Biomass and Bioenergy 128 (2019) 105287

Contents lists available at ScienceDirect

Biomass and Bioenergy

journal homepage: www.elsevier.com/locate/biombioe

Research paper

The development of a cost model for two supply chain network scenarios for T
decentralized pyrolysis system scenarios to produce bio-oil
Madhumita Patel, Adetoyese Olajire Oyedun, Amit Kumar*, John Doucette
Department of Mechanical Engineering, 10-263 Donadeo Innovation Centre for Engineering, University of Alberta, Edmonton, Alberta, T6G 1H9, Canada

ARTICLE INFO ABSTRACT

Keywords: Bio-oil, produced through fast pyrolysis, can act as an intermediate, which can be further upgraded to biofuels.
Supply chain model development Fast pyrolysis is done in either a centralized or a decentralized (mobile pyrolysis system [MPS]) system. In a
Decentralized system (MPS) centralized system, biomass is transported to a plant to produce bio-oil, which is upgraded in the same unit,
Techno-economic analysis while in a decentralized system, the mobile plant is moved to the forest to produce bio-oil, which is transported
Bio-oil
to an upgrading facility. The objective of this study is to develop supply chain models and conduct a techno-
Radial and truncated configuration
economic assessment of bio-oil production in these two systems using the developed models. For both systems,
the bio-oil plant capacity is considered to be 2000 dry t d−1. Because the capacity of an MPS ranges from 10 to
100 dry t d−1, several mobile pyrolysis units are required to achieve the target bio-oil produced in the base case
capacity. In this study, models were developed for each supply chain network allocation scenario/configuration
for MPS site location (the radial scenario and the truncated scenario) and four relocation time scenarios were
assumed (yearly, semi-annually, quarterly, and monthly). Then, bio-oil production costs were evaluated and
compared those with a centralized plant. For a 2000 dry t d−1 biomass plant, bio-oil production costs are 0.241$
L−1, 0.349 $ L−1, and 0.407 $ L−1 for the centralized plant, the MPS truncated scenario (100 dry t d−1 MPS
unit, yearly relocation), and the MPS radial scenario (100 dry t d−1 MPS unit, yearly relocation), respectively.

1. Introduction [14,15]. Bio-oil is considered an intermediate for transportation fuel

and can be further upgraded through hydroprocessing technology to
Greenhouse gas (GHG) emissions and global warming due to fossil produce a petro-fuel equivalent [16,17].
fuel use, the sustainability of biomass (considered carbon neutral), and In general, centralized plants are used for conventional fuels such as
the availability of local biomass are the main reasons for the growing coal and natural gas because of economies-of-scale benefits in capital
interest in bioenergy and biofuels. cost at higher capacities. Usually biomass is transported from the
In North America, there is growing interest in biofuel as a trans- nearest forest to a centralized pyrolysis plant to produce biofuel. The
portation fuel replacement [1]. Canada has 347 million hectares of economies of scale concept is also true for biomass [18]. Kumar et al.
forest, an estimated 9% of the world's forests [2]. In the last several [18] investigated the production costs of electricity from biomass for
decades, biomass has been used primarily in the pulp and paper in- different plant capacities. They found that electricity costs initially
dustry, but due to the popularity and adoption of electronic media, decreased with an increase in capacity as the economy of scale benefits
demand for pulp and paper is shrinking which was using about 57% of outweigh the increase in the biomass transportation cost and reach a
biomass energy [3–5]. minimum. With further increases in plant capacity, electricity costs
Governments around the world support the bioenergy sector and are increase. This is because of increases in biomass transportation costs,
encouraging the use of biomass resources for the production of biofuels which outweigh the benefits from the economies of scale in capital cost.
and chemicals [6–10]. Fast pyrolysis is a well-investigated commercial Another study by Kumar et al. [19] estimated the cost of power gen-
technology for the conversion of biomass to liquid fuels in a centralized eration from the combustion of Canadian mountain pine-infested wood
plant. Fast pyrolysis takes place in a pyrolyzer in the absence of oxygen for a range of plant capacities from 50 to 450 MW and observed that
at 400–600 °C and at atmospheric pressure [11–13]. The outputs from power production costs were highest at low capacity and, with in-
this process are bio-oil, gas, and biochar. Generally, bio-oil yields are creases in plant capacity, the cost was lowest at 300 MW. But with
from 50 to 75 wt% depending on the feedstock and process parameters further increases in plant capacity to 450 MW, costs remained the same.

*
Corresponding author.
E-mail address: [email protected] (A. Kumar).

https://doi.org/10.1016/j.biombioe.2019.105287
Received 17 October 2018; Received in revised form 23 April 2019; Accepted 4 July 2019
Available online 24 July 2019
0961-9534/ © 2019 Elsevier Ltd. All rights reserved.
M. Patel, et al. Biomass and Bioenergy 128 (2019) 105287

Nomenclature M Number of relocations by an MPS in a year

As Area required for an MPS in relocation time, M ha (re-
MPS Mobile pyrolysis system location time)−1
N No of MPS units T Lifetime of an MPS in years
A Capacity of centralized plant, dry t d−1 R Radius of the entire area required for the harvesting for T
A1 Bio-oil produced in a centralized plant, dry t d−1 years
B1 Capacity of an MPS unit, dry t d−1 A′ Area required for entire lifetime of all MPS units for T
Z Bio-oil used by an MPS to produce electricity, t d−1 years

Kumar et al. found these same results [18]. Other key challenges for a • The development of supply chain networks to evaluate the dis-
large-capacity centralized pyrolysis plant are feedstock availability and tributed mobile system's bio-oil production through pyrolysis pro-
the need for large capital investment. These challenges might be cess using Canadian hardwood as the feedstock for a base case plant
overcome by using mobile pyrolysis system [20]. capacity of 2000 dry t d−1,
The concept of the decentralized or MPS is to reduce the raw bio- • The development of techno-economic models for each MPS config-
mass handling and transportation costs. The characteristics and ad- uration (radial and truncated) to assess their economic feasibility,
vantages of an MPS over a centralized plant are portability, mobility, • The comparison of bio-oil production costs from both centralized
simplicity, and adaptability to regions and weather. and distributed pyrolysis systems for the base case capacity of 2000
The technical and economic aspects of a mobile pyrolysis system dry t d−1,
have been studied and are summarized in Table 1 [21–27]. Brown et al. • The evaluation of the optimum plant capacity for each configuration
[22] investigated economic feasibility of two capacities of MPS units, at the point at which they intersect with the centralized system.
10 and 50 dry t d−1 for small diameter conifer trees found in northern
New Mexico forests. A series of experiments were performed simulta- In this study, the terms scenario and configuration are used inter-
neously to enhance pyrolysis oil yield in remote locations. The authors changeably, just as decentralized and mobile pyrolysis system are.
concluded that a larger capacity MPS unit is more cost effective, and the
cost of bio-oil can be optimized by reducing labor costs and increasing 2. Method
fixed costs such as suppliers for pre-chipped or pre-sized biomass. A
2013 study focused on two distributed mobile conversion facilities (fast 2.1. Centralized pyrolysis plant
pyrolysis and torrefaction) for a biomass plant capacity of 50 dry t d−1
converting forest residue to higher energy density material that can be A centralized pyrolysis plant handles the large-scale production of
further processed in a biofuel facility and compared the levelized cost of bio-oil from whole tree biomass feedstock at a single facility (a fixed
delivery for these pathways [23]. Badger and Fransham [24] conducted facility where biomass is brought from nearby), as shown in Fig. 1 (c). A
a preliminary techno-economic analysis to estimate the capital cost, centralized facility is located at the center of a circular area and bio-
including installation cost, of a bio-oil handling system for a mobile mass is transported to the plant gate for processing. The biomass is
pyrolysis system and concluded that the capital cost was comparable to harvested, chipped, and then trucked to the plant gate. The chipped
a 50 MWe biomass handling system at the power plant. A techno-eco- biomass is then ground to a particle size of less than 2 mm and dried to
nomic assessment of 100 dry t d−1 southern pine wood chip trans- a moisture content of less than 10 wt%, which is suitable for the pyr-
portable pyrolysis system was studied by Badger et al. [27] using life olyzer [12]. Biomass pyrolysis takes place in a fluidized bed reactor in
cycle operating costs provided by Renewable Oil International (ROI) the absence of oxygen with nitrogen as the fluidizing gas to produce
LLC. To study Fischer-Tropsch fuel production in a centralized catalytic vapor and biochar. Biochar is separated through a cyclone and com-
synthesis facility, Wright et al. [21] compared two pathways, the tra- busted to generate energy. The pyrolysis vapor is quenched in a series
ditional centralized biomass gasification with Fischer-Tropsch synthesis of condensers to produce bio-oil. For the centralized plant, it was as-
and the distributed pyrolysis system to produce bio-oil followed by bio- sumed that the pyrolysis unit and upgrading unit are located at the
oil gasification and Fischer-Tropsch synthesis. For distributed pyrolysis, same place, so bio-oil transportation cost was not considered in the
the authors considered three biomass facilities: on-farm pyrolyzer, a analysis. The detailed process and operating conditions of a centralized
small cooperative pyrolyzer, and a large cooperative pyrolyzer. A few plant are further explained by Sarkar and Kumar [28].
researchers have used the geographic information system (GIS) to op-
timize the pathway and movement of the distributed pyrolysis system
2.2. Mobile pyrolysis system
to reduce biomass transportation costs based on feedstock availability
and subsequently bio-oil cost [25,26].
As a mobile pyrolysis plant is convenient, easy to assemble and use,
All of the studies cited above focus on the technical and economic
and portable, it is considered a strong alternative to a centralized plant.
feasibility of a mobile pyrolysis system to produce bio-oil. But there is
An MPS unit is loaded on a trailer that is attached to a truck or tractor.
no study available in the literature describing the development of a
The unit can be moved to a remote area where feedstock is available
supply chain network model for the allocation of MPS units over the
and the produced bio-oil can be transported to an upgrading facility. As
lifetime of the operation. The models developed here assess MPS units
the density of bio-oil (1.2–1.3 kg L−1) is almost three times greater than
of five different capacities (10, 20, 40, 50 and 100 dry t d−1) and four
the density of raw biomass, transportation costs are low [28]. An MPS
relocation times (yearly, semi-annually, quarterly, and monthly) to
unit consists of a dryer, grinder, biomass feeder, auger screw feeder,
estimate bio-oil production costs. Relocation time refers to the fre-
auger reactor, cyclone separator, condenser, biochar collector, and bio-
quency with which a decentralized unit moves in a year (i.e., semi-
oil collection chamber (see Fig. 2). Each MPS unit requires a biomass
annually relocation means that MPS unit will move 2 times in a year).
field storage area.
To create a framework for the distributed pyrolysis system, two con-
Auger pyrolysis technology was developed by ABRI-Tech [29]. The
figurations are considered and are shown in Fig. 1. The results are
reactor mechanically mixes the biomass and uses a heat transfer
compared with those from the centralized pyrolysis facility.
medium. In the pyrolysis system, two independent rotating devices
The key objectives of this study are:
rotate inside the horizontal auger reactor; the reactor does not auto-
rotate [20,22]. The heating medium is heated prior to pyrolysis, so no

