I think I’ve got the best career ever. I wake up every morning, go to the airport and fly some of the coolest airplanes on the planet. In a typical year, I’ll fly every variant of the Piper PA46, the Socata TBM and the Beechcraft King Air. Whether it’s initial or recurrent training, flying one of our managed King Airs or ferrying an airplane to the far reaches of our spinning globe, I get to fly nearly every day. It’s a privilege that I don’t take lightly. I love my job!
The bottom line is I get to fly single-engine and multi-engine turbines nearly every week, and there is one major difference between flying the two: the MOR. MOR stands for “manual override,” and it is a critical aspect of flying a single-engine turbine. But a multi-engine turbine does not have a MOR. Why?
To answer that question, we need to understand why a single-engine turbine has a MOR. A MOR (usually either a switch or a lever) allows the pilot to control the engine in the event of a power rollback. A power rollback is the nemesis of single-engine turbine operation and one of the least understood aspects of flying a multi-engine PT6-powered turbine. During a power rollback, the engine rolls back to idle, and the pilot cannot control the engine with the power lever. A rollback to idle is deadly because the propeller could windmill. A windmilling propeller causes an immense amount of drag that must be understood to handle properly.
In King Air emergency procedure training, the focus has traditionally been on a single engine failure. A failure occurs when the Ng stops turning, when the ITT drops off the cliff, when the oil pressure reads below minimums. A traditional single engine failure usually has a whole host of caution/advisory lights illuminated and the gauges read so low that the offending engine is easy to identify. It is certainly a terrible potential, especially if the failure occurs at a critical time in flight. An engine failure just after takeoff can be deadly and the training community is right to train pilots to handle engine failures. But can an engine lose the ability to produce torque and not fail? Can a PT6 engine lose the ability to produce thrust but still have normal oil pressure and normal (albeit lower) ITT readings? Can a PT6 engine lose the ability to respond to pilot input? Yes, it can. It is called a power rollback, and it is a potential that I believe is one of the least understood emergencies in the King Air community.
Two of my favorite airplanes to fly are the Piper Mirage with the JetPROP conversion (hereafter referred to as JetPROP) and the TBM 7XX/850. Both are equipped with a PT6 engine and both have a prop lever to feather the prop. Some other single-engine airplanes (Meridian, M500, M600, M700) lack a prop lever, meaning the prop can only be feathered by shutting down the engine. However, in the JetPROP and TBM, the prop can be feathered by pulling the prop lever all the way aft, making it a useful feature in a training environment.
If you fly a JetPROP to a safe altitude (VMC conditions), you can bring the power lever to idle and feather the prop. If the pilot then pitches for 110 KIAS (an ideal gliding speed in this airplane) in a no-wind scenario, the rate of descent will be about 700 feet per minute, and the descent angle (observed with the flight path marker) will be 3.5 degrees down. This demonstrates that the JetPROP is a phenomenal glider with remarkable gliding characteristics in the event of an engine failure.
In a training event, I’ll then advance the prop lever back to full forward and adjust the power lever to achieve 700 fpm and a 3.5 descent angle. Usually, this torque setting is 160 feet/lbs, equating to zero thrust (equitable drag as an airplane with a feathered prop). This is important because it is possible to know zero-thrust with the prop spinning at 2,200 rpm.
If the pilot pulls the power lever back to idle, the descent rate will increase dramatically to nearly 2,000 fpm, and the descent angle will plummet to over 10 degrees down. For just a 160 feet/lbs change in torque, there will be a 1,300+ fpm increase in descent rate, causing the airplane to dive another 7 degrees downward. That is HUGE.
The change occurs because the prop is now windmilling. In normal flight, the engine drives the propeller, but with a windmilling prop, the wind drives the prop and drag increases dramatically. In a JetPROP with a windmilling prop, the drag makes the airplane nearly impossible to land. Simply put, if the airplane were to approach the ground at 2,000+ fpm and 10+ degrees of downward movement, depending on airspeed there’s not enough energy to arrest the rate of descent, and the airplane will crash.
