Acta Astronautica 168 (2020) 104–115

Contents lists available at ScienceDirect

Acta Astronautica
journal homepage: www.elsevier.com/locate/actaastro

A techno-economic analysis of asteroid mining T

Andreas M. Hein , Robert Matheson, Dan Fries
∗

Initiative for Interstellar Studies, 27/29 South Lambeth Road, London, SW8 1SZ, United Kingdom

ARTICLE INFO ABSTRACT

Keywords: Asteroid mining has been proposed as an approach to complement Earth-based supplies of rare earth metals and
Asteroid mining to supply resources in space, such as water. However, existing studies on the economic viability of asteroid
Techno-economic analysis mining do not provide much guidance on which technological improvements would be needed for increasing its
Space economics economic viability. This paper develops a techno-economic analysis of asteroid mining with the objective of
Platinum
providing basic recommendations for future technology development and performance improvements. Both,
providing water in space as well as returning platinum to Earth are considered. Starting from first principles of
techno-economic analysis, gradually additional economic and technological factors are added to the analysis
model. Applied to mining missions involving spacecraft reuse, learning curve effect, and multiple spacecraft,
their economic viability is assessed. A sensitivity analysis with respect to throughput rate, spacecraft mass, and
resource price is performed. Furthermore, a sample asteroid mining architecture for volatiles based on small
CubeSat-class spacecraft is presented. It is concluded that key technological drivers for asteroid mining missions
are throughput rate, number of spacecraft per mission, and the rate in which successive missions are conducted.
Furthermore, for returning platinum to Earth, market reaction strongly influences its economic viability and it
seems to be economically viable only under unlikely conditions.

1. Introduction made, in particular regarding platinum group metals [30–32].

The exploitation of asteroids, and in particular Near Earth Asteroids
(NEAs) has been repeatedly proposed as a source of resources for Earth
and space [1–3]. Ross [4] distinguishes between metals and volatiles as
resources along with their use in a variety of applications such as
construction, life support systems, and propellant. In particular, vola-
tiles have received attention for in-space use, due to their relative ease
of extraction. For example, Calla et al. [5] explored the technological
and economic viability of supplying water from NEAs to cis-lunar orbit.
Regarding the supply of resources to Earth, only resources with a
high value to mass ratio are interesting, due to the additional cost of the
heat shield and other equipment for atmospheric entry, descent, and
landing on Earth. Therefore, high-value metals such as rare earth metals
and in particular the subgroup of platinum group metals have been the
subject of asteroid mining studies [6]. The supply of rare earth metals is
crucial for many “green technologies” such as fuel cells, catalyzers,
high-capacity batteries, and solar cells [7–9].
Existing research on asteroid mining has mainly looked into its
economic viability [2,6,10,11], technological feasibility [2,12–24],
cartography of asteroids [14,25,26], and legal aspects [27–29]. More
recently, environmental arguments for asteroid mining have been
In the following, we will focus on the techno-economic viability of
asteroid mining using techno-economic analysis. Techno-economic
analysis commonly brings together technical and economic aspects of a
product or service in order to assess its performance. Previously,
Andrews et al. [6] have analysed the techno-economic viability of a
specific asteroid mining architecture using Net Present Value (NPV)
analysis. They propose an economically viable asteroid mining program
for supplying volatiles and platinum group metals (PGMs) to Earth and
space. Busch [10] provides an analysis of an asteroid mining archi-
tecture using magnetic sails and high-efficiency solar cells. Sonter [2]
provides a generic NPV model for asteroid mining, linking economic
parameters to technical parameters. He concludes that asteroid mining
is economically viable, given an in-space market for volatiles of a few
thousand tons per year, competing with resources launched from Earth.
Figures of merit for comparing asteroid mining architectures are dis-
cussed in Lewis [33,34]. A key figure of merit is the Mass Payback ratio
(MPBR) [35]. Oxnevad [36] proposes NPV to go beyond MPBR by in-
cluding development cost and cost of capital. Diot and Prohom [37]
provide an overview of various strategic, financial, and supply chain
implications for asteroid mining. Gertsch and Gertsch [38] apply return
on investment (ROI) analysis to a sample asteroid mining venture for

∗
Corresponding author.
E-mail addresses: Andreas.hein@i4is.org (A.M. Hein), robmatheson95@gmail.com (R. Matheson), dan_fries@web.de (D. Fries).

https://doi.org/10.1016/j.actaastro.2019.05.009
Received 9 October 2018; Received in revised form 30 January 2019; Accepted 4 May 2019
Available online 10 May 2019
0094-5765/ © 2019 IAA. Published by Elsevier Ltd. All rights reserved.
A.M. Hein, et al.
returning platinum to Earth. They conclude that the main difference
between a terrestrial and space mining venture is the associated risk,
which is much larger for asteroid mining, rendering it difficult to fi-
nance. Craig et al. [39] provide a year-by-year cash flow analysis for a
mining mission, using the M-type NEA 1986 DA as a case study. Similar
to Gertsch and Gertsch [38], they conclude that a mining venture would
currently be too risky to commercially succeed.
It seems that the NPV model developed in Sonter [2] is currently the
most detailed model which links economic to technical parameters.
Nevertheless, the model has shortcomings:
• Only a single mining mission is considered;
• Cost of return to Earth surface is omitted;
• Development cost is not considered;
• Focused on impulsive transfers.
In this article, we will provide a techno-economic analysis of as-
teroid mining in order to identify key criteria for economic viability.
For this purpose, we specifically focus on economic viability compared
to immediate alternatives, such as terrestrial mining and direct delivery
of resources from Earth to space. We derive a generic profitability
model for analysing the economic viability of asteroid mining and apply
it to the case of water and platinum mining. In addition, we analyse
supply-demand effects that could follow from the injection of platinum
into the terrestrial market, taking price elasticity and substitution into
account.
2. Material and methods
We modify and extend the techno-economic model from Sonter [2],
first looking at the profit from a single mission, ignoring the time value
of money. Subsequently, a profit equation with spacecraft reuse is
proposed. Next, a profit equation for multiple spacecraft is developed,
including reuse. We further add a learning curve for the production cost
of spacecraft. Subsequently, the time-value of money is taken into ac-
count. The resulting profit equation addresses several of the short-
comings of Sonter's [2] model and is capable of:
• Assessing multiple mining missions as part of an entire asteroid
mining venture;
• Taking the cost of return to Earth into account;
• Including development cost;
• Taking transfer time into account via a generic mission start date
and a date for resource return.
The resulting profit equation is applied to the case of water and
platinum mining.
For the supply-demand analysis for platinum, a linear supply-de-
mand curve is developed, including price elasticity. In addition, an
equation for the total amount of platinum in the global terrestrial
market is defined, which is used for expressing the substitution of ter-
restrial platinum by platinum from space. Further, using a simplified
version of a profit equation, the profitability of a platinum mining
venture is assessed.
3. Theory and calculation
3.1. Asteroid mining breakeven analysis without mission duration and time-
value of money
We start with a simple equation for selling material from space in
space or on Earth. In this equation, profit P from a single mission is
revenue R minus cost C. We neglect all upfront costs such as develop-
ment costs and limit R to revenue from material sold, writing Rmat _ sold .
As shown in equation (1), we can further expand Rmat _ sold to the mass of
material m, multiplied by the mass-specific price cprice _ mat of that
105
Acta Astronautica 168 (2020) 104–115
material. We further break down C into production cost Cprod , trans-
portation cost from Earth into space and in space Ctransport , and cost of
operations Cops .
P = Rmat _ sold C=
mmat cprice _ mat (Cprod + Ctransport + Cops )
(1)
The expression in equation (1) is sufficiently general that it can also
be used for the case where resources are brought into space from Earth,
for example, water. In such a case, the price of the resource sold in
space needs to at least cover the incurred cost of producing the
spacecraft for transporting the resource, the transportation cost, and
cost of operations. For material mined in space and sold in space, the
revenue needs to at least cover the cost of production, transportation,
and operations of the spacecraft needed for that purpose. For the case
where material is mined in space and brought back to Earth, for ex-
ample, platinum, the price of the resource would be the market price on
Earth.
We express production and transportation cost in terms of the
spacecraft mass and the mass-specific production and launch cost cprod
and ctransport respectively. ctransport represents the mass-specific cost of
transportation service providers such as the cost for launch into low
Earth orbit (LEO).
P = mmat cprice _ mat mSC (cprod + ctransport ) Cops (2)
The mass of material mmat is expressed in terms of the spacecraft dry
mass mSC , the throughput rate f in kg processed per second per kg of
spacecraft mass (kg/s/kg), time of mining t, and the fraction of ex-
tracted resources from the total material processed Mp .
mmat = mSC f t Mp (3)
Contrary to distinguishing between the mass of the processing
equipment and other subsystems of the spacecraft in Sonter [2], we
treat the overall dry mass of the spacecraft as a reference. The reason is
that the dry mass is less susceptible to change during a mission than the
wet mass. Note that the choice of reference is arbitrary and we could
similarly choose the initial wet mass of the spacecraft as a reference. In
that case, the value for f would change with respect to the reference.
Inserting equation (3) into (2) and rearranging gives:
P = mSC [f t Mp cprice _ mat (cprod + ctransport )] Cops (4)
Note that in case the spacecraft dry and wet mass are different,
ctransport has to be adjusted by a mass ratio MR , indicating the ratio be-
tween spacecraft wet mass and dry mass [40]. In particular, for rocket
propulsion systems, the mass ratio is determined by the Tsiolkovsky
equation. Rearranging (4) yields:
cprod + ctransport
P = mSC cprice _ mat ft Mp Cops
cprice _ mat (5)
As a minimum condition, the expression in [] must be > 0. This
means that f t Mp needs to be sufficiently high, and in addition, the
expression in () as small as possible.
Equation (5) is important, as it defines criteria for minimum eco-
nomic viability. If the bracket [] is not positive, no profit is generated.
We can also see that we could, in principle, increase the mining dura-
tion t arbitrarily and always yield a profit for positive f and Mp, no
matter the associated cost. This is obviously not realistic. Realistic va-
lues for t are likely to be less than 5 years. Another observation is that
there are essentially two key technical parameters that are under the
direct influence of the asteroid mission architect. The first is the
throughput rate f , which should be as high as possible, and second,
cprod , which should be as low as possible. Note that the duration of
transportation has been neglected in the equation.
Up to this point, only ongoing costs, also called operational ex-
penditures (OpEx) were considered. We have excluded the capital costs
(or capital expenditures CapEx) such as the development cost of the
A.M. Hein, et al. Acta Astronautica 168 (2020) 104–115

