A New Perspective on Battery-Electric Aviation, Part II:

Conceptual Design of a 90-Seater

Reynard de Vries∗, Rob E. Wolleswinkel†,

Elysian Aircraft, Rendementsweg 2, 3641SK Mijdrecht, The Netherlands

Maurice F. M. Hoogreef‡ and Roelof Vos§

Delft University of Technology, Kluyverweg 1, 2629HS Delft, The Netherlands

Battery-electric aviation is commonly believed to be limited to small aircraft and is therefore

expected have a negligible impact on the decarbonization of the aviation sector. In this paper
we argue that, with the correct choice of design parameters and top-level aircraft requirements,
the addressable market is actually substantial. To demonstrate this, the Class-II sizing of a
battery-electric 90-seater is performed, and the environmental impact is assessed in terms of
well-to-wake CO2 -equivalent emissions per passenger-kilometer. The resulting 76-ton aircraft
achieves a battery-powered useful range of 800 km for a pack-level energy density of 360
Wh/kg. For this range, it has an energy consumption of 167 Wh per passenger-kilometer
and an environmental impact well below that of kerosene, eSAF, or hydrogen-based aircraft
alternatives and comparable to land-based modes of transport. These results indicate that, to
successfully reduce the climate impact of the aviation sector, battery-electric aircraft should
not be designed as a niche product operating from small airfields but as commercial transport
aircraft competing with fuel-based regional and narrowbody aircraft.

Nomenclature
AEO = All engines operative 𝐶𝐷 = Drag coefficient, 𝐷/(0.5𝜌𝑉 2 𝑆) [-]
BM = Battery mass 𝐶𝐷0 = Zero-lift drag coefficient [-]
CASK = Cost per available seat-kilometer 𝐶𝐿 = Lift coefficient, 𝐿/(0.5𝜌𝑉 2 𝑆) [-]
DoD = Depth of discharge 𝐶 𝐿,max = Maximum lift coefficient [-]
ECS = Environmental control system 𝐷 = Drag [N]
EoL = End of life 𝐷 fus = Fuselage diameter [m]
eSAF = Synthetic (power-to-liquid) SAF 𝑒 = Oswald factor [-]
eVTOL = Electric vertical take-off and landing ℎ = Altitude [m]
EV = (Ground based) electric vehicle 𝑙 = Tail moment arm [m]
HT = Horizontal tail 𝐿 = Lift [N]
IPS = Ice protection system 𝑙 fus = Fuselage length [m]
LTO = Landing and take-off cycle 𝑀 = Mach number [-]
MTOM = Maximum take-off mass 𝑁 = Number of propellers [-]
OEI = One engine inoperative 𝑃s = Shaft power [W]
OEM = Operating empty mass 𝑄 = Interference factor [-]
PLM = Payload mass 𝑆 = Wing planform area [m2]
PMAD = Power management & distribution 𝑠L = Landing distance [m]
RAM = Regional air mobility 𝑠TO = Take-off distance [m]
SAF = Sustainable aviation fuel 𝑇 = Thrust [N]
STOL = Short take-off and landing 𝑉 = Flight speed [m/s], volume [m3]
TMS = Thermal management system 𝑉app = Approach speed [m/s]
VT = Vertical tail 𝑊TO = Maximum take-off weight [N]
𝐴 = Wing aspect ratio [-] 𝜂p = Propeller efficiency [-]
𝑏 = Wing span [m] 𝜌 = Ambient air density [kg/m3]
∗Head of Design & Engineering, reynard@[Link], AIAA member.
†Co-CEO and CTO, rob@[Link], AIAA member.
‡Assistant Professor, Faculty of Aerospace Engineering, [Link]@[Link], AIAA member.
§Associate Professor, Faculty of Aerospace Engineering, [Link]@[Link], Associate Fellow AIAA.

1
Copyright © 2024 by Reynard de Vries, Rob E. Wolleswinkel, Maurice F. M. Hoogreef and Roelof Vos.
This document has been published by the American Institute of Aeronautics and Astronautics, Inc., with permission, and is available at [Link]
 I. Introduction
ifferent pathways are currently being investigated to reduce the climate impact of the aviation sector [1]. Many
D studies conclude that electric aircraft are only feasible for very short ranges (200∼400 km) with both current
and future battery technology (200∼500 Wh/kg) [2–5]. This has led to the common perception of battery-electric
aviation being limited not only to short ranges, but also to small aircraft. In this two-part paper series, we challenge this
perspective to show that, with the right design choices, the useful range achievable with batteries is substantially higher
than literature suggests. And while the range is still significantly lower than fuel-based aircraft, this increase justifies a
fundamental shift in perspective regarding the potential role of battery-electric propulsion in the decarbonization of the
aviation sector: rather than designing electric aircraft for niche or emerging markets, they can be designed to compete
with large fuel-based aircraft.
Here, the concept of “large fuel-based aircraft” refers to aircraft in the CS-25/FAR Part 25 category with more than
e.g. 50 seats: regional turboprops, regional jets, narrowbodies, and widebodies. These aircraft cover the vast majority
of passenger-kilometers travelled by air and are also responsible for the majority of aviation-related emissions [5, 6].
Among these aircraft types, the narrowbodies—and not the regionals—dominate the market on short ranges (< 2000
km) [7]. This is largely driven by their low cost per available seat-kilometer (CASK) compared to regional jets and
turboprops, as a result of their higher passenger count and flexibility to also operate on longer routes within the airline’s
network. Thus, to maximize the impact of electric aviation on the sector as a whole, electric aircraft would have to be
designed to compete with aircraft such as the Airbus A320neo and Boeing 737MAX in terms of operating costs, while
achieving a useful range which is as high as possible.
Part I of this paper series [8] describes four key design principles that enable an increase in the useful range of
electric aircraft. First, electric aircraft have a high empty mass but a low empty mass fraction (where “empty mass”
excludes the batteries). This is a result of their intrinsically high energy mass (i.e. battery mass) fraction. Second,
electric aircraft inherently have a high lift-to-drag ratio due to their low wetted-area-to-reference-area ratio. Third,
using a fuel-based reserve energy system to cover reserves is a much lighter solution than to carry that energy in the
form of batteries, without impacting the day-to-day emissions and energy costs of the aircraft. And finally, increasing
the passenger count of an electric aircraft does not limit its range. To the contrary—for the same energy efficiency,
designing an electric aircraft for more passengers often leads to an increase in maximum range.
This follow-up paper, Part II, has two purposes: to verify these design principles by performing a more in-depth
design of an electric aircraft, and to demonstrate that the impact of such aircraft on the climate footprint of the aviation
sector can be substantial. For this, the question is not: “How do we design an electric aircraft to mimic existing
aircraft as much as possible and fit into existing networks?” But, instead: “What should the top-level requirements and
operations of such an aircraft look like, if we want to maximize the decarbonization potential of this new technology?”
Therefore, rather than providing the design requirements as input, we first discuss how they are selected in Sec. II. Then,
in Sec. III, the conceptual design of a battery-electric aircraft is performed for the selected set of requirements. Finally,
Sec. IV performs an environmental comparison to other means of transport. In this process, we identify a series of
technical challenges that must be addressed in the development of such aircraft, which are finally summarized in the
Conclusions (Sec. V).

II. Selection of Top-Level Aircraft Requirements

A. Design Objective
The design objective is to achieve an aircraft with zero emissions during day-to-day operation which can operate
cost-effectively and from the same airports as narrowbodies. With this objective in mind, the top-level aircraft
requirements and aircraft configuration were iteratively evaluated. The following sections describe the top-level aircraft
requirements which were finally selected.

B. Payload-Range Requirements
When designing an aircraft, most mass (and CASK) contributions scale less than linearly with payload mass. As a
result, increasing the passenger count is generally beneficial for the energy efficiency per passenger-kilometer. There
are, however, several reasons why it is not practical to indefinitely increase the number of passengers, such as market
considerations (insufficient demand to cover a route effectively), infrastructure considerations (e.g. airport span
constraints), physical limits (e.g. square-cube law [9]), or technology availability (e.g. components for multi-MW,

2
multi-kV electrical power transmission). Based on these considerations and preliminary calculations for 19, 40, and
90-seater configurations, a payload requirement of 90 passengers is selected. In this trade-off, the airport gate span
constraint is found to play a particularly important role, as evidenced in earlier work [10].
Contrary to fuel-based aircraft, where the harmonic range for which the aircraft is designed is based on economic
considerations and not on physical limitations of the maximum range, for electric aircraft maximizing the achievable
range is key to ensure a large market potential (see Sec. IV.D). Thus, in this study we investigate what battery
characteristics would be required to reach a 1000 km target, and then assess which ranges are achievable with more
“near-term” battery technologies (see Sec. III.C). Moreover, for reserves, three components are considered: diversion
range, loiter time, and contingency. These are covered by a dedicated fuel-based reserve energy system. The aircraft is
therefore technically a hybrid, although it is described here as “battery-electric” since it only uses the reserve energy
system for emergencies. Thirty minutes loiter and 5% contingency are assumed to be applicable based on current
regulations [11]. Note that, while it is unclear whether electrically-driven large aircraft will have to satisfy a 30 mins or
45 mins loiter time, this does not appreciably affect the design proposed in this paper due to the use of a fuel-based
reserve energy system. For the diversion to alternate, a range of 150 km is assumed, flown at 𝑀 = 0.4 and 1800 m
altitude. These values are summarized in Table 1.

Table 1 Summary of main top-level aircraft requirements.

Parameter Value Parameter Value

Number of passengers [-] 90 AEO climb gradient [-] 8%
Mass per passenger [kg] 100 Contingency energy [-] 5%
Range (target) [km] 1000 Diversion range [km] 150
Take-off field length [m] 2000 Diversion Mach number [-] 0.4
Landing distance [m] 2000 Diversion altitude [m] 1800
Approach speed [m/s] 74.6 Loiter time [mins] 30
Cruise Mach number [-] 0.6 Battery charging time [mins] <45
Cruise altitude [m] >7000 Wing span at gate [m] <36

C. Performance Requirements
Unlike gas turbines and reciprocating engines, electric motors do not present a power lapse with altitude. As a result,
the installed power is generally constrained by the take-off distance requirement (see Sec. III.A.6). Reducing the power
required for take-off and climb is therefore key to reducing the powertrain mass, which is especially important given the
high powertrain mass fraction of (hybrid-)electric propulsion systems [12]. In other words, a longer take-off distance is
more beneficial for the empty-mass fraction of an electric aircraft than for a turbofan or turboprop aircraft. We therefore
argue that, for maximum energy efficiency, the electric aircraft should be designed for the same “long” runways as
narrowbody aircraft, and not for small airfields.
Based on this, a typical narrow-body take-off distance of 2000 m is selected as take-off balanced field length and
as landing distance requirement. This is contrary to most design studies on electric aviation, which generally size
the aircraft to match the take-off and landing distance of turboprop aircraft [13, 14], or even for short take-off and
landing (STOL) capabilities [15]. Given the amount of narrowbody-compatible runways in regions like Europe and
North America, we hypothesize that the reduction in empty weight fraction—and thus, the increase in range and energy
efficiency—enabled by a “long” runway is more relevant for the overall market potential of the electric aircraft, than the
ability to operate from airfields with runways below 2000 m length. This implies that the aircraft will not be able to
compete with e.g. turboprop aircraft on specific routes. Moreover, to limit the approach speed to values corresponding
to existing large passenger aircraft, an approach speed of 145 kts (74.6 m/s) is specified.
The cruise speed is selected to be as high as possible, while remaining in the subsonic regime to allow for
aerodynamically-efficient unswept wings and propeller propulsion. This trade-off results in a cruise Mach number
requirement of 𝑀 = 0.6. From an operational perspective one could argue that the cruise speed should be derived from a
block time requirement instead. However, given the importance of maximizing the range in the case of electric aircraft,
a cruise speed which enables high aero-propulsive efficiency is prioritized over a cruise speed which reduces block time.