In the PA46 community, the accident history in the last 15 months has been horrific. I’ve personally counted 29 deaths in this time, with 11 fatal accidents. Of those fatal accidents, at least four were from a probable/potential power rollback when the pilot did not feather the prop and did not advance the MOR to control the engine (NTSB reports not final). The impact forces with the excessive descent rate and descent angles rendered the crash unsurvivable or nearly unsurvivable. But those accidents did not need to happen. Had the pilot engaged the MOR, the engine could have been easily managed.
There is a gap in understanding of a power rollback in the PA46 and TBM communities. Sadly, I have had pilots come to me for recurrent training in their turbine airplanes who have never touched the MOR. How anyone could get through an initial or recurrent training event and NOT have flown their turbine with the MOR is unfathomable to me. Yet it happens.
Based on purely anecdotal evidence from my decades of training in the single-engine turbine market, I believe the chances of a power rollback as compared to an engine failure to be a 10-1 ratio (I don’t have exact numbers, and I don’t think anyone else has accurate numbers either). Said another way: I believe a power rollback is 10 times more likely to happen than an engine failure in PT6-powered airplanes. Why? Most successfully handled power rollback events are not reported, and in the worst of power rollback fatal crashes, the engine is so mangled that the NTSB cannot determine the cause of the event. Many of the worst crashes are simply reported, with “subsequent power loss” as a contributing cause to the accident. My point? Power rollbacks do occur and are probably/arguably the primary cause of “loss of engine thrust” in a PT6 engine.
We have a lot of work to do in the PA46 community in teaching the dangerous flight characteristics of an airplane with a windmilling prop and how to use a MOR if we are going to improve the safety record in the turbine PA46. I hope we can reverse the present trend.
Here’s where this discussion applies to the King Air community. The King Air does not have a MOR, but it has the same PT6 engine that is susceptible to a power rollback. Beechcraft (and the FAA) decided that a multi-engine airplane does not need a MOR because the offending engine can simply be shut down and the airplane flown with the good engine to a safe landing. Yet, the dastardly effects of a windmilling prop are still present. If the offending windmilling prop is not feathered, the excessive drag of the windmilling prop can cause huge aerodynamic problems. It can even cause a King Air to lose control if improperly flown.
A Piper Meridian, M500 and M600 have the identical engine as many of the King Air 200s. The only difference between the two engines is the King Air has the PT6-42, and the Meridian has the PT6-42A. That “A” means the fuel control for the Meridian has provisions for the MOR. Otherwise, the two engines are identical. Both are susceptible to a power rollback.
A power rollback occurs when there is a loss of Py pressure. Py pressure is modified P3 air, and P3 air is bleed air tapped from the engine’s compressor section. P3 air is bled (or ported) from the engine and used for all sorts of important in-flight functions. It is used to pressurize the cabin, “poof the boots,” create vacuum for pressurization control, and P3 air also is sent to the fuel control unit.
In the FCU, this P3 air is downgraded slightly and called Px air, and then it is modified again slightly and called Py air. This air is used to control the fuel that goes into the engine.
Fuel flow in the FCU is ultimately controlled by the metering valve, which is like the valve that is on the hand wand of a water pressure sprayer you might use to clean your home’s siding, deck or concrete. You control the valve with your hand in a pressure sprayer but not in your PT6-powered King Air. Beechcraft knew a pilot could not be trusted to control the metering valve directly. They knew that a pilot might cause a flameout or a compressor stall if the metering valve were mishandled.
A flameout could occur if the metering valve were allowed to close too quickly and not allow enough fuel to flow to keep the flame burning in the combustion section of the engine. A flameout results in an engine failure, a bad potential for any regime of flight.
A compressor stall occurs when the metering valve opens too quickly, allowing fuel to be dumped into the engine at a higher rate than the compressor can provide compressed air. If you want the engine to speed up, you must also make the compressor turn faster. If not, the compressor will stall and the excessive pressures can destroy an engine internally.