2
M. Patel, et al. Biomass and Bioenergy 128 (2019) 105287

fluidizing medium is required, thereby further increasing the efficiency

Labour cost and the increased use of fixed MPS unit are the major driving factors to control bio-oil cost. 50 dry t d−1

A preliminary techno-economic analysis was conducted to estimate the capital cost, including the installation cost, of

annually, quarterly, monthly and stationary) were analyzed. The probability of success increases with reductions in
For a 40 dry t d−1 mobile pyrolysis system, three biomass feedstocks and five relocation frequencies (yearly, semi-
Forest residues can be converted to a useful form (bio-oil/torrefied wood) through a distributed conversion facility
of the condensers as they handle only the pyrolysis vapor. Biochar is

(where they undergo fast pyrolysis or torrefaction) and transported to a biofuel facility. Initial moisture content,

The production cost of bio-oil was reported to be 1.08 $ gal−1 (US dollar). Bio-oil production cost is drastically
transport distance, and the haul cost are the deciding factors for the levelized delivered cost of feedstocks [23].

a bio-oil handling system for a mobile pyrolysis system and the researchers concluded that the capital cost was
collected from two places: at the end of the reactor and from the cy-
clone char collector. In this study, it is assumed that non-condensable
gases are the heating carrier for the reactor. The pyrolysis vapor pro-
ducts are separated from the reactor due to the pressure difference and

influenced by feedstock cost, feedstock quality. and pretreatment of biomass (ash content) [27].
passed through the cyclone and a series of condensers where the bio-oil
is collected and non-condensable gases are combusted to produce heat
(and used as the heating carrier). As the MPS unit is located in the
forest, some bio-oil is used to produce electricity using a flex fuel
generator [20].
The main objective of this paper is to develop generic models to
comparable to a 50 MWe biomass handling system at a power plant [24].

investigate the technical and economic feasibility of bio-oil production

from two MPS configurations. A base case of a 2000 dry t d−1 biomass
facility was considered and Canadian hardwood were assumed as
feedstock. The capacity of an MPS unit is from 10 to 100 dry t d−1, thus
many units are required to achieve the target base case capacity. The
actual number of units required is higher than expected because some
of the produced bio-oil is used to produce electricity, which affects the
overall revenue. For each MPS, a restricted area is assigned for the
lifetime of the harvesting operation so that no two MPSs intersect. In
performs better than 10 dry t d−1 [22].

this study, five MPS capacities (10, 20, 40, 50, and 100 dry t d−1) and
four relocation times (yearly, semi-annually, quarterly, and monthly)
were considered. Two scenarios (configurations) were assumed in the
allocation of harvesting areas, truncated and radial; they are shown in
relocation frequency [25].

Fig. 1 (a) and (b), respectively. Two generic equations were developed
to estimate bio-oil production costs for the two configurations, and the
Comment/References

results were compared with those reported by Sarkar et al. [28] for a
centralized pyrolysis system.

2.2.1. Generic model development for an MPS

Fig. 1 (a) and (b) show two MPS scenarios in which all the units
operate simultaneously to achieve the target base case plant capacity.
Texas, Illinois and Nebraska USA

To derive generalized equations, each parameter is considered as a

Renewable Oil International,

Renewable Oil International,

variable that can be changed depending on the requirements. It is as-

sumed that a centralized plant of A dry t d−1capacity produces A1 dry
Northern New Mexico

tonnes of bio-oil each day. One MPS unit of B dry t d−1 capacity yields
B1 dry tonnes of bio-oil each day. Because an MPS operates in a remote
location, some of the produced bio-oil, z t d−1 of bio-oil, is used to
generate electricity. To ensure equal amounts of bio-oil from each
Location

Canada

framework (centralized and MPS) for comparison, N numbers of MPS

USA

USA

units are required and can be calculated from Eqn. (1):

A1
Small diameter conifer tree

Southern pine wood chips

N=
B1 z (1)
Corn stover and energy

The required biomass area for one MPS in each relocation time is M
Green woodchips

ha (represent by As) and the lifetime of each MPS is T years. It is im-

Forest residue

portant to note that the area required for each MPS for the lifetime of
Feedstock

sorghum

the operation is assigned at the beginning of the operation to avoid

overlapping between MPS units. In addition to M ha area, each MPS
requires ¾ acre (0.3035 ha) for its set-up [25]. Therefore, the total area
required by each MPS is M + 0.3035 ha. For example, in the yearly
relocation scenario, the MPS requires M ha/year, and the total area
required for a 10-year MPS lifetime is 10*(M + 0.3035) ha. The same
Fast pyrolysis and

area is required by each MPS.

Fast pyrolysis

Fast pyrolysis

Fast pyrolysis

Fast pyrolysis

The total area A′ required for N MPS units for the lifetime of the
torrefaction

operation for both scenarios can be estimated using Eqn. 2:

Process

A = (M + 0.3035)*N *T *m (2)
where m is the number of MPS relocations in a year (e.g., m = 1, 2, 4,
10 and 50 dry t d−1

20000 gallon bio-oil

and 12 for yearly, semi-annually, quarterly, and monthly relocations,

Research on MPSs.

respectively).
100 dry t d−1

−1
MPS capacity

50 dry t d−1

The area required for the lifetime of all MPSs is considered to be

40 dry t d

circular because for a centralized plant, it is also assumed that the area
Table 1

is circular and the plant is situated at the center of the region. The
radius of the circular area R is calculated using Eqn. (3):

3
M. Patel, et al. Biomass and Bioenergy 128 (2019) 105287

Fig. 1. Two MPS allocation configurations: (a) truncated and (b) radial and (c) centralized unit.

