Process Design and Economics

for the Conversion of

Lignocellulosic Biomass to
Hydrocarbons:
Dilute-Acid and Enzymatic
Deconstruction of Biomass to Sugars
and Catalytic Conversion of Sugars to
Hydrocarbons
R. Davis, L. Tao, C. Scarlata, and E.C.D. Tan
National Renewable Energy Laboratory

J. Ross, J. Lukas, and D. Sexton

Harris Group Inc.

NREL is a national laboratory of the U.S. Department of Energy

Office of Energy Efficiency & Renewable Energy
Operated by the Alliance for Sustainable Energy, LLC
This report is available at no cost from the National Renewable Energy
Laboratory (NREL) at www.nrel.gov/publications.

Technical Report
NREL/TP-5100-62498
March 2015

Contract No. DE-AC36-08GO28308

 Process Design and
Economics for the Conversion
of Lignocellulosic Biomass to
Hydrocarbons:
Dilute-Acid and Enzymatic
Deconstruction of Biomass to Sugars
and Catalytic Conversion of Sugars to
Hydrocarbons
R. Davis, L. Tao, C. Scarlata, and E.C.D. Tan
National Renewable Energy Laboratory

J. Ross, J. Lukas, and D. Sexton

Harris Group Inc.
Prepared under Task No. BB14.2410

NREL is a national laboratory of the U.S. Department of Energy

Office of Energy Efficiency & Renewable Energy
Operated by the Alliance for Sustainable Energy, LLC
This report is available at no cost from the National Renewable Energy
Laboratory (NREL) at www.nrel.gov/publications.

National Renewable Energy Laboratory Technical Report

15013 Denver West Parkway NREL/TP-5100-62498
Golden, CO 80401 March 2015
303-275-3000 www.nrel.gov
Contract No. DE-AC36-08GO28308
 NOTICE

This report was prepared as an account of work sponsored by an agency of the United States government.
Neither the United States government nor any agency thereof, nor any of their employees, makes any warranty,
express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of
any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately
owned rights. Reference herein to any specific commercial product, process, or service by trade name,
trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation,
or favoring by the United States government or any agency thereof. The views and opinions of authors
expressed herein do not necessarily state or reflect those of the United States government or any agency thereof.

This report is available at no cost from the National Renewable Energy

Laboratory (NREL) at www.nrel.gov/publications.

Available electronically at http://www.osti.gov/scitech

Available for a processing fee to U.S. Department of Energy

and its contractors, in paper, from:

U.S. Department of Energy

Office of Scientific and Technical Information
P.O. Box 62
Oak Ridge, TN 37831-0062
phone: 865.576.8401
fax: 865.576.5728
email: mailto:[email protected]

Available for sale to the public, in paper, from:

U.S. Department of Commerce

National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
phone: 800.553.6847
fax: 703.605.6900
email: [email protected]
online ordering: http://www.ntis.gov/help/ordermethods.aspx

Cover Photos: (left to right) photo by Pat Corkery, NREL 16416, photo from SunEdison, NREL 17423, photo by Pat Corkery, NREL
16560, photo by Dennis Schroeder, NREL 17613, photo by Dean Armstrong, NREL 17436, photo by Pat Corkery, NREL 17721.

NREL prints on paper that contains recycled content.

 Executive Summary
The U.S. Department of Energy (DOE) promotes the production of an array of liquid fuels and
fuel blendstocks from lignocellulosic biomass feedstocks by funding fundamental and applied
research that advances the state of technology in biomass collection, conversion, and
sustainability. As part of its involvement in this program, the National Renewable Energy
Laboratory (NREL) investigates the conceptual production economics of these fuels.

Over the past decade, NREL conducted a campaign to quantify the economic implications
associated with observed and future targeted performance for the biochemical conversion of corn
stover to ethanol through techno-economic modeling. This effort served to set state of
technology benchmarks and to guide research and development by setting cost targets and
tracking progress towards final achievement of these targets in 2012. Beginning in 2013, NREL
began transitioning from the singular focus on ethanol to a broad slate of products and
conversion pathways, ultimately to establish similar benchmarking and targeting efforts. The
first product from this expanded focus was a technical report documenting a conceptual process
pathway for biological (i.e., microbial) conversion of cellulosic sugars to hydrocarbon fuels. The
analysis in this design report expands from this initial baseline to consider an approach via
catalytic conversion of sugars and other hydrolysate components to hydrocarbons.

This report describes one potential conversion process to hydrocarbon products by way of
catalytic conversion of lignocellulosic-derived hydrolysate. This model leverages expertise
established over time in biomass deconstruction and process integration research at NREL, while
adding in new technology areas for sugar purification and catalysis. The overarching process
design converts biomass to diesel- and naphtha-range fuels using dilute-acid pretreatment,
enzymatic saccharification, purifications, and catalytic conversion focused on deoxygenating and
oligomerizing biomass hydrolysates. Ancillary areasfeed handling, hydrolysate conditioning
and upgrading (hydrotreating) to final fuel products, wastewater treatment, lignin combustion,
and utilitiesare also included in the design. Detailed material and energy balances and capital
and operating costs for this baseline process are also documented.

This techno-economic analysis models a production cost for cellulosic hydrocarbon biofuels that
can be considered as a baseline to assess the competitiveness and market potential for the
technology. It can also be used to quantify the economic impact of individual conversion
performance targets and prioritize them in terms of their potential to reduce cost. The analysis
presented here also includes consideration of the life-cycle implications of the process model by
tracking sustainability metrics for the modeled biorefinery, including greenhouse gas emissions,
fossil energy demand, and water consumption.

Building on prior design report practices, NREL, together with Harris Group, performed a
feasibility-level analysis for a plausible catalytic conversion process to meet an intermediate
DOE cost goal of $5/gallon gasoline equivalent (GGE) in a timeframe prior to 2022. The
modeled biorefinery processes 2,205 dry tons biomass/day and achieves a fuel selling price of
$4.05/GGE in 2011-year dollars as determined by modeled conversion targets and nth-plant
project costs and financing, associated with a total fuel yield of 78.3 GGE/dry ton when
assuming hydrogen is purchased from off-site production. Alternatively, if hydrogen is instead
produced in situ by diverting a fraction (41%) of hydrolysate carbon towards reforming

iii
This report is available at no cost from the National Renewable Energy Laboratory (NREL) at www.nrel.gov/publications.
reactions, fuel yield drops to 45.3 GGE/dry ton and selling price increases to $5.48/GGE. If
instead hydrogen is produced by diverting a fraction of biomass to a gasification process to
produce and subsequently refine syngas to hydrogen, such a scenario would require 36% of the
available biomass and would translate to a fuel yield of 50.1 GGE/dry ton and selling price of
$4.95/GGE. Preliminary assessment of sustainability metrics associated with these three
hydrogen scenarios indicates a more favorable greenhouse gas profile for the in situ and
gasification cases as it avoids the use of natural gas for off-site hydrogen production. However,
additional benefits are likely to be seen for a case including on-site hydrogen production via
steam methane reforming (rather than off-site as in the base case evaluated here), given
additional integration advantages that may be brought about through heat/power integration and
carbon recycles which were beyond the scope of this effort; in any case, a full well-to-wheel life-
cycle analysis is required to better understand greenhouse gas implications for the fully
integrated technology pathway. Finally, the report includes a brief discussion on improvements
needed to achieve a final 2022 DOE target of $3/GGE moving forward, focused on coproducts
from lignin.

Figure ES-1. Minimum fuel selling price results for external (purchased) hydrogen base design
versus internal hydrogen production via in situ reforming of hydrolysate and biomass gasification

iv
Figure ES-2. Economic summary for catalytic conversion base case process (purchased H2)

v
Table of Contents
Executive Summary ................................................................................................................................... iii
Acronyms .................................................................................................................................................. viii
1 Introduction ........................................................................................................................................... 1
1.1 Background and Motivation .......................................................................................................... 1
1.2 Process Overview .......................................................................................................................... 2
1.3 Techno-Economic Analysis Approach .......................................................................................... 5
1.4 About nth-Plant Assumptions......................................................................................................... 6
1.5 About the NREL Aspen Model ..................................................................................................... 7
2 Design Basis and Conventions ........................................................................................................... 8
2.1 Plant Size ....................................................................................................................................... 8
2.2 Feedstock Composition ................................................................................................................. 8
2.3 Updated Modeling Basis ............................................................................................................... 9
2.4 Design Report Conventions......................................................................................................... 10
2.4.1 Units ............................................................................................................................... 10
2.4.2 Total Solids Loading ...................................................................................................... 10
2.4.3 Product Yield and Heating Value................................................................................... 10
3 Process Design and Cost Estimation Details .................................................................................. 11
3.1 Area 100: Feedstock Storage and Handling ................................................................................ 11
3.1.1 Overview ........................................................................................................................ 11
3.1.2 Design Basis ................................................................................................................... 12
3.1.3 Cost Estimation .............................................................................................................. 13
3.2 Area 200: Pretreatment and Conditioning ................................................................................... 13
3.2.1 Overview ........................................................................................................................ 13
3.2.2 Design Basis ................................................................................................................... 14
3.2.3 Cost Estimation .............................................................................................................. 17
3.3 Area 300: Enzymatic Hydrolysis, Hydrolysate Conditioning, and Purification ......................... 18
3.3.1 Overview ........................................................................................................................ 18
3.3.2 Design Basis ................................................................................................................... 19
3.3.3 Cost Estimation .............................................................................................................. 24
3.4 Area 400: Cellulase Enzyme Production..................................................................................... 25
3.4.1 Overview ........................................................................................................................ 25
3.4.2 Design Basis ................................................................................................................... 26
3.4.3 Cost Estimation .............................................................................................................. 28
3.5 Area 500: Catalytic Conversion and Upgrading ......................................................................... 29
3.5.1 Overview ........................................................................................................................ 29
3.5.2 Design Basis ................................................................................................................... 30
Hydrogenation and Aqueous Phase Reforming ........................................................................ 32
Condensation ............................................................................................................................ 36
Hydrotreating ............................................................................................................................ 38
3.5.3 Cost Estimation .............................................................................................................. 40
3.6 Area 600: Wastewater Treatment ................................................................................................ 43
3.6.1 Overview ........................................................................................................................ 43
3.6.2 Design Basis ................................................................................................................... 44
3.6.3 Cost Estimation .............................................................................................................. 45
3.7 Area 700: Product and Feed Chemical Storage........................................................................... 46
3.7.1 Overview ........................................................................................................................ 46
3.7.2 Design Basis ................................................................................................................... 46
3.7.3 Cost Estimation .............................................................................................................. 46
3.8 Area 800: Combustor, Boiler, and Turbogenerator ..................................................................... 47

vi
 3.8.1 Overview ........................................................................................................................ 47
3.8.2 Design Basis ................................................................................................................... 47
3.8.3 Cost Estimation .............................................................................................................. 48
3.9 Area 900: Utilities ....................................................................................................................... 48
3.9.1 Overview ........................................................................................................................ 48
3.9.2 Design Basis ................................................................................................................... 49
3.9.3 Cost Estimation .............................................................................................................. 51
4 Process Economics ........................................................................................................................... 52
4.1 About Cost-Year Indices ............................................................................................................. 52
4.2 Total Capital Investment ............................................................................................................. 53
4.3 Variable Operating Costs ............................................................................................................ 55
4.4 Fixed Operating Costs ................................................................................................................. 57
4.5 Discounted Cash Flow Analysis and the Minimum Selling Price of Fuel .................................. 58
4.5.1 Discount Rate ................................................................................................................. 58
4.5.2 Equity Financing ............................................................................................................ 59
4.5.3 Depreciation ................................................................................................................... 59
4.5.4 Taxes .............................................................................................................................. 59
4.5.5 Construction Time .......................................................................................................... 60
4.5.6 Start-Up Time................................................................................................................. 60
4.5.7 Working Capital ............................................................................................................. 60
5 Analysis and Discussion ................................................................................................................... 64
5.1 Carbon and Water Balances ........................................................................................................ 64
5.2 Cost Sensitivity Analysis ............................................................................................................ 67
5.2.1 Single-Point Sensitivity Analysis ................................................................................... 67
5.2.2 Evaluation of Alternative Hydrogen Sourcing Options ................................................. 70
In Situ Hydrogen Production .................................................................................................... 70
Biomass Gasification Hydrogen Production............................................................................. 72
Model Results ........................................................................................................................... 74
5.3 Sustainability Metrics .................................................................................................................. 78
5.4 Paths Forward to $3/GGE ........................................................................................................... 83
6 Concluding Remarks .......................................................................................................................... 86
6.1 Summary ..................................................................................................................................... 86
6.2 Future Work ................................................................................................................................ 88
References ................................................................................................................................................. 90
Appendix A. Individual Equipment Costs Summary ............................................................................. 96
Appendix B. Discounted Cash Flow Rate of Return Worksheet ........................................................ 103
Appendix C. Process Parameters/Operating Summary ...................................................................... 105
Appendix D. Aspen Properties .............................................................................................................. 106
Appendix E. Process Flow Diagrams.................................................................................................... 111

vii
Acronyms
APR aqueous phase reforming MFSP minimum fuel selling price
BETO Bioenergy Technologies MM million
Office MVR mechanical vapor
BFW boiler feed water recompression
CIP clean-in-place MYPP BETOs Multi-Year Program
COD chemical oxygen demand Plan
CSL corn steep liquor NPV net present value
DAP diammonium phosphate NREL National Renewable Energy
DB declining balance Laboratory
DCFROR discounted cash flow rate of OSBL outside battery limits (of the
return plant)
DOE U.S. Department of Energy PFD process flow diagram
FCI fixed capital investment P&ID piping and instrumentation
FGD flue gas desulfurization design
GBT gravity belt thickener PNNL Pacific Northwest National
GGE gallon gasoline equivalent Laboratory
GHG greenhouse gas PSA pressure swing adsorption
H/D height-to-diameter ratio R&D research and development
HDO hydrodeoxygenation RDB renewable diesel blendstock
HHV higher heating value RFS Renewable Fuel Standard
HMF 5-hydroxymethyl furfural RO reverse osmosis
INL Idaho National Laboratory SCF standard cubic feet
IS insoluble solids SMR steam-methane reforming
IRR internal rate of return SOT annual State of Technology
ISBL inside battery limits (of the case
plant) TAN total acid number
IX Ion Exchange TCI total capital investment
LCA life cycle assessment TDC total direct cost
LCI life cycle inventory TEA techno-economic analysis
LHSV liquid hour space velocity TS total solids
LHV lower heating value WHSV weight hourly space velocity
MACRS IRS Modified Accelerated WWT wastewater treatment
Cost Recovery System
MESP minimum ethanol selling
price

viii
1 Introduction
1.1 Background and Motivation
The U.S. Department of Energy (DOE) Bioenergy Technologies Office (BETO) promotes the
production of liquid fuels from lignocellulosic feedstocks by sponsoring programs in
fundamental and applied research that aim to advance the state of biomass conversion
technology. These programs include laboratory research to develop improved cellulose
hydrolysis enzymes and metabolic conversion microorganisms through synthetic biology,
catalyst development and testing, chemical and mechanical pretreatment work, detailed
engineering studies of potential processes, and construction of pilot-scale demonstration and
production facilities. This research is conducted by national laboratories, universities, and private
industry in conjunction with engineering and construction companies.

To support the DOE program, the National Renewable Energy Laboratory (NREL) investigates
the process design and economics of cellulosic biofuel manufacturing in order to develop a plant
gate price for fuels and fuel blendstocks based on process and plant design assumptions
consistent with applicable best practices in engineering, construction, and operation. This plant
gate price is referred to as the minimum fuel selling price or MFSP. The MFSP can be used by
DOE to assess the cost-competitiveness and market penetration potential of a given cellulosic
biofuel technology pathway in comparison with petroleum-derived fuels and established biofuel
technologies such as starch- or sugar-based ethanol.

The techno-economic analysis (TEA) effort at NREL also helps to direct biomass conversion
research by examining the sensitivity of the MFSP to process alternatives and research advances.
Proposed research and its anticipated results can be translated into a new MFSP that can be
compared to the benchmark case documented in this report. Such comparison helps to quantify
the economic impact of core research targets at NREL, and elsewhere, and to track progress
toward meeting competitive cost targets. It also allows DOE to make more informed decisions
about research proposals that claim to reduce MFSP, and to better quantify trade-offs between
process metrics (yields, chemical inputs, carbon efficiencies), economics (MFSP), and
sustainability (greenhouse gas, fossil energy, water profiles) among process options for a given
technology pathway or amongst multiple pathways.

For more than 10 years, NREL has developed design case models and associated reports [1, 2]
that documented process and cost targets for ethanol production from cellulosic feedstocks via
biochemical conversion (e.g., deconstruction to monomeric sugars followed by fermentation to
ethanol), based on the best understanding of the technology and equipment costs at the time. As
understanding evolved on process and economic metrics, models were refined and re-
benchmarked relative to updated targets. The final update to the ethanol design targets was
published in 2011 (referred to hereafter as the 2011 design report) [2] and established an
MFSP cost goal of $2.15/gal ethanol to be achieved in 2012 based on performance data from
pilot-scale demonstration runs conducted at NREL. This goal was subsequently achieved through
2012 pilot plant trials, with the pertinent experimental data input to the model, which calculated
an ethanol MFSP of $2.15/gal for a commercial-scale nth-plant facility processing 2,000 dry
metric tons/day of corn stover [3-5]. Further context on the history of NRELs ethanol program,
including individual year-by-year modeled production costs, may be found in Tao et al. and
Davis [4, 6].

1
While further room for improvement exists to continue reducing ethanol costs, the achievement
of the 2012 minimum ethanol selling price (MESP) target represented the culmination and
closeout of NRELs primary charter to demonstrate technology improvements required to de-risk
the technology in support of moving it toward commercialization, and to provide a benchmark
for private industry to expand upon. Moving forward, in 2013 DOE-BETO began transitioning
its core platform focus from ethanol to hydrocarbon fuel and blendstock products
(fungible/infrastructure-compatible fuels), including a new pathway MFSP cost target of
$3/gallon gasoline equivalent (GGE) in 2022 for the biochemical (sugar conversion) platform
[7]. Within the biochemical platform, two options were established to consider for sugar
conversion to hydrocarbon fuels, namely biological (microbial fermentation) conversion and
catalytic conversion. The biological conversion pathway was first modeled, analyzed, and
documented in a design report published in 2013 (hereafter referred to as the 2013 design
report) [8], which presented a modeled MFSP of $5.10/GGE and hydrocarbon fuel yield of 45
GGE/ton as an intermediate target case prior to 2022, with additional insight provided for a path
forward to ultimately meet the final 2022 target of $3/GGE.

As described in the 2013 design report, a key benefit of the biological conversion approach is the
ability to produce specific, easily-characterized and potentially high-value hydrocarbon products
afforded by metabolic engineering. However, a drawback of the pathway lies in the challenging
carbon efficiency and fuel yields inherent to metabolic conversion of sugars (rejecting oxygen as
CO2), as well as challenges for large-scale aerobic bioconversion. Building from the basis
established for the 2013 design report, the focus of this report is to similarly document a
plausible pathway model for the catalytic conversion of cellulosic sugars to hydrocarbon
blendstock fuels, both to enable achieving similar future cost targets, as well as to compare and
contrast key differences in this technology pathway relative to the biological pathway. The
overall scope of the report focuses on model details attributed to achieving intermediate MFSP
costs of $5/GGE or less prior to 2022, but with additional brief insight provided for a path
forward to ultimately meet the final 2022 target of $3/GGE.

1.2 Process Overview

The process modeled in this report uses preprocessing (deacetylation) and cocurrent dilute-acid
pretreatment of lignocellulosic biomass feedstock (primarily corn stover), followed by enzymatic
hydrolysis (saccharification) of the remaining cellulose, followed by hydrolysate conditioning
and catalytic conversion/upgrading of the resulting hydrolysate soluble carbon components to
naphtha- and diesel-range fuel products. The process design also includes feedstock handling and
storage, product upgrading, wastewater treatment (WWT), lignin combustion, product storage,
and utilities. The process is divided into nine areas (Figure 1).

Area 100: Feed handling. The feedstock, in this case milled corn stover blended with
switchgrass, is delivered to the feed handling area from a uniform-format feedstock supply
system. Only minimum storage and feed handling are required. From there, the biomass is
conveyed to the pretreatment reactor (Area 200).

Area 200: Pretreatment and conditioning. In this area, the biomass is processed in an alkaline
deacetylation step to solubilize and remove acetate as well as portions of ash, lignin, and other
components. It is then drained and treated with dilute sulfuric acid catalyst at a high temperature
for a short time to liberate the hemicellulose sugars and break down the biomass for enzymatic

2
hydrolysis. Ammonia is then added to the whole pretreated slurry to raise its pH to
approximately 5 for enzymatic hydrolysis.

Area 300: Enzymatic hydrolysis, hydrolysate conditioning, and purification. Enzymatic

hydrolysis is initiated in a high-solids continuous reactor using a cellulase enzyme produced on
site. The partially hydrolyzed slurry is batched to one of several parallel reactors. Hydrolysis is
completed in the batch reactor with a total time of 3.5 days between the continuous and batch
steps. The slurry is then fed to a vacuum filter press to remove insoluble solids (primarily lignin)
and the solids fraction exiting the filter press is sent to the boiler (Area 800). The hydrolysate is
then concentrated in a mechanical vapor recompression evaporator. A polishing filter is used to
remove any residual insoluble particulates carried over from the filter press and any solids that
may have precipitated in the evaporator. Finally, the filtered and concentrated hydrolysate is
treated by ion exchange to remove soluble cations and anions prior to catalytic upgrading.

Area 400: Cellulase enzyme production. An on-site enzyme production section was maintained
in this design, consistent with the details provided in the 2011 ethanol design report. Purchased
glucose (corn syrup) is the primary carbon source for enzyme production. Media preparation
involves a step in which a portion of the glucose is converted to sophorose to induce cellulase
production. An enzyme-producing fungus (modeled after Trichoderma reesei) is grown
aerobically in fed-batch bioreactors. The whole fermentation broth, containing the secreted
enzyme, is fed to Area 300 to carry out enzymatic hydrolysis.

Area 500: Catalytic conversion and upgrading. The catalytic upgrading stage begins after the
hydrolysate stream is concentrated and purified in Area 300. The catalytic conversion process
consists of four stages: hydrogenation, aqueous phase reforming (APR), condensation and
oligomerization, and hydrotreating. Each stage consists of packed-bed reactor vessels. Hydrogen
is added to the reactors in each stage which operate at varying process conditions and have
varying catalyst composition. The goal of these successive catalytic steps is to remove oxygen or
de-functionalize carbohydrates and other carbon components and oligomerize them to
primarily diesel-range hydrocarbons.

Area 600: Wastewater treatment. Wastewater streams are treated by anaerobic and aerobic
digestion. The methane-rich biogas from anaerobic digestion is sent to the boiler (Area 800),
where sludge from the digesters is also burned for the production of steam and power. The
treated water is suitable for recycling and is returned to the process.

Area 700: Storage. This area provides bulk storage for chemicals used and produced in the
process, including sulfuric acid, ammonia, nutrients, water, and products.

Area 800: Combustor, boiler, and turbogenerator. The solids from the filter press and WWT are
combusted along with the biogas from anaerobic digestion and the process off-gases to generate
high-pressure steam for electricity production and process heat. Most of the process steam
demand is in the pretreatment reactor. The boiler produces excess steam beyond facility heating
demands that is converted to electricity for use in the plant.

Area 900: Utilities. This area includes a cooling water system, chilled water system, process
water manifold, and power systems.

3
 AMMONIA

GLUCOSE
STEAM A400
ENZYME
ACID PRODUCTION
NUTRIENTS
CAUSTIC
A300
CELLULASE ENZYMATIC
A200 HYDROLYSIS,
RAW A100 FEEDSTOCK PRETREATMENT HYDROLYSATE HYDROLYSATE
FEEDSTOCK
FEED HANDLING
& CONDITIONING CONDITIONING, &
PURIFICATION

RECYCLE

CONCENTRATED
HYDROLYSATE
WATER

& PURIFIED
BLACK LIQUOR,
EVAPORATION FLASH
CONDENSATE

LIGNIN

A600 PURIFICATION A500

NUTRIENTS WASTEWATER RECYCLE
BLOWDOWN CATALYTIC
TREATMENT HYDROGEN CONVERSION &
UPGRADING

HYDROCARBON
WASTEWATER

PRODUCTS
PROCESS
OFF-GAS

AUX. FUEL

A800
STEAM A700
BURNER/BOILER
STORAGE
TURBOGENERATOR
ANAEROBIC
ELECTRICITY
BIOGAS

A900
UTILITIES
OVERALL PROCESS:
BIOMASS TO HYDROCARBON
Nov 2014 PFD-P120-A000

Figure 1. Simplified flow diagram of the overall process. (Key streams only. See Appendix E for more detailed schematic and process flow diagrams.)

4
1.3 Techno-Economic Analysis Approach
Figure 2 describes the engineering approach used for modeling the conversion of biomass to biofuels,
including process design, process modeling, and economic analysis. This approach was largely
followed for this study as well, albeit under a condensed timeline and with additional external inputs
from literature, patents, consultants, and vendors, for areas of the process in which NREL does not yet
have sufficient in-house data (primarily the purification and catalytic conversion steps). As such, this
report is less prescriptive in some sections than are previous reports, because of the early stage of
understanding for new areas of the process and the somewhat more preliminary nature of the
associated models.

Engineering Companies
Consulting on Process Process Flow Diagrams
NREL or Other
Configuration
DOE-Sponsored
Research

Rigorous Published Literature

Estimates of Other
Material & Energy Balances
Commercial Technology
(Aspen Plus )
Outside Engineering
Studies (e.g., feed handling,
Cost Estimation separations)
Software (e.g., ICARUS )

Capital & Project Cost

Vendor Cost Quotations
Estimation

Engineering Company
Cost Estimations

Discounted Cash Flow

Economic Model

Minimum Fuel Selling Price

Figure 2. NRELs approach to process design and economic analysis

Starting from the general process flow diagram shown in Figure 1 and the more detailed diagrams
contained in Appendix E, a process simulation is developed using Aspen Plus software [9]. This
process model computes thermodynamically rigorous material and energy balances for each unit
operation in this conceptual biorefinery.

5
The material and energy balance data from the Aspen simulation are next used to assist in determining
the number and size of capital equipment items. As process conditions and flows change, baseline
equipment costs are automatically adjusted in an Excel spreadsheet using a scaling factor. These
baseline costs come from vendor quotes (a favored procedure for larger or nonstandard unit operations
and packaged or skid-mounted subsystems) or from Harris Groups historical cost data (for secondary
equipment such as tanks, pumps, and heat exchangers). Harris Group provided updated design and
cost estimates based largely on vendor budgetary quotations for critical new or modified areas of the
current process model, primarily hydrolysate polishing filtration, ion exchange purification, and
catalytic conversion operations, as new operations relative to NRELs prior 2013 design report
focused on biological conversion. Final equipment costs for this report are tabulated in Appendix A.

Once equipment costs are determined, direct and indirect overhead cost factors (e.g., installation costs
and project contingency) are applied to determine a feasibility-level estimate of total capital
investment (TCI) in 2011 U.S. dollars. The TCI, along with the plant operating expenses (also
developed using flow rates from the Aspen model), is used in a discounted cash flow rate of return
(DCFROR) analysis to determine a plant gate price for total hydrocarbon fuel for a given discount
rate. In this model, the fuel product is primarily refined renewable diesel blendstock (RDB) with a
lesser fraction of naphtha. The plant gate price is also called the minimum fuel selling price (MFSP, in
$/GGE), required to obtain a net present value (NPV) of zero for a 10% internal rate of return (IRR)
after taxes, associated with an nth-plant model (discussed in the next section).

The product of the analysis described above is a techno-economic model that estimates a product price
for a pre-commercial process. The resultant MFSP is unique for the set of process conditions
simulated, and it should be emphasized that some amount of uncertainty always exists around these
chosen conditions, as well as around the assumptions made for capital and raw material costs. Without
a detailed understanding of the basis behind it, the computed MFSP carries a risk of being taken out of
context. While the MFSP can be used to assess the marketplace competiveness of a given process, it is
best suited for comparing technological variations against one another or for performing sensitivity
analyses that indicate where economic or process performance improvements are needed.

1.4 About nth-Plant Assumptions

The techno-economic analysis reported here uses what are known as nth-plant economics. The key
assumption implied by nth-plant economics is that our analysis does not describe a pioneer plant;
instead, it assumes several plants using the same technology have already been built and are operating.
In other words, it reflects a mature future in which a successful industry of n plants has been
established. Because the techno-economic model is primarily a tool for studying new process
technologies or integration schemes in order to comment on their comparative economic impact, nth-
plant analysis avoids artificial inflation of project costs associated with risk financing, longer startups,
equipment overdesign, and other costs associated with first-of-a-kind or pioneer plants, lest these
overshadow the economic impact of research advances in conversion or process integration. At the
same time, NREL also continues to work on quantifying economic factors associated with first-of-a-
kind implementation. At the very least, these nth-plant economics should help to provide justification
and support for early technology adopters and pioneer plants.

The nth-plant assumptions in the present model apply primarily to the factored cost model used to
determine the total capital investment from the purchased equipment cost and to the choices made in
plant financing. The nth-plant assumption also applies to some operating parameters, such as process

6
uptime of 90%. These assumptions were agreed upon by BETO and reflect standard metrics applied
universally across multiple national laboratory TEA models to allow for a consistent basis for analysis.
It should be emphasized, however, that these assumptions carry a degree of uncertainty and are subject
to refinement.

1.5 About the NREL Aspen Model

While Aspen Plus can be thermodynamically rigorous, such detail is not always warranted in the
simulation, whether for lack of data or introduction of additional complexity for little gain in accuracy.
Some unit operations, particularly solid-liquid separation and hydrolysate purification (polishing
filtration and deionization), were modeled with a fixed performance determined by experimental
testing or by standard engineering practices confirmed with vendors. Catalytic reactors were modeled
using published values from patent literature for carbon yields to product classes and carbon number
ranges (discussed further in Section 3.5) rather than rigorous kinetics or rate expressions, dictated by a
dearth of publicly available guiding data and the granularity of the data which is available. This simple
stoichiometric model still satisfies mass and energy balances, and corresponds closely with published
yield targets.

The Aspen Plus simulation uses component physical properties internal to the software, as well as
property data developed at NREL or from the literature [10, 11]. Similar to the 2011 and 2013 models,
the current model does not rely on external property databanks and minimizes the number of custom-
defined components within reason. A discussion of components and properties used is given in
Appendix D.

7
2 Design Basis and Conventions
2.1 Plant Size
The plant size in the present design is the same as that used in prior designs: 2,205 dry U.S. ton/day
(2,000 metric tonne/day). With an expected 7,880 operating hours per year (90% uptime), the annual
feedstock requirement is 724,000 dry U.S. ton/year. The present model maintains consistency with the
2013 design report basis and assumes that feedstock is delivered under a uniform-format logistics
system, which includes a blended feedstock consisting of multi-pass corn stover, single-pass corn
stover, and switchgrass, to enable achieving cost and composition requirements set for 2017 targets by
Idaho National Laboratory (INL) [12]. This blending strategy assumes that the formulation of multiple
feedstocks will perform as well as, or better than, a singular feedstock, while allowing for ultimately
achieving stipulated feedstock cost targets. There are research plans for FY 2014 to test various
blended feedstocks in multiple technology research pathways. Under the new uniform-format system,
opportunities may exist to justify a larger facility scale with an increased feed rate to the biorefinery,
thus allowing for economy of scale advantages; however, the specifics are not yet defined and we
retain the 2,000 dry tonne/day basis here.

2.2 Feedstock Composition

The conversion target feedstock composition remains consistent with what was assumed in the 2011
and 2013 design cases; however, the biomass supply composition is updated relative to the original
2011 case. As noted by INL, the previous requirements for feedstocks assumed in the 2011-era ethanol
models included volume and cost targets only. There were no quality requirements; further analysis
indicated a mismatch on biomass characteristics and conversion in-feed specifications. The update
applied in the 2013 design case, and maintained here, improves on the 2011 design basis and the
associated original feedstock price of $58.50/dry ton (2007$), which included a more uncertain grower
payment, to a more reasonable target at an increased price of $80/dry ton, including grower payment
(2011$). This new price is more appropriate for a large commodity scale going beyond a niche
market price.

As has been described in prior design reports, the feedstock composition (see Table 1) plays a critical
role in overall process design and economics. Of particular importance are carbohydrate components
(cellulose and hemicellulose), as well as lignin, acetate, and ash, given the modifications being made
to the pretreatment strategy such as the use of deacetylation as well as high sensitivity to impurity
components such as ash and metals in the catalytic reactor section of this design. The blended
uniform-format feedstock composition assumed here for purposes of future design case targets is
shown below, with supporting details (in the context of corn stover compositional variability)
described in the 2011 ethanol report [2]. Also consistent with prior design cases, the moisture content
for the delivered feedstock is 20%.

When converting the analytical composition to components used in the Aspen Plus model, the
nonstructural component fractions from the compositional analysis were combined under
extractives. The extractives component is assumed to be primarily organic, with an average
composition of CH2O, and consists primarily of sugars, sugar alcohols, and organic acids, as well as
some nonstructural inorganics [13]. The presence of extractives in the biomass depends on the time of
harvest and, in part, to how much microbial degradation of the material occurs after harvest; the
amount of extractives in a given sample may therefore be indicative of its age. Additionally, where the

8
mass balance did not sum to 100%, the extractives component was used to close it by difference.
Sucrose is another extractive component, but it is measured separately in laboratory analysis and has
been added as a separate feedstock component in the present design. The amount of sucrose present in
the biomass is dependent on harvesting and handling practices. In pretreatment, this sucrose is
assumed to be fully hydrolyzed to glucose and fructose. The labile fructose is further converted to
degradation products in pretreatment, but the less reactive glucose resists degradation [14].

Table 1. Delivered Feedstock Composition Assumed in the Present Design

Component Composition
(dry wt %)
Glucan 35.1
Xylan 19.5
Lignin 15.8
Ash 4.9
Acetatea 1.8
Protein 3.1
Extractives 14.7
Arabinan 2.4
Galactan 1.4
Mannan 0.6
Sucrose 0.8
Total structural carbohydrate 59.0
Total structural carbohydrate + sucrose 59.8
Moisture (bulk wt %) 20.0
a
Represents acetyl groups present in the hemicellulose polymer; converted to acetic acid in pretreatment.

In the present design, the lignin fraction will also become essentially of equal economic importance as
the carbohydrates because of the need to improve overall carbon efficiency to product(s) in order to
ultimately achieve the 2022 $3/GGE target. This is described further in Section 5.4.

2.3 Updated Modeling Basis

Similar to other biofuel pathways recently published in design reports under DOE-BETO efforts [8,
15], a number of key modeling assumptions are fixed constant to allow for comparable modeling
outputs. These parameters include:

1. Facility start-up time: 0.5 year (6 months)

2. Cost-year dollar basis: 2011 dollars (all cost results presented here will be in 2011 dollars)
3. Facility on-stream time: 90% (7,884 hours/year); note, this is still intended to represent an nth-
plant facility, but allows for slightly more operational downtime in part to reflect the increased
process complexity of the new pathway models considered for hydrocarbon production,
relative to the prior 2011 ethanol basis of 96% on-stream time.