mining spacecraft. Furthermore, we have only considered a single where cprod _ n is the specific production cost of the nth unit, cprod _ 1 the
mission. By a single mission, we understand a single spacecraft moving specific production cost of unit 1, n the number of units produced and a
from its initial location to the NEA, performing the mining operation, the learning factor, defined as [34, p.809]:
and transporting the refined material to the location where it is trans-
ln(1/S )
ferred to a client. a=1
ln(2) (10)
However, in a realistic scenario, the CapEx would also need to be
taken into account, which we introduce in the following. In addition, This form of the learning curve has been proposed in the literature
the case of several missions is also taken into account. The most due to its good fit to empirical data [42]. S indicates the slope of the
dominant contribution to CapEx is expected to be the development cost learning curve. With S typically 0.85 for aerospace systems [34, p.809],
of the spacecraft. We then yield for the total profit Ptot of the mining we get a = 0.77. Inserting Equation (9) into (8) yields:
venture:
T
cprod + ctransport ni
Ptot = n mSC cprice _ mat f t Mp Cops Cdev Ptot = ni mSC cpricemat f t Mp ceil
cprice _ mat i=1 si + 1

(6)

( )
a
The factor n indicates the number of mining missions under the cprod _ 1 ceil
ni
+ ctransport
si + 1
assumption that one spacecraft is used for one mission and then dis- Cops mSC cdev
carded. Cdev indicates the development cost. In the following, we will cprice _ mat
express development cost Cdev in the form of mSC cdev , where cdev is the (11)
mass-specific development cost. For the case where spacecraft are re-
For the case where multiple spacecraft per mission are used, we
used s-times, we yield:
introduce a factor p that indicates the number of spacecraft per mission,
n cprod + ctransport assuming spacecraft of the same type. In the case of multiple spacecraft
Ptot = n mSC cprice _ mat f t Mp ceil Cops mSC cdev,
s+1 cprice _ mat per mission, mSC is the total dry mass of all spacecraft participating in a
mission. The resulting equation is:
(7)
where ceil (x ) , the ceiling function, maps to the least integer greater than T
or equal to x . We introduce the ceiling function in order to make cprod Ptot = ni mSC cprice _ mat f t Mp
and ctransport disappear for missions where the spacecraft is reused. For i=1

example, if there are three missions (n = 3) and the spacecraft is reused

( )
a
once (s = 1), the value in the bracket becomes 1.5. Applying ceil () to cprod _1 p ceil
ni
+ ctransport
1.5 yields 2. Correspondingly, the floor (x ) function would yield 1. ni si + 1 mSC
ceil Cops cdev
Hence, production and transportation costs are only incurred twice si + 1 cprice _ mat p
instead of three times, compared to the case where the spacecraft is not
reused. (12)
The way reuse is treated here is simplified and does not take addi- From equation (12), other things constant, it can be seen that in-
tional costs such as refurbishing, processing, and reduced reliability creasing the number of spacecraft per mission tends to decrease pro-
into account, which would diminish the effect of reuse. duction cost due to the learning curve effect and decrease development
Note that Equation (7) does not contain the duration of the mining cost. We assume that advanced automation or autonomy keeps the cost
venture, in order to be valid independently of the mission duration to a of operations constant, even with an increased number of spacecraft per
specific NEA. mission. However, we acknowledge that the effect of advanced auto-
We introduce time in terms of years by summing up ni and si per mation would diminish if its development costs are high. In case the
year, counting the number of missions delivering resources in a year i, assumption of constant cost of operations is not valid, Cops needs to be
counted from t 0 and analogously for the number of missions with a multiplied by a function depending on p.
reused spacecraft si .
T
ni cprod + ctransport 3.2. Asteroid mining net present value analysis
Ptot = ni mSC cprice _ mat f t Mp ceil Cops mSC cdev
si + 1 cprice _ mat
i=1
In the breakeven analysis in 3.1, we have omitted the time value of
(8)
money. For a more realistic analysis, we introduce the time value of
Breakeven can be calculated by putting Ptot = 0 and solving the re- money via (1 + r )t , where r is the interest rate for a project with
sulting implicit function for n. Note that an underlying assumption of comparable risk. t is the number of years, counted as number of years
equation (8) is that cprod and ctransport are incurred in the same year in from a defined t 0 , in which the revenue stream is received or cost is
which the mined resources are sold on the market. Otherwise, these incurred. The expression is used in the following to discount profit
cost factors would need to be accounted for in a different year. As we which is received in the future, resulting in the present value of the
are assuming here that there is no time value of money, it does not profit.
make a difference with regards to Ptot . We will consider the more gen- Similar to Sonter [2], the present value of the profit from an asteroid
eral case with the time value of money in Section 3.2. mining mission can be calculated via:
Going one step further, manufacturing the same or similar types of
R C
spacecraft multiple times leads to the reduction of production cost for PPV =
the next unit produced: This is due to the effect of learning how to (1 + r )t (13)
produce more efficiently. The effect of learning on production cost is We can extend the equation to multiple missions,
commonly described by the experience curve or learning curve. We can
introduce the learning curve by a typical power law relation [34, Rj Cj
PPV = ,
p.809]. j
(1 + r )t j (14)
cprod _ n = cprod _ 1 na (9) where j indicates the j-th mission and t j denotes after how many years,