2 (M + 0.3035)*N *T *m 2 Rd d2 d
R= *1.27* As = * = d2
3 (3) 2
(
R
d
2 ) (4)
R is constant for the two MPS scenarios (truncated and radial), but the
calculations for the other parameters, such as biomass transportation From Eqn. (4), the distance d can be calculated using Eqn. (5):
distance to the MPS unit, bio-oil transportation distance from the MPS
d= (M + 0.3035) (5)
to the bio-oil processing facility, and relocation distance, are explained
in the subsequent section. In Eqn. (3), 1.27 is the winding factor (used 2.2.1.1.1. Bio-oil transportation distance. The bio-oil transportation
to account the winding of the roads in a remote area) and 2/3 is the distance is the distance bio-oil is transported from the MPS unit to the
average transportation displacement of the biomass collection area central bio-oil processing facility at the center O′ (see Fig. 4(a)). It is
[30]. important to estimate the number of relocations (As) based on N MPS
units and allocate the position of each MPS so that the MPS units do not
2.2.1.1. Scenario 1 – truncated. In the truncated scenario, the total interact with each other during the lifetime of each plant. Fig. 4(a)
circular area of radius R is divided into rings (see Fig. 1 (a)) and each shows the formation of rings and the division of each ring location area
MPS is placed in the truncated d segment. The upgrading unit would be (As) for each MPS unit and ‘•’ shows the position of the MPS units in the
located at the center O′ of the circular area of radius R. Therefore, the As area. The innermost circular area is for the upgrading unit to process
truncated d segments in the innermost ring do not touch the center (it the bio-oil to produce renewable diesel.
will still be a truncated area). The section framed by the letters A, B, F, The number of rings (N ) can be calculated from Eqn. (6):
C, D, and E is the location area, As, for one MPS for a particular
R
relocation time in which AD = BC = EOF = d, as shown in Fig. 3. In the N =
d (6)
segment, the MPS unit is located at the center O of the polygon
ABFCDEA. Eqn. (4) is the derived expression of As to calculate d. The The number of As in the circle with the radius R (NN′) can be cal-
detailed calculations steps are given in Appendix A (Fig A1) culated using Eqn. (7), which is the sum of the all the As in all the rings.

4
M. Patel, et al. Biomass and Bioenergy 128 (2019) 105287

Biomass
Desired particle
handling and
size biomass
storage
Dryer Grinding Screening
Hopper

Condensers Cyclone separators Screw feeder

Auger reactor
Biochar
Bio-oil collector
Biochar

Upgrading unit to
convert transportation
Truck fules

Non-condensable
gases
Gas burner
Gas
circulating
pump
Fig. 2. Schematic of an MPS unit.

different relocations. The distance covered each year can be calculated

by taking 1/10 of the Di since 10 years is considered the MPS lifetime in
this study.

DN =
N
Ni * R
(2i 1) d
=
N 2 (R (2i
2
1) d
)
i=1 2 i=1 d

* R
(2i 1) d
=
N 2 (R (2i
2 )
1) d 2

2 i=1 d (8)

2.2.1.1.2. Biomass transportation distance. Biomass transportation

distance varies with relocation time and MPS capacity. In each As,
the MPS is located at the center (O) and biomass is transported from the
surrounding area, as shown in Fig. 4(b). To estimate average biomass
transportation distance D for a relocation period, the truncated area
ABFCDEA is divided into two segments: one is ABFEA (outer arc to
MPS) and the second is EFCDE (MPS to inside arc). To estimate the
average biomass transportation distance, a small strip of area dx located
at a distance of x from the center O′is considered for both segments as
shown in Fig. 4(b). The small area dx, known as PQRST, is divided into
Fig. 3. MPS allocation in area As for scenario 1: Truncated. two equal polygons, PQR and RST. Q and S are the biomass collection
centers following harvesting, skidding, and chipping. One 10 dry tonne
The detailed calculation is given in Appendix A (Figure A2), capacity truck per trip is used to deliver the feedstock to the MPS from
the biomass collection point [31]. Generally for a centralized plant,

Ni =
N
Perimeter of dotted circle Ni
=
N 2 (R (2i
2
1) d
) 20–40 tonnes of biomass are transported per trip from the roadside to
d d the plant gate [32]. But for a decentralized plant, biomass need to be
(7)
i 1 1
transported from a remote forest location to the MPS unit. Therefore, in
where i = 1, 2, 3 … N . this study, a small capacity truck of 10 dry tonne is considered [30–32].
The bio-oil produced in the MPS needs to be transported to a bio-oil The paths followed are QR and RO for the small polygon PQR and RS
processing facility. Bio-oil transportation distances vary with the loca- and RO for the other side of the polygon RST. The average biomass
tion of the MPS. transportation distance for the truncated area ABFCDEA is given in Eqn.
Eqn. (8) is the generalized expression for the bio-oil transportation (9). The detailed derivation steps are given in Appendix A.(Figure A3)
distance for all As located in the N rings. DN is the total bio-oil trans-
portation distance for N MPS units for the entire lifetime of the plant for

5
M. Patel, et al. Biomass and Bioenergy 128 (2019) 105287

travelled by the MPS from point X to point Y is d.

2.2.1.2. Scenario 2 – radial. Fig. 1 (b) shows the radial configuration for
MPS operations. In this scenario, the number of MPSs required and the
radius of the circular area are estimated through Eqn. (1) and Eqn. (3),
respectively. BO'C is the area required by one MPS throughout its
lifetime and is shown in Fig. 5.
Fig. 5 also shows the distribution of an MPS unit's area into sub-
sectors based on relocation frequency. An MPS will relocate with time
in the assigned area. In each relocation, the MPS settles in the centroid
position (the average distance from every point in a specified area) to
minimize biomass and bio-oil transportation distance. In this config-
uration, each sector is assigned by dividing the total angle (2π) with the
number of MPS units and relocation frequencies. The angle β is sub-
tended by each sector and can be calculated by Eqn. (10).
2
=
Nm (10)
2.2.1.2.1. Bio-oil transportation distance. The bio-oil produced from
each MPS is transported to the bio-oil upgrading facility, as shown in
Fig. 6(a). In sector PO'Q, the MPS is situated at the centroid (O) before
being moved. Therefore, the bio-oil transportation distance, O′O, is
two-thirds of the radius of the entire area, as shown in Eqn. (11).
2R
OO= R =
3 (11)
2.2.1.2.2. Biomass transportation distance. Fig. 6(b) shows the
biomass transportation distance in the radial configuration for a
subsector at a particular relocation period. For the transportation
distance, the sector PO՜Q is divided into two segments: one is from
the center O′ to the centroid and the second is from the centroid to the
end of sector PO՜Q. To calculate the average biomass transportation
distance, a small strip of area dr is considered in both segments, shown
in Fig. 9. The small area dr, known as polygon WXBNELKMW, is located
at distance r from the center. The biomass in the area dr is divided into
two equal polygons, WXBNMW and MNELKM. B and E are the biomass
collection centers following harvesting, skidding, and chipping. A 10
dry tonne capacity truck is used to deliver the feedstock to the MPS
from the biomass collection points. The total average distance travelled
by truck can be estimated by summing the two integrations for the
sector PO'Q, which are estimated in Eqn. (12). The deatiled calculation
is given in Appendix A.(Figure A4)

2*84*2 R3 9*2 8
D= * +
20*Nm 27 4Nm 3 (12)
2.2.1.2.3. Relocation distance. Fig. 6(c) shows the two adjacent
segments, PO'Q and QO'S, where one MPS unit is moved following
the completion of a harvest. The relocation distance from PO'Q to QO'S
is the sum of KC and CL and is represented by Eqn. (13). We assumed
that O'CK and O'CL are a right-angled triangle.
2
MPS Relocation distance = X = KC + CL = 2r = 2R sin = 2*(2R/3)*sin
2 2Nm
(13)

Fig. 4. Bio-oil transportation distance (a), biomass transportation distance (b) 2.3. Description of economic parameters
and MPS relocation distance (c) for the truncated scenario.
This section describes the techno-economic parameters used in the
7d 3 7d 3 d MPS analysis. All the cost numbers are reported in 2017 US dollars (1
D= [13d 2 + 40dR 0 + 36R 02] + + 3R 0
10(d + 2R 0 )2 5(d + 2R0) 2 USD = 1.33 CAD). Biomass properties are given in Table 2 [18,33]. The
(9) cost model inputs used to estimate bio-oil production costs from the
MPS are shown in Tables 3 and 4. The characteristics of the centralized
2.2.1.1.3. Relocation distance. Relocation distance refers to the 2000 dry t d−1 plant were taken from Sarkar and Kumar [28].
distance travelled by the MPS between two consecutive allocated
points where it operates and is shown in Fig. 4(c). X and Y represent 2.3.1. Field cost
the MPS locations for the respective relocation times. The distance The field cost includes harvesting (cutting, skidding, and chipping),