Additionally, the feedstock cost for the biochemical pathway is updated from a previous basis of
$58.50/dry ton (2007$) as utilized in the 2011 design case to $80/dry ton (2011$) for the 2013
biological case and present catalytic design target cases, to reflect a more commercially relevant
pricing model at a larger national scale (described in more detail in the 2013 report).

9
2.4 Design Report Conventions
2.4.1 Units
The Aspen Plus model described in this report is, by legacy, based on the set of units required by
Aspen for specifying custom component properties: kg, kmol, atm, and C for materials, and MMkcal
(Gcal) for energy. Values in this report that were drawn directly from the Aspen model therefore tend
to be reported in these units. Harris Group preferred to use U.S. standards (lb, BTU,F, gal, etc.) when
communicating with equipment vendors. Therefore, equipment specifications tend to be cited in these
U.S. units.

Note that in the present report, certain quantities (e.g., yields and costs) are computed and reported in
terms of tons. To avoid ambiguity, tonne will denote a metric tonne (1,000 kg) and ton will denote a
short or U.S. ton (2,000 lb). In general, the U.S. ton is the standard for this document.

2.4.2 Total Solids Loading

The process described here converts a solid feedstock (corn stover blended with switchgrass) into
liquid fuel products (diesel/RDB + naphtha). Most material streams in the process therefore have a
solid fraction and a liquid fraction. The relative amount of solids in a given stream is called its solids
loading. Total solids (TS) loading is defined as the total weight percent of soluble solids (e.g., sugars
and salts) and insoluble solids (e.g., cellulose and lignin) in a given material stream. Where useful, the
TS loading and the insoluble solids (IS) loading will be reported together. Note that in our convention,
sulfuric acid, acetic acid, and ammonia are not considered soluble solids, but ammonium acetate and
ammonium sulfate are. Therefore, around some unit operations, e.g., hydrolysate conditioning, TS
loading is not a conserved quantity.

2.4.3 Product Yield and Heating Value

The results from this analysis are reported primarily in terms of energy yields in gallons- gasoline
equivalent: $/GGE, GGE/yr, GGE/ton, etc. This includes energy yield in both the diesel (RDB) and
naphtha product cuts from hydrotreating distillation, rather than yield of a single product with the
other treated as a coproduct; this is done to maintain focus on total fuel (energy) yield from
hydrolysate carbon, to avoid introducing unnecessary subjectivity in over-interpreting the granularity
of available public data regarding absolute yields to diesel- versus naphtha-range products from
catalytic conversion/upgrading. Additionally, as modeled, the naphtha product is a non-trivial fraction
of total product yield on a volume basis (approximately 20 vol % of total fuel produced), and as such
is more appropriate to consider the summative energy-equivalent (GGE) yield for both products, as
dramatic sensitivity to MFSP could be introduced when assigning naphtha a coproduct value; this is
consistent with other recently-published TEA work [16]. Lower heating values (LHVs) for both diesel-
and naphtha-range products were calculated by the Aspen model, and correspond to similar values for
standard petroleum-equivalent products [17]. To translate to a GGE basis, a conventional gasoline
heating value of 116,090 BTU/gal (LHV basis) was applied [17].

10
3 Process Design and Cost Estimation Details
3.1 Area 100: Feedstock Storage and Handling
3.1.1 Overview
Area 100 processes incoming biomass feedstock. Herbaceous biomass is the preferred feedstock for a
biochemical conversion pathway focused on biomass deconstruction to monomeric sugars and other
carbohydrates. The present design maintains all feedstock assumptions as consistent with NRELs
2013 biological hydrocarbon design report [8] and is summarized here. The design assumes a blend of
corn stover and switchgrass delivered according to the specifications detailed in the INL design report
for the uniform-format feedstock supply system [18]. In this envisioned design, biomass is stored in a
central depot and is preprocessed, densified, stabilized, and homogenized to a degree before delivery,
such that the biorefinery receives feedstock with known, uniform-format specifications including
particle size, ash content, and moisture content. All analysis and cost implications associated with
Area 100, and with feedstock logistics in general, are outside the scope of the present effort focusing
on the conversion biorefinery, but are summarized here to convey information attributed to the
updated feedstock cost of $80/dry ton delivered to Area 200.

To address the high cost of biomass procurement, lower cost resources could be supplemented from
other regions beyond the U.S. Midwest, but this strategy would increase transportation costs. Another
strategy to reduce the cost of obtaining biomass is to blend or use numerous types of feedstocks
within a region to design an aggregated or formulated feedstock. This strategy assumes that the
formulation of multiple feedstocks will perform as well as or better than a singular feedstock. As
illustrated in Figure 3, the advanced Terrestrial Feedstock Processing Supply and Logistics System is
envisioned to draw in presently inaccessible and/or underused resources via local biomass
preprocessing depots that format biomass into a stable, bulk, densified, and flowable material. The
formatted biomass will be transported to one or more networks of much larger supply terminals, where
the material aggregated from a number of depots may be blended and/or further preprocessed to meet
the specification required by each biorefinery conversion process. The advanced biomass logistics
system design incrementally incorporates technology and other system improvements as the industry
matures. The introduction of advanced preprocessing operations and sequences, including the
implementation of blending and formulation strategies, and preprocessing strategies, is critical to
achieving feedstock cost, quality, and volume targets to meet BETO 2017 and 2022 goals. In
supporting this effort, partners at NREL and INL are working to understand the impact that blended
feedstocks have on operability and convertibility through the various conversion unit operations, to
ultimately demonstrate that the targets stipulated in this design report are achievable for such blended
feedstock formulations rather than merely a single corn stover feedstock.

11
 Figure 3. The depot concept behind the advanced Terrestrial Feedstock Processing Supply and
Logistics System. Courtesy INL.

Blending strategies could also improve overall feedstock quality to help meet biofuel quality
specifications and conversion in-feed requirements. Quality targets can have a large impact on whether
or not a particular feedstock is cost effective. Inherent spatial and temporal variability of biomass
resource quality translates to variability in biomass compositional attributes including ash, xylan,
glucan, and lignin. Formulation allows the use of low-cost, but typically low-quality, biomass blended
with higher cost and quality biomass to achieve minimum conversion process quality specifications.
One particularly important consideration in the present analysis is that catalytic conversion processes
are more sensitive than biological conversion processes to the amount, type, and variability of ash
(part of which is dependent on different feedstock sources and types). It is known that metal ions,
silica, and other species present in herbaceous biomass ash may poison, deactivate, or otherwise
precipitate and clog catalyst systems and related equipment [19]. In the base analysis, to protect
catalysts a fraction of the ash delivered to the biorefinery (see Table 1) is assumed to be removed up
front during pretreatment deacetylation (Area 200), and the remainder removed during hydrolysate
purification (Area 300). However, additional opportunities may exist to reduce ash content earlier on
in the process through other logistics preprocessing mechanisms. This will be a continued subject of
further analysis in collaboration with partners at INL.

3.1.2 Design Basis

In the new uniform-format feedstock supply system, feedstock would be stored in a satellite depot
location with delivery to the biorefinery occurring 6 days per week by truck, or possibly by rail. At the
depot, material would be milled to a mean size of 0.150.25 in. (with a high content of fines), and then
densified into pellet format in order to achieve a mean bulk density of 30-40 lb/ft3 to maximize the
biomass load per trailer. The incoming uniform-format feedstock (corn stover/switchgrass blend) is
assumed to have 20% moisture when it reaches the biorefinery; this is representative of a mixture of
field-dried material having less than 15% moisture and co-harvested material having greater than 20%

12
moisture. Because the preprocessing operation is designed to lose very little dry matter and does not
include any rinsing of biomass, it is assumed that the biomass composition discussed in Section 2.2 is
valid for the delivered material.

The as-received feedstock requirement for a plant is 2,750 U.S. ton/day (104,200 kg/h; 229,700 lb/h)
including moisture. In the projected design, refinery receiving operates on the same schedule as the
biomass depot: 24 hours per day, 6 days per week. Incoming trucks are weighed by electronic scale
and unloaded using a whole-truck dumper capable of a 710 minute unloading time. The dumpers
empty into dedicated hoppers that meter the biomass to a series of conveyors. The conveyors carry
material from the truck tipper to short-term storage. The minimum receiving rate is 250 ton/h to
maintain 110 ton/h of continuous processing. In order to process 250 ton/h, and assuming a relatively
constant flow of trucks, a pair of scales (one inbound and one outbound) and two truck dumpers are
required.

On-site storage is kept to a minimum of 72 hours to allow for a weekend buffer. Open piles are not
favored due to concerns of fire, rodent infestation, and moisture degradation. Instead, the unloaded
feedstock is stored in concrete domes. Two domes (each with a 36-hour capacity) are required so that
one can be loaded while the other empties to the conversion process.

Conveyors connect the storage domes to the feedstock receiving bins on the pretreatment reactor in
Area 200. A dust collection system integrated with the conveyors and domes handles airborne particles
released during the unloading and conveying processes. No dry matter is assumed lost in Area 100.

3.1.3 Cost Estimation

Consistent with the 2013 biological hydrocarbon design report, the feedstock cost assumed in this
report is $80/dry ton (2011$) attributed to the new uniform-format logistics system described above.
This cost comes from the Multi-Year Program Plan (MYPP) published by DOE-BETO [7]. This is a
rolled-up cost which includes both grower payment as well as logistics considerations for all
collection, processing, storage, and transportation costs between the field and the receiving bin on the
pretreatment reactor (e.g., Table B-2 of the referenced MYPP document). It should be stressed that
these costs are future (20172022) DOE research targets, similarly to the conversion performance
targets used in the Aspen Plus model.

3.2 Area 200: Pretreatment and Conditioning

3.2.1 Overview
Pretreatment of biomass for biofuel production is a crucial step. Its primary role is to disrupt the
matrix of polymeric compounds that are physically and chemically bonded within lignocellulosic
biomass cell wall structures, including cellulose microfibrils, lignin, and hemicellulose. Pretreatment
has a significant impact not only on enzymatic hydrolysis, conversion, and downstream processing,
but also on overall process economics and sustainability [20]. Acid pretreatments tend to hydrolyze
significant fractions of hemicellulose and small fractions of cellulose.

Figure 4 shows a simplified flow diagram of the pretreatment area that uses a two-step process
(deacetylation and dilute-acid pretreatment). Milled biomass (corn stover and switchgrass blend) is
conveyed to the pretreatment area and treated first with dilute sodium hydroxide (for deacetylation),
followed by sulfuric acid at moderate to high temperature (see Table 2) to liberate the hemicellulose

13
sugars and break down the biomass for enzymatic hydrolysis. Solids exiting the alkaline deacetylation
step are sent on to dilute-acid pretreatment. Alkaline pretreatments break bonds between lignin and
carbohydrates and disrupt lignin structure, making carbohydrates more accessible to enzymatic
hydrolysis [20]. The dilute-acid pretreatment process converts most of the hemicellulose
carbohydrates in the feedstock to soluble sugars (xylose, arabinose, galactose, and mannose) by
hydrolysis reactions. This is followed by ammonia conditioning to raise the pH of the whole pretreated
slurry from near 1 to 5, which is then sent to enzymatic hydrolysis.

Uniform formatted feedstock

NaOH
steam

To wastewater
Deacetylation
reactor Black liquor

Acid
Vertical presteamer
Dilution water
& NH3
Horizontal reactor
To enzymatic hydrolysis

High-pressure steam
Blowdown Ammonia
tank conditioning

Figure 4. Simplified flow diagram of the deacetylation, pretreatment, and conditioning process

3.2.2 Design Basis

The model used in this report employs the same design and yield assumptions as the 2013 design case
[8] and is summarized below. Deacetylation has been used to improve conversion of oligomeric xylan
to monomeric xylose, with one study showing a decrease in oligomeric xylose yield from 21%23%
for non-deacetylated corn stover to 7%10% for deacetylated corn stover [21]. Biological upgrading
of sugars to fuel can benefit from the conversion of oligomers to monomeric sugar due to the
metabolic limits of microorganisms; however, research suggests that catalytic conversion of biomass
may also utilize oligomers equally to monomers and that they may even be desirable [19]. The current
base case design is identical to the 2013 case for two reasons: (1) it provides consistency between
design reports and (2) there is a lack of public data from which to build a detailed, redesigned
pretreatment system that targets oligomer production. Additionally, maintaining the deacetylation step
is useful in the present design as it also removes a significant fraction of ash up front, which is a
problematic component that must be removed to protect conversion reactor catalysts. There are
potential opportunities for cost improvements by targeting oligomers, including lower cost reactor
metallurgy, fewer chemical inputs to pretreatment, and lower conditioning and purification costs as
discussed further in Section 5.2.

The deacetylation step uses a sodium hydroxide soaking process at 80C for 1 hour, with a NaOH
loading of 17 mg/g dry biomass (see Table 2). The deacetylated material is dewatered by draining
through screens at the bottom of the deacetylation reactor. A 316 stainless steel vertical pressure vessel

14
is used for the deacetylation tank. Three identical batch reactors are used with square trough live-
bottoms consisting of four 24-inch horizontal discharge screws each. Each tank volume is 44,000 gal
with a total cycle time of 2.4 hours. The drained liquor, often referred to as black liquor, contains
20%25% of the original dry biomass constituent material, including water-soluble extractives (100%
solubilization), soluble ash constituents (75% ash solubilization), 20% of the lignin, 2% of the xylan,
50% of the sucrose, and 88% of the acetate that was originally present in the feedstock (dry basis).
This liquor stream is pumped to wastewater treatment (A600) in the baseline process design. The
remaining biomass solids are discharged from the deacetylation reactor and transported to the acid
pretreatment system. Moving forward, there may be opportunities to identify conditions which
optimize deacetylation more towards ash removal than acetate removal, as acetate may be utilized as
an additional carbon source in the downstream catalytic conversion operations in this case.
Additionally, more room for further investigation exists regarding the implications that solubilized
lignin components have on performance of the wastewater treatment organisms, which may become
problematic at a certain threshold of soluble lignin removal.

The acid pretreatment reactor system was described in detail in the 2011 ethanol design report [2]. The
system includes a feedstock receiving system, followed by a vertical vessel with a long residence time
for steam heating and potential acid impregnation of the biomass, followed by the horizontal
pretreatment reactor, which operates at a higher pressure and a short residence time. The 26-inch plug
screw feeder is a rugged, high-compression screw device designed to form a pressure-tight plug of
material through axial compression. Dilute sulfuric acid is metered at the discharge spool of each plug
screw feeder. Feedstock drops from the plug screw discharge into a mixing and heating screw, which
discharges the feedstock into the top of the presteamer. High-pressure steam is injected into this vessel
to maintain temperature, while hot water is added at this point to control the pretreatment effluent at
30 wt % TS; at this concentration, the material is not easily pumpable and is primarily transported
between unit operations within Area 200 using transfer conveyors.

The current model assumes operation of the presteamer at 100C such that no significant hydrolysis
reactions occur in this section. It can be used if additional hold up is required for acid hydrolysis.
Feedstock flows downward through the vertical presteamer with uniform temperature throughout,
discharging through a dual screw outlet device to two plug screw feeders. The plug screw feeders
meter feedstock to the horizontal pretreatment reactor, with a horizontal reactor configuration chosen
to allow tighter residence time distribution control than with a vertical reactor.

Upon deacetylation and solid-liquid separation by draining the black liquor, the actual sulfuric acid
loading present in the horizontal acid pretreatment reactor is only 9 mg/g dry feedstock. The reactor
pressure is held at the bubble point for the mixture. Heat losses from the reactor are not accounted for
in the energy balance calculations. The residence time in the pretreatment reactor is nominally 5
minutes. The effective sulfuric acid concentration in the pretreatment reactor (estimated at 0.30.4 wt
% after dilution by condensing steam is accounted for) may allow for the use of lower-cost
metallurgies in the reaction zone (such as 904L or other duplex stainless alloys) instead of Incoloy-
clad carbon steel, which could reduce pretreatment equipment costs. However, relevant corrosion data
under these conditions are not available, so in the model, Incoloy825 cladding is still conservatively
assumed for the material of construction (consistent with the 2011 and 2013 design reports). The
potential impact of lower-cost metallurgy and other pretreatment reactor cost savings is addressed in
the Sensitivity Analysis section.

15
Additional acid is used to pre-impregnate the deacetylated biomass (the cost of this usage is included
in the purchased sulfuric acid cost in the techno-economic model), but much of this acid is removed in
pilot plant operation upon dewatering of the impregnated biomass and does not enter the pretreatment
reactor. Acid pre-impregnation and its associated dewatering step are not explicitly modeled in our
design, but it is envisioned that these could take place in the vertical presteamer, similar to the
rationale in the 2011 design report [2]. The pretreatment reactor operating conditions are summarized
in Table 2, and are maintained consistently with the 2013 design report targets.

Table 2. Deacetylation and Pretreatment Conditions Applied in this Design

Parameter (DA = deacetylation; PT = pretreatment) Current design
DA Temperature 80 C
DA Residence time 1 hour
DA Caustic (NaOH) loading 17 mg/g dry biomass
PT Total sulfuric acid loading present in pretreatment reactor 9 mg/g dry biomass
PT Residence time 5 minutes
PT Temperature 160C
PT Pressure 5.5 atm
PT Total solids loading 30 wt %

Table 3 summarizes the reactions and conversions that take place in pretreatment. Glucan contained in
the hemicellulose is converted to glucose along with a small portion of the cellulose. Minor
hemicellulose carbohydrates (arabinan and galactan) are assumed to have the same reactions and
conversions as xylan. The xylan-to-xylose conversion is a total hydrolysis value, which also may
include an enzymatic component that will be discussed later. The sucrose reaction to HMF and
glucose reflects 100% hydrolysis of sucrose to fructose and glucose, followed by complete
degradation of the fructose to HMF.

Table 3. Pretreatment Hydrolysis Reactions and Assumed Conversions

Reaction Reactant % Converted to Product
(Glucan)n + n H2O n Glucose Glucan 9.9%
(Glucan)n + n H2O n Glucose Oligomera Glucan 0.3%
(Glucan)n n HMF + 2n H2O Glucan 0.3%
Sucrose HMF + Glucose + 2 H2O Sucrose 100%
(Xylan)n + n H2O n Xylose Xylan 90.0%
(Xylan)n + m H2O m Xylose Oligomera Xylan 2.4%
(Xylan)n n Furfural + 2n H2O Xylan 5.0%
Acetate Acetic Acid Acetate 100%
(Lignin)n n Soluble Lignin Lignin 5.0%
a
Sugar oligomers are considered soluble and convertible via downstream catalytic conversion.

Table 4 shows demonstrated performance results achieved during NRELs 2012 pilot-scale
demonstration runs, as well as the 2011 ethanol report assumptions and the present pathway
conversion targets (i.e., those used in the current Aspen model) for the pretreatment area. The present
model target is 90% conversion of xylan to monomeric xylose with 5% loss to degradation products.

16
 Table 4. Research Status and 2017 Targets in the Pretreatment Area

2011 Design 2012 State of Present Design

Report [2] Technology [6] Targets
Pretreatment
Solids loading (wt %) 30% 30% 30%
Xylan conversion to xylose (%) 90% 82% 90%
Xylan conversion to furfural (%) 5% 6% 5%
Conditioning
Ammonia loading (g/L of hydrolysate) 4.8 1.6 1.6
Hydrolysate solid-liquid separation no no no
Xylose sugar loss (% entering conditioning) 1% 0% 0%
Glucose sugar loss (% entering conditioning) 0% 0% 0%

The pretreatment reactor is discharged to a flash tank. The pressure of the flash is controlled to keep
the temperature at 100C (212F). The flash is condensed and routed to WWT (Area 600), containing
more than 30% of the furfural along with other volatile organics from pretreatment. While this stream
could potentially be routed to the conversion reactors (Area 500) as an additional source of carbon, it
would first require a separation step to remove a large amount of water at added cost, as sending the
full flash stream to the conversion section would greatly re-dilute the hydrolysate stream fed to
conversion by more than doubling the total amount of water; thus, the model assumes it is instead
routed to wastewater treatment. After the flash, the hydrolysate whole slurry containing 30% TS and
16% IS is sent to conditioning, namely neutralization by ammonia in stoichiometric quantities.
Ammonia gas is mixed into dilution water to raise the hydrolysate pH to 5. The residence time for
neutralization is 30 minutes and the dilution cools the slurry to 75C (167F). The slurry is diluted
with water to slightly greater than 20 wt % TS to ensure miscibility through enzymatic hydrolysis and
conversion. The composition of the stream at this point, and other major points throughout the process,
is shown in the stream table information in Appendix E. The material from conditioning is conveyed
to a saccharification storage tank.

3.2.3 Cost Estimation

Each deacetylation tank was quoted by Harris Group at $890,000 for this design, including reactor
tanks and conveyors. The acid pretreatment reactor design and cost basis was updated in the 2013
design report to allow for a more robust cost estimate with higher flexibility in operational variables.
Namely, an evaluation of pretreatment reactor pricing in terms of metallurgy, reactor conditions, and
reaction residence time was conducted by Harris Group based on input from a vendor. In addition to
the base case Incoloy-clad metallurgy used as the baseline assumption, a less costly duplex stainless
steel option was considered as an alternative. Further details behind the cost study may be found in the
2013 report, but in summary, the effort established correlation equations that allow for capital cost
estimation of the pretreatment reactor system as a function of residence time for the two reactor
metallurgy choices:

Total Equipment Cost (2011$ MM); Incoloy825 = 16.4+7.4m (1)

Total Equipment Cost (2011$ MM); Duplex steel = 16.8+3.9m (2)

Where, m = integer (minutes of resident time/3.3)

17
The above equations are based on an original design basis of 1,516 dry tonne/day biomass feed rate to
the pretreatment operation (reflecting fractional biomass solubilization upstream during deacetylation
from an initial feed rate of 2,000 dry tonne/day), and may be scaled to a different basis following
standard scaling factor adjustments as discussed in Section 4.2. For the conditions used in this design,
which are milder than earlier designs as employed in the 2011 ethanol report and other related work
under the ethanol program, the recommended metallurgy from the vendor continues to be Incoloy
825. However, further opportunity exists to continue pursuing more mild pretreatment conditions,
particularly for this catalytic conversion pathway, given the ability to utilize sugar oligomers rather
than strictly monomers as noted above. This may translate to lower acid and caustic neutralization
demands (lower operating costs as well as lower ammonium salt removal demands during downstream
hydrolysate cleanup), as well as allowing for the use of the lower-cost metallurgy presented above.
The total installed capital cost for the pretreatment system is approximately $46.8 MM (2011$)
including feedstock receiving bin, pre-steaming, pressurized heating, reaction, and flash cooling, with
another $4.6 MM for the deacetylation equipment. The reactor schematic remains consistent with the
configuration detailed in the 2011 and 2013 reports.

Area 200 contributes roughly $0.44/GGE to the MFSP, including deacetylation, acid pretreatment, and
conditioning. About 55% of this is attributed to capital cost, of which deacetylation and acid
pretreatment equipment accounts for 96% of total capital expenses. Given the capital-intensive nature
of the acid pretreatment step combined with the problematic ammonium sulfate salts introduced in the
process which must be subsequently removed for downstream catalyst protection, an alternative
pretreatment operation based on mechanical disc refining is considered later in the Sensitivity Analysis
section. This operation eliminates the need for steam heating as well as acid and caustic chemicals,
and may enable production of cleaner hydrolysate sugars; however, it also incurs a substantial power
demand as well as potentially requiring the addition of hemicellulase enzymes, as it does not
significantly hydrolyze hemicellulose to sugars like acid pretreatment does. This alternative will also
be investigated in the future for biological conversion pathways. Further discussion on this option is
presented in Section 5.2.

3.3 Area 300: Enzymatic Hydrolysis, Hydrolysate Conditioning, and

Purification
3.3.1 Overview
In this process area, cellulose is converted to glucose using cellulase enzymes. This process is known
as enzymatic saccharification or enzymatic hydrolysis. A cellulase enzyme preparation is a mixture of
enzymes that work together to break down cellulose fibers into cellobiose and soluble gluco-oligomers
and ultimately into glucose monomers. The resulting glucose and other sugars hydrolyzed from
hemicellulose during pretreatment are then filtered to remove insoluble solids, concentrated, and
purified by microfiltration and ion exchange.

The enzymatic hydrolysis portion of this process area follows the same fundamental process schematic
described in the 2011 ethanol report and 2013 biological hydrocarbon report [2, 8], namely, carrying
out hydrolysis at elevated temperature (48C) so the reaction proceeds faster and requires lower
enzyme loading, with hydrolysis split between two sequential reactions: a high-solids continuous flow
reactor followed by batch hydrolysis in a stirred tank.

18
The hydrolysate is then treated to remove problematic soluble and insoluble material, and to reduce
water content of the stream prior to catalytic upgrading steps. Figure 5 presents a schematic diagram of
the purification and concentration steps used in this design, which were established based on
recommendations from Harris Group and equipment vendors. First, insoluble lignin and other solids
are removed from hydrolysate using vacuum belt filtration with a high wash ratio to minimize soluble
sugar/carbon losses carried over into the solids fraction. The washed lignin-rich solids fraction is
routed to the boiler for combustion. The hydrolysate is then concentrated in a mechanical vapor
recompression evaporator under mild vacuum. A polishing filter (microfiltration) is used to remove
any residual particulates carried over from the belt filter and solids that may have precipitated in the
evaporator to protect downstream catalysts. Finally, the filtered and concentrated hydrolysate is treated
by ion exchange to remove soluble cations and anions that may otherwise deactivate catalysts, before
sending to catalytic upgrading.
Enzymatic Hydrolysis, Hydrolysate Conditioning & Purification

Condensate to process water

Enzyme Solids to recycle
combustor
From pretreatment
Lignin
filter Hydrolysate concentration

CW

Storage
tank

High-solids
hydrolysis Recycle purge
(continuous flow) Batch hydrolysis
reactor

Storage
Polishing Recycle from polishing filter
tank
filter

Storage
To catalytic conversion and
tank
upgrading

Storage
tank
Ion exchange

Figure 5. Simplified flow diagram of the hydrolysis, hydrolysate purification, and concentration unit
operations utilized in Area 300

3.3.2 Design Basis

The process and design basis for enzymatic hydrolysis is the same as described in the 2011 ethanol
report and 2013 biological hydrocarbon report, thus will only be briefly summarized here. In short,
enzymatic hydrolysis is initiated in a continuous, high-solids vertical tower reactor with the slurry
flowing down the reactor by gravity; this first step is required as the feed material at 20% solids (or
more) is not pumpable until the cellulose has been partially hydrolyzed. After mixing in the cellulase
enzyme, the total solids loading into the continuous column reactor is 20 wt% (12.2 wt% insoluble)
and the temperature is 48C (118F). The residence time in the continuous reactor is 24 hours. After
this point, the slurry is pumpable and is batched to one of six 1 MM gal vessels (950,000 gal working
volume), where enzymatic hydrolysis continues for another 60 hours. The batch reactors are agitated
and temperature controlled at 48C using a pump-around loop with cooling water heat exchange.

19
The amount of enzyme used (the enzyme loading) is determined based on the amount of cellulose
present in the hydrolysate and the specific activity of the enzyme. Consistent with the 2013 biological
hydrocarbon design report [8], enzyme loading is targeted at 10 mg enzyme protein/g cellulose with a
90% conversion target of cellulose to glucose. This represents a roughly 50% reduction in enzyme
loading and concomitant 12% increase in cellulose-to-glucose conversion relative to NRELs
demonstrated 2012 state of technology pilot-scale trials, whose best-case performance runs
demonstrated 78%82% cellulose to glucose conversion at 19 mg/g enzyme loading, or 89% cellulose
to glucose at 26 mg/g loading [3]. Given ongoing advancements in enzyme preparations, discussions
with enzyme manufacturers, and preliminary NREL data, utilizing new enzyme packages suggest the
target loading of 10 mg/g is not an unreasonable projection within several years; however, an
additional risk is posed in the new shift towards blended biomass feedstocks rather than a single
feedstock (corn stover), which may prove more recalcitrant to enzymatic hydrolysis. In addition, new
enzyme cocktails with increasing xylanase activity may become available to allow for a greater degree
of xylan-to-xylose conversion during enzymatic hydrolysis, taking some of the burden off the
pretreatment section to achieve extremely high levels of xylan deconstruction to xylose (potentially
allowing milder pretreatments at reduced cost). Furthermore, this particular technology pathway does
not require breaking carbohydrates down all the way to monomeric sugars, but also may catalytically
convert soluble sugar oligomers; thus, potentially reducing cost burdens on the biomass deconstruction
operations even further. The target design conditions for enzymatic hydrolysis are summarized in
Table 5.

Table 5. Enzymatic Hydrolysis Conditions

Temperature 48C (118F)
Initial solids loading 20 wt % total solids (12.2% insoluble/7.8% soluble)
Residence time 3.5 days total (84 h)
Number and size of continuous vessels 8 @ 950 m3 (250,000 gal) each
Number and size of batch vessels 6 @ 3,600 m3 (950,000 gal) each
Cellulase loading 10 mg protein/g cellulose

The reactions and conversions taking place during enzymatic hydrolysis are listed in Table 6. As noted
above, although the use of more advanced enzymes may achieve some xylan conversion during
enzymatic hydrolysis, for modeling purposes, all xylan conversion is assumed to be achieved upstream
during dilute-acid pretreatment.

Table 6. Enzymatic Hydrolysis Reactions and Assumed Conversions

Reaction Reactant % Converted to Product
(Glucan)n n Glucose Oligomer Glucan 4%
(Glucan)n + n H2O n Cellobiose Glucan 1.2%
(Glucan)n + n H2O n Glucose Glucan 90%
Cellobiose + H2O 2 Glucose Cellobiose 100%

Once biomass deconstruction is complete, the hydrolysate is processed through a series of

conditioning steps to purify and concentrate the sugars and other soluble carbon components prior to
conversion. Some of these operations are consistent with those utilized in NRELs 2013 biological
design report (namely lignin filtration and hydrolysate concentration), but additional purification steps
are added in this case, given more stringent purity requirements dictated by catalyst tolerance.

20
Lignin is the largest source of insoluble solid residue from the process, along with smaller amounts of
insoluble material from feedstock deconstruction. Insoluble solids can plug catalytic reactors and are
primarily separated from the liquid portion of the hydrolysate by horizontal vacuum belt filters. This
operation includes a wash step to recover soluble sugars carried over into the solids fraction. The
washed, lignin-rich solids are sent to the boiler for combustion. This step is similar to the process
described in the 2013 design report [8] and the vacuum belt filter specifications are summarized in
Table 7. The design of the belt filter was based on extrapolating from preliminary NREL data as well
as input from Harris Group and vendors. The vacuum belt filter was chosen because it was shown to
accommodate a larger solids processing capacity, thus requiring lower total filtration area, relative to
pressure filtration. Further details on trade-offs between pressure filtration, vacuum filtration, and
centrifugation options for solids separation are discussed in the 2013 design report.

As discussed in the 2013 report, the vacuum filter is expected to provide the most cost-efficient
processing operation for hydrolysate solids separation, while maintaining high sugar recovery (up to
99% at a wash ratio of 2.5 L water/L of liquor remaining in the filter cake). Although this is a
relatively high wash ratio and incurs an associated cost penalty, the stipulation for 99% sugar recovery
is also fairly aggressive, and not yet well understood in the context of separating solids in post-
saccharification slurries, which is more challenging than separating prior to enzymatic hydrolysis. The
economic impact incurred by varying the assumption around sugar recovery is considered in the
Sensitivity Analysis section. Alternatively, to achieve sufficient solids separation in this step after 90%
hydrolysis of cellulose to glucose, additional filtration area could plausibly be required relative to the
baseline assumption (depending on the ability of the cake to be filtered). The cost impact associated
with varying the vacuum filter capital cost, attributed to varying the total filtration area, is also
investigated in the Sensitivity Analysis section. Another constraint also imposed on the belt filter in
this design that was not present in the 2013 process is the addition of a recycle stream from a new
downstream microfiltration step (discussed below), consisting of particle fines as small as 0.1 microns.
This could translate to more challenging operability of the belt filter, including more frequent removal
of filter cake and potentially increased replacement maintenance costs for the belt itself. While this
issue requires experimental validation to better understand, for modeling purposes this is addressed by
assuming a 50% purge rate of the microfilter retentate recycle (discussed below), coupled with a
sensitivity analysis considering the addition of periodic belt replacement costs beyond standard
maintenance factors otherwise assumed (see Section 5.2).

Table 7. Vacuum Belt Filter Specifications for Hydrolysate Solids Separation

Number of stages 2
Wash ratio 2.5:1 (L water : L liquor in filter cake)
Capacity (insoluble solids basis) 20 kg/hr-m2 (IS basis)
Soluble sugar recovery 99.0%
Maximum unit size 170 m2
Power demand 490 KW (660 hp) per unit

The clarified filtrate is then pumped from the belt filter to hydrolysate concentration, consisting of a
four-stage mechanical vapor recompression (MVR) evaporation system, which produces a
concentrated sugar stream of 46% (monomeric) sugar, 50% water, and 4% other dissolved solids. The
evaporator design and cost basis is consistent with the 2013 design report [8]. MVR evaporators were
considered to be preferable over standard thermal (i.e., steam-driven) evaporators given preliminary

21
heat and power balance information, where MVR evaporators are primarily electricity-driven rather
than steam-driven. Four effects are required to concentrate the sugars from 14% to 46% (i.e., reduce
water content to 50%). To avoid the possibility of sugar degradation at high temperatures [22], the
evaporators are assumed to be operated under slight vacuum to keep the maximum temperature below
80C (176F). Even so, some fraction of monomeric sugars could revert back to disaccharides or
higher monomers across the concentration step, but this is not anticipated to be a concern here as this
pathway is assumed to utilize oligomers as an additional carbon source in the conversion step. The
vapor exiting the evaporation system is nearly all water (approximately 99.9% purity) and can be
recycled directly to the process water manifold after being condensed by water cooling. This has an
added benefit of reducing the total water flow rate to the WWT section (A600), which reduces cost for
WWT units that are dependent on the hydraulic flow rate. However, it also incurs a loss of convertible
volatile carbon, primarily furfural, constituting the majority of the remainder of the vaporized stream.
Recovery methods may be plausible to recapture the vaporized carbon species for subsequent
conversion in the catalytic step, however the relatively low flow rate of such species in the vapor
product (310 kg/hr) likely does not justify the added cost for such a system. The vendor-provided
evaporation system design specifications are summarized in Table 8.

The monomeric sugar concentration of the filtrate stream from the horizontal belt filter is calculated by
the model to be 14% by weight. The solids product stream from the horizontal vacuum belt filter has a
total solids content of 36% and is conveyed to the boiler. A filtered hydrolysate storage tank provides
20 minutes of intermediate storage of filtrate from the belt filter. A concentrated hydrolysate storage
tank, which provides a residence time of 20 minutes, is located downstream of the evaporator.

