Ask a multi-engine student why they are paying for the rating and most give some version of the same answer. If one engine quits, the other one gets you home.
That sentence is true at sea level on a cool morning. It quietly stops being true above a certain density altitude, and Kenya's training aerodromes sit closer to that number than most students are ever shown.
The Short Version
- Vmc, the speed marked with the red radial line, only guarantees you can keep the aircraft pointed straight after an engine fails. It says nothing about whether the aircraft can climb.
- The number that actually answers "can I climb" is the single-engine service ceiling, and the FAA does not require light twins under 6,000 lb with a stall speed under 61 knots to have any single-engine climb performance at all.
- Density altitude erodes whatever single-engine climb margin exists, and Kenya's main multi-engine training base, Wilson, sits at 5,536 ft before the day even gets warm.
- A peer-reviewed accident study found that in visual conditions, the single largest cause of fatal twin-piston accidents was a malfunction combined with a failure to follow single-engine procedures.
- KCAA already treats the multi-engine transition as a serious training event, with a dedicated 40-hour instrument rating requirement. The performance-planning discipline around it deserves the same seriousness.
What Vmc actually promises, and what it does not
Every multi-engine student learns the red line early. Vmc is defined in 14 CFR 23.2135(c) as the calibrated airspeed at which, following the sudden critical loss of thrust, it is still possible to maintain control of the aircraft, and the FAA's Airplane Flying Handbook, Chapter 13 is explicit about what that definition leaves out.
There is no requirement that the aircraft be capable of climbing at Vmc. The speed addresses directional control only, full stop.
Most students absorb Vmc as the whole engine-out story because it is the number with a red line on the airspeed indicator and a dedicated demonstration in the syllabus. It is one part of a bigger performance picture, and it is not the part that tells you whether you are going to hit the ground.
The number that actually matters: single-engine service ceiling
The FAA's own pamphlet for light twin pilots, FAA-P-8740-66, Flying Light Twins Safely, defines the figure that does answer the climb question. The single-engine service ceiling is the maximum density altitude at which the aircraft will climb at 50 feet per minute, in smooth air, at maximum weight, with the other engine feathered.
Read that definition twice, because the second half is where the surprise lives. The pamphlet states plainly that the Federal Aviation Regulations do not require any single-engine climb performance whatsoever for light twins weighing 6,000 lb or less with a stall speed of 61 knots or less.
That covers a large share of the piston twins actually flying training syllabuses today.
A "multi-engine rating" does not, by itself, certify that the aircraft you trained in can climb away from a bad day on one engine. It certifies that you can keep it under control while you find out whether it can.
Critical density altitude: where the two numbers meet
Density altitude does not just eat into your single-engine climb margin. It moves Vmc itself.
The Airplane Flying Handbook notes that reduced air density lowers Vmc, while conditions that increase available power raise it. Push density altitude high enough and Vmc keeps falling until it meets the aircraft's stall speed, a point the handbook calls the critical density altitude.
Above that altitude, the published Vmc number stops being usable at all, because you would stall the aircraft before you reached a speed at which asymmetric power could be controlled. That is not an exotic edge case reserved for mountain flying courses.
Our piece on density altitude in the Kenya highlands lays out how far above field elevation the real number climbs on an ordinary warm afternoon. Every degree of that same arithmetic applies to a multi-engine aircraft, and it applies on top of a single-engine climb margin that may already be thin or nonexistent.
Wilson is not sea level, and neither is the cross-country
Kenya's multi-engine training is not happening in a low, cool, forgiving environment by accident of geography. It is happening at altitude by default.
Our piece on negative transfer between aircraft types documents that Kenya School of Flying runs the Piper PA-34 Seneca as its multi-engine trainer, based at Wilson (HKNW). As the companion piece on highland density altitude sets out, Wilson sits at 5,536 ft AMSL, and a multi-engine cross-country from there routinely reaches Eldoret at 6,936 ft or Nanyuki at 6,250 ft.
None of those fields are unusually high by Kenyan standards. They are simply not the sea-level or near-sea-level environments that most multi-engine performance charts, training folklore, and internet forum wisdom were quietly written around.
The margin you assume is the margin you have not checked
A single-engine service ceiling published in a Pilot's Operating Handbook is a number, not a guarantee that follows the aircraft around. It was calculated at a specific weight, in a specific configuration, at standard temperature.
A hot afternoon departure from Eldoret, at maximum training weight with an instructor, a student, and full fuel aboard, can sit meaningfully closer to that published ceiling than the same aircraft ever gets during training flown out of a lower, cooler aerodrome. The gap between "the POH says we can climb" and "we can climb today, at this weight, in this heat" is exactly the gap a proper preflight performance calculation is supposed to close.
What the accident record actually shows
This is not a theoretical concern dressed up to sound serious. A peer-reviewed study in Accident Analysis & Prevention, Causes and risk factors for fatal accidents in non-commercial twin engine piston general aviation aircraft, examined 376 accidents in twin piston aircraft of 4 to 8 seat capacity between 2002 and 2012.
Under visual weather conditions, the single largest identified cause of fatal accidents, 27 percent of them, was a malfunction combined with a failure to follow single-engine procedures.
That is not an engine failure alone doing the damage. It is an engine failure followed by the pilot not doing, in the moment, what the single-engine procedure requires.