6
M. Patel, et al. Biomass and Bioenergy 128 (2019) 105287

Fig. 5. Distribution areas into subsectors, radial scenario.

the premium paid to the land owner, road construction cost, and sil- unit, or 17 h altogether [20]. The rental truck charge is considered to be
viculture cost [34]. In this study, it was assumed that harvesting 78 $ h−1. Electricity from bio-oil, noncondensable gases, and propane
equipment (such as a feller bencher for cutting, a skidder to skid whole are used as the energy in the plant. Thermal energy for pyrolysis is
trees, and a chipper for making chips) is rented and was considered provided by the non-condensable gases and propane in the ratio of 3:1,
operating cost. A premium of 5.26 $ t−1 of feedstock is paid above the respectively. The details of the operating costs are in Table 4.
forest biomass cost for the collection and use of the forest biomass [35].
The silviculture cost is the cost to cultivate and grow new trees after 2.3.5. Biomass and bio-oil storage costs
harvesting and is 119.04 $ ha−1 [35]. Road constuction costs are the It is assumed that biomass is stored for at least 15 days. On-field
costs to build a tertiary road network that can be used during harvesting storage costs are 1.1 $ t−1 [35]. The bio-oil produced from the MPS unit
and for truck capacities of 10 dry tonnes [35,36]. The details of these is stored on site for 3 days in a 100 m3 capacity stainless steel tank and
costs are listed in Table 3 [18,23,24,31,34,35,37–39]. then transported to the bio-oil processing plant in a 30 m3 capacity
liquid tank [38]. Storage cost details are provided in Table 3.
2.3.2. Biomass and bio-oil transportation cost
Biomass and bio-oil transportation costs comprise two components, 2.3.6. Biochar transportation cost
the fixed cost for loading and unloading and the variable cost for the The biochar produced during pyrolysis is spread out in the har-
round-trip transportation of biomass and bio-oil [34]. The variable cost vested area to preserve soil nutrients for reforestation [41]. Costs to
is a function of plant capacity. In both scenarios, both biomass and bio- transport the biochar from the plant gate to the field are 4.5 $ t−1 of
oil transportation distance were shown to be a function of the radius of biochar [37].
the whole circular area. The radius depends on the yield of the biomass
as a higher yield reduces the area required for the plant. 2.3.7. Product cost
Although bio-oil is ultimately the source of revenue from the MPS
2.3.3. Capital cost unit, some of it is used to produce electricity for the portable pyrolysis
The capital cost is the investment to fabricate an MPS unit to pro- unit. The bio-oil production cost is calculated using a discounted cash
duce bio-oil in a remote location. For this analysis, five capacities are flow method.
considered: 10, 20, 40, 50, and 100 dry t d−1. Capital cost data were
drawn from publications and industrial reports [20,25,40]. From the 2.3.8. Uncertainty analysis
available data, a scale factor was derived to calculate the capital cost at In this study, a detailed robust approach has been employed to es-
various capacities. As the MPS concept is new and operation is available timate bio-oil production cost from two distributed pyrolysis facility
only on a small scale in a laboratory setting, there is very limited configurations. But, as we know, the mobile pyrolysis process is an
published research available. The MPS unit characteristics and cost emerging technology and not yet commercialized, and so a degree of
inputs are given in Table 4 [18,20,22,28]. uncertainty is associated with the field cost and operating parameters of
an MPS unit. To reduce the uncertainty and risk associated with the
2.3.4. Operational cost input parameters, a uncertainty analysis was performed using the
The operational cost includes labor, maintanance, energy, and re- Monte Carlo application [42] on the base case plant capacity of 2000
location costs. Each MPS unit operates continuously for 24 h per day. dry t d−1. The idea behind this simulation is to use the randomness of
There are 3 eight-hour shifts and each is run by 2 employees (one for the input variables to obtain an accurate result without propagating
controlling and one for fuel handling) for an MPS capacity of 10–50 dry errors. Each variable is changed in a range depending on its sensitive-
t d−1, and one adminstrative staff is assigned to each MPS. But for 100 ness on the output and large numbers of iterations are used to produce
dry t d−1 MPS unit, one extra employee is provided per shift, which accurate results.
increases the number of employees to 3 (one for controlling and two for
fuel handling). An average wage of 30 $ h−1 is assumed; this includes 3. Results and discussion
benefits [18,20,28]. The maintanance cost is assumed to be 3% of the
initial capital cost investment including all the MPSs (equivalent to a This study develops techno-economic models to investigate the
base case capacity of 2000 dry t d−1) [18]. The relocation cost is the economic and technical feasibility of replacing centralized pyrolysis
cost to move an MPS unit from one point to another by two rental plants with a decentralized system. A centralized plant benefits from
trucks. Relocation involves three activities: dismantling, moving the economies of scale, but mobile pyrolysis, because it can reduce biomass
unit to the next processing point, and reassembling. It is assumed that it transportation costs and improve accessibility in remote locations and
takes 6 h to dismantle the unit, 5 h to move it, and 6 h to reassemble the in extreme weather conditions, is gaining some interest. For the base

7
M. Patel, et al. Biomass and Bioenergy 128 (2019) 105287

3.1. Estimate of the number of MPS units required for different capacities

Fig. 7 shows the number of MPS units required with changes in

plant capacity. Both scenarios require the same number regardless of
plant capacity. A significant part of the bio-oil produced is used to
generate electricity for biomass pretreatment in the MPS. 31.25 gallons
of bio-oil produces 0.75 MM BTU of electricity per hour [20]. This re-
duces the bio-oil revenue from 57% to 50% by weight for each MPS.
Therefore the actual number of MPS units required for the various ca-
pacities is not an exact divisible relation to the base capacity of 2000
dry t d−1. When the MPS capacity increases, the required number of
MPS units decreases.

3.2. MPS capital costs

Few studies have been done on the mobile pyrolysis process. To

develop a scale factor to calculate the capital cost of an MPS unit, a
curve was plotted showing the capital cost and MPS capacity using data
from the literature [20,27,40] (see Fig. 8). The calculated scale factor is
0.61 with an R2 value of 0.9933 and indicates that capital cost increases
at a slower rate than it should with an increase in capacity. Fig. 8 shows
the capital cost for different capacities of MPS units. The capital cost is
the cost for the biomass pretreatment unit, the pyrolysis unit, and the
bio-oil condensing and bio-oil storage unit.

3.3. Biomass transportation distance in scenarios 1 and 2

In the two configurations (truncated and radial), the MPS unit is

treated like a centralized unit for the specific times. In this study, four
relocation times are considered: yearly, semi-annually, quarterly, and
monthly. For the specific relocation time, biomass is transported to the
MPS by 10 dry tonne capacity trucks. As stated earlier, the average
transportation distance for the two scenarios depends on the MPS area.
The MPS area varies with MPS capacity and relocation time. Fig. 9 (a)
and (b) shows the biomass transportation distance for the truncated and
radial scenarios, respectively. In both scenarios, biomass transportation
distance increases with increases in capacity and relocation time. This
distance is significantly higher in the radial scenario than in the trun-
cated scenario. For a yearly relocation for a 100 dry t d−1MPS capacity,
transportation distance is 129 km for the radial scenario and 30 km for
the truncated scenario. This is due to the nature of the assigned area. In
the radial scenario, MPSs are located at the centroid of each sector and
the centroid is located two-thirds the distance of the radius. But in the
truncated scenario, the MPS is located in the center of the truncated d-
segment, thus the transportation distance is considerably smaller than
in the radial scenario.
Fig. 6. Bio-oil transportation distance (a), biomass transportation distance (b),
and MPS relocation distance (c) for the radial scenario.
3.4. Bio-oil transportation distance in scenarios 1 and 2

Table 2 After the pyrolysis operation, bio-oil is transported to the upgrading

Whole tree biomass properties. facility in the center of the harvest area. The transportation distance
Parameters Value Reference/Comments varies with the MPS allocation configuration, as explained earlier.
Fig. 10 (a) and (b) shows the average bio-oil transportation distance for
Feedstock type Hard wood This study both scenarios. In both cases, with changes in relocation time, a small
Moisture content (%) 50 [18]
Heating value (MJ kg−1, HHV) 20 [18]
change in bio-oil transportation distance is observed, in the range of
Fuel density (kg m−3) 486 [33] 0.5–1 km. With variations in relocation frequency from yearly to
Biomass yield (t ha−1) 84 [18] monthly for a 100 dry t d−1 MPS capacity, the transportation distance
changes from 10.80 to 10.90 km for both scenarios, which is insignif-
icant. However, the change is somewhat significant at lower capacities.
case comparison, a centralized plant with a capacity of 2000 dry t d−1 This is because the number of units increases the radius of the circle a
producing 72% by weight of bio-oil was assumed [28]. To produce a little due to the addition of the settle-down area for each MPS. At lower
comparable amount of bio-oil per day from an MPS, several units op- capacities, the number of MPS units is comparatively higher than at
erating simultaneously are required. higher capacities, and thus the bio-oil transportation distance increases
slightly with the additional settle down areas.

