Articulo Procesos

Subscriber access provided by UB + Fachbibliothek Chemie | (FU-Bibliothekssystem)
Article
Design and assessment of advanced thermochemical
plants for second generation biobutanol production
considering mixed alcohols synthesis kinetics
Chinedu Obiora Okoli, and Thomas A. Adams
Ind. Eng. Chem. Res., Just Accepted Manuscript • DOI: 10.1021/acs.iecr.6b03196 • Publication Date (Web): 19 Jan 2017
Downloaded from http://pubs.acs.org on January 22, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Industrial & Engineering Chemistry Research is published by the American Chemical

Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.
Page 1 of 58 Industrial & Engineering Chemistry Research
1
2
3
4
5
6
7 Design and assessment of advanced
8
9
10
11 thermochemical plants for second generation
12
13
14
15 biobutanol production considering mixed
16
17
18
19 alcohols synthesis kinetics
20
21
22
23
24 Chinedu Okoli and Thomas A. Adams II*
25
26
27 Department of Chemical Engineering, McMaster University, 1280 Main Street West,
28
29 Hamilton, Ontario, L8S 4L7
30
31
32 Abstract
33
34
35 In prior work, the authors' developed a novel process for producing second generation
36
37
biobutanol through a thermochemical conversion route and mixed alcohols synthesis
38
39
40 (MAS) process and showed that it was economically competitive to the biochemical
41
42 route, and to gasoline under certain market scenarios. However, the prior work made
43
44 use of simplistic conversion models for the MAS reactor based on the U.S. National
45
46 Renewable Energy Laboratory future conversion targets for MAS catalysts. This work
47
48
49
improves on the past work by replacing the simplistic MAS reactor model with a
50
51 detailed kinetic model using a pilot-scale demonstrated catalyst leading to a more
52
53 realistic representation of the process, and enabling the impact of design parameters to
54
55
56 1
57
58
59
60
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research Page 2 of 58
1
2
3 be investigated for various novel process designs. Furthermore, economic and
4
5 environmental metrics such as the minimum butanol selling price (MBSP), and cost of
6
7 CO2 emissions avoided (CCA) were used to assess the potential of all the process
8
9
10 designs.
11
12
13 Keywords: second generation biobutanol; thermochemical; economic analysis;
14
15 minimum butanol selling price; cost of CO2 emissions avoided; mixed alcohol synthesis
16
17
18 1. Introduction
19
20 Biofuels are increasingly seen by policy makers globally as an important strategy to
21
22 help address the challenges posed by climate changing greenhouse gas (GHG)
23
24
25
emissions. This is highlighted by the continual increase in global biofuels production.
26
27 For example biofuels contributed 3 % to total road-transport fuel demand in 2013, with
28
29 this value expected to hit 8 % by 2035 1. Key to achieving this potential for biofuels is
30
31 the utilization of second generation biofuel feedstocks which are primarily
32
33
cellulosic/lignocellulosic terrestrial biomass such as wood chips and grasses. Second
34
35
36 generation biofuel feedstock are preferred over first generation biofuel feedstock
37
38 (derived from corn and other food crops) because unlike first generation biofuel
39
40 feedstock they are significantly less competitive with food production for land and their
41
42 feedstocks are or can be made widely available. In fact, the U.S. Congress Renewable
43
44
45
Fuel Standard (RFS) mandates a minimum production of 21 billion gal/year of
46
47 cellulosic/lignocellulosic biofuels by 2022 2. They aim to achieve this by both
48
49 technological improvements of biomass to biofuel processes, as well as improved
50
51 production and supply of biomass. As a positive step towards achieving the U.S. RFS
52
53
mandate, the U.S. Department of Energy and the U.S. Department of Agriculture
54
55
56 2
57
58
59
60
1
2
3 collectively champion a goal of developing a biomass supply chain which by 2030 is
4
5 able to produce 1 billion tons per year of biomass, solely to be used for second and third
6
7 generation biofuels production 3.
8
9
10 In terms of the biofuels of choice, researchers into second generation biofuels are
11
12
13 increasingly looking at biobutanol as a preferable option to bioethanol as a gasoline
14
15 replacement in automobiles because it offers advantages over bioethanol such as a
16
17 higher energy content, lower water miscibility, and better compatibility with existing
18
19 internal combustion engines and pipeline networks 4,5.
20
21
22 There are two routes for the production of butanol from second generation biomass
23
24
25
feedstocks, namely the biochemical route and the thermochemical route. The
26
27 biochemical route proceeds mainly through the acetone-butanol-ethanol (ABE)
28
29 fermentation process which was developed in UK in 1912 and was a commercial route
30
31 for acetone and butanol production from first generation biofuel feedstock (such as corn
32
33
and potatoes) up until the 1950s before the advent of petrochemical derived butanol 6,7.
34
35
36 In this route, bacteria species (mainly species of Clostridia) are used to ferment the
37
38 sugars in biomass into butanol. Despite its long history the ABE process faces a number
39
40 of key challenges such as the difficulty in handling second-generation biofuel
41
42 feedstocks such as forest residue because of their high lignin content which is difficult
43
44
45
to break down by fermentation organisms 8; low productivity of the fermentation
46
47 process 6; and the very challenging downstream separation process because of the low
48
6,9,10
49 concentration of the product in the fermentation broth . Review papers by
50
51 researchers further discuss these challenges and mention that research efforts into
52
53
metabolic engineering to improve fermentation bacteria strains and novel reactor and
54
55
56 3
57
58
59
60
1
2
3 separation technologies can help address these challenges in the future 6,7,10–12. However
4
5 these challenges mean that production of second generation biobutanol through the
6
7 biochemical route is not commercially ready 13.
8
9
10 One of the major pathways for the production of biobutanol from the thermochemical
11
12
13 route is the mixed alcohol synthesis (MAS) process. In this process, biomass is gasified
14
15 to produce mainly carbon monoxide (CO) and hydrogen (H2), together called syngas.
16
17 The syngas from biomass also contains impurities which are removed in a syngas
18
19 cleanup step before being sent to a MAS reactor in which the syngas is catalytically
20
21
22
converted to butanol and other alcohols. The MAS catalysts used in this process can be
23
24 subdivided into four groups namely, modified high-pressure methanol synthesis
25
26 catalysts, modified low-pressure methanol synthesis catalysts, modified Fischer-
27
28 14,15
Tropsch catalysts, and alkali-doped molybdenum catalysts . Of major interest for
29
30 butanol production are the modified methanol synthesis catalysts because they have
31
32
33 better selectivity to butanol than the other catalyst groups 15. The thermochemical route
34
35 has a number of advantages over the biochemical route such as the ability to handle
36
37 lignin-rich second generation biofuel feedstocks because of the ease of gasifying the
38
39 carbon rich lignin into syngas, and an easier separation step because the mixed alcohols
40
41
42
are at a high concentration in the liquid feed to the downstream separation section. The
43
44 major disadvantage of the thermochemical route is the low CO conversion of current
45
14,16,17
46 catalysts, with research currently ongoing into improvements in MAS catalysts .
47
48 Furthermore, this route requires capital-intensive equipment thus might not be
49
50
economically feasible at small processing scales.
51
52
53
54
55
56 4
57
58
59
60
1
2
3 Despite the merits of second generation biobutanol, quantitative metrics are required to
4
5 demonstrate both its economic and environmental potential. In this regard only very few
6
7 studies exist in the peer reviewed literature for both the biochemical and
8
9 18–21 18
10 thermochemical routes . Qureshi et al. assessed the techno-economics of the
11
12 biochemical conversion of wheat straw to n-butanol using the ABE fermentation
13
14 process based on a Clostridium beijenrickii fermentation bacteria specie. The plant
15
16 capacity was 150,000 tonnes of butanol/year with assumptions such as a 20 % mass
17
18
yield of butanol per sugars, feedstock cost of 24 $/tonne, and co-products price of 800
19
20
21 $/tonne for ethanol and 1,300 $/kg for acetone. The result of the analysis was that
22
23 butanol can be produced at 1.05 $/L in 2012 dollars. Kumar et al. 20 studied the techno-
24
25 economics of switchgrass to n-butanol via the ABE biochemical route. The plant was
26
27 designed to produce 10,000 tonnes of butanol/year with an assumed mass yield of 23.4
28
29
30 % butanol per sugars. Switchgrass was assumed to be purchased at 40 $/tonne, with co-
31
32 products ethanol sold at 890 $/kg and acetone sold at 810 $/kg. This results in a butanol
33
19
34 production price of 0.59 $/L, in 2012 dollars. Tao et al. analyzed the techno-
35
36 economics of i-butanol production from corn stover via the biochemical route using an
37
38
improved strain of Escherichia coli bacteria. The plant was designed to process 2,000
39
40
41 dry tonnes of feedstock with assumptions such as 65 $/dry tonne feedstock and butanol
42
43 yield of 31.45 g butanol/g sugars resulting in a butanol production price of 0.78 $/L in
44
45 2007 dollars. The assumption around butanol yield is an important one to note for this
46
47 study, as it is 85 % of the theoretical butanol yield of 37 g butanol/g sugars 22
, a very
48
49
50 high value in comparison to the other studies reviewed. Furthermore, this conversion is
51
52 quite optimistic as fermentation using this E. coli bacteria strain has not even been
53
54 demonstrated in bench scale studies 19. The study addressed the uncertainty around the
55
56 5
57
58
59
60
1
2
3 selected butanol yield by carrying out sensitivity analysis with results showing that in
4
5 the butanol yield range of 20 - 24 g butanol/g sugars (55 - 65 % of the theoretical
6
7 butanol yield) which is typical of other biochemical butanol studies, the butanol
8
9
10 production price is 1.20 - 1.02 $/L in 2007 dollars. These analyses highlight the
11
12 importance of taking into consideration techno-economic assumptions made on
13
14 feedstock costs, co-product prices, cost year of analysis, and butanol product yields,
15
16 among other factors when butanol production price results are assessed.
17
18
19 For second generation biobutanol production via the thermochemical route, Okoli and
20
21
22
Adams 21 is the only techno-economic study in the peer-reviewed literature. The authors
23
24 designed a 2,000 tonnes per day of wood chips production plant that converts syngas
25
26 from the gasification section to butanol and other mixed alcohols over a modified low-
27
28 pressure methanol synthesis catalyst. The MAS reactor was modeled using simple yield
29
30 and conversion data of the modified low-pressure methanol synthesis catalyst from
31
32 23
33 Herman . Furthermore, the CO to alcohols conversion was set at a 40 % (CO2-free
34
35 basis), which was motivated by U.S. National Renewable Energy Laboratory (NREL)
36
37 future targets of 50 % CO conversion (CO2-free basis) for MAS catalysts 24 but has not
38
39 yet been achieved in practice. The resulting butanol production price was 0.83 $/L in
40
41
42
2012 dollars. Due to the simplistic catalyst assumptions, a sensitivity analysis was
43
44 carried out on the CO conversion of the catalyst which showed that at the literature
45
46 reported CO conversion value of the catalyst at 8.5 % 23, the butanol selling price is 1.56
47
48 $/L in 2012 dollars.
49
50
51 As very little techno-economic studies have been carried out on second generation
52
53
biobutanol and on the thermochemical route in particular, more studies are required in
54
55
56 6
57
58
59
60
1
2
3 other to get a better economic and environmental understanding of the process. Thus
4
5 this current work aims to improve on the study carried out by Okoli and Adams 21. First,
6
7 the simplistic MAS reactor and catalyst model is replaced with a detailed kinetic model
8
9
10 using a high-pressure methanol synthesis catalyst that has been demonstrated at the pilot
11
12 scale 25. This change depicts a more realistic representation of the process, and allows
13
14 the impact of different design decisions, such as the impact of unreacted syngas recycle
15
16 choices on the butanol production, to be evaluated. Furthermore, this study also
17
18
improves on Okoli and Adams 21 by assessing other novel process configurations which
19
20
21 consider the impact of high temperature utility and power importation on both the
22
23 economic and environmental feasibility of second generation biobutanol production
24
25 using metrics such as the minimum butanol selling price (MBSP), and cost of CO2
26
27 equivalent emissions avoided (CCA). The CCA is an important environmental metric
28
29
30 for assessing GHG emission reduction technology as it informs biofuel technology
31
32 decision makers how much financial cost is incurred to reduce GHG emissions when a
33
34 particular biofuel technology is implemented over a fossil based fuel. To the best of the
35
36 authors' knowledge this metric has not been appropriately studied nor quantified in the
37
38
peer reviewed literature for second generation biobutanol.
39
40
41
42
Another area identified for improvements is the separation section in which a
43
44 conventional distillation sequence is used for butanol recovery. Potential exists for
45
46 improvements of this section by utilizing process intensification technology based on
47
48 dividing wall columns (DWCs) instead of conventional distillation columns. The use of
49
50 DWCs can potentially be significant as prior research has shown that DWCs can
51
52
53 improve operating and capital costs of conventional distillation sequences by up to 30 %
54
26,27
55 . However, these savings have never been investigated on a plant-wide level for