To prevent both a flameout and a compressor stall, Pratt & Whitney devised an ingenious system: They married the compressor’s speed with the movement of the metering valve. The engine uses bleed air (Py air) from the compressor section to move the metering valve. When you move the power lever on your King Air, you are not controlling fuel. You are controlling air (Py air) that subsequently controls fuel. When you push the power lever forward, you increase the Py pressure in the fuel control, which opens the metering valve and sends more fuel to the thirsty engine. Pulling back on the power lever means lessening Py air and decreasing fuel flow.
If you’ve ever noticed, there’s a bit of a delay between power lever movement and engine response. Turbine pilots become accustomed to this delay in response. This delay is because it takes a while for the compressor to speed up and provide the air pressure that moves the metering valve.
The FCU on a PT6 is a mechanical device, and mechanical things can break. If there is a break (leak) in the air lines bringing the Py air to the fuel control (or forward to the prop control), then there will be no Py pressure to move the metering valve, and the engine will roll back to idle. The pilot can move the power lever as much as desired, but there will be no engine response. The engine will remain at idle.
This power rollback causes the engine to roll back to idle, creating the question, “Did the prop feather?” If you’re like most pilots in training, there will be some confusion in the cockpit. The engine will still be running, so you should have otherwise normal engine indications. However, the torque will be near zero, and if your power lever is forward (arming the autofeather function with the power lever angle switch), the autofeather system should feather the prop.
But what if the power lever isn’t set to forward arming autofeather? What if you pulled the power lever back in an attempt to diagnose the issue? What if your King Air doesn’t have autofeather installed?
Your King Air emergency procedure checklist will direct you to try to add power, check the Ng speed (which will be at idle if you are experiencing a power rollback), feather the propeller and eventually shut down the engine. The critical action is to feather the prop, as the windmilling prop is your adversary. You won’t need to restart the offending engine in flight since you won’t be able to regain control of the power lever until the Py air leak is repaired.
In the King Air market, I have noticed a lack of understanding regarding the impact of a power rollback when conducting recurrent training. Furthermore, I’ve had few discussions about a power rollback within the King Air community during my 22 years of flying King Airs, and those have only occurred with the best training providers. It is as if there’s a dark gap in understanding of the power rollback in this community.
Another system within the King Air that I find consistently misunderstood is the autofeather system. King Air pilots truly need to understand these two systems to operate a King Air safely. If you have a windmilling prop, you won’t be airborne for long. Ensuring the prop is feathered in a loss of thrust scenario is critical.
When I administer an airline transport pilot (ATP) multi-engine practical test, the airman certification standards (ACS) require the candidate to shut down the engine and restart it in flight. I’ve completed hundreds of in-flight engine shutdowns and restarts in every type of King Air, and I’ve witnessed the unfortunate consequences of a windmilling prop.
During the ATP-ME practical test, I can pull the power lever of one engine to idle, prompting a response from the pilot candidate and providing an ideal moment to consider the effects of a windmilling prop. When the prop is feathered, the King Air will surge forward due to the reduction of drag. Depending on the model of the King Air, a windmilling prop can lead to a decrease of more than 35 knots in airspeed. That’s HUGE! One of the most important considerations when dealing with an engine failure (or a power rollback) is to ensure the prop is feathered!
It is true. You may need to feather a prop even if the engine has not failed. A power rollback can indicate an engine issue that doesn’t lead to a failure but does necessitate feathering the prop.
Knowledge is power. Understanding the power rollback and a windmilling prop can be the difference between a successful single-engine landing (and a subsequent minor engine repair after landing) and a fatal crash after the King Air flips over on its back. Directional control is everything; handling a power rollback is one skill that every King Air pilot should fully understand.
If you seek additional information about the nuances of the power rollback and how it occurs within the fuel control system, I have a tutorial video on my website at flycasey.com/videos.