Inverter Lockout Cited In ZeroAvia Hydrogen Test Bed Accident


The forced landing of a hydrogen-powered Piper PA-46-350P Malibu Mirage test bed operated by hydrogen-electric aircraft developer ZeroAvia last year resulted from an inverter lockout, according to the final accident report published by the U.K.’s Air Accidents Investigation Branch (AAIB). Investigators found that “the loss of power occurred during an interruption of the power supply when, as part of the test procedure, the battery was selected off with the intention of leaving the electrical motors solely powered by the hydrogen fuel cell. During this interruption the windmilling propeller generated a voltage high enough to operate the inverter protection system, which locked out the power to the motors.”

Attempts to restart the system were unsuccessful and the aircraft, which was being flown as part of ZeroAvia’s Project HyFlyer I flight test program, was severely damaged when it landed off-airport and impacted a ditch and a hedge. As previously reported by AVweb, the accident took place on April 29, 2021, near Cranfield Airport. No post-crash fire occurred and no one was injured in the accident.

Issues identified in the AAIB’s report include the positioning of the pilot’s display unit, which was obscured by the pilot’s hand on the throttle, along with warning and caution indications that had no attention-getting properties. The AAIB further concluded that “sufficient ground testing had not been carried out to determine the effect of the back voltage from a windmilling propellor on the inverter protection system” and that the emergency procedures established for the situation were ineffective. In addition, it found that a loss of power due to an inverter lockout had occurred on several previous flights and had not been properly investigated nor had risk assessments been reviewed following those flights.

The AAIB’s safety recommendations related to the accident call on the U.K. Civil Aviation Authority (CAA) to take actions such as developing guidance on the use of existing materials related to the design and positioning of controls and displays on experimental aircraft, clarifying “the scope of projects considered suitable to be carried out” under experimental conditions and requiring an independent review of the dossier for aircraft operating under experimental conditions. The AAIB also recommended that CAA require that individuals nominated as a competent person under CAP1220, Operation of Aircraft Under E Conditions, have the knowledge, skills, experience and capacity to manage and oversee the registered experimental test program and that CAA “enhance the guidance for the competent person and principal test pilot in the organisation, management, and conduct of the flight of an experimental aircraft project.”

“While the report does not make specific recommendations for ZeroAvia within its conclusions, many of the issues identified in the report were similarly noted in our internal investigation and have subsequently been addressed robustly,” the company said in a statement. “We have embedded key learnings into both our organisational culture and structure, as well as our future technical designs.” ZeroAvia also noted that it has established a Safety and Security Review Board along with a safety management system, which includes occurrence reporting, investigation and corrective actions functions, since the accident.  

Kate O'Connor
Kate O’Connor works as AVweb's Editor-in-Chief. She is a private pilot, certificated aircraft dispatcher, and graduate of Embry-Riddle Aeronautical University.

Other AVwebflash Articles


  1. …and once all the test procedure improvements are made…fly those test points that could potentially interrupt power above a suitable runway with a pilot proficient in power off landings.

    • Poor planning predicts poor performance.

      Even as such HFC are a better power source if electric airplanes are to be developed than lithium batteries.

    • They are trying to make a distinction between things that were actually learned, as opposed to things that were taught (where perhaps the lesson was not learned).

    • The report was written by inhabitants of the UK. There are a lot of differences between UK and USA grammar.

  2. “….it found that a loss of power due to an inverter lockout had occurred on several previous flights and had not been properly investigated nor had risk assessments been reviewed following those flights.” In hindsight, perhaps reviewing these inverter lockouts might have revealed unintended issues to address before continuing tests.

    • A wag said years ago:
      “If someone invited you to ride in a conveyance that contained a quantity of a very flammable substance you’re run away screaming.
      Oh, wait a minute – did you say ‘traditional car?'”

      I think the hydrogen push is a crock, but could be handled with proper study and quality handlers.

      (FTR, Humans cannot cause runaway climate warming, which is not and cannot happen.

      Climate has been warming slowly since the end around 1750AD of a cool era, shown by accurate thermometers like weather balloons and satellite sensors, and by tide gages.

      Earth was warmer and climate stable in the Medieval Warm Period when Vikings farmed southwest Greenland.

      The effect CO2 can have is small, limited by the ‘saturation’ effect of overlap of spectra of carbon dioxide and dihydrogen monoxide vapour, most of the increase has already been realized.

      Read Alex Epstein’s book ‘Fossil Future’ for facts of climate and the benefits of fossil fuels and CO2 to humans, and more at

      • Despite the absurdity of the “Climate Change” hysteria, it’s worth comparing the simplicity and safety of an electric motor, battery, controller, and fuel cell with the complexity of, say, a TIO-360.
        * Manufacturers such as Magnix and Yana are using a new design called Axial Flux, yielding a light but powerful motor. Some motors are designed with a hollow shaft, enabling two motors to be stacked for redundancy/fault tolerance.
        * Hydrogen fuel storage is becoming increasingly ready for commercialization as R&D in metal hydride storage progresses.
        * A small battery in the serial hybrid setup would provide sufficient power to return to land safely in the event of disruption of power from the fuel cell(s).
        * Automotive fuel cells are too heavy for aircraft. HyPoint has a promising cell using new technology and is in partnership with Piasecki to commercialize it.