3
In any case, given the relatively short range of electric aircraft, higher cruise Mach numbers would reduce the block
time of the mission by several minutes at best.
The cruise altitude is treated as dependent variable, and computed to ensure cruise flight near maximum 𝐿/𝐷.
No constraints are included in terms of one-engine-inoperative (OEI) ceiling, since early analyses indicated that this
would not be a limitation for electrically-driven propulsors without power lapse. Additionally, the AEO and OEI climb
performance requirements of EASA CS-25 [16] are applied. Since it is currently unclear how regulations will evolve for
such aircraft, “OEI” is treated here as the failure of just one of the 𝑁 propellers of the aircraft, while the most restrictive
value (corresponding to four-engine aircraft) is selected for climb-gradient requirements. Finally, to verify that the
aircraft has enough power to climb after take-off, an AEO climb gradient of 8% at sea level is added to the performance
constraints.

D. Operational Requirements
To compete effectively with narrowbody aircraft, several narrowbody-like constraints are imposed on the design. Firstly,
the aircraft must fit in category C airport gates, i.e. the span must be below 36 m. To limit the turnaround time, a
maximum battery charging time of 45 minutes is imposed. Note that this time is only necessary if the aircraft fully
depleted its energy reserves in the previous mission. In most cases, the aircraft does not fly the harmonic mission, and
therefore only part of the battery capacity has to be recharged. Thus, the average charging time is less than 45 mins. The
battery system must be designed such that battery replacement can be performed during planned maintenance intervals
in hangars with dedicated tooling. This allows for battery-pack designs with a higher degree of integration with the
airframe, compared to swappable or line-replaceable units.
Due to the relatively short range of the aircraft, a web-like network of short routes would greatly enhance its
decarbonization potential, as discussed in Sec. IV.D. In such a network, the aircraft would often operate from secondary
airports (with a runway of at least 2000 m) and with frequent stopovers. This implies that several aspects of the aircraft,
such as the cabin layout and luggage compartments, must be designed with stopovers in mind. These aspects are not
discussed in further detail in this paper. However, both when operating from large international and secondary regional
airports, noise is key for societal acceptance. Therefore, the aircraft should comply with ICAO Annex 16, Chapter 14
[17] limits, and ideally produce less noise than the latest-generation turboprop and turbofan aircraft. While no detailed
noise assessment has been performed at this stage, several qualitative considerations are discussed in Sec. IV.E.

III. Conceptual Design of a Battery-Electric 90-Seater

To verify the parametric analyses of Part I [8], an electric aircraft is sized for the top-level aircraft requirements specified
in Sec. II, using a Class-II design method. The basis of the method is described in Ref. [18], is validated in Ref. [19],
and is expanded to account for the Class-II weight breakdown of Torenbeek [9] and the drag component buildup of
Raymer [20]. The energy requirements for the battery and fuel are computed using a time-stepping mission analysis
which accounts for climb, cruise, descent, diversion, and loiter. Assumptions regarding the efficiency and specific power
of the powertrain components, take-off and landing performance modeling, energy consumption of non-propulsive
systems, etc., are provided in Appendix A.

A. Key Design Choices

In the design-space exploration phase, several technologies (boundary-layer ingestion, ultra-high aspect-ratio wings,
tip-mounted propellers, etc.) and configurations choices (high vs. low wing, V-tail vs. T-tail, 4 vs. 5 abreast, etc.)
were evaluated. For brevity, only the main design choices leading to the final configuration are discussed here. Most of
these design choices focus on minimizing the empty-mass fraction. Despite evaluating several radical technologies, the
exploration phase converged towards a relatively conventional tube-and-wing design, as can be seen in Fig. 1. The main
difference compared to conventional configurations becomes evident when comparing the proportions of the aircraft to
a fuel-based counterpart. Figure 1 shows how the fuselage of the electric aircraft is significantly smaller than that of an
A320, since it carries half the amount of passengers. However, the wing is significantly larger. In other words, the
electric aircraft has a large wing relative to its fuselage—as explained by some of the scaling effects discussed in Part I
[8]. In the following subsections, the design choices made for this aircraft are further elaborated.

4
 15o

36.0

3.7

9o

42.0

7.5o

Fig. 1 Three-view of the 90-seater battery-electric aircraft. Top view includes A320 planform for scale in gray
and the location of the reserve energy system, indicated with dotted lines.

1. Wing Design
The wing combines several characteristics to minimize the empty weight fraction of the aircraft:

• Batteries in wing: Placing the batteries in the wing is crucial to minimize the structural weight of the aircraft,
since it reduces the root bending moment compared to having batteries in the fuselage, and thereby reduces the
wing structural weight—assuming the wing structure is predominantly sized by the 2.5g pull-up maneuver, as is
generally the case with commercial aircraft [21]. This design choice comes with its own challenge regarding access
to batteries for replacement and ensuring safety in case of cell or module failure (e.g. thermal runaway∗). This
requires an integrated design of the battery-wing system where several compromises have to be made (Challenge
1). However, the substantial weight reduction enabled by placing batteries in the wing gives ample room for
new integration solutions. For the design proposed in this paper, the battery pack occupies less than 50% of the
volume of the wing box (i.e. the space available between the front and rear spars) if a pack-level mass density of
3000 kg/m3 is assumed. This provides room for a smart packaging of the batteries that facilitates inspection and
replacement when necessary.
• Low-wing configuration: A low wing leads to a shorter, and thus lighter, main landing gear. The combination of
multiple, smaller propellers and wing dihedral allows for a low wing while maintaining an acceptable ground
clearance, as indicated in Fig. 1. A high wing configuration with batteries located in the wing is expected to
be particularly heavy, since the fuselage would have to be reinforced to transmit loads to the landing gear (for
fuselage-mounted landing gear) or absorb loads in case of a belly landing (for wing-mounted landing gear).
• Modest aspect ratio: generally, a high aspect ratio is targeted for electric aircraft, due to the importance of
achieving a high lift-to-drag ratio (L/D) (see e.g. [23]). However, there are several reasons why a more modest
aspect ratio is more beneficial in this case than for conventional aircraft. First, because a lower A implies a
smaller wing span, making it easier to satisfy the airport span constraint. Second, because the wing volume
scales with 𝑉 ∝ 𝑆 3/2 𝐴 −1/2 , leaving more volume for batteries and other powertrain components inside the wing.
And third, because a lower aspect ratio is expected to lead to less critical aeroelastic and landing loads, which is
beneficial for wing structural weight. At this stage it is unclear if these load cases are strongly limiting for the
wing structural weight (Challenge 2), since placing the batteries in the wing alleviates the 2.5g pull-up maneuver
∗Note that for the battery to be certified in the first place, the module itself must prohibit failure propagation. See for example Ref. [22] and its
list of references.

5
 load but deteriorates the landing loads, compared to conventional aircraft. Based on these considerations, an
aspect ratio of 𝐴 = 12 is selected, similar to existing turboprop aircraft.
• Folding wing tips: even with a “modest” aspect ratio, the high maximum take-off mass (MTOM) combined with
a low wing loading (Sec. III.A.6) leads to a large wingspan, which makes it challenging to fit within the gate span
constraint while maintaining an acceptable aerodynamic performance. To circumvent this at the expense of a
slight increase in wing weight, foldable wingtips are employed, similar to ones on the Boeing 777X.

2. Landing-Gear Placement
The main landing gear is installed behind the rear spar and is retracted rearwards into the pods visible in Fig. 1. There
are two main reasons why a wing-mounted landing gear is preferred over a fuselage-mounted gear. First, since it is
placed further outboard, a shorter gear is needed to satisfy the propeller ground clearance constraint in case of a banked
landing. Second, it minimizes bending moments in the wing during landing or a taxi bump. This configuration also
comes with drawbacks, such as reduced space for trailing-edge high-lift devices, or a larger turn radius on the ground.
The landing-gear pod also contributes to weight and friction drag. However, the alternative fuselage-mounted gear
would also require a large fairing: since the fuselage is relatively small (driven by payload weight) and the gear is
relatively large (driven by max landing weight), it is impossible to store the main gear inside the fuselage itself.

3. Propulsion System Integration

The number of propellers is determined in a trade-off between disk loading, propeller noise, landing gear length, OEI
power requirements, motor rotational speed, nacelle drag, and slipstream effects on wing performance. This leads to a
configuration with eight propellers covering the full available wing span, while maintaining sufficient clearance between
propellers to limit interaction noise, which decreases rapidly as the separation between adjacent propellers is reduced
[24]. The resulting low disk loading (377 N/m2 in cruise, compared to typical twin-turboprop values of 500 − 800 N/m2)
ensures a high propeller efficiency in cruise and, especially, during take-off. The latter is particularly important due to its
impact on the required motor power in the sizing condition. Moreover, while some studies suggest that using a gearbox
is lighter than a direct-drive solution [25], we initially opt for a direct-drive arrangement to reduce the complexity of the
system. This implies that the electric motors have to operate at low rotational speeds, in the order of 800 to 1400 rpm.

4. Reserve Energy System

A gas-turbine-based turbogenerator is located in the tail cone to provide power to the batteries and motors in the case
of diversion or loiter. The term “range extender” is avoided here, since the primary purpose of the reserve energy
system is to cover reserves, and not enhance mission range. For the same reason, the aircraft is described as “battery
electric”, although technically it is a serial hybrid architecture. Note that the turbogenerator could in fact be used as
“range extender” without major changes to the aircraft but, in that case, the emissions and operating costs of the aircraft
are found to increase. Several decisions remain regarding the design of the reserve energy system (Challenge 3). For
example, it is undecided whether for the envisioned application, two gas turbines would be required for redundancy.
Although that would have a substantial impact on aircraft weight and costs, it remains a much lighter solution than to
use the batteries to cover reserves [8]. In this paper, a single turbogenerator is assumed.

5. Tail Configuration & Fuselage Cross Section

The required horizontal-tail surface area scales with 1/𝑙, with 𝑙 being the distance between the aerodynamic center of the
wing and the aerodynamic center of the tailplane. Since the electric aircraft has a large wing relative to the fuselage, the
tail is closer to the wing (in a non-dimensional sense) than in conventional aircraft. Therefore, to minimize horizontal tail
size—and thus, drag—a T-tail configuration with a swept vertical tail is selected. During the design-space exploration
phase, the choice for a 4-abreast arrangement (heavier fuselage) over a 5-abreast arrangement (lighter fuselage) was
also primarily driven by the reduction in tail size enabled by the more slender fuselage. A T-tail configuration also
ensures that the propeller slipstreams do not impinge on the horizontal tail at low to moderate angles of attack, although
additional analyses are required to assess the impact at high angles of attack and mitigate deep stall.