106
A.M. Hein, et al.
after t 0, the revenue cash flow occurs.
Next, the equation is further extended to missions that do incur cost
at different points in time, for example production and launch cost of
the spacecraft. We, therefore, add in tm , which is subtracted from t j . tm is
usually the mission duration. t j tm is, therefore, the point in time
where the mission was launched. For simplicity, we assume that pro-
duction and launch cost are incurred at this point. This leads to:
Rj Cj
PPV = . that is inelastic, which restricts the range of elasticities to
(1 + r )t j (1 + r )t j tm
j 0 < PED < 1 [43]. In order to stay within this range we restrict the
We insert the respective revenue and cost factors from (12) into (15)
and get for the present value of the total profit over all missions Ptot _ PV :
T
f t Mp
Ptot _ PV = ni mSC cprice _ mat
i=1 j
(1 + r )t j
( )
a
ni
ni
cprod1 p ceil
si + 1
+ ctransport
ceil
si + 1 cprice _ mat (1 + r )t j tm
From exploring these basic relationships, we can find several tech-
nical parameters we can tune for maximizing profit:
• Use multiple spacecraft per mission in order to exploit the learning
curve effect, resulting in lower production cost. Furthermore, use
multiple small spacecraft resulting in lower development cost;
• Increase the throughput rate;
• Reduce mission duration;
• Compress the sequence of missions (launch follow-up spacecraft
before previous spacecraft returns);
• Increase reuse of spacecraft, although we acknowledge that reuse
may induce reliability issues. It may also increase development and
production costs for the spacecraft to be reusable;
•
Reduce other cost factors such as cost of operations and transpor-
tation cost.
Obviously, the magnitude of these factors depends on the concrete
parameter values that are plugged in. Sample values will be given in
Section 4. However, reducing development cost via smaller spacecraft
and increasing throughput rate seem to be the most promising ap-
proaches for increasing profitability.
3.3. Supply and demand-dependent profitability
In the following, we develop the framework for analysing the
profitability of an asteroid mining venture with respect to supply and
demand. The first assumption in the supply-demand model is that there
is only one firm that is mining asteroids for platinum, rather than
multiple competing firms. It is clear that it is more likely that more than
one firm would engage in platinum asteroid mining. However, we
consider such a model to be beyond the scope of this paper.
In the model, we estimate that the demand curve for platinum is
linear with the equation from Ref. [43]:
Cprice _ mat = + Mtot
where Cprice _ mat is the price of platinum in the global market as covered
earlier, is a constant, the intercept of the curve on the price axis, is a
constant and the slope of the curve which is less than zero as demand
curves slope downward and Mtot is the total quantity of platinum sup-
plied to the world markets in one year, in kg.
We can approximate the platinum demand curve to being linear as
we are trying to capture how sensitive profitability is to the price
elasticity of platinum in the local vicinity of the global platinum de-
mand curve. Elasticity is the negative of the percentage change in price
Acta Astronautica 168 (2020) 104–115
that occurs with a percentage change in quantity. The slope of the
demand curve is directly proportional to the elasticity at any point on
the demand curve, so is the first input to the model.
To practically estimate this demand curve, five-year averages of the
price of platinum per kg are used [44] and global demand for platinum
[45] to estimate the location of a single point on the global platinum
demand curve (Cmat 1, Mtot 1) = ($40,000 per kg, 255 t). It is known that
platinum, like most rare earths has a price elasticity of demand (PED)
(15)
range of to −0.5 > > −6.0. Given we know and that all de-
mand curves go through point (Cmat 1,Mtot 1), we can derive corre-
sponding values of for each .
The equation for Mtot is
Mtot = Mearth + n Mmat (18)
where Mearth is the mass of platinum supplied into the global markets in
Cops mSC
a year from sources on Earth, from mining and recycling. n is the
(1 + r )t j
cdev
p number of platinum asteroid mining missions run in one year by our
asteroid mining company, and Mmat as stated earlier is the mass of
(16) platinum returned to Earth from one mission. However, Mearth is de-
pendent on n Mmat . We estimate this in the following manner:
Mearth = Mtot1 k n Mmat (19)
Mtot 1 = 255 t is the quantity of platinum that would be supplied by
Earth-based platinum producers to the global market if the asteroid
miner did not enter the market. k is the mass of platinum by which
Earth production decreases if asteroid mining production increases by
1 kg. For example, were k = 0.5 then every kg of platinum the asteroid
miner produced would result in Earth-based production of platinum
decreasing by 0.5 kg.
The demand curve faced by the asteroid miner is obtained by
plugging Eq. (18) and Eq. (19) into Eq. (17):
(20)
Cprice _ mat = (Mtot 1 + n Mmat (1 k ))
The production decisions of Earth-based platinum producers depend
on the quantity of platinum supplied by the asteroid miner into the
global market. This is particularly relevant, as the quantity of platinum
on the global market is relatively small and on the order of hundreds of
tons. If n Mmat increases, the total quantity of platinum in the global
market increases and so the price of platinum decreases. This de-
creasing price will cause less profitable Earth-based platinum mining
producers to cease or reduce platinum production once again increasing
the price.
Here an assumption is made in the model that Earth-based platinum
producers believe that the asteroid miner will continue to produce in
the long-term and not go bankrupt. Given the long delay between in-
vestment in platinum extraction and the eventual return on investment,
it is reasonable to believe that Earth-based platinum producers will not
change their production behaviour in response to the asteroid miner's
production level unless they believe the asteroid miner will continue to
operate in the market in the long run.
This assumption is reasonable since asteroid mining is such a large
feat it would likely only be successfully brought to fruition by a large
and profitable company with enough funds for the R&D. Also, due to
(17)
the political implications of rare earths, the venture may also be backed
by a government.
k is the second lever in the model, it determines how platinum
production of asteroid miners affects the quantity of platinum in the
global markets, therefore the price of platinum and therefore the
profitability of the asteroid mining venture. The closer Earth-based
platinum producers are operating to their profit margins the larger k
will be. A reasonable range for k is 0 < k < 1.
Eq. (21) is inputted into the below profit function for the asteroid
miner:
107
A.M. Hein, et al.
max P (n) = max n (Mmat Cprice _ mat
n n
MSC (Cprod + Ctransport ) Cops )
The asteroid miner chooses the number of missions to launch per year,
n , to achieve the maximum profit per year, P . We modelled the asteroid
miner as having borrowed the funds necessary to finance the cost of
development, Cdev . As a result, the asteroid miner pays 10% fixed rate
interest on the loan per year, 0.1 Cdev , which is generously low given
how risky an asteroid mining venture would be. The asteroid miner
does not pay off any of the principle of the loan, which is a reasonable
assumption since the asteroid miner might like to keep its profits in a
liquid form in case unexpected costs occurred such as a mission failure.
and k are the only factors extrinsic to the model, the levers of our
model. All other figures are either generated within the model, such as
and n , or are inputted constants from either the conservative costs
column of Table 2 in this paper or from platinum market data from the
World Bank and the World Platinum Investment Council.
In this model, it is assumed that the demand curve is constant and
does not shift to the left or right. However, there may be a feedback
loop between supply and demand. Agents in the economy might start
investing in capabilities that require platinum under the following
conditions. First, they judge that due to asteroid mining the supply of
platinum would be able to keep pace with growth in demand. Second,
the price remains relatively stable rather than rising wildly as would
happen if demand grew whilst supply stayed fixed. For example, petrol
car manufacturers might consider using platinum rather than palladium
in their catalytic converters. However, the timescales over which this
might happen are large enough that other factors will likely be of
greater significance, such as electric cars becoming more prominent and
taking market share from the traditional automobile industry which in
the period 2013–2017 constituted 41% of global platinum demand
[45].
One limitation of this analysis is that platinum recycling has not
been taken into account. An analysis with recycling, therefore, remains
for future work.
4. Results
4.1. Profitability of scenarios
In order to provide sample values for asteroid mining profitability,
we use Equation (12) and omit interest rate. We use both conservative