The same study found twin-piston accidents carry a materially higher fatality risk than single-engine accidents, and identified specific risk multipliers: night operations raised the odds of a fatal outcome roughly 3.7 times, an off-airport landing roughly 14.8 times, and a post-impact fire roughly 7.2 times. The authors point to the twin's own flight profile, higher airspeed, higher service ceiling, more fuel aboard, and the yawing moment an engine failure introduces, as reasons single-engine accident research does not transfer cleanly to twins.
None of that research was conducted in Kenya, and it should not be read as a Kenya-specific statistic. What it establishes, from a large enough sample to mean something, is that "follow the single-engine procedure correctly" is already the single biggest lever in a light twin accident, before density altitude enters the picture at all.
The regulation already takes this transition seriously
Kenya is not silent on the multi-engine step. KCAA's Civil Aviation (Personnel Licensing) Regulations, 2018 underpin a multi-engine instrument rating that, as our piece on negative transfer notes, requires a minimum of 40 hours of dedicated training split between simulator and aircraft time.
That is a real, hour-based commitment, more structured than the class-rating transitions students make between single-engine types. It tells you KCAA already recognises the multi-engine step as materially different, not a bigger version of the same licence.
What the regulation does not do, and cannot realistically do from a hours-based rule, is mandate that every one of those 40 hours includes an honest single-engine performance calculation for the actual density altitude of the actual training day. That part is a school and instructor discipline, not a paperwork requirement, and it is exactly the part most likely to get skipped when the hour count is already satisfied on paper.
What a Kenyan multi-engine syllabus should add
None of this argues against multi-engine training in the highlands. It argues for treating the performance chart the way the checklist already treats the fuel sample: as a step you actually do, not a step you assume was fine last time.
- Calculate density altitude before every single-engine training exercise, not only before cross-country departures. A practice engine failure flown at a density altitude close to the aircraft's single-engine service ceiling is not a realistic simulation, it is a real performance limitation wearing a training label.
- Check the single-engine service ceiling against today's density altitude, at today's weight, before any flight that will include single-engine work above field elevation. If the margin is thin, that is information for the go/no-go decision, not a detail to discover in the air.
- Brief the critical density altitude explicitly for the type in use. A student who has never heard the term will not think to ask whether today's conditions are anywhere near it.
- Treat a cross-country to a higher field as a separate performance problem, not an extension of the departure calculation. Nanyuki and Eldoret are not Wilson with a different radio frequency.
The honest opinion: "it's a twin, we'll be fine" is the dangerous sentence
Here is the opinion this article is willing to commit to. The most dangerous phrase in multi-engine training is not a specific error, it is the general confidence a second engine quietly produces in everyone around it, instructors included.
A twin feels safer because it has a spare engine, and that feeling is doing real work against the accident study above, which found the procedure failure, not the engine failure, was the largest fatal factor. The aircraft category most associated with a safety margin is the one where getting the response wrong costs the most, and Kenya's elevation adds a second way for the same overconfidence to bite: assuming climb performance that the density altitude has already quietly removed.
Instructors: the brief that costs one extra minute
An instructor cannot recompute a full performance chart from memory at the hold point, but they can ask one question out loud before every single-engine exercise. What is today's density altitude, and where does that put us relative to the single-engine service ceiling.
If the honest answer is "close" or "I don't know," the exercise gets flown differently or not at all. Naming the number, rather than trusting a general sense that "the Seneca can handle it," is the entire discipline this article is asking for.
What pilots admit on the forums
Multi-engine forum threads have a recognisable shape once you have read enough of them. Almost nobody posts that they calculated single-engine service ceiling against density altitude before a routine training flight, because a calculation that ordinary rarely feels worth mentioning.
What does come up, repeatedly, is the near miss discovered afterwards: the instructor who realised mid-debrief that the single-engine practice that afternoon had been flown closer to the aircraft's real ceiling than anyone had checked beforehand, or the pilot who only started running the numbers on every leg after a hot, heavy departure that climbed noticeably slower than expected on both engines, let alone one. The regret in those threads is rarely about a genuine engine failure gone wrong.
It is about how many training flights went by first, unchecked, before the number got taken seriously.
AngaBrief's Aircraft section of the PAVE wizard is exactly where a performance-limiting note like a marginal single-engine climb margin belongs, recorded before departure rather than reconstructed afterwards. It does not calculate Vmc margins or make the dispatch decision itself, and it is not a dispatch authority.
The decision to fly, on one engine or two, rests with the pilot in command and their instructor.
Key Takeaways
- Vmc only guarantees directional control after an engine failure. It says nothing about whether the aircraft can climb.
- Single-engine service ceiling is the number that answers the climb question, and light twins under 6,000 lb with a stall speed under 61 knots have no FAA-mandated single-engine climb performance at all.
- Density altitude lowers Vmc and erodes single-engine climb margin simultaneously, and Kenya's multi-engine training base at Wilson starts at 5,536 ft before the day warms up.
- A peer-reviewed study found that in visual conditions, malfunction combined with a failure to follow single-engine procedures was the largest cause of fatal twin-piston accidents, at 27 percent.
- KCAA's 40-hour multi-engine instrument rating already treats this transition seriously. Checking single-engine service ceiling against real density altitude before every single-engine exercise is the training discipline that should match it.