6
6. Wing & Power Loading
Figure 2 shows the wing/power loading diagram of the aircraft. The area in gray indicates the combinations of wing
loading (i.e. wing size) and power loading (i.e. powertrain size) which satisfy all performance requirements. Assuming
a maximum lift coefficient of 𝐶 𝐿max = 2.5, the landing distance required is met with a wing loading of approximately
5 kN/m2. This modest 𝐶 𝐿max is lower than typical narrowbody aircraft, and may be attainable with a plain flap
with propeller slipstream effects, though that raises the question whether future regulations will be adapted to allow
determination of the stall speed in power-on conditions, in the case of distributed-propulsion aircraft. The 𝐶 𝐿max can
also be increased using more complex high-lift devices, but this is counterproductive: when evaluating the design at
higher wing loadings, the MTOM increases. Previous studies [12] have already shown that this occurs because the
take-off requirement is the active constraint, and therefore a higher wing loading leads to a lower power loading and,
subsequently, a higher powertrain weight. This is contrary to other electric aircraft design studies which attempt to use
distributed propulsion to maximize the wing loading (see e.g. Ref. [23]). On the other hand, an excessively low wing
loading leads to a large wing, which in turn makes it difficult to satisfy the maximum span constraint. In this study, we
assume a maximum folding wingtip span of 3 m per side to ensure that the ailerons, which are located inboard of the
folding location, retain sufficient control authority (for context, note that the Boeing 777X has folding wingtips of 3.5 m
length). Further research is required to determine whether larger or smaller folding wingtips are possible. With this
assumption and the gate span constraint of 36 m, the maximum wingspan of the aircraft can be 42 m.

0.20

Vapp= 145 kts, CLmax = 2.5

Vapp= 145 kts, CLmax = 3.1

sL= 2000 m, CLmax = 2.5

sL= 2000 m, CLmax = 3.1

Feasible design space
Approach speed
Landing distance
Cruise speed
Take-off distance
0.15 AEO take-off climb grad.
Shaft power loading WTO/Ps [N/W]

Diversion cruise speed

CS25 performance reqs.
CS25
Span constraint .121a
Selected design point CS25
.121c
0.10 CS25
.121b
m)
00

CS25.1
18

) CS25.12 19
=

0m 1d
(h

0
76
.4

AEO 8%
=0

(h=
M

0.6
b ≤ 42 m

0.05 =
M sTO =
2000
m

0.00
0 1 2 3 4 5 6 7 8
Wing loading WTO/S [kN/m2]

Fig. 2 Wing-loading/power-loading diagram of the battery-electric aircraft. The power values refer to the
maximum continuous shaft power of the electric motors.

Based on these considerations, the wing loading of approximately 5 kN/m2 is selected as design point. At this
wing loading, the power loading is determined by the take-off constraint, while the cruise speed constraint can be met
with much higher 𝑊TO /𝑃s values. The difference between take-off power and cruise power implies that the cruise
speed can be increased beyond 𝑀 = 0.6 without imposing any penalty on powertrain weight. While some high-speed
propeller designs have shown constant efficiencies up to a cruise Mach number of 0.75 [26], a more detailed analysis

7
of the propeller and wing aerodynamics is required to determine how much the Mach number can be increased with
deteriorating the aero-propulsive efficiency of the aircraft, 𝜂p (𝐿/𝐷).

B. Sizing Results
The resulting mass breakdown of the aircraft is shown in Fig. 3. If we consider the fuel of the reserve energy system as
part of the empty mass for bookkeeping purposes, we see that an empty-mass fraction of 42% is obtained, including the
reserve energy system. This is in line with the Class-I analyses of Part I [8]. The figure also indicates a battery mass
fraction of 46%. Placing 46% of the aircraft mass inside the wing is highly unusual for turboprop aircraft, but energy
fractions of around 45% are not unheard of for jet aircraft—see e.g. the Boeing 707-320 or DC-8-63.

Packaging

Cells

Payload
Battery
46.0% Payload
11.8%
Operational items
MTOM = 76 t
Miscellaneous
ECS & IPS
Furnishing
Electrical systems
Hydraulics/pneumatics
OEM Avionics
42.2% Thermal management system
Wing PMAD
Motors (incl. inverters)
Horizontal tail Generator (incl. rectifier)
Vertical tail Propellers
Fuselage
Landing Oil & fuel systems
gear Fuel Gas turbine
Surface Nacelles
controls

Fig. 3 Class-I (inner circle) and Class-II (outer circle) mass breakdown of the 90-seater battery-electric aircraft
at maximum take-off mass. Note that for bookkeeping purposes, the OEM includes fuel mass of the reserve-energy
system.

Although these results are encouraging, there are also several uncertainties which require further verification. For
example, the wing structural mass is computed using handbook methods [9], which are partially based on statistical data.
The question is whether such correlations remain accurate in a configuration where different load cases may be driving
the structural mass (Challenge 2). Analogously, there is uncertainty regarding the masses of the high-voltage (Challenge
4) and thermal management system (Challenge 5) components of the powertrain, which are estimated using simple
assumptions for power density based on literature (see Appendix A). The same applies for non-propulsive systems such
as the flight control system (FCS), ice protection system (IPS), or environmental control system (ECS), for which an
increased degree of electrification may lead to different architectural choices and weights, compared to conventional
aircraft (Challenge 6). More detailed investigations are required to confirm the accuracy of these assumptions. And
finally, there is uncertainty in the assumptions made with respect to battery cells and packaging. These characteristics
have a large impact on MTOM and are discussed in the following section.
To illustrate how scaling effects in electric aircraft influence the lift-to-drag ratio, Fig. 4 compares the drag breakdown
of the electric aircraft to a fuel-based turboprop reference aircraft designed with the same tool for the same mission, in
cruise conditions. The figure shows how the lower fuselage drag coefficient is the main reason for lower drag coefficient
in the electric aircraft. A slightly lower induced drag coefficient is also achieved, due to the higher Oswald factor (less

8
fuselage disturbance on the wing). Overall, this leads to a relatively high lift-to-drag ratio of 𝐿/𝐷 = 23. However, also
here there are some uncertainties. For example, the drag due to heat exchangers cannot be accurately modeled at this
stage (Challenge 5), and a more detailed control & stability analysis is required to calculate the required tail area for this
distributed-propulsion aircraft (Challenge 7). Moreover, the effect of the propellers and—especially—the nacelles on
wing performance must be assessed in more detail. This has been investigated to a certain extent in previous studies
[27], but is especially critical for high-lift performance (Challenge 8). In any case, the resulting lift-to-drag ratio is
significantly higher than the fuel-based counterpart.

350 Induced drag 25

Heat exchangers
300 Leakage & protub.
Gear fairing 20
Drag coefficient CD × 104 [-]

Lift-to-drag ratio L/D [-]

250 Nacelles
Tail
Wing 15
200
Fuselage cruise condition
150
10

100
5
50
Electric aircraft
Reference turboprop
0 0
Reference Electric 0.0 0.5 1.0 1.5 2.0
turboprop aircraft Lift coefficient CL [-]
a) Drag breakdown in cruise b) Lift-to-drag ratio polars

Fig. 4 Drag breakdown and lift-to-drag ratio of the battery-electric aircraft compared to a conventional
twin-turboprop reference configuration, designed for the same mission. Propeller interaction effects neglected.

To summarize, Table 2 presents an overview of some of the main characteristics of the aircraft. Note the high
estimated propeller efficiency as a result of the low disk loading.

Table 2 Overview of main design results. The quoted empty-mass fraction, OEM/MTOM, includes the mass of
the reserve energy system (i.e. gas turbine, generator, fuel, and accessories).

Parameter Value Parameter Value

MTOM [t] 76.0 𝐿/𝐷 (cruise) 23.0
OEM/MTOM 42% Max cont. power per motor [MW] 1.4
BM/MTOM 46% Propeller efficiency (cruise) 0.87
PLM/MTOM 12% Propeller diameter [m] 3.7

C. Range Versus Battery Technology Scenarios

The achievable range depends on the battery energy density, and the achievable energy density depends on other
characteristics of the battery. In this study, we make the following assumptions:

• A maximum (dis)charge rate of 1.35C stems from the top-level aircraft requirement of 45 mins charging time.
Since the aircraft carries a significant amount of energy in the batteries—of the order of 11 MWh—the peak
discharge rate during take-off is relatively low (∼1.3C). These C-rate requirements are substantially lower than
those of ground-based vehicles and eVTOL aircraft [28].

9
 • A packaging is required to connect, manage, and ensure the structural integrity and safety of the cells. We
assume a 25% packaging mass overhead (i.e. a cell-to-pack ratio of 0.8). This target is aligned with the targets
established in several roadmaps [29] and eVTOL manufacturers have been said to achieve overheads below 25%,
though the question is whether this is also achievable with higher-energy-density cells [30]. For this 25% overhead,
a smart integration solution of the battery pack on the aircraft is required, with airframe elements potentially
contributing to structural, thermal, or other safety functions of the pack (Challenge 1).
• The battery has a maximum depth-of-discharge of 90%, as suggested by Ref. [31]. In other words, the
state-of-charge can vary from approximately 95% to 5%. Note that this maximum discharge is only required
on very limited occasions, when the aircraft flies its maximum range and is additionally required to perform a
go-around.
• An end-of-life capacity of 90%, i.e. the battery is replaced after 10% capacity degradation. Although this is
important for sizing the thermal management system, it is more of an operational choice than an assumption.
Retiring the battery “ahead of time” is required to maintain range capabilities. However, an operator that only flies
short routes may opt to retire the battery pack after >20% degradation in order to extract more cycles.
• Battery cycle life is an important driver for operating costs and environmental impact. In our calculations, we
assume the battery is replaced after 1500 cycles. However, we must realize that an aircraft does not always fly the
design range, but a distribution of ranges instead. For the distribution of flights provided to this study by a major in-
ternational airline, roughly 30% of flights (i.e., only 450 cycles) would correspond to “deep” discharges above 70%
DoD. These assumptions imply that the battery pack has to be replaced once or twice per year, depending on the us-
age of the aircraft. This should be seen as a planned base maintenance operation, and not a line maintenance activity.

The resulting relation between the energy density of a new cell and the useful, end-of-life pack energy density is shown
in Fig. 5.

100
Energy density relative to new cell [%]

Packaging
80
Max DoD
Degradation
60
New cell New pack
(100%) (80%) New pack EoL pack
40
usable energy usable energy
(72%) (65%)
20

Fig. 5 Battery energy-density waterfall chart.

With the packaging overhead and other characteristics of the battery specified, we define three cell energy-density
scenarios in Table 3: a conservative scenario of 300 Wh/kg (corresponding to current state-of-the-art cell technology), a
baseline scenario of 450 Wh/kg for a 1st generation aircraft, and a future scenario of 550 Wh/kg for a 2nd generation
aircraft (which would be required to achieve the targeted 1000 km range). Note that the cells must be able to provide
these energy densities at the C-rates quoted above; if not, an additional knock-down factor must be considered [32].
The aircraft achieves ranges of 500 km, 800 km, and 1000 km for cell energy densities of 300 Wh/kg, 450 Wh/kg,
and 550 Wh/kg, respectively. This is shown in Table 3, where the range values refer to the useful air range, which takes
into account the energy required for taxi, take-off, reserves, power off-takes, etc. While the 450 Wh/kg and 550 Wh/kg
cases require further cell technology development, they are much less aggressive requirements than the pack-level 500
to 1000+ Wh/kg that literature suggests is necessary to make an electric passenger aircraft viable [5, 28, 31, 33]. The
development of cells with these energy densities and the other performance metrics listed above is one of the key drivers
to further enhance electric aircraft range (Challenge 9).