and optimistic reference scenarios for water mining and platinum
mining. Additional assumptions are then added to these reference sce-
narios to generate a set of scenarios derived from them. The con-
servative scenario uses current cost values and the optimistic scenario
uses cost values for mass production, similar to those which have been
observed for the OneWeb constellation spacecraft. We assume a pro-
pellant-less propulsion system such as solar sails or electric sails, which
have been proposed for asteroid exploration missions [46–48]. As
mentioned in Section 4.1, the use of a rocket propulsion system would
Table 1
Water mining conservative and optimistic reference scenarios.
cprod Specific production cost [$/kg]
ctransport Specific transportation cost to asteroid [$/kg]
cprice _ mat Price of water [$/kg]
Cops Cost of operations [$]
cdev Specific development cost [$/kg]
t Mining duration [s] (100 days)
Mp Ratio refined resources vs total processed material
f Throuput rate [kg/s/kg equipment]
mSC Spacecraft dry mass [kg]
Material processed per mission [kg]
Bootstrapping factor
Acta Astronautica 168 (2020) 104–115
0.1 Cdev
(21) require the introduction of a mass ratio, which would impact trans-
portation cost and bootstrapping factor, which we will introduce in the
next paragraph.
We further add different assumptions to the reference case: two-
times reuse of spacecraft (2 reuse in Fig. 1), learning curve effect
(learning in Fig. 1), and multiple spacecraft per mission (10 spacecraft in
Fig. 1). Only the number of missions is shown on the x-axis and mission
duration is not indicated. However, according to Table 1 and 2, mining
duration is fixed to 100 days and getting to and back from the asteroid
(s) is likely to take between three to eight years [21]. Mission duration
can, therefore, be assumed to be below 10 years. This is consistent with
values found in the literature such as in Gertsch and Gertsch [38] and
Andrews et al. [52]. Furthermore, if the same spacecraft is reused, the
spacecraft has to make a full cycle (outbound and inbound) before
delivering resources. For the case of expendable spacecraft, launching
spacecraft in tighter sequences could be imagined (launching sub-
sequent spacecraft before the return of precedent), compressing the
sequence of missions. For the case of using multiple spacecraft, we
assume 10 spacecraft per mission with a mass of 15 kg each.
An overview of the input values for water mining is shown in
Table 1. A value of 5% for the fraction of extracted resources versus
processed material is used, which is a rather conservative value for the
water content of C-type asteroids [2], combined with the efficiency of
extraction approaches proposed in the literature [2,49]. More opti-
mistic values of about 10% might be feasible [50]. However, as Sonter
[2] notes, extraction efficiencies are subject to large uncertainties, as
the technology has not yet been tested in space. The mining duration is
rather arbitrarily set to 100 days for each individual mission. A small
spacecraft with a mass of 150 kg is assumed as a reference, consistent
with a recent proposal of using small spacecraft for asteroid mining [5].
The throughput rate to arrive at a reasonable duration to breakeven is
2.3 10−4 kg/s/kg, meaning that for each kg of spacecraft dry mass,
0.23 g of material is processed per second. This value is consistent with
the values given in Sonter [2], where throughput rates for the mining
equipment alone on the order of 2.3 10−3 kg/s/kg are given. As we use
the total spacecraft dry mass instead of equipment mass only, we think
it is reasonable to assume that the equipment mass is roughly 10% of
the total spacecraft dry mass. This value is also consistent with space-
craft payload mass fractions given in Larson and Wertz [41]. We further
introduce a bootstrapping factor, where the total mass of water or
platinum delivered is divided by the spacecraft mass.
We assume a price of water of about 2 104 $/kg, which is an ap-
proximate value for the launch cost of 1 kg of water from Earth into cis-
lunar space. For example, the cost of launching 1 kg on an Earth escape
trajectory using a Falcon Heavy launcher is approximately 104 $ [51].
However, the cost for the spacecraft that delivers the payload to the
client needs to be included as well. Hence, we estimated that the cost
would be roughly twice as high. This value might be optimistic, as
developing the spacecraft for delivering water to the client might re-
quire significant up-front investments.
Conservative scenario [5,41] Optimistic scenario (mass production)
1,000,000 10,000
20,000 20,000
20,000 20,000
5,700,000 100,000
5,453,333 500,000
8,640,000 8,640,000
0.05 0.05
2.3 10−4 2.3 10−4
150 150
300,000 300,000
99 99
108
A.M. Hein, et al.
Table 2
Platinum mining conservative and optimistic reference scenarios.
cprod Specific production cost [$/kg]
ctransport Specific transportation cost to asteroid [$/kg]
cprice _ mat Price of platinum [$/kg]
Cops Cost of operations [$]
cdev Specific development cost [$/kg]
t Mining duration [s] (100 days)
Mp Ratio refined resources vs total processed material
f Throuput rate [kg/s/kg equipment]
mSC Spacecraft dry mass [kg]
Material processed per mission [kg]
Bootstrapping factor
The results of the conservative scenario can be seen in Fig. 1. The
profit for cases with reference, reuse, and learning all start with the
same negative profit, as the development cost for all these cases is
equal. However, using multiple, lighter spacecraft results in a lower
development cost and breakeven is reached earlier. Learning curve ef-
fects seem to have a rather minor impact on profitability, as the com-
parison of the data points for two-times reuse and two-times reuse plus
learning curve effect cases reveal.
Two sensitivity analyses are conducted, where the conservative re-
ference case is varied by reducing the throughput rate by 50% or re-
ducing the spacecraft mass by 50%. It can be seen that the reduction of
the throughput rate would render this case not profitable.
As previously mentioned, one advantage of expendable spacecraft is
that missions could be launched in much tighter sequence than in the
case of reuse. In the case of reuse, the spacecraft has to make a full cycle
of getting to and back from the asteroid before it can be refurbished and
relaunched. One can see that if we would be launching missions twice
as often as for the case of reuse, expendable spacecraft would be
competitive with the reuse and reuse plus learning curve effect cases.
A similar result is obtained for the optimistic scenario in Fig. 2,
where the reference case and those with reuse, learning curve effects,
and multiple spacecraft are very close to each other. Likewise, the 50%
smaller spacecraft and 50% throughput rate cases are very close to each
other. The reason is the lower cost of development, production, and
operations. This makes the cases mainly dependent on throughput rate,
price of resource sold, and spacecraft mass. An interesting observation
is that for drastically lower costs, compressing the launch schedule for
Fig. 1. Water asteroid mining conservative scenario.
Acta Astronautica 168 (2020) 104–115
Conservative cost [5,41] Optimistic cost (mass production)
1,000,000 10,000
20,000 20,000
70,000 70,000
5,700,000 100,000
5,453,333 500,000
8,640,000 8,640,000
10–5 10–5
0.035 0.035
1500 1500
4.5 108 4.5 108
3 3
Fig. 2. Water asteroid mining optimistic scenario.
expendable spacecraft would lead to significant gains in profit. Hence,
using expendable spacecraft would be better than reusing spacecraft for
the optimistic scenario.
For platinum mining, the situation is quite different, simply due to
the much lower ratio of platinum to asteroid material. We use the ratio
of 10−5 from Blair [53]. The throughput rate needs to be about two
orders of magnitude higher than for water mining. We use the value of
0.035 kg/s/kg. The spacecraft would need to be able to process the
equivalent of its own mass within 30 s. For the optimistic scenario, this
value could be reduced by one order of magnitude due to the one order
of magnitude lower cost values.
To include the cost of returning the material to Earth, the spacecraft
mass needs to be increased by the mass of the propulsion system (sail-
based propulsion system) and other spacecraft subsystems. In parti-
cular, this includes aerobraking and Earth re-entry systems, which are
needed to transport a mass of 4.5 tons of platinum back to Earth as-
suming a 1.5 t spacecraft. The cost could be reduced by manufacturing
a heat shield from asteroid material [54].
As no detailed studies on small spacecraft for platinum mining exist,
we use a spacecraft dry mass value 10 times as high as for water mining
(1.5 t), including equipment for producing a heat shield for aerobraking
and re-entry, using in-situ materials, where we assume that the heat
shield would weigh roughly 50% of the platinum (2.25 t).
As for the case of water, we further add different assumptions to the
reference case: two-times reuse of spacecraft, learning curve effects,
and multiple spacecraft per mission. The results for the conservative
scenario can be seen in Fig. 3 and Fig. 4. For the conservative scenario,
all cases do not reach breakeven after 10 missions. For the optimistic
109
A.M. Hein, et al. Acta Astronautica 168 (2020) 104–115

