Every multi-engine takeoff begins with an unspoken question: if an engine fails at the worst possible moment, what happens next? The answer depends on two carefully defined performance concepts—accelerate-stop distance and accelerate-go distance. Together they frame the central OEI (one-engine-inoperative) decision a twin pilot must make during the takeoff roll, and they explain why runway selection, weight management, and speed discipline are so critical in light-twin operations.
Unlike large transport-category aircraft, most light twins are certified under 14 CFR Part 23 rather than Part 25. Part 25 requires balanced field-length analysis and guarantees a flyable airplane after an engine failure at V1. Part 23 imposes no such guarantee for light twins. That gap in regulation makes understanding these distances not just academic—it is a genuine safety imperative that every multi-engine pilot must internalize before the first takeoff roll.
What Accelerate-Stop Distance Means
Accelerate-stop distance is the total runway length required to accelerate to a specified speed, recognize an engine failure, and then bring the airplane to a complete stop using maximum braking (and, where applicable, thrust reversers or spoilers—rarely relevant on light piston twins). Think of it as the answer to: "If I abort right now, will I fit on this runway?"
The sequence of events the calculation captures is straightforward. The airplane accelerates from a standing start to the decision speed. The pilot recognizes the failure and initiates the abort. There is an inherent reaction-time delay—the airplane continues moving at roughly its current speed during that recognition phase. Finally, braking is applied and the airplane decelerates to a stop. Every one of those phases consumes pavement.
A critical practical point: the higher the speed at which the abort is initiated, the dramatically more distance is needed. Because kinetic energy rises with the square of velocity, an abort from 80 knots requires roughly four times the stopping distance of one from 40 knots. This is why beginning the abort as early as possible—the moment a failure is confirmed—is so important. Hesitation is pavement.
What Accelerate-Go Distance Means
Accelerate-go distance is the total distance required to accelerate to a specified speed, experience an engine failure, and then continue the takeoff on the remaining engine until a defined obstacle clearance height (typically 50 feet) is reached. It answers: "If I continue after this failure, how much runway and clear ground do I need?"
This calculation is inherently more optimistic-looking than reality will feel in the cockpit. It assumes the pilot correctly identifies and controls the failure immediately, retracts the gear at the appropriate time, establishes the correct OEI climb attitude, and reaches Vyse (blue line)—the best single-engine rate-of-climb speed. Any delay, any incorrect control input, any failure to promptly feather the propeller on the affected engine (if the airplane is so equipped) stretches the actual distance well beyond the published figure.
An especially sobering fact for light-twin pilots: many under-powered light twins at gross weight on a hot day at a high-elevation airport will have a negative single-engine climb rate. In that environment the accelerate-go option is not really an option—the airplane cannot maintain altitude on one engine. The pilot continuing the takeoff after an engine failure may be committing to a controlled crash landing ahead rather than a climb.
The Decision Speed Problem in Light Twins
In transport-category operations, V1 is a legally defined and operationally enforced decision speed. Below V1 you abort; at or above V1 you go. Light twins have no FAA-mandated equivalent. The pilot must personally calculate or determine the speed below which aborting is safer and above which continuing may be viable—and that speed is not stamped in the flight manual as a single number.
A commonly used concept is comparing accelerate-stop and accelerate-go distances for a given runway and condition set. If the accelerate-stop distance is shorter than the available runway and the accelerate-go distance also fits, the pilot has a genuine choice. If only the abort fits, that should inform the go/no-go decision before brake release. The Airplane Flying Handbook (FAA-H-8083-3C, Chapter 13) emphasizes that this analysis should be done on the ground, not in the cockpit during the roll.
The Role of Vmc and Engine-Failure Speed
Where these distances intersect with directional control is at Vmc—the minimum control speed with the critical engine inoperative, established under 14 CFR 23.149. Vmc is the calibrated airspeed below which the pilot cannot maintain directional control when the critical engine (the left engine on a conventional light twin with both propellers rotating clockwise as seen from the pilot's seat) suddenly fails at full power.