Table 8. Hydrolysate Evaporator Specifications

Feed sugar concentration 14 wt %
Product sugar concentration 46 wt % (50% water, 4% other solubles)
Maximum operating temperature 80C
Evaporator technology MVR
Number of effects 4
Electricity usage 3,600 KW (4,830 hp)
Steam usage (low-pressure steam) 1,550 kg/hr (0.8 MMkcal/hr)

Beyond the insoluble solids separation (lignin filtration) and hydrolysate concentration steps, which
were also utilized in the 2013 design report and whose design and cost assumptions were derived from
the 2013 basis, additional hydrolysate purification steps are utilized in this design, given more
stringent specifications for the downstream catalytic operations. First, a polishing filter is included to
protect the downstream catalytic conversion reactors from fouling by residual insoluble solids that
were not removed in the vacuum belt filtration step. The polishing filter consists of four parallel
crossflow microfiltration skids with a pore size of 0.1 microns and other specifications as listed in
Table 9. This is a conservative specification since hydrotreating reactors, which are expected to
operate similarly to the catalytic conversion reactors, typically have suspended particulate size limits
between 1 and 70 microns [23, 24]. In this case, a considerably smaller particle size for microfilter
design was selected to err on the conservative side, given the early level of understanding for the
catalytic reactors required here. At an assumed maximum particle size of 50 microns in the feed to the
microfiltration step (based on typical specifications in the upstream belt filter press operation as
furnished by Harris Group), the vendor suggested that a moderate increase in the pore size of the
polishing filter would have a minimal impact on the price of the system.

22
The microfilters were specified to allow for a soluble solids recovery rate of approximately 90% in the
filtrate. To retain as much sugar as possible in the process, the retentate, which includes the remaining
10% of sugars and other convertible species, is recycled back to the vacuum belt filtration feed.
However, as the retentate contains particle fines down to the 0.1 micron size, recycling this stream in
its entirety could alter the particle size distribution on the belt filter and cause operability issues either
by fouling the belt and requiring frequent replacement, requiring short cycle times, or fouling the
microfilter units, due to a buildup of particle fines in the recycle loop. To mitigate this concern, 50%
of the retentate recycle stream is removed as a purge and routed to wastewater treatment and the
remaining 50% reaches the belt filter step for a subsequent pass. Thus, at a 99% targeted sugar
recovery across the vacuum belt filters as discussed above, the overall net recovery of sugars and other
soluble/convertible carbon species from the initial hydrolysate is still reasonably high at 93.7%, but
lower than the 2013 design case reflecting the increased loss of soluble carbon attributed to more
stringent purification demands. A polished hydrolysate storage tank is located downstream and
provides a residence time of 20 minutes.

Table 9. Polishing Filter Specifications

Parameter Value Units
Total feed flow rate 111,930 kg/hr
Total feed flow rate 356 gpm (1,348 L/min)
Total solids concentration 50% wt %
Soluble solids flow rate 55,585 kg/hr
Feed max particle size 50 micron
Feed total suspended solids (TSS) 706 ppm by wt.
Filtrate max particle size 0.1 micron
Filter type Microfiltration
Filter recovery rate of soluble solids 90 % wt.
Filter metallurgy 316L stainless steel

Next, the polished filtrate is routed to ion exchange for further purification. This is an important step
as certain soluble solids from process- or feedstock-related sources can poison catalysts and diminish
their performance. The two expected sources of catalyst poisons are chemicals and enzymes used in
biomass deconstruction, and components of the biomass feedstock itself. Sulfur from the acid used in
pretreatment has been shown to poison metal-doped catalysts like those used in hydrogenation, Fisher-
Tropsch synthesis, and methanation [25]. Nitrogen present in enzymes and in the feedstock can also
pose a potential issue. Ammonia (e.g., ammonium salts produced during the pretreatment
neutralization step) or organic bases like amines poison acidic catalysts like those used in this design.
Other inorganic anions and cations can poison catalysts as well. For example; nickel, lead, vanadium,
and zinc are poisons for hydrotreating catalysts [25], as are other metals more common to biomass,
such as alkali metals. Additionally, ash components, such as silica, can accumulate and potentially
plug fixed bed catalytic reactors or foul heat exchangers. Any halogen ions like chloride and fluoride
can also corrode process equipment [19].

A separate-bed ion exchange system is utilized to remove a range of ionic species. Two resin bed
trains are includedone for anions and one for cations, with pertinent specifications listed in Table
10. There are two units installed in parallel to allow for regeneration of the resin, which is expected to
be required every 17 hours. Resin is regenerated with acid and caustic, whose estimated usage
requirements are listed in Table 10. While specific components expected to be most problematic for

23
downstream catalysts and their concentration limits were considered proprietary and were not made
available, it is expected that a separate-bed system utilizing both anionic and cationic resins will be
sufficiently flexible to remove a wide variety of ionic impurities as required for downstream catalyst
protection. A deionized sugar storage tank is located downstream and provides a residence time of 20
minutes.

Table 10. Ion Exchange System Specifications

Parameter Value Units

Total feed flow rate 100,766 kg/h
Soluble solids flow rate 50,127 kg/hr
Bed type Separate-bed
Cation Strong acid cation
Anion Weak base anion
Time between regenerations 17.4 hours
Caustic (NaOH) demand for regeneration 12.8 g/kg total soluble solids feed
Acid (HCl) demand for regeneration 22.4 g/kg total soluble solids feed

3.3.3 Cost Estimation

The design and cost basis assumptions for all enzymatic hydrolysis equipment were left unchanged
from the 2011 and 2013 design cases [2, 8], namely empty towers for the continuous hydrolysis
reactor based on a vendor quotation for flat-bottomed plug-flow reactors with a 10:1 height to
diameter ratio, as well as 1 MM-gal batch hydrolysis reactors and agitators. The material of
construction for enzymatic hydrolysis equipment is 304SS. Similar to the 2013 design case, six batch
hydrolysis vessels are utilized in the present model versus eight vessels in the 2011 ethanol report, due
to a reduction in total volumetric throughput into the hydrolysis section associated with the addition of
up-front deacetylation to solubilize and remove a portion of non-convertible biomass components.

The vacuum belt filter design and cost information was provided by a vendor, who provided a design
basis of 170 m2 maximum filtration area per individual filter, with filter sizing based on a filtration
rate of 20 kg/h m2 based on insoluble solids. The cost includes a two-stage countercurrent water wash
step at a wash ratio of 2.5 L/L liquor in the filter cake. A filtered hydrolysate storage tank with a
residence time of 20 minutes is included to provide intermediate storage for filtrate from the belt filter.
The sugar concentration (evaporation) equipment was also based on vendor-provided design and cost
estimates. The system utilizes MVR evaporators in a series of four effects to concentrate the filtered
hydrolysate from 14 wt % to 46 wt % (reducing water content to 50%). A concentrated sugar storage
tank with a residence time of 20 minutes is included downstream of the evaporator. The vacuum belt
filter and the hydrolysate concentration systems as well as associated storage tanks are consistent with
the 2013 design report basis, with costs scaled here according to system throughputs in the present
model.

The polishing filtration system includes four parallel skid microfiltration systems, including CIP and
backwash. Installed equipment cost is $3.4 MM total for the four skids. The ion exchange system uses
a separate-bed configuration, with strong acid cation and weak base anion resins. Estimated installed
equipment cost is $10.0 MM including the resins. Capital costs for the key operations utilized for the
hydrolysate purification and concentration steps are summarized in Table 11.

24
 Table 11. Installed Equipment Costs for Purification and Concentration
Equipment Number of Units Installed Cost
Vacuum Belt Filter 5 $18,331,000
MVR Evaporator System 1 $16,778,000
Polishing Filter 4 $3,361,000
Ion Exchange 1a $10,025,000
a
Ion exchange system consists of parallel cation and anion resin beds, with parallel redundancy provided for
resin regeneration.

Although a TEA assessment on the cost of sugar production (as the final product) is beyond the
intended scope of this effort, an earlier analysis for drawing the product boundary after sugar
production/concentration was presented in NRELs 2011 ethanol design report and the reader is
referred there for that information [2]. To summarize, the 2011 report evaluated a minimum sugar
selling price associated with the front-end operations for lignocellulosic biomass deconstruction to
sugars, followed by lignin/insoluble solids removal and optional sugar concentration. The resulting
sugar selling price was estimated to be 11.6 cents/lb for dilute sugars (127 g/L total sugar
concentration) or 14.3 cents/lb if including concentration via evaporation (487 g/L total sugar
concentration); as noted in the ethanol design report, these prices were in 2007-year dollars. Updating
to 2011 dollars, the same sugar selling price would be 12.7 cents/lb or 16.0 cents/lb for the dilute and
concentrated cases respectively. These values provide some insight into what the intermediate sugar
costs could look like in the present work, although the present design presented here contains a
number of process deviations including (a) the addition of deacetylation, (b) revised pretreatment
reactor costs now as a function of residence time, (c) additional sugar purification steps beyond merely
lignin separation (namely microfiltration and ion exchange), and (d) differences in design basis
assumptions including higher feedstock costs and lower targeted enzyme loadings.

3.4 Area 400: Cellulase Enzyme Production

3.4.1 Overview
This process area produces cellulase enzyme that is used in Area 300 to hydrolyze cellulose into
glucose. Cellulase refers to a mixture of enzymes (catalytic proteins) that includes: (1)
endoglucanases, which attack randomly within the cellulose fiber, reducing polymer chain length
rapidly; (2) exoglucanases, which attack the ends of highly crystalline cellulose fibers; and (3) -
glucosidase, which hydrolyzes the small cellulose fragments (cellobiose, a glucose dimer) to glucose.
Cellulase is produced industrially using (among other microorganisms) T. reesei, a filamentous fungus
that secretes high levels of cellulase enzymes when grown aerobically in the presence of cellulose or
other cellulase inducers.

The present analysis maintains consistency with the fundamental assumptions for enzyme production
and cost estimation as detailed in the 2011 ethanol report and 2013 biological hydrocarbon report,
most importantly, the use of on-site enzyme production rather than a purchased-enzyme model. As
stipulated in prior reports, we again note that by including an on-site enzyme production section,
NREL and DOE are not making a judgment about whether or not the cellulosic biofuel industry should
align to this mode of enzyme distribution. Rather, the model on-site enzyme section is intended to
improve transparency in determining the true cost of cellulase enzymes for large-scale cellulosic
biofuel production.

25
Consistent with earlier design cases, the present design considers submerged aerobic cultivation
(aerobic fermentation) of a T. reesei-like fungus on a feedstock of glucose and fresh water.
Producing cellulase enzymes directly with glucose is not straightforward, because the absence of
cellulose does not encourage the microorganism to secrete cellulase enzymes. Here, we have assumed
a media preparation step where a small fraction of glucose is converted to sophorose, a powerful
inducer of cellulase, using a small amount of the cellulase enzyme itself. When grown on this
substrate, T. reesei has been shown to productively secrete cellulase [26]. Using glucose as the
substrate is not necessarily more expensive than using hydrolysate slurry from pretreatment because
the expected enzyme titer is significantly higher with glucose, reducing the capital and utility costs
compared with using hydrolysate. The whole broth product is transferred to the hydrolysis tanks
without an intermediate enzyme isolation step. Figure 6 is a simplified flow diagram of the enzyme
production section.

Seed train
(Trichoderma reesei)
Glucose + water Glucose/sophorose Air
Sophorose conversion enzyme
Media prep
Holding tank
Aerobic culture
(fed-batch) To hydrolysis

Figure 6. Simplified flow diagram of the enzyme production process

3.4.2 Design Basis

The design for the enzyme production operations is described in detail in the 2011 ethanol report and
will not be repeated in such detail here. In summary, the design for the system was based in part on the
claims in U.S. Patent 4,762,788 (Example 5) [27, 28] in combination with a number of reasonable
assumptions to develop a rudimentary process. The key assumptions used in the current design are
summarized in Table 12.

Table 12. Area 400 Guiding Design Basis Assumptions

Parameter Assumption
Protein loading to enzymatic hydrolysis 10 mg protein/g cellulose
Reactor size 300,000 L @ 80% final working volume
Enzyme titer at harvest 50 g/L
Mass yield of enzyme from glucose 0.24 kg enzyme/kg glucose
Enzyme production cycle time 120 h online, 48 h offline, 168 h total

The assumed cellulase loading to enzymatic hydrolysis is reduced in the present pathway model to 10
mg of enzyme protein per g of cellulose as also assumed in the 2013 biological conversion report,
down from 20 mg/g in the 2011 ethanol report. The rationale for this decision (keeping in mind this is
a target for 20172022) is described above in Section 3.3.2. Here, protein refers to the total amount
of high molecular weight protein in the enzyme broth as determined by assay; not all of this protein is
active cellulase. The total protein demand was thus calculated to be 260 kg/h (570 lb/h). An additional
10% is produced to account for the slipstream provided to the media preparation tank to make the

26
glucose/sophorose mixture. The size of the cellulase production vessels was set at 300 m3 (80,000 gal)
with a height-to-diameter ratio (H/D) of 2. Fermentation is assumed to be a fed-batch process starting
at 50% working volume and ending at 80%. Over one week, each bioreactor will see a 24-hour cell
growth period, a 96-hour protein production period, and a 48-hour offline period for draining,
cleaning, and refilling. With a one-week total cycle time and the production parameters listed in Table
12, one bioreactor is capable of producing 12,000 kg of protein in a week, or 71.4 kg/h (157 lb/h). The
equivalent enzyme volumetric productivity is 0.30 g protein/L-h. Five reactors were therefore required
to deliver the 260 kg/h of protein needed for enzymatic hydrolysis.

The reactors are loaded initially with the glucose/sophorose carbon source and nutrients, including
corn steep liquor (CSL), ammonia, and SO2. After the initial cell growth period, additional substrate is
added to maintain protein production. The bioreactors are sparged with compressed and cooled air and
corn oil is added as antifoam. The reactors are temperature-controlled by chilled water flowing
through internal coils.

Aeration and agitation requirements for the production bioreactors are functions of oxygen transfer
rate and oxygen uptake rate (OTR and OUR), which have been discussed in detail in the 2011 ethanol
report and again in the 2013 report. Thus, the details and governing principles for these metrics will
not be repeated again here. In the modeled bioreactors, the reaction stoichiometry balances the
reactions of substrate, oxygen, ammonia, and SO2 to cell mass and enzyme (plus CO2 and water) using
an elemental composition for commercial cellulase provided by Novozymes [28]. The composition of
cell mass was taken as the average of a generic cell mass composition [29] and the enzyme
composition, with the assumption that cell mass includes some unreleased protein.

Cellulase protein: CH1.59N0.24O0.42S0.01

T. reesei cell mass: CH1.645O0.445N0.205S0.005

In the production bioreactors, it is assumed that 90% of the carbon source is converted via the protein
reaction and 10% is converted via the cell mass reaction. In the seed reactors, 85% of the carbon is
converted via the cell mass reaction and 5% via the protein reaction, with 10% unreacted. This
represents an overall molar selectivity of glucose to 31% protein, 4% cell mass, and 65% CO2,
yielding 0.24 kg enzyme protein/kg glucose. The average aeration rate to the bioreactors remains the
same as in the 2011 and 2013 models at 0.83 standard m3/s (including two days of offline time), as
does the average impeller power P at 260 KW (350 hp), with a maximum of 600 KW (800 hp). The
final specifications for the enzyme production reactors are shown in Table 13.

Table 13. Specifications of the Enzyme Production Bioreactors

Total volume 300 m3/80,000 gal
Maximum working volume 80%
H/D ratio 2
Height 11.5 m
Diameter 5.75 m
Operating pressure 1 atm
Operating temperature 28C (82F)
Material 316SS
Agitator 800 hp
Total electricity demand per kg protein 9 KWh/kg
(air compressors, agitators, chillers, pumps)

27
Four trains of three seed fermentors provide inoculum to the main enzyme production bioreactors.
Each vessel in the seed trains is run batchwise on the same substrate as the production vessels. Air is
also sparged through each of the seed vessels, which are cooled with chilled water. The seed
bioreactors are each sized at 10% of the next bioreactor volume, i.e., 0.3 m3, 3 m3, and 30 m3. The
aeration demand is assumed to be 10% of the production aeration rate. Four trains were chosen
because each production fermentor has a total cycle time of 7 days; each seed fermentor should have a
cycle time of 2 days (including cleaning and sterilization) to get through the cell growth phase only.

Like the oxygen uptake rate, the glucose demand is also computed stoichiometrically from the
required protein production rate. Ammonia and SO2 are fed to the reactors stoichiometrically and CSL,
trace nutrients, and antifoam (corn oil) are added to the substrate based on flow rate. The required
nutrient concentrations are based on Schell et al. [30] and remain the same as presented in prior design
reports. Glucose, the carbon source for cell mass and protein, is the most significant enzyme
production expense in this model. The cost for glucose was maintained consistently with the 2011
ethanol design case (after adjusting to 2011$), but may be revisited moving forward as it may err
slightly on the conservative side. Electricity also remains a significant contributor due to the power
requirements of air injection, agitation, and refrigeration. Total electricity demand for these operations
is shown in Table 13.

3.4.3 Cost Estimation

The cost estimation for all equipment in A400 was left unchanged from the basis values provided in
the 2011 ethanol report and maintained in the 2013 case. Most equipment in this area is stainless steel.
The air compressor and some of the nutrient delivery equipment items are specified as carbon steel.
Quotes for the production bioreactors, internal cooling coils, production agitators and motors, skid-
mounted seed fermentors, and air compressor were provided by vendors through Harris Group, which
developed costs for the pumps in this area using their historical database.

Not included in the enzyme production model are any costs for concentration, stabilization, or
transportation of the enzyme to the plant. At the very least, one expects to have to pay licensing fees
for the cellulase production microorganism, but these costs are not included because we lack
information on what they might be. Additionally, while the design and cost assumptions for the
enzyme production vessels themselves were left unchanged from those described in the 2011 ethanol
report, the 300,000 L vessel size could potentially be on the large side for realistic commercial
operation of a highly viscous system such as that for T. reesei. The implications for variations in
assumed vessel size (e.g., capital cost) are considered in the Sensitivity Analysis section. The enzyme
production system is also sized strictly to provide the amount of enzyme required for the hydrolysis
step (e.g., is not oversized to accommodate occasional poor-yielding runs). A reasonable backup
measure for this risk would be to either over-design the enzyme system or to add supplemental
purchased enzyme capacity. These strategies are not considered here for consistency with the nth-plant
approach taken elsewhere (e.g., avoiding over-design of equipment operations), but may be considered
in future analyses exercising this design model.

Based on the economics of the on-site enzyme section described above, the predicted cost of enzymes
to the facility is $0.21/GGE of total fuel product. As enzyme loading (10 mg/g) and production
design/cost assumptions were maintained identically with the 2013 biological conversion design case,
the resulting enzyme cost on a per-kg basis is the same, e.g., $5.38/kg protein, with consistent cost
contribution breakdowns as presented in the 2013 report (and thus, will not be shown again here).

28
However, the enzyme cost allocation to overall fuel MFSP is lower at $0.21/GGE in this design
relative to the 2013 biological conversion case ($0.37/GGE), given the higher fuel yields calculated in
this case (discussed further in Section 4).

The caveat discussed in the 2011 and 2013 reports bears repeating, that the enzyme cost contribution
modeled here is lower than one would expect for an enzyme preparation purchased from a separate,
non-adjacent production facility. Transportation of the enzyme to the biorefinery facility could add a
non-trivial amount to enzyme costs, even if formulation costs could be avoided. Furthermore, by
lumping the enzyme production equipment in with the biorefinery, some key items are inherently
shared, e.g., the land and buildings, cooling tower, and utilities infrastructure. Overhead and fixed
costs, especially labor and management, would also be higher for a standalone facility. Eliminating the
shared aspects between the enzyme unit and the biorefinery could easily add another $0.05/GGE of
product to the enzyme contribution. Additionally, an external enzyme production facility would
probably demand a higher rate of return than the 10% IRR assumed for the biorefinery plant because it
is a higher-risk and lower-volume business.

Including all these extra costs would likely bring the total enzyme cost contribution in line with
numbers cited publicly (on the order of $0.50/gal for ethanol under todays loading near 20 mg/g as
demonstrated in NRELs 2012 State of Technology [SOT] work, which could translate to $0.25
$0.50/GGE of fuel products in this pathway for 1020 mg/g loading, respectively). Such differences
between off-site and on-site production make an argument for eventual integration and on-site
production. However, in the near term it is more likely that cellulosic biorefineries (whether targeting
sugars, ethanol, or hydrocarbons) will purchase enzyme from an external supplier with an organization
dedicated to improving enzyme performance and reducing costs.

A400 systems and costs in this design were originally designed for performance specifications that are
suitable for microbial conversion of biomass sugars. Most of those conversion processes are intended
to use a microbe that primarily metabolizes monosaccharides. Therefore, there are opportunities to
optimize biomass deconstruction steps, pretreatment, and enzymatic hydrolysis for catalytic
conversion processes. One area for potential cost savings in the current model is that the catalytic
conversion process may tolerate soluble oligomeric sugars [19]. This was not explicitly modeled due
to a lack of publicly available data. However, costs could potentially be lowered by decreasing enzyme
loading or by using lower performance/lower cost enzyme packages.

3.5 Area 500: Catalytic Conversion and Upgrading

3.5.1 Overview
The catalytic conversion process begins after the hydrolysate stream from pretreatment and enzymatic
hydrolysis is processed through upstream filtration, concentration, and purification, leaving a
relatively pure stream of water and convertible carbon species. The purified and concentrated
hydrolysate is converted to fuel/blendstock products (primarily RDB with a lesser amount of naphtha)
through four catalytic reactor stages. Figure 7 presents a high level schematic of the catalytic
conversion and upgrading stages employed. The process begins with a set of two reforming reactors
allocated into hydrogenation and aqueous phase reforming steps, followed by condensation and
oligomerization, and a final hydrotreating step. The goal of these successive catalytic steps is to
remove oxygen or de-functionalize carbohydrates and oligomerize them to primarily diesel-range
hydrocarbons.

29
Each of the catalytic reactors shares a number of similarities. All of the reactors operate at elevated
temperature and pressure, up to 350C and 1,050 psig in the presence of added hydrogen, with
reactions occurring in the liquid and gas phases [31]. Reactors in the base case model described in this
section are supplied with purchased hydrogen from an external source to maximize carbon yield to
fuels. Ex situ hydrogen is assumed to be produced by off-site steam-methane reforming (SMR) from
natural gas. Hydrogen from SMR exhibits a large contribution to operating expenses as well as
carrying sustainability implications, as will be shown later. Making use of externally supplied
hydrogen, the catalytic conversion/upgrading process leads to very good overall carbon efficiency to
fuels, roughly 85%, relative to starting carbon in the purified hydrolysate (discussed further in Section
5.1). However, the fossil CO2 penalty attributed to natural gas-based SMR hydrogen production may
detrimentally impact the overall life cycle of the process given the resulting amount of hydrogen
import required. Information from patent literature has described alternative systems to produce
hydrogen in situ from aqueous phase reforming of biomass hydrolysates [32, 33]. Alternatively, a
fraction of the biomass may also be diverted and gasified to produce hydrogen from syngas [34].
Hydrogen via in situ or gasification production can improve the sustainability profile for the process
relative to SMR-derived hydrogen, but at the cost of fuel yield. While external SMR-based hydrogen
is considered for the base case, the in situ and gasification hydrogen alternatives are presented as
important comparative cases in the Sensitivity Analysis section (see Section 5.2).

Hydrogen

Purified
Hydrolysate
Hydrogenation Aqueous Condensation
Hydrotreater
Reactor Phase Reactor
Reformer

Hydrocarbons

Figure 7. A simplified block diagram of overall catalytic conversion and upgrading process stages

3.5.2 Design Basis

Acid catalysis is a widely used approach for deconstructing biomass, as well as upgrading process
intermediates. Section 3.2 described the well-known approach of dilute-acid pretreatment to solubilize
hemicellulose. Dilute-acid pretreatment is one example of aqueous-phase, acid-catalyzed processing of
biomass. The acid is in solution in pretreatment. In contrast, solid-phase acid catalysts are used in the
upgrading process described here with the reactants in aqueous solution. Acid catalysts are attractive
due to their ability to deoxygenate biomass-derived compounds like carbohydrates [35]. As an
example, consider the acid-catalyzed conversion of glucose to HMF. Three oxygen atoms are
removed from the carbohydrate in this reaction.

(Under acidic conditions): C6H12O6 C6H6O3 + 3 H2O

30
The acidity of solid catalysts can be much greater than mineral acids in solution and is dependent on
factors such as catalyst composition, surface topology, and interaction with reaction solvents [36, 37].
Detailed understanding of the interaction between catalyst active sites, reactants, and water remains
incomplete [35]. However, there have been a number of journal articles and patents published on the
effectiveness of solid-phase acid catalyst for the aqueous-phase conversion of lignocellulose (primarily
based on bench-scale experimental studies with clean commodity sugar substrates) which help to form
the basis for the current design targets [31-33, 35, 38-51]. Aside from patent information, literature or
other data in the public domain remains scarce on catalytic conversion of sugars to fuels, particularly
in the context of cellulosic hydrolysate. As such, details and results presented in this analysis are
limited to a similar level of detail as the underlying data presented in the cited patent literature, e.g.,
discussing yields based on modeled carbon conversions to component classes (alcohols, ketones,
cyclic mono-oxygenates, etc.) as well as carbon number ranges and overall carbon efficiencies as
described in patents, rather than rigorous kinetics or rate expressions as more comprehensive datasets
might allow for.

As designed and costed by Harris Group based on patent information (details discussed below), the
catalytic conversion process of hydrolysate to hydrocarbons consists of four stages (Figure 7):
hydrogenation, aqueous phase reforming, condensation, and hydrotreating. Each stage consists of
packed-bed pressure vessels, clad in 317L SS, and includes internals that support the catalyst and
distribute the process fluid. Hydrogen is added to the reactors which operate at differing process
conditions and catalyst compositions for each reaction step. The composition of products from these
reactors is dependent on the composition of the feed stream, the reaction temperature and pressure,
space velocity, and catalyst type. As modeled, element balance closure is maintained for all conversion
reactors using stoichiometric reactions based on yields presented in literature, with model convergence
based on the use of the NRTL property package given the presence of oxygenated/polar components.
As discussed in A300, catalyst performance can be sensitive to a number of impurities that can poison
the catalyst bed or reduce the performance of the reactors; primary examples include sulfur, cations,
anions, solids, and coke [36]. Purification and filtration described in A300 is intended to minimize the
amount of catalyst poisons and the selected operating conditions are assumed to minimize the
formation of coke in order to maximize catalyst lifetimes and minimize frequency of catalyst
regeneration.

This discussion begins with a focus on the first three catalytic conversion steps (hydrogenation
through condensation), followed by the final hydrotreating operation. A schematic diagram depicting
the envisioned design and stream routing for these first conversion steps is shown in Figure 8. It
should be noted, however, that while this basis was used to assist in developing equipment costing by
Harris Group, some details regarding heat exchange and stream routing were different in the Aspen
model, and are discussed below.

31
 Make-up
Hydrogen
Hydrogen
Make-up
Purified Recycle
Hydrogen Purge Gas To
Hydrolysate Compressor
Compressor Boiler
from Ion
Exchange

Product
Decanter

Aqueous
Hydrogenation Reactor APR Reactor Condensation Reactor Phase to WWT

Organic Phase
Combined Hydrogenation & APR Reactor System as Modeled to Hydrotreater

Figure 8: An overview of the three catalytic reactor stages used to upgrade purified biomass
hydrolysate to longer chain oxygenates and hydrocarbons, as envisioned for commercial production
and utilized for equipment design/costing purposes. (The intermediate products from this stage are sent to
the hydrotreater for final conversion into deoxygenated diesel- and naphtha-range fuels.)

Hydrogenation and Aqueous Phase Reforming

In the hydrogenation step, the feed stream of purified, concentrated hydrolysate is mixed with fresh
and recycled hydrogen from the process, and then heated to the inlet reactor operating temperature of
205C [33]. Hydrogenation reduces sugars to sugar alcohols (e.g., sorbitol, xylitol) or other polyols
(e.g., glycerol, ethylene, or propylene glycol) [31, 32]. Most biomass-derived sugars are aldoses,
primarily glucose and xylose, although ketoses (e.g., fructose) may be found in some feedstocks.
These monosaccharides undergo mutarotation to predominantly form pyranose and furanose ring
structures in solution. These ring structures are less reactive than the open, linear form of sugars [52].
Hydrogenation converts sugars to their corresponding linear sugar alcohol form which has been shown
to be more reactive during APR [31, 32, 53]. Catalyst formulations for the hydrogenation of sugars,
organic acids, furans, and ketones often include rhenium and/or ruthenium on titania, carbon, or Raney
nickel support and may include a variety of transition metal promoters [48, 54-59].

Following hydrogenation reactions, aqueous phase reforming is used to reform oxygenated

hydrocarbons (i.e., polyhydric alcohol or other materials) in water to produce hydrogen, carbon
dioxide, light alkanes, and oxygenates [32, 33, 41, 42, 44, 60, 61]. APR reactions tend to cleave C-C
and C-O bonds. APR catalysts may be multi-functional where acidic supports catalyze dehydration
reactions of oxygenated compounds and hydrogenation reactions are catalyzed by metallic sites
incorporated into the catalysts [31, 32, 39, 40, 46, 49]. Examples of APR catalysts include: tungstated,
titania, or sulfated zirconia; acidic or silica-alumina zeolites; or heteropolyacids alone or doped with
metals including palladium (Pd), platinum, copper (Cu), cobalt, ruthenium (Ru), chromium, nickel, or
silver.

32
One example of an APR reaction is the reforming of sorbitol. APR of sorbitol begins with the
formation of an aldehyde [33]. The aldehyde reacts with water to form H2, carbon monoxide (CO), and
a smaller polyol which can react with hydrogen to form other compounds. The CO can react with
water to form CO2 and H2 by the water-gas shift reaction [32]. On a high level, this general example is
summarized by the following stoichiometry: C6H14O6 C6H12O6 + H2 + 6H2O 6 CO2 + 13H2.
The products from the APR reactor are a variety of oxygenates including alcohols, ketones, aldehydes,
furans, diols, triols, and organic acids which may be upgraded by condensation reactions in the next
reactor stage. Other organic compounds in biomass hydrolysates (e.g., organic acids, furans, and
solubilized lignin) may follow similar reactions to cleaving of C-C or C-O bonds, hydrogenation,
rearrangement, and dehydration [31]. Methanation and Fisher-Tropsch reactions also occur in APR to
produce light hydrocarbons. Off-gases containing light hydrocarbons are separated downstream and
are sent to the boiler by way of a purge split stream from the gas recycle line, which is useful for
providing heat and power to the plant but incurs some loss of potential fuel-range carbon. The outlet
stream from the APR is further heated and pumped to the condensation reactor.

The hydrogenation and APR reactors were designed and priced as separate units, both given their
different functionalities (although different catalysts could also be utilized in different beds of a single
reactor depending on catalyst and support types), as well as for better control over heat management in
a commercial configuration. However, for modeling purposes these two steps were combined into a
single set of reactions and yields. This choice was dictated by the limited degree of data availability in
publications and patent literature from which the model was established. As data from patent literature
was only available for a single reaction step taking sugars to partially deoxygenated APR products
[33], the conditions and yields employed in the model were set accordingly and thus were not modeled
independently with consideration of internal recycle flows or intermediate heat exchange. The
hydrogen consumption of the two reactors was combined, as was the conversion of biomass sugars
and other hydrolysate carbon compounds to oxygenates through the reactors. The hydrogenation
reaction is a relatively mild step intended to chemically reduce sugars prior to reforming, thus could be
viewed as a catalytic preconditioning step that could also take place in the first catalyst beds of a
combined reactor.

Additionally, the heat integration as conducted in the Aspen model utilizes a feed-effluent exchanger
to heat the hydrogenation/APR feed stream up to reaction temperature by cross-exchange with the
condensation reactor effluent (rather than a feed-effluent exchanger only around the condensation
reactor, as indicated in Figure 8). The APR effluent is then further heated to 262C utilizing high-
pressure steam generated from the steam system (A800), such that the condensation effluent
temperature reaches the targeted 300C [33]. The condensation effluent is then further cooled via air-
and water-cooling to a temperature of 30C to maximize fuel-range carbon recovery into the liquid
phase and minimize losses into the off-gas. Thus, the general process concept is similar to the
equipment costing/design basis presented in Figure 8, but with small differences in stream and heat
exchanger placement to best match the available public data for use in establishing the model.

Table 14 lists the composition of the feed stream into the catalytic reactor block (exiting hydrolysate
purification) starting with the hydrogenation/APR reactor system. All carbon-containing components
shown in the table were assumed to be convertible to fuels via the catalytic steps employed [19],
highlighting a key benefit of this pathway in the ability of the catalyst to utilize additional carbon
rather than strictly monomeric sugars, including oligomers, degradation products (furfurals), and
solubilized lignin produced from upstream deacetylation. Of course, this benefit comes with the trade-

33
off of increased costs for more stringent hydrolysate purification requirements to remove solids, ash,
and ionic species than is necessary for biological conversion. As modeled, monomeric and oligomeric
sugars were converted stoichiometrically to a number of representative components (CO2 and
oxygenates) discussed below. Other minor components, including soluble lignin, HMF, and furfural,
were modeled using simplifying assumptions. In this case, minor components were assumed to be
converted to a single representative component (e.g., HMF was converted to a 6-carbon oxygenate) for
simplicity in maintaining carbon balance closures; while some loss of carbon from such minor
components to side reactions and other products is likely, the overall impact on yields or costs would
be marginal as these minor components represent 2% of all convertible carbon in the feed stream to
the conversion section. The total modeled flow of hydrolysate feed is 99,620 kg/hr with 49% of that as
convertible species.

Table 14. Mass Flow of Feed Stream Components into the Catalytic Reactor Section. (All carbon
containing components listed below were assumed to be convertible across the catalytic reactor system and
account for 49% of the total mass flow with the balance as water.)
Component Mass Flow (kg/h)
Water 50,639
Total Convertible Components 48,981
Glucose 27,754
Galactose 1,106
Mannose 464
Xylose 15,289
Arabinose 1,882
Cellobiose 311
Gluco-oligomers 1,066
Galacto-oligomers 27
Manno-oligomers 11
Xylo-oligomers 359
Arabino-oligomers 45
Soluble lignin 492
5-Hydroxymethyl furfural 170
Furfural 6
Total Mass Flow 99,620

The hydrogenation and APR reactor operating conditions employed in the model are listed in Table
15. The operating temperature and pressure for the single reaction step as modeled was set at 254C
and 1,050 psig based on the most relevant set of conditions specified in patent literature [33].
Additionally, the total hydrogen rate in the gas stream (sum of fresh and recycle hydrogen streams)
was set at a molar ratio of 7.9 mol H2/mol reactor feed, based on the same patent data. As the fresh
makeup hydrogen is assumed in the base case to be produced from a nearby off-site SMR plant, the
makeup hydrogen was assumed to be delivered to the facility at a typical SMR product pressure of 18
atm (265 psia), and the makeup hydrogen compressor was designed accordingly based on estimates
provided by Harris Group and a vendor. Weight hourly space velocity (WHSV) was used as the
guiding basis for reactor design as employed by Harris Group. Although todays state of technology
experimental baseline for hydrogenation/APR was presented in the referenced patent literature as 0.7
hr -1 [33], the target case for this value was increased to 1.2 hr -1 for the hydrogenation reactor and 1.0 -
1
for the APR reactor, based on guidance from Harris Group, the vendor, and communication with
Harris Groups industrial contacts. This value translates to a total reactor catalyst bed volume of 2,325
ft3 for the hydrogenation step and 1,441 ft3 for the APR step. As a general note, WHSV values are

34
targeted to increase by roughly 40-70% relative to benchmarks demonstrated in the above-cited patent
literature (from 0.7 to 1.0-1.2 hr -1), coupled with targeted catalyst lifetimes of 1-2 years across all
three conversion reactor steps (see Table 20); while these represent considerable R&D improvements
over currently demonstrated performance which has primarily been on clean commodity sugar
substrates, they are in line with typical values demonstrated in the petroleum industry. Additionally,
WHSV alone was not seen to exhibit strong impact on overall MFSP, as demonstrated by the
sensitivity analysis below in Figure 15.