56 7
57
58
59
60
1
2
3 biofuel applications, and thus this work also aims to quantify these savings for
4
5 thermochemical second generation biobutanol applications.
6
7
8 2. Methods
9
10
11
2.1 Biomass feedstock and plant size
12
13 Forest biomass can come from a variety of sources such as: trees that are of harvestable
14
15 age but are not useful for lumber, trees killed by disturbances (such as fire, diseases or
16
17 insects), harvest residues, and trees from plantations grown specifically to provide
18
19
biomass for conversion to bioenergy. Furthermore, industrial forest processes such as
20
21
22 harvesting and milling operations as well as pulping processes provide additional
23
24 sources of biomass i.e. sawdust, bark and chips, and lignin-rich "black liquors". The
25
26 thermochemical conversion route is preferred over the biochemical route for the
27
28 conversion of forest biomass because of the high-lignin content of forest biomass 8. This
29
30
31 is because fermentation bacteria used for biochemical routes find great difficulty in
32
33 converting lignin which typically has to be removed in pre-treatment processes, while in
34
35 the thermochemical route the carbon rich lignin is easily converted to syngas through
36
37 the gasification process. Thus for biochemical biofuel routes agricultural residues have
38
39
been the preferred feedstock because of their lower lignin content in comparison to
40
41
42 forest resources 8.
43
44
45 The forest biomass used as the base case for this study is wood chips obtained from pine
46
47 trees. The plant gate ultimate and proximate analyses used to model it are shown in
48
49 Table 1. Also shown is the lower heating value (LHV) and the higher heating value
50
51 (HHV).
52
53
54
55
56 8
57
58
59
60
1
2
3 The plant is designed to handle a gasifier inlet flow of 2,000 tonnes per day (10 wt %
4
5 moisture content) of wood chips. This is a reasonable size for the plant leveraging on
6
7 other second generation biofuel studies 8,28 as well as the availability of forest resources
8
9
10 (existing and unexploited) in the U.S. put at 334 million dry tonnes in 2005 3.
11
12
13 Table 1. Ultimate analysis of pine wood chips used in this study 8.
14
15 Ultimate analysis wt % moisture
16 free basis
17 Carbon 50.94
18 Hydrogen 6.04
19
20 Nitrogen 0.17
21 Sulphur 0.03
22 Oxygen 41.9
23 Ash 0.92
24
25
HHV [MJ/kg] 19.99
26 LHV [MJ/kg] 18.59
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56 9
57
58
59
60
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37 Figure 1. Process flow diagram of the proposed thermochemical biomass-to-butanol process superstructure. See supporting information for stream conditions.
38
10
39
40
41
42
43
44
45
46 ACS Paragon Plus Environment
47
48
1
2
3
4
5
6
7
2.2 Process simulation overview
8
9 In this work, three major configurations for the production of butanol from woody
10
11 biomass are designed and assessed. The general process flow for all the designs is as
12
13
14 follows: first, woody biomass is gasified to produce syngas, which is then cleaned before
15
16 being used for alcohols production in the MAS reactor. The alcohols are then separated
17
18 downstream of the MAS reactor into butanol and a mixed alcohols co-product.
19
20
21
22
The three configurations are differentiated by their requirements for the provision of high
23
24 temperature process heat, and power. Configuration 1, which is also called the "self-
25
26 sufficient" configuration, is designed such that the plant does not utilize any external
27
28
29 sources of hot utilities, including high temperature heat, and power, thus it is totally
30
31 powered by renewable biomass. This design requirement means that any extra high
32
33 temperature heat required by processes such as the endothermic tar reforming process is
34
35
36 met by combusting a portion of the syngas from the gasifier. Furthermore, an adaptation
37
38 of configuration 1 called "configuration 1a", in which a DWC is used in place of a
39
40
conventional column for alcohols separation is also considered. In configuration 2,
41
42
43 referred to as the "natural gas (NG) import" configuration, combustion of NG is used for
44
45 the provision of high temperature process heat in place of biomass syngas combustion
46
47
48 because the cost per unit of heating value of NG is lower than that of biomass. The
49
50 argument against this approach is that the "greenness" of the biofuel is reduced as a result
51
52 of the fossil fuel input. The final configuration (configuration 3) also called the "NG and
53
54
55
56 11
57
58
59
60
1
2
3
4
5
6
7
power import" configuration, uses NG for high temperature heat as is done in
8
9 configuration 2. However, electricity is also imported for process use instead of power
10
11 generation through expansion of hot gases in turbines. This configuration is motivated by
12
13
14 past work which showed that costs associated with turbines can contribute as much as 25
15
16 % of the capital costs of a thermochemical biobutanol process 21, thus power importation
17
18 from the grid though increasing the operating costs of the plant might help reduce the
19
20
21 upfront capital costs. These decisions behind the three different configurations can be
22
23 assessed by considerations such as GHG emissions, renewable energy usage, capital and
24
25
operating costs.
26
27
28
29 All the design configurations considered in this work are simulated in Aspen Plus V8.4
30
31 software so as to estimate their mass and energy balances. The selection of physical
32
33 property packages, and unit operation specifications were done to be consistent with
34
35
36 Okoli and Adams 21, in which a second generation biobutanol plant via a thermochemical
37
38 route was designed and assessed. However, major differences between this work and that
39
40
work arise in the design of the gas cleanup, alcohol synthesis, and separation sections as a
41
42
43 result of the kinetic MAS reactor model and different MAS catalyst used for this work, as
44
45 well as considerations regarding NG use, power importation and the unreacted syngas
46
47
48 recycle configuration. These differences will be discussed in more detail in the
49
50 proceeding sections.
51
52
53
54
55
56 12
57
58
59
60
1
2
3
4
5
6
7
2.3 Process description
8
9 In Figure 1 a simplified process flow diagram superstructure is used to illustrate the
10
11 design configurations considered. It shows all the major steps required for the conversion
12
13
14 of woody biomass to butanol including gasification, syngas cleanup, mixed alcohols
15
16 synthesis, and alcohols separation. These processing steps are further discussed in
17
18 proceeding sub-sections, with Table 2 showing the major design parameters for these
19
20
21 steps.
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56 13
57
58
59
60
1
2
3
4
5
6
7
Table 2. Major design parameters of process areas
8
9 Gasification Gas cleanup (acid gas removal)
10 Feed rate per train (gasifier inlet) 1000 tonnes/ day Amine used Monoethanolamine
11 Parallel trains required Two (2) Amine concentration, wt % 35
12 o
Gasifier operating pressure 2.28 bar Amine temperature in absorber ( C) 43.33
13
o
14 Gasifier operating temp. 800 C Absorber pressure (bar) 31
15 Char combustor pressure 2 bar Stripper pressure (bar) 4.12
16 Char combustor temp. 850 oC Heat duty to remove CO2 (kJ/kg) 5337
17
18 Gas cleanup (tar reforming) Alcohol synthesis reactor
19 Reformer operating pressure (bar) 1.86 Gas hourly space velocity (h-1) 5000
20 o o
Reformer operating temp. ( C) 910 Reactor temperature ( C) 440
21
22 Reformer space velocity (h-1) 2,476 Pressure (bar) 120
23 Tar reformer conversions (%) Inlet CO2 concentration (wt %) <5
24 Methane (CH4) 80% Inlet sulphur concentration (ppmv) < 0.1
25
Ethane (C2H6) 99% Steam system and power generation
26
27 Ethylene (C2H4) 90% Turbine design Three stage turbine
28 Tars (C10+) 99% High pressure inlet conditions 58 bar, 482 oC
29
Benzene (C6H6) 99% Medium pressure inlet conditions 12 bar, 303 oC
30
31 Ammonia (NH3) 90% Low pressure inlet conditions 4.5 bar, 210 oC
32 Condenser outlet conditions 0.304 bar, saturated
33
Alcohol separation (distillation columns) Alcohol separation (distillation columns)
34
35 Conventional column 1 Dividing wall column (configuration 1a only)
36 Propanol recovery in overhead 99.3 mol% Methanol recovery in overhead 97.4 mol%
37 Butanol recovery in bottoms 99.3 mol% Butanol purity in side stream 96 mol%
38
Number of trays 56 Pentanol recovery in bottoms 92.4 mol%
39
40 Conventional column 2 Number of trays (wall section) 23
41 Butanol purity in overhead 96 mol% Number of trays (main section) 45
42 Pentanol recovery in bottoms 99 mol%
43
44 Number of trays 20 Cooling water system
45 Alcohol separation (Molecular sieve) Supply temperature (oC) 32
46 o
47 Outlet water content (wt %) 0.5 Return temperature ( C) 43
48
49
50
51
52
53
54
55
56 14
57
58
59
60
1
2
3
4
5
6
7
2.3.1 Feed handling and drying
8
9 In our analysis, we have chosen to use the following biomass pathway. Southern pine
10
11 trees are cut and piled on the ground, which have a moisture content of 50 wt%. They are
12
13
14 air dried over time to 35 wt% and then processed into wood chips. The wood chips are
15
16 sent to the biorefinery, where a five-day inventory is kept. Chips pulled from this pile
17
18 before use are dried again using forced ambient-air, bringing them down to a 30 wt%
19
20
21 moisture content with characteristics as shown in Table 1 8. The chips are subsequently
22
23 dried in a biomass dryer to 10 wt. % using hot flue gas from the char combustor and
24
25
catalyst regenerator via the steam generation system.
26
27
28
29 In this work, our mass and energy balance model begins at the point at which the chips
30
31 have a moisture content of 30 wt%, and so the biomass drying process using hot flue gas
32
33 is modeled in all the simulations to enable accurate and complete mass and energy
34
35
36 balance information for the steam generation system. However, the techno-economic
37
38 analysis model does not include the capital cost of the biomass dryer or other process
39
40
units upstream of the gasifier directly. Instead, the costs upstream of the gasifier inlet are
41
42
43 integrated into the price of biomass. This is in line with the feedstock supply and cost
44
45 model of the Idaho National Laboratory (INL) which accounts for the logistical, capital,
46
47
48 and operating costs associated with feed delivery, handling and drying of the biomass to
49
50 the up to the gasifier inlet 8. For example, the base case price used in our analysis
51
52 ($75.01/tonne) is the cost of biomass received at the plant gate plus the additional costs of
53
54
55
56 15
57
58
59
60
1
2
3
4
5
6
7
the drying and other upstream steps, as reported by the INL model which uses all of the
8
9 same biomass drying and moisture content assumptions as in our analysis.
10
11
12 If the actual moisture content at the point of cutting is higher or lower than the 50% used
13
14
in this study, the effective price of biomass would be affected by a small amount. This
15
16
17 would change the economics of the biorefinery slightly but not affect the technical
18
19 outcomes of the model. Although we did not do a sensitivity analysis on the moisture
20
21
22
content assumption directly, a sensitivity of the effect of changes in biomass price were
23
24 performed as a part of this analysis and are discussed in section 3.5.3.
25
26
27 The temperature for the flue gas leaving the steam generation section (stream 3 in Figure
28
29
30
1) is adjusted to ensure that the biomass is dried to 10 wt. % moisture content and that the
31
32 humidified flue gas leaves the system above its dew point temperature. In Aspen Plus this
33
34 is done by using a simple custom model that calculates the heat duty required to dry the
35
36
37 biomass and the corresponding temperature and humidity change of the flue gas stream.
38
39
40 2.3.2 Gasification
41
42 The dried biomass with 10 wt. % moisture content is fed to the gasifier, where it is
43
44
45
gasified with low-pressure steam to produce syngas. The gasifier used for this study is a
46
47 low-pressure allothermal indirect circulating fluidized bed gasifier whose output product
48
49 composition is modeled with temperature correlations from the Batelle Columbus
50
51
52 Laboratory test facility as reported by Dutta et al. 8. The gasification system is made up of
53
54 two beds, the gasifier and the char combustor. In the gasifier bed, the overall gasification
55
56 16
57
58
59
60
1
2
3
4
5
6
7
reaction process is endothermic, thus the required heat is supplied by circulating heated
8
9 olivine (calcined magnesium silicate, the gasifier bed material) from the char combustor
10
11 into the gasifier making the gasification system adiabatic. The products from the
12
13
14 gasification process are mainly CO, H2, CH4, tars and solid char. The olivine, which is an
15
16 inert, also exits from the gasifier alongside the products. The gaseous products are
17
18 separated from the olivine and char with a cyclone, with the solids recycled back to the
19
20
21 char combustor while the raw syngas is sent to the gas cleanup section. In the char
22
23 combustor, the char is combusted with air thereby producing hot combustion gases and
24
25
ash as well as heating up the olivine. The olivine, ash, and hot combustion gases exit from
26
27
28 the char combustor and are separated from each other through a series of cyclones. The
29
30 hot olivine is recycled back to the gasifier, while the hot combustion gases are used to
31
32
33
generate steam in the steam system and dry the biomass before it enters the gasifier.
34
35 2.3.3 Gas cleanup
36
37 After gasification, a gas cleanup step is necessary to remove impurities in the raw syngas,
38
39
40 as these impurities such as tars, sulphur and CO2 can foul process equipment and poison
41
42 the MAS catalyst. The gas cleanup step comprises of tar reforming, syngas quenching,
43
44 and amine scrubbing. The raw syngas is first sent to the tar reformer, where unreacted
45
46
47 tars, methane, and other hydrocarbons are reformed to CO and H2. The tar reformer is a
48