On a conventional light twin, the left engine is critical because of P-factor: the descending blade of each propeller produces more thrust than the ascending blade. The right engine's descending blade operates at a greater distance from the aircraft's centerline (longer moment arm), generating a stronger yawing moment. If the left engine fails, the right engine's powerful asymmetric thrust must be controlled entirely by the rudder. Losing the right engine instead leaves the left engine—with its less advantageous moment arm—as the operating engine, creating a more manageable (though still demanding) control problem. Aircraft with counter-rotating propellers eliminate this asymmetry entirely.
This matters for accelerate-stop and accelerate-go analysis because an engine failure below Vmc is categorically more dangerous than one above it. If the failure occurs below Vmc during the roll and the pilot attempts to continue, directional control is lost. The immediate correct action is to reduce power on the operating engine to eliminate the asymmetric thrust, and lower the nose to regain speed—not to hold full power and fight with the rudder. If this action is taken promptly on the ground before lifting off, the abort is still viable. Once airborne below Vmc with an engine failure, the outcome is almost certainly catastrophic.
This is one reason many multi-engine instructors teach that the mental decision point for continuing versus aborting should be tied not just to runway remaining but also to whether Vmc has been reached and whether the airplane is accelerating through Vsse—the safe single-engine speed, which is the minimum speed at which intentional OEI training maneuvers should be initiated.
Why These Distances Matter Operationally
A pilot who has not done this analysis before taxiing out is operating on hope. Consider a realistic scenario: a light twin at or near gross weight, a 5,000-foot runway, a density altitude of 4,000 feet due to heat and elevation, and an engine failure at 70 knots during the roll. The accelerate-stop distance might be 3,800 feet—manageable. The single-engine climb rate at that density altitude might be 50 feet per minute—barely positive and certainly not enough to clear trees off the departure end. Continuing the takeoff on one engine is survivable only if the terrain ahead is flat and unobstructed.
This analysis changes with weight. A lighter aircraft not only shortens accelerate-stop distance but also improves single-engine climb performance. It does, however, raise Vmc slightly—because the lighter weight means a smaller lift vector when banked into the operating engine, reducing the rudder's effective authority. This tradeoff is exactly why the handbook specifies that Vmc is determined at the most unfavorable conditions: light weight, aft CG, maximum power, gear up, and with the critical engine's propeller windmilling rather than feathered.
Key Numbers and Rules
- Vyse (blue line): Best single-engine rate-of-climb speed. Target immediately after a confirmed engine failure and gear retraction on departure.
- Vmc: Minimum directional control speed with critical engine inoperative, bank ≤5°. An engine failure below this speed on the ground demands immediate power reduction on the good engine and abort.
- Vsse: Safe single-engine speed. Never intentionally simulate OEI conditions below this speed.
- Bank angle for OEI climb: Approximately 2° into the operating engine with coordinated rudder (zero sideslip) yields the best OEI climb performance—not wings level, not a full 5° bank.
- Abort decision timing: The earlier the abort begins, the shorter the stopping distance. Reaction time alone can add several hundred feet of rollout at typical takeoff speeds.
- Part 23 vs. Part 25: Light twins under Part 23 are not guaranteed to maintain altitude or clear obstacles on one engine. Always verify single-engine performance at actual weight and density altitude.
Common Test Traps
- Assuming lighter weight always helps: Lighter weight improves single-engine climb and shortens stop distance, but it raises Vmc—the two effects must be considered together, not in isolation.
- Identifying the wrong critical engine: On a conventional twin, the left engine is always critical. The right engine's more outboard descending blade creates the worst asymmetric yaw when the left fails. Never reverse this.
- Thinking Vmc guarantees a flyable airplane: Vmc is purely a directional control standard, not a climb guarantee. Maintaining control at Vmc does not mean the airplane can climb or even maintain altitude on one engine.
- Delaying abort action after recognizing failure: Reaction time is built into published figures, but additional hesitation is not. Every extra second at speed adds disproportionate stopping distance due to the square relationship between speed and kinetic energy.
- Confusing accelerate-go with a guaranteed outcome: The published accelerate-go distance assumes immediate, correct OEI technique. Real-world performance is almost always worse. Single-engine performance must be verified at actual density altitude and weight before assuming a continued takeoff is viable.