Table 15. Hydrogenation and APR Reaction Conditions

Overall Modeling Parameters Operating Condition Units
Total Feed Stream Flow Rate 99,620 kg/h (modeled)
Convertible Species in Feed Stream 49% (modeled)
Hydrogen Feed Molar Ratio 7.9 mol hydrogen/
mol convertible feed
Operating Temperature a 205-254 (inlet-outlet) C
Operating Pressure a 1,050 psig
Hydrogenation Reactor Parameters b
WHSV 1.2 hr-1
Total Catalyst Bed Volume 2,325 ft3
Catalyst Type Precious metal on carbon support
APR Reactor Parameters b
WHSV 1.0 hr-1
Total Catalyst Bed Volume 1,441 ft3
Catalyst Type Base metal on mixed oxide support
a
Process conditions for single hydrogenation/APR reactor as modeled
b
Design specifications for separate hydrogenation/APR reactors as designed and costed

As noted above, the combined yields through the modeled hydrogenation and APR reactors were set
based on patent literature data [33]. For this step of the process, the model made use of stoichiometric
reactions to target representative components within each class of molecules, primarily less than C7
carbon chain length. The modeled yields from the hydrogenation/APR reaction steps are summarized
in Table 16 alongside the original patent data guidance, and show close agreement with targeted
yields. In some instances yields fluctuated slightly due to the stoichiometric reactions selected as well
as the composition of the recycle gas stream co-fed to the hydrogenation/APR reactors along with the
hydrolysate material.

35
 Table 16. Product Carbon Yields from the Hydrogenation and APR Reactors. (Data from recent patent
literature and the current model results are provided for comparison [33].)
Component Patent Data (% of feed C)a Model Results (% of feed C)
Carbon Dioxide 2 2
Alkanes 2 2
Total Mono-oxygenates 62 64
Alcohols 15 14
Ketones 16 17
Cyclic Ethers 17 18
Cyclic Mono-oxygenates 15 14
Organic Acids 2 2
Di-oxygenates 9 9
Poly-oxygenates 3 3
Unknown Aqueous 17 20
a
Blommel, P., et al., US20120198760 A1, Method and systems for making distillate fuels from biomass. 2012, Virent Inc.
Madison, WI: United States. Table 9.

Condensation
In the condensation processing step, the intermediates from the previous APR step are reacted to form
longer continuous carbon chains through C-C bond forming reactions. Several types of reactions occur
in this single catalytic step in the presence of water, including dehydration, oligomerization,
cyclization, aromatization, and hydrogenation. The average carbon chain length increases from less
than 6 in the feed to include a substantial amount of molecules within the C8C24 range most suitable
for diesel fuel. Products of this step include normal and iso-paraffins, olefins, ketones, aromatics, and
cycloparaffins [32, 33].

The APR product stream is heated and fed into the condensation reactor. APR, as designed in the
current model, tends to produce oxygenates with shorter carbon chain lengths [33], primarily less than
C7 given similar carbon chain lengths of the starting carbohydrate and other (minor) convertible
components. Aldol condensation is a useful route to couple light oxygenates to form higher molecular
weight compounds, as is the focus here for a diesel-predominant production process (e.g., RDB
targeting carbon chain lengths in the C10-C24 range). Aldol condensation can be catalyzed by solid acid
or solid base catalysts. However, basic catalysts can be poisoned by organic acids in the biomass
hydrolysate feed stream. Acid catalysts are not affected by organic acids and have an advantage when
processing biomass hydrolysates [35, 62]. Examples of acidic condensation catalysts include zirconia,
alumina, silica, aluminasilicates, aluminum phosphates, silica-alumina phosphates alone or doped with
metals including Ga, Ru, Cu, or Pd [32, 35].

In the model the hydrogenation/APR product stream is first further heated to 262C, calculated by the
Aspen model to correspond to a target condensation reactor product temperature of 300C [33]. This
duty is supplied by high-pressure steam generated in the steam section (A800) and removed from the
steam loop prior to being sent through the first turbine stage. The output from the condensation reactor
is cooled to 141C by cross-exchange with the hydrogenation/APR feed stream, then further cooled by
air- and water-cooling to a final temperature of 30C to recover light naphtha-range components, and
then flashed in a three-phase separator. The flash stream containing hydrogen, CO2, and light
hydrocarbons is routed to the recycle gas compressor for recycle through the system, with a purge
slipstream removed to mitigate buildup of produced CO2. The purge split is set to achieve a minimum
90 mol % hydrogen purity after combining with the fresh hydrogen makeup stream (which is sent
through a separate multi-stage makeup compressor prior to combining with the recycle gas and

36
feeding to the hydrogenation/APR feed stream at 1,050 psig). The purge stream is routed to the boiler.
The aqueous phase contains a fraction of miscible or partially miscible carbon components along with
water from the hydrolysate feed, as well as additional produced water as a product of oxygen rejection,
and is sent to wastewater treatment. The organic phase product stream contains mostly C6+
hydrocarbons and mono-oxygenates which are further processed by hydrotreating.

The condensation reactor operating conditions are listed in Table 17. The operating temperature and
pressures were set based on the same above-referenced patent literature source as utilized for the
hydrogenation/APR step [33], namely 300C and 900 psig. While the majority (80%) of the combined
fresh/recycle gas hydrogen stream is fed to the hydrogenation/APR step, 20% of the stream is split and
routed to the condensation step following the hydrogen molar feed ratio stipulated in the same patent
experimental basis of 1.9 mol H2/mol feed to condensation (e.g., product stream exiting APR).
Additional hydrogen is present in the APR product stream, thus the total hydrogen molar ratio is
higher than this basis. Similar to the WHSV values discussed above for the hydrogenation/APR steps,
the WHSV for the condensation step was targeted to increase from a state of technology
experimental baseline of 0.7 hr -1 to a projected value of 1.2 hr -1 based on guidance from Harris
Group, the vendor, and Harris Groups industry contacts. This translates to a total reactor catalyst bed
volume of 2,059 ft3.

Table 17. Condensation Reactor Conditions

Condensation Parameters Operating Condition Units
Hydrogen Molar Ratio 1.9 mol hydrogen/
mol condensation reactor feed
Operating Temperature 262-300 (inlet-outlet) C
Operating Pressure 900 psig
WHSV 1.2 hr-1
Total Catalyst Bed Volume 2,059 ft3
Catalyst Type Precious metal on mixed oxide support

Similar to the hydrogenation/APR reaction, the condensation reaction was modeled based on
stoichiometric equations to target yields presented in literature. Table 18 lists the relative yield split of
intermediate products from the condensation reactor. The modeled results are contrasted with data
from the same recent patent literature as above [33]. The products are listed in broad groups by carbon
number, and consist largely of mono-oxygenates and deoxygenated hydrocarbons. The modeled yields
from condensation correspond closely with targets in patent literature information [33], with remaining
carbon closure allocated to CO2 (4% of convertible feed carbon in the hydrolysate sent to
hydrogenation/APR). The total modeled flow rate of the condensation reactor product stream is
349,850 kg/hr. Of this rate, a large fraction (71%) is flashed into the gas phase with the majority
recycled back to the hydrogenation/APR inlet, and roughly 1% split off as a purge stream. Such a large
recycle flow translates to considerable power demand for the recycle gas compressor to accommodate
the 150 psi (10.2 atm) pressure drop across the recycle loop, at nearly twice the power consumption of
the makeup compressor (more details discussed in Section 3.9). The purge stream flow rate is 2,630
kg/hr, and consists primarily of CO2 (roughly 53 wt %), hydrogen (20%), off-gases (2%), and light
naphtha-range components (25%) whose loss detracts from ultimate fuel yields, but are unavoidable
unless more complex and costly recovery schemes are employed. The aqueous product fraction
comprises roughly 22% of the total condensation product stream, and contains slightly more than 1%
carbon species with marginal to high water miscibility including light (carbon numbers 15) organic

37
acids, alcohols, ketones, cyclic ethers, di- and poly-oxygenates, and other aqueous-partitioning
components (e.g., the unknown aqueous fraction as listed in Table 16 exiting the condensation
reactor). Finally, roughly 6.5% or 22,874 kg/hr of the total condensation product stream is recovered
as the organic liquid product phase and is routed to hydrotreating.

Table 18. Overall Product Carbon Yields from the Condensation Reactor. (Data from recent patent
literature and the current model results are provided for comparison [33].)
Products by Carbon Patent Data (% of feed C to Model Results (% of feed C to
Number hydrogenation/APR)a hydrogenation/APR)
C1-C7 23% 22%
C8-C14 50% 50%
C15-C24 23% 23%
C24+ 1% 1%
a
Blommel, P., et al., US20120198760 A1, Method and systems for making distillate fuels from biomass. 2012, Virent Inc.
Madison, WI: United States. Figure 6.

Hydrotreating
The organic liquid fraction exiting the condensation reactor system is routed to hydrotreating as a final
catalytic upgrading step to complete the deoxygenation to fuel-range hydrocarbons. The design
assumptions and operating conditions for this step were set to be consistent with the hydrotreating
section described in NRELs 2013 design report, given their origins based on vendor-supplied
information for a similar concept focused on deoxygenation to primarily diesel-range fuel/blend stock
products [8]. While it is recognized that a number of compositional details differ for the feed material
to hydrotreating 1, the associated design and cost information provided for the hydrotreating facility in
the 2013 design report [8] is still viewed as the most credible basis available to continue utilizing in
this case, and is still largely appropriate for use here with minor adjustments discussed below. While
full details for the hydrotreating facility are provided in the associated 2013 design report, the process
and key differences will be summarized here.

For the 2013 fatty acid upgrading design, Harris Group consulted with a number of vendors, including
catalyst suppliers and process licensors, to obtain cost estimates for a hydrotreating facility that may
be expected to process a moderately oxygenated (1015 wt % oxygen content) feed into deoxygenated
diesel-range components. While a number of cost estimates were provided from vendors for such a
facility, associated operational specifics were not available without confidentiality restrictions, thus
operating conditions such as temperature, pressure, and hydrogen partial pressure were based on
Harris Group estimates combined with literature values. The primary literature basis utilized was a
report jointly authored by UOP, NREL, Pacific Northwest National Laboratory (PNNL), and Michigan
Technological University [63]. This report described relatively mild hydrotreating temperature and
pressure conditions at 350C and 500 psig, but a high hydrogen feed ratio at 6,000 standard cubic
feet/barrel of feed (this translates to a hydrogen partial pressure of approximately 430 psig at reactor
inlet), where high hydrogen feed is necessary to control the heat generation (exotherm) incurred for
deoxygenation. A reasonable liquid hourly space velocity (LHSV) associated with the original fatty

1
The Davis et al. 2013 design report was based on hydrotreating free fatty acid material with a high TAN
number and potentially higher oxygen content, versus a cleaner feed with lower impurity levels that may look
more similar to standard petroleum hydrotreater feedstocks in the present design.

38
acid upgrading process was estimated to be in the range of 0.91.4 hr-1 as presented in the above-cited
literature basis [63], which was confirmed by an estimated value of 1.2 hr-1 from Harris Group.

A schematic diagram of the hydrotreating process operations is shown in Figure 9. The material is first
heated to near-reaction temperature in a reactor feed-effluent heat exchanger, then further heated to
350C in a furnace which is supplied by a portion of the recycle gas purge split during normal
operation (rather than natural gas or another fossil heat source). The reactor product is then cooled and
split into a gas recycle stream, a liquid product stream, and a small wastewater stream. The 2013
design made use of a pressure swing adsorption (PSA) unit to purify hydrogen from a relatively clean
mixture of hydrogen and CO2. The present design contains higher levels of light hydrocarbon off-
gases in the gas phase mixture beyond hydrogen and CO2, which make the use of a PSA unit
impractical as such components would translate to poor PSA performance and higher cost. Instead, a
fraction (5%) of the recycle gas is split as a purge slipstream to mitigate buildup of CO2 and light
gases. The purge is again split to first satisfy heat demands in the feed furnace, with the remainder
routed to the boiler.

Hydrogen from
Off-site SMR Makeup Off-gases to
Hydrogen Recycle Boiler
Compressor Hydrogen
Comressor

Purge

Makeup Natural
Gas
(for startup and
as needed)
Preheater

Hydrotreating
Reactor
Separator
Produced
Water to WWT

Organic Liquid
Phase from Partial
Off-gas
Condensation Condenser
Feed/Effluent
Reactor Heat
Exchanger Water Naphtha
Air
Separator Cooler
Cooler Product
Distillation Water
Column

Reboiler
Diesel (RDB)
Hydrotreating Unit

Figure 9. Schematic diagram of hydrotreating operations included in vendor design estimate. (All
operations shown are included in ISBL cost estimate for the hydrotreating facility.)

Hydrotreating yields and associated hydrogen consumption were assumed to proceed by

hydrodeoxygenation (HDO) reactions for oxygen removal rather than
decarboxylation/decarbonylation. HDO removes oxygen while leaving all carbon chain lengths intact.
This increases hydrogen consumption but achieves maximum carbon retention efficiency. Given the

39
relatively low oxygen content of the feed stream attributed to the mixture of hydrocarbons and mono-
oxygenates noted above, the amount of oxygen removal and associated hydrogen demand is still
relatively low, even for the 100% HDO assumption. There would be some sustainability trade-offs
between higher hydrogen consumption (HDO pathway rejecting oxygen as H2O) versus lower carbon
yields to products (decarboxylation/decarbonylation pathway rejecting oxygen as CO2 and CO), but
such trade-offs are expected to be minor given the low feed oxygen content as modeled for this case.
The liquid product is fractionated in a distillation column (including heat and cooling demand
considerations in overall heat balances), yielding cuts of diesel- and naphtha-range products, as well as
a light overhead gas stream routed to the boiler. The relative carbon yields to both the diesel and
naphtha streams are reported in Table 19; however as discussed previously, given the lack of
information specifics in patent literature and to avoid unnecessary subjectivity in interpreting this
information in a way that could introduce additional uncertainty into the economic results, the primary
focus for reported yields is on overall total fuel yields, i.e. both diesel and naphtha, translated to
energy or carbon outputs. This said, the analysis does include considerations for capital and energy
(heating/cooling) costs to fractionate the product into these cuts. The primary diesel product cut
exiting the distillation column was also seen to exhibit a boiling point curve that closely resembles
information presented publicly for this technology pathway, also resembling standard petroleum diesel
[64]. The operating assumptions discussed above for the hydrotreating section, as well as modeled
yields and overall hydrogen consumed for all catalytic conversion steps (starting from the APR feed),
are summarized below in Table 19.

Table 19. Summary of Hydrotreating Process Assumptions and Modeled Yields for Overall Process
Hydrotreater feed rate 191,610 gal/day (725 m3/day)
Average reaction temperature 660 F (350 C)
Operating pressure 525 psig (37 atm) at outlet
Hydrogen partial pressure 430 psig (30 atm)/6,000 SCF/bbl
Reactor LHSV ~1.2 hr-1 (estimated) a
Hydrogen loss across purge split 5%
Overall carbon yield to fuel-range products from
85.6% (172,637 GGE/day)
hydrotreater vs. APR feed components
C yield to RDB vs. APR feed components 69.4% (139,962 GGE/day)
C yield to naphtha vs. APR feed components 16.2% (32,675 GGE/day)
Total H2 consumption across all catalytic
6.5% (reaction only; 7.9% total H2 makeup) b
conversion steps, wt % of APR feed components
Total makeup hydrogen rate 38.2 MM SCFD (1.0 MM Nm3/day)
a
Reactor LHSV is not used in the present model. (Total hydrotreater facility cost was used based on vendor
feedback, which implicitly includes hydrotreater sizing which was not provided by vendor.)
b
Value is for consumption across reaction steps only (sum of hydrogenation/APR, condensation, and hydrotreating),
not including additional losses due to purge streams, etc. (increases to 7.9% of APR feed components for total H2
makeup rate).

3.5.3 Cost Estimation

Harris Group developed capital and operating cost estimates for the unit operations utilized in the
catalytic reactor scheme as presented in Figure 8, i.e. the hydrogenation, APR, and condensation
reactor steps, as well as supporting equipment including compressors and heat exchangers. Capital
costs for each new piece of major equipment are based on budgetary pricing from vendors, adjusted to
2011 dollars. Each catalytic reactor in this section (excluding the hydrotreating section) is clad with
3/16 inch 317SS for the material of construction. In general, higher combinations of temperature and

40
hydrogen partial pressure require higher metallurgy. In the base process design, the reactors are clad
with 317L stainless steel for corrosion protection.

Table 20 provides a summary of the catalytic conversion reactor designs. Similar to the constraints
imposed on inclusion of reactor details around the hydrotreating operation as presented in the 2013
design report and maintained in this case (which vendors typically consider to be proprietary), limited
details were made available to NREL attributed to the underlying cost estimates for the hydrogenation,
APR, and condensation reactors. Summarizing on a high level, overall reactor sizing and associated
catalyst demands are dictated by reactor WHSV values for each of the three steps. As discussed above,
based on guidance from Harris Group as well as expertise provided to Harris Group by the vendor and
other industry consultants, reactor WHSV values were targeted to increase from the experimental
baseline of 0.7 hr-1 provided in patent literature [33] to aspirational projections of 11.2 hr-1 for the
three reactor steps. This translates to total reactor catalyst bed volumes of 2,325 ft3, 1,441 ft3, and
2,059 ft3 for the hydrogenation, APR, and condensation steps respectively, based on catalyst loading
densities as shown in Table 20. Based on a 33% allowance for reactor internals per the design
furnished by Harris Group, this equates to a total (internal) reactor volume of 3,470 ft3, 2,151 ft3, and
3,073 ft3 for the respective steps. The resulting installed capital cost estimates for the three steps are
approximately $18.3 MM, $4.7 MM, and $6.1 MM ($29.1 MM in 2011 dollars for the total conversion
system reactors), which includes a 6% allowance factor for operations which were not costed
explicitly by Harris Group, namely pumps and separation vessels. Additionally, Harris Group (with
guidance from vendors) furnished cost estimates for the key heat exchangers employed in the model,
including the feed-effluent exchanger, condensation feed heater, and condensation product coolers; as
well as both the makeup and recycle hydrogen compressors, based on modeled stream conditions.
These items translate to a total installed cost of $23.7 MM.

In addition to reactor design and capital cost estimates for the three conversion steps prior to
hydrotreating, Harris Group also furnished estimates for catalyst cost, lifetime, and time between
regenerations, the net sum of which translates to amortized annual catalyst operating expenses for a
given overall catalyst demand dictated by reactor WHSV targets discussed above. Again, similar to
limitations on reactor design details provided by the vendor, specific catalyst composition details were
considered proprietary and were not provided to NREL, but are consistent with performance expected
to be commensurate with the above-cited patent literature basis which formed the majority of the
modeling assumptions here [32, 33]. In summary, Table 20 presents amortized catalyst operating costs
for the hydrogenation, APR, and condensation reactors including the annual catalyst consumption and
cost per kilogram. The hydrogenation reactors are initially filled with 40,818 kg of precious metal on
carbon support catalyst at a cost of $121/kg (2011$). The estimated catalyst lifetime is 1 year with
regeneration at 6 months, as a targeted nth-plant projection for anticipated target performance based on
guidance from Harris Group and Harris Groups industry contacts. Being based on a carbon support, it
is assumed that the catalyst is regenerated by hot hydrogen stripping techniques as utilized in the
petroleum industry for catalysts which cannot be regenerated by standard combustion regeneration
[65-67]. The APR and condensation reactors are filled with 48,981 kg and 40,818 kg of catalyst
respectively, each with projected target 2-year lifetimes with regeneration every 12 months. As noted
above, the 1-2 year targeted catalyst lifetimes for all three reactor steps correspond with standard
catalyst lifetimes for established refining technologies in the petroleum industry. The APR catalyst is a
base metal on mixed oxide support at a cost of $16/kg (2011$). The condensation catalyst is a precious
metal on mixed oxide support catalyst with a cost of $43/kg (2011$). The catalyst consumption
amortized to an annual basis for each reactor is based on the catalyst loading, time between

41
regenerations, and total lifetime for each catalyst, and translates to annual operating costs of $4.9 MM,
$0.4 MM, and $0.9 MM for the three respective reactors.

Table 20. Catalytic Reactor Design and Cost Summary for Hydrogenation, APR, and Condensation
Reactors
Parameter Units Hydrogenation APR Condensation

Cladding material 317SS 317SS 317SS

Weight hourly space velocity hr-1 1.2 1.0 1.2
Catalyst loading (total) kg 40,818 48,981 40,818
Catalyst bulk density g/mL 0.62 1.2 0.70
Total catalyst bed volume ft3 2,325 1,441 2,059
Reactor internal allowance % 33% 33% 33%
Total reactor (internal) volume ft3 3,470 2,151 3,073
Reactor internals Inlet diffuser, inlet distribution tray, catalyst support grid
Catalyst lifetime Years 1 2 2
Time between regenerations Years 0.5 1 1
Catalyst cost (2011$) $/kg 121 16 43
Initial catalyst charge cost (2011$) $MM 4.9 0.8 1.8
Amortized annual catalyst cost (2011$) $/year 4,940,000 390,000 880,000

As presented in NRELs 2013 design report, the hydrotreating facility cost is an inside battery limits
(ISBL) cost estimate, including costs for reactors, makeup and recycle gas compressors, fired heater,
separation vessels, and distillation [8]. Additionally, while catalyst replacement cost was not explicitly
specified for the hydrotreating section, one vendor noted that the catalyst would be an insignificant
part of the total cost estimate for this particular system, at a typical catalyst lifetime of 2 years
(standard for petroleum hydroprocessing) and the given feed specification. This guidance was based
on a likely more difficult feed in the 2013 design basis, including higher oxygen content, higher total
acid number (TAN), and higher level of impurities; thus, this assumption would be even more accurate
here as catalyst costs would likely be more marginal in this design than the 2013 basis.

Based on a number of separate vendor-supplied cost estimates, combined with Harris Groups
expertise on the subject as well as typical industry costs for petroleum hydroprocessing, an original
hydrotreater cost estimate of $23 MM was used (installed cost) in the 2013 design case, which
translates to $30 MM installed in the present analysis scaling to the increased hydrotreater feed rate
modeled here. The uncertainty range associated with this cost estimate was quoted as -50%/+100%,
and thus is most appropriately viewed as an order-of-magnitude cost for processing this type of mildly
to moderately oxygenated feedstock. The overall cost impact associated with this given range of
uncertainty in the hydrotreating cost estimate is evaluated in Section 5.2, Cost Sensitivity Analysis.
Even with such a relatively high margin of uncertainty in the hydrotreating cost estimate, we feel it is
more prudent to leverage this information than attempt a cost estimate on our own (e.g., using cost
estimation software) as the vendor estimates bring valuable insight as to important considerations such
as reactor design, number of stages, and operating conditions.

42
3.6 Area 600: Wastewater Treatment
3.6.1 Overview
All the wastewater generated in the conversion process is sent to the wastewater treatment system in
Area 600. The treated water is assumed clean and fully reusable by the process, which reduces both
the fresh makeup water requirement and discharge to the environment. This WWT design adopted the
same design used in the 2013 biological hydrocarbon design report using the design from Harris
Group and Brown and Caldwell (a WWT technology vendor). Each unit was costed out on an
individual basis with associated scaling factors both on hydraulic rate and chemical oxygen demand
(COD) load. The simplified flow diagram is shown in Figure 10.

Figure 10. Simplified flow diagram of the WWT process

Condensed pretreatment flash vapor, black liquor from deacetylation, boiler blowdown, cooling tower
blowdown, and the wastewater streams from hydrolysate purification (i.e., ion exchange backwash)
and conversion operations (i.e., aqueous product streams from the condensation reactor flash and
hydrotreater produced water) are mixed together. The combined wastewater stream is processed in
A600 and consists of anaerobic and aerobic digestion, membrane filtration, reverse osmosis,
evaporation, dewatering, and gravity belt thickening. Anaerobic digestion produces a biogas stream
that is rich in methane, so it is fed to the combustor. The aerobic system utilizes a membrane
bioreactor which is essentially an aeration tank followed by an ultrafiltration membrane system with
biomass recycle. The effluent of the membrane system is of sufficient quality to feed a reverse osmosis
system directly (i.e., without additional filtration). The dewatered sludge and biogas are sent to the
boiler in A800. The dewatering process for the anaerobic sludge is centrifugation. The aerobic sludge
is thickened by gravity belt thickeners (GBTs) prior to centrifugation in order to reduce the number of
centrifuges required. In general, centrifugation provides higher solids capture and a drier biomass cake
compared to other dewatering systems thus reducing the size of the biomass burner system. It also
provides a margin of safety against changes in sludge quality which could impact dewatering
processes relying upon filtration (e.g., belt filter press). Reverse osmosis is required to remove
dissolved salts so that cleaned water can be recycled to upstream processes while mitigating salt
accumulation.

This process produces a relatively clean water stream that can be reused in the process. The water
stream is routed to the common process water header (A900) to be used in other process areas, such as
dilution water in pretreatment or enzymatic hydrolysis.

43
3.6.2 Design Basis
The WWT system as designed by Harris Group and Brown and Caldwell includes anaerobic digestion,
aerobic digestion, sludge dewatering, reverse osmosis, evaporation, and centrifugation. Evaporated
water from the sugar concentration step is of high purity (> 99.5% water), and thus is not routed to
A600 for further cleanup but is recycled directly for process water use. This helps reduce the total
hydraulic flow and cost of a number of WWT unit operations. The deacetylation black liquor,
condensed pretreatment flash vapor, ion exchange backwash, condensation/upgrading reactor aqueous
product phases, boiler blowdown, and cooling tower blowdown are mixed together and pumped
through a heat exchanger and cooled to 35C before being sent to the anaerobic digester.

The total COD entering anaerobic digestion in the current design case is approximately 100 g/L
(20,000 kg/h), shown in Table 21. This COD concentration represents a reduction from the 2013
design report (120 g/L), primarily because of improved carbon efficiency in the present catalytic
conversion pathway and less residual unconverted carbon (COD) ultimately reaching this section,
although total volume flow is similar. Along with COD, total (suspended + soluble) solids and
suspended solids have also decreased in concentration in the wastewater feed stream again due to less
unconverted material ultimately reaching wastewater treatment. Total solids in the wastewater feed
include unconverted sugars, extractives, solubilized lignin, salts, and insoluble solids such as ash,
lignin, and cellulose. While the COD concentration has decreased, the wastewater material in this case
has the potential to include more problematic and potentially toxic components for the microbes,
including light oxygenates and hydrocarbons produced in the conversion section; given the lack of
operational data for treatment of wastewater generated in this process, the assumptions for the WWT
system were left unchanged for the time being. Low-rate bulk volume fermentors (BVF reactors) are
used as the anaerobic digester vessels, capable of removing 89% of incoming COD. 85% of the COD
is converted to biogas (methane and carbon dioxide) and 5% is converted to cell mass. Methane is
produced from the organic matter at 281 g/kg COD removed. Carbon dioxide is also produced at a
nearly equimolar rate to methane, such that the biogas from the digester is 51% CH4/49% CO2 on a
dry molar basis; this ratio is lower than typical digester biogas at 60%70% CH4, but is consistent with
the vendor suggestions for such an application. Cell mass is produced at a yield of 58 g cell mass per
kg COD removed. Biogas blowers pull a negative pressure to remove the produced gas.

The preferred use of the biogas is in the on-site combustor for combined heat and power benefits.
However, economic factors may drive the sale of the gas via pipelines. In either case, an emergency
biogas flare is required and designed into the process to burn the biogas in case of a process upset in
the combustor. Sulfate is converted to hydrogen sulfide (H2S) and leaves the system in the biogas.
Since the COD-to-sulfate ratio is approximately 2,000, the methane-producing organisms will out-
compete the sulfate-reducing organisms, so upstream removal of sulfate is unnecessary. The H2S and
any other sulfur species in the biogas are converted to SO2 downstream in the combustor, which is
mitigated using flue gas desulfurization (FGD) (Area 800).

44
 Table 21. WWT System Design Basis

2013 WWT design

This design (catalytic conversion)
(biological conversion)
Hydraulic load 1.6 MMgal/d 1.3 MMgal/d
Wastewater inlet temperature 60C 60C
Total COD 120 g/L 100 g/L
Total solids 130 g/L 120 g/L
Total suspended solids 27 g/L 17 g/L
Total Kjeldahl N 724 mg/L 710 mg/L

The liquid from the anaerobic digester is pumped to the aerobic activated-sludge basin with surface
aerators. In the concrete and steel basin, 96% of the remaining soluble organic matter is removed, with
74% producing water and carbon dioxide and 22% forming cell mass. With the 89% COD reduction in
anaerobic digestion followed by 96% reduction of the remaining COD in aerobic digestion, the total
COD reduction is 99.6%. The fully digested material is pumped to a membrane bioreactor for
clarification. The membrane unit removes additional COD along with colloidal particles (especially
silica). Coarse- and ultra-filtration upstream of the membrane unit separate the aerobic biomass sludge.
The sludge is pumped to a dewatering device to reduce the sludge volume and water content to the
boiler. A gravity belt thickener applied first produces sludge with a solid content of 4%. The
centrifuge units are then periodically operated to provide a solids capture rate of 95% and a cake
concentration of 20%. The centrate flow is recycled with polymer addition to the activated-sludge
aerobic system. Most of the dewatered sludge is recycled to the aeration basin to maintain a high cell
mass loading. The remaining portion is removed to a holding tank, where it is mixed with sludge from
anaerobic digestion. Dewatered sludge is then routed to the combustor in A800. The sludge solid is
primarily cell mass and unconverted biomass components. The treated water is pumped to a reverse
osmosis (RO) membrane system for salt removal. RO produces brine stream containing primarily
sodium nitrate along with all remaining ions and organics. The brine is further concentrated in a
mechanical-vapor-recompression evaporator to 50% solids and the condensate is also recycled to the
process. We do not further process the brine in our model at this time, and it is assumed to be waste.

3.6.3 Cost Estimation

The cost estimation using the 2013 design basis allows for greater detail to adjust individual unit costs
(relative to prior estimates utilized in the 2011 case) either based on hydraulic flow rate or by COD,
thus resulting in improved accuracy (e.g., to accommodate differences in COD concentration even at
similar total hydraulic flow rate). The anaerobic system is scaled based on total COD and the installed
cost in this design is reduced to $24 MM due to lower COD feed rate as discussed above. For the
aerobic digestion step, liquid volume per basin is 4.2 MM gal with mean cell residence time of 15
days. The installed capital cost of the aeration basin is estimated at $10.2 MM here, including capital
costs for the surface aerators and air blowers. The membrane bioreactor, dewatering units, reverse
osmosis system, and evaporators are also scaled based on hydraulic flow rate, contributing another
$13.7 MM capital costs. Influent Total Kjeldahl Nitrogen (TKN) is estimated at 710 mg-N/L in this
design by the Aspen Plus simulation, and it is low enough that a nitrification step is not needed
(similar to the 2013 case as well). This is due to the shift toward increasingly more mild pretreatment
conditions and acid loading (with subsequent lower ammonia neutralization demand), consistent with
the 2013 report. Instead, supplemental ammonia is needed to make up for a slight nitrogen deficit now

45
present in the model to meet the optimum nitrogen demand (922 mg-N/L) for microorganisms to
metabolize organic compounds in the aerobic digestion unit. The resulting ammonia demand in the
aerobic basin is calculated to be 109 kg/h. In addition, polymer is added to the centrifuge to increase
dewatering efficiency, specified at 9 lb polymer per ton total solids feeding into the aeration basin. The
model estimates a polymer loading at 2 kg/h, with operating expenses discussed in more detail in
Section 4.

3.7 Area 700: Product and Feed Chemical Storage

3.7.1 Overview
This portion of the plant provides bulk storage for process chemicals and the fuel products. The
chemicals stored in this area include ammonia, CSL, sulfuric acid, sodium hydroxide, and purchased
glucose for enzymes. Water for fire suppression is also stored here.

3.7.2 Design Basis

Table 22 shows the major storage requirements for the present design. The design and cost bases for
all tanks maintained in the current design were left unchanged relative to the 2013 basis, and are
scaled to a new cost if needed. The only exception is the product storage tanks, whose total storage
volume is increased to 1.2 MM gal for the diesel (RDB) product given a higher RDB production rate
relative to the 2013 basis, as well as the addition of a second product storage tank for the naphtha
product cut, sized at 300,000 gallons; both tanks are sized for seven days of storage. Other
supplemental tanks and pumps were left unchanged relative to the 2011 ethanol design basis, with
more details provided in the 2011 report.

3.7.3 Cost Estimation

The costs for the A700 storage section were left unchanged relative to the 2011 ethanol and
subsequent 2013 biological hydrocarbon design basis (with the exception of the main product tanks as
noted above), and costs are scaled according to new material flow rates to estimate new tank prices if
applicable.

Table 22. Storage Requirements

Material Size (size basis maintained from 2013 biological pathway design report)
Diesel product Sufficient to contain 7 days of production: 1 carbon steel tank @1,200,000 gal
Naphtha product Sufficient to contain 7 days of production: 1 carbon steel tank @ 300,000 gal
Sulfuric acid (93%) 1 carbon steel tank @ 12,600 gal
Caustic (as pure) Duplicated from sulfuric acid tank (includes caustic for deacetylation + IX regeneration)
Ammonia 2 SA-516-70 tanks @ 28,000 gal: ammonia is stored anhydrous at 250 psig
HCl (IX regeneration) Duplicated from sulfuric acid tank
Fire water 4 hours of fire suppression @ 2,500 gpm: 1 glass-lined carbon steel tank @ 600,000 gal
Corn steep liquor 1 glass-lined carbon steel tank @ 70,000 gal
Glucose syrup for A400 Duplicated from CSL tank: 1 glass-lined carbon steel tank @ 70,000 gal
Diammonium phosphate
1 SS304 tank @ 12,800 gal
(DAP)

46
3.8 Area 800: Combustor, Boiler, and Turbogenerator
3.8.1 Overview
The purpose of the combustor, boiler, and turbogenerator subsystem is to burn various organic
byproduct streams to produce steam and electricity. Combustible byproducts include insoluble lignin
and the unconverted cellulose and hemicellulose from the feedstock, biogas from anaerobic digestion,
biomass sludge from WWT, and off-gases from the conversion/upgrading operations. Burning these
byproduct streams to generate steam and electricity allows the plant to be self-sufficient in heating
energy (thermal-neutral), reduces solid waste disposal costs, and minimizes the amount of electricity
that must be purchased from the grid. Sulfur oxide (SOx) emissions are mitigated from the combustor
via flue-gas desulfurization.