49 circulating, heterogeneous, fluidized catalyst bed system that is made up of a reforming
50
51
bed and a catalyst regeneration bed. In the reforming bed, the endothermic reforming
52
53
54 reactions which is catalyzed by a fluidizable Ni/Mg/K catalyst take place between steam
55
56 17
57
58
59
60
1
2
3
4
5
6
7
and raw syngas 8. The tar reforming catalyst is separated from the syngas through a
8
9 cyclone at the reformer exit and sent to the catalyst regenerator. The catalyst regeneration
10
11 occurs in the exothermic catalyst regenerator by combusting entrained coke on the
12
13
14 catalyst, after which the regenerated catalyst is separated from the hot flue gases and sent
15
16 back to the tar reformer, thus completing the loop. The heat requirement of the reformer
17
18 bed, which is maintained at isothermal conditions, is supplied by heat transfer from the
19
20
21 hot flue gases produced in the exothermic catalyst regenerator. Furthermore, the hot
22
23 combustion gases stream also has its heat recovered for steam generation and biomass
24
25
drying. If there is insufficient heat supply from the catalyst regenerator to the tar
26
27
28 reformer, this can be addressed by combusting unreacted syngas from downstream, and a
29
30 portion of the raw syngas feed from the gasifier as is done in configurations 1 and 1a
31
32
33
(stream 9 in Figure 1), or by using heat from combusted NG as in configurations 2 and 3
34
35 (stream 12 in Figure 1). The NG specifications used are shown in Table S6.
36
37
38 The syngas from the tar reformer is cooled to 60 °C before it is quenched and water
39
40
41 scrubbed to remove remaining particulate matter, tars, and other impurities. The purge
42
43 water stream is sent for treatment to a downstream waste water treatment facility, while
44
45 makeup water is also added. The cooled syngas is compressed to 30 bar in a five-stage
46
47
48 compressor before being sent for H2S and CO2 (acid gas) removal in the amine scrubber
49
50 system. The amine scrubbers as well as the subsequent ZnO bed are used to remove acid
51
52
53
gas in the syngas so it meets the MAS catalyst specifications of less than 0.1 ppm H2S
54
55
56 18
57
58
59
60
1
2
3
4
5
6
7
and less than 5 wt % CO2 15. The H2S and CO2 from the amine scrubber exit are sent to a
8
9 LO-CAT® system where elemental sulphur and CO2 are generated 15
. A portion of the
10
11 generated CO2 can be recycled back to the tar reformer to regulate the H2/CO ratio of the
12
13
14 tar reformer exit syngas. Increasing CO2 in the feed to the tar reformer favours the reverse
15
16 reaction of the water-gas shift reaction, shown in equation (1), leading to a reduction in
17
18 H2 and an increase in CO, while the forward reaction is favoured if CO2 is reduced in the
19
20
21 tar reformer feed.
22
23
24 + ↔ + (1)
25
26
27 2.3.4 Alcohol Synthesis
28
29
30
The clean syngas from the syngas cleanup section is compressed to 120 bar in a multi-
31
32 stage compressor and then heated to 440 °C before being passed into the MAS reactor.
33
34 The MAS reactor is an isothermal fixed bed reactor which uses a modified high pressure
35
36 25
37 methanol synthesis catalyst (K-promoted Zn/Cr/O catalyst) . The reactor products
38
39 consist mainly of C1 – C4 alcohols, water, methane, C5+ alcohols and other hydrocarbon
40
41
42 products. The reactor products are cooled to 43 °C by heat exchange with process streams
43
44 and cooling water, then using a series of flash drums the unconverted syngas and other
45
46
47
light gases are separated from the liquid alcohols. A portion of the unconverted syngas,
48
49 which is still at high pressure, is recompressed and recycled back to the MAS reactor
50
51 while the rest is expanded to 30.4 bar through a turbine to recover power in
52
53
54 configurations 1, 1a and 2, and through a flash valve in configuration 3. A portion of this
55
56 19
57
58
59
60
1
2
3
4
5
6
7
expanded stream is recycled to the acid gas removal section to remove CO2 before being
8
9 sent back to the MAS reactor, while the remaining stream is expanded (through a turbine
10
11 in configurations 1, 1a and 2, and a flash valve in configuration 3) to 2.3 bar and sent to
12
13
14 the indirect tar reforming system where it is either reformed in the tar reforming bed or
15
16 combusted in the catalyst regenerator to help meet the heat requirements of the plant. The
17
18 liquid alcohols stream is then sent to the alcohols separation section.
19
20
21 21
22
One major improvement of this work over past work is the use of a kinetic model to
23
24 predict the production of butanol and other alcohols from the MAS reactor. This allows
25
26 the impact of varying operating conditions such as feed composition, feed flowrate,
27
28
29 temperature, pressure etc. on the prediction of reactor products, particularly butanol, to be
30
31 studied. The kinetic model used to predict the mixed alcohols production over the high
32
33 pressure modified methanol synthesis catalyst was developed by Beretta et al. 25
of the
34
35
36 Snamprogetti SpA research laboratories and the Polytechnic University of Milan using
37
38 data from both laboratory and pilot scale experiments. Their work provides data on the
39
40
whole set of reactions for mixed alcohols synthesis, and the accompanying rate
41
42
43 expressions and kinetic parameter estimates, and is adapted for use in this study through
44
45 model regression of kinetic parameters to improve alcohols prediction. Note that the
46
47
48 kinetic model is a lumped parameter model that accounts for the effects of reactant and
49
50 product concentrations, reactor pressure, and temperature on the MAS reactor output
51
52 predictions.
53
54
55
56 20
57
58
59
60
1
2
3
4
5
6
7
The reacting system is reduced to a selected number of components and pseudo-
8
9 components as follows, CO, H2, CO2, H2O, methanol, ethanol, propanol, isobutanol, C4+
10
11 higher alcohols, methane and C2+ hydrocarbons. For further simplicity in this work, the
12
13
14 C4+ higher alcohols are approximated as pentanol and the C2+ hydrocarbons as ethane.
15
16 The mixed alcohols reaction scheme is shown in equations (2) - (10):
17
18
19 + 2 ↔ (2)
20
21
22 + ↔ + (3)
23
24
25
2 → + (4)
26
27 + → + (5)
28
29
30 + → + (6)
31
32
33
2 + → + 2 (7)
34
35 2 → ( ) + (8)
36
37
38 + 3 → + (9)
39
40
41
2 + 5 → + 2 (10)
42
43 From the reaction set it can be seen that methanol is the key building block for the
44
45
46
production of higher alcohols. In addition, methanol production from CO and H2 is
47
48 considered reversible and chemical equilibrium limited. Furthermore, CO and H2 react to
49
50 also produce methane and ethane, while the water gas shift reaction accounts for the
51
52
53 formation of CO2 which is equilibrium limited. Finally, the model also accounts for the
54
55
56 21
57
58
59
60
1
2
3
4
5
6
7
production of other oxygenated organic compounds such as dimethyl ether in equation
8
9 (8).
10
11
12 In Aspen Plus a plug flow reactor (RPlug) model is used to model the MAS reactor, with
13
14
the reaction set, rate expressions and kinetic parameters implemented using the Aspen
15
16
17 Plus Langmuir Hinshelwood Hougen Watson and power law kinetic models. More details
18
19 of the kinetic model study carried out for this work are available in the appendix.
20
21
22 For simulation simplicity, the operating temperature, pressure and gas hourly space
23
24
25 velocity (GHSV) of the MAS reactor are fixed. The GHSV is fixed at 5,000 h-1 based on
26
27 8,24
values reported in other methanol and alcohols production studies , while the
28
29
30
temperature and pressure are set at the upper bound (440 °C) and mid range value (120
31
25
32 bar) of the experiment test range reported by Beretta et al. . From the experiments by
33
34 25
Beretta et al. , it is known that butanol yield is maximized at higher temperatures and
35
36
37 pressure, and lower GHSV, therefore the choice of temperature and pressure is done with
38
39 the idea of maximizing the production of butanol from the MAS reactor while limiting the
40
41 costs associated with operation at high pressures such as the capital cost of reactors and
42
43
44 compressors, and the energy penalty associated with syngas compression to high
45
46 pressures. The selection of temperature and pressure used in this work is thus seen as a
47
48
good compromise. Other variables that significantly affect the MAS reactor product yield
49
50
51 such as the selected unreacted syngas recycle scheme and H2/CO ratio are determined
52
53 using an optimization framework that will be discussed in the results section.
54
55
56 22
57
58
59
60
1
2
3
4
5
6
7
2.3.5 Alcohol Separation
8
9 The raw mixed alcohols stream from the alcohol synthesis section is flashed to 4 bar to
10
11 remove absorbed gases which are subsequently recycled to the tar reformer in the gas
12
13
14 cleanup section. The liquid alcohols are then superheated before being dehydrated by
15
29
16 passing through a molecular sieve which adsorbs water . The dehydrated alcohols are
17
18 cooled down to a liquid state at 45 °C before they are separated into final products
19
20
21 through two conventional distillation columns in series. In column 1, methanol, ethanol,
22
23 propanol and any remaining light gases are removed in the distillate while butanol and
24
25
higher alcohols (C5+) are removed in the bottoms. In column 2, butanol is recovered in the
26
27
28 distillate at ≥ 96 wt% purity to meet ASTM fuel specification standards 30, while the C5+
29
30 alcohols are recovered in the bottoms and can be sold as a mixed alcohols co-product.
31
32
33 The methanol rich liquid distillate from column 1 is compressed and recycled to the
34
35 molecular sieve where it is used as a sweep gas to recover adsorbed water from the
36
37 molecular sieve in its desorption phase. As the molecular sieve is cyclic, two molecular
38
39
40 sieves are used in parallel such that one is always adsorbing while the other is desorbing.
41
42 The methanol and recovered water vapour stream is recycled back to the tar reformer,
43
44
where it is reformed back to syngas, as an analysis showed that recycling back to the
45
46
47 MAS reactor reduced the catalyst selectivity to butanol (though not significantly).
48
49
50
51
52
53
54
55
56 23
57
58
59
60
1
2
3
4
5
6
7
The DWC used in configuration 1a in place of the two conventional distillation columns
8
9 is a ternary product DWC designed using a methodology discussed in the authors' past
10
11 work 27. A summary of the design is shown in Table 2.
12
13
14
15 2.3.6 Utilities (steam system, power generation and cooling)
16
17 Steam production is necessary in all the assessed configurations to help meet process
18
19 heating requirements, for power generation, and for direct process needs. In the design of
20
21
22 configurations 1, 1a and 2, high pressure (HP) steam is produced for power generation,
23
24 while low pressure (LP) steam is produced in all the configurations for direct injection
25
26
into the biomass gasifier and tar reformer. Furthermore, LP steam is used for indirect
27
28
29 heating in the reboilers of the distillation columns and amine system. Heat exchange
30
31 between water and hot process streams like the flue gases from the char combustor and
32
33
34
catalyst regenerator, as well as the exothermic heat from the MAS reactors are used to
35
36 generate HP steam, while LP steam is produced when the HP steam is expanded through
37
38 steam turbines, to produce work as in configurations 1, 1a, and 2, or through an expansion
39
40
41 valve as in configuration 3. Further power is produced in configurations 1, 1a, and 2 by
42
43 the expansion of high pressure unconverted syngas through turbines in the alcohol
44
45 synthesis and gas cleanup sections. It is reiterated here that all the power requirements for
46
47
48 configuration 3 are met by importing power from the grid, thus creating a trade-off
49
50 between eliminating the high cost of capital associated with purchasing turbines and
51
52
53
expanders, and increasing the operating costs associated with purchasing grid power.
54
55
56 24
57
58
59
60
1
2
3
4
5
6
7
All configurations have their cooling requirements met by using forced-air heat
8
9 exchangers and cooling water after process stream to process stream heat exchange has
10
11 been carried out. The idea behind using forced-air heat exchangers is to reduce the water
12
13
14 demand of the processes by replacing cooling water in the provision of cooling for the
15
16 amine system condensers and steam turbine exhaust. Aspen Plus calculator blocks
17
18 31,32
utilizing correlations from literature are used to compute the power requirements of
19
20
21 the forced-air heat exchangers for the simulations.
22
23
24 2.4 Economic analysis
25
26
An economic analysis using the MBSP serves the purpose of providing a metric to assess
27
28
29 the economic merits of the various plant configurations with respect to themselves as well
30
31 as to other second generation biobutanol processes and conventional gasoline. The
32
33
34
economic analysis was done on an "nth-plant" basis, meaning that the learning curve
35
36 associated with building new plants of this type has been surmounted. The MBSP is the
37
38 unit selling price of butanol over the plant's lifetime for which the net present value
39
40
41 (NPV) is zero. A discounted cash flow rate of return analysis is used to compute the
42
43 MBSP and takes into consideration capital and operating costs of the processes as well as
44
45 other assumptions of economic parameters which are detailed in Table S7.
46
47
48
49
50
51
52
53
54
55
56 25
57
58
59
60
1
2
3
4
5
6
7
Data used for estimating the capital costs of the various process units are based on mass
8
9 and energy results of converged Aspen Plus simulations, cost data from Aspen Capital
10
11 8,24
Cost estimator software and literature sources, particularly U.S. NREL reports .
12
13
14 Literature reported values are scaled to the required size by using the capacity power law
15
16 expression as shown in equation (11), with m varying from 0.48 to 0.87. The resulting
17
18 cost (Cost2) is adjusted to 2014 U.S. dollars by using the Chemical Engineering Plant
19
20
21 Cost Index 33.
22
"#"$%&
23