10
Table 3 Useful air range obtained for various 2030+ battery technology scenarios. Ranges are quoted at
approximately 3% cell capacity degradation (i.e. one-third into the useful life on the aircraft).

Conservative 1st gen aircraft 2nd gen aircraft

Cell energy density [Wh/kg] 300 450 550
Pack energy density [Wh/kg] 240 360 440
Ramge [km] 500 800 1000

IV. Environmental Assessment

A. Alternative Energy Sources for Aircraft Propulsion

Three main energy carriers are generally considered for a future “zero-emission” air transport system: batteries,
hydrogen, and sustainable aviation fuel (SAF) [1]. For a fair environmental assessment of electric aircraft, they should
not only be compared to current kerosene-based aircraft, but also to other aircraft with new technologies that may appear
in the coming decades. For this comparison, we assume that hydrogen is used in liquid form, which offers a higher
volumetric density than gaseous hydrogen and is considered more beneficial for passenger transport aircraft [6]. Among
the different types of SAF, here the comparison is performed with power-to-liquid synthetic fuels (“eSAF”) including
direct air capture, which—to the best of our knowledge—is the only SAF pathway to achieve a net-zero-CO2 -emission
energy carrier that can be scaled up to power the complete aviation sector in the future. While other types of SAF such
as biofuels are also beneficial for emissions, they are an unlikely candidate for decarbonization of the aviation sector as
a whole due to challenges associated with scaling up production to satisfy the demand in a sustainable manner [1, 34].
In this analysis, electricity can be used to produce eSAF, charge batteries, or produce liquid hydrogen for direct
combustion or use in a fuel cell. For an apples-to-apples comparison, these four pathways use electricity from the same
grid†. For zero CO2 emissions, this electricity has to be produced with renewable (“green”) electricity. However, most
energy produced around the world today comes from non-renewable sources. Assuming that 100% green electricity can
be used for battery, eSAF, or hydrogen-powered aircraft in a world where the electrical grid is not 100% renewable, and
implying that other industries must use the non-renewable part, is a misleading argument. Therefore, instead, in Sec.
IV.C we will analyze how different grid cleanliness-scenarios affect the comparison.
Figure 6 shows typical efficiencies associated with the various steps of the four pathways to bring electricity from
the grid to the shaft of propulsor, based on discussions with experts and values encountered in literature [6, 34, 35]. The
figure shows that, for a battery-powered aircraft, roughly 1.3 kWh of grid energy is required per kWh of energy at the
shaft. For the hydrogen propulsion options, roughly 4 kWh of grid energy is required per kWh of shaft energy, and for
eSAF this increases up to 5−9 kWh. In other words, bringing electricity directly to shaft is roughly three to six times
more efficient than using hydrogen or eSAF, in terms of grid energy required per unit energy at the shaft. The following
sections discuss how these ratios translate into differences in energy consumption and emissions.

B. Tank and Grid Energy Consumption

As reflected on the right-hand side of Fig. 6, the powertrain efficiency of a battery-electric aircraft is high compared to
fuel-based aircraft, but a battery-electric aircraft is also very heavy on a per-passenger basis (low payload mass fraction;
see Sec. III.B). Therefore, a priori it is not clear whether the aircraft is energy-efficient on a per-passenger-kilometer
basis. To dive into this and illustrate the importance of accounting for the various pathway efficiencies, in this section
we compare the energy consumption per passenger-kilometer of different modes of transport, for a mission range of the
“1st gen” electric aircraft of 800 km. We compare the aircraft to existing modes of transport (narrowbodies, cars, and the
train), and to hypothetical future competitors of the aircraft (hydrogen aircraft and a next-gen turboprop design). While
these different forms of transport are designed for different ranges, here we are interested in how they perform on ranges
where the electric aircraft may be an alternative. More specifically, we compare:

†One could argue that hydrogen or eSAF can be produced in a part of the world with abundance of renewable energy, and then transported to
other parts of the world. However, in that case the energy (and costs) required for that transport must be included in the analysis.

11
 Power-to-liquid synthetic fuel (eSAF)
Green Grid H2 CO2 direct e-Fuel Transport Gas Propulsion
electricity transporta electrolysis air captureb synthesis turbine
63−68%
5 ~ 9 kWh 94−100% 70−71% 65−73% 98−99% 38−42% 1 kWh
or 100%

Hydrogen turbine
Green Grid H2 Lique- Transport Gas Propulsion
electricity transportc electrolysis faction & boil-off c turbine
4 ~ 5 kWh 94−97% 70−71% 70−83% 97−98% 38−42% 1 kWh

Hydrogen fuel cell

Green Grid H2 Lique- Transport Fuel Electric Propulsion
electricity transportc electrolysis faction & boil-off c cell motord
3 ~ 4 kWh 94−97% 70−71% 70−83% 97−98% 50−60% 85−95% 1 kWh

Battery electric
Green Grid Battery Battery Electric
Propulsion
electricity transport charging discharging motord
~1.3 kWh 94−97% 95−96% 95−96% 85−95% 1 kWh
a
100% if renewable electricity is produced at fuel production site
b
100% if carbon is available from elsewhere
(e.g. non-clean steel or cement production) Grid-to-tank Tank-to-shaft
c
Assuming hydrogen is produced at airport
d
Includes losses of cables, inverters, etc.

Fig. 6 Conversion efficiency of four pathways to bring renewable energy to the propeller shaft: eSAF, H2
turbine, H2 fuel cell, and battery electric.

• The “1st gen” electric 90-seater designed in this study, with a harmonic range of 800 km.
• A hypothetical next-gen, fuel-based, twin-turboprop reference aircraft designed for the same payload (90
passengers), range (800 km), and entry-into-service as the electric aircraft, to demonstrate the effect of battery-
electric propulsion technology while keeping the requirements constant. This is the “apples to apples” comparison
from a technical perspective.
• An A320neo-like narrowbody, which can carry around 180 passengers up to 6500 km. While this aircraft is
designed for much longer ranges and therefore has a suboptimal energy consumption on such a short range, it is
still the most commonly used type of aircraft for these ranges due to economic and operational reasons (see Sec.
IV.D). Therefore, this would be the relevant comparison from a business-case perspective.
• A hydrogen fuel-cell regional aircraft concept of the Aerospace Technology Institute (ATI) [36], designed to carry
75 passengers a distance of roughly 1500 km.
• A hydrogen gas-turbine powered narrowbody aircraft concept designed by the ATI [36], designed to carry 180
passengers a distance of roughly 4500 km.
• A ground-based electric vehicle (EV, i.e. a car).
• An electrically-driven train operating on a densely traveled network.

For the ground-based modes of transport, stops or stopovers may be required to complete a 800 km mission, but here
we focus the discussion exclusively on energy consumption per passenger-kilometer, which we assume to be independent
of range for the car and train. For a fair comparison, we correct the energy consumption of all modes of transport for the
average occupancy (85% load factor on planes, 1.2 passengers in a car, and the data of the train already accounts for this).
The full list of assumptions can be found in Appendix B. The resulting energy consumptions per passenger-kilometer
are shown in Fig. 7 in terms of vehicle (i.e. tank-to-wake) and grid (i.e. well-to-wake) energy consumption. Note that
for the fuel-based aircraft we assess both Jet A1 and eSAF as alternatives, which becomes relevant in the next section.
Figure 7 illustrates that the battery-electric (167 Wh/pax-km) and conventional propeller (187 Wh/pax-km) aircraft
designed for 800 km range are comparable in terms of vehicular energy consumption per passenger kilometer, shown
in dark blue. While the electric aircraft (76 t) is more than twice as heavy as the reference turboprop (28 t), this

12
 1000
Energy consumption [Wh/pax-km] Vehicle (tank-to-wake)
Grid (well-to-wake)
800

600

400

200
N/A

N/A
0
ft

1)

ll
F)

1)

EV

n
F)

in

ai
ce
ra

A
SA

SA

rb

Tr
irc

el
et

et

tu
(e

(e

fu
ca

(J

(J
p

dy

2
H
p

dy
tri

2
ro

H
ro

bo

dy
ec

bo
op

al
op

bo
El

on
w
rb
rb

ro
ro

w
tu

gi
tu

ar
ar

ro
Re
f.

N
f.

ar
Re
Re

N
Fig. 7 Comparison of vehicle and grid energy consumption per passenger-kilometer for various aircraft
configurations and land-based modes of transport. Comparison performed for a mission range of 800 km.

is compensated by a much higher powertrain efficiency and slightly higher lift-to-drag ratio. Both aircraft have a
substantially lower energy consumption per passenger kilometer than the narrowbody at these ranges. The figure also
shows that vehicle energy consumption of the electric aircraft is in the same ballpark as the H2 narrowbody and an
electric car, on a per-passenger-kilometer basis. In this comparison and based on the values reported in Ref. [36], the
H2 narrowbody has a substantially lower tank energy consumption than the fuel-based narrowbody and the H2 fuel cell,
though the reason for this is unclear.
However, the vehicle energy consumption per passenger kilometer at tank level is neither representative of the energy
costs nor of the life-cycle emissions. For that, the energy required from the grid to produce the energy used on the
aircraft is more representative. This is also shown in Fig. 7 and is obtained by multiplying the tank energy consumption
per passenger kilometer by the average grid-to-tank conversion efficiency shown in Fig. 6. The graph shows that
the battery-electric aircraft requires approximately 70% and 80% less grid energy per passenger kilometer than the
eSAF-based regional and narrowbody aircraft, respectively. Compared to the hydrogen-based aircraft, the reductions
are approximately 75% and 40% for the regional (fuel cell) and narrowbody (turbine) configurations, respectively. The
energy consumption of the electric aircraft remains comparable to the ground-based electric vehicle and is roughly twice
as high as the train, on a per-passenger-kilometer basis. These numbers suggest that in terms of both environmental
impact and energy costs, the electric aircraft is a highly efficient mode of transport.

C. Comparison of Emissions
Now that we know the tank and grid energy consumptions per passenger kilometer of the different vehicles, we can
make a preliminary assessment of their environmental impact. Accurately quantifying the environmental impact is very
challenging and thus here we perform only a simplified analysis, the assumptions of which can be found in Appendix B.
While this analysis only gives a first indication of the environmental impact of the various modes of transport, the stark
differences between the different options are larger than the uncertainty bands and allow for meaningful conclusions.
For this simplified analysis, we use the well-to-wake “CO2 -equivalent” per passenger-kilometer as an indicator of the
impact on global warming. In this CO2 -equivalent we account for emissions produced during the mission, emissions
generated in the production of the energy consumed during the mission, and emissions produced in the manufacturing of
batteries. Emissions generated in the manufacturing process of the vehicle or the infrastructure are not included in this
analysis. For in-flight emissions, we consider both CO2 and non-CO2 effects, which can be significant [37]. Since there
is significant uncertainty regarding several of these assumptions, the results are presented in Fig. 8 with conservative
error bars that give a notional indication of the possible error band.