The fuel streams are fed to a combustor capable of handling the wet solids. A fan moves air into the
combustion chamber. Treated water enters the heat exchanger circuit in the combustor and is boiled
and superheated to high-pressure steam. A multi-stage turbine and generator are used to generate
electricity. Steam is extracted from the turbine at two different pressures for use in the process, and
additional saturated steam is also extracted prior to the superheater. In the final stage of the turbine, the
remaining steam is taken down to a vacuum and condensed with cooling water for maximum energy
conversion. The condensate is returned to the boiler feed water system along with condensate from the
various process heat exchangers. The steam turbine turns a generator that produces AC electricity to
supply users in the plant. After considering all facility power demands, a net power surplus remains
which is exported to the grid for coproduct revenue.

3.8.2 Design Basis

For the A800 section equipment, the fundamental design and cost basis assumptions as provided by a
boiler vendor employed in the 2011 ethanol design report were maintained in the present model. Thus,
details of the boiler and turbogenerator design will not be repeated here, but can be found in the 2011
report. To briefly summarize, the combustor system features a live-bottom grated fuel bin to ensure
drying and complete combustion of the wet solid fuel, with a boiler efficiency near 80%. Flue gas
from the combustor preheats the entering combustion air then enters a spray dryer absorber for flue
gas desulfurization (FGD) (with the majority of sulfur entering the boiler system in the form of
ammonium sulfate salt introduced upstream during pretreatment and conditioning operations). Ash
and residual lime and calcium sulfate from the spray dryer absorber are removed in a baghouse and
disposed to a landfill. Finally, the boiler raises steam at 875 psig, a portion of which (18%) is split off
for use in satisfying high-temperature steam utility heat demands in the conversion/upgrading process
(namely, heating the feed stream to the condensation reactor as well as the hydrotreater distillation
column reboiler). The remainder is superheated and then sent through a multi-stage steam turbine with
two extraction ports and a final condenser.

Similar to the 2013 biological design case, a number of process parameters are somewhat different for
the boiler and turbogenerator system than the original design basis established in the 2011 report.
Primarily, this includes a relatively high moisture content of 60% in the combined solid feed to the
combustor versus 44% in the original 2011 ethanol model. This is primarily due to the lignin
separation step, which is now placed upstream of the conversion step. This necessitates the use of a
solids wash step to recover carryover losses of sugars and other soluble carbon components into the
solids material from the filter press. This wash step adds more water to the lignin stream, which diverts
more heat toward drying and reduces the amount of heat generated in the combustor. Compared to the

47
2013 biological design case, this process sends less combustible material to the boiler by way of the
liquid sludge stream from WWT (due in turn to higher utilization of carbon in the hydrolysate
including sugar oligomers, solubilized lignin, and furfurals), but also produces more combustible
offgas through the conversion and upgrading process, thus resulting in similar overall heat generation
from the boiler (146 versus 143 MMkcal/hr in the present design and the 2013 biological design
respectively, after accounting for boiler heat losses). This translates to a similar amount of total steam
generated as the biological design case, however as noted above the present design diverts nearly 20%
of the high-pressure steam immediately away from the turbine system for use in high-temperature
utility heat demands required for the thermochemical catalytic operations which was not required in
the low-temperature bioconversion pathway. Consequently, a lower amount of steam is sent through
the turbine stages, which reduces the net power generation from the turbine by roughly 13% (from
53.1 MW to 46.4 MW). Aside from the added steam heating demands for the conversion/upgrading
section, remaining steam demands stay consistent with the 2013 case, namely high-pressure steam
extracted after the first turbine stage for injection to the pretreatment reactor as well as low-pressure
steam extracted after the second turbine stage to provide minor supplemental heating to the MVR
evaporation unit.

After the intermediate extraction point, the remainder of the steam is sent through a final turbine stage
to vacuum pressure, and is condensed at -13 psig (0.1 atm) and pumped back to the boiler. The turbine
generator efficiency is assumed to be 85% as in prior designs. The resulting power generation is 46
MW. The process requires 35 MW of power, resulting in a net power surplus of 11 MW which is
exported to the grid as a coproduct.

3.8.3 Cost Estimation

The cost basis for the A800 equipment remains the same as described in the 2011 and 2013 reports.
Namely, the combustor/boiler system is based on a vendor quote and includes the boiler feed water
preheater, FGD spray dryer, and baghouse. For the baghouse, bag replacement appears as a periodic
charge in the cash flow worksheet. The turbogenerator is also based on a vendor quote for an industrial
generator which remains suitable for this project. Harris Group also obtained quotes from a third
vendor for support equipment including the deaerator, chemical injection system, tanks, and pumps.

3.9 Area 900: Utilities

3.9.1 Overview
Area 900 contains the utilities required by the biorefinery facility (except for steam, which is provided
by Area 800). Area 900 tracks cooling water, chilled water, plant and instrument air, process water,
and the clean-in-place (CIP) system. In the model, Area 900 also tracks the electricity usage
throughout the plant.

The process water manifold in Area 900 mixes fresh water with treated wastewater and condensate
from the sugar evaporation system (assumed suitable for all plant users) and provides this water at a
constant pressure to the facility. Water is provided to the cellulase production unit, boiler and cooling
tower makeup, the CIP system, and the wash for the lignin filter press. Fresh water is also mixed with
some internally-recycled water for dilution before pretreatment and enzymatic hydrolysis. The plant
and instrument air systems provide compressed air for general use (pneumatic tools and cleanup) and
instrument operation. Larger users of compressed air, namely the cellulase system, have their own

48
compressors specified. The CIP system provides hot cleaning and sterilization chemicals to hydrolysis
and the enzyme production section.

3.9.2 Design Basis

The cooling water system is designed for a 28C supply temperature with a 9C temperature rise in
coolers throughout the facility. This is an assumed average rise; the actual cooling water rises across
each exchanger are not explicitly modeled in Aspen. The primary cooling water users in this process
are listed in Table 23. The percentage of cooling duty contributed by each user is shown in Figure 11.
Compared to the 2013 biological pathway design basis, the total cooling water demand has decreased
by 14% from 130 to 112 MMkcal/h, due primarily to reduced steam turbine condenser duty for a
smaller steam flow reaching the condensing turbine (noted above in Section 3.8), as well as reduced
chiller duty and elimination of a bioreactor air cooler which had been required in the 2013 biological
pathway utilizing aerated bioreactors. This is discussed and analyzed in further detail in Section 5.

Table 23. Cooling Water Users

Hydrolysate flash Condenses residual pretreatment flash vapor before it enters WWT. Most of the vapor
condenser is condensed by heat exchange with the incoming dilution water and boiler feed water.

Hydrolysis cooler Cools the pretreated slurry to enzymatic hydrolysis temperature (48C).
Evaporator
Provides cooling of the sugar evaporator condensate.
condensate cooler
Enzyme air
Uses cooling water to cool the compressed air used in enzyme production.
compressor cooler
Conversion/ Trim cooling duty is used in the catalytic conversion/upgrading area, including
upgrading trim distillation column condenser duty, hydrotreating cooling duty, and cooling to reduce
coolers both naphtha and diesel products to 43C.
Condensation rx
Cools the product stream from the catalytic condensation reactor stage prior to flashing.
product cooler
Turbine condenser Condenses the steam turbine exhaust at a vacuum.
The chilled-water loop requires cooling water to condense the refrigerant. The cooling
Chiller condenser
water duty to M-908 is set equal to the total load on the chilled-water loop.

As was the case in prior design models, the largest user of cooling water continues to be the
condensing turbine. Aspen computes the cooling tower evaporation rate based on a temperature drop
from 37C to 28C. It was assumed that windage would be 0.005% of the total flow to the tower. The
tower blowdown was assumed to be 0.15% of the flow leaving the tower basin.

Chilled water is provided by two 2,350-ton Trane centrifugal chillers. Per the chiller spec sheet, the
compressor electricity demand for the chiller was estimated at 0.56 KW/ton of refrigeration. The
cooling water demand for the chiller system was assumed to be equal to the heat removed in the
chilled-water loop. The chiller provides cooling to the enzyme production bioreactor in Area 400. The
total chiller duty has decreased considerably relative to the 2011 and 2013 cases, which both required
chilled water to maintain bioreactor operating temperatures and have now been replaced with high-
temperature catalytic reactors.

Fresh water is assumed to enter the facility at 13C and is used to cool the wastewater entering Area
600 to digestion temperature before entering the process water tank. The fresh water is mixed with the
treated wastewater effluent and the evaporation condensate in the process water tank (T-914) and then

49
split several ways. Clean water must be provided to biomass dilution in the front end, to the cellulase
production unit, to the lignin filter press wash step, and to the boiler and cooling tower water makeup.
The process water tank is designed for an 8-hour residence time. The process water pump (P-914)
pumps water from the tank into the facility and is designed to handle 1.5 times the process water flow
requirement.

The plant and instrument air systems provide compressed air for pneumatic tools and cleanup and
instrument operation (not including major air demands such as enzyme bioreactors). The plant air
compressor is sized for 400 standard cubic feet per minute (SCFM) at 125 psig. An instrument air
dryer and surge tank were designed to provide clean dry air at a consistent pressure to the instrument
air system. The surge tank was sized at 3,800 gal.

Hydrolysis cooler
6%
Chiller condenser Hydrolysate flash
4% condenser
1% Conversion/upgrading
trim coolers
2% Enzyme production
compressor
Condensation reactor 1%
product cooler
11%

Sugar evaporator
condensate
Condensing turbine 7%
68%

Total Cooling Duty = 112 Mkcal/hr

Figure 11. Cooling water heat duty distribution between major users

About 76% of the electricity generated by the boiler in Area 800 is used throughout the plant to power
pumps, agitators, compressors, etc. The surplus is sold to the grid for credit. The distribution of total
plant power utilization among all areas is shown in Figure 12. The conversion/upgrading process area
contributes the most to overall facility power demand, requiring nearly 9 MW of power. This is largely
attributed to high compressor power demand for the recycle gas compressor and to a lesser degree the
makeup hydrogen compressor, given the high hydrogen demand for the process coupled with
relatively high operating pressures (up to 1,050 psig for the conversion step). Note that the cost of the
power required by Area 100 is already assumed to be included in the feedstock cost but must be
subtracted from the plants electricity balance. This is reflected in the economics by an operating cost
credit equal to this amount of electricity.

50
 Total Power Generated = 46 MW
to A800
Boiler/Turbogen
3%
to A900
to A700 Storage Utilities
0% 8%
Excess electricity to
grid
24%
to A600 Wastewater
to A100 Feedstock
13%
Handling
2%
to A500 Conversion +
Upgrading to A200 Pretreatment
19% 17%

to A300 Hydrolysis/
to A400 Enzyme Conditioning/
Production Purification
4% 10%

Figure 12. Distribution of plant electricity utilization by process area

3.9.3 Cost Estimation

All cost estimates for the utility equipment in A900 were maintained consistent with the basis values
used in the 2011 report. To summarize, the cooling tower was based on a cost estimate from a vendor
for a fiberglass cooling tower capable of handling 44,000 gpm; this cost is scaled to the increased
cooling water throughput estimated here. Harris Group provided estimated costs for the cooling water
circulation and makeup pumps from their historical database. The material of construction for the
cooling water piping is carbon steel. The cost for the chiller came from a recent quote for a similarly-
sized system. Harris Group had also used their historical database to estimate costs for the remaining
equipment: the process water tank and pump; the plant/instrument air compressor, dryer, and surge
tank; and the CIP system.

51
4 Process Economics
The ultimate purpose for developing such a detailed process design, simulation model, and cost
estimate is to determine the economics of biofuel production. This information is used either as an
absolute cost to assess the products potential in the marketplace or as a relative cost that can be used
to guide research by examining the change in production cost associated with a process modification
or other core research activity.

The total capital investment (TCI) is first computed from the total equipment cost. Next, variable and
fixed operating costs are determined. With these costs, we use a discounted cash flow analysis to
determine the minimum fuel selling price required to obtain a zero net present value (NPV) with a
fixed internal rate of return (IRR). This section summarizes the assumptions made in completing the
discounted cash flow analysis, with more details and supporting description available in the 2011
ethanol design report [2] and 2013 biological hydrocarbon design report [8] for assumptions that are
unchanged. Our analysis does not take into account any policy factors such as subsidies, mandates, or
carbon credits, because these would be purely speculative. The purpose of this analysis is to
demonstrate the process requirements needed to achieve specific DOE cost targets which are set from
the top down, and to demonstrate how the technology pathway described here is able to achieve such
targets on its own merits (through bottom-up TEA modeling) and, if it cannot, to give policymakers a
sense of the magnitude of incentive required to make it so.

4.1 About Cost-Year Indices

The cost-year of 2011 was chosen for this analysis to provide more updated and relevant cost output
information relative to the 2007-year basis, which had been consistently utilized for a number of years
in prior analyses [2, 68]. This new basis is being applied consistently across all DOE-BETO platforms
for which similar design case target reports are being established during 20132014 efforts, and it is
expected that performance goals and TEA outputs will remain in 2011$ through 2017 to permit
comparison of future feedstocks, conversion technologies, and other alternative scenarios. However,
the present equipment costs were obtained in 2012, 2013, or 2014 dollars for new pieces of key
equipment which were added or replaced in the current design model (as part of the subcontract with
Harris Group), and in 2009 or 2010 dollars for unit operations that were previously provided by Harris
Group and vendors in support of NRELs 2011 ethanol design report and 2013 biological design
report, and are still maintained in the present model. Cost-years for chemicals range from 1999 to
2012.

The methods used for determining MFSP in another years dollar value and for scaling capital,
operating, and labor cost estimates to a desired target year, remain the same as described in the
previous NREL design reports [2, 8, 69]. Thus, the details will not be repeated here, but will be
summarized briefly. Capital costs provided in a year other than 2011$ were adjusted using the Plant
Cost Index from Chemical Engineering Magazine [70] to a common basis year of 2011. The final cost
index for a given year is generally not made available until the spring of the following year. Therefore,
for the small number of equipment items that were quoted in 2014$, we assumed the same Plant Cost
Index value from 2013 (all cost quotes which fall in this category were provided in the first half of
2014). Similarly, for chemical costs we used the Industrial Inorganic Chemical Index from SRI
Consulting [71]. Employee salaries were maintained from 2009$ and were scaled using the labor
indices provided by the U.S. Department of Labor Bureau of Labor Statistics [72]. The general
formula for cost-year dollar back-casting is:

52
 2011 Cost Index
2011 Cost = (Base Cost)
Base Year Index
To maintain consistency with the 2013 and 2011 reports, capital costs and material prices for items
which were unchanged in the present analysis were not modified except to update to 2011 dollars. The
only new feed materials that were added to the present design that were not required in the previous
2013 biological hydrocarbon case were chemicals and catalysts used in the sugar purification and
catalytic conversion steps. The electricity export price, taken to be the average wholesale price as
determined by the North American Electric Reliability Corporation, was left unchanged from the 2013
report at $0.0572/KWh [73-74].

4.2 Total Capital Investment

Section 3 of this report describes the details of the conceptual process design and how the purchased
cost of the equipment was determined. The next step is to determine the installed cost of that
equipment. The installation cost can be determined by performing a detailed study of everything
required to install the necessary equipment and make it operational (e.g., foundation, piping, and
wiring). This type of detail is not warranted at this level of process development, and a factored
approach in which multipliers are applied to the purchased equipment cost is considered satisfactory.
The methodology and rationale for applying unit-level installation costs remain the same as described
in the 2011 and 2013 reports, and again further detail can be found there that will not be repeated here.

In summary, each type of equipment utilizes a different installation factor to scale the given direct
equipment purchased cost to a final installed cost, with these factors generally varying between 1.5
and 3.0. A complete list of the equipment is provided in Appendix A, along with equipment purchased
and installed costs. As described in the 2013 report, a number of recent cost estimates furnished by
Harris Group utilized budgetary vendor quotes for system package costs, whereby a given unit
operation and all of its supporting equipment were quoted under one price; the installation factor for
such packages can be relatively low because much of the engineering is already included in the price.
Additionally, equipment designed as a pre-fabricated skid generally has a lower construction cost.
Also, components that are more highly machined or have higher metallurgy tend to have a higher
purchase price relative to the cost of installation than a less sophisticated component, and therefore
have a lower installation factor.

The purchased cost for a given component reflects a baseline equipment size. As changes are made to
the process, the equipment size required may be different than what was originally designed and
costed. Instead of re-costing in detail, an exponential scaling expression was used:

New Size
New Cost = (Base Cost)
Base Size
In this equation, the scaling exponent n varies depending on the type of equipment to reflect economy-
of-scale dependencies (more detail on reasonable scaling values for different types of equipment is
provided in the 2011 ethanol report [2]). The basis for scaling is typically some characteristic of the
equipment related to production capacity, such as flow or heat duty. Some equipment does not follow
such a scaling-factor approach, namely when the capacity for a given operation is exceeded and
requires multiple units in parallel, thus losing economy of scale benefits which are captured in the
exponential expression above.

53
Once the total equipment cost has been determined in the year of interest, several other direct and
indirect costs were added to determine the total capital investment (TCI). Site development and
warehouse costs are based on the inside-battery-limits (ISBL) equipment costs (Areas 200, 300, 400,
and 500) and are considered part of the total direct cost (TDC). Beyond ISBL Areas 200500, the
other process areas are considered outside battery limits (OSBL), including Areas 100 (rolled up into
feedstock costs) and 600-900. Project contingency, field expenses, home office engineering and
construction activities, and other costs related to construction are computed relative to the TDC and
give the fixed capital investment (FCI) when summed. The sum of FCI and the working capital for the
project is the TCI. Table 24 summarizes these categories and additional factors. The values assumed
for each respective factor were maintained consistently with those discussed in the 2011 ethanol report
[2] as well as the 2013 biological hydrocarbon report [8].

Table 24. Additional Costs for Determining TCI

Item Description Amount
Additional direct costs
Warehouse On-site storage of equipment and supplies. 4% of ISBL a
Site Includes fencing, curbing, parking lot, roads, well drainage, rail 9% of ISBL
development system, soil borings, and general paving. This factor allows for
minimum site development assuming a clear site with no unusual
problems such as right-of-way, difficult land clearing, or unusual
environmental problems.
Additional To connect ISBL equipment to storage and utilities outside battery 4.5% of ISBL
piping limits.
Indirect costs
Prorateable This includes fringe benefits, burdens, and insurance of the 10% of total
costs construction contractor. direct cost (TDC)
Field expenses Consumables, small tool and equipment rental, field services, 10% of TDC
temporary construction facilities, and field construction
supervision.
Home office Engineering plus incidentals, purchasing, and construction. 20% of TDC
and
construction
Project Extra cash on hand for unforeseen issues during construction. 10% of TDC
contingency
Other costs Start-up and commissioning costs. Land, rights-of-way, permits, 10% of TDC
surveys, and fees. Piling, soil compaction/dewatering, unusual
foundations. Sales, use, and other taxes. Freight, insurance in
transit, and import duties on equipment, piping, steel,
instrumentation, etc. Overtime pay during construction. Field
insurance. Project team. Transportation equipment, bulk shipping
containers, plant vehicles, etc.
a
ISBL = installed cost of equipment inside battery limits (A200,300,400,500).

54
 Table 25. Project Cost Worksheet Including TDC and TCI
Process Area Purchased Cost Installed Cost
a
Area 100: Feedstock Storage and Handling $ 15,800,000 $ 26,900,000
Area 200: Pretreatment $ 33,900,000 $ 51,400,000
Area 200: Neutralization $ 1,100,000 $ 2,200,000
Area 300: Enzymatic Hydrolysis, Hydrolysate Conditioning & Purification $ 52,700,000 $ 68,800,000
Area 400: Enzyme Production $ 7,300,000 $ 12,400,000
Area 500: Catalytic Conversion & Upgrading $ 39,600,000 $ 82,900,000
Area 600: Wastewaterb $ 34,400,000 $ 47,700,000
Area 700: Storage $ 2,800,000 $ 4,800,000
Area 800: Boiler $ 42,200,000 $ 76,300,000
Area 900: Utilities $ 4,000,000 $ 7,000,000
Totals (Excl. Area 100) $ 217,900,000 $ 353,500,000
Warehouse 4.0% of ISBL $ 8,700,000
Site Development 9.0% of ISBL $ 19,600,000
Additional Piping 4.5% of ISBL $ 9,800,000
Total Direct Costs (TDC) $ 391,500,000
Prorateable Expenses 10.0% of TDC $ 39,200,000
Field Expenses 10.0% of TDC $ 39,200,000
Home Office & Construction Fee 20.0% of TDC $ 78,300,000
Project Contingency 10.0% of TDC $ 39,200,000
Other Costs (Start-Up, Permits, etc.) 10.0% of TDC $ 39,200,000
Total Indirect Costs $ 234,900,000

Fixed Capital Investment (FCI) $ 626,500,000

Land $ 1,800,000
Working Capital 5.0% of FCI $ 31,300,000
Total Capital Investment (TCI) $ 659,600,000

Lang Factor (TCI/Purchased Equip Cost) 3.4

TCI per Annual Gallon Gasoline Equivalent $11.63/GGE
2011 Dollars
BC1411A
a
Feedstock handling not included in this calculation.
b
Area 600 not included in Lang Factor.

4.3 Variable Operating Costs

Variable operating costs, which include raw materials, waste handling charges, and byproduct credits,
are incurred only when the process is operating. Quantities of raw materials used and wastes produced
were determined using the Aspen material balance. Table 26 documents the costs and sources of
chemicals used in the process and Table 27 summarizes the variable costs on a per-year and per-GGE
basis. As noted above, the cost basis for all material costs in the present model, which were also used
in the 2013 biological hydrocarbon model, were left unchanged.

Hydrogen was assumed here to be purchased as a product from standard natural gas-derived steam
methane reforming (SMR) consistent with the 2013 report which utilized hydrogen for hydrotreating
fatty acids. However, the hydrogen demands required for upgrading the fatty acid material in the 2013

55
 case are substantially lower than total hydrogen demands in the present pathway, so two alternative
on-site hydrogen production options are considered as alternate cases using in situ hydrogen
production from a fraction of the hydrolysate carbon, as well as gasification of a fraction of the
biomass to syngas and subsequently to purified hydrogen. The implications for cost and sustainability
metrics associated with these alternative scenarios were evaluated to compare with externally supplied
hydrogen sourced from natural gas; results for the alternate hydrogen cases are presented in Section 5.
For the base case, the purchased hydrogen price was set based on a recent DOE Hydrogen Program
report, which presented a current price for natural gas-based SMR hydrogen of $1.57/kg (assumed in
2012$) associated with a natural gas price of $4/MM BTU [75], which was applied consistently
following the 2013 biological hydrocarbon report. The overall cost sensitivity to the assumed
hydrogen price is presented in the Sensitivity Analysis section, using a reasonable maximum and
minimum hydrogen cost range as presented in the referenced DOE report, namely $1.10$2.00/kg
associated with a natural gas price range of $2$7.37/MM BTU, respectively (a larger hydrogen price
range is also considered separately in Section 5). Prices for all the raw materials are summarized
below.

Table 26. Chemical Costs and Sources [8]

Component Cost (2011$) Source

Biomass uniform-format feedstock $0.0320/lb 2013 MYPP, $80/dry ton @ 20% moisture [7]
Sulfuric acid, 93% $0.0499/lb Basic Chemical of Omaha via Harris Group
Ammonia $0.2496/lb Terra Industries via Harris Group
Acid (HCl) for IX regeneration $0.0489/lb Harris Group
Caustic (NaOH) for IX regeneration $0.1601/lb Harris Group
Corn steep liquor $0.0316/lb Corn Products via Harris Group
Corn oil (antifoam) $0.6077/lb SRI Chemical Economics Handbook (CEH)
Glucose $0.3230/lb USDA ERS [76]
SO2 $0.1690/lb SRI Chemical Economics Handbook (CEH)
Enzyme nutrients $0.4570/lb SRI CEH (See 2011 design report for details)
Hydrogen $0.6838/lb DOE report, SMR H2 @ $4/MM BTU NG [75]
APR-1 catalyst (hydrogenation) $54.8730/lb Harris Group
APR-2 catalyst (APR reactor) $7.2626/lb Harris Group
Condensation catalyst $19.3670/lb Harris Group
Caustic for WWT $0.0832/lb Brown and Caldwell 2011 WWT design [77]
Polymer for WWT $2.4806/lb Brown and Caldwell 2012 WWT design [78]
Lime $0.1109/lb Harris Group Report 2013
Boiler chemicals $2.7788/lb 2002 Design Report [69]
Cooling tower chemicals $1.6653/lb 2002 Design Report [69]
Fresh water $0.0001/lb Peters & Timmerhaus [79]

56
 Table 27. Variable Operating Costs
Process Usage Usage Cost MM$/yr Cent/GGE
Stream Description $/hour
Area (kg/hr) (lb/hr) ($/ton) (2011$) (2011$)
Raw Materials
N/A Feedstock 104,167 229,688 64.00 7,350.00 57.95 102.18
A200 Sulfuric Acid, 93% 2,240 4,940 99.85 246.62 1.94 3.43
Caustic (as pure) 1,406 3,101 166.42 258.02 2.03 3.59
Ammonia 310 683 499.27 170.55 1.34 2.37
A300 HCl for IX Regeneration 1,122 2,474 97.86 121.04 0.95 1.68
Caustic for IX Regeneration 640 1,410 320.28 225.82 1.78 3.14
A400 Glucose 1,213 2,675 645.94 863.86 6.81 12.01
Corn Steep Liquor 83 182 97.86 8.90 0.07 0.12
Corn oil 7 15 1,215.42 8.98 0.07 0.12
Ammonia 58 127 499.27 31.75 0.25 0.44
Host Nutrients 34 75 913.94 34.05 0.27 0.47
Sulfur Dioxide 8 18 388.08 3.07 0.02 0.04
A500 Hydrogen 3,847 8,483 1,367.51 5,800.17 45.73 80.63
APR-1 Catalysta 5 11 109,746.10 626.42 4.94 8.71
APR-2 Catalysta 3 7 14,525.22 49.75 0.39 0.69
Condensation Catalysta 3 6 38,733.92 110.54 0.87 1.54
A600 Ammonia 109 239 499.27 59.78 0.47 0.83
Polymer 2 5 4,961.25 11.16 0.09 0.16
A800 Boiler Chems 0.2 0.5 5,557.64 1.42 0.01 0.02
FGD Lime 180 397 221.90 44.04 0.35 0.61
A900 Cooling Tower Chems 3 6 3,330.66 10.09 0.08 0.14
Makeup Water 159,026 350,653 0.29 50.51 0.40 0.70
Subtotal 16,086.56 126.83 223.63
Waste Disposal
A800 Disposal of Ash 4,506 9,936 35.39 175.84 1.39 2.44
Subtotal 175.84 1.39 2.44
By-Products and Credits
Grid Electricity 11,045 KW $0.0572/KWh 631.43 4.98 8.78
Area 100 Electricity 859 KW $0.0572/KWh 49.13 0.39 0.68
Subtotal 680.57 5.37 9.46
Total Variable Operating Costs 15,582 122.85 216.62
a
Catalyst usage amortized to kg/hr-basis for consistency with rest of table. APR-1 = hydrogenation reactor, APR-2 = APR
reactor

4.4 Fixed Operating Costs

Fixed operating costs are generally incurred in full whether or not the plant is producing at full
capacity. These costs include labor and various overhead items. The assumptions on fixed operating
costs were maintained consistently from the 2011 design basis (after updating to 2011$), which in turn
were based in large part on NRELs 2002 ethanol design report [69] and/or Peters and Timmerhaus
[79].

Table 28 shows the recommended number of employees and associated salaries. The number of
employees was estimated by considering the likely degree of automation for each area and adding a
reasonable number of management and support employees. Details behind the assumed number of
employees and associated salaries are provided in the 2011 ethanol report. Because the model
feedstock is predominately corn stover, salaries were estimated for rural regions of the U.S. Midwest
(Iowa, Missouri, etc.). These estimates may vary depending on location. While beyond the scope of
this analysis, some economy of scale advantages could further be gained with respect to labor costs by
considering multiple units at the same site rather than a single stand-alone unit, such that some

57
overlapping positions could be shared (i.e. one plant manager); however, such savings would likely be
marginal in comparison to overall facility costs.

Table 28. Fixed Operating Costs

2011 # MM$/yr
Position 2011 Cost Cent/GGE(2011$)
Salary Required (2011$)
Labor and supervision
Plant Manager 155400 1 $155,000
Plant Engineer 74000 2 $148,000
Maintenance Supr 60257 1 $60,000
Maintenance Tech 42286 12 $507,000
Lab Manager 59200 1 $59,000
Lab Technician 42286 2 $85,000
Lab Tech-Enzyme 42286 2 $85,000
Shift Supervisor 50743 4 $203,000
Shift Operators 42286 20 $846,000
Shift Oper-Enzyme 42286 8 $338,000
Yard Employees 29600 4 $118,000
Clerks & Secretaries 38057 3 $114,000
Total Salaries $2,719,000 2.72 4.79
Labor Burden (90%) $2,447,000 2.45 4.31
Other overhead
Maintenance 3.0% of ISBL $6,530,000 6.53 11.51
Property Insurance 0.7% of FCI $4,385,000 4.39 7.73
Total fixed operating costs 16.08 28.36

A 90% labor burden is applied to the salary total and covers items such as safety, general engineering,
general plant maintenance, payroll overhead (including benefits), plant security, janitorial and similar
services, phone, light, heat, and plant communications. The 90% estimate is the median of the general
overhead range suggested in the 2008 PEP Yearbook produced by SRI Consulting [71]. Annual
maintenance materials were estimated as 3% of the installed ISBL capital cost and property insurance
and local property tax were estimated as 0.7% of the fixed capital investment, based on the 1994 Chem
Systems report described in NRELs 2011 ethanol report. These factors are all consistent with those
used in the 2011 and 2013 design reports.

4.5 Discounted Cash Flow Analysis and the Minimum Selling Price of
Fuel
4.5.1 Discount Rate
For this analysis, the discount rate (which is also the IRR in this analysis) was set to 10% and the plant
lifetime was set to 30 years. The discount rate was also used in previous design reports and was based
on the recommendation in Short et al. [80] on how to perform economic evaluations of renewable
energy technologies for DOE. His view was that, In the absence of statistical data on discount rates
used by industrial, transportation and commercial investors for investments with risks similar to those
of conservation and renewable energy investments, it is recommended that an after tax discount rate of
10%be used.

58
4.5.2 Equity Financing
For this analysis, it was assumed that the plant would be 40% equity financed. The terms of the loan
were established at 8% interest for 10 years. The principal is taken out in stages over the 3-year
construction period. Interest on the loan is paid during this period, but principal is not paid back (this
is another nth-plant assumption, which says that this cash flow comes from the parent company until
the plant starts up). This is all consistent with the assumptions used in the 2013 biological hydrocarbon
and the 2011 ethanol reports. Figure 13 illustrates the sensitivity of MFSP to the percentage of equity
financing and the after-tax discount rate (the IRR).

$6.50
Minimum Fuel Selling Price ($/GGE)

10% IRR
$6.00 15% IRR
20% IRR
$5.50

$5.00

$4.50

$4.00

$3.50
0% 20% 40% 60% 80% 100%
Equity %

Figure 13. Sensitivity of MFSP to IRR and % equity. (8% interest on a 10-year loan.)

4.5.3 Depreciation
To determine the capital depreciation amount for the calculation of federal taxes to be paid, we used
the IRS Modified Accelerated Cost Recovery System (MACRS). Within the MACRS system is the
General Depreciation System (GDS), which allows both the 200% and 150% declining balance (DB)
methods of depreciation. This offers the shortest recovery period and the largest tax deductions.
According to IRS publication 946 [81], a cellulosic biorefinery plant would fall under Asset Class
49.5, Waste Reduction and Resource Recovery Plants. This class uses a 7-year recovery period, not
including the steam plant equipment, which has a 20-year recovery period (Asset Class 49.13). IRS
publication 946 contains a special provision for cellulosic biofuels plants that allows them to write off
50% of the capital investment in the first year. This was not implemented in our cost model because it
does not ultimately affect the MFSP (although the provision affects the cash flow in the first few years
of the analysis, it does not change the year in which the plant goes into the black and must start paying
taxes).

4.5.4 Taxes
The federal corporate tax rate used in our analysis is 35%. Income tax is averaged over the plant life
and that average is calculated on a per-GGE basis. The amount of income tax to be paid by a potential
fuel producer varies annually due to changes in the volume of product produced and the allowable
depreciation deduction. In fact, no income tax is paid in the first eight years of operation because the

59
depreciation and loan interest deductions are greater than the net income. State taxes are not
considered, primarily because the location of the plant has not been determined and tax rates vary from
state to state (from 0% to 12%).

4.5.5 Construction Time

The construction time is important to the cash flow analysis because no income is earned during
construction, but huge sums of money are being expended. Construction time assumptions were left
unchanged from the 2011 and 2013 design basis assumptions. Perry and Green [82] indicate that small
projects (less than $10 million investment) can be constructed in fewer than 18 months and that larger
projects can take up to 42 months. An overview of petroleum refining economics indicates that large
refineries (on the order of $1.5 billion investment) can be constructed in 24 months [83]. Certainly this
facility is much smaller than a petroleum refinery, so using a construction time of 24 months fits
within these references, although an important difference between this type of facility and a refinery is
the large number of field-erected vessels. These are constructed on-site and have a longer construction
time than if the tanks were delivered finished. Table 29 summarizes the schedule for construction and
the cash flow during that time. Twelve months are added before construction for planning and
engineering.

Table 29. Construction Activities and Cash Flow

Project Project % of
Start End Activity Description Project
Month Month Cost
0 12 Project plan and schedule established; conceptual and basic design 8%
engineering, permitting completed. Major equipment bid packages issued,
engineering started on selected sub-packages, P&IDs complete, preliminary
plant and equipment arrangements complete.
12 24 All detailed engineering including foundations, structure, piping, electrical, site, 60%
etc., complete; all equipment and instrument components purchased and
delivered; all site grading, drainage, sewers, rail, fire pond, foundation, and
major structural installation complete; 80% of all major process equipment set
(all except longest-lead items), all field fabricated tanks built, and the majority
of piping and electrical materials procured.
24 36 Complete process equipment setting, piping, and instrumentation installation 32%
complete; all electrical wiring complete; all building finishing and plumbing
complete; all landscaping complete; pre-commissioning complete; and
commissioning, start-up, and initial performance test complete.
TOTAL 100%
Note: The above assumes no utility or process equipment orders placed prior to month seven. Expenditures are based on a
typical 60 MMgal/yr grain-to-ethanol facility.

4.5.6 Start-Up Time

Perry and Green [82] indicate that for a moderately complex plant, startup should be about 25% of the
construction time, or 6 months in this case. Similar to 2013 design basis, a start-up time of 6 months
under an nth plant assumption was assumed in this analysis. The start-up period is not completely
wasted, however. We expect that an average of 50% production could be achieved during that period
while incurring 75% of variable expenses and 100% of fixed expenses.