= ("#"$%&)' (11)
24
25
26
27 Operating costs can be grouped into fixed and variable operating costs. Fixed operating
28
29
30 costs are calculated by using correlations from Seider et al. 34, and include items such as
31
32 maintenance, operating overhead, labour related operations, property tax and insurance.
33
34 On the other hand variable operating costs include items such as feedstock costs, cooling
35
36
37 water costs etc. which vary with production rates. The values of the variable operating
38
39 costs used for this study are shown in Table S8. All the values shown in the table have
40
41 been adjusted to 2014 U.S. dollars from their reference values by using an inorganic
42
43 35
44 index obtained from the U.S. Bureau of Labor Statistics . In addition to the sale of
45
46 butanol as a product, mixed alcohols and electricity (for configurations 1, 1a, and 2) are
47
48
49
sold as co-products to generate additional revenue with the selling price of mixed
50
51 alcohols calculated as 90% of the price of gasoline on an HHV equivalent basis.
52
53
54
55
56 26
57
58
59
60
1
2
3
4
5
6
7
2.5 Cost of CO2 equivalent emissions avoided (CCA) analysis
8
9 The reduction of GHG emissions in the transportation sector is one of the major
10
11 objectives driving policy for the use of biofuels as a replacement for fossil derived fuels
12
13
14 in vehicles. However, consideration has to be given to the cost associated with reducing
15
16 GHG emissions, as on an energy density basis fossil fuels are usually cheaper to produce
17
18 than biofuels. This cost can be evaluated using the CCA metric, which is defined as the
19
20
21 marginal cost required to avoid the emission of a unit of GHG emissions when a biofuel
22
23 is combusted as a replacement for a fossil derived fuel. The unit of GHG emissions is
24
25
"tonne per CO2 equivalent emissions (CO2e)". The lower the CCA, the better the biofuel
26
27
28 is for reducing GHG emissions to the environment. In addition, the CCA serves as a good
29
30 way to compare biofuel processes to each other and to other GHG emission reduction
31
32
33
technologies because it factors in both cost and life cycle impacts. The CCA is computed
34
35 in this study by using equation (12), and conventional gasoline as a baseline.
36
37
38 $
$ /%01"+2 '"45%+"2 $ ) . B/CD E 5"2%+, #4%$,
67
39 ( )++,- ,. = <=>>?@A ? = F8 E F/
(12)
40 898 ,'%%+ ":%;,; ) .
67
41
42
43 The carbon intensity of gasoline (CIG) is its total wells-to-wheels life cycle emissions per
44
45
46
unit of energy. It is made up of the sum of the direct GHG emissions from combustion of
47
48 gasoline in a vehicle, and the indirect GHG emissions of its entire upstream supply chain
49
50 including oil drilling, production, refining, and distribution. The carbon intensity of
51
52
53 biobutanol (CIB) is also the wells-to-wheels life cycle emissions of biobutanol per unit of
54
55
56 27
57
58
59
60
1
2
3
4
5
6
7
energy which includes all indirect GHG emissions associated with upstream biomass
8
9 production, the indirect GHG emissions of utilities used in the plant (NG and electricity
10
11 for example), and the direct GHG emissions from the combustion of biobutanol in a
12
13
14 vehicle. For this analysis all GHG related chemicals are evaluated using the IPCC 100-
15
16 year metric 36. A summary of all direct and indirect GHG emissions data along the wells-
17
18 to-wheels life cycle considered in this work are shown in Table S9 for a U.S. based plant.
19
20
21 Note that butanol combustion in a vehicle was estimated by assuming 100% conversion
22
23 of all carbon atoms into CO2.
24
25
26
27
3. Results and discussion
28
29 3.1 Particle swarm optimization of recycle configurations
30
31 The selection of the percentage of unreacted syngas stream to be recycled directly back to
32
33
34
the MAS reactor, to the amine absorber or to the tar reforming system is not an arbitrary
35
36 decision. Recycling directly back to the MAS reactor is the best option to improve the
37
38 butanol yield from the MAS reactor, but the amount that can be recycled is constrained by
39
40
41 the requirement to have less than 5 wt. % of CO2 in the reactor feed, as greater than this
42
43 amount of CO2 poisons the catalyst 15. The next option to recycle to the amine absorber in
44
45 the gas cleanup section allows CO2 to be removed to meet the MAS reactor CO2
46
47
48 specification, albeit with a penalty of high recompression costs as the syngas flowrate
49
50 through the multi-stage compressor is increased. Finally, recycling back to the tar
51
52
reforming section allows higher hydrocarbons formed in the MAS reaction process to be
53
54
55
56 28
57
58
59
60
1
2
3
4
5
6
7
reformed back to syngas, and/or the unreacted syngas combusted in the catalyst
8
9 regenerator to help meet the heat requirements of the process. The downside of this is that
10
11 it reduces the butanol yield.
12
13
14
The determination of the optimal unreacted syngas recycle configuration can thus be set
15
16
17 up as an optimization problem. A simple objective function used for this study was to
18
19 maximize the contribution margin of the biobutanol plant, where the contribution margin
20
21
22
is defined as the operating revenue minus the variable operating cost. The constraints
23
24 used in the optimizer are feasibility constraints to ensure that the simulation runs without
25
26 errors and mass and energy is conserved, a MAS reactor feed constraint to limit the
27
28
29 amount of CO2 in the reactor feed stream to less than 5 wt %, and a butanol product
30
31 purity constraint to have greater than 96 mol % purity of butanol. Another decision
32
33 variable included in the optimization is the syngas H2/CO ratio at the MAS reactor inlet,
34
35
36 because experimental studies by Beretta et al. 25 showed that butanol yield from the MAS
37
38 reactor is optimal when the H2/CO ratio is in the range of 0.5 to 1.0.
39
40
41 Mathematically, the optimization problem is formulated as follows:
42
43
Maximize Contribution Margin
44
45 Decision variables:
46 1. RF1: Fraction of unreacted syngas recycled to the MAS reactor - (stream 26/ stream 25) of
47 Figure 1.
48
49 2. RF2: Fraction of unreacted syngas recycled to the amine absorber - (stream 28/ stream 27) of
50 Figure 1.
51
52 3. H-C: H2/CO ratio of syngas at the tar reformer exit - stream 14 of Figure 1.
53
54 Inequality constraints:
55
56 29
57
58
59
60
1
2
3
4
5
6 1. CO2 in MAS reactor feed ≤ 5 wt %
7
8 2. Purity of butanol in butanol product stream ≥ 96 mol %
9
10 Equality constraints:
11 1. Contribution margin = Revenue − Variable Operating Cost (VOC)
12
13 2. Revenue = butanol product flowrate × butanol selling price + mixed alcohols product flowrate ×
14 mixed alcohols selling price + electricity exported × electricity selling price
15
16 3. VOC = [biomass feed flowrate × biomass purchase price + NG flowrate × biomass purchase
17 price + electricity exported × electricity selling price] × 1.05
18
4. Material and energy balance equations of process flowsheet
19
20
21 The selling prices of butanol and mixed alcohols are calculated based on their HHV
22
23 adjusted gasoline prices (90 % of that value in the case of mixed alcohols). The prices
24
25 used for biomass, NG, gasoline and electricity in the optimization are given in Table S8.
26
27
28 Note also that the variable operating cost (VOC) calculation is increased by 5% to
29
30 account for miscellaneous VOC items such as makeup water, catalyst refills etc.
31
32
33 A particle swarm optimization (PSO) algorithm was used for this analysis. The PSO
34
35
36 algorithm is a nature-inspired derivative-free optimization technique that has found wide
37
38 use in the peer-reviewed literature for optimizing complicated process systems or systems
39
40
with no gradient information available, such as Aspen Plus sequential modular
41
42
43 flowsheets. Although PSO cannot guarantee that it will find the global optimum, the
44
45 global optimum is not strictly required in this case. In-depth discussions about PSO and
46
47
48
other similar derivate-free optimization algorithms can be found in books, such as those
49
50 written by Gendreau and Potvin 37, and Kaveh 38.
51
52
53
54
55
56 30
57
58
59
60
1
2
3
4
5
6
7
The PSO optimization framework used here is adapted from the authors’ previous work
8 27
9 on optimizing DWCs in Aspen Plus in which MATLAB (which executes the PSO
10
11 algorithm) is linked to the Aspen Plus simulation file using an Excel Visual Basic for
12
13
14 Applications interface. The decision variables are calculated by the PSO algorithm in
15
16 MATLAB and passed to Aspen Plus which uses them to run the plant simulations. At the
17
18 completion of the simulation, Aspen Plus returns the values of the variables required to
19
20
21 compute the objective function, as well as values of the constraints. If the constraints are
22
23 violated the objective function in the PSO algorithm is given a large penalty so the
24
25
algorithm's search is driven away from the infeasible region. The algorithm ends when a
26
27
28 convergence criterion is met or a maximum number of iterations have been reached. In
29
30 this work 10 particles with a maximum iteration of 100 for each particle is used. The
31
32
33
resulting optimal values of the decision variables for the base case configurations are
34
35 shown in Table 3.
36
37 Table 3. Results of optimum decision variables
38
39
RF1 RF2 H-C
40
41 Config. 1 0.2247 0.1828 0.9050
42 Config. 2 0.1373 0.4414 0.8153
43 Config. 3 0.1556 0.6028 0.7319
44
45 3.2 Process modeling results
46
47 All the assessed configurations were simulated in Aspen Plus to enable quantification of
48
49
50
their mass and energy balances, as well as size processing units. Stream conditions which
51
52 correspond to Figure 1 for all the different configurations are detailed in the supporting
53
54 information available on the journal website, while in Table 4 the major process modeling
55
56 31
57
58
59
60
1
2
3
4
5
6
7
results for the different configurations are summarized. The table shows the major feed
8
9 and product flows of the configurations, as well as their net power and energy efficiency.
10
11 The plant energy efficiencies shown in the table are computed on an HHV basis, and