Fig. 3. Platinum mining conservative scenario. Fig. 5. Platinum mining conservative scenario with $30 k platinum price.

scenario, profitability is reached after three missions for all cases except
for the case of 50% lower throughput, for which breakthrough is
reached after 7 missions.
As fluctuations to platinum price are expected, we generate results
for a platinum price of $30,000 per kg instead of $70,000. These results
can be seen in Fig. 5 and Fig. 6. For the conservative scenario in Fig. 5,
no case reaches breakeven within 10 missions. For the optimistic case,
breakeven is reached for reuse, reuse with learning, and 10 spacecraft
after six missions. These results indicate that long-term prices of pla-
tinum would have a significant impact on asteroid platinum mining
viability.
An additional question we investigated is at which throughput rate
value one of the cases for the conservative scenario would reach
breakeven after the third mission. Fig. 7 shows the results. The case
with 10 spacecraft, 2 reuse, and learning reaches breakeven after three
missions at a throughput rate of 0.064 kg/s/kg.
To summarize, for all analysed cases, using multiple small space-
craft combined with reuse and learning curve effect outperforms its
alternatives. However, similar performance levels could be reached by
the reference case for the optimistic scenario by compressing the launch
Fig. 6. Platinum mining optimistic scenario with $30 k platinum price.

Fig. 4. Platinum mining optimistic scenario.
schedule for expendable spacecraft by e.g. twice the rate of its alter-
natives. Overall, the technical parameter with the largest impact on
economic viability is the throughput rate, followed by the number of
spacecraft used per mission, which has a significant impact on devel-
opment cost.
A sample mission architecture for mining volatiles, using multiple
small spacecraft is shown in Fig. 8. The spacecraft are expected to be 26
unit CubeSats. Two types of spacecraft are used: A laser spacecraft that
heats the regolith with a laser and a volatile capturing spacecraft that
captures and condenses the volatiles and transports them to their target
destination using propellant-less propulsion such as an electric sail or a
solar sail [46,55–60]. As shown in Fig. 8, the asteroid is first de-spun if
necessary (step 1) [61–65], then mined (step 2 and 3) and the mined
resources transported to the target destination (step 4). In this archi-
tecture, lasers are used to extract volatiles from the asteroid regolith.
The volatiles are then captured by funnels which are transparent for the
lasers and condensed with cold traps attached at their ends. Transpor-
tation is carried out via electric sails that do not consume any pro-
pellant. A mission architecture for mining platinum group metals would
look significantly different, due to the use of the Mond process [6].