4.5.7 Working Capital

Peters and Timmerhaus [79] define working capital as money available to cover (1) raw materials and
supplies in inventory, (2) finished product in storage, (3) accounts receivable, (4) cash on hand for
monthly payments such as wages and maintenance supplies, (5) accounts payable, and (6) taxes

60
payable. The present analysis applies the same basis for working capital as was used in the 2011
ethanol report, namely 5% of FCI; this translates to a slightly higher working capital cost value given
the increased direct capital costs and FCI in the present model.

Table 30 summarizes the parameters used in the discounted cash flow analysis. Using these
parameters, plus the cost information in Table 25, Table 27, and Table 28, the resulting MFSP of total
fuel products is $4.05/GGE (2011$), including both naphtha- and diesel-range product cuts adjusted
by heating values (calculated in the Aspen model) to gasoline-equivalents. Table 31 summarizes the
yields and conversion costs for the present design. According to the methodology of Cran [84], the
expected accuracy of the overall TCI analysis is 25% (although some specific pieces of equipment
carry a higher degree of uncertainty in underlying cost quotations as identified above). If we apply this
uncertainty to the TCI, the impact on the cost of total fuel is $0.40/GGE. The complete discounted
cash flow summary worksheet is shown in Appendix B. The MFSP can be further broken down into
the cost of each process area. Figure 14 illustrates the contribution to the overall cost by process area
and capital, operations, and fixed costs (the bar for feedstock plus handling reflects the single
feedstock cost of $80.00/dry U.S. ton delivered to pretreatment and has not been decomposed).

Table 30. Discounted Cash Flow Analysis Parameters

Plant life 30 years
Discount rate 10%
General plant depreciation 200% declining balance (DB)
General plant recovery period 7 years
Steam plant depreciation 150% DB
Steam plant recovery period 20 years
Federal tax rate 35%
Financing 40% equity
Loan terms 10-year loan at 8% APR
Construction period 3 years
First 12 months expenditures 8%
Next 12 months expenditures 60%
Last 12 months expenditures 32%
Working capital 5% of fixed capital investment
Start-up time 6 months
Revenues during startup 50%
Variable costs incurred during startup 75%
Fixed costs incurred during startup 100%

61
 Table 31. Summary of Yields, Rates, and Conversion Costs

Feedstock rate 2,205 dry U.S. ton/day

Online time 7,884 h/yr (90% online factor)
Total fuel yield 78.3 GGE/dry U.S. ton feedstock
Total fuel production rate 56.7 MM GGE/yr
Diesel-range production rate 46.0 MM GGE/yr
Naphtha-range production rate 10.7 MM GGE/yr
Total equipment cost $354 MM
Total capital investment (TCI) $660 MM
TCI per annual gallon $11.63/GGE
Minimum Fuel Selling Price $4.05/GGE
Feedstock contribution $1.02/GGE
Enzyme contribution $0.21/GGE
Non-enzyme conversion contribution $2.81/GGE

Capital Recovery Charge Raw Materials & Waste Process Electricity

Grid Electricity Total Plant Electricity Fixed Costs

Feedstock + Handling: 102.2

Pretreatment/Conditioning: 44.2

Hydrolysis, Hydrolysate Conditioning: 45.3

Cellulase Enzyme: 21.3

Conversion + Upgrading: 142.5

Wastewater Treatment: 31.3

Storage: 2.6

Boiler/Turbogenerator: 8.0 $4.05/GGE

MFSP
Utilities: 7.4

-$0.60 -$0.40 -$0.20 $0.00 $0.20 $0.40 $0.60 $0.80 $1.00 $1.20 $1.40 $1.60

Figure 14. Cost contribution details from each process area. (Per GGE total fuel products.)

Figure 14 highlights a number of clear differences relative to NRELs 2013 design report for
biological conversion of sugars to hydrocarbons. Primarily, it presents a rare case where feedstock is
not the primary cost driver at $80/ton (translating to a cost contribution of $1.02/GGE to overall
MFSP), but is surpassed by the conversion/upgrading section contributing $1.43/GGE or 35% to
overall MFSP. In turn, this is driven by purchased hydrogen cost, contributing $0.81/GGE attributed to
the total hydrogen demand of 7.9 wt % (net, including purge losses) relative to the total convertible
component feed rate to the conversion system. Although purchased hydrogen is clearly a key cost
driver, because it is derived from low-cost external natural gas it is ultimately more economical to

62
source hydrogen this way than to produce it in situ via reforming a fraction of biomass hydrolysate or
by diverting a fraction of biomass feedstock to gasification to produce hydrogen from syngas, and
taking a yield reduction penalty accordingly (i.e., sourcing hydrogen from biomass which is much
more costly than natural gas). Details for this trade-off are presented in Section 5.2.2. As a point of
reference, the 2013 biological design case required hydrogen at a rate that equated to $0.08/GGE
contribution to overall MFSP, as hydrogen was only required during the hydrotreating step to upgrade
fatty acids to RDB but not in the production of the fatty acid intermediate itself from sugars. Or, to
state another way, in the 2013 biological case, oxygen from biomass carbohydrates is rejected as CO2
inherent to biological metabolism (resulting in lower carbon retention to fuels), while in the present
design, oxygen is primarily rejected as water with the addition of hydrogen (resulting in higher
carbon-to-fuel yield but higher hydrogen demand). Indeed, overall fuel yield is roughly 78 GGE/ton in
the present pathway associated with 85.6% carbon retention efficiency from hydrolysate, compared
with 45 GGE/ton and 49.5% carbon efficiency from sugars to product in the 2013 biological case.

Beyond costs for conversion and feedstock, the next highest cost drivers are hydrolysis and
hydrolysate conditioning/concentration as well as pretreatment/neutralization, each contributing a
similar amount to MFSP at $0.45/GGE and $0.44/GGE respectively. Both process areas are driven
primarily by capital expenses, in turn associated with the hydrolysate purification operations and the
pretreatment reactor respectively. Similar to the 2013 case, enzyme costs are not a large burden on the
overall MFSP, at the sixth highest cost driver of the nine process areas considered. This result is driven
by the low enzyme loading target of 10 mg/g cellulose; alternative enzyme loadings are considered in
the Sensitivity Analysis section.

While the analysis presented here demonstrates a favorable result for both fuel yields and economics,
with the base case MFSP of $4.05/GGE well on the way to the final 2022 target of $3/GGE, it may do
so at the expense of sustainability given the large amount of natural gas-derived hydrogen utilized for
this case. A full well-to-wheel life cycle assessment (LCA) evaluation of the process (outside the
scope of this work) must be conducted to understand the overall greenhouse gas profile of the process
and associated implications for Renewable Fuel Standard (RFS) qualifications [85], but to frame the
outermost boundaries for the options, the alternatives based on producing 100% of the required
hydrogen internally via in situ reforming of hydrolysate or routing a fraction of biomass to a
gasification train are considered below in Section 5.2.2. A plausible scenario could in fact fall between
the boundaries, e.g., sourcing as much hydrogen as possible from external SMR-derived natural gas to
still qualify for RFS criteria, while sourcing the remainder internally by an alternative route such as in
situ or gasification. However, it is acknowledged that the purchased SMR hydrogen base case likely
leaves more room for optimization if an alternative scenario were considered for an on-site integrated
SMR plant, which would allow for heat/power integration opportunities as well as benefits associated
with routing light gases produced elsewhere to the SMR rather than burning them in the boiler as in
the present design for off-site SMR production. Considering any of the hydrogen scenarios, further
improvements would be required to ultimately reach final 2022 cost goals of $3/GGE; a brief
discussion on paths toward achieving this final goal is presented in Section 5.4.

63
5 Analysis and Discussion
5.1 Carbon and Water Balances
Table 32 shows the overall flow of carbon inputs and outputs, with a carbon balance closure very near
unity. As shown in the table, roughly 99% of all carbon in the process enters in the biomass feed, with
small amounts of additional carbon coming from glucose for enzyme production. Given the low target
enzyme loading (10 mg/g cellulose), the glucose contribution to total input carbon flow was only
1.3%.

Table 32. Biorefinery Overall Carbon Balance

% of Carbon
Stream Carbon Flow (kmol/h)
Flow
Carbon inlets
Biomass feedstock 3,117 99%
Glucose 40 1%
A400 nutrients <1 0%
Total 3,158 100%
Carbon outlets
Diesel-range (RDB) product 1,145 36%
Naphtha-range product 267 8%
Combustion exhaust 1,629 52%
WWT brine 19 1%
Aerobic lagoons 58 2%
A400 vent 26 1%
Combustor ash 8 0%
Hydrotreater furnace flue gas 3 0%
Total 3,155 100%

Of the carbon inputs to the process, a total of 44.7% of carbon atoms leave as total fuel products
(either naphtha- or diesel-range hydrocarbons) after hydrotreating and product distillation. The major
exit point for the remainder of the carbon is primarily the combustor stack (52%). We expect the
amount of carbon exiting in the combustion exhaust to be rather large because most byproducts
(lignin, other unconverted sugars, etc.) of this process are burned to form CO2 with a small amount of
CO2 also vented from the aerobic basin in the WWT section.

In addition to the carbohydrate components of the biomass (both monomeric and oligomeric sugars),
other components such as furfural, HMF, solubilized lignin, and acetate may also be convertible by the
catalytic process (although furfural and HMF themselves are degradation products from starting
carbohydrates). Of the carbon in the feed stream, 1,857 kmol/h (59.6%) are carbohydrates (glucan,
xylan, arabinan, mannan, galactan, and sucrose); 690 kmol/h (22.2%) are lignin, and the remaining
570 kmol/hr (18.2%) are the sum of acetate, extractives, and protein. Theoretically all the components
with carbon could plausibly be converted to hydrocarbons using appropriate catalysts and reaction
conditions. In the case of lignin, this would first require solubilizing the lignin which is possible for a
fraction of lignin but unlikely to be practical for 100% of the starting lignin in the biomass feed.
According to Table 32, 1,412 kmol/h of the feed stream carbon is actually converted by the catalytic

64
process, so the actual carbon yield is 45.3% of the biomass feedstock, without considering the carbon
footprint associated with externally supplied hydrogen. When also including carbon in off-site natural
gas SMR associated with the required hydrogen utilization for catalytic conversion/upgrading, overall
carbon yield decreases to 35.7% based on a translation of 0.44 mol natural gas feed to SMR per mol
H2 produced [86]. The carbon retention efficiency is 85.6%, relative to carbon in the hydrolysate feed
stream to the catalytic reactor system (after lignin/solids separation and hydrolysate
concentration/purification). Further improvements to carbon efficiency from biomass feedstock could
be achieved through improving pretreatment and enzymatic hydrolysis technologies to solubilize more
carbon components, such as lignin, while minimizing loss of extractives and carbohydrates.

The overall flow of water throughout the model is presented below in Table 33. The primary point of
entry for dedicated process water demand is the makeup water stream, shown in the current model to
be 159,026 kg/h or 5.8 gal/GGE hydrocarbon product. On an absolute flowrate basis, this is similar to
the 2011 ethanol report demand of 147,100 kg/h (5.4 gal/gal or 8.2 gal/GGE) but considerably lower
than the 2013 biological hydrocarbon report demand of 217,100 kg/h (14.4 gal/gal or 13.7gal/GGE).
This is primarily due to a lower cooling water usage and associated lower cooling tower losses
(typically the largest consumer of facility process water), due in turn to a lower amount of steam sent
through the steam turbine condenser (typically the largest contributor to overall cooling water
demands). Specifically, total cooling water demand in the present design is 112 MMkcal/h, compared
to 97 MMkcal/h in the 2011 ethanol design and 130 MMkcal/h in the 2013 biological hydrocarbon
design. Furthermore, a substantial amount of water is generated in the catalytic conversion reactions of
the present design model (associated with high hydrogen consumption rejecting oxygen as water),
which ultimately is routed to wastewater treatment and made available for subsequent recycle to the
process. Finally, on a per-GGE basis, water consumption also appears lower than the 2011 and 2013
designs given the much higher product yield in this case (approximately 78 GGE/ton in this model,
versus 52 GGE/ton in the 2011 ethanol report or 45 GGE/ton in the 2013 biological hydrocarbon
case).
Table 33. Biorefinery Water Balance

Water Balance
Inputs kg/hr gal/GGE Outputs kg/hr gal/GGE
Moisture in feedstock 20,833 0.76 Water in fuel product 5 0.00
Water in glucose syrup 214 0.01 Cooling tower evap. 158,773 5.82
Stripped in enzyme
Water in raw chemicals 198 0.01 388 0.01
aeration
Generated in enzyme prodn. 648 0.00 Consumed in WWT 332 0.01
Generated in combustor 29,086 1.07 Consumed in pretreatment 2,394 0.09
Enzyme air intake 320 0.01 Consumed in enz. hydr. 2,640 0.10
Generated in APR &
26,297 0.96 WWT evap. 4,535 0.17
Condensation
WWT air intake 4,595 0.17 WWT brine 3,362 0.12
Combustion air intake 6,225 0.23 Combustor stack 71,871 2.64
Makeup water 159,026 5.83 Boiler blowdown vent 2,705 0.10
Upgrading flue gas 241 0.01
Sum of Inputs 247,443 9.08 Sum of Outputs 247,247 9.07

65
Similar to the 2013 case, a vacuum filter press is utilized after hydrolysis to remove lignin and other
insoluble solids, which again requires a wash step to recover lost sugars; this wash step results in a
lignin material with considerably higher moisture content, and thus water loss, as described previously
in the 2013 report. Additional water enters the process via the feedstock and raw chemicals, both as
free water and as potential water, i.e., the combustion product of lignin and unconverted sugars.
Water is also consumed in pretreatment and hydrolysis reactions, and these are accounted for as well.

As noted above, utilities are responsible for a large majority of the water loss, with the cooling tower
evaporation rate nearly matching the makeup water demand (leading to an apparent makeup water
demand of approximately 100% allocated to cooling tower evaporation alone). The remainder of water
makeup is met by the large generation during catalysis, or put differently, the required process water
makeup rate would be higher if the generation during the conversion step were lower. Table 34
expresses the individual cooling water users in terms of their water loss responsibility in gal/GGE.

Table 34. Individual Contributors to Cooling Water Evaporation

Cooling Water User MMKcal/hr % of duty gal/GGE

Condensing turbine 76.3 67.9% 3.96
Chiller condenser (enzyme prod.) 3.9 3.5% 0.20
Hydrolysis cooler 6.7 6.0% 0.35
Hydrolysate flash condenser 1.4 1.3% 0.07
Enzyme production compressor 0.6 0.5% 0.03
Conversion/upgrading trim coolers 2.6 2.3% 0.13
Condensation reactor effluent cooler 12.5 11.1% 0.65
Sugar evaporator condensate cooler 8.3 7.4% 0.43
Total 112.3 100.0% 5.82

The largest cooling water user is the condensing turbine, which is responsible for 3.96 gal/GGE of
water demand (68% of cooling water duty). A condenser on the turbine allows the steam to discharge
at a vacuum (about 0.1 atm) for maximum conversion of compressive energy to electricity. With the
condensing turbine in place, the process generates a significant amount of electricity to supply all
users in the plant, leaving roughly 11 MW of excess electricity exported to the grid as a coproduct.
Without the turbine condensers ability to let the steam down to such low pressure (and temperature),
the total amount of generated electricity would decrease and could potentially lead to a net power
deficit that would have to be purchased from the grid. As noted above, the amount of heat that must be
removed from the condensing turbine is less than in the 2013 biological design (76.3 versus 86.8
MMkcal/hr), due to higher heating steam demands for this system given the larger temperature swings
in the back-end conversion steps; particularly an added demand for high-pressure steam that must be
removed from the steam system prior to being sent through any of the turbine stages.

As discussed in the 2013 report, one potential possibility to reduce the biorefinery water demand could
be to employ direct air-cooled condenser (ACC) units on the steam turbine exhaust. In contrast to
standard fin-fan wet evaporative cooling, an ACC unit is a completely closed-circuit heat exchange
system in which the turbine steam is condensed on the inside of the tubes as a result of colder ambient
air flowing across the outside of the finned tube surfaces. Because the heat rejection is achieved using
only air, no water is consumed in the condensing process while achieving lower process temperatures
compared to standard fin-fan air coolers. Further study is warranted to confirm that replacing the

66
water-cooled steam turbine condenser with an air-cooled condenser (either in full or in part) to reduce
consumptive water use is technically and economically feasible.

5.2 Cost Sensitivity Analysis

5.2.1 Single-Point Sensitivity Analysis
A single-point sensitivity was performed on the Aspen model using the variables and limits shown in
Table 35. The baseline for all variables used in the design case is described previously in this report.
Reasonable minima and maxima for each variable were chosen to understand and quantify the
resulting cost impact on overall MFSP. Each variable was changed to its maximum and minimum
value with all other factors held constant. The sensitivity on MFSP is displayed as a tornado chart in
Figure 15.

Table 35. Assumptions Varied in the Sensitivity Analysis

Assumption Name Min Baseline Max
Pretreatment Deacetylation xylan loss 0% 2% 10%
PT reactor metallurgy Stainless steel High alloy
PT residence time 2 5 10
PT acid loading (mg/g) 5 9 20
PT temperature C 150 158 170
PT xylan to xylose 80% 90% 95%
PT xylan to furfural 3% 5% 8%
PT glucan to glucose 6% 10% 12%
Disc refining option Disc refining Acid pretreatment -
Enz Hydrolysis EH % solids 17.5% 20% 25%
+ Conditioning EH cellulose to glucose 75% 90% 95%
EH enzyme loading mg/g 5 10 20
EH time (d) 2 3.5 6
Sugar loss in S/L separation - 1% 5%
S/L separation capex -50% - +50%
Included in Replace every
Belt filter replacement -
maintenance factor other year
Microfilter recycle purge 0% 50% 100%
Microfilter retentate loss 5% 10% 15%
Ion exchange capital -25% - +50%
Enz Production A400 capex -50% - +50%
Catalytic Catalytic upgrading capital -50% - +100%
Upgrading Carbon yield from sugars 0.7 0.86 -
WHSV-APR-1 (hydrogenation) 0.7 1.2 2
WHSV-APR-2 (APR reactor) 0.7 1 2
WHSV-condensation 0.7 1.2 2
Total H2 feed molar ratio 6 9.8 15
H2 price ($/kg) 1.1 1.57 2
Catalyst price -20% - +20%
Capital Total capital investment -25% - +25%

67
The metrics of carbon efficiency from hydrolysate, uncertainty in total capital investment resulting
from the factored approach we used ( 25%), as well as the quoted catalytic conversion capital cost
uncertainty range, have the largest impact on MFSP; additionally, glucose yield from hydrolysis and
hydrogen purchase price also weigh heavily on costs. The critical role that hydrogen and the
associated hydrogen price plays is noteworthy to emphasize, given the large amount of hydrogen
import required in the base case. While the single-point sensitivity to hydrogen price is intended to
capture a reasonable range of expected costs in the context of U.S. natural gas price fluctuations over
recent years (i.e., $2$7.4/MM BTU as discussed previously), hydrogen price variances are further
expanded over a larger range of values as shown in Figure 16 to consider potential implications for
higher natural gas prices. The upper end of prices shown in Figure 16 corresponds to other countries
(such as Japan) with significantly higher natural gas prices ($1417/MM BTU) than the United States
[87], which has fluctuated closer to $4/MM BTU in recent years [88]. The associated correlation
between hydrogen prices and natural gas prices is based on the above-cited H2A report for current
SMR technology [75].

Beyond drivers associated with capital costs, hydrogen cost, and overall carbon yields to fuels,
cellulose-to-monomer glucose yield and enzyme loading play a less critical role in overall cost impacts
on a cost-per-GGE basis than in the 2013 biological hydrocarbon report, but still carry a higher cost
impact over the remainder of the parameters evaluated here. Additionally, pretreatment and hydrolysis
operations may have room to be further tailored to produce more oligomeric sugars and less monomers
(leading to potential cost savings for deconstruction operations) given the ability of the catalyst to also
convert oligomer components beyond strictly monomers. This scenario was not considered here, but
would be an alternative consideration to the cellulose-to-monomeric glucose evaluation. Conversion
parameters factor heavily into production cost, because these parameters not only affect product final
yield, but also affect capital and other operating costs. We note that this is an imperfect comparison
because the TCI and other cost-related bars represent an actual uncertainty while the other process-
related bars are more accurately called a risk, in the sense that they represent a range of results that
might be expected in future pilot-scale demonstration runs.

The majority of the remaining parameters are self-explanatory; however, a brief discussion is
warranted on the inclusion of pretreatment disc refining as an alternative option. In recent years,
NREL has begun investigating mechanical disc refining as a means to improve digestibility while
maintaining high process yields [21, 89]. Mechanical refining is a relatively established technology
that historically has been used in the pulp and paper industry to improve fiber strength but also has
been found to be an attractive approach to biomass deconstruction processes as a means to increase
accessibility of the cellulose in biomass and improve process sugar yields [90]. More recently, a new
process technology has been investigated using deacetylation followed by disc refining only,
eliminating the acid pretreatment step. In this process, biomass is first subjected to a deacetylation step
where about 70%80% of the acetyl groups in the biomass are removed by a dilute alkali solution
heated to approximately 80C. The deacetylated biomass is mechanically refined through a single
stage of disc refining. The refined biomass is then hydrolyzed using cellulase and hemicellulase
enzymes. The process leaves relatively high purity sugar syrups that can then be biologically or
catalytically upgraded to produce biofuels and chemicals; this may be advantageous in the present
pathway as it avoids introduction of acid and caustic (neutralization) chemicals, thus reducing
ammonium salt formation and subsequent removal.

68
As a cursory feasibility analysis, disc refining (specifically deacetylation and disc refining or DDR) is
considered here as an alternative to acid pretreatment. For a 2,000 dry ton biomass facility, four 54-
inch double disc refiners were quoted with motor, surge bins, and feed and discharge conveyors. The
selected disc refiners are 54 inches in diameter with a maximum 5,000 connected horsepower and a
no-load power demand of 453 hp each. Based on NREL test data for corn stover refining, we assumed
a specific refining energy of 200 KWh/bone dry ton of biomass. Thus, while minimizing operating
expenses by way of acid and caustic neutralization chemicals relative to the dilute-acid pretreatment
base case, the alternative utilizing disc refining requires considerably more electrical energy. No
change in cellulose or hemicellulose sugar yields were assumed, thus this analysis strictly considers
trade-offs in capital and operating cost estimates. Although cost impact for replacing dilute-acid
pretreatment with disc refining (Figure 15) was not found to be significant, additional benefits may be
seen that were not included here; namely, production of cleaner sugars with lower ammonium salt
levels (thus likely lower purification demands through ion exchange in this pathway), as well as
expected lower production levels of components known to be fermentation inhibitors, such as furfurals
and acetic acid [90-92] (likely a benefit for biological conversion pathways, although such species are
assumed to be convertible in the present catalytic design).

Figure 15. Single-point sensitivity tornado chart for purchased hydrogen base case

69
Figure 16. MFSP sensitivity scan as a function of hydrogen price and correlated SMR natural gas prices

5.2.2 Evaluation of Alternative Hydrogen Sourcing Options

The carbon yield from hydrolysate through finished fuels in the base case design is very high at
85.6%, attributed to the use of externally-supplied SMR hydrogen. However, as mentioned previously,
the greenhouse gas and fossil energy penalties attributed to natural gas SMR hydrogen production may
detrimentally impact the overall life cycle of the process, given the resulting amount of hydrogen
import required. There are a large number of alternative methods to source hydrogen from more
sustainable sources including electrolysis of water, solar-thermal and photoelectrochemical water
splitting, high-temperature electrolysis coupled with nuclear energy, or biological pathways [93];
while most of these are expected to increase hydrogen costs considerably over standard SMR
production (today), they are noteworthy to mention as future opportunities for improving sustainability
for a large hydrogen consuming process such as this. Additionally, two other options also exist that
may be more sensible in this context by offering the potential to leverage the biorefinery
infrastructure, namely in situ hydrogen production from the produced sugars, or gasification of a
fraction of biomass feed. Both of these approaches tap into the potential to use a fraction of the
biomass feedstock resource otherwise devoted entirely to fuel production in the base case purchased
hydrogen scenario. To preserve consistency across modeled hydrogen scenarios, both alternative cases
considered here maintain a total biomass feed rate to the biorefinery of 2,000 dry tonne/day.

In Situ Hydrogen Production

In situ hydrogen production offers the potential to improve sustainability over natural gas-derived
hydrogen, but at the cost of fuel yield. While external SMR-based hydrogen is considered for the base
case, the in situ hydrogen approach is presented here as one alternative to frame the issue from a high-
level cost and sustainability standpoint. Information from patent literature has described alternative
systems to produce hydrogen in situ from aqueous phase reforming of biomass hydrolysates [32, 33];
a key advantage of APR is that it enables flexibility in choosing the source of hydrogen supplied to the
plant.

70
Early published work on APR focused on generation of hydrogen from renewable biomass feedstocks
with the implication that it would be a more viable process than electrolysis of water, gasification,
steam-reforming of bio-oils, or enzymatic processes [31]. The initial focus was to produce hydrogen
from biomass-derived material for use in fuel cells [94]. Information from patent literature has
suggested that hydrogen produced by APR (APR-H2) can also be integrated into a catalytic sugar
reforming process like the base case model in this report [32, 33]. The selectivity of APR reactions to
hydrogen can be improved by using reduced substrates like sugar alcohols (for example, the basic
reaction from sorbitol: C6H14O6 + 6 H2O 13 H2 + 6 CO2) [31] and therefore APR-H2 could use a
stream diverted from the hydrogenation reactor feed (i.e., first considering reduction of glucose to
sorbitol: C6H12O6 + H2 C6H14O6) [31, 95-98]. In situ hydrogen by APR has other potential
advantages over SMR. APR-H2 production takes place at lower temperature (less than 400C) versus
SMR which takes place at 400C to 700C [94]. SMR requires steam whereas APR-H2 production
takes place in the liquid phase eliminating the energy penalty due the heat of vaporization of water.

A fraction of the hydrolysate after enzymatic hydrolysis, sugar concentration, and cleanup is used to
produce hydrogen for sugar catalytic conversion to fuels. Of course, the more hydrolysate is diverted
to produce hydrogen, the lower the amount of hydrogen then required to satisfy the decreased amount
of hydrocarbons produced from the lower remaining amount of available sugars (carbon), leading to
an optimum split. In order to provide the required hydrogen demands for hydrogenation/APR,
condensation, and hydrotreating reactions to hydrocarbon fuels, approximately 41% of the initial
purified hydrolysate material is required, thereby reducing carbon and overall fuel yields accordingly
by a similar fraction relative to the purchased hydrogen base case (shown further below in Table 38).
This yield optimization is based on assuming stoichiometric conversion of all pertinent hydrolysate
species to hydrogen and CO2 using simplified reforming reactions as shown in Table 36; thus, this
presents a likely best-case scenario based on complete conversion of all hydrolysate components and
does not consider possible subsequent recombination of H2 and CO2 to alkanes and water or other
hydrogen yield losses [31].
Table 36. Governing Stoichiometry Assumed for Components in Hydrolysate to in situ Hydrogen
Production (APR H2 Train)

Reaction
GLUCOSE + 6 H2O --> 12 H2 + 6 CO2
XYLOSE + 5 H2O --> 10 H2 + 5 CO2
GALACTOSE + 6 H2O --> 12 H2 + 6 CO2
MANNOSE + 6 H2O --> 12 H2 + 6 CO2
ARABINOSE + 5 H2O --> 10 H2 + 5 CO2
GLUCOSE OLIGOSACCHARIDE + 7 H2O --> 12 H2 + 6 CO2
XYLOSE OLIGOSACCHARIDE + 6 H2O --> 10 H2 + 5 CO2
GALACTOSE OLIGOSACCHARIDE + 7 H2O --> 12 H2 + 6 CO2
MANNOSE OLIGOSACCHARIDE + 7 H2O --> 12 H2 + 6 CO2
ARABINOSE OLIGOSACCHARIDE + 6 H2O --> 10 H2 + 5 CO2
CELLOBIOSE + 13 H2O --> 24 H2 + 12 CO2
SOLUBLE LIGNIN + 13 H2O --> 17 H2 + 8 CO2
HMF + 9 H2O --> 12 H2 + 6 CO2
FURFURAL + 8 H2O --> 10 H2 + 5 CO2

71
To model the in situ process train, the diverted hydrolysate stream is first heated to near-reaction
temperature in a feed-effluent heat exchanger (by cross-exchange with the APR-H2 reactor effluent),
then heated further using high pressure steam from the A800 steam system. The reaction takes place at
227C and 30 atm (420 psig) [31], and the reactor product stream is then further cooled to 30C to
knock out water (routed to wastewater treatment). Hydrogen is purified by a PSA system to remove
CO2, then compressed to 1,050 psig and routed to the hydrocarbon catalytic reactor train. The design
and cost estimates for the PSA unit were adopted from the 2013 biological design report utilizing
zeolite-based sorbent with 95% hydrogen recovery. The PSA tailgas, containing mainly CO2 with 5%
H2 loss, is combined with the purge gas stream from the hydrocarbon catalytic train and routed to the
boiler (A800). The APR-H2 reactor capital cost and operating expenses for annualized catalyst cost
(and associated catalyst composition) were set consistently with the hydrocarbon train APR reactor
(APR-2, i.e., the second of the three individual conversion reactors quoted by Harris Group and the
vendor).

Biomass Gasification Hydrogen Production

The second alternative hydrogen sourcing option considered is to gasify a fraction of biomass supplied
to the biorefinery in a separate gasification process train operated in parallel to the primary sugar/fuel
train which otherwise maintains all process assumptions consistently with the purchased hydrogen
base case. In this scenario, the overall biorefinery feedstock rate remains fixed at 2,000 dry tonne/day,
but the split towards hydrogen production occurs further upstream in the process (i.e. at the
biorefinery plant gate) relative to the in situ pathway which instead takes the split from the produced
sugar stream. However, another difference is that the gasification process is more favorable towards
woody biomass feedstocks as the standard for design rather than herbaceous feedstocks such as corn
stover, due to lower ash and higher carbon content for woody feedstocks as preferential for
thermochemical processes such as gasification [68]. As such, the gasification scenario employs a dual-
feedstock basis delivered to the biorefinery, with the same herbaceous blend utilized for the sugar/fuel
train (see Section 2.2) and a woody feedstock blend utilized for the gasification/hydrogen train, with
the sum of the two feedstocks fixed at 2,000 dry tonne/day.

The gasification train model was constructed based primarily on the process and design assumptions
as documented in NRELs 2011 thermochemical ethanol design report [68]. This includes the use of
indirect steam gasification coupled with a fluidized bed char combustor with olivine circulating
between the two operations as a heat transfer medium, followed by syngas cleanup based on catalytic
tar reforming and then cooling, quench, and scrubbing of the syngas to remove impurities such as
particulates and ammonia. The tar reformer is designed similar to a small-scale fluidized catalytic
cracker (FCC) unit, whereby tars, methane, and light hydrocarbons are reformed to additional syngas
over a catalyst, with continuous regeneration of the catalyst by combustion of the deposited coke. The
syngas is then compressed in a multi-stage compressor to a pressure of 29 atm (430 psig). From this
point, a number of additional unit operations were added for hydrogen production and recovery which
were not part of the thermochemical ethanol design, but rather were based on the biomass gasification-
to-hydrogen design report published by NREL in 2005 [34]; this includes removing sulfur species in a
zinc oxide (ZnO) guard bed, increasing hydrogen yield through catalytic high-temperature shift (HTS)
and low-temperature shift (LTS) reactors (i.e. following the water gas shift reaction H2O + CO H2
+ CO2), and purifying the hydrogen via a PSA unit. All capital and operating costs for the process are
based on the above-referenced NREL thermochemical ethanol and hydrogen reports for the respective
unit operations. The reader is referred to these reports for more information on process and design

72
details beyond the brief high-level summary provided here. Key operational and yield parameters for
the gasification/hydrogen production train are summarized in Table 37.

Table 37: Key Operational and Yield Parameters for Gasification/Hydrogen Process Train
Overall Modeling Parameters Value
Delivered feedstock moisture content 10%
Gasifier temperature 870 C
Char combustor temperature 981 C
Tar reformer temperature 910 C
Tar reformer catalyst type Ni/Mg/K/Al2O3
H2:CO molar ratio across tar reformer 0.57 (in) 1.61 (out)
Tar reformer component conversions:
Methane (CH4) 80%
Ethane (C2H6) 99%
Ethylene (C2H4) 90%
Tars (C10+) 99%
Benzene (C6H6) 99%
Ammonia (NH3) 90%
Syngas compressor outlet pressure 29.3 atm
H2S concentration across ZnO guard bed 414 ppm (in) 0 ppm (out)
HTS temperature 468 C
LTS temperature 285 C
Catalyst used in shift reactors Copper-zinc oxide
PSA H2 recovery 85%
H2 product stream purity 99.99%
Overall hydrogen yield across H2 process train 82.0 kg/dry tonne biomass

Beyond unit operation design and cost assumptions taken from the above-referenced thermochemical
ethanol and biomass-to-hydrogen design reports, assumptions pertaining to woody feedstock delivery
to the gasifier train were updated to maintain consistency with NRELs most recent modeling efforts
making use of this feedstock blend, namely NRELs 2014 indirect gasification of biomass-to-high
octane gasoline design report [99]. Based on recent updates from partners at INL, the feedstock
delivery details documented in this new 2014 report stipulate a feedstock cost of $80/dry ton delivered
to the throat of the gasifier system, associated with a delivered moisture content of 10% after upstream
preprocessing; at this moisture content, the new 2014 design also makes use of a new cross-flow
heater prior to the gasifier in place of a biomass dryer that had been used in prior designs to
accommodate higher-moisture feedstocks. Similar to the herbaceous feedstock blend on the
biochemical process train (discussed above in Section 3.1), the woody feed to the gasifier process train
is also based on a blending strategy that combines pulpwood, wood residues, construction and
demolition waste, and switchgrass [100]. Based on the modeled hydrogen yield of 82.0 kg/dry metric
tonne biomass to the gasifier (Table 37), this scenario requires allocating 720 dry tonne/day of woody
biomass to the gasifier train (36% the total 2,000 dry tonne/day available for the overall system) to
produce the required amount of hydrogen, leaving 1,280 dry tonne/day of herbaceous feedstock
available for the primary sugar/fuel process train, thereby reducing fuel yields by 36% relative to the
purchased hydrogen base case; see Table 38.

73
Model Results
Results for fuel and carbon yields for the in situ and gasification scenarios for hydrogen sourcing are
presented in Table 38, alongside the purchased hydrogen base case.

Table 38. Comparison of Fuel and Carbon Yields between Base Design and Alternative Hydrogen Cases

Base Case In situ Hydrogen Biomass Gasification

41% split of 36% split of total
purchased
Hydrogen source hydrolysate feed biomass feed
Fuel yield (GGE/ton total biomass) 78.3 45.3 50.1
Carbon efficiency to fuels from total
85.6% 49.5% 85.6%
hydrolysate a
Carbon efficiency to fuels from total
45.3% 26.2% 27.7%
biomass a
a
Note, calculations do not include carbon from natural gas for purchased hydrogen case (reduces carbon yield down to
35.7%); additionally, calculations are based on total starting hydrolysate carbon (and biomass); actual carbon efficiency
of the in situ hydrocarbon train from the fraction of hydrolysate sent to this train is equal to the purchased H2 case.