12
13
14 defined as the total HHV of the output products (butanol, mixed alcohols and electricity)
15
16 divided by the total HHV of the input feedstocks (wood biomass, NG, and electricity),
17
18 noting the exception of course that since HHV has no meaning with regards to electric
19
20
21 power, the actual electric energy is used instead.
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56 32
57
58
59
60
1
2
3 Table 4: Major flowrates and process energy efficiency of all configurations
4
5 Config. 1- Config. Config. 2 - Config. 3 -
6 Self 1a- Self NG import NG &
7 sufficient sufficient plant power
8 plant + DWC import
9 plant plant
10
11 Biomass flow rate (kg/h) 83,384 83,384 83,384 83,384
12 Natural gas requirement (kg/hr) - - 23,730 22,161
13
14 Makeup water requirement (kg/hr)
15 Boiler feed water makeup 4,093 4,049 4,190 2,105
16 Cooling water makeup 16,253 16,377 37,081 287,680
17 Product yields (kg/hr)
18
19
Butanol 10,659 10,864 22,063 23,513
20 Mixed alcohols 5,154 4,930 5,541 4,212
21 Total product yield, mass basis 15,813 15,793 27,604 27,725
22 % products yield per feed (Biomass 19.0 18.9 25.8 26.3
23 +NG), mass basis
24
25 Power (MW)
26 Power generation 57.83 58.64 115.94 -
27 Power use 57.17 57.31 111.55 126.67
28 Net power produced (MW) 0.66 1.33 4.39 (126.67)
29
Biomass HHV (MW) 463 463 463 463
30
31 Natural gas HHV (MW) - - 348 325
32 Butanol HHV (MW) 113 116 235 250
33 Mixed alcohols HHV (MW) 48 46 51 39
34
Total input HHV + electricity import 463 463 812 915
35
36 Total output HHV + electricity export 162 163 291 289
37 Plant energy efficiency (%) 35.0 35.1 35.8 31.6
38
39
It can be seen from Table 4 that using heat from NG combustion as a high temperature
40
41
42 heat source leads to an increase in the liquid product yields of configurations 2 and 3.
43
44 This happens because the syngas flowrate to the MAS reactor is increased thus
45
46 increasing the yields of butanol and mixed alcohols. It can also be seen from the table
47
48 that there is a huge power import requirement of 127 MW for configuration 3 in
49
50
51 contrast to all the other configurations which export electricity. Besides resulting in a
52
53 huge power import requirement, the decision to not generate power in configuration 3
54
55 leads to a huge makeup water requirement as excess heat from the process is not used to
56 33
57
58
59
60
1
2
3 generate power but is instead wasted through cooling with cooling water. The large
4
5 amount of makeup water required is because of increased evaporative losses in the
6
7 cooling tower due to the increased water flowrate required for cooling the unused heat.
8
9
10 The excess heat available in configuration 3 means that there is a potential for the plant
11
12 to provide off-plant district heat as a way to minimize the cooling water losses. For a
13
14 similar reason, the makeup water requirement for configuration 2 is twice that of
15
16 configuration 1 because of the larger flowrates through the plant and thus higher cooling
17
18
water requirements leading to higher evaporative water losses.
19
20
21
22 Due mainly to the higher total liquids production from configuration 2, its plant energy
23
24 efficiency is higher than configuration 1. However configuration 3 which has the
25
26 highest total liquids products (slightly higher than configuration 1) has the lowest plant
27
28 energy efficiency because of the huge power import requirement. Note though that the
29
30
31 plant energy efficiency of the assessed configurations are lower than the comparable
32
33 thermochemical second generation biobutanol plant by Okoli and Adams 21, which is 46
34
35 % on a HHV basis, because the plants in the present work have lower once-through CO
36
37 conversions (CO2-free basis). For example the once-through CO conversion (CO2-free
38
39
basis) of configuration 1, which is a self-sufficient plant similar to that of Okoli and
40
41 21 21
42 Adams is 22 % while prior work assumes 40 % conversion based on U.S. NREL
43
44 MAS catalysts development targets 24.
45
46
47 When configuration 1 is compared to configuration 1a, in which a DWC is used in place
48
49 of conventional columns in the separation section, the product yields and plant thermal
50
51
52
efficiency are very similar though configuration 1a has twice the net power production
53
54 of configuration 1. However, it is not very clear from the process results if there are any
55
56 34
57
58
59
60
1
2
3 overall plant improvements obtained by utilizing a DWC in place of conventional
4
5 columns. Thus any potential improvements have to be investigated through an economic
6
7 analysis which factors in revenues, capital, and operating costs. This is discussed in the
8
9
10 next section.
11
12
13 3.3 Economic analysis results
14
15 An economic assessment of all the configurations was carried out with the results
16
17 summarized in Table 5. From the table it can be seen that amongst the base
18
19
configurations, configuration 2 has the highest total capital investment (TCI) followed
20
21
22 by configuration 3 and then configuration 1. The higher TCI of configuration 2 over
23
24 configuration 1 is because the combustion of NG over syngas for high temperature heat
25
26 in the tar reforming system of configuration 2 enables a higher flowrate of syngas to be
27
28 available for conversion downstream. This cascades down to higher flowrates of other
29
30
31 accompanying streams through the plant sections downstream of the gasifier leading to
32
33 the requirement of larger process equipment and thus capital costs. However, as
34
35 configurations 2 and 3 have similar flowrates, the lower TCI of configuration 3 is
36
37 because it has no electricity generating turbines thus leading to capital cost savings.
38
39
40 In terms of total operating costs (TOC), configuration 3 has the highest TOC because
41
42
just like configuration 2 it has higher flowrates downstream of the gasifier in
43
44
45 comparison to configuration 1, but unlike the other configurations there is a significant
46
47 additional cost associated with electricity purchase from the grid.
48
49
50 When revenue from co-products is considered, configuration 2 has the highest revenue
51
52 because it produces the largest amount of mixed alcohols and exports the largest amount
53
54
55
56 35
57
58
59
60
1
2
3 of electricity. In contrast configuration 3 has the lowest co-products revenue because it
4
5 produces the least amount of mixed alcohols and has no electricity export.
6
7
8 The MBSP, computed through a DCFROR analysis, unifies all the co-product revenue,
9
10 capital and operating cost results into a single economic value thus allowing the
11
12
13 economic potential of producing butanol through each considered configuration to be
14
15 assessed. As can be seen from Table 5 configuration 2 has the lowest MBSP of 0.92 $/L
16
17 meaning that it is the most economically viable plant in comparison to the others for
18
19 producing biobutanol. The reason configuration 2 has a lower MBSP in comparison to
20
21
22
configuration 1(which has the highest MBSP) is because the higher cost per energy of
23
24 biomass in comparison to NG mean that it is more valuable for producing butanol than
25
26 being used for high temperature heat production as is done in configuration 1.
27
28 Configuration 3 has the next best MBSP at 1.10 $/L. Compared to configuration 1, its
29
30 result shows that the capital cost savings benefit of not investing in electricity
31
32
33 generation infrastructure is far outweighed by the higher operating costs attributable to
34
35 electricity import from the grid.
36
37
38 As discussed in the introduction it is difficult to compare the MBSP of second
39
40 generation biobutanol because of the variety of assumptions made in the cost
41
42 computation. For example in the thermochemical second generation biobutanol study
43
44 21
45
carried out by Okoli and Adams an MBSP of 0.83 $/L is obtained when a catalyst
46
47 once-through CO conversion of 40 % is assumed but 1.56 $/L when the catalyst
48
23 19
49 literature conversion value of 8.5 % is used. In the study by Tao et al. on
50
51 biochemical second generation biobutanol, they obtained an MBSP of 0.78 $/L when
52
53
they assumed a very high butanol yield from sugars of 85 % the theoretical yield.
54
55
56 36
57
58
59
60
1
2
3 However, when they used more realistic values of 55 - 65 % theoretical yield the MBSP
4
5 drops to 1.20 - 1.02 $/L. The advantage of this work over other studies is that it does not
6
7 make futuristic technological assumptions but uses data from current technology that
8
9
10 have either been demonstrated on a pilot scale or commercially, thus the MBSP
11
12 numbers obtained here are a better representation of the current economic potential of
13
14 thermochemical second generation biobutanol. Note though that the MBSP values
15
16 obtained in this work (0.92 - 1.13 $/L for all configurations) are still competitive with
17
18
literature values for second generation biobutanol.
19
20
21
22
When configuration 1a (with DWC) is compared to configuration 1, configuration 1a
23
24 has a slightly higher TCI. This is primarily because the savings obtained in the alcohol
25
26 separation section (reduced capital cost and LP steam usage) are offset by the
27
28 requirement for slightly larger power generating equipment (specifically the LP steam
29
30 turbine). The need for a bigger LP steam turbine arises because the LP steam savings of
31
32
33 the DWC means that extra LP is available for expansion through the steam turbine to
34
35 generate electricity. Overall, the decision to use a DWC in configuration 1a leads to
36
37 only a 1 ¢/L MBSP saving over configuration 1. This is primarily because the
38
39 separation section makes up only a small portion of the capital cost of the
40
41
42
thermochemical biobutanol plant.
43
44
45
46
47
48
49
50
51
52
53
54
55
56 37
57
58
59
60
1
2
3 Table 5. Economic summary of all configurations (a more detailed breakdown is
4
5 provided in the supporting information)
6
7 Plant design Config. 1 - Config. 1a Config. 2 - Config. 3 -
8 Self - Self NG import NG &
9 sufficient sufficient + plant power
10 plant DWC plant import plant
11
Capital Investment ($'000)