110
A.M. Hein, et al.
Fig. 7. Platinum mining conservative scenario with 0.064 kg/s/kg throughput
rate.
4.2. Supply-demand dependent profitability
Table 3 shows the outputs of the supply-demand model based on its
two inputs:
– directly proportional to the price elasticity of demand for pla-
tinum and determines how the total supply of platinum sold in the
market per year affects the price of platinum. Therefore, it influences
how the quantity of platinum produced by the asteroid miner affects the
price it receives for it, and hence the profit of the asteroid miner.
k – The decrease in Earth-based platinum production per kg increase
in asteroid mining platinum production. k determines how the asteroid
miner's platinum production affects the total quantity of platinum sold
on world markets, which affects the price of platinum, which in con-
sequence affects the asteroid miner's profitability.
Note: Breakeven is defined as the number of years of operation
necessary for NPV = 0 to be true, with a discount rate of 10%.
It can be seen from Table 3 that there is only a slim range of and k
values for which asteroid mining is profitable, and an even slimmer
range in which it breaks even before 30 years of operation.
Fig. 8. Small spacecraft asteroid mining architecture for volatiles using lasers for extraction and electric sails for transportation (Image credit: Efflam Mercier).
111
Acta Astronautica 168 (2020) 104–115
Where k = 1 the asteroid miner is profitable and breaks even in a
year regardless of the value of . However, this is an unlikely corner
solution. There is a good probability that k is not greater than 0.9, it
may be in the region of 0.6 or 0.4. Moreover, there is a good probability
that elasticity is not lower than 0.25.
This model has assumed that the quantity of platinum in the global
market is the only factor affecting its price. However, platinum is a
speculative asset as well as a useful rare earth and precious metal. Its
price is subject not only to supply and demand but to speculation. For
example, the prices of precious metals are sensitive to market ex-
pectations of global, especially U.S., interest rates. The higher interest
rates are the more returns you receive from holding assets such as
bonds. This increases the opportunity cost of holding precious metals
such as platinum which do not yield any interest or returns. If the
market expects with a high probability that the Federal Reserve will
increase interest rates, then the demand from speculators for precious
metals such as platinum will fall and with it the price of platinum.
Platinum is not a self-contained market unaffected by other markets
and factors in the global economy. For example, the supply of platinum
can be affected by movements in currency values. In general platinum
miners and refiners pay their costs such as wages, equipment, and
electricity in the local currency of the country in which they are mining.
But they receive their revenue in foreign currency as their platinum is
largely sold abroad as exports. This means that if the local currency
decreases in value relative to other currencies that profit margins for
the platinum miner increase because the revenue of foreign currency
they receive for their platinum increases in local currency terms, whilst
their costs remain unchanged in local currency terms.
An example of this is occurring in South Africa, which over the
period 2013–2017 produced 53% of the global supply of platinum [45].
The global supply includes mined and recycled platinum. Despite the
platinum prices being at their lowest point in a decade, platinum pro-
ducers in South Africa are not decreasing production, in fact, their
largest producer, Anglo American Platinum Ltd., is increasing its pro-
duction [66]. This has partly been made possible due to the South
African Rand being weak and so supporting the profit margins of South
African producers.
The global demand for platinum can also be affected by the move-
ment of currency values. For example, changes in the strength of the
U.S. dollar can change the effective price Americans face for platinum
in dollar term, and so the U.S. demand for platinum. Since it constituted
A.M. Hein, et al.
Table 3
Profitability analysis of asteroid mining, taking terrestrial production decrease into account.
11% of global demand for platinum on average in the period
2013–2017 [45], changes in the relative strength of the U.S. dollar can
affect global platinum demand and therefore prices.
There are also black swans that can affect the price of platinum. An
example of this is how the Volkswagen scandal affected consumers'
opinions and government treatment of diesel cars. The scandal con-
tributed to reduced diesel car sales, meaning reduced demand from
diesel car producers for the platinum that is used in catalytic converters
[67]. Given that in the years 2013–2017 the demand for platinum as an
autocatalyst for diesel cars constituted 39% of the global demand for
platinum [45], anything which affects this element of platinum demand
will therefore strongly affect global platinum demand and therefore
platinum prices.
To conclude, the model indicates that a platinum asteroid miner
would only be profitable under a relatively small and unlikely set of
conditions for the global platinum market. This draws into question
whether a platinum asteroid miner would be able to survive the vola-
tility that can occur in the global platinum market.
5. Discussion
The results from the profitability model show that first and fore-
most, the economic viability of an asteroid mining venture depends on
the throughput rate, which depends on the mining process. This raises
interesting questions about the properties of the mining and refining
process. Furthermore, development cost is an important factor and
could be reduced via developing smaller spacecraft but using several of
them per mission. This seems to be critical in reducing the capital ex-
penditure and risk of an asteroid mining venture, which have been
identified as key to its economic viability [38,39]. Using smaller but
more spacecraft also leads to the learning curve effect, where producing
each subsequent spacecraft gets cheaper. However, our results indicate
that the learning curve effect does not make a significant difference for
achieving profitability. Reuse can also help to achieve profitability
more quickly but needs to be applied with care, as expendable space-
craft may be more advantageous as their mission sequence can be
compressed.
Regarding specific mining missions, our results have confirmed that
mining volatiles is easier due to the higher fraction of volatiles in C-type
112
Acta Astronautica 168 (2020) 104–115
asteroids. Hence, lower throughput rates are sufficient. This is a sup-
plement to the common argument that volatiles are simply easier to
mine due to their lower boiling point compared to metals.
The supply-demand analysis has shown that supplying platinum
from space is only economically viable in a very narrow set of values for
price elasticity and substitution. For realistic price elasticity values, the
mining venture would only be economically viable when the quantity
of platinum from space would substitute an equal quantity of terrestrial
platinum.
The underlying models to obtain these results come with important
limitations and assumptions. We have listed those we consider most
important in the following:
- Reuse: Reuse, on the one hand, would reduce production costs but
on the other hand, would generate additional costs for refurbishing
and processing the spacecraft for reuse. Furthermore, due to the
longer duration of operations, development cost might be higher, in
order to ensure the same reliability over its lifetime as the expend-
able spacecraft. Otherwise, reduced reliability would be the con-
sequence. These trade-offs have not been taken into account in the
current model.
- Propulsion: We have assumed that the spacecraft are operated with
sail-based propulsion systems, which do not require propellant. For
including propulsion systems such as solar-thermal propulsion [16],
the propellant mass needs to be added to the spacecraft mass. Fur-
thermore, if mined volatiles are used as propellant, the respective
propellant mass needs to be subtracted from the mass of mined
volatiles.
- Advanced automation (for operations): We have assumed that ad-
vanced automation is able to significantly reduce the cost of oper-
ating multiple spacecraft. However, we did not take into con-
sideration the potentially more expensive and technically difficult
development of such a system. The debate on on-orbit servicing with
and without human intervention might provide insights into the
tradeoffs of using advanced automation.
- Reliability: In the models presented here, spacecraft reliability and
the reliability of the mining process have not been taken into con-
sideration. Existing reliability models such as in Castet and Saleh
[68] or extrapolations thereof could help adjust the presented
A.M. Hein, et al.
models for taking reliability into account.
- Scalability of equipment and processes: Scalability in this context
means that increasing or shrinking the size of equipment is possible
and does not lead to a significant decrease in performance. We have
assumed that the performance of mining equipment for water and
platinum mining scales linearly. This means that the processes
would have the same throughput rate, independently of the size of
the equipment. In reality, performance usually does not scale line-
arly with size. We leave it to future work to characterize these
processes in terms of scalability. Nevertheless, for mining volatiles
from C-type asteroids with approaches such as optical mining [69]
or using lasers instead of concentrated sunlight, there does not seem
to be an obstacle for scaling the spacecraft down. Fries et al. [5]
have conducted an assessment of the scalability of processes for
mining volatiles. They have identified a combined drilling and mi-
crowave system for mining volatiles with small spacecraft.
Things might be quite different for mining metals from M-type as-
teroids, such as platinum group metals using the Mond process.
However, no characterization of the process in a simulated or in-
space environment has been conducted [16, p.14]. This makes it
difficult to estimate if the Mond process could be scaled down
without losses in efficiency.
For mechanically extracting material from the asteroid, various
technologies have been presented in Sonter [2] and Fries et al. [5]. It
is out of the scope of this article to assess for each technology if
scaling down might be feasible. This question is also difficult to
answer, as scaling laws for these technologies might be quite dif-
ferent.
Other areas where scalability issues are present are the potential de-
spinning of the asteroid and attachment. As far as de-spinning of the
asteroid is concerned, approaches for distributed de-spinning using
small spacecraft and lasers have been proposed and analysed in the
literature [62–64,70,71]. Regarding attachment, the miniaturiza-
tion of mechanisms has also been assessed for small spacecraft in
Fries et al. [5] and grippers are selected. We see no principal ob-
stacle for miniaturizing such a gripping system to a smaller scale.
- In-space manufacturing: In our scenarios, we have not taken into
consideration the effect of in-space production of equipment, for
example, using additive manufacturing at the asteroid, as proposed
by Dunn et al. [72]. An exception is the use of asteroid material for
heat shield production, which we have taken into consideration for
platinum mining. The effect of in-space manufacturing on asteroid
mining is primarily in reducing transportation and production cost.
Furthermore, the throughput rate can be increased, in case mining
equipment is manufactured at the asteroid. With this additional
equipment, more material can be processed. The models in Section 2
can cover in-space manufacturing. For equipment, which is pro-
duced in space, production and transportation cost for the equip-
ment are set to 0 in the model, as it is not produced on Earth and
does not need to be launched. Furthermore the throughput rate in-
creases. The reason is that the dry mass of the spacecraft decreases.
It decreases, as the mass of the equipment produced at the asteroid
is not counted in the dry mass. Due to the higher throughput rate, in-
space manufacturing would improve the economic viability of as-
teroid mining if it does not take an excessive amount of time to
manufacture the equipment. As already mentioned, another use case
for in-space manufacturing would be manufacturing a heat shield
for bringing material down to Earth, for example, platinum. This
would reduce the transportation cost for equipment into space [54].
- Interest rates: We have omitted interest rates for analysing the eco-
nomic viability of the various scenarios, in order to rather focus on
the relative differences between scenarios than the absolute profit
values. However, it would be interesting for future work to see if, for
example, different reliability values for spacecraft would lead to
different risk estimates for scenarios. Lower risk translates into
lower interest rates. For example, using multiple spacecraft
113
Acta Astronautica 168 (2020) 104–115
introduces redundancy, which might be regarded as a de-risking
strategy by investors and thereby leads to lower interest rate values.
- Risk: Another important aspect that has not been fully considered in
the paper is risk. It is known that inaccuracies in risk assessment can
jeopardize large scale projects [73–76]. Risk factors are multiple
and include financial (cost of capital, interest rates, governmental
instruments), technical (maturity, scalability, reliability), economic
(demand, supply, price, cost, market size, adaptation rates, will-
ingness to pay, etc.), political (governmental instruments such as
tax-exempts, subsidies, construction of a legal framework, tech-
nology policy, local returns), social (acceptance of the technology by
the public), legal (existence of a clear legal framework at a national
and international level; see Tronchetti [28] for a discussion), en-
vironmental (impact on Earth and space environments [32]). A
thorough enumeration of the risk factors is out of the scope of this
article but could be done using typical approaches from strategic
management such as political, economic, sociological, technology,
environmental, and legal (PESTEL) analysis [77]. Apart from these
risk factors, Flyvbjerg points at delusion and deception as important
intrinsic factors in why large infrastructure projects fail in general
[73].
Hence, the results presented in this paper might need to be revisited
once more detailed models are available. Nevertheless, the results
should provide rough estimates for the conditions of economic viability.
6. Conclusions
In this article we analysed the economic viability of asteroid mining,
focusing on supplying water in space and returning platinum to Earth
under idealized conditions for determining lower bounds and tenden-
cies. We find that from a profitability perspective, the throughput rate
and using smaller but multiple spacecraft per mission are key technical
parameters for reaching breakeven quickly. Hence, the development of
efficient mining processes and small spacecraft that are mass-produced
seem to be key to economic viability. In particular, the use of small but
multiple spacecraft might contribute to de-risking asteroid mining
ventures. De-risking is considered critical by the literature for financing
the mining venture. Reuse may contribute to profitability by reducing
production costs. However, it limits the rate at which successive mis-
sions are conducted, as spacecraft need to return and be refurbished.
Hence, another way to improve profitability is to increase the rate at
which successive missions are conducted by using expendable space-
craft. In terms of technological feasibility, platinum mining requires
very high throughput rates, which could be difficult to achieve.
Furthermore, for returning resources from space to Earth, the reaction
of the Earth-based market is critical for economic viability. Returning
platinum to Earth seems to be only economically viable under an un-
likely set of conditions. As in previous studies, mining volatiles and
supplying them to cis-lunar orbit seems to be economically viable. Also,
from a technical point of view, no development of mining and refining
processes with very high throughput rates is required. These results
have been obtained under simplifying assumptions such as the feasi-
bility of miniaturization of equipment and processes and their dis-
tribution to multiple spacecraft, ignoring additional cost factors for
reuse, and without taking in-space manufacturing into consideration.
For future work, we propose to explore more specific mining mis-
sion architectures and the role of regulatory instruments for supplying
resources from space to Earth. In addition, the estimation of possible
throughput rates of known mining and refining processes would merit
further investigation.
Acknowledgements
We would like to thank the anonymous reviewers for their detailed
comments that helped considerably improving this paper.
A.M. Hein, et al. Acta Astronautica 168 (2020) 104–115