As over 41% of the hydrolysate stream is diverted to produce hydrogen for the in situ pathway, carbon
retention efficiency from total hydrolysate carbon is reduced from 85.6% to 49.5%, and the carbon
efficiency from biomass feedstock is reduced from 45.3% to 26.2% (this only considers total initial
carbon in hydrolysate and biomass respectively, excluding carbon in natural gas used to produce H2 in
the purchased scenario). Correspondingly, total fuel yield decreases from 78.3 GGE/ton in the ex
situ base case to 45.3 GGE/ton in the in situ scenario. While this saves on purchased hydrogen cost,
the yield penalty outweighs this benefit and results in an increase in MFSP from $4.05 to $5.48/GGE,
shown in Figure 17. Likewise, the gasification scenario requires 36% of the biomass originally
available for the ex situ base case, reducing fuel yield to 50.1 GGE/ton of total biomass which
corresponds to a carbon efficiency from total biomass to fuels of 27.7% (including biomass diverted to
the gasifier train). Although the gasification case is estimated to be slightly more capital cost-intensive
than the in situ case ($345 MM versus $331 MM total installed costs), the higher fuel yield for this
scenario ultimately translates to a roughly 10% lower MFSP at $4.95/GGE, although still markedly
higher than the $4.05/GGE ex situ base case. Put differently, both of these alternative hydrogen cases
replace hydrogen sourced from low-cost natural gas with much higher-cost biomass. However, as
noted above, a key rationale for such alternative scenarios is the ensuing sustainability benefits for
eliminating natural gas SMR demands. While a full well-to-wheel LCA of the system is beyond the
scope of this analysis, sustainability metrics associated strictly with the modeled biorefinery
conversion process are reported in the next section for all three hydrogen scenarios. Cost breakdowns
for all three scenarios are provided in Figure 17, with more details for the two alternate cases shown in
Figures 18-19.

It is noteworthy that the in situ scenario coincides closely with NRELs 2013 biological hydrocarbon
design case, both in terms of cost and yields. Namely, the 2013 design basis presented modeled fuel
(RDB) yields of 45.4 GGE/ton and 26.2% carbon retention efficiency from biomass to finished fuels
[8] compared to 45.3 GGE/ton and 26.2% carbon retention efficiency from biomass carbon in the
present in situ case. Likewise, the MFSP for the 2013 design was $5.10/GGE, compared to $5.48/GGE
in the in situ case here. Thus, the case for in situ hydrogen production coupled with catalytic
conversion of remaining hydrolysate to fuels presents a more comparable basis relative to the
2013 biological conversion base case, both in terms of TEA results and the process concept itself
being self-sufficient in how the fuel product is produced across the conversion step.

74
Figure 17. Cost comparison of the ex situ hydrogen base case with in situ and gasification hydrogen
alternatives

75
Figure 18: Economic and process summary results for in situ H2 scenario

76
Figure 19: Economic and process summary results for H2 via biomass gasification

77
5.3 Sustainability Metrics
An important aspect of evaluating biofuel processes is the quantification of life cycle resource
consumption and environmental emissions. Life cycle assessment (LCA) provides a framework from
which the environmental sustainability of a given process may be quantified and assessed. This section
presents the salient sustainability metrics of the current conceptual process at the conversion stage.
Direct biorefinery emissions (i.e., CO2, NO2 and SO2), water consumption, and other process-related
metrics were derived from the conversion process model described above.

SimaPro v.8.0.2 software [101] was used to develop and link units quantifying life cycle impacts as
previously documented by Hsu et al. [102]. Greenhouse gas (GHG) basis values for purchased
hydrogen (derived via natural gas steam methane reforming) and the U.S. average electricity mix from
the grid were applied consistently with the basis from Argonne National Laboratorys GREET model
software; this includes GHG emission factors of 11.2 kg CO2e/kg H2 and 0.65 kg CO2e/KWh
purchased electricity [103]. Likewise, the basis from GREET was also applied for non-renewable
(fossil) consumption values of 183 MJ/kg H2 and 7.46 MJ/KWh purchased electricity [103]. All other
process input/output inventory estimates were derived from Ecoinvent v.2.2 [104] and the U.S. Life
Cycle Inventory (LCI) [105] processes to fill the data gaps, similar to the methods applied in NRELs
2013 biological conversion design report. The Ecoinvent processes were modified to reflect U.S.
conditions and the U.S. LCI processes were adapted to account for embodied emissions and fossil
energy usage. The material and energy flows of the conversion step capture the impacts of input raw
materials, and outputs, such as emissions, wastes, and coproducts as predicted by the process model,
and are shown in Table 39. This table and subsequent discussion of sustainability metrics consider the
base case model utilizing purchased hydrogen and the alternative cases with internal hydrogen
generation via in situ and gasification routes. Additionally, the LCI data presented below in Table 39
is translated to an average hourly basis, i.e., dividing the annual rates by the operating factor of 7,884
hours/year. In some cases the hourly rates for certain items, such as catalyst usage, do not reflect an
actual hourly consumption but rather are amortized as such for purposes of evaluating the resulting
sustainability metrics on a consistent basis.

78
 Table 39. Input and Output Data Associated with the Modeled Conversion Facility
Purchased Hydrogen In Situ Hydrogen Gasification Hydrogen
Products Production Rate Production Rate Production Rate
kg/hr MJ/hr (LHV) kg/hr MJ/hr (LHV) kg/hr MJ/hr (LHV)
Total Hydrocarbon Fuel 20,047 874,291 11,601 505,353 12,833 559,715
Electricity Export (Import) 11,045 KW (914) KW 1,588 KW
Resource Consumption a Flow Rate Per MJ Fuel Flow Rate Per MJ Fuel Flow Rate Per MJ Fuel
kg/hr kg/MJ kg/hr kg/MJ kg/hr kg/MJ
Biomass Feedstock 104,167 1.2E-01 104,167 2.1E-01 99,998 2.0E-01
Sulfuric Acid (93%) 2,240 2.6E-03 2,240 4.4E-03 1,950 3.5E-03
Caustic (NaOH as pure) 1,406 1.6E-03 1,406 2.8E-03 900 1.6E-03
Ammonia 310 3.5E-04 310 6.1E-04 198 3.5E-04
HCl for IX Regeneration 1,122 1.3E-03 1,122 2.2E-03 718 1.3E-03
Caustic for IX regeneration 640 7.3E-04 640 1.3E-03 409 7.3E-04
c c c
Imported Hydrogen 3,847 4.4E-03 NA NA NA NA c
Total APR Catalysts 8 9.5E-06 6 1.2E-05 5 8.9E-06
Condensation Catalyst 3 3.0E-06 2 4.0E-06 2 3.6E-06
Polymer 2 2.1E-06 2 3.6E-06 1 1.8E-06
Ammonia (WWT) 109 1.2E-04 128 2.5E-04 104 1.9E-04
Boiler Chemicals 2.3E-01 2.6E-07 2.3E-01 4.5E-07 2.3E-01 4.0E-07
FGD Lime 180 2.1E-04 184 3.6E-04 115 2.1E-04
Cooling Tower Chemicals 2 2.8E-06 2 4.8E-06 2 4.3E-06
Makeup Water 159,026 1.6E-01 154,837 3.1E-01 199,495 3.8E-01
Air Demand 568,809 0.7 542,388 1.1 472,942 8.4E-01
Glucose 1,213 1.4E-03 1,213 2.4E-03 776 1.4E-03
Corn Steep Liquor 83 9.4E-05 83 1.6E-04 53 9.4E-05
Corn Oil 7 7.7E-06 7 1.3E-05 4 7.7E-06
Host Nutrients 34 3.9E-05 34 6.7E-05 22 3.9E-05
Sulfur Dioxide 8 9.4E-06 8 1.6E-05 5 9.4E-06
Ammonia 58 6.6E-05 58 6.6E-05 37 6.6E-05
Gasifier Train Chemicals NA NA NA NA 95 1.7E-04
Air Emissions kg/hr kg/MJ kg/hr kg/MJ kg/hr kg/MJ
Water (H2O) 260,437 3.0E-01 230,219 4.6E-01 232,737 4.2E-01
Nitrogen (N2) 429,467 4.9E-01 421,070 8.3E-01 362,830 6.5E-01
Oxygen (O2) 66,379 8.5E-02 64,369 1.3E-01 46,721 8.3E-02
Carbon Dioxide (CO2) b 74,741 7.6E-02 101,595 2.0E-01 103,914 1.9E-01
Methane (CH4) 1.6 1.8E-06 1.4 2.7E-06 1.4 2.5E-06
Nitrogen Dioxide (NO2) 53 6.0E-05 53 1.0E-04 42 7.5E-05
Carbon Monoxide (CO) 53 6.0E-05 53 1.0E-04 34 6.0E-05
Sulfur Dioxide (SO2) 11 1.2E-05 11 2.1E-05 12 2.1E-05
Waste Streams kg/hr kg/MJ kg/hr kg/MJ kg/hr kg/MJ
Disposal of Ash/Spent Catalyst 4,506 5.2E-03 4,512 8.9E-03 3,442 6.2E-03
Wastewater (brine) 6,720 7.7E-03 6,356 1.3E-02 4,305 7.7E-03
a
See discussion above for details on chemical inputs listed here.
b
Direct carbon dioxide (CO2) emission is 100% biogenic CO2.
c
Hydrogen provided internally via in situ or gasification generation.

79
Details of contributions to GHG emissions and fossil energy consumption at the conversion stage are
presented in Table 40, which corresponds to the information in Table 39. GHG emissions associated
with the conversion stage for the purchased hydrogen base case, as well as the alternative in situ and
gasification cases, are estimated to be 6.0 kg CO2-eq/GGE, 1.9 kg CO2-eq/GGE, and 0.9 kg CO2-eq/GGE
respectively. The higher GHG profile associated with the purchased hydrogen case is largely attributed
to the relatively large hydrogen import, derived from natural gas via external SMR processing
(corresponding to a fossil GHG emissions footprint of 11.2 kg CO2e/kg H2). Alternatively, the
gasification scenario for internal hydrogen generation results in a lower GHG profile than the in situ
scenario primarily due to the reduction in feed rate to the sugar/fuel process train which has a higher
GHG footprint due to chemicals, enzyme usage, etc. relative to the gasifier train for a given biomass
input; additionally, the gasification scenario also results in a marginally higher fuel yield and also
maintains a small electricity export, while the in situ scenario requires a small net power import.
Direct CO2 emissions from the conversion facility itself (shown in Table 39) consist of only biogenic
CO2 (i.e., CO2 absorbed from the atmosphere during feedstock growth). With its biomass origin,
biogenic CO2 does not contribute to the increase of GHG in the atmosphere [106] and is not
considered in the IPCC global warming methodology [107]. Hence, the contributions to GHG at the
conversion stage are solely from the fossil GHG profile associated with the underlying processes (e.g.,
material inputs/outputs to and from the facility to support process operations). For the purchased
hydrogen base case, beyond external hydrogen it is evident that underlying processes associated with
sodium hydroxide (NaOH), HCl, and enzyme production contribute the most to the overall GHG
emissions estimate, while excess electricity exported to the grid (largely displacing coal-derived
electricity) provides a considerable credit to reduce overall GHG emissions.

Table 40 also shows the fossil energy consumption for the conversion process. Fossil energy
consumption associated with the purchased, in situ, and gasification hydrogen scenarios are estimated
to be 99.3 MJ/GGE, 23.8 MJ/GGE, and 12.2 MJ/GGE respectively. Similar to the GHG emissions, the
difference in total fossil energy consumption for the purchased versus internal hydrogen cases is
largely driven by hydrogen sourced from natural gas in the purchased hydrogen case; as this is itself a
fossil fuel, the resulting fossil energy profile for the purchased hydrogen scenario is higher. Similarly,
the gasification scenario results in roughly 50% reduction in fossil energy demand relative to the in
situ case due to the diversion of a portion of the biomass feedstock away from the more chemical-
intensive sugar/fuel process train to the less-intensive gasification train, coupled with slightly higher
fuel yields and lower power demand for the gasification scenario.

80
 Table 40. Conversion Process GHG Emissions and Fossil Energy Consumption per GGE Total Fuel
Product

GHG Emission Fossil Energy Input

kg CO2-eq/GGE MJ/GGE
H2 Sourcing: Purchased In situ Gasification Purchased In situ Gasification
Direct refinery emission - - - - - -
Sulfuric acid 0.03 0.04 0.04 0.36 0.62 0.49
Ammonia 0.09 0.16 0.09 1.80 3.11 1.79
Sodium hydroxide 0.28 0.48 0.28 2.97 5.13 2.97
HCl for IX regeneration 0.23 0.40 0.23 3.07 5.30 3.07
Caustic for IX regeneration 0.09 0.16 0.09 1.37 2.37 1.37
Hydrogen 6.00 NA a NA a 97.98 NA a NA a
APR/condensation catalysts 0.01 0.01 0.00 0.09 0.12 0.07
Enzyme production 0.19 0.34 0.21 2.09 3.62 2.39
Biorefinery infrastructure 0.01 0.02 0.02 0.14 0.18 0.17
Electricity -1.00 0.14 -0.23 -11.46 1.64 -2.57
Cooling water chemicals 6.0E-04 1.0E-03 9.4E-04 0.01 0.01 0.01
WWT chemicals 0.03 0.07 0.05 0.64 1.30 0.95
Boiler chemicals 0.02 0.04 0.02 0.11 0.19 0.11
Waste disposal 4.8E-03 0.01 0.01 0.12 0.21 0.15
Gasifier chemicals NA NA 0.09 NA NA 1.25
Total 5.98 1.86 0.91 99.29 23.79 12.21
a
Hydrogen provided internally via in situ or gasification generation.

Table 41 summarizes the key sustainability metrics for the conversion process across the three
hydrogen sourcing scenarios. On an energy basis, the GHG emissions at the conversion stage for the
purchased, in situ, and gasification hydrogen cases are 49.2 g CO2-eq/MJ, 15.3 g CO2-eq/MJ, and 7.5 g
CO2-eq/MJ respectively. Similarly, the fossil energy consumption for the three scenarios is 0.82
MJ/MJ, 0.20 MJ/MJ, and 0.10 MJ/MJ respectively. In addition to GHG emissions and fossil energy
consumption, water consumption (i.e., net water use during the biorefinery operation), total fuel
efficiency (yield), and carbon-to-fuel efficiency are also summarized again in Table 41. Biorefinery
net water consumption includes, but is not limited to, water that is incorporated into the product and
evaporation directly from process operations or indirectly from cooling and heating processes (e.g.,
cooling tower evaporative losses). By this definition, net water consumption for the modeled
biorefinery is 5.8 gal/GGE total fuels for the purchased hydrogen case (discussed above in Section
5.1), increasing to 9.8 gal/GGE for the in situ case and 11.4 gal/GGE for the gasification case, which is
primarily attributed to lower fuel yields for both alternative hydrogen scenarios but also to higher
overall water demand for the gasification case (water demand for the in situ case is similar to the
purchased hydrogen basis) . As discussed previously, total modeled fuel yields are 78.3, 45.3, and 50.1
GGE/dry ton for the purchased, in situ, and gasification hydrogen scenarios respectively, associated
with a total biomass carbon-to-fuel retention efficiency of 45.3%, 26.2%, and 27.7% respectively. For
the purchased hydrogen base case, when including the carbon inherent to natural gas for SMR
hydrogen production (0.44 mol/mol H2 [75]), overall carbon efficiency decreases to 35.7%.

81
 Table 41. Summary of Sustainability Metrics for the Biorefinery Process

Purchased H2 In situ H2 Gasification H2

GHG emissions (g CO2e/MJ) 49.2 15.3 7.5
3
Water consumption (m /day) 3817 3716 4788
Water consumption (gal/GGE fuel) 5.8 9.8 11.4
Total fuel yield (GGE/dry ton) 78.3 45.3 50.1
Carbon-to-fuel efficiency (C in fuel/C in biomass) a 45% 26% 28%
Net fossil energy consumption (MJ/MJ) 0.82 0.20 0.10
a
Carbon efficiency is defined here as the ratio of carbon in primary fuel products (RDB + naphtha) relative to carbon in
biomass feed; when also including carbon in natural gas for SMR H2 production, the purchased hydrogen case
decreases to 36%.

Based on this analysis, the key sustainability metrics for GHG and fossil energy profiles are notably
influenced in this technology pathway by the choice for hydrogen sourcing. Although the purchased
hydrogen case results in substantially higher fuel yields (which benefits sustainability as it does
economics by ultimately amortizing inputs over a larger denominator), this benefit is outweighed by
the larger sustainability penalty attributed to fossil-derived hydrogen. However, an important caveat is
noted here that the purchased hydrogen case assumes off-site SMR hydrogen production and thereby
does not allow for realizing any potential benefits gained by on-site integration. An on-site SMR unit
integrated with the rest of the facility is likely to improve the sustainability profile for the purchased
hydrogen case, as it would allow for additional heat and power integration opportunities, as well as
offsetting a portion of the required natural gas by routing light gases produced in
conversion/upgrading to the reformer unit rather than the combustor as is done currently. Such an
option should be evaluated moving forward. A complete well-to-wheel LCA evaluation of all
scenarios is required to more fully understand the sustainability implications for the full supply chain
based on this technology pathway, such as how the results may weigh into qualification for credits
under the Renewable Fuel Standard (RFS) set by the U.S. Environmental Protection Agency (EPA)
[108]. Such an analysis is outside the scope of this work; however, this preliminary analysis around
the conversion stage suggests an optimum could exist between the two extremes. Namely, a
commercial facility utilizing this technology would be likely to import purchased hydrogen up to the
point of reaching a threshold for GHG limits allowed by RFS requirements, given the low cost of
natural gas (and thereby hydrogen) and ensuing economic benefits for MFSP as presented above. The
remaining hydrogen deficit could then be made up via in situ hydrogen production from a fraction of
hydrolysate carbon and/or process off-gases, or from gasifying a portion of the incoming biomass
feedstock, or other renewable sources. This trade-off between economics and sustainability is depicted
in Figure 20, which presents the boundary conditions for the hydrogen scenarios considered here, and
assumes linear interpolation between these boundaries (taking the averages of the two internal
hydrogen scenarios for illustrative purposes). It is recognized that the system may not behave linearly
between these boundaries, but a linear interpolation is assumed strictly to illustrate the opposing trends
between the two metrics and the possibility of an optimum existing in between. Again, a different set
of tradeoffs may be observed for a process utilizing on-site SMR hydrogen production integrated with
the rest of the facility, which could plausibly eliminate any RFS constraints altogether without needing
to include alternative hydrogen sourcing options. This will be an important area for further analysis
moving forward, to better understand and quantify such details.

82
 7

MFSP ($/GGE) and GHG Emissions

6 Optimization
In Situ
5
(kg CO2-eq/GGE)
Gasification
4 MFSP ($/GGE)

3
GHG Emissions (kg CO2-
2 eq/GGE)
In Situ
1
Gasification
0
100% External 100% Internal
(Purchased)
Hydrogen Sourcing

Figure 20. Trade-offs depicting conceptual optimization between economics (MFSP) and sustainability
(fossil GHG emissions) based on hydrogen sourcing as modeled here (illustrative only). (Purchased
hydrogen via external natural gas SMR versus internal production via in situ or gasification scenarios.)

5.4 Paths Forward to $3/GGE

The modeled overall fuel selling price for this conversion pathway is projected to range between
$4.05$5.48/GGE as an nth-plant target case, depending on the mode of hydrogen sourcing (purchased
versus in situ hydrogen respectively, with biomass gasification falling within this range at
$4.95/GGE). While it is somewhat speculative to assign a specific target year by which process
performance will have been shown publicly to increase to a level consistent with these cost targets
(due to a large lack of publicly available data from which to draw such conclusions around technology
progress), it is expected that the targets documented here would be achievable prior to year 2022.
Similar to the base case results presented for the 2013 biological conversion pathway, this still leaves
further room for improvement required to ultimately achieve final DOE-BETO cost goals of $3/GGE
by 2022. A number of possibilities to allow for such improvements are briefly discussed here.

Moving beyond the process and economic performance metrics described here, the primary means to
further improve MFSP is either to further reduce costs or to increase yields (i.e., overall carbon
efficiency from biomass to useful products). Additional cost reductions are possible for this
technology pathway that have not been considered in this analysis. For example, further reducing
sugar costs and optimizing process integration will continue to be important areas for further R&D and
may be achieved through the use of alternative or milder pretreatment options and/or improved
enzyme performance (higher conversion yields and/or lower enzyme doses or cost), including the
incorporation of new enzyme classes and enzymatic hydrolysis mechanisms [109]. Furthermore, the
flexibility of the catalytic conversion operations towards utilizing a variety of carbon components
beyond just monomeric sugars, including oligomers, sugar degradation products, and solubilized
lignin, is a key benefit that may have more opportunity to be leveraged than what was considered here;
namely, pursuing more synergistic integration to further tailor biomass deconstruction steps to produce
less costly sugar (hydrolysate) consisting of more oligomers and less monomers at reduced chemical
and enzyme demands [19].

83
Aside from cost reductions, further opportunities also exist for further improving fuel yields, primarily
by way of solubilizing additional lignin up front during deacetylation for subsequent catalytic
conversion and upgrading. However, the majority of solubilized lignin and other solubilized
components (e.g., acetate and extractives) from deacetylation partition into the separate liquor product
stream, which is then routed directly to the boiler in the base case. This highlights the primary
opportunity for ultimately achieving $3/GGE goals, which is to utilize additional lignin and potentially
other carbon components in this separate liquor stream for conversion to value-added coproducts. The
possibility for utilization of lignin for conversion to coproducts was investigated in detail in the 2013
biological pathway design report, including process modeling scenarios and consideration of economic
and sustainability implications, and will not be repeated here, but will be briefly summarized. The
reader is encouraged to refer to the 2013 report for additional detail and context for the lignin-to-
coproduct pathway analysis.

As described in the 2013 design report, there are a number of possible approaches for isolating lignin
for subsequent conversion to coproducts, including: (1) deconstructing and solubilizing lignin up front
through more severe alkaline pretreatment operations than the deacetylation base case; (2)
fractionating biomass into lignin and carbohydrate fractions utilizing organic solvents with an acid
catalyst; and (3) isolating and solubilizing lignin after it has passed through carbohydrate
deconstruction steps and separated out from soluble hydrolysate, e.g., via belt filtration as employed in
the base case design. The 2013 report focused on the first option, utilizing alkaline pretreatment for
purposes of conducting a feasibility TEA assessment for subsequent upgrading to commodity
chemicals or fuels. The alkaline pretreatment operation was modeled to take place at 130C with a
caustic (NaOH) loading of roughly 40 mg/g dry biomass and a small amount of added anthraquinone
to minimize carbohydrate losses into the soluble liquor stream. Under these conditions, 80% of the
lignin was assumed to be solubilized for illustrative purposes, and subsequently converted to one of
four example coproducts (among many other coproduct options): (1) 1,3- butadiene, (2) 1,4-
butanediol, (3) adipic acid, and (4) cyclohexane. All of these coproducts were deemed reasonable
examples in light of sufficient market volumes (greater than 1 MM tons/year world market volume) so
as to not risk flooding a particular product market from a small number of biorefinery facilities. The
economic analysis included consideration of all pertinent steps, including lignin solubilization
(alkaline pretreatment), conversion to the given product, and product purification to more than 95%
purity.

The 2013 analysis showed that a number of the selected coproduct options supported achieving or
even surpassing $3/GGE MFSP targets, based on a starting baseline MFSP of $5.10/GGE prior to
including a lignin coproduct train (i.e. solely focused on bioconversion of sugars to hydrocarbons).
Namely, adipic acid and 1,4-butanediol both supported achievement of $3/GGE at 60%80%
conversion of available lignin to coproduct. The results of the TEA assessment are repeated below in
Figure 21. Although no new analysis was conducted here to repeat this study, this figure is included
below to illustrate that a similar result would be expected in this pathway. Specifically, given that the
incoming biomass feedstock composition remains the same as the assumptions employed in the 2013
analysis for the purchased and in situ hydrogen scenarios, both of these cases would track in parallel to
the original cost curves presented in Figure 21 for the biological pathway in terms of economic
implications for achieving ultimate $3/GGE goals, but shifted up by $0.38/GGE for the in situ case
($5.48/GGE versus $5.10/GGE for the 2013 biological case), or shifted down by $1.05/GGE for the ex
situ purchased hydrogen case ($4.05/GGE versus $5.10/GGE). This concept is highlighted in Figure
21 by overlaying parallel lines for two of the example coproducts, shifted by these respective

84
differences; note, the original curves for the 2013 biological case originate at $5.57/GGE rather than
$5.10/GGE reflecting the increased capital costs for adding a lignin coproduct train prior to
introducing any coproduct yields (i.e. 0% conversion of lignin to coproducts). This suggests a lower
lignin conversion around 40% would be required for the purchased hydrogen scenario to ultimately
achieve $3/GGE for either the adipic acid or 1,4-butanediol coproducts; alternatively, the required
lignin conversion would increase to 80-90% for the in situ hydrogen scenario to support achieving
$3/GGE. The biomass gasification case would fall between these example boundaries.

in situ H2 estimates

ex situ purchased H2 estimates

Figure 21. NREL 2013 design report analysis of lignin-to-coproduct scenarios [8]. (Minimum fuel selling
price as a function of lignin conversion to selected coproducts, excerpted from 2013 report. Note: 0% point
($5.57/GGE) was higher than 2013 base case ($5.10/GGE) given the initial cost premium of adding a new lignin
conversion train without any coproduct yield. Dashed-green (1,4-butanediol) and blue (adipic acid) lines are
overlaid here to approximate ex situ purchased hydrogen scenario and in situ hydrogen scenario for the present
catalytic conversion pathway analysis.)

Beyond economic considerations for coproduct pathways, it is also critical to understand trade-offs
against sustainability metrics (including greenhouse gas and fossil energy profiles) for pursuing such a
strategy rather than lignin combustion. NRELs 2013 design report also included quantification of
sustainability metrics attributed to the above coproduct scenarios from lignin and found that the 1,4-
butanediol and adipic acid coproduct options fared better than the base case for lignin combustion in
terms of greenhouse gas emissions, while the other two coproducts fared worse. A similar result would
be expected here; however, it is more important to reiterate the primary point, which is that a much
larger array of coproduct options are possible than merely the four options discussed above, and some
(but not all) coproducts are likely to support ultimately achieving $3/GGE cost targets for the
integrated biorefinery, and promising options must be weighed against considerations for
sustainability as well as market volume capacity. These scenarios and associated economic,
sustainability, and market drivers will continue to be investigated for both the biological and catalytic
conversion pathways moving forward.

85
6 Concluding Remarks
6.1 Summary
This report establishes a plausible case for achieving an intermediate cost goal prior to 2022 of
approximately $4$5/GGE of upgraded renewable diesel- and naphtha-range blendstocks via a
hybrid biochemical/catalytic process, based on a single set of process R&D metrics and associated
cost estimates run across two scenarios for hydrogen sourcing. Namely, the pathway discussed here
assumes continued improvements are made in the biomass deconstruction areas including
deacetylation, dilute-acid pretreatment, and enzymatic hydrolysis, beyond performance demonstrated
during NRELs 2012 State of Technology trials. This includes demonstrating high removal of acetate,
extractives, ash, and other non-convertible or otherwise problematic components at low loss of
carbohydrates under reasonable conditions during deacetylation, as well as high sugar yields under
mild conditions with low chemical and enzyme loadings during pretreatment and hydrolysis. These
front-end unit operations leverage NREL expertise and prior work in these areas around biomass
deconstruction, and will continue to be important areas of future research for continued improvement.

In addition to biomass deconstruction to sugars, the design also sets targets for sugar and other
hydrolysate carbon upgrading to fuels, primarily carbon yield (85.6% carbon retention efficiency from
hydrolysate carbon to fuels exiting hydrotreating), hydrogen consumption for catalytic reactions (6.5
wt % relative to hydrolysate components converted to fuels), and reactor WHSVs for conversion-stage
operations (11.2 hr-1 across three conversion reactor stages). These metrics apply equally to the
performance of the catalytic reaction stages whether sourcing hydrogen externally via off-site natural
gas SMR (ex situ), or internally by (a) diverting a fraction of hydrolysate through separate reforming
reactions (in situ) or (b) diverting a fraction of delivered biomass feedstock to a separate gasification
process train to produce hydrogen for the fuel production train. The primary difference is that while
the in situ route maintains the same carbon efficiency and reaction hydrogen consumption per input
rate of hydrolysate carbon converted to fuels, the amount of carbon converted to fuels is roughly 40%
lower given the requirement to divert 41% of the hydrolysate stream to the hydrogen production train
to satisfy all hydrogen requirements for both the conversion fuel train as well as the hydrotreating step;
similarly, the gasification train requires diverting 36% of the total available biomass towards syngas
production for this purpose. While the above conversion process targets are derived from patent
literature data with reasonable extrapolations made as to ultimately achievable performance in the
future, it is subjective to speculate at what point these targets may be achieved given the lack of
publication data and granularity of patent data available today (particularly related to catalytic
conversion of lignocellulosic hydrolysate, rather than commodity sugars). At a minimum, it is
anticipated these targets would be achievable prior to year 2022 and thus would represent an
intermediate achievement case on the way to ultimate cost goals of $3/GGE by 2022. To further
improve costs towards this ultimate $3/GGE goal, additional improvements would be required by way
of improved process integration with lower costs for biomass deconstruction to sugars, as well as
increasing carbon yield beyond what is achievable from only the carbohydrate fraction (e.g., pursuing
opportunities for value-added coproducts from lignin).

The end result of the techno-economic analysis was a predicted minimum fuel selling price (MFSP) of
$4.05/GGE (2011$) at a final upgraded total fuel yield of 78.3 GGE/dry ton of biomass or overall
carbon efficiency from biomass to fuels of 45.3% (35.7% when also considering the carbon implicit
for natural gas-derived hydrogen) for the base case pathway, assuming purchase of off-site
hydrogen. This reflects a $1.02/GGE contribution from feedstock, a $0.21/GGE contribution from

86
enzymes, and a $2.81/GGE contribution from the remainder of the conversion process (including
$0.81/GGE attributed to hydrogen costs). Although this case results in favorable fuel yields and MFSP
results, it does so at the expense of substantial natural gas-derived hydrogen import, which carries
sustainability implications. To begin understanding cost and sustainability drivers associated with
hydrogen sourcing, cursory models were also run for in situ hydrogen production directly from a
fraction of biomass hydrolysate, as well as gasification of a fraction of incoming biomass feedstock.
Given the requirement for 41% diversion of hydrolysate carbon to in situ hydrogen, the resulting
MFSP cost for the in situ approach was found to be $5.48/GGE with a fuel yield of 45.3 GGE/ton or
overall carbon efficiency from biomass of 26.2%. As the gasification scenario required a slightly
lower diversion (36%) of biomass feedstock for production of hydrogen from syngas, fuel and carbon
yields were slightly higher than the in situ scenario at 50.1 GGE/ton and 27.7% overall carbon yield,
translating to an MFSP cost of $4.95/GGE. These modeled selling prices are strictly representative of
nth-plant assumptions regarding biorefinery design, operation, and financing, and are not intended to
reflect first-of-a-kind or early-entry commercial facilities. Consistent with the biological conversion
pathway discussed in the 2013 report, the technology pathway described here carries a distinct
advantage of producing a mixture of components which may be directly utilized by the existing fuel
infrastructure and blended into existing fuel markets (namely diesel and naphtha in this case).

Beyond the detailed modeling and analysis conducted to establish the $4$5/GGE intermediate targets
for this pathway, this report also reiterates viable improvements to ultimately achieve the $3/GGE
target by 2022, based on lignin conversion to value-added coproducts as originally detailed in the 2013
report. While the analysis for lignin coproduct pathways was not repeated here, the baseline MFSP
costs from this work suggest it is reasonable to conclude that lignin deconstruction and conversion to a
number of possible coproduct options still exhibits the potential to ultimately achieve the $3/GGE
target at least similarly to the 2013 biological conversion pathway (when comparing to the in situ
hydrogen scenario here), if not more easily (when considering the ex situ purchased hydrogen
scenario).

In addition to providing an economic analysis, this report also considers sustainability metrics for the
baseline purchased hydrogen and alternative (in situ and gasification) hydrogen scenarios by tracking
and reporting on greenhouse gas emissions, fossil energy demand, and direct consumptive water use at
the modeled biorefinery facility. The sustainability analysis for the purchased hydrogen scenario
indicates a GHG profile of 6.0 kg CO2-eq/GGE, fossil energy consumption of 99.3 MJ/GGE, and water
demand of 5.8 gal/GGE. In comparison, the in situ hydrogen scenario results in a GHG profile
reducing to 1.9 kg CO2-eq/GGE, fossil energy consumption of 23.8 MJ/GGE, and water demand of 9.8
gal/GGE, while the gasification scenario is estimated at a GHG profile of 0.9 kg CO2-eq/GGE, fossil
energy consumption of 12.2 MJ/GGE, and water demand of 11.4 gal/GGE . While this is strictly
associated with the biorefinery conversion process and does not represent a full well-to-wheel LCA
analysis (which is required to more completely understand the implications for these scenarios on
overall sustainability), it does indicate notable differences between external versus internal hydrogen
sourcing and highlights the key benefit for considering the alternative internal hydrogen options. As
discussed previously, a plausible case for commercial-scale production could fall between the two
extremes, where a facility would import the maximum amount of natural gas-derived hydrogen as
practical given restrictions imposed by applicable policy guidelines, while producing the rest internally
either via hydrolysate carbon reforming and/or biomass gasification; however, additional sustainability
benefits could also be gained by considering on-site SMR hydrogen production which could alter these
trade-offs. Given these important results in both cost and sustainability implications for hydrogen

87
sourcing, as well as additional trade-offs that would become apparent when also introducing
coproduction of value-added products from lignin (or other potential carbon sources), NREL will
continue to evaluate these scenarios as further experimental data becomes available.

Similar to the 2013 biological pathway, it is worthwhile to reiterate that cellulosic ethanol production
via biochemical processing is supported by decades of research and process understanding, both at
NREL and elsewhere, thus carrying a level of certainty and credibility when making target projections
for biochemical ethanol TEA modeling, as in the case of the 2011 and even 2002 ethanol design
reports. In contrast, catalytic hydrocarbon production via aqueous phase reforming of sugars is a much
newer and more novel approach to biofuel production, with fairly limited understanding of technology
status and progress offered by information in the public domain, particularly in the context of
cellulosic hydrolysate conversion. Thus, the absolute cost values established here, as well as the
timeframe required to achieve these outcomes, inherently carry a somewhat higher degree of
uncertainty given the nascent stage of research as presented in the public domain.

