maintenance personnel and review of maintenance actions proved to be
most productive and revealing. Prior to the deployment of the operator’s Longbows to
Frosinone Italy, the Petroleum-Oil-Lubricant (POL) section was instructed to “drain and
clean the fuel truck”. Complying with instructions, the fuel truck was “emptied and
cleaned out per the manual”. The truck was then “sealed” and driven to Italy via
Switzerland. Upon arrival, the fuel truck was filled with F-34 Jet Fuel from the Italian Air
Force to await the arrival of the four Longbows.
Normal procedures required the fuel handlers to take a fuel sample from the truck prior to
dispensing into any aircraft. According to the supervisor of the POL section, fuel samples
were taken in accordance with procedures. After the incidents, another fuel sample was
drawn from the truck and analyzed. Local Aqua-Glow testing did not reveal anything out
of the ordinary. A second sample was sent to the Shell Oil Company for analysis, and
again, the sample test results were good. Information that was obtained by the
investigation team revealed that a very large quantity of water (at least five, 5 gallon
buckets) was drained off before getting a clean sample, which was the sample tested
locally and the sample sent to Shell for analysis. It was also learned that by procedure,
fuel samples were generally not taken from the aircraft, except after fuel cell related
maintenance. With this information in hand, a closer look at the fuel systems of each of
the four aircraft was conducted.
A review of each aircraft’s log book was conducted and did not reveal any fuel related
maintenance actions had been conducted with the exception of the installation of
Robinson Extended Range Fuel System installed on three of the four aircraft for the ferry
to Italy. Additionally, all four of the aircraft were ferried with 230 gallon external fuel
A closer inspection of the aircraft was facilitated by hand turning of each of the engines.
A very noticeable grinding noise was heard emanating from the fuel control unit (hydro-
mechanical unit (HMU)) on three of the eight engines. The HMU’s were removed in
accordance with appropriate maintenance publications, and each engine was hand turned
again. This time, the grinding noise was not present. There was no apparent indication of
external damage to the HMU’s, HMU drive shafts or the mounting plate assemblies onto
the accessory sections of the engines. The maintenance manual describes a break away
torque of at least 24 in-lbs possibly required to turn the HMU shaft. However, when a
break away torque of 50 in-lbs was required to turn the shaft, with an accompanying
grinding noise, the HMU’s were suspected to be bad. Fuel samples were obtained from
the fuel cells, the HMUs, and the fuel filter bowls. Close examination of the fuel samples
revealed a heavier than water or fuel cellulose substance free-floating in the fuel that had
the appearance of the white stringy stuff in egg drop soup. Additionally, a gel-like
substance was observed to be adhering to the bottom and sides of the fuel filter bowls that
required scraping to remove. As the samples were allowed to settle, the gel-like substance
began to coagulate into a larger ball of material. The gel-like material was sticky to the
touch. Samples of the fuel with the cellulous material were sent to Shell Oil for analysis.
The HMUs were sent to their original manufacturer for tear down evaluation.
An initial attempt of remedying the aircraft to return them to flight consisted of draining
the fuel from one of the aircraft and refilling with known good fuel. After this was
accomplished, the aircraft was ground-run while varying the power setting and
manipulating fuel control components. At the completion of the ground run, fuel samples
were drawn from the aircraft for analysis. A visual inspection of the fuel samples
revealed the same type substance in both the cellulous form as well as the gel form were
present in the fuel samples initially obtained from the fuel filter bowls.
Initial reports from the vendor indicated that the HMU sent for analysis passed the initial
Production Acceptance Test Procedure (PATP) but failed when fuel was hooked up.
Several valves were stuck which would have prevented the HMU from operating
normally. Additionally, the #1 engine HMU on the ground test aircraft would not start the
engine and was also sent in for analysis. When consulted, General Electric, the engine
manufacturer, responded that if the contamination caused valves within an engines HMU
to stick, the indications we had seen in the MDR data could occur. While nothing could
be positively identified as to the cause of the engine oscillation events, it appeared more
and more likely that fuel contamination was the culprit.
Even though we were fairly convinced that the problem was with the fuel, we still wanted
to eliminate the possibility of EMI. A hand-held Spectrum Analyzer was used to measure
any and all RF energy and broadband field strength that might be in the area. There were
no emission in the area that could produce enough energy that would effect even the
reception on a TV, let alone a helicopter in flight 12 kilometers away.
Upon return to the US, we learned that the gel-like substance was a variation of “Apple
Jelly”. No, not the apple jelly you spread on your biscuits in the morning, but a gel-like
substance that predominantly affects fixed wing aircraft fuel systems, and is developed
when jet fuel with an anti-ice additive combines with an excessive amount of water.
Several instances of contamination of “Apple Jelly” have been documented inside fuel
cells, but not in fuel control system components on engines.
The moral of this story is to take fuel samples seriously. This operator was very lucky
that no one was injured and that no aircraft were lost. Kudos to the flight crews involved,
as there could have been far more serious consequences to their flight with “Apple Jelly”.
Experimental Test Pilot
The Boeing Company