Appendix A. Supplementary data [16] K. Zacny, P. Chu, J. Craft, M. Cohen, Asteroid mining, Proceeding (2013).
[17] K. Erickson, Optimal architecture for an asteroid mining mission: equipment details
and integration, Space 2006, 2006, p. 7504.
Supplementary data to this article can be found online at https:// [18] Z. Hasnain, C. Lamb, S. Ross, Capturing near-Earth asteroids around Earth, Acta
doi.org/10.1016/j.actaastro.2019.05.009. Astronaut. 81 (2) (2012) 523–531.
[19] C. McInnes, Near Earth object orbit modification using gravitational coupling, J.
Guid. Control Dyn. 30 (3) (2007) 870–873.
Nomenclature [20] J. Sanchez, C. McInnes, Assessment on the feasibility of future shepherding of as-
teroid resources, Acta Astronaut. 73 (2012) 49–66.
a Learning factor [21] D. Yárnoz, J. Sanchez, C. McInnes, Easily retrievable objects among the NEO po-
pulation, Celestial Mech. 116 (4) (2013) 367–388.
C Total cost [$] [22] A. Carlson, M. Date, M. Duarte, N. Erian, G. Gafka, P. Kappler, S. Patano, M. Perez,
Cj Total cost of j-th mission [$] E. Ponce, B. Radovich, Asteroid exploration and utilization: the Hawking explorer,
cdev Mass-specific development cost [$/kg] Proceedings of the 8th Summer Conference NASA/USRA Advanced Design
Program, 1992.
Cops Cost of operations [$]
[23] R.P. Mueller, L. Sibille, G.B. Sanders, C.A. Jones, Concepts of operations for asteroid
cprice _ mat Mass-specific price of material [$/kg] rendezvous missions focused on resources utilization, Earth and Space 2014, 2014,
Cprod Production cost of spacecraft [$] pp. 468–477.
cprod Mass-specific production cost of spacecraft [24] R. Gertsch, Asteroid mining, NASA. Johnson Space Center, Space Resources.
Volume 3: Materials, 1992, pp. 111–120.
cprod _ 1 Mass-specific production cost of first spacecraft [$/kg] [25] M. Elvis, How many ore-bearing asteroids? Planet. Space Sci. 91 (2014) 20–26.
Ctransport Transportation cost [$] [26] J. Sanchez, C. McInnes, Asteroid resource map for near-Earth space, J. Spacecr.
ctransport Mass-specific transportation cost [$/kg] Rocket. 48 (1) (2011) 153–165.
[27] L. Shaw, Asteroids, the new western frontier: applying principles of the general
f Throughput rate per kg of spacecraft mass [kg/s/kg] mining law of 1872 to incentive asteroid mining, J. Air L. Com. (2013).
mmat Mass of material that can be sold on the market [kg] [28] F. Tronchetti, Private property rights on asteroid resources: assessing the legality of
Mp Mass ratio of material that can be sold versus total material the ASTEROIDS Act, Space Policy (2014).
[29] R. Lee, Law and Regulation of Commercial Mining of Minerals in Outer Space,
processed (2012).
MR Mass ratio [30] A. Hennig, Policy Recommendations for Economically and Socially Valuable
mSC Spacecraft dry mass [kg] Asteroid Mineral Resource Exploitation Activities, (2016).
[31] K. MacWhorter, Sustainable mining: incentivizing asteroid mining in the name of
n Number of mining missions environmentalism, Wm. Mary Envtl. L. Pol’y Rev. 40 (2015) 645–676.
ni Number of missions delivering resources in a year i [32] A.M. Hein, M. Saidani, H. Tollu, Exploring potential environmental benefits of as-
p Number of spacecraft per mission teroid mining, 69th International Astronautical Congress (IAC), 2018.
[33] J.S. Lewis, Logistical implications of water extraction from near-Earth asteroids,
P Profit [$]
Space Manufacturing 9- The High Frontier Accession, Development and Utilization,
r Interest rate 1993, pp. 71–78.
R Revenue [$] [34] J.S. Lewis, Non-terrestrial resources of economic importance to earth, International
Rj Revenue from j-th mission [$] Astronautical Congress, 42nd, 1991.
[35] C.L. Gerlach, Profitably exploiting near-Earth object resources, Proceedings of the
Rmat _ sold Revenue from material sold on the market [$] 2005 International Space Development Conference, 2005.
s Number of times a spacecraft is reused [36] K.I. Oxnevad, An investment analysis model for space mining ventures, AACE
si Number of times a spacecraft is reused in a year i International Transactions (1991) K7-1.
[37] P. Diot, E. Prohom, Is There a Business Opportunity for the Exploration of Resources
S Learning factor parameter on Celestial Bodies? Toulouse Business School, 2018.
t Duration of mining operation [s] [38] R. Gertsch, L. Gertsch, Economic analysis tools for mineral projects in space, Space
tj Year in which revenue is generated for mission j, counting Resources Roundtable, 2005.
[39] G.A. Craig, S. Saydam, A.G. Dempster, Mining off-Earth minerals: a long-term play?
from t 0 = 0 J. South. African Inst. Min. Metall. 114 (12) (2014) 1039–1047.
tm Mission duration [40] G.P. Sutton, O. Biblarz, Rocket Propulsion Elements, John Wiley & Sons, 2016.
[41] W.J. Larson, J.R. Wertz, Space Mission Analysis and Design, third ed., Microcosm,
1999.
References
[42] T.P. Wright, Factors affecting the cost of airplanes, J. Aeronaut. Sci. 3 (4) (1936)
122–128.
[1] J.S. Lewis, Mining the Sky: Untold Riches from the Asteroids, Comets, and Planets, [43] P. Krugman, R. Wells, Economics, fourth ed., Worth Publishers, 2015.
Addison-Wesley Pub. Co., Reading, Massachusetts, 1996. [44] The World Bank, CMOHistoricDataAnnual, (2018) [Online]. Available http://www.
[2] M. Sonter, The technical and economic feasibility of mining the near-earth aster- worldbank.org/en/research/commodity-markets , Accessed date: 15 August 2018.
oids, Acta Astronaut. 41 (4-10) (1997) 637–647. [45] WPIC_PQ_tables_Q1_2018, World platinum investment Council, 2018 [Online].
[3] C. Lewicki, P. Diamandis, E. Anderson, C. Voorhees, F. Mycroft, Planetary re- Available: https://www.platinuminvestment.com/supply-and-demand/platinum-
sources—the asteroid mining company, New Sp. 1 (2) (2013) 105–108. quarterly , Accessed date: 15 August 2018.
[4] S.D. Ross, Near-Earth Asteroid Mining, (2001). [46] A. Slavinskis, P. Janhunen, P. Toivanen, K. Muinonen, A. Penttilä, M. Granvik,
[5] P. Calla, D. Fries, C. Welch, Asteroid Mining with Small Spacecraft and its Economic H. Ehrpais, “Nanospacecraft fleet for multi-asteroid touring with electric solar wind
Feasibility, (2018) arXiv:1808. sails, 2018 IEEE Aerospace Conference, 2018, pp. 1–20.
[6] D. Andrews, K. Bonner, A. Butterworth, H. Calvert, Defining a successful com- [47] B. Dachwald, W. Seboldt, Multiple near-Earth asteroid rendezvous and sample re-
mercial asteroid mining program, Acta Astronaut. 108 (2015) 106–118. turn using first generation solar sailcraft, Acta Astronaut. 57 (11) (2005) 864–875.
[7] X. Du, T.E. Graedel, Global in-use stocks of the rare earth elements: a first estimate, [48] A. Peloni, M. Ceriotti, B. Dachwald, Solar-sail trajectory design for a multiple near-
Environ. Sci. Technol. 45 (9) (2011) 4096–4101. earth-asteroid rendezvous mission, J. Guid. Control Dyn. 39 (12) (2016)
[8] T.E. Graedel, E.M. Harper, N.T. Nassar, P. Nuss, B.K. Reck, Criticality of metals and 2712–2724.
metalloids, Proc. Natl. Acad. Sci. 112 (14) (2015) 4257–4262. [49] C. Dreyer, J. Sercel, L. Gertsch, A. Lampe, Optical mining subscale testing, Earth Sp.
[9] T. Graedel, E. van der Voet, Linkages of Sustainability, MIT Press, Cambridge, USA, (2016) 493.
2010. [50] J. Ganguly, S.K. Saxena, Theoretical Predictions of Volatile Bearing Phases and
[10] M. Busch, Profitable asteroid mining, J. Br. Interplanet. Soc. (2004). Volatile Resources in Some Carbonaceous Chondrites, (1989).
[11] J.S. Kargel, Market value of asteroidal precious metals in an age of diminishing [51] SpaceX, Falcon Heavy, SpaceX Website, 2019 [Online]. Available: https://www.
terrestrial resources, Proceedings of the 1996 5th International Conference on spacex.com/falcon-heavy.
Engineering, Construction, and Operations in Space. Part 2 (of 2), Vol. 2 1996, pp. [52] D.G. Andrews, K.D. Bonner, A.W. Butterworth, H.R. Calvert, Defining a successful
821–829. commercial asteroid mining program, Acta Astronaut. 108 (2015) 106–118.
[12] J. Brophy, B. Muirhead, Near-earth Asteroid Retrieval Mission (ARM) Study, [53] B.R. Blair, The role of near-earth asteroids in long-term platinum supply, Space
(2013). Resources Roundtable II, 2000, pp. 1–15.
[13] H. Benaroya, Architecture for an asteroid-mining spacecraft, Asteroids, Springer, [54] M.D. Hogue, R.P. Mueller, P.E. Hintze, L. Sibille, D.J. Rasky, Regolith-derived heat
Berlin, Heidelberg, 2013, pp. 403–413. shield for planetary body entry and descent system with in situ fabrication, Earth
[14] V. Badescu, Asteroids: Prospective Energy and Material Resources - Google Scholar, and space 2012: Engineering, Science, Construction, and Operations in Challenging
(2013). Environments, 2012, pp. 526–536.
[15] K. Zacny, P. Chu, G. Paulsen, A. Avanesyan, Mobile In-Situ Water Extractor [55] P. Janhunen, Electric sail for spacecraft propulsion, J. Propuls. Power 20 (4) (2004)
(MISWE) for mars, moon, and asteroids in situ resource utilization, AIAA Sp. 763–764.
(2012). [56] P. Janhunen, P. Toivanen, J. Envall, K. Muinonen, A. Slavinskis, J. Gonzalez del