13
 350
Battery manufacturing
Electrical energy production
300 Hydrogen production
eSAF production
gCO2 -equivalent / pax-km

250 Jet A1 production

In-flight CO2 emissions
In-flight non-CO2 emissions
200

150

100

50

0
ft

1)

F)

ll

EV

n
1)

F)

in

ai
ce
ra

A
SA

SA

rb

Tr
irc

el
et

et

tu
(e

(e

fu
ca

(J

(J
p

dy

2
H
p

dy
tri

2
ro

H
ro

bo

dy
ec

bo
op

al
op

bo
El

on
w
rb
rb

ro
ro

w
tu

gi
tu

ar
ar

ro
Re
f.

N
f.

ar
Re
Re

Fig. 8 Breakdown of equivalent CO2 emissions per passenger-kilometer, for various aircraft configurations
and land-based modes of transport. Results presented for a “clean” EU 2030 target grid emission index of 114
gCO2 /kWh. Comparison performed for a mission range of 800 km.

The emissions presented in Fig. 8 correspond to a grid with an emission index of 114 gCO2 /kWh, which is targeted
for 2030 in Europe [38]. This emission index is representative of a grid with a high degree of renewables, and is
comparable to e.g. Denmark in 2019 [39]. For the conventional turboprop and narrowbody, both eSAF and kerosene
(Jet A1) options are included. The figure shows how, for electric vehicles, the contribution of battery manufacturing to
total lifecycle emissions is smaller than the contribution of the electrical energy consumed during operation. The figure
also shows how, for the selected grid emission index, eSAF is not necessarily better than kerosene for conventional
aircraft. Moreover, the hydrogen aircraft present higher CO2 -equivalent emissions than the advanced turboprop. This is
partially due to increased non-CO2 effects and a relatively heavier aircraft (additional mass due to e.g. the hydrogen
tank and the airframe mass required to accommodate it), but also partially because, like the fuel-based narrowbody, they
are over-sized for an average trip length of 800 km.
Two important conclusions can be drawn from Fig. 8. First, despite the significant uncertainties, the battery-electric
configuration presents a clear advantage over the other aircraft configuration in terms of CO2 -equivalent/pax-km. In other
words, “if the mission can be flown electrically, then it should be flown electrically”. From this perspective, kerosene,
hydrogen, or eSAF should only be used for higher ranges that are not achievable with battery-electric propulsion. And
second, if we compare the battery-electric aircraft to the car and train, we see that these lie in the same ballpark in terms
of CO2 -equivalent per passenger-kilometer. However, this does not consider infrastructure requirements, such as the
impact of railway and road construction and maintenance on emissions and land use, nor the fact that cars and trains
generally travel in a less straight line from origin to destination. This suggests that such an aircraft is comparable to
electric land-based modes of transport from an environmental perspective. It also implies that bans on flights for short
distances, such as the ban imposed in France [40], are unnecessary for a battery-electric aircraft.
The aforementioned conclusions are valid for the selected “EU 2030” grid scenario. However, on one hand the
current world grid uses much less renewable energy, and on the other, we strive to obtain a fully renewable grid in the
future. So, what does the comparison look like in those scenarios? For this, Fig. 9 shows the CO2 -equivalent emissions
of the various modes of transport for the 2019 world average emission index (475 gCO2 /kWh [41]), the EU 2030 target

14
(114 gCO2 /kWh; used in the previous figure), and a hypothetical fully-renewable grid (0 gCO2 /kWh). Note how the Jet
A1 aircraft are slightly affected by the grid scenario due to the fuel production phase. The figure shows that, with the
2019 electricity mix, the electric 90-seater is comparable to a next-gen turboprop designed for the same mission, but
already presents less CO2 -equivalent emissions than a latest-generation narrowbody or hydrogen-based alternatives.

600
2019 world average (475 gCO2/kWh)
EU 2030 target (114 gCO2/kWh)
500 Fully renewable (0 gCO2/kWh)
gCO2 -equivalent / pax-km

400

300

200

100

0
ft

1)

F)

1)

F)

ll

EV

n
in

ai
ce
ra

A
SA

SA

rb

Tr
rc

el
et

et

tu
(e

(e
ai

fu
(J

(J
c

dy

2
H
p

dy
tri

2
ro

H
ro

bo

dy
ec

bo
op

al
op

bo
El

on
w
rb
rb

ro
ro

w
tu

gi
tu

ar
ar

ro
Re
f.

N
f.

ar
Re
Re

Fig. 9 Comparison of equivalent CO2 emissions for various grid cleanliness scenarios. Errorbars are omitted to
improve readability; see Fig. 8 for an indication of the uncertainty.

Another relevant takeaway from Fig. 9 is what the authors refer to as the “SAF paradox”: while SAF is generally seen
as a short-term solution because the technology is relatively easy to implement, it only becomes a scalable sustainable
solution on the long term, when the grid is (almost) completely renewable. On the short term, it is either not scalable
(biofuel), or not sustainable (synthetic fuel). Although the use of biofuels does help to reduce global warming and
should therefore be encouraged as much as possible, the overall “decarbonization” potential of SAF will remain limited
until there is an abundance of green energy. Until then, the development of battery-electric and hydrogen technologies is
key to reduce emissions. In fact, an extensive use of battery-electric propulsion for short ranges would allow us to save a
scarce H2 or SAF supply for longer-range missions.

D. Potential Impact on Aviation Sector

Figure 8 shows that the electric aircraft produces approximately 90% lower CO2 -equivalent emissions than the
latest-generation narrowbody flying on Jet A1, for the EU 2030 grid cleanliness scenario. This reduction increases to
nearly 100% if the grid becomes fully renewable. Since the Airbus A320neo and Boeing 737MAX are currently the
latest-generation narrowbodies, we consider their emissions (be it due to the combustion of kerosene or the production
of SAF) to be representative of the average narrowbody fleet emissions in the 2030+ timeframe, as they will gradually
replace existing older aircraft models. However, these aircraft—and especially widebody aircraft—have substantially
higher design ranges than the electric aircraft. This raises the question whether electric aircraft, despite the potential
emissions reduction of over 90% on a per-passenger-kilometer basis, will have a significant effect on the aviation sector
as a whole, or whether the effect is negligible due to the relatively short ranges to which they are limited.
To put this into perspective, Fig. 10 shows the distribution of CO2 emissions produced by commercial passenger
transport, for different mission ranges. Note that this figure excludes non-CO2 and other life-cycle emissions. The data

15
 shown in this figure corresponds to the year 2019, prior to the COVID-19 pandemic. While most studies agree that the
amount of passenger-kilometers flown globally will increase substantially in the coming decades (for context: several
studies suggest that the amount of passenger-kilometers will roughly triple by 2050, relative to 2020 [42]), in this simple
analysis we assume that the distribution of flights in the 2030+ timeframe will be comparable to the one shown in Fig.
10. Looking at this distribution and focusing on regional flights below 2000 km, we can draw two important conclusions.
First, we see that these ranges are dominated by narrowbody aircraft, and not by regional aircraft. The narrowbody
aircraft emit less CO2 on a per-passenger-kilometer basis than regional aircraft [7] due to the higher seat count, and
therefore their share of passenger-kilometers is even higher than the bars suggest. And second, the figure shows that a
significant portion of the total CO2 emissions of the aviation sector are generated on ranges below 2000 km. In other
words, while the contribution of regional aircraft to overall CO2 emissions is relatively low, the contribution of the
regional market to the overall CO2 emissions is relatively large. And to assess the overall decarbonization potential of
electric aircraft we must not focus on the regional aircraft, but on the regional market.

120
CO2 from passenger transport [Mt]

Regionals Narrowbodies Widebodies

100

80

60

40

20

0
10 0
15 0
20 0
25 0
30 0
35 0
40 0
45 0
50 0
55 0
60 0
65 0
70 0
75 0
80 0
85 0
90 0
95 0
10 00
10 00
11 00
11 00
12 00
12 00
13 00
13 00
14 00
>1 00
00
50
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

0
5
0
5
0
5
0
5
0
40
Range R [km]

Fig. 10 Distribution of CO2 emissions due to commercial passenger transport versus flight distance and aircraft
type, in 2019. Adapted from Ref. [7].

To be more specific, Table 4 shows the share of CO2 emissions and number of flights corresponding to ranges below
500 km, 800 km, 1000 km, and 2000 km. The number of flights is relevant not only for the market size, but also for the
LTO emissions (landing and take-off cycle emissions, such as NOx and soot) and noise produced in the vicinity of
airports. If a battery-electric aircraft replaces a share of these flights, the associated LTO emissions are reduced to
zero and the noise per flight may be reduced (see Sec. IV.E); however, note that a single narrowbody would have to
be replaced by roughly two electric 90-seaters to transport the same volume of passengers, and therefore the required
number of flights for short distances would increase noticeably.
The first three columns of Table 4 correspond to the ranges achievable with the electric aircraft for the three
battery-technology scenarios collected in Table 3, while the fourth corresponds to the range a 1000 km electric
aircraft could cover with one stopover. Although an effective use of stopovers would require a shift from a traditional
point-to-point or hub-and-spoke network of operations to a more “web” based network where a fleet of aircraft operates
from not only primary but also secondary airports (with 2000 m runway), the development of a zero-emission 90-seater
would provide an opportunity to take advantage of currently under-served regional airports. This may allow passengers
to cover 2000 km with a single stopover without noticeably increasing the overall door-to-door travel time, if the airports
are closer to their original departure location and final destination. This line of thought has some parallels with the
concept of Regional Air Mobility (RAM) [43], though further research is encouraged to investigate advantages and
drawbacks of such operations in terms of economics and passenger experience.
In any case, Table 4 shows that, independently of the range scenario, the decarbonization potential of large electric
aircraft is substantial. Note that the values do not consider non-CO2 effects such as contrails, which are more difficult to
quantify and are higher for longer-distance flights. The values of Table 4 also reflect a non-linear increase in emissions
with range: the share of CO2 emissions for 1000 km (19%) is more than double the share corresponding to 500 km (6%),

16
and the share corresponding to 2000 km (43%) is in turn more than double the share corresponding to 1000 km. This
occurs because the 1000 km − 1500 km interval is the largest contributor (see Fig. 10), and reconfirms the importance
of maximizing electric aircraft range to maximize their market potential. For example, a “1st gen” 90-seater with 800
km range has the potential to replace 41% of all flights and thereby reduce the CO2 of the entire aviation industry by
14%, while the use of a “2nd gen” aircraft with 1000 km range and stopovers would increase these shares to 82% of all
flights and 43% of all CO2 emissions. These numbers evidence that, to truly maximize the decarbonization potential of
this new propulsion technology, the aircraft should be designed in a way that facilitates a network of operations with
stopovers.

Table 4 Share of CO2 emissions and number of flights for four different market sizes, as of 2019 and as a
fraction of the total amount of commercial passenger transport flights.