        So no, hydrogen-fueled hybrid power trains are not fully developed yet, but the R&D looks promising and given how many people are injured and killed each year in loss-of-engine accidents they can’t get here too soon. Remember, the P&W R-2800 didn’t exist until it did.

      • Endless repetition of the “CO2 saturation” fallacy does not make it any less false.

        Radiational cooling of the earth occurs only above the tropopause, because IR absorption of the atmosphere as a whole is already saturated. Heat transfers from the surface to the tropopause by convection, which is subject to adiabatic cooling.

        Adding more CO2 (and other greenhouse gases) to the atmosphere raises the tropoause, causing more adiabatic cooling, so the surface temperature has to increase to maintain the same temperature at the tropopause. There is no limit to this effect. See

        The “medieval warm period” and “little ice age” were local variations and are dwarfed by the temperature rise over the past century. See

        See for correlation with recorded CO2 levels

        • Endless peddling of the false claim that humans are causing runaway warming of climate does not make it any less false.

          Reality of the MWP is that it was warmer than today but climate stable, look at Viking farming remains in southwest Greenland, look at tree remains north of the present tree line in Canada.

          Examine the theories call ‘models’ with their far-too-large grid and omission of various factors.

    • If you are thinking of the Hindenberg, they did not use hydrogen in an exothermic reaction setting.

  3. IOW:
    – lack of electrical technology knowledge
    – not thinking things through
    – evasion of a repeat problem

    I credit Elon ‘The Mouth’ Musk for one good maxim – ‘FixIt! don’t futz around.’ (Perhaps learned the hard way in his rocket efforts, Boeing should listen.)

    • Consulting with diesel-electric locomotive design engineers would have prevented this lock-out situation. They have been using electric motors for years in the alternating power phases of electric motor vs electric generator.

  4. There’s significant detail in the AAIB report, of how the system and fuel cell works

    Noteworthy is fuel cell temperature has to be maintained within limits.

    Ballard from the Vancouver BC area were pioneers of fuel cells, which have been used in specialty applications including a remote unmanned device by International Submarine Engineering. (Who seemed thorough designers.) However attempts to make a ‘reformer’ to convert petroleum based fuels into whatever fuel cells used have not succeeded. (There may be fuel cells that use plant-based liquid alcohols, ethanol from corn/sugar and methanol from trees and such, check an auxiliary power unit in RVs in Europe.)

    (I note they weren’t very high for testing margin, especially given past problems.)

  5. Hey I just had a great idea! Instead of all this complicated “inverter lockout” stuff, why not just put all that hydrogen in a big bag, hang your Malibu underneath it, and then off you go! You could call it the Malidenburg.

  6. Another good solution would have been to leave the piston or turbine in the Mirage to begin with and go have fun with it. It was a nice airplane.

  7. This whole situation is a “simple” electronic design problem. Railroad diesel-electric locomotive designers fixed the problem almost 30 years ago, calling it “dynamic braking:” switching from electric motor power-using to electric motor power-generating.

    • Diesel-electric locomotives are just an engine , a generator, some series-parallel controls and some traction motors geared to the truck axles.
      No inverters used.

  8. So the inverter doesn’t like to be ‘back driven’…! I’m sure that the manufacturer knows this, but wasn’t aware that the prop driven motors could generate some excessive voltage, and shit them down.

    These high power electric aircraft are much more complex than an avgas engine, which is mostly mechanical, vs a ‘whole lot of electronics and software’… and multiple cooling systems !!

    I’m an electrical engineer and I’d never want an electric car, let alone and e-aircraft.!

  9. It is interesting that so many people seem to be afraid of hydrogen, but think nothing of riding around with many gallons of an equally flammable liquid in their cars and airplanes. From a safety standpoint, hydrogen is actually safer than gasoline. It burns with a low-luminosity flame that emits little radiant heat. Being much lighter than air, any that is spilled or leaks out will quickly rise and disperse. Gasoline is heavier than air, its vapors hang close to the ground, and it burns with a highly radiant flame that can burn you even if you are some distance from the flames. Hydrogen’s biggest problem is that it must be stored as a cryogenic liquid, requiring exotic materials for containment, and specialized handling techniques. You can’t fill up your plane and then let it sit for days before flying again. Do that and the fuel will vaporize and disappear. So, it’s a great fuel, but not something your average line jockey would know how to handle.

  10. I find their choice of testbeds highly questionable, why stick an early development powerplant in a high performance single? How about something a bit more forgiving like a Cessna 206 or Cherokee 6, give the pilot a sporting chance when something goes wrong. I suppose using a stodgy old airframe is less shiny and attractive to those willing to bleed cash for the Green effort.