114
A.M. Hein, et al.
Amo, Asteroid touring nanosat fleet with single-tether E-sails, European Planetary
Science Congress, vol. 11, 2017.
[57] J.T. Grundmann, R. Boden, M. Ceriotti, B. Dachwald, E. Dumont, C.D. Grimm,
J. Riemann, Soil to sail-asteroid landers on near-term sailcraft as an evolution of the
GOSSAMER small spacecraft solar sail concept for in-situ characterization, 5th IAA
Planetary Defense Conference PDC 2017, 2017.
[58] J.T. Grundmann, W. Bauer, J. Biele, R. Boden, M. Ceriotti, F. Cordero, T.M. Ho,
Small spacecraft based multiple near-earth asteroid rendezvous and landing with
near-term solar sails and ‘Now-Term‘ technologies, 69th International Astronautical
Congress, 2018.
[59] L. McNutt, L. Johnson, D. Clardy, Near-earth Asteroid Scout, (2014).
[60] L. Johnson, J. Castillo-Rogez, J. Dervan, L. McNutt, Near Earth Asteroid (NEA)
Scout, (2017).
[61] M. Vasile, C. Maddock, Design of a formation of solar pumped lasers for asteroid
deflection, Adv. Space Res. 50 (7) (2012) 891–905.
[62] M. Vasile, A. Gibbings, I. Watson, J. Hopkins, Improved laser ablation model for
asteroid deflection, Acta Astronaut. 103 (2014) 382–394.
[63] M. Vetrisano, C. Colombo, M. Vasile, Asteroid rotation and orbit control via laser
ablation, Adv. Space Res. 57 (8) (2016) 1762–1782.
[64] N. Thiry, M. Vasile, Recent advances in laser ablation modelling for asteroid de-
flection methods, Nanophoton. Macrophoton. (2014).
[65] M. Vasile, M. Vetrisano, A. Gibbings, Light-touch 2: a Laser-Based Solution for the
Deflection, Manipulation and Exploitation of Small Asteroids, IAA Planet., 2013.
[66] F. Njini, Top Platinum Mines Keep Digging Even as Price at Decade Low,
Bloomberg, 2018 [Online]. Available: https://www.bloomberg.com/news/articles/
2018-09-14/top-platinum-mines-keep-up-supply-even-with-price-at-decade-low ,
Accessed date: 14 September 2018.
115
Acta Astronautica 168 (2020) 104–115
[67] H. Sanderson, Platinum Prices Hit by Volkswagen Scandal, Financial Times, 2018
[Online]. Available: https://www.ft.com/content/97d5cbb8-61c5-11e5-a28b-
50226830d644 , Accessed date: 23 September 2018.
[68] J.F. Castet, J.H. Saleh, Satellite and satellite subsystems reliability: statistical data
analysis and modeling, Reliab. Eng. Syst. Saf. 94 (11) (2009) 1718–1728.
[69] J.C. Sercel, C.B. Dreyer, A. Abbud-Madrid, D.T. Britt, R. Jedicke, L.S. Gertsch,
S.G. Love, A coordinated research program to develop the technology to optical
mine asteroids, Earth and Space, (2016) Orlando, FL, USA.
[70] A. Gibbings, J.M. Hopkins, D. Burns, On testing laser ablation processes for asteroid
deflection, IAA Planetary Defense Conference, Protecting Earth from Asteroids:
From Threat to Action, 2011.
[71] A. Gibbings, M. Vasile, J. Hopkins, D. Burns, I. Watson, Potential of laser-induced
ablation for future space applications, Space Policy 28 (3) (2012) 149–153.
[72] J. Dunn, M. Fagin, M. Snyder, E. Joyce, Project RAMA: Reconstructing Asteroids
into Mechanical Automata, (2017).
[73] B. Flyvbjerg, M. Garbuio, D. Lovallo, Delusion and deception in large infrastructure
projects: two models for explaining and preventing executive disaster, Calif. Manag.
Rev. 51 (2) (2009) 170–193.
[74] B. Flyvbjerg, M.S. Holm, S. Buhl, What causes cost overrun in transport infra-
structure projects? Transp. Rev. 24 (1) (2004) 3–18.
[75] B. Flyvbjerg, “Survival of the unfittest: why the worst infrastructure gets built—and
what we can do about it, Oxford Rev. Econ. Policy 25 (3) (2009) 344–367.
[76] B. Flyvbjerg, From Nobel Prize to project management: getting risks right, Proj.
Manag. J. 37 (3) (2006) 5–15.
[77] L. Shilei, W. Yong, “Target-oriented obstacle analysis by PESTEL modeling of en-
ergy efficiency retrofit for existing residential buildings in China's northern heating
region, Energy Policy 37 (6) (2009) 2098–2101.