Flights up to Flights up to Flights up to Flights up to

500 km 800 km 1000 km 2000 km
Share of total aviation CO2 emissions 6% 14% 19% 43%
Share of total number of scheduled flights 22% 41% 52% 82%

E. Aircraft Noise
The societal acceptance of the aircraft depends not only on its impact on the climate, but also on the noise produced.
For this study, no detailed noise assessment was performed, with the exception of simple one-equation estimations
for propeller noise based on Ruijgrok [44] when assessing the impact of the number of propellers. Therefore, this
final section is limited to a qualitative discussion. However, it is important to manage expectations in this regard.
After hearing electric cars and small electrically-driven planes, one may expect a large electric passenger aircraft to be
quiet compared to a fuel-burning counterpart. But for such aircraft, the propellers are the dominant source of noise,
particularly in take-off. And since the electric aircraft weighs two to three times more than current fuel-based turboprop
aircraft, it produces substantially more thrust and features more propellers. Likewise, if we compare it to a narrowbody,
the MTOM is the same, but the payload is cut in half. This implies that double the amount of movements are required to
displace the same volume of passengers. This, combined with the limited excess power after take-off for a steep climb,
means that the noise footprint will be a serious challenge for these aircraft if no action is taken.
Fortunately, the use of distributed electric propulsion also opens the space to novel low-noise designs. The
aforementioned low disk loading of the propellers reduces the loading noise, and may allow for lower rotational speeds
while maintaining an acceptable propeller efficiency. This can significantly reduce thickness noise, which is driven
by the tip Mach number. The fact that electric motors can operate at different rotational speeds while maintaining an
acceptable efficiency also provides more design freedom in this regard, particularly to reduce noise during take-off and
landing—though a gearbox may be more beneficial than a direct-drive configuration if the propeller rotational speeds
are too low. During cruise, cabin noise can be reduced by synchrophasing, similarly to conventional turboprop aircraft
[45, 46]. And in landing, the combination of a high approach speed, relatively low wing loading, and propeller effects
may allow for reductions in airframe noise by using, for example, slotless flaps. Additional research on these topics is
required to ensure low-noise operation of large electric aircraft (Challenge 10).

V. Conclusions & Outlook: The Hot Potatoes

In this second paper of the two-part paper series, we present the conceptual design of a battery-electric 90-seater using
Class-II handbook design methods. The design-space exploration converges on a tube-and-wing design with distributed
propellers and a gas-turbine-based reserve energy system. The resulting design features a maximum take-off mass of 76
tons, a vehicular energy consumption of 167 Wh/pax-km, and achieves a useful range of 500 km, 800 km, or 1000 km,
for cell energy densities of 300 Wh/kg, 450 Wh/kg, and 550 Wh/kg, respectively. This corresponds to pack-level energy
densities of 240 Wh/kg, 360 Wh/kg, and 440 Wh/kg, respectively. On these ranges, the aircraft presents approximately
70% to 80% lower grid energy consumption than eSAF-based alternatives, and 40% to 75% less than hydrogen-based
alternatives, respectively. The climate impact in terms of CO2 -equivalent emissions generated in flight, in the production
of energy, and in the production of batteries, is substantially lower than for hydrogen, kerosene, or eSAF-based aircraft,
and is comparable to land-based modes of transport. These results imply that battery-electric aircraft can, in fact, play
an important role in the decarbonization of the aviation sector: replacing flights up to 800 km can reduce the CO2

17
emissions of the aviation sector by up to 14%, while further developments in battery technology and operations would
allow electric aircraft to tackle the market up to 2000 km range, which is responsible for 43% of all CO2 emissions. To
maximize this decarbonization potential, a shift in the design objectives (compete with narrowbodies vs. enable new
forms of mobility) and operations (web-based regional transport using stopovers vs. point-to-point) of electric aircraft is
required.
This perspective is contrary to the majority, if not all, of existing literature, including some previous work of the
authors. However, although this paper demonstrates feasibility from an aircraft-design perspective, sub-system-level
assumptions have been made which need further verification. Throughout this paper, several technical challenges have
been highlighted. These so-called “hot potatoes” require further research to ensure technical feasibility at sub-system
level. In the order of appearance:

Challenge 1. Wing/battery-pack integrated design, including battery replacement

Challenge 2. Wing structural sizing for aeroelastic and landing loads with batteries in wing
Challenge 3. Design and certification of the reserve energy system
Challenge 4. High-voltage component and system design
Challenge 5. Thermal management system sizing for powertrain components and batteries
Challenge 6. Mass and energy consumption of non-propulsive systems
Challenge 7. Sizing of tail surfaces and control surfaces with distributed propulsion
Challenge 8. Propeller/wing aerodynamic performance in cruise and high-lift conditions
Challenge 9. Battery cell development for large aircraft applications
Challenge 10. Low-noise propeller design for distributed-propulsion aircraft

Future work will focus on these ten research topics. Based on that, a second design iteration can be performed to
confirm if, or with what technologies, the aircraft meets the results set out in this study. Readers are also encouraged to
confirm or disprove the findings of this work by performing their own Class-II (or higher) design of similar aircraft.
All things considered—technical feasibility, required technology and infrastructure developments, and climate
impact—we believe that battery-powered aircraft for commercial air transport deserve at least the same amount of
attention as hydrogen, hybrid-electric, and SAF-based propulsion systems. The opportunity that electric aircraft offer
to substantially reduce the negative impact of aviation on global warming warrants substantial investments into the
maturation of the technological building blocks that enable electric aviation. By developing these technologies, we can
make not just evolutionary steps, but also revolutionary steps in the path towards decarbonization of the aviation sector.

Acknowledgments
This research is funded by Elysian Aircraft Company BV. R. de Vries and R. E. Wolleswinkel declare that they have
competing interests, and that they have performed this research in an objective way to the best of their abilities. The
authors would like to thank Joaquin Exalto for contributions to the figures presented in this paper and Pieter-Jan
Proesmans for input and feedback on the calculations of emissions.

Appendices

A. Aircraft Sizing Methods and Assumptions

This appendix provides an overview of the main assumptions and methods used in the conceptual aircraft design study,
for transparency and to indicate where the main uncertainties lie in the component modeling.

A. Performance Modeling
The flight-performance constraints of the wing loading/power loading diagram, as well as the mission analysis, are
modeled with physics-based point-performance equations as described in Ref. [18]. Two exceptions in this regard are the
take-off performance constraint, which is modeled semi-empirically following Ref. [47], Ch. 9, and the landing-distance
constraint, which is modeled semi-empirically following Ref. [48], Part I, Ch. 3. Propeller interaction effects were
assessed in exploratory analyses using the method of Ref. [18] and were found to have a small positive effect, but are
neglected in the conceptual design study presented here for simplicity. The power lapse of the gas turbine with altitude
is modeled according to Ref. [49].

18
B. Mass Estimation
The semi-empirical component mass build-up of Torenbeek [9], Ch. 8 is used. The following paragraphs list additional
assumptions and exceptions that are made, or additional data that is required to estimate the mass of components that
are not found in conventional fuel-based aircraft. For the powertrain components, the following approach is taken:

• The mass of the gas turbine of the reserve energy system is estimated using turboshaft mass data of Roskam, Part
V, Ch. 6 [48], from which the gearbox mass calculated according to Ref. [50] is subtracted. Accessories and fuel
system mass are added following Ref. [9]. A gas-turbine efficiency of 0.4 is assumed.
• Propeller mass is estimated based on Ref. [51].
• The assumed efficiencies and specific powers of the components of the electric drivetrain are listed in the table
below. These data are based on a combination of public literature and discussions with manufacturers. Note that
the battery efficiency is only used as a proxy to estimate the heat load of the battery, since the cell energy density
assumed in Sec. III.C already corresponds to the useful energy that can be extracted from the cell at the required
C-rate. A direct-drive electric motor, without gearbox, is chosen in this study. Electric motor specific power
refers to maximum continuous power, and it is assumed that 20% over-powering is possible for 3 minutes (during
take-off).
• For the thermal management system (TMS), a mass of 1 kg is assumed per kW of heat load rejected, based on
values reported in Refs. [25, 52]. The spread in values encountered in literature show that this value is highly
uncertain and that a more detailed definition of the TMS is required for accurate mass estimates.

Table 5 Efficiencies and specific powers assumed for the components of the electric drivetrain

Component Efficiency Specific power [kW/kg]

Electric motor + inverter 0.95 5
Generator + rectifier 0.93 11
PMADa 0.96 20
Battery 0.95 -
aPower management and distribution system, including
cables, circuit breakers, fuses, buses, and converters.

In addition, the following assumptions are made for other elements of the aircraft’s operating empty mass:
• A 10% and 20% reduction in fuselage mass and tail mass are assumed due to the use of composites, respectively,
based on Ref. [47]. For the wing, no weight reduction is assumed in this regard. The choice of materials is to be
investigated in more detail.
• When estimating the wing weight according to Ref. [9], Ch. 8, the max zero fuel mass is taken as the total aircraft
mass excluding battery mass. This is based on the assumption that the battery contributes to a wing bending relief,
similarly to fuel. A more detailed structural analysis of the wing is required to verify this approach. Additionally,
the 10% reduction in wing mass due to bending relief of the engines as suggested by Ref. [9] is applied, and a 4%
wing mass penalty is assumed for the folding wingtips.
• A 50% reduction in hydraulic and pneumatic systems mass is assumed as a consequence of the electrification of
aircraft subsystems. For the same reason, a 25% reduction in the mass of environmental control system (ECS) and
ice protection system (IPS) are assumed, which may share functions with the TMS of the aircraft. Conversely, a
100% increase in (non-propulsive) electrical system mass is assumed. Further investigation into the electrification
of these aircraft subsystems is required to verify these assumptions.

C. Tail Sizing
The horizontal and vertical tail surfaces are estimated based on volume coefficients. For the horizontal tail, a
volume coefficient of (𝑆HT 𝑙 HT )/(𝑐 mac 𝑆) = 0.9 is selected. For the vertical tail, a modified volume coefficient of
(𝑆VT 𝑙 VT )/(𝑏𝑙fus 𝐷 fus ) = 0.07 is selected. The vertical tail coefficient based on the fuselage projected area (𝑙fus 𝐷 fus )
instead of the wing reference area (𝑆) is used because initial sizing studies showed that the original formulation based
on wing reference area led to unrealistically large vertical tails as a result of the large wing relative to the fuselage in the

19
case of electric aircraft. Due to the destabilizing effect of the fuselage, its lateral projected area is also considered a
more representative parameter for vertical tail sizing than the wing reference area, and the modified volume coefficient
was found to present a slightly stronger correlation with the reference data provided in Ref. [48], Part 2, Ch. 8, than the
original volume coefficient based on wing reference area. Moreover, for the vertical tail a 30% reduction in area is
assumed due to the reduced yawing moments in case of motor failure and possible use of differential thrust for yaw
control in distributed-propulsion configurations, based on Ref. [53]. For the horizontal tail, a 10% reduction in tail area
was assumed as a result of the low center-of-gravity excursion present in the electric aircraft. A more detailed control &
stability analysis of the aircraft is required to verify these claims.

D. Energy Sizing
The energy required for the nominal mission is computed with a time-stepping mission analysis. In addition to the
power required for propulsion, the power required for non-propulsive systems such as the environmental control system
and avionics is taken to be 3 kW/passenger, based on Ref. [54]. For the thermal management system of the electrical
drivetrain, 0.1 kW of power is assumed to be required per kW of heat rejected, which is roughly halfway between the
upper and lower bounds estimated in Ref. [52]. For taxi, take-off, and landing, the energy fractions shown in Table 6 are
assumed based on Ref. [48], Part I.