12
13 Direct costs breakdown
14 Gasification 47,364 47,364 47,364 47,364
15 Gas cleanup 100,622 100,853 140,527 135,281
16 Mixed Alcohol synthesis 18,079 18,159 41,310 35,470
17 Alcohol separation 9,905 9,270 13,145 12,435
18
Steam system & power gen. 35,692 36,115 61,400 10,306
19
20 Cooling water & other utilities 20,606 20,603 32,909 26,449
21 Total Direct Costs 232,268 232,364 336,654 267,307
22 Engineering and supervision 32,802 32,835 47,731 36,064
23 Construction expenses 34,852 34,887 50,714 38,318
24
Contractor's fee & legal expenses 23,577 23,600 34,307 25,921
25
26 Contingency 20,911 20,932 30,429 22,991
27 Royalties 6,888 6,892 9,997 7,812
28 Land 6,888 6,892 9,997 7,812
29 Working Capital 17,565 17,576 25,492 19,921
30 Total Capital Investment 375,752 375,979 545,320 426,143
31 Operating costs ($'000/year)
32
33 Woody Biomass 52,571 52,571 52,571 52,571
34 Natural gas - - 51,741 48,321
35 Catalysts & chemicals 2,435 2,439 3,310 3,394
36 Waste stream treatment 848 848 1,001 995
37 Water makeup 80 80 162 1,137
38
Electricity import - - - 39,139
39
40 Labour related costs 23,870 23,870 23,870 23,870
41 Maintenance costs 24,040 24,050 34,844 27,666
42 Operating overheads 7,610 7,611 8,948 8,059
43 Property taxes and Insurance 4,645 4,647 6,733 5,346
44 Total Operating Costs 116,098 116,116 183,181 210,499
45
Co-prod. revenues ($'000/year)
46
47 Mixed Alcohols 38,597 36,932 41,523 31,491
48 Electricity export 371 744 2,448 -
49 Total co-prod. revenue 38,969 37,677 43,971 31,491
50 MBSP ($/L) 1.13 1.12 0.92 1.10
51
52
53
54
55
56 38
57
58
59
60
1
2
3 3.4 Cost of CO2 equivalent emissions avoided (CCA)
4
5 The results of the CCA computations are shown in Table 6. Note that it is assumed that
6
7 all carbon in the biomass originated from atmospheric CO2, and thus the biogenic CO2
8
9
10 uptake can be computed from the biomass ultimate analysis as shown in Table 1.
11
12 Furthermore, an energy basis allocation factor which is computed as the fraction of
13
14 butanol product in the total product mix on a HHV basis (see Table 4) is used to allocate
15
16 GHG emissions from the well-to-gate exit emissions of the process to butanol.
17
18
Configuration 1 has the cheapest CCA of 134.65 $/tonneCO2e among all the base case
19
20
21 configurations making it the "greenest" plant. This low value can be attributed to the
22
23 fact that it has zero emissions associated with NG and electricity imports (because they
24
25 do not exist), thus its well-to-wheel emissions are the lowest leading to the highest CCA
26
27 values. In contrast, the GHG emission penalties associated with both NG and electricity
28
29
30 imports in configuration 3 mean that it has the highest CCA.
31
32
33 The reason configuration 2 has a lower CCA in comparison to configuration 3 is
34
35 because configuration 2 has a lower biofuel marginal cost and higher avoided GHG
36
37 emissions than configuration 3. However, despite configuration 2 having a lower
38
39 biofuel marginal cost than configuration 1, its avoided GHG emissions are relatively
40
41
42
much lower meaning that it ends up having a higher CCA than configuration 1.
43
44 Replacing conventional distillation columns with a DWC as is done in configuration 1a
45
46 leads to savings of 3 $/tonneCO2e over configuration 1 which could be significant if
47
48 these technologies are implemented on a large scale.
49
50
51 The CCA of configurations 1 and 2 are very competitive with the range of values
52
53
estimated for other biofuels in the literature. For example, the CCA for European
54
55
56 39
57
58
59
60
1
2
3 biofuels is put at between 277 - 2,524 $/tonneCO2e (Euro converted to U.S. dollars
4
5 39 40
using December 2014 exchange rate) by Ryan et al. while Fulton et al. estimates
6
7 this cost at 180 - 874 $/tonneCO2e for ethanol from different biomass sources. Ethanol
8
9 40
10 from sugarcane in Brazil with a value of around 30 $/tonneCO2e is a notable
11
12 exception because sugarcane crops in Brazil have low costs as a result of very high
13
14 yields, and the sugarcane to ethanol conversion processes used have near zero fossil fuel
15
16 requirements 39,40.
17
18
19 The target value of CCA for GHG emission reduction technologies generally discussed
20
21 40
22
by policy makers in western countries is 50 $/tonneCO2e . Though the technologies
23
24 for biobutanol production studied in this work are higher than this value, it is the
25
26 authors' opinion that this target can be met and even surpassed by these technologies if
27
28 improvements in areas such as the biomass supply chain (to reduce biomass costs),
29
30 biomass-to-butanol processing technology and MAS catalysts are obtained.
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56 40
57
58
59
60
1
2
3 Table 6. Summary of CCA calculations
4
5 Plant Config. 1 - Config. 1a Config. 2 Config. 3
6 Self - Self - NG - NG &
7 sufficient sufficient import power
8 plant + DWC plant import
9 plant plant
10 Biogenic CO2 sequestered during biomass -1,867 -1,867 -1,867 -1,867
11 growth (from ultimate analysis)
12 Well-to-gate GHG emissions for biomass 96 96 96 96
13 wood chips import
14 Biomass to butanol plant emissions (from 1,355 1,356 1,819 1,764
15 simulation results)
16 Well-to-gate GHG emissions for NG use - - 140 131
17 Well-to-gate GHG emissions for - - - 129
18 electricity import
19 Well-to-gate exit emissions (kgCO2e/dry tonne -416 -415 187 253
20 biomass)
21 Well-to-gate exit emissions allocated to butanol -52.18 -51.93 13.41 18.24
(kgCO2e/GJ), product energy basis
22
23 Gate-to-wheel GHG emissions for biobutanol 65.07 65.07 65.07 65.07
(kgCO2e/GJ)
24
Well-to-wheel emission for butanol (kgCO2e/GJ) 12.89 13.14 78.47 83.31
25
26 Avoided GHG emissions (kgCO2e/GJ) [A] 80.73 80.49 15.15 10.32
27 MBSP ($/GJ) 37.13 36.85 30.25 36.21
28 Biofuel marginal cost ($/GJ) [B] 10.87 10.59 4.00 9.95
29 CCA ($/tonneCO2e) [B/A] 134.65 131.59 263.73 964.37
30
31
32 3.5 Sensitivity analysis - one variable
33
34 3.5.1 Sensitivity analysis - impact of gasoline price changes
35
36 The effect of changing gasoline prices on the MBSP and CCA of configurations 1 and 2
37
38
39
are shown in Figure 2 and Figure 3 respectively. The gasoline price is varied between a
40
41 ten-year (January 2005 to December 2014) maximum and minimum historical price
42
41
43 range . The minimum price of gasoline in that time was 0.43 $/L (December 2008),
44
45 while its maximum price was 1.08 $/L (June 2008). Both Figure 2 and Figure 3 show
46
47 that changes in gasoline price have an inverse effect on MBSP and CCA for the
48
49
50 configurations assessed. This is because increasing the gasoline price increases the
51
52 revenue from selling mixed alcohols and thus lowers the revenue that is needed from
53
54 selling biobutanol to get a zero NPV, meaning that the MBSP and CCA are lowered.
55
56 41
57
58
59
60
1
2
3 The opposite effect also holds true when gasoline prices are reduced. As can be seen
4
5 from the slopes of the lines in Figure 2, changing the gasoline price has slightly more
6
7 effect on the MBSP of configuration 1 than configuration 2. This is because the mixed
8
9
10 alcohols co-product, whose pricing is directly correlated to gasoline price, contributes a
11
12 higher percentage to the total product yield of configuration 1 than configuration 2 (see
13
14 Table 4), thus changes in gasoline price affect configuration 1 more than configuration
15
16 2. However, the much smaller avoided GHG emissions value of configuration 2 (see
17
18
Table 6) means that changes in gasoline price have a larger effect on its CCA in
19
20
21 comparison to configuration 1 as is shown in Figure 3. Also shown in Figure 3 is the
22
23 price of gasoline above which the CCA becomes lower than the 50 $/tonneCO2e
24
25 benchmark. This corresponds to approximately 1 $/L for both configurations 1 and 2.
26
27
28
29
30
1.30
31
32
33
34
MBSP ($/L)
35
1.10
36
37
38
39
40
0.90
41
42
43
44
45
0.70
46
47 0.40 0.60 0.80 1.00 1.20
48 U.S. average retail gasoline price ($/L)
49
50
51
52
53
54
55
56 42
57
58
59
60
1
2
3 Figure 2. Impact of changes in gasoline price ($/L) on MBSP ($/L). Symbols:
4
5 maximum gasoline price [••], minimum gasoline price [∎], base case gasoline price [▲
6
7
8
].
9
10
11 1600
Cost of CO2e avoided ($/tonne CO2 ea)
12
13
14 1200
15
16
17
18 800
19
20
21 400
22
23
24
25 0
26
27
28
-400
29
30 0.40 0.60 0.80 1.00 1.20
31 U.S. average retail gasoline price ($/L)
32
33
Figure 3. Impact of changes in gasoline price ($/L) on CCA ($/tonneCO2e). Symbols:
34
35
36 maximum gasoline price [••], minimum gasoline price [∎], base case gasoline price [▲
37
38 ].
39
40
41 3.5.2 Sensitivity analysis - impact of natural gas price changes
42
43 Figure 4 and Figure 5 respectively show the effect of changing NG prices on the MBSP
44
45
46
and CCA of configurations 1 and 2. Similar to the sensitivity analysis on gasoline price,
47
48 the NG price is varied between a ten-year (January 2005 to December 2014) maximum
49
42
50 and minimum historical price range . The minimum price of NG in that time was
51
52 141.98 $/tonne (May 2012), while its maximum price was 613.97 $/tonne (July 2008).
53
54 As NG is an operating cost item, the MBSP and CCA are directly correlated to it.
55
56 43
57
58
59
60
1
2
3 However, note that changes in NG have no impact on configuration 1 as that plant
4
5 makes no use of NG. It can be seen from Figure 4 that at the maximum ten-year NG
6
7 historical price, the MBSP of configuration 2 is 1.31 $/L while it is 0.82 $/L at the
8
9
10 minimum NG historical price. At the same maximum and minimum NG prices, the
11
12 CCA is 942.33 $/tonneCO2e and 39.33 $/tonneCO2e respectively for configuration 2 as
13
14 is shown in Figure 5. Furthermore, Figure 5 shows that with an NG price below 153.13
15
16 $/tonne, the CCA is lower than the 50 $/tonneCO2e benchmark.
17
18
19
20
21
22 1.3
23
24
25
MBSP ($/L)
26
27 1.1
28
29
30
31
32 0.9
33
34
35
36
37 0.7
38 100 200 300 400 500 600 700
39
40 U.S. average industrial NG price ($/tonne)
41
42
43 Figure 4. Impact of changes in NG price ($/tonne) on MBSP ($/L). Symbols: maximum
44
45 NG price [•], minimum NG price [∎], base case NG price [▲ ].
46
47
48
49
50
51
52
53
54
55
56 44
57
58
59
60
1
2
3
4
1100
Cost of CO2 e avoided ($/tonne CO2 ea)