8
M. Patel, et al. Biomass and Bioenergy 128 (2019) 105287

Table 3
Field cost, transportation cost, and storage cost for an MPS unit.
Parameters Value Reference/Comments

Field cost
Harvesting cost (cutting, chipping, and skidding) ($ t−1) 24.42 [34,35]
Premium paid to the land owner ($ t−1) 5.26 [18,35]
Road constuction and infrastucture cost ($ ha−1) 3.68 [18,35]
Silviculture cost ($ ha−1) 119.04 [18,35]
Transportation cost]
Biochar transportation cost ($ t−1 of biochar) 4.5 [37]
Biomass transportation cost ($ m−3) 2.35 + 0.03D D is the round-trip distance travelled by the truck from
the forest to the plant gate [18,23,24,35].
Truck capacity for the ransportation of biomass dry tonne 10 [31,39]
Bio-oil transportation cost ($ m−3) 6.824 + 0.0.060D D is the round-trip distance travelled by the liquid tank
truck from the MPS point to the bio-oil processing plant
[38].
Capacity of liquid tank truck (m3) 30 [38]
Storage system
Biomass storage type On-field storage [35]
Biomass stoarage cost ($ t−1) 1.1 [35]
No. of biomass storage days 15 Assumed
Bio-oil storage capacity (m3) 100 This study
Cost of storage tank for bio-oil ($) 137,689.02 [24,38]
Material type for bio-oil storage tank Stainless steel [38]
No. of days bio-oil can be stored 3 [38]

Table 4
MPS plant characteristics and cost input.
Parameters Value Reference/Comments

MPS characteristics
MPS capacity (dry t d−1) 10,20, 40, 50 &100 Assumed for this study
Lifetime of plant (years) 10
Running time of MPS (days) 273
Use rate (%) 90
Finalcial analysis method Discounted cash flow
Internal rate of return (%) 10
No. of shifts per day 3 [20]
No. of employees per shift 2 or 3 [22]
Adminstarative staff 1 Assumed
Employee salary including 30 [18,20,28]
benefits ($ h−1)
Scale factor for MPS capacity 0.61 Calculated. The details Fig. 7. Change in the number of MPS units with plant capacity.
from 10 to 100 dry t d−1 are included in section
3.2.
Maintanance costs 3% of initial capital [18]
investment
Relocation frequency Yearly, semi-annually Assumed.
, quarterly, and
monthly
Fast pyrolysis results (wt%)
Bio-oil 57% [20]
Biochar 27%
Noncondensable gases 16%
Energy used in MPS
Bio-oil used to produce 11% of the bio-oil This reduces the bio-oil
electricity (wt%) yield yield to 50% from 57%
by weight.
Thermal energy
Propane 25% [20]
Noncondensable gases 75%
Propane cost ($ L−1) 0.456 [43]

Fig. 8. The development of the scale factor for an MPS plant and changes in
3.5. Relocation distance capital cost with MPS capacity.

The relocation distances for scenarios 1 and 2 for different MPS

relocation distance is significantly lower at that capacity for both sce-
capacities with different relocation times are shown in Fig. 11 (a) and
narios. Therefore the relocation cost is considered an operating cost. To
(b). In both scenarios, the relocation distance is less than 2 km. The
disassemble, relocate, and reassemble every MPS unit, 17 h of labor are
relocation distance is highest for the 100 dry t d−1 capacity yearly
required. Hence with an increase in the number of MPS units and re-
relocation for both scenarios. Because many MPSs work concurrently to
location frequency comes an increase in relocation cost.
achieve the base case target of 2000 dry t d−1 in a circular area, the

9
M. Patel, et al. Biomass and Bioenergy 128 (2019) 105287

Fig. 11. MPS relocation distance with variations in MPS capacity: (a) truncated
and (b) radial configurations.

Fig. 9. Biomass transportation distance with changes in MPS capacity: (a)

truncated and (b) radial configurations. 3.6. Bio-oil production cost

This study investigates the bio-oil production costs of a 2000 dry t

d−1 biomass plant for a mobile pyrolysis system and compares them
with results for a centralized plant from work by Sarkar and Kumar
[28]. The results are summarized in Table 5. Bio-oil produced in an
MPS costs significantly more than bio-oil produced in a centralized
plant because of economies of scale.
Of the two MPS scenarios, the truncated scenario shows better re-
sults than the radial scenario. Bio-oil production costs are drastically
higher for low-capacity MPS units and decrease with increases in plant
capacity for both MPS scenarios. The reason behind this is the labor
cost. Even if the MPS capacity size changes, two employees are required
to run the unit in each shift. Therefore, for lower capacity mobile units,
far more units are required than for a higher capacity MPS unit. In a
centralized unit, however, labor cost is not a significant component
because the facility is fixed.
Table 6 shows the distribution of bio-oil production cost compo-
nents for the radial and truncated scenarios for yearly relocations of a
2000 dry t d−1 plant. Of all the cost components, capital, labor, har-
vesting, and biomass transportation cost have a significant effect on the
bio-oil production cost. With an increase in MPS capacity, labor costs
fall from 76% to 46% and 75%–39%, respectively, for the truncated and
radial scenarios. For a 10 dry t d−1MPS unit, 350 units are active at one
time. Each unit is operated by 2 employees per shift, which increases
the labor cost significantly, resulting in very high labor costs compared
to other cost components. Harvesting and biomass transportation costs
increase with the increase in MPS capacity. For the harvesting cost, as
the MPS capacity increases, it can handle more biomass than a lower
capacity MPS. Therefore it requires more biomass as feed, which in-
creases the harvesting cost. In the case of biomass transportation costs,
a 100 dry t d−1 MPS unit has a greater transportation distance than the
10 dry t d−1 unit. But the biomass transportation distance is 12% for
Fig. 10. Bio-oil transportation distance with changes in MPS capacity: (a) the truncated scenario and 23% for the radial scenario. The overall bio-
truncated and (b) radial configuration. oil production costs are lower for a 100 dry t d−1 plant with a yearly

10
M. Patel, et al. Biomass and Bioenergy 128 (2019) 105287

Table 5
Bio-oil production costs in two frameworks, centralized and MPS.
Bio-oil production cost ($ L−1) for 2000 t d−1 plant

Capacity of MPS (t d−1) Yearly relocation Semi-annually relocation Quarterly relocation Monthly relocation Centralized plant [28]

Truncated Radial Truncated Radial Truncated Radial Truncated Radial

10 1.455 1.462 1.459 1.461 1.466 1.466 1.496 1.500 0.241

20 0.828 0.841 0.830 0.838 0.834 0.841 0.854 0.864
40 0.502 0.528 0.503 0.523 0.506 0.526 0.520 0.543
50 0.432 0.467 0.436 0.461 0.438 0.464 0.451 0.480
100 0.349 0.407 0.345 0.398 0.345 0.400 0.353 0.414

relocation: 0.294 $ L−1 and 0.334 $ L−1 for the truncated and radial
scenarios.
Bio-oil production costs in the decentralized scenarios for a 100 dry
t d−1 capacity and four relocation times are shown in Fig. 12. The bio-
oil production cost decreases when the relocation time changes from
yearly to semi-annually and with further reductions in location time
from semi-annually to quarterly and monthly, bio-oil production in-
creases. The cost distributions of different components are given in
Table A1 in Appendix A. (Table A1)

3.7. Bio-oil cost variations with changing centralized plant capacity

Fig. 13 shows the variations in bio-oil production cost in the mobile Fig. 12. Bio-oil production costs for a 100 dry t d−1 MPS plant in the truncated
pyrolysis system (100 dry t d−1for yearly relocation) and the cen- and radial configurations.
tralized system. It is interesting to observe that at lower capacities (less
than 500 dry t d−1), both MPS scenarios perform better than the cen-
tralized facility. With increases in plant capacity in the centralized
plant, bio-oil production costs decrease due to economies of scale. But
for the mobile system, the results are quite different. In the radial
configuration, bio-oil production costs increase with an increase in base
case plant capacity, and in the truncated configuration, production
costs are almost constant with variations in plant capacity. The plot of
the centralized plant intersects the radial and truncated scenarios at
410 and 500 dry t d−1, respectively.
From this assessment it is reasonable to conclude that at lower ca-
pacities, it is good to adopt the MPS to produce bio-oil. In general, plant
capacity in a region depends on the availability of biomass and the
accessibility of the trees. If biomass resources are limited, it is better to
use a mobile pyrolysis plant. And of the two MPS configurations con-
sidered, the truncated scenario is economically more attractive.
Fig. 13. Bio-oil production costs for centralized and decentralized pyrolysis
plants.