Table 6 Energy required for ground phases of the mission, as a fraction of the total energy carried on board.

Mission phase Energy fraction

Taxi out 1%
Take-off 0.5%
Landing 0%
Taxi in 0.5%

E. Energy Sizing
A symmetric parabolic drag polar, 𝐶𝐷 = 𝐶𝐷0 + 𝐶 𝐿2 /(𝜋 𝐴𝑒), is used to model the drag of the aircraft. The Oswald
factor 𝑒 is computed as a function of the aspect ratio using the correlation presented in Ref. [55], Ch. 40. An increase
of +0.025 and a decrease of -0.025 is applied to the Oswald factor of the electric and reference turboprop aircraft,
respectively, to model the sensitivity of having a small or large fuselage relative to the wing span. The zero-lift drag
coefficient 𝐶𝐷0 is estimated using the component drag buildup method of Raymer [20]. For the drag buildup, smooth
paint is assumed as surface finish and a 5% increase 𝐶𝐷0 is assumed for miscellaneous (e.g. leakage & protuberance)
drag. The fraction of laminar flow and the interference factors assumed for the various aircraft components are given in
Table 7. For the extent of laminar flow on the wing, it is assumed that 10% of the wing surface is laminar downstream
of the propellers (note that a propeller does not lead to a fully turbulent boundary-layer; see e.g. Ref. [56]), and that
50% laminar flow is achieved on the folding wingtip.

Table 7 Percentages of laminar flow and interference factors assumed in the zero-lift drag buildup.

Component Percentage laminar flow Interference Factor Q

Fuselage 15% 1.0
Wing 16% 1.0
Tail 35% 1.03
Motor nacelles 10% 1.5
Landing gear fairing 10% 1.0

Furthermore, for the thermal management system, a drag penalty of 0.17 N per kW of heat rejected is assumed,
based on Ref. [52]. Again, the large spread encountered in literature highlights that a more careful assessment of the
TMS is required to produce more accurate values of the drag due to heat exchangers.

20
 Finally, the propeller efficiency is estimated by calculating the propulsive efficiency of an ideal actuator disk as a
function of the thrust coefficient, and multiplying it by a constant factor 𝑘 = 0.88 representing non-ideal losses. This
approach is similar to Ref. [57].

B. Assumptions Made in Environmental Assessment

This appendix lists the main assumptions made in Sec. IV to estimate the grid energy consumption and emissions of
various vehicles. For the comparison of energy consumption:

• The fuel-based reference turboprop is designed for the same mission and using the same sizing tool as the electric
aircraft, with the same assumptions regarding component technologies.
• For the A320neo-like narrowbody, an energy consumption of 250 Wh per passenger-kilometer is assumed for
max payload and short range, based on Refs. [7, 36] and various online sources. The spread in data encountered
indicates more reliable sources are required for an accurate estimate.
• For the regional H2 fuel-cell aircraft, an energy consumption of 300 Wh/pax-km at full payload is selected for
an 800 km mission based on Ref. [36], which reports an anergy consumption of 300 Wh/pax-km for a 700 km
mission, and 262 Wh/pax-km for a 1480 km mission.
• For the narrowbody H2 turbine aircraft, an energy consumption of 150 Wh/pax-km at full payload is selected for
an 800 km mission based on Ref. [36], which reports an anergy consumption of 138 Wh/pax-km for a 1575 km
mission, and 130 Wh/pax-km for a 4445 km mission.
• For all aircraft, an average occupancy of 85% is assumed, which is representative of narrowbody economy class
[7].
• For the car (EV), an energy consumption of 152 Wh/km is taken, based on data for a Tesla Model 3 in Ref. [58].
For the EV, an average occupancy of 1.2 passengers and a battery pack capacity of 75 kWh is assumed.
• For the train, an energy consumption of 75 Wh/pax-km is selected, corresponding to the densely-travelled,
electrically-powered train network in the Netherlands in 2019 [59].

In order to get a first estimate of the emissions of the various modes of transport, the following assumptions are made:

• The CO2 emitted in the production of hydrogen, eSAF, or electricity to charge the batteries is computed with
the grid energy consumption and a given grid emission index, in grams of CO2 emitted per kWh of grid energy
produced. The spread in grid-to-tank efficiencies shown in Fig. 6 is taken as the uncertainty band. For eSAF
aircraft, direct air capture is assumed to be required to achieve a net-zero carbon cycle.
• For conventional kerosene-based aircraft, the energy required to produce kerosene is assumed to be 0.26 kWh per
kWh of fuel energy produced, based on Ref. [60]. A generic ±25% uncertainty is assumed on this value. During
the combustion of kerosene, 3.16 kg of CO2 is emitted per kg of kerosene consumed.
• Battery manufacturing: the CO2 emitted in the production of batteries is difficult to quantify because it depends
on where the batteries are made, where the energy comes from, and where the materials come from. Here we
assume that 35 to 87.5 kg of CO2 is emitted per kWh of battery capacity produced, depending on the cleanliness
of the energy used. Note that even for a fully-renewable energy source, CO2 may be emitted in the extraction and
chemical transformation of raw materials. These values are based on Ref. [61], assuming a 50% reduction in
emissions on a per-kWh basis for a high-energy-density battery ( 350 Wh/kg) compared to a traditional battery
( 100 Wh/kg). This assumption may be conservative since it corresponds to a 75% increase in emissions on a
per-kg basis for the advanced battery. This contribution to emissions is spread over 1500 cycles with an average
range of 700 km per cycle for the electric aircraft (note that the average range is lower than the harmonic range),
and over 320,000 km distance in the case of the electric vehicle [58]. Second-life applications and emissions
created in the disposal or recycling of the battery are not accounted for. On one hand, a second-life application of
equal number of cycles would halve the emissions due to battery production on a per-passenger-kilometer basis.
On the other hand, some energy is required to dispose or disassemble the batteries at their true end of life. To be
conservative, a ±90% uncertainty is assumed for these values.
• The non-CO2 effects of aircraft on the climate, such as water vapor and contrails, are particularly difficult
to quantify. The work of Lee et al. [37] indicates that in the current aviation system non-CO2 effects have
approximately twice the impact on global warming, compared the CO2 effects. This factor 2 comes with a 95%
confidence interval of ±1.3. Based on this, we assume the following ratios of non-CO2 effects to CO2 -effects:

21
 – For a kerosene-based, A320neo-like narrowbody: a factor 2, based on Ref. [37]. An uncertainty of ±65%
(1.3÷2) is applied based on the 95% confidence interval of Ref. [37].
– For an eSAF-based, A320neo-like narrowbody: an additional 25% reduction with respect to the kerosene-
based variant is assumed. This is based on 50% reduced soot emissions and ice-crystal formation [62]
leading to an approximately 25% reduction in radiative forcing [63], which is in line with Ref. [6].
We assume the same uncertainty interval width as for the kerosene-based aircraft: while the non-CO2
contributions of eSAF aircraft are lower than kerosene aircraft and therefore one could expect the error
band to be smaller in an absolute sense, there are also additional unknowns with respect to the impact of
eSAF on non-CO2 emissions.
– For the fuel-based reference turboprop: a factor 0.5, since contrails are not generated at the lower cruise
altitude and CO2 effects dominate [64], but NOx and other emissions are still produced. We apply the same
uncertainty interval as above. The same factors are applied for Jet A1 and eSAF, since in this case the
reduction in contrails with eSAF is not applicable.
– For hydrogen gas-turbine aircraft: 40% less than the non-CO2 effects that would occur if the aircraft burned
kerosene at the same energy. This mimics the 40% reduction suggested in Ref. [6] if one considers the
combined effect of NOx , water vapor, and contrail/cirrus. To be on the conservative side, we assume a
±90% uncertainty on the non-CO2 effects.
– For hydrogen fuel-cell aircraft: same approach as for hydrogen gas-turbine aircraft but with a 70% lower
value than the kerosene equivalent instead of 40%, due to the reduced contrail/cirrus formation and zero
NOx emissions [6]. Again, to be on the conservative side, we assume a ±90% uncertainty on the non-CO2
effects.

References
[1] Friedrichs, J., Radespiel, R., Werij, H., and Vos, R., “Accelerating the path towards carbon-free aviation,” report of the CoE
“Sustainable and Energy Efficient Aviation” (SE2A) and Aeronautics Research Centre Niedersachsen (NFL), 2022.
[2] Mukhopadhaya, J., and Graver, B., “Performance Analysis of Regional Electric Aircraft,” International Council of Clean
Transportation white paper, 2022.
[3] Staack, I., Sobron, A., and Krus, P., “The potential of full electric aircraft for civil transportation: from the Breguet range
equation to operational aspects,” CEAS Aeronautical Journal, Vol. 12, 2021, pp. 803–819. doi:10.1007/s13272-021-00530-w.
[4] Webber, H., and Job, S., “Realising Zero-Carbon Emission Flight,” Aerospace Technology Institute report FZ-0-6.1, September
2021.
[5] Epstein, A. H., and O’Flarity, S. M., “Considerations for Reducing Aviation’s CO2 with Aircraft Electric Propulsion,” Journal
of Propulsion and Power, Vol. 35(3), 2019, pp. 572–582. doi:10.2514/1.B37015.
[6] Fuel Cells and Hydrogen Joint Undertaking, Clean Sky 2 Joint Undertaking, “Hydrogen-powered aviation: A fact-based study
of hydrogen technology, economics, and climate impact by 2050,” , May 2020.
[7] Graver, B., Rutherford, D., and Zheng, S., “CO2 emissions from commercial aviation: 2013, 2018, and 2019,” International
Council of Clean Transportation report, 2020.
[8] Wolleswinkel, R. E., de Vries, R., Hoogreef, M. F. M., and Vos, R., “A New Perspective on Battery-Electric Aviation, Part I:
Reassessment of Achievable Range,” AIAA Scitech 2024 Forum, Orlando, FL, USA, January 8-12 2024.
[9] Torenbeek, E., Synthesis of Subsonic Airplane Design, Delft University Press, 1982.
[10] Bonnin, V., Hoogreef, M., and de Vries, R., “Distributed Hybrid-Electric Propulsion Benefits for Span-Limited Aircraft,” AIAA
SciTech 2023 Forum, National Harbor, MD, USA, January 23-27 2023. doi:10.2514/6.2023-2098.
[11] European Aviation Safety Agency, “Easy Acces Rules for Air Operations,” [Link] 180-182, European Aviation Safety
Agency, Cologne, Germany, September 2023.
[12] de Vries, R., and Vos, R., “Aerodynamic Performance Benefits of Over-the-Wing Distributed Propulsion for Hybrid-Electric
Transport Aircraft,” Journal of Aircraft, Vol. 60(4), 2023. doi:10.2514/1.C036909.
[13] Karpuk, S., and Elham, A., “Influence of Novel Airframe Technologies on the Feasibility of Fully-Electric Regional Aviation,”
Aerospace, Vol. 8(6), 2021. doi:10.3390/aerospace8060163.
[14] Monjon, M. M. M., and Freire, C. M., “Conceptual Design and Operating Costs Evaluation of a 19-seat All-Electric Aircraft for