5
6
7
8 800
9
10
11
12
500
13
14
15
16
17 200
18
19
20
21
-100
22
23 100 200 300 400 500 600 700
24 U.S. average industrial NG price ($/tonne)
25
26 Figure 5. Impact of changes in NG price ($/tonne) on CCA ($/tonneCO2e). Symbols:
27
28
29
maximum NG price [•], minimum NG price [∎], base case NG price [▲ ].
30
31
32 3.5.3 Sensitivity analysis - impact of biomass price changes
33
34 The impact of changes in biomass price on the MBSP and CCA of configurations 1 and
35
36 2 were assessed with the results presented in Figure 6 and Figure 7 respectively. As can
37
38 be seen in Figure 6 the MBSP of configuration 1 is more sensitive to biomass price
39
40
41
changes than that of configuration 2. This is because in configuration 1, biomass cost
42
43 makes up a greater percentage of the TOC unlike in configuration 2 in which it has a
44
45 smaller percentage, primarily because of the additional significant cost of NG.
46
47 However, as shown in Figure 7 the CCA of configuration 2 is more sensitive to biomass
48
49
price changes than that of configuration 1. As previously mentioned for gasoline price
50
51
52 changes, this is because of the much smaller avoided GHG emissions value of
53
54 configuration 2 in comparison to configuration 1 (see Table 6). Furthermore, Figure 7
55
56 45
57
58
59
60
1
2
3 shows that a minimum 41 % reduction in biomass price from the base case value of 75
4
5 $/tonne to 44 $/tonne is needed for both configurations before their CCA is below the
6
7 policy threshold of 50 $/tonneCO2e.
8
9
10
11 1.30
12
13
14
15 1.20
16
17
MBSP ($/L)
18 1.10
19
20
21
22 1.00
23
24
25
26 0.90
27
28
29 0.80
30
31
40 50 60 70 80 90 100
32 Biomass price ($/tonne)
33
34 Figure 6. Impact of changes in biomass price ($/tonne) on MBSP ($/L). Symbol: base
35
36
37
case biomass price [▲ ].
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56 46
57
58
59
60
1
2
3
4
480
Cost of CO2 e avoided ($/tonne CO2 ea