Table 6
Cost distribution of bio-oil production costs for an MPS.
Cost distribution of bio-oil production costs ($ L−1) for a 2000 dry t d−1 plant

Cost parameters Yearly relocation- Scenario 1:truncated Yearly relocation- Scenario 2:radial

−1 −1 −1
10 t d 50 t d 100 t d 10 t d−1 50 t d−1 100 t d−1

Capital cost 10.19% 18.73% 17.79% 10.14% 17.39% 15.12%

Labor and overhead cost 75.31% 51.87% 46.10% 74.94% 48.17% 39.20%
Harvesting cost 3.22% 11.10% 13.81% 3.21% 10.31% 11.74%
Transportation cost of biomass to MPS 1.68% 7.18% 12.46% 2.16% 13.79% 25.57%
Maintenance cost 2.04% 3.75% 3.56% 2.03% 3.48% 3.02%
Bio-oil transportation cost 0.58% 1.99% 2.48% 0.58% 1.85% 2.11%
Propane cost 5.73% 3.94% 2.45% 5.70% 3.66% 2.09%
Relocation cost 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%
Bio-oil storage cost 0.17% 0.12% 0.07% 0.17% 0.11% 0.06%
Biomass storage cost 0.88% 0.61% 0.38% 0.88% 0.56% 0.32%
Biochar transportation cost 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%
Bio-oil production cost ($ L−1) 1.45 0.43 0.35 1.46 0.47 0.41

11
M. Patel, et al. Biomass and Bioenergy 128 (2019) 105287

3.8. Uncertainty analysis were considered and a triangular distribution was used. The input
parameters considered were capital cost, labor and overhead cost,
A Monte Carlo analysis was performed on the 100 dry t d−1MPS harvesting cost, transportation cost of biomass to the MPS, maintenance
unit for a yearly relocation for the two configurations where bio-oil cost, bio-oil transportation cost, propane cost, relocation cost, bio-oil
production cost was optimum. For these analyses, 10,000 iterations storage cost, biomass storage cost, and biochar transportation cost. The

Fig. 14. Uncertainty analysis for bio-oil production cost for a 100 dry t d−1 capacity.

12
M. Patel, et al. Biomass and Bioenergy 128 (2019) 105287

uncertainties considered for these variables were between 70 and modifications. Biomass cost, labor cost, truck capacity, energy cost, and
130%. The production cost of bio-oil was found to be 0.41 ± 0.0198 $ capital cost will change with location. In general, biomass cost is a
L−1 and 0.35 ± 0.0189 $ L−1 at a 95% confidence level, respectively, combination of biomass harvesting cost, biomass collection cost, and
for radial and truncated configurations of a 100 dry t d−1 MPS unit for biomass transportation cost. Harvesting technology varies by feedstock
yearly relocation and is shown in Fig. 14. The uncertainty associated and makes a significant contribution to bio-oil cost. In this study, a
with the radial configuration was little higher than the truncated, which generic equation was developed for the biomass transportation cost
can be attributed to the higher standard deviation associated with the based on the configuration of the selected area. To make the estimation
cost at 95% confidence. of biomass transportation distance real, a geographical information
system (GIS) can be used to calculate the exact distance between the
4. Conclusion biomass location and the MPS. Finally, labor cost and inflation are
completely driven by the location of the plant and the benefits provided
Bio-oil can be produced from both centralized and decentralized by the employer. These are the cost components that need to be
pyrolysis plants and be further processed to produce transportation changed when this technology is adopted in other jurisdictions. The
fuel. In this study, two supply chain networks were explored, radial and concept and application of technology will be same irrespective of the
square, in which a number of decentralized units work simultaneously. location.
From the cost distribution it was determined that by reducing the MPS A centralized plant performs better at higher capacities, but it might
capacity from 100 to 10 dry t d−1, labor costs would increase from not be preferred given the limited availability of large amounts of
around 40%–75%. This is because with the decrease in capacity, the biomass feedstock. Biomass availability is the driver for the selection of
number of MPS units required increases, which increases labor costs. Of the pyrolysis framework. The outcomes from this paper can also be used
the two MPS scenarios, the truncated scenario is more economical than to encourage local governments to provide decentralized units to
the radial scenario because it has a shorter biomass transportation farmers, municipalities, and forest communities to divert waste and
distance. woody biomass to valuable products and generate employment at a
Bio-oil production costs of the centralized and mobile frameworks small-scale pyrolysis unit.
were compared. For higher plant capacities, the centralized plant per-
forms better because of economies of scale. For a 2000 dry t Acknowledgments
d−1biomass plant, bio-oil production costs are 0.24 $ L−1, 0.35 $ L−1,
and 0.41 $ L−1 for a centralized plant, an MPS unit in the truncated We thank NSERC/Cenovus/Alberta Innovates Associate Industrial
scenario (100 dry t d−1 unit, yearly relocation), and an MPS unit in the Research Chair in Energy and Environmental Systems Engineering and
radial scenario (100 dry t d−1 unit, yearly relocation), respectively. For Cenovus Energy Endowed Chair in Environmental Engineering for
plant capacities below or equal to 500 dry t d−1, an MPS performs providing financial support for this project. We also thank re-
better than a centralized plant. At low capacities, the capital and ad- presentatives from Alberta Innovates (AI), Cenovus Energy, Suncor
ministrative costs in a centralized plant are significantly higher than Energy, Environment and Climate Change Canada (ECCC) and Natural
those for an MPS. Therefore, at lower base case plant capacities, an MPS Resources Canada (NRCan) for their inputs in various forms. As a part
is recommended as it is movable, provides access to remote locations, of the University of Alberta's Future Energy Systems (FES) research
and tolerates extreme weather conditions. initiative, this research was made possible in part thanks to funding
The MPS configurations developed to estimate bio-oil production from the Canada First Research Excellence Fund (CFREF). The authors
costs in this study are generic and can be applied to any real-time de- thank Astrid Blodgett for editorial assistance. Dr. Abayomi Oni is also
velopment scenario or jurisdiction through location-specific thanked for his contributions to this paper.

Appendix A

Scenario 1: Truncated

Calculation of angle :

Fig. A1. Truncated scenario.

13
M. Patel, et al. Biomass and Bioenergy 128 (2019) 105287

The perimeter of the dotted circle is given by Eqn. (14):

d
P=2 * R
2 (14)
The perimeter P creates an angle 2π at the center. So, angle (radian) is estimated by Eqn. (15):
d d d d
=2 * = 2 * = =
P 2 ( R
d
2 ) ( R
d
2 ) ( R0 +
d
2 ) (15)

where R = R ( = R0 +
d
2 ) (
as shown in Fig. A3.
d
2 )
The area of the ring created by R and R-d is Ad as represented by Eqn. (16).
Ad = R2 (R d ) 2 = 2 Rd d2 (16)
The area of segment ABFCDEADEA is As (Fig. A1) and is given in Eqn. (17)
Ad 2 Rd d2 d
As = = * = d2
2 2
(
R
d
2 ) (17)

Bio-oil transportation distance

Fig. A2. Bio-oil transportation distance in the truncated scenario.

The number of all truncated segments in the 1st ring can be calculated using Eqn. (18).

N1 =
Perimeter of dotted circle N1
=
2 (R d
2 )
d d (18)
The number of all truncated segments in the 2nd ring can be calculated by using Eqn. (19).

N2 =
Perimeter of dotted circle N2
=
2 (R 3d
2 )
d d (19)
The bio-oil transportation distance for all truncated segments located in the 1st ring can be calculated by using Eqn. (20).

D1 = N1 * R
d
=
2 (R d
2 )* R d
=
2 (R 2)
d 2

2 d 2 d (20)
The bio-oil transportation distance for the d squares located in the 2nd ring can be calculated by using Eqn. (21).

D2 = N2* R
3d
=
2 (R 3d
2 )* R 3d
=
2 (R 2)
3d 2

2 d 2 d (21)

14
M. Patel, et al. Biomass and Bioenergy 128 (2019) 105287

Biomass transportation distance

Fig. A3. Biomass transportation distance in the truncated scenario.