22
 Regional Aviation,” AIAA Propulsion and Energy Forum (virtual event), August 24-28 2020. doi:10.2514/6.2020-3591.
[15] Courtin, C., Mahseredjian, A., Dewald, A. J., Drela, M., and Hansman, J., “A Performance Comparison of eSTOL and eVTOL
Aircraft,” AIAA Aviation 2021 Forum, Virtual Event, August 2-6 2021. doi:10.2514/6.2021-3220.
[16] European Aviation Safety Agency, “Certification Specifications and Acceptable Means of Compliance for Large Aeroplanes
(CS-25),” Amendment 27, European Aviation Safety Agency, Cologne, Germany, November 2021.
[17] International Civil Aviation Organization, “Annex 16, Environmental Protection, Volume I – Aircraft Noise,” 8th edition, ICAO,
Montreal, Canada, 2017.
[18] de Vries, R., Brown, M., and Vos, R., “Preliminary Sizing Method for Hybrid-Electric Distributed-Propulsion Aircraft,” Journal
of Aircraft, Vol. 56(6), 2019, pp. 2172–2188. doi:10.2514/1.C035388.
[19] Finger, D. F., de Vries, R., Vos, R., Braun, C., and Bil, C., “Cross-Validation of Hybrid-Electric Aircraft Sizing Methods,”
Journal of Aircraft, Vol. 59(3), 2022, pp. 742–760. doi:10.2514/1.C035907.
[20] Raymer, D. P., Aircraft design: A conceptual approach, 6th Edition, AIAA Education Series, 2018.
[21] Torenbeek, E., “Development and Application of a Comprehensive, Design-senstive Weight Prediction Method for Wing
Structures of Transport Category Aircraft,” Delft University of Technology, Department of Aerospace Engineering Report
LR-693, September 1992.
[22] European Aviation Safety Agency, “Third Publication of Means of Compliance with the Special Condition VTOL,” MOC-3
SC-VTOL, Issue 2, European Aviation Safety Agency, Cologne, Germany, June 2023.
[23] Borer, N. K., Patterson, M. D., Viken, J. K., Moore, M. D., Clarke, S., Redifer, M. E., Christie, R. J., et al., “Design and
Performance of the NASA SCEPTOR Distributed Electric Propulsion Flight Demonstrator,” 16th AIAA Aviation Technology,
Integration, and Operations Conference, Washington, DC, USA, June 13-17 2016. doi:10.2514/6.2016-3920.
[24] Lee, H., and Lee, D.-J., “Rotor interactional effects on aerodynamic and noise characteristics of a small multirotor unmanned
aerial vehicle,” Physics of Fluids, Vol. 32, 2020. doi:10.1063/5.0003992.
[25] Biser, S., Filipenko, M., Boll, M., Kastner, N., Atanasov, G., Hepperle, M., Keller, D., et al., “Design Space Exploration Study
and Optimization of a Distributed Turbo-Electric Propulsion System for a Regional Passenger Aircraft,” AIAA Propulsion and
Energy Forum (virtual event), August 24-28 2020. doi:10.2514/6.2020-3592.
[26] Mitchell, G. A., “Experimental Aerodynamic Performance of Advanced 40o-Swept, 10-Blade Propeller Model at Mach 0.6 to
0.85,” NASA Technical Memorandum 88969, 1988.
[27] Keller, D., “Towards higher aerodynamic efficiency of propeller-driven aircraft with distributed propulsion,” CEAS Aeronautical
Journal, 2021. doi:10.1007/s13272-021-00535-5.
[28] Argonne National Laboratory, “The U.S. Department of Energy Vehicle Technologies Office and National Aeronautics and
Space Administration Joint Assessment of the R&D Needs for Electric Aviation,” Argonne National Laboratory white paper,
September 2021.
[29] European Comission, “Strategic Research Agenda for Batteries,” European Technology and Innovation Platform on Batteries –
Batteries Europe, December 2020.
[30] Chin, J. C., Look, K., McNichols, E., Hall, D. L., Gray, J. S., and Schnulo, S. L., “Battery Cell-to-Pack Scaling Trends
for Electric Aircraft,” 2021 AIAA/IEEE Electric Aircraft Technologies Symposium, Denver, CO, USA, August 11-13 2018.
doi:10.23919/EATS52162.2021.9704819.
[31] Viswanathan, V., Epstein, A. H., Chiang, Y.-M., Takeuchi, E., Bradley, M., Langford, J., and Winter, M., “The challenges and
opportunities of battery-powered flight,” Nature, Vol. 601, 2022, pp. 519–525. doi:10.1038/s41586-021-04139-1.
[32] McDonald, R., “Batteries Are Not Fuel,” , 2023. doi:10.31224/2803.
[33] Bills, A., Sripad, S., Fredericks, W. L., and Singh, M., “Performance Metrics Required of Next-Generation Batteries to Electrify
Commercial Aircraft,” ACS Energy Letters, Vol. 5(2), 2020, pp. 663–668. doi:10.1021/acsenergylett.9b02574.
[34] The Royal Society, “Net zero aviation fuels: resource requirements and environmental impacts policy briefing,” , February
2023.
[35] Agora Verkehrswende, Agora Energiewende and Frontier Economics, “The Future Cost of Electricity-Based Synthetic Fuels,” ,
September 2018.
[36] Debney, D., Beddoes, S., Foster, M., James, D., Kay, O., Karim, S., Stubbs, E., Thomas, D., Weider, K., and Wilson, R.,
“Zero-Carbon Emission Aircraft Concepts,” Aerospace Technology Institute Fly Zero report FZO-AIN-REP-0007, March 2022.

23
[37] Lee, D. S., Fahey, D., Skowron, A., Allen, M. R., Burkhardt, U., Chen, Q., J., D. S., et al., “The contribution of
global aviation to anthropogenic climate forcing for 2000 to 2018,” Atmospheric Environment, Vol. 244, 2021. doi:
10.1016/[Link].2020.117834.
[38] European Environment Agency, “Greenhouse gas emission intensity of electricity generation in Europe,” , 24 October 2023. URL
[Link] accessed
December 2023.
[39] European Environment Agency, “CO2-emission intensity from electricity generation,” , 30 June 2023. URL [Link]
[Link]/data-and-maps/daviz/sds/co2-emission-intensity-from-electricity-generation-6/, accessed
December 2023.
[40] Ledsom, A., “France Legally Bans Short-Haul Flights—Environmentalists Want More,” , 4 June 2023.
URL [Link]
environmentalists-want-more/, accessed December 2023.
[41] International Energy Agency, “Global Energy & CO2 status Report: The latest trends in energy and emissions in 2018,” , 2019.
URL [Link]
[42] Schafer, A., and Victor, D. G., “The future mobility of the world population,” Transportation Research Part A: Policy and
Practice, Vol. 34(3), 2000, pp. 171–205.
[43] Antcliff, K., Borer, N., Sartorius, S., Saleh, P., Rose, R., Gariel, M., Oldham, J., et al., “Regional Air Mobility: Leveraging Our
National Investments to Energize the American Travel Experience,” NASA Report, April 2021.
[44] Ruijgrok, G. J. J., Elements of Aviation Acoustics, Delft University Press, 2004.
[45] Pascioni, K. A., Rizzi, S. A., and Schiller, N. H., “Noise Reduction Potential of Phase Control for Distributed Propulsion
Vehicles,” AIAA Scitech 2019 Forum, San Diego, CA, USA, January 7-11 2019. doi:10.2514/6.2019-1069.
[46] Magliozzi, B., “Synchrophasing for cabin noise reduction of propeller-driven airplanes,” 8th Aeroacoustics Conference, Atlanta,
GA, USA, April 11-13 1983. doi:10.2514/6.1983-717.
[47] Torenbeek, E., Advanced aircraft design: conceptual design, analysis and optimization of subsonic civil airplanes, John Wiley
& Sons, 2013.
[48] Roskam, J., Airplane Design, DARcorporation, 1985.
[49] Ruijgrok, G. J. J., Elements of airplane performance, Delft University Press, 1996.
[50] Johnson, W. R., “NDARC NASA Design and Analysis of Rotorcraft,” NASA Technical Publication 2009-215402, December
2009.
[51] Filippone, A., Advanced aircraft flight performance, Cambridge University Press, 2012.
[52] Chapman, J. W., Hasseeb, H., and Schnulo, S., “Thermal Management System Design for Electrified Aircraft Propulsion
Concepts,” AIAA Propulsion and Energy 2020 Forum, Virtual Event, August 24-28 2020. doi:10.2514/6.2020-3571.
[53] Nguyen, E., Alazard, D., Pastor, P., and Döll, C., “Towards an Aircraft with Reduced Lateral Static Stability Using Electric
Differential Thrust,” 18th AIAA Aviation Technology, Integration, and Operations Conference, Atlanta, GA, USA, June 25-29
2018.
[54] Gnadt, A. R., Speth, R. L., Sabnis, J. S., and Barrett, S. R. H., “Technical and environmental assessment of all-electric 180-
passenger commercial aircraft,” Progress in Aerospace Sciences, Vol. 105, 2019, pp. 1–30. doi:10.1016/[Link].2018.11.002.
[55] Obert, E., Aerodynamic Design of Transport Aircraft, IOS Press, 2009.
[56] Howard, R. M., and Miley, S. J., “An investigation of the effects of the propeller slipstream on a laminar wing boundary layer,”
SAE Technical Paper Series, 850859, 1985.
[57] Zamboni, J., Vos, R., Emeneth, M., and Schneegans, A., “A Method for the Conceptual Design of Hybrid Electric Aircraft,”
2019 AIAA Aerospace Sciences Meeting, San Diego, CA, USA, January 7-11 2019.
[58] Tesla, “Impact Report 2019,” , 2019. URL [Link]
accessed December 2023.
[59] Nationale Spoorwegen, “Annual Report 2020: Zero-emission enterprise,” , 2020. URL [Link]
nl/annual-report-2020/our-activities-and-achievements-in-the-netherlands/performance-on-
sustainability/zeroemission-enterprise, accessed December 2023.

24
[60] Kolosz, B. W., Luo, Y., Xu, B., Maroto-Valer, M. M., and Andresen, J. M., “Life cycle environmental analysis of ‘drop in’
alternative aviation fuels: a review,” Sustainable Energy & Fuels, Vol. 4, 2020. doi:10.1039/C9SE00788A.
[61] Hall, D., and Lutsey, N., “Effects of battery manufacturing on electric vehicle life-cycle greenhose gas emissions,” International
Council on Clean Transportation Briefing, February 2018.
[62] Voigt, C., Kleine, J., Sauer, D., Moore, R. H., Bräuer, T., Le Clercq, P., Kaufmann, S., et al., “Cleaner burning aviation fuels
can reduce contrail cloudiness,” Nature Communications & Earth Environment, Vol. 2(114), 2021. doi:10.1038/s43247-021-
00174-y.
[63] Burkhardt, U., Bock, L., and Bier, A., “Mitigating the contrail cirrus climate impact by reducing aircraft soot number emissions,”
Climate and Atmospheric Science, Vol. 1(37), 2018. doi:10.1038/s41612-018-0046-4.
[64] Thijssen, R., Proesmans, P., and Vos, R., “Propeller Aircraft Design Optimization for Climate Impact Reduction,” 33rd Congress
of the International Congress of the Aeronautical Sciences, Stockholm, Sweden, September 4-9 2022.

25