5
6
7
8 360
9
10
11
12
13
240
14
15
16
17 120
18
19
20
21
22 0
23 40 50 60 70 80 90 100
24
25
Biomass price ($/tonne)
26 Figure 7. Impact of changes in biomass price ($/tonne) on CCA ($/tonneCO2e).
27
28
Symbol: base case biomass price [▲ ].
29
30
31
32 3.6 Sensitivity analysis - two variable
33
34 In this sensitivity analysis the impact of simultaneously changing a combination of two
35
36 decision variables from a selection of the gasoline price, NG price and biomass price on
37
38 the CCA and MBSP of configurations 1 and 2 are considered. The results of three
39
40
41 selected cases are presented as "colour maps" illustrating the optimal configurations.
42
43 Similar to the one variable sensitivity analysis, ten year historical data between January
44
45 2005 and December 2014 are used as the range for energy prices (gasoline and NG),
46
47 while the biomass price is varied between +/- 30% of its base value.
48
49
50 Figure 8 and Figure 9 illustrate the impacts of simultaneously changing gasoline and
51
52
NG prices on the MBSP and CCA respectively. The values of MBSP and CCA in both
53
54
55 figures are shown as contour lines. Also annotated on the figures are points which
56 47
57
58
59
60
1
2
3 represent real energy price conditions in the historical range considered. Figure 8 shows
4
5 that the MBSP of the biomass and NG import plant (configuration 2) is lower than the
6
7 biomass only plant (configuration 1) in the majority of scenarios, and especially for
8
9
10 lower NG prices. As can be seen from the figure, the biomass only plant is favoured in
11
12 all scenarios when the price of NG is above 600 $/tonne. However, the picture is
13
14 different for CCA as shown in Figure 9. In Figure 9, the CCA of the biomass only plant
15
16 is lower in most scenarios (about 85%), with the biomass & NG import plant favoured
17
18
for high gasoline price and low NG price scenarios. It is also interesting to note that
19
20
21 points [c], [e] and the base case [▲] fall in different regions in Figure 8 and Figure 9,
22
23 meaning that for those points the plant choice depends on what objective is of interest.
24
25 If low MBSP is the objective, the biomass & NG import plant is preferred, while the
26
27 biomass only plant is preferred if the objective is low CCA.
28
29
30 In Figure 10, the impact on CCA (shown as contour lines) of simultaneously varying
31
32
33 biomass and gasoline prices is shown. As can be seen from the figure, the optimal CCA
34
35 is provided by the biomass only plant in majority of the cases. However, at high
36
37 gasoline prices the biomass & NG import plant starts being preferred. Also note from
38
39 this figure that the biomass & NG import plant is preferred in almost all scenarios when
40
41
42
the CCA is below 50 $/tonneCO2e.
43
44
45
46
47
48
49
50
51
52
53
54
55
56 48
57
58
59
60
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25 Figure 8. Optimal plant choice for lowest MBSP, considering variations in gasoline and
26
27 NG prices only. Symbols: maximum gasoline price - Jun 2008 [a], maximum NG price
28
29 - Jul 2008 [b], minimum gasoline price - Dec 2008 [c], minimum NG price - May 2012
30
31 [d], energy prices - Dec 2014 [e], base case [▲].
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56 49
57
58
59
60
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25 Figure 9. Optimal plant choice for lowest CCA, considering variations in gasoline and
26
27 NG prices only. Symbols: maximum gasoline price - Jun 2008 [a], maximum NG price
28
29 - Jul 2008 [b], minimum gasoline price - Dec 2008 [c], minimum NG price - May 2012
30
31 [d], energy prices - Dec 2014 [e], base case [▲].
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56 50
57
58
59
60
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Figure 10. Optimal plant choice for lowest CCA, considering variations in gasoline and
26
27 biomass prices only. Symbol: base case [▲].
28
29
30 4. Conclusion
31
32 In this work, novel configurations for producing second generation biobutanol through a
33
34 thermochemical route and MAS process are designed and assessed based on economic
35
36
37 and environmental metrics. The different configurations vary based on choices
38
39 regarding the use of NG combustion for high temperature heat, import of electricity to
40
41 meet plant demands, and the use of DWCs for alcohols separation.
42
43
44 The economic results showed that the NG import configuration has the lowest MBSP at
45
46 0.92 $/L. This is because this configuration avoids the high operating costs associated
47
48
49
with electricity import and has a high butanol yield because syngas is not diverted for
50
51 meeting high temperature heat requirements. However the environmental results show
52
53
54
55
56 51
57
58
59
60
1
2
3 that the self-sufficient configuration has the lowest CCA of 134.65 $/tonneCO2e among
4
5 the base configurations.
6
7
8 The use of DWC technology over conventional columns in the alcohols separation
9
10 section leads to CCA savings of 3 $/tonneCO2e, which could be significant if these
11
12
13 technologies are implemented on a large scale. Furthermore, one variable sensitivity
14
15 analysis showed that the CCA of the NG import configuration is more sensitive than the
16
17 self-sufficient configuration to changes in gasoline, NG and biomass prices, while two
18
19 variable sensitivity analysis on energy prices show that the optimal plant selection
20
21
22
depends on whether the objective is to have a lower MBSP or CCA.
23
24
25
Finally, the economic results show that the assessed thermochemical plants produce
26
27 second generation biobutanol that are competitive with second generation biobutanol
28
29 produced through a biochemical route.
30
31
32 Supporting Information
33
34
35 Additional supporting information as noted in the text. This information is available free
36
37 of charge via the Internet at http://pubs.acs.org/.
38
39
40 Parameter estimation study to improve mixed alcohols synthesis kinetic model
41
42
43
predictions of alcohols production (pdf); tables of reference data (Tables S6 - S9) used
44
45 in the manuscript (pdf); stream conditions for Figure 1 for all configurations (excel
46
47 workbook); equipment cost breakdown for all configurations (excel workbook);
48
49 DCFROR analysis for all configurations (excel workbook)
50
51
52
53
54
55
56 52
57
58
59
60
1
2
3 Author Information
4
5 *
Corresponding author. Department of Chemical Engineering, McMaster University,
6
7 1280 Main Street West, Hamilton, Ontario, Canada; Email: tadams@mcmaster.ca; Ph:
8
9
10 (905) 525-9140 x24782.
11
12
13 Acknowledgement
14
15 This work was funded by an Ontario Research Fund – Research Excellence Grant (ORF
16
17 RE-05-072).
18
19
20 Abbreviations
21
22 ABE Acetone-butanol-ethanol
23
24 ASTM American society for testing and materials
25
26 CCA Cost of CO2 equivalent emissions avoided
27 CIB Carbon intensity of biobutanol
28
29 CIG Carbon intensity of gasoline
30
31 CO2e CO2 equivalent emissions
32 DCFROR Discounted cash flow rate of return
33
34 DWC Dividing wall columns
35
36 GHG Greenhouse gas
37 GHSV Gas hourly space velocity
38
39 HHV Higher heating value
40
41 HP High pressure
42 IRR Internal rate of return
43
44 LP Low pressure
45
46 MAS Mixed alcohol synthesis
47 MBSP Minimum butanol selling price
48
49 NG Natural gas
50
51 NPV Net present value
52 NREL National renewable energy laboratory
53
54 PSO Particle swarm optimization
55
56 53
57
58
59
60
1
2
3 RFS Renewable fuel standard
4
5 TAC Total annualized cost
6 TCI Total capital investment
7
8 TOC Total operating costs
9
10 VBA Visual Basic for Applications
11 VOC Variable operating cost
12
13
14 References
15
16 (1) Annual Energy Outlook 2014 with Projections to 2040; Report DOE/EIA-
17 0383(2014); U.S. Energy Information Administration: Washington, DC, Apr
18 2014; available at http://www.eia.gov/forecasts/aeo/pdf/0383(2014).pdf
19
(accessed March 20, 2016).
20
21 (2) U.S. Energy Independence and Security Act of 2007. Public Law 2007, 110–140.
22
23 (3) Perlack, R. D.; Wright, L. L.; Turhollow, A. F.; Graham, R. L.; Stokes, B. J.;
24 Erbach, D. C. Biomass as feedstock for a bioenergy and bioproducts industry: the
25
technical feasibility of a billion-ton annual supply; Oak Ridge National
26
27 Laboratory: Oak Ridge, TN, April 2005.
28
29
(4) Kumar, M.; Gayen, K. Developments in biobutanol production: New insights.
30 Appl. Energy 2011, 88 (6), 1999–2012.
31
32 (5) Ranjan, A.; Moholkar, V. S. Biobutanol: science, engineering, and economics.
33 Int. J. Energy Resour. 2012, 36 (3), 277–323.
34
35 (6) Green, E. M. Fermentative production of butanol—the industrial perspective.
36 Curr. Opin. Biotechnol. 2011, 22 (3), 337–343.
37
38 (7) García, V.; Päkkilä, J.; Ojamo, H.; Muurinen, E.; Keiski, R. L. Challenges in
39 biobutanol production: How to improve the efficiency? Renew. Sustain. Energy
40 Rev. 2011, 15 (2), 964–980.
41
42 (8) Dutta, A.; Talmadge, M.; Hensley, J.; Worley, M.; Dudgeon, D.; Barton, D.;
43 Groenedijk, P.; Ferrari, D.; Stears, B.; Searcy, E. M.; Wright, C. T.; Hess, J. R.
44 Process Design and Economics for Conversion of Lignocellulosic Biomass to
45 Ethanol; National Renewable Energy Laboratory: Golden, CO, May 2011.
46
47 (9) Kraemer, K.; Harwardt, A.; Bronneberg, R.; Marquardt, W. Separation of butanol
48 from acetone–butanol–ethanol fermentation by a hybrid extraction–distillation
49 process. Comput. Chem. Eng. 2011, 35 (5), 949–963.
50
51 (10) Abdehagh, N.; Tezel, F. H.; Thibault, J. Separation techniques in butanol
52 production: Challenges and developments. Biomass and Bioenergy 2014, 60,
53 222–246.
54
55
56 54
57
58
59
60
1
2
3 (11) Ndaba, B.; Chiyanzu, I.; Marx, S. n-Butanol derived from biochemical and
4 chemical routes: A review. Biotechnol. Reports 2015, 8, 1–9.
5
6 (12) Lee, S. Y.; Park, J. H.; Jang, S. H.; Nielsen, L. K.; Kim, J.; Jung, K. S.
7 Fermentative butanol production by clostridia. Biotechnol. Bioeng. 2008, 101 (2),
8 209–228.
9
10 (13) Kazi, F. K.; Fortman, J.; Anex, R.; Kothandaraman, G.; Hsu, D.; Aden, A.; Dutta,
11 A. Techno-economic analysis of biochemical scenarios for production of
12 cellulosic ethanol; National Renewable Energy Laboratory: Golden, CO, June
13
2010.
14
15 (14) Surisetty, V. R.; Dalai, A. K.; Kozinski, J. Intrinsic Reaction Kinetics of Higher
16
Alcohol Synthesis from Synthesis Gas over a Sulfided Alkali-Promoted
17
18 Co−Rh−Mo Trimetallic Catalyst Supported on Multiwalled Carbon Nanotubes
19 (MWCNTs). Energy & Fuels 2010, 24 (8), 4130–4137.
20
21 (15) Nexant. Equipment Design and Cost Estimation for Small Modular Biomass
22 Systems, Synthesis Gas Cleanup, and Oxygen Separation Equipment Task 9:
23 Mixed Alcohols From Syngas — State of Technology; National Renewable
24 Energy Laboratory: Golden, CO, 2006.
25
26 (16) Fang, K.; Li, D.; Lin, M.; Xiang, M.; Wei, W.; Sun, Y. A short review of
27 heterogeneous catalytic process for mixed alcohols synthesis via syngas. Catal.
28 Today 2009, 147 (2), 133–138.
29
30 (17) Surisetty, V. R.; Dalai, A. K.; Kozinski, J. Alcohols as alternative fuels: An
31 overview. Appl. Catal. A Gen. 2011, 404 (1), 1–11.
32
33 (18) Qureshi, N.; Saha, B. C.; Cotta, M. A.; Singh, V. An economic evaluation of
34 biological conversion of wheat straw to butanol: A biofuel. Energy Convers.
35 Manag. 2013, 65, 456–462.
36
37 (19) Tao, L.; Tan, E. C. D.; McCormick, R.; Zhang, M.; Aden, A.; He, X.; Zigler, B.
38 T. Techno-economic analysis and life-cycle assessment of cellulosic isobutanol
39 and comparison with cellulosic ethanol and n-butanol. Biofuels, Bioprod.
40
Biorefining 2014, 8 (1), 30–48.
41
42 (20) Kumar, M.; Goyal, Y.; Sarkar, A.; Gayen, K. Comparative economic assessment
43
44
of ABE fermentation based on cellulosic and non-cellulosic feedstocks. Appl.
45 Energy 2012, 93, 193–204.
46
47
(21) Okoli, C.; Adams, T. A. Design and economic analysis of a thermochemical
48 lignocellulosic biomass-to-butanol process. Ind. Eng. Chem. Res. 2014, 53 (28),
49 11427–11441.
50
51 (22) Gapes, J. R. The Economics of Acetone-Butanol Fermentation: Theoretical and
52 Market Considerations. J. Mol. Microbiol. Biotechnol. 2000, 2 (1), 27–32.
53
54 (23) Herman, R. Advances in catalytic synthesis and utilization of higher alcohols.
55 Catal. Today 2000, 55 (3), 233–245.
56 55
57
58
59
60
1
2
3 (24) Phillips, S.; Aden, A.; Jechura, J.; Dayton, D.; Eggeman, T. Thermochemical
4 Ethanol via Indirect Gasification and Mixed Alcohols Synthesis o
5 Lignocellulosic Biomass; National Renewable Energy Laboratory: Golden, CO,
6 April 2007.
7
8 (25) Beretta, A.; Micheli, E.; Tagliabue, L.; Tronconi, E. Development of a Process
9 for Higher Alcohol Production via Synthesis Gas. Ind. Eng. Chem Res. 1998,
10 5885 (97), 3896–3908.
11
12 (26) Yildirim, Ö.; Kiss, A. a.; Kenig, E. Y. Dividing wall columns in chemical process
13
industry: A review on current activities. Sep. Purif. Technol. 2011, 80 (3), 403–
14
15
417.
16
(27) Okoli, C. O.; Adams, T. A. Design of dividing wall columns for butanol recovery
17
18 in a thermochemical biomass to butanol process. Chem. Eng. Process. Process
19 Intensif. 2015, 95, 302–316.
20
21 (28) Zhu, Y.; Gerber, M. A.; Jones, S. B.; Stevens, D. J. Analysis of the Effects of
22 Compositional and Configurational Assumptions on Product Costs for the
23 Thermochemical Conversion of Lignocellulosic Biomass to Mixed Alcohols - FY
24 2007 Progress Report; Pacific Northwest National Laboratory: Richland, WA,
25 February 2009.
26
27 (29) Cohen, A. P.; Reynolds, T. M.; Davis, M. M. Adsorbent for Drying Ethanol. US
28 2010/0081851 A1, 2012.
29
30 (30) ASTM Standard D7862-15. Standard Specification for Butanol for Blending with
31 Gasoline for Use as Automotive Spark-Ignition Engine Fuel; ASTM
32 International: West Conshohocken, PA 2015.
33
34 (31) Hicks, T. G.; Wills, K. D. Handbook of Mechanical Engineering Calculations;
35 McGraw-Hill: New York, 2006.
36
37 (32) Perry, R. H.; Green, D. W. Perry’s Chemical Engineers' Handbook, 7th ed.;
38 McGraw-Hill: New York, 1997.
39
40 (33) CE. Economic indicators. Chem. Eng. 2015, 122 (12), 192.
41
42 (34) Seider, W. D.; Seader, J. D.; Lewin, D. R. Product & Process Design Principles:
43 Synthesis, Analysis and Evaluation, 3rd ed.; John Wiley & Sons: Hoboken, New
44 Jersey, 2009.
45
46 (35) Joseph, K.; Conforti, L.; Hergt, B. PPI Detailed Report Data for December 2015.
47 U.S. Bureau of Labor Statistics 2015, 19 (12).
48
49 (36) IPCC Fourth Assessment Report: Climate Change 2007; available at
50 https://www.ipcc.ch/publications_and_data/ar4/wg1/en/ch2s2-10-2.html
51
52
(accessed Apr 19, 2016).
53
(37) Gendreau, M.; Potvin, J.-Y. Handbook of Metaheuristics Vol. 2; Springer: New
54
55
York, 2010.
56 56
57
58
59
60
1
2
3 (38) Kaveh, A. Advances in Metaheuristic Algorithms for Optimal Design of
4 Structures; Springer: New York, 2014.
5
6 (39) Ryan, L.; Convery, F.; Ferreira, S. Stimulating the use of biofuels in the
7 European Union: Implications for climate change policy. Energy Policy 2006, 34
8 (17), 3184–3194.
9
10 (40) Fulton, L.; Howes, T.; Hardy, J. Biofuels for transport: An international
11 perspective; International Energy Agency: Paris, 2004.
12
13 (41) Petroleum & Other Liquids - U.S. All Grades All Formulations Retail Gasoline
14 Prices; U.S Energy Information Administration: Washington, DC, 2016;
15 available at
16
http://www.eia.gov/dnav/pet/hist/LeafHandler.ashx?n=PET&s=EMM_EPM0_PT
17
18 E_NUS_DPG&f=A (accessed Jan 31, 2016).
19
(42) United States Natural Gas Industrial Price; U.S Energy Information
20
21 Administration: Washington, DC, 2016; available at
22
23 https://www.eia.gov/dnav/ng/hist/n3035us3m.htm (accessed Jan 31, 2016).
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56 57
57
58
59
60
1
2
3
4
5
6
7
Regenerated
8
catalyst
catalyst
Used
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
For Table of Contents Only
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56 58
57
58
59
60

Articulo Procesos

Cargado por

Información del documento

Derechos de autor

Formatos disponibles

Compartir este documento

Compartir o incrustar documentos

Opciones para compartir

¿Le pareció útil este documento?

¿Este contenido es inapropiado?

Copyright:

Formatos disponibles

Articulo Procesos

Cargado por

Copyright:

Formatos disponibles

Subscriber access provided by UB + Fachbibliothek Chemie | (FU-Bibliothekssystem)

Industrial & Engineering Chemistry Research is published by the American Chemical

Cost of CO2 e avoided ($/tonne CO2 ea)

Cost of CO2 e avoided ($/tonne CO2 ea

También podría gustarte