For the small area dx in the first segment, ABFEA (outer arc to MPS), the distance travelled by the truck to transport the biomass to the MPS unit
is calculated by Eqn. (22):
d
Biomass transportation distance by truck = d1 = (QR + RO) + (SR + RO) = 2*(QR + RO) = 2* x* + x R0
4 2

d x d
= 2* * + x R0
(R 0 +
d
2 ) 4 2
(22)
For the small area dx in the second segment, EFCDE (MPS to inside arc), the path followed by the truck to transport the biomass to the MPS unit is
calculated by Eqn. (23):

d x d
Biomass transportation distance by truck = d2 = 2* * + R0 + x
(R 0 +
d
2 ) 4 2
(23)
The area of the small strip dx is given in Eqn. (24):
d
dA = (PR + RT )dx = 2*PR*dx = 2* *x *dx = *x *dx
2 R0 + ( d
2 ) (24)
The truck capacity for biomass transportation is 20 green tonnes per trip and a yield of 84 dry tonnes of forest biomass is assumed. The area
required to collect biomass for one trip is estimated with Eqn. (25) and the number of trips required to deliver the biomass is given in Eqn. (26).
Capacity of truck 10 ha
Area required for each truck trip = =
Whole tree yield 84 trip (25)

dA 84 dA
Number of truck trips required for the area = = trip/ ha
10/84 10 (26)
The total distance travelled by truck and the number of trips for the area dA in segments 1 and 2 are given in Eqn. (27) and Eqn. (28),
respectively.

84 dA d x d 84 d x d d
dd1 = *2* * + x R0 = 2* * * + x R0 * *x dx
10
(R 0 +
d
2 ) 4 2 10
(R 0 +
d
2 ) 4 2
(R 0 +
d
2 ) (27)

84 d x d d
dd2 = 2* * * + R0 + x * *x dx
10
( R0 +
d
2 ) 4 2
( R0 +
d
2 ) (28)
To calculate the average biomass transportation distance for the whole area ABFCDEA, Eqn. (27) and Eqn. (28) are integrated from R0 + d/2 to
R0 + d and R0 to R0 + d/2, respectively. The total average distance travelled by truck can be estimated by summing the two integrations, which are
estimated in Eqn. (29)

15
M. Patel, et al. Biomass and Bioenergy 128 (2019) 105287

R0+d R0 + d2 R0+d R 0 + d2
D=
R 0 + d2
dd1 +
R0
dd2 =
R 0 + d2
84
2* 10 *
(
d
R0 +
d
)
x
(
* 4 + x R0
d
2 ) *( d
R0 +
d
)
*x dx +
R0
84
2* 10 *
(R + )
0
d
d
x
* 4 + R0 +( d
2
x ) *( d
R0 +
d
)
*xdx
2 2 2 2

7d3 (6d2 + 17dR0 + 12 R02) 7d5 7d 4 21d 4R0 42d3R02 63d3R0 84d2R 02 42d2R 0
D= + + + + + +
5(d + 2R0 )2 10(d + 2R 0)2 10(d + 2R0) 5(d + 2R0 )2 5(d + 2R 0)2 5(d + 2R0 ) 5(d + 2R0) 5

7d3 7d3 d
D= [13d 2 + 40dR0 + 36R02] + + 3R 0
10(d + 2R 0)2 5(d + 2R0) 2 (29)

Scenario 2: Radial scenario

Biomass transportation distance

For the small area dr in the first segment (from the center to the centroid), the distance travelled by the truck to transport the biomass to the MPS
unit is calculated by Eqn. (30):

Biomass transportation distance by truck = d1 = (BN + NO) + (EN + NO) = 2*(BN + NO ) = 2* r* + (r R)

4 (30)

Fig. A4. Biomass transportation distance in the radial scenario.

For the small area dr in the second segment (from the centroid to the end of the arc), the path followed by the truck to transport the biomass to
the MPS unit is calculated by Eqn. (31):

Biomass transportation distance by truck = d2 = 2* r* + (R r)

4 (31)
The area of the small strip dr is given in Eqn. (32):
(r + dr ) 2 r2 2 r dr
dA = =
Nm Nm Nm (32)
As for the truncated configuration, the number of truck trips is estimated using Eqn. (26). The total distance travelled by truck and the number of
trips for the area dA in segments 1 and 2 are given in Eqn. (33) and Eqn. (34), respectively.

84 dA 84 2 r dr 2 r 2R
dd1 = *2* r* + (r R ) = 2* * * + r
10 4 20 Nm 4Nm 3 (33)

84 dA 84 2 r dr 2 r 2R
dd2 = *2* r* + (R r) = 2* * * + r
10 4 20 Nm 4Nm 3 (34)
To calculate the average biomass transportation distance for the whole area QO′P, Eqn. (33) and Eqn. (34) are integrated from R′ to R and 0 to R’,
respectively. The total average distance travelled by truck can be estimated by summing the two integrations, which are estimated in Eqn. (35).

R R = 2R R R = 2R
3 3
84 2 r 2 r 2R 84 2 r 2 r 2R
d = D= dd1 + dd2 = 2* * * + r dr + 2* * * + r dr
10 Nm 4Nm 3 10 Nm 4Nm 3
R = 2R 0 R = 2R 0
3 3

2*84*2 R3 9*2 8
= * +
10*Nm 27 4Nm 3 (35)

16
M. Patel, et al. Biomass and Bioenergy 128 (2019) 105287

Table A1
Cost distribution

Scenario: Truncated 100 dry t d−1

Monthly Quarterly Semi-annually Yearly

Capital cost 16.7% 17.2% 17.1% 17.8%

Labour and overhead cost 46.0% 47.2% 47.1% 46.1%
Harvesting cost 13.8% 14.1% 14.1% 13.8%
Transportation cost of biomass to MPS 8.6% 10.0% 11.1% 12.5%
Maintenance cost 3.5% 3.6% 3.6% 3.6%
Bio-oil transportation cost 6.8% 3.8% 2.9% 2.5%
Propane cost 2.4% 2.5% 2.5% 2.5%
Relocation cost 0.0% 0.0% 0.0% 0.0%
Storage cost of bio-oil 0.9% 0.3% 0.1% 0.1%
Storage cost of biomass 0.4% 0.4% 0.4% 0.4%
Biochar transportation cost 0.0% 0.0% 0.0% 0.0%
Bio-oil production cost 0.9% 0.9% 0.9% 0.9%

Scenario 2: Radial 100 dry t d−1

Monthly Quarterly Semi-annually Yearly

Capital cost 13.7% 15.4% 14.7% 15.1%

Labour and overhead cost 49.9% 39.8% 40.5% 39.2%
Harvesting cost 9.9% 11.1% 11.3% 11.7%
Transportation cost of biomass to MPS 11.1% 24.0% 24.5% 25.6%
Maintenance cost 3.6% 3.1% 3.1% 3.0%
Bio-oil transportation cost 5.3% 3.2% 2.5% 2.1%
Propane cost 3.8% 2.1% 2.2% 2.1%
Relocation cost 0.0% 0.0% 0.0% 0.0%
Storage cost of bio-oil 1.4% 0.3% 0.1% 0.1%
Storage cost of biomass 0.6% 0.3% 0.3% 0.3%
Biochar transportation cost 0.0% 0.0% 0.0% 0.0%
Bio-oil production cost 0.7% 0.8% 0.8% 0.8%

References Sustain. Energy Rev. 4 (2000) 1–73.

[16] M. Patel, A. Kumar, Production of renewable diesel through the hydroprocessing of
lignocellulosic biomass-derived bio-oil: a review, Renew. Sustain. Energy Rev. 58
[1] World Energy Council, World Energy Resources Bioenergy, (2016) available at: (2016) 1293–1307.
https://www.worldenergy.org/wp-content/uploads/2017/03/WEResources_ [17] A.V. Bridgwater, Principles and practice of biomass fast pyrolysis processes for li-
Bioenergy_2016.pdf , Accessed date: 29 July 2017. quids, J. Anal. Appl. Pyrolysis 51 (1999) 3–22.
[2] Natural Resources Canada, How much Forest does Canada Have? available at: [18] A. Kumar, J.B. Cameron, P.C. Flynn, Biomass power cost and optimum plant size in
http://www.nrcan.gc.ca/forests/report/area/17601 , Accessed date: 23 May 2017. western Canada, Biomass Bioenergy 24 (2003) 445–464.
[3] Historica Canada, Pulp and Paper Industry, available at: http://www. [19] A. Kumar, P. Flynn, S. Sokhansanj, Biopower generation from mountain pine in-
thecanadianencyclopedia.ca/en/article/pulp-and-paper-industry/ , Accessed date: fested wood in Canada: an economical opportunity for greenhouse gas mitigation,
22 May 2017. Renew. Energy 33 (2008) 1354–1363.
[4] N.R. Canada, Benchmarking Energy Use in Canadian Pulp and Paper Mills, available [20] C.B. Sorenson, A Comparative Financial Analysis of Fast Pyrolysis Plants in
at: https://www.nrcan.gc.ca/sites/www.nrcan.gc.ca/files/oee/pdf/industrial/ Southwest Oregon, (2010).
technical-info/benchmarking/pulp-paper/pdf/benchmark-pulp-paper-e.pdf , [21] M.M. Wright, R.C. Brown, A.A. Boateng, Distributed processing of biomass to bio‐oil
Accessed date: 8 June 2018. for subsequent production of Fischer‐Tropsch liquids, Biofuels Bioprod Bioref 2
[5] K. Bradburn, CanBio Report on the Status of Bioenergy in Canada, (2014) available (2008) 229–238.
at: http://www.fpac.ca/wp-content/uploads/2014_CanBio_Report.pdf , Accessed [22] A.L. Brown, P.D. Brady, C.D. Mowry, T.T. Borek, An Economic Analysis of Mobile
date: 8 June 2018. Pyrolysis for Northern New Mexico Forests, Report No. SAND2011-9317 Sandia
[6] M.F. Demirbas, M. Balat, Recent advances on the production and utilization trends National Laboratories, US Department of Energy, Albuquerque, New Mexico, 2011.
of bio-fuels: a global perspective, Energy Convers. Manag. 47 (2006) 2371–2381. [23] D. Brown, A. Rowe, P. Wild, A techno-economic analysis of using mobile distributed
[7] M. Guo, W. Song, J. Buhain, Bioenergy and biofuels: history, status, and perspec- pyrolysis facilities to deliver a forest residue resource, Bioresour. Technol. 150
tive, Renew. Sustain. Energy Rev. 42 (2015) 712–725. (2013) 367–376.
[8] L.R. Lynd, R.A. Aziz, C.H. de Brito Cruz, A.F.A. Chimphango, L.A.B. Cortez, A. Faaij, [24] P.C. Badger, P. Fransham, Use of mobile fast pyrolysis plants to densify biomass and
N. Greene, M. Keller, P. Osseweijer, T.L. Richard, J. Sheehan, A. Chugh, L. van der reduce biomass handling costs—a preliminary assessment, Biomass Bioenergy 30
Wielen, J. Woods, W.H. van Zyl, A Global Conversation about Energy from Biomass: (2006) 321–325.
the Continental Conventions of the Global Sustainable Bioenergy Project, Interface [25] M.A. Palma, J.W. Richardson, B.E. Roberson, L.A. Ribera, J. Outlaw, C. Munster,
Focus, 2011. Economic feasibility of a mobile fast pyrolysis system for sustainable bio-crude oil
[9] M. Kumar, A.O. Oyedun, A. Kumar, Hydrothermal liquefaction of biomass for the production, Int. Food Agribus. Manag. Rev. 14 (2011).
production of diluents for bitumen transport, Biofuels Bioprod Bioref 11 (5) (2017) [26] M. Ha, M.L. Bumguardner, C.L. Munster, D.M. Vietor, S. Capareda, M.A. Palma,
811–829. T. Provin, Optimizing the logistics of a mobile fast pyrolysis system for sustainable
[10] M. Kumar, A.O. Oyedun, A. Kumar, A comparative analysis of hydrogen production bio-crude oil production, Proccedings of Conference Optimizing the Logistics of a
from the thermochemical conversion of algal biomass, Int. J. Hydrogen Energy 44 Mobile Fast Pyrolysis System for Sustainable Bio-Crude Oil Production, vol. 1, 2010.
(2019) 10384–10397. [27] P. Badger, S. Badger, M. Puettmann, P. Steele, J. Cooper, Techno-economic analysis:
[11] A.V. Bridgwater, Review of fast pyrolysis of biomass and product upgrading, preliminary assessment of pyrolysis oil production costs and material energy bal-
Biomass Bioenergy 38 (2012) 68–94. ance associated with a transportable fast pyrolysis system, BioResour 6 (2010)
[12] M. Patel, A.O. Oyedun, A. Kumar, R. Gupta, A techno-economic assessment of re- 34–47.
newable diesel and gasoline production from aspen hardwood, Waste Biomass [28] S. Sarkar, A. Kumar, Large-scale biohydrogen production from bio-oil, Bioresour.
Valorization 9 (2018) 1–16. Technol. 101 (2010) 7350–7361.
[13] M. Patel, A.O. Oyedun, A. Kumar, R. Gupta, What is the production cost of re- [29] A.S.J. Marshall, Commercial application of pyrolysis technology in agriculture,
newable diesel from woody biomass and agricultural residue based on experi- (2013) available at: https://ofa.on.ca/uploads/userfiles/files/pyrolysis%20report
mentation? A comparative assessment, Fuel Process. Technol. 191 (2019) 79–92. %20final.pdf , Accessed date: 22 May 2017.
[14] M. Patel, X. Zhang, A. Kumar, Techno-economic and life cycle assessment on lig- [30] A. Wong, H. Zhang, A. Kumar, Life cycle assessment of renewable diesel production
nocellulosic biomass thermochemical conversion technologies: a review, Renew. from lignocellulosic biomass, Int. J. Life Cycle Assess. 21 (2016) 1404–1424.
Sustain. Energy Rev. 53 (2016) 1486–1499. [31] H. Mahmudi, P.C. Flynn, Rail vs truck transport of biomass, Proccedings of
[15] A.V. Bridgwater, G.V.C. Peacocke, Fast pyrolysis processes for biomass, Renew. Conference Rail vs Truck Transport of Biomass, (20060, pp. 88-103.

17
M. Patel, et al. Biomass and Bioenergy 128 (2019) 105287

[32] J.D. Thompson, J. Klepac, W. Sprinkle, Trucking characteristics for an in-woods 05/A_Kumar.pdf , Accessed date: 8 June 2017.
biomass chipping operation, Proccedings of Conference Trucking Characteristics for [38] T. Pootakham, A. Kumar, Bio-oil transport by pipeline: a techno-economic assess-
an In-Woods Biomass Chipping Operation, 2012. ment, Bioresour. Technol. 101 (2010) 7137–7143.
[33] V. Francescate, E. Antonini, L.Z. Bergomi, Wood fuels handbook, (2008) available [39] A. Kumar, J.B. Cameron, P.C. Flynn, Pipeline transport of biomass, in:
at: http://www.biomasstradecentre2.eu/scripts/download.php?file=/data/pdf_ C.O. Breckenridge (Ed.), Proceedings of the Twenty-Fifth Symposium on
vsebine/literature/wood_fuels_handbook.pdf , Accessed date: 1 June 2017. Biotechnology for Fuels and Chemicals Held May 4–7, 2003, Springer, 2004, pp.
[34] H. Shahrukh, A.O. Oyedun, A. Kumar, B. Ghiasi, L. Kumar, S. Sokhansanj, Techno- 27–39.
economic assessment of pellets produced from steam pretreated biomass feedstock, [40] A.S.J. Marshall, P.F. Wu, S.H. Mun, C. Lalonde, Commercial application of pyrolysis
Biomass Bioenergy 87 (2016) 131–143. technology in agriculture, Proceedings of Conference Commercial Application of
[35] E. Agbor, A.O. Oyedun, X. Zhang, A. Kumar, Integrated techno-economic and en- Pyrolysis Technology in Agriculture, 2014, p. 1.
vironmental assessments of sixty scenarios for co-firing biomass with coal and [41] D.A. Laird, J.M. Novak, H.P. Collins, J.A. Ippolito, D.L. Karlen, R.D. Lentz,
natural gas, Appl. Energy 169 (2016) 433–449. K.R. Sistani, K. Spokas, R.S. Van Pelt, Multi-year and multi-location soil quality and
[36] N. Mahbub, A.O. Oyedun, A. Kumar, D. Oestreich, U. Arnold, J. Sauer, A life cycle crop biomass yield responses to hardwood fast pyrolysis biochar, Geoderma 289
assessment of oxymethylene ether synthesis from biomass-derived syngas as a diesel (2017) 46–53.
additive, J. Clean. Prod. 165 (2017) 1249–1262. [42] VoseSoftware, Model risk – Monte Carlo simulation, available at: https://www.
[37] A. Kumar, Techno-economic assessment of biochar production for carbon seques- vosesoftware.com/index.php , Accessed date: 12 February 2018.
tration, (2013) available at: http://albertabiochar.ca/wp-content/uploads/2013/ [43] Natural Resources Canada.

18