6.2 Future Work

Moving forward, to ultimately achieve cost goals as well as reduce uncertainty in key areas for the
catalytic conversion pathway, a number of important bottlenecks, uncertainties, and areas for further
development are summarized below. These points are repeated from a recently published technical
memo for catalytic hydrocarbon production [19]:

Optimize upstream biomass deconstruction processes for downstream upgrading.

Catalytic conversion routes allow for more flexibility in the hydrolysate composition that can
be utilized for fuel production. Unlike biological routes which require primarily monomeric
sugars for fuel production, catalytic routes can utilize oligomeric species, carbohydrate
degradation products, and solubilized lignin intermediates. To help reduce the overall costs and
maximize biomass carbon utilization, future work is needed to develop opportunities to lower
downstream upgrading costs by tailoring the biomass deconstruction processes for catalytic
upgrading. This includes investigating alternative deconstruction routes that maximize the
desirable slate of components and minimize the cost of hydrolysate purification and
conditioning steps for catalytic upgrading. One such option could include an alternative
approach based on dehydrating sugars to produce furfurals which may phase separate out of
the aqueous stream, thereby bypassing multiple hydrolysate cleanup steps. Opportunities for
biomass preprocessing to remove components (i.e., ash) should also be explored as a means to
improve catalytic performance and minimize hydrolysate conditioning costs.
Minimize cost of hydrolysate conditioning steps. Alternative, lower cost hydrolysate
conditioning steps, including solid/liquid separations and improved purification process steps,
should be investigated. This includes developing a better understanding for particle size
requirements associated with the catalytic conversion step and associated implications on
optimizing separations designs and costs with increased certainty.
Maximize carbon selectivity towards desired fuel products. Catalysts could be designed to
have increased selectivity to hydrocarbon fuel products and to minimize side- or byproduct
steps. An understanding of catalyst fundamentals is needed to design better catalysts and
understand catalyst selectivity. Rational design and development of catalysts could be
accomplished using a combination of computational and experimental techniques.

88
 Improve catalyst lifetime and durability. Understanding catalyst deactivation mechanisms
for hydrolysate poisoning will be critical to improving catalyst lifetimes and minimizing
processing costs. Using biomass-derived hydrolysates can help identify problematic
components. Determining the impact of lignin deconstruction products on catalyst lifetimes as
a function of deconstruction routes will improve process integration and carbon efficiency for
fuel production. A better fundamental understanding of both chemical conversion and
poisoning mechanisms may help to design better catalysts that are resistant to contaminants
and poisons. Additionally, optimizing catalysts and conditions to improve catalyst WHSV and
reduce reactor volumes will also be important.
Pursue opportunities for lignin utilization. Catalytic processes are more tolerant to the
introduction of a range of carbon species than biological systems are. For these catalytic
processes, lignin utilization has great potential for improving total hydrocarbon product yields
and increasing overall carbon efficiency for biomass conversion compared to pathways focused
on carbohydrate carbon conversion alone. The key challenges for lignin usage are difficulty
solubilizing the components and avoiding side reactions that form more recalcitrant
compounds. Development and demonstration of a catalyst that can effectively convert lignin
products without severe deactivation can significantly improve overall economic viability.
Furthermore, parallel conversion pathways from lignin to other value-added coproducts will
also be an important consideration for ultimately achieving $3/GGE production costs for the
integrated pathway.
Expand analysis on hydrogen sourcing tradeoffs. While the analysis presented here
considered economic and sustainability implications for a number of possible hydrogen
sourcing scenarios strictly within the context of the biorefinery facility alone, this work could
be expanded upon to conduct a more thorough LCA analysis for the full well-to-wheel supply
chain. This would enable a more complete understanding of TEA and LCA tradeoffs
associated with hydrogen sourcing, in light of the large amount of hydrogen consumed in the
catalytic steps for this process. This expanded modeling work should also consider added
benefits that could be brought about by considering a fully integrated on-site SMR hydrogen
plant, as this would allow for additional optimization and recycle opportunities beyond model
cases that assume hydrogen import from off-site.

89
References
NREL milestone reports cited below cannot be accessed outside of NREL and DOE. Readers may
contact the authors of the specific reference to determine if this information has been made public
since publication of this design report.

1. Wooley, R.J., et al. Lignocellulosic Biomass to Ethanol Process Design and Economics Utilizing
Co-Current Dilute Acid Prehydrolysis and Enzymatic Hydrolysis Current and Futuristic Scenarios.
NREL/TP-580-26157. Golden, CO: National Renewable Energy Laboratory, 1999. Available from:
http://www.nrel.gov/docs/fy99osti/26157.pdf.
2. Humbird, D., et al. Process Design and Economics for Biochemical Conversion of Lignocellulosic
Biomass to Ethanol: Dilute-Acid Pretreatment and Enzymatic Hydrolysis of Corn Stover.
NREL/TP-5100-47764. Golden, CO: National Renewable Energy Laboratory, 2011. Available
from: http://www.nrel.gov/docs/fy11osti/47764.pdf.
3. Schell, D. "Biochemical Processing Integration," 2013 Peer Review Report: Review Panel
Summary Report and Project Results. Alexandria, VA: U.S. Department of Energy (DOE),
Bioenergy Technologies Office (BETO), 2014.
4. Davis, R. "Biochemical Platform Analysis," 2013 Peer Review Report: Review Panel Summary
Report and Project Results. Alexandria, VA: U.S. Department of Energy (DOE), Bioenergy
Technologies Office (BETO), 2014.
5. Tao, L., et al. NREL Milestone Report: FY 2012 Biochemical Platform State of Technology Report.
September 2012 (internal only).
6. Tao, L., et al. NREL 2012 Achievement of Ethanol Cost Targets: Biochemical Ethanol
Fermentation via Dilute-Acid Pretreatment and Enzymatic Hydrolysis of Corn Stover. NREL/TP-
5100-61563. Golden, CO: National Renewable Energy Laboratory, 2014. Available from:
http://www.nrel.gov/docs/fy14osti/61563.pdf.
7. U.S. Department of Energy (DOE). Bioenergy Technologies Office Multi-Year Program Plan: May
2013. Washington, DC: U.S. Department of Energy, 2013. Available from:
http://www1.eere.energy.gov/biomass/pdfs/mypp_may_2013.pdf.
8. Davis, R., et al. Process Design and Economics for the Conversion of Lignocellulosic Biomass to
Hydrocarbons-Dilute-Acid and Enzymatic Deconstruction of Biomass to Sugars and Biological
Conversion of Sugars to Hydrocarbons. NREL/TP-5100-60223. Golden, CO: National Renewable
Energy Laboratory, 2013. Available from: http://www.nrel.gov/docs/fy14osti/60223.pdf.
9. ASPEN, Release 7.2, in AspenPlus. Cambridge, MA: Aspen Technology Inc., 2007.
10. Wooley, R.J.; Putsche, V.; Ibsen, K. Development of an ASPEN PLUS Physical Property Database
for Biofuels Components. NREL Technical Memo, Document #4141. Golden, CO: National
Renewable Energy Laboratory, 1999. 18 pp. Available from:
http://www.nrel.gov/docs/fy11osti/47764.pdf.
11. Wooley, R.J.; Putsche, V. Development of an ASPEN PLUS Physical Property Database for
Biofuels Components. NREL/MP-425-20685. Golden, CO: National Renewable Energy
Laboratory, 1996. Available from: http://biodev.nrel.gov/pdfs/3955.pdf.
12. Muth, D.; Jacobson, J.; Cafferty, K.; Jeffers, R. Feedstock Pathway for Cellulosic Sugar
Conversion Pathway. INL Technical Memo, INL/MIS-13-29309. 2013. (internal only).
13. Chen, S.F., et al. "Compositional analysis of water-soluble materials in corn stover." Journal of
Agricultural and Food Chemistry (55:15), 2007 ; pp. 5912-5918.
14. Bower, S., et al. "Modeling sucrose hydrolysis in dilute sulfuric acid solutions at pretreatment
conditions for lignocellulosic biomass." Bioresource Technology (99:15). 2008: pp. 7354-7362.

90
15. Jones, S., et al. Process Design and Economics for the Conversion of Algal Biomass to
Hydrocarbons: Whole Algae Hydrothermal Liquefaction and Upgrading. PNN-L23227. Richland,
WA: Pacific Northwest National Laboratory, 2014. Available from:
https://hydrogen.gov/bioenergy/pdfs/pnnl_whole_algae_liquefaction.pdf.
16. Jones, S., et al. Process Design and Economics for the Conversion of Lignocellulosic Biomass to
Hydrocarbon Fuels: Fast Pyrolysis and Hydrotreating Bio-oil Pathway. PNNL-23053;NREL/TP-
5100-61178. Richland, WA: Pacific Northwest National Laboratory; Golden, CO: National
Renewable Energy Laboratory (NREL), 2013. Available from:
http://www.osti.gov/scitech/biblio/1126275.
17. "Lower and higher heating values of fuels." U.S. DOE Hydrogen Analysis Resource Center, 2012.
Available from:
http://hydrogen.pnl.gov/cocoon/morf/hydrogen/site_specific/fuel_heating_calculator.
18. Hess, J.R. "Uniform Format Design and Depot Preprocessing." Presented at the PNWER 2010
Annual Summit, July 16-20, 2010, Calgary, Alberta.
19. Biddy, M.; Jones, S. Catalytic Upgrading of Sugars to Hydrocarbons Technology Pathway.
NREL/TP-5100-58055 and PNNL-22319. Richland, WA: Pacific Northwest National Laboratory;
Golden, CO: National Renewable Energy Laboratory (NREL), 2013. Available from:
http://www.nrel.gov/docs/fy13osti/58055.pdf.
20. Galbe, M.; Zacchi, G. "Pretreatment of lignocellulosic materials for efficient bioethanol
production," Olsson, L., ed. Biofuels. Berlin, Germany: Springer, 2007; pp. 41-65.
21. Chen, X., et al. "Improved ethanol yield and reduced Minimum Ethanol Selling Price (MESP) by
modifying low severity dilute acid pretreatment with deacetylation and mechanical refining: 1)
Experimental." Biotechnology for Biofuels (5:1), 2012; pp. 60.
22. Fagerson, I.S. "Thermal degradation of carbohydrates; a review." Journal of Agricultural and Food
Chemistry (17:4), 1969; pp. 747-750.
23. "Filtration of Hydrotreater Feed Streams for Catalyst Process Protection." 3M, October 2011.
Available from:
http://multimedia.3m.com/mws/mediawebserver?mwsId=SSSSSuH8gc7nZxtUNY_GOx29evUqe1
7zHvTSevTSeSSSSSS--&fn=70-0201-8620-4.pdf.
24. Lindstrom, T.H.; Levesque, F.; Cathcart, N.G. "Protect desulfurization catalyst with prefiltration
systems." Hydrocarbon Processing 2003; pp. 49-51. Available from:
http://www.pall.com/pdfs/Fuels-and-Chemicals/HCP203.pdf.
25. Bartholomew, C. "Catalyst Deactivation and Regeneration," McGraw Kirk-Othmer Encyclopedia
of Chemical Technology. 4th edition. New York, NY: John Wiley & Sons, Inc., 2000.
26. Meerman, H.J.; Kelley, A.S.; Ward, M. "Advances in Protein Expression in Filamentous Fungi,"
Baneyx, F., ed. Protein Expression Technologies: Current Status and Future Trends. Norfolk, UK:
Horizon Bioscience, 2004; p. 345.
27. Warzywoda, M., " Process for producing cellulolytic enzymes," U.S. Patent No. 4,762,788, 1984.
28. Emme, B., personal communication to Humbird, D., editor, Novozymes, Inc., 2009.
29. Wiseman, A. "Biochemical Engineering and Biotechnology Handbook, Atkinson, B. and
Mavituna, F., eds." Chemistry in Britain (19:5), 1983; pp. 423-424.
30. Schell, D., et al. Technical and Economic Analysis of an Enzymatic Hydrolysis Based Ethanol
Plant. Golden, CO: Solar Energy Research Institute, 1991. Available from:
http://biodev.nrel.gov/bcfcdoc/3892.pdf.
31. Cortright, R.D.; Davda, R.R.; Dumesic, J.A. "Hydrogen from catalytic reforming of biomass-
derived hydrocarbons in liquid water." Nature (418:6901), 2002; pp. 964-967.
32. Cortright, R.D.; Blommel, P.G., "Synthesis of liquid fuels and chemicals from oxygenated
hydrocarbons," U.S. Patent No. 8,455,705, 2013.

91
33. Blommel, P., et al., "Method and systems for making distillate fuels from biomass," U.S. Patent
No. 2012/0198760 A1, 2012.
34. Spath, P., et al. Biomass to Hydrogen Production Detailed Design and Economics Utilizing the
Battelle Columbus Laboratory Indirectly Heated Gasifier. NREL/TP-510-37408. Golden, CO:
National Renewable Energy Laboratory, 2005. Available from:
http://www.nrel.gov/docs/fy05osti/37408.pdf.
35. Gurbuz, E., et al. "Role of acid catalysis in the conversion of lignocellulosic biomass to fuels and
chemicals," Triantafyllidis, K.S.; Lappas, A.A.; Stcker, M., eds. The role of catalysis for the
sustainable production of bio-fuels and bio-chemicals. Amsterdam/Boston: Elsevier, 2013.
36. Twigg, M.V., ed. Catalyst Handbook. Victoria, Australia: CSIRO, 1989.
37. Benesi, H.A. "Acidity of Catalyst Surfaces. I. Acid Strength from Colors of Adsorbed Indicators."
Journal of the American Chemical Society (78:21), 1956; pp. 5490-5494.
38. West, R.M., et al. "Catalytic conversion of biomass-derived carbohydrates to fuels and chemicals
by formation and upgrading of mono-functional hydrocarbon intermediates." Catalysis Today
(147:2), 2009; pp. 115-125.
39. Weingarten, R., et al. "Conversion of glucose into levulinic acid with solid metal(IV) phosphate
catalysts." Journal of Catalysis (304), 2013; pp. 123-134.
40. Saha, B.; De, S.; Fan, M.H. "Zr(O)Cl-2 catalyst for selective conversion of biorenewable
carbohydrates and biopolymers to biofuel precursor 5-hydroxymethylfurfural in aqueous medium."
Fuel (111), 2013; pp. 598-605.
41. Kubicka, D.; Kubickova, I.; Cejka, J. "Application of Molecular Sieves in Transformations of
Biomass and Biomass-Derived Feedstocks." Catalysis Reviews-Science and Engineering (55:1),
2013; pp. 1-78.
42. Gallo, J.M.R., et al. "Production of Furfural from Lignocellulosic Biomass Using Beta Zeolite and
Biomass-Derived Solvent." Topics in Catalysis (56:18-20), 2013; pp. 1775-1781.
43. Gallo, J.M.R., et al. "Production and upgrading of 5-hydroxymethylfurfural using heterogeneous
catalysts and biomass-derived solvents." Green Chemistry (15:1), 2013; pp. 85-90.
44. Weingarten, R., et al. "Design of solid acid catalysts for aqueous-phase dehydration of
carbohydrates: The role of Lewis and Bronsted acid sites." Journal of Catalysis (279:1), 2011; pp.
174-182.
45. Palkovits, R. "Cellulose and Heterogeous Catalysis - A Combination for Future." Chemie Ingenieur
Technik (83:4), 2011; pp. 411-419.
46. Crisci, A.J., et al. "Add-Functionalized SBA-15-Type Silica Catalysts for Carbohydrate
Dehydration." Acs Catalysis (1:7), 2011; pp. 719-728.
47. Xing, R., et al. "Production of jet and diesel fuel range alkanes from waste hemicellulose-derived
aqueous solutions." Green Chemistry (12:11), 2010; pp. 1933-1946.
48. Li, N.; Huber, G.W. "Aqueous-phase hydrodeoxygenation of sorbitol with Pt/SiO2-Al2O3:
Identification of reaction intermediates." Journal of Catalysis (270:1), 2010; pp. 48-59.
49. Huber, G.W.; Dumesic, J.A. "An overview of aqueous-phase catalytic processes for production of
hydrogen and alkanes in a biorefinery." Catalysis Today (111:1-2), 2006; pp. 119-132.
50. "Virent and Fellow NABC CLS Collaborators Make BioFormate Gasoline from Biomass."
National Advanced Biofuels Consortium (NABC), May 23, 2011. Available from:
http://www.nabcprojects.org/pdfs/virent_nabc_make_bioformate_gasoline.pdf.
51. "CLS Proven in Stage I, Catalyst Optimization Ahead." National Advanced Biofuels Consortium
(NABC), April 19, 2012. Available from:
http://www.nabcprojects.org/pdfs/cls_proven_in_stage_i.pdf.
52. Vogel, J.; Westphal, G.; Pippig, C. "Mutarotation von D-Glucose in Abhngigkeit vom
Reaktionsmilieu." Food / Nahrung (32:7), 1988; pp. 709-714.

92
53. Huber, G.W., et al. "Production of liquid alkanes by aqueous-phase processing of biomass-derived
carbohydrates." Science (308:5727); 2005; pp. 1446-1450.
54. Jin-Woong, K., et al., "Method for preparing sugar alcohols using ruthenium zirconia catalyst,"
Patent No. WO 2006093364 A1, 2006.
55. Jin-Woong, K., et al., "Method for preparing sugar alcohols by catalytic hydrogenation of sugars,
Patent No. WO 2005021475 A1, 2005.
56. Arena, B.J., "Hydrogenation of saccharides," U.S. Patent No. 4380679 A, 1983.
57. Elliott, D.C., et al., "Converting sugars to sugar alcohols by aqueous phase catalytic
hydrogenation," U.S. Patent No. 6570043 B2, 2003.
58. Yoshino, A.; Izumori, K.; Takahashi, T., "Hydrogenation a ketohexose using Group 8 catalyst,"
U.S. Patent No. 20040143024 A1, 2004.
59. Werpy, T.A., et al., "Using a rhenium containing catalyst; catalytic hydrogenation of sugars, sugar
alcohol forming propylene glycol," U.S. Patent No. 7038094 B2, 2006.
60. Huber, G.W.; Cortright, R.D.; Dumesic, J.A. "Renewable alkanes by aqueous-phase reforming of
biomass-derived oxygenates." Angewandte Chemie-International Edition (43:12), 2004; pp. 1549-
1551.
61. Huber, G.W., et al. "Aqueous-phase reforming of ethylene glycol over supported Pt and Pd
bimetallic catalysts." Applied Catalysis B-Environmental (62:3-4), 2006; pp. 226-235.
62. Len, M., et al. "Consequences of the ironaluminium exchange on the performance of
hydrotalcite-derived mixed oxides for ethanol condensation." Applied Catalysis B: Environmental
(102:3), 2011; pp. 590-599.
63. Marker, T.L., et al. "Opportunities for Biorenewables in Oil Refineries." UOP LLC, 2005.
Available from: http://www.osti.gov/bridge/servlets/purl/861458-Wv5uum/861458.pdf.
64. "Products: Diesel Fuel. Products." Virent. Available from:
http://www.virent.com/products/diesel-fuel/.
65. Bailor, J.P.; Dauber, L.S., "Catalyst regeneration process," U.S. Patent No. 4406775, 1983.
66. Cerven, L., ed. Catalytic Hydrogenation. Elsevier Science, 1986.
67. Froment, G.F.; Delmon, B., eds. Catalyst Deactivation 1994. Elsevier Science, 1994.
68. Dutta, A., et al. Process Design and Economics for Conversion of Lignocellulosic Biomass to
Ethanol, Thermaochemical Pathway by Indirect Gasification and Mixed Alcohol Synthesis.
NREL/TP-5100-51400. Golden, CO: National Renewable Energy Laboratory, 2011. Available
from: http://www.nrel.gov/docs/fy11osti/51400.pdf.
69. Aden, A., et al. Lignocellulosic Biomass to Ethanol Process Design and Economics Utilizing Co-
Current Dilute Acid Prehydrolysis and Enzymatic Hydrolysis for Corn Stover. NREL/TP-510-
32438. Golden, CO: National Renewable Energy Laboratory, 2002. Available from:
http://www.nrel.gov/docs/fy02osti/32438.pdf.
70. "Chemical Engineering Magazine Plant Cost Index." Chemical Engineering Magazine, 2011.
71. "U.S. Producer Price Indexes Chemicals and Allied Products/Industrial Inorganic Chemicals
Index." Chemical Economics Handbook. Menlo Park, CA: SRI Consulting, 2008.
72. "National Employment, Hours, and Earnings Catalog, Industry: Chemicals and Allied Products,
1980-2009." U.S. Bureau of Labor Statistics, 2009. Available from:
http://data.bls.gov/cgi-bin/srgate.
73. "Electricity Wholesale Market Data." U.S. Energy Information Admninistration (EIA), 2013.
Available from: http://www.eia.gov/electricity/wholesale.
74. "Average Retail Price of Electricity to Ultimate Customers: Total by End-Use Sector, 2004 - April
2014." U.S. Energy Information Admninistration (EIA), 2014. Accessed July 11, 2014:
http://www.eia.gov/electricity/monthly/epm_table_grapher.cfm?t=epmt_5_3.

93
75. Dillich, S.; Ramsden, T.; Melina, M. " Hydrogen Production Cost Using Low-Cost Natural Gas"
DOE Hydrogen and Fuel Cells Program Record #12024, 2012. Available from:
http://www.hydrogen.energy.gov/pdfs/12024_h2_production_cost_natural_gas.pdf.
76. "Table 7U.S. wholesale list price for glucose syrup, Midwest markets, monthly, quarterly, and
by calendar and fiscal year." U.S. Department of Agriculture, 2013. Available from:
http://www.ers.usda.gov/datafiles/Sugar_and_Sweeteners_Yearbook_Tables/World_and_US_Suga
r_and_Corn_Sweetener_Prices/TABLE07.XLS.
77. Brown and Caldwell. Process Design of Wastewater Treatment for the NREL Cellulosic Ethanol
Model. NREL/SR-5100-51838. Golden, CO: National Renewable Energy Laboratory, 2011.
Availale from: http://www.nrel.gov/docs/fy11osti/51838.pdf
78. Brown and Caldwell. Process Design of Wastewater Treatment for the NREL Cellulosic Ethanol
Model. Golden, CO: National Renewable Energy Laboratory, 2012.
79. Peters, M.S., et al. Plant design and economics for chemical engineers. New York, NY: McGraw-
Hill, 2003.
80. Short, W.; Packey, D.J.; Holt, T. A Manual for the Economic Evaluation and Energy Efficiency
and Renewable Energy Technologies. NREL/TP-462-5173. Golden, CO: National Renewable
Energy Laboratory, 1995.
81. "How to Depreciate Property." U.S. Internal Revenue Service, 2009. Available from:
http://www.irs.gov/pub/irs-pdf/p946.pdf.
82. Perry, R.H.; Green, D.W. Perry's Chemical Engineers' Handbook 7th Edition. New York, NY:
McGraw-Hill, 1997.
83. Gary, J.H.; Handwerk, G.E. Petroleum Refining: Technology and Economics: Third Edition. New
York, NY: Marcel Dekker Inc., 1994.
84. Cran, J. "Improved factored method gives better preliminary cost estimates." Chemical Engineer (
88:7), 1981; pp. 65-79.
85. "Fuels & Fuel Additives: Renewable Fuel Standard (RFS)." U.S. Environmental Protection
Agency, 2012. Available from: http://www.epa.gov/otaq/fuels/renewablefuels.
86. "H2A Production Model. Version 3." U.S. Department of Energy Hydrogen and Fuel Cells
Program. Available from: http://www.hydrogen.energy.gov/h2a_production.html.
87. "World LNG Estimated November 2013 Landed Prices." Federal Energy Regulatory Commission,
2013. Available from: http://www.ferc.gov/market-oversight/mkt-gas/overview/ngas-ovr-lng-wld-
pr-est.pdf.
88. "Natural Gas Prices." U.S. Energy Information Administration, 2013. Accessed June 18, 2013:
http://www.eia.gov/dnav/ng/ng_pri_sum_dcu_nus_m.htm.
89. Chen, X., et al. "Improved ethanol yield and reduced Minimum Ethanol Selling Price (MESP) by
modifying low severity dilute acid pretreatment with deacetylation and mechanical refining: 1)
Experimental." Biotechnology for Biofuels (5:1), 2012; pp. 60.
90. Chen, X., et al. "Comparison of different mechanical refining technologies on the enzymatic
digestibility of low severity acid pretreated corn stover." Bioresource Technology (147:0), 2013;
pp. 401-408.
91. Chen, X., et al. "A highly efficient dilute alkali deacetylation and mechanical (disc) refining
process for the conversion of renewable biomass to lower cost sugars." Biotechnology for Biofuels
(7:1), 2014; pp. 98.
92. Tao, L., et al. "Improved ethanol yield and reduced MESP by modifying low severity dilute acid
pretreatment with deacetylation and mechanical refining: 2) Techno-economic Analysis."
Biotechnology for Biofuels (5:69), 2012.
93. "DOE H2A Analysis: Production Case Studies." U.S. Department of Energy (DOE) Hydrogen and
Fuel Cells Program. Available from: http://www.hydrogen.energy.gov/h2a_prod_studies.html.

94
94. Cortright, R.D.; Dumesic, J.A. "Low-temperature hydrogen production from oxygenated
hydrocarbons," U.S. Patent No. 6964757 B2, 2005.
95. Cortright, R.D.; Dumesic, J.A. "Hydrocarbon reactivity over tin-modified platinum catalysts."
Abstracts of Papers of the American Chemical Society (223), 2002; pp. U652-U652.
96. Huber, G.W.; Shabaker, J.W.; Dumesic, J.A. "Raney Ni-Sn catalyst for H-2 production from
biomass-derived hydrocarbons." Science (300:5628), 2003; pp. 2075-2077.
97. Shabaker, J.W., et al. "Aqueous-phase reforming of methanol and ethylene glycol over alumina-
supported platinum catalysts." Journal of Catalysis (215:2), 2003; pp. 344-352.
98. Davda, R.R.; Dumesic, J.A. "Renewable hydrogen by aqueous-phase reforming of glucose."
Chemical Communications (1), 2004; pp. 36-37.
99. Tan, E. Process Design and Economics for the Conversion of Lignocellulosic Biomass to High
Octane Gasoline: Thermochemical Research Pathway with Indirect Gasification and Methanol
Intermediate (forthcoming).
100. Jacobson, J.J.; Cafferty, K.; Roni, M.S.; Lamers, P.; Kenney, K. Feedstock and Conversion Supply
System Design and Analysis--"The Feedstock Logistics Design Case". INL/EXT-14-32377. Idaho
Falls, ID: Idaho National Laboratory, September 2014.
101. SimaPro, v.7.3. Amersfoort, Netherlands: Product Ecology Consultants, 2011.
102. Hsu, D.; Heath, G.; Wolfrum, E.; Mann, M.K.; Aden, A. "Life Cycle Environmental Impacts of
Selected U.S. Ethanol Production and Use Pathways in 2022." Environ. Sci. Technol (44:13), 2010;
pp. 52895297.
103. "GREET Publications Listing." Argonne National Laboratory, 2013. Accessed July 14, 2014:
https://greet.es.anl.gov/publications.
104. Ecoinvent, v.2.2. Duebendorf, Switzerland: Swiss Center for Life Cycle Inventories, 2010.
105. U.S. Life Cycle Inventory Database, v. 1.6.0. Golden, CO: National Renewable Energy Laboratory,
2008.
106. Biorecro. Global Status of BECCS Projects 2010. Canberra, Australia: Global CCS Institute, 2010.
107. Fischer, B.S., et al. "Issues related to mitigation in the long-term context," Chapter 3. Climate
Change 2007: Mitigation of Climate Change: Contribution of Working Group III to the Fourth
Assessment Report of the Intergovernmental Panel on Climate Change, 2007. Cambridge, MA:
Cambridge University Press, 2007; pp. 169-250.
108. "Regulation of Fuels and Fuel Additives: RFS Pathways II, and Technical Amendments to the RFS
Standards and E15 Misfueling Mitigation Requirements." U.S. Environmental Protection Agency,
2014. Available from: http://www.gpo.gov/fdsys/pkg/FR-2014-07-18/pdf/2014-16413.pdf.
109. DOE. "Conversion Technologies for Advanced Biofuels: Preliminary Roadmap & Workshop
Report." Washington, DC: U.S. Department of Energy, 2012. Available from:
https://www.fedconnect.net/fedconnect/?doc=DE-FOA-0000775&agency=DOE.
110. Keller and Heckman LLP. Assessment Plan for Corn Steep Liquor (CAS #66071-94-1) in
Accordance with the U.S. EPA High Production Volume Chemical Challenge Program. Prepared
for The Corn Refiners Association, 2006.

95
Appendix A. Individual Equipment Costs Summary
The following table shows abbreviated specifications, purchased cost, and installed cost for each piece
of equipment in this process design. Although each piece of equipment has its own line, many were
quoted as part of a package so their scaling calculations are not shown. NREL and Harris Group would
like to acknowledge the equipment vendors who assisted us with the cost estimation effort for this
design report.

96
 Mechanical Equipment List: Advanced Biofuels Engineering Design: Dilute Acid
Pretreatment, Enzymatic Hydrolysis, Hydrolysate Catalytic Conversion, Scaled Installed Costs
11/11/2014 Hydrotreating to HC

NUM REQD

Scaling Exp
Scaling Val

Inst Factor

Size Ratio
EQPT NO

New Val
Year of
Quote
Purch Cost in Scaled Purch Purch Cost in Inst Cost in

Units
EQUIPMENT TITLE DESCRIPTION HP MATERIAL $ Scaling Variable
Base Yr Cost Proj year Proj year

C- 101 Transfer Conveyor 160 MTPH ea., enclosed, 60 in. x 65 ft. 20 hp CS 2 $5,397,000 2009 $5,397,000 STRM.101 94697 kg/hr 0.60 1.7 104167 1.10 $5,714,628 $6,413,216 $10,902,467
C- 102 High Angle Transfer Conveyor 160 MTPH ea., enclosed, 72 in. wide 50 hp CS 2 INCLUDED
C- 103 Reversing Load-in Conveyor 320 MT / hr, enclosed, 84 in. wide 20 hp CS 1 INCLUDED
C- 104 Dome Reclaim System 100 MTPH ea. 45 KW CS 2 $3,046,000 2009 $3,046,000 STRM.101 94697 kg/hr 0.60 1.7 104167 1.10 $3,225,265 $3,619,540 $6,153,218
C- 105 Reclaim Conveyor 100 MTPH ea., enclosed, 48 in. x 125 10 hp CS 2 INCLUDED
C- 106 High Angle Transfer Conveyor 100 MTPH, enclosed, 72 in. wide 20 hp CS 1 INCLUDED
C- 107 Elevated Transfer Conveyor 100 MTPH, enclosed, 48 in. x 200 ft. 10 hp CS 1 INCLUDED
C- 108 Process Feed Conveyor 70 MTPH ea., enclosed, 42 in. x 25 ft. 5 hp ea. CS 1 INCLUDED
M- 101 Truck Scale 10' x 70', 200,000 lb CONCRETE 2 $110,000 2009 $110,000 STRM.101 94697 kg/hr 0.60 1.7 104167 1.10 $116,474 $130,712 $222,211
M- 102 Truck Dumper 70' x 55 ton x 63 degree 2 x 50 hp CS 2 $484,000 2009 $484,000 STRM.101 94697 kg/hr 0.60 1.7 104167 1.10 $512,485 $575,134 $977,727
M- 103 Truck Dumper Hopper 3500 cu.ft. hopper w/ drag chain 50 hp CS 2 $502,000 2009 $502,000 STRM.101 94697 kg/hr 0.60 1.7 104167 1.10 $531,544 $596,523 $1,014,089
M- 104 Concrete Feedstock Storage 98 ft. dia., 160 ft. high., 4000 MT CONCRETE 2 $3,500,000 2009 $3,500,000 STRM.101 94697 kg/hr 0.60 1.7 104167 1.10 $3,705,984 $4,159,025 $7,070,342
M- 105 Belt Scale Scale plus processor CS 2 $10,790 2009 $10,790 STRM.101 94697 kg/hr 0.60 1.7 104167 1.10 $11,425 $12,822 $21,797
M- 106 Dust Collection System 8500 ACFM 25 hp CS 6 $279,900 2009 $279,900 STRM.101 94697 kg/hr 0.60 1.7 104167 1.10 $296,373 $332,603 $565,425
Area 100 Totals $14,114,178 $15,839,574 $26,927,276

97
 NUM REQD

Scaling Exp
Scaling Val

Inst Factor

Size Ratio
EQPT NO

New Val
Year of
Quote
Purch Cost in Scaled Purch Purch Cost in Inst Cost in

Units
EQUIPMENT TITLE DESCRIPTION HP MATERIAL $ Scaling Variable
Base Yr Cost Proj year Proj year

Mechanical Equipment List Scaled Installed Costs

S- 201 In-line Sulfuric Acid Mixer Kynar lined - 600 gpm H2O - 5 gpm SS304 1 $6,000 2009 $6,000 strm.a200.214 136260 kg/hr 0.50 1.0 806 0.01 $462 $518 $518
A- 203 Blowdown Tank Agitator Side-mounted, 3 x 75 hp. ( 170 KW) 170 KW 316LSS 1 INCLUDED
A- 204 Flash Tank Agitator Side-mounted, 3 x 75 hp. ( 170 KW) 170 KW 316LSS 3 $90,000 2009 $90,000 strm.a200.254 252891 kg/hr 0.50 1.5 216084 0.85 $83,193 $93,363 $140,045
A- 209 Ammonia Addition Tank 10 hp SS 1 $21,900 2009 $21,900 strm.a200.228 410369 kg/hr 0.50 1.5 324669 0.79 $19,479 $21,861 $32,791
A- 210 Ammonia Static Mixer SS 1 $5,000 2009 $5,000 strm.a200.275 157478 kg/hr 0.50 1.0 108584 0.69 $4,152 $4,659 $4,659
C- 201 Transfer Conveyor 800 mm x 7600 mm (2'-8" x 25') 60 hp ea 316LSS 2 INCLUDED
C- 202 Distribution Conveyor 800 mm x 7600 mm (2'-8" x 25') 60 hp ea 316LSS 2 INCLUDED
C- 203 Overfeed Conveyor 800 mm x 7600 mm (2'-8" x 25') 60 hp ea 316LSS 4 INCLUDED
C- 204 Pressurized Heating Screw 2500 mm x 9500 mm (8' x 31') 100 hp Dup. 2205 1 INCLUDED
C- 205 Pressurized Pre-heater 850 mm x 3500 mm (2'-10" x 12') 50 hp Dup. 2205 2 INCLUDED
C- 206 Pressurized Transport - No. 1 900 mm x 3500 mm (3' x 12') 25 hp Incoloy 825- 1 INCLUDED
C- 207 Pressurized Transport - No. 2 1200 mm x 3500 mm (4' x 12') 50 hp Incoloy 825- 1 INCLUDED
H- 201 Pretreatment Water Heater 29.9 MMBTU 304SS 1 $92,000 2010 $92,000 Heat.A200.QH201 -8 Gcal/hr 0.70 2.2 -11 1.41 $117,262 $124,692 $274,323
H- 244 Waste Vapor Condenser Copied H-201 304SS 1 $34,000 2009 $34,000 Heat.A200.QH244 2 Gcal/hr 0.70 2.2 1 0.77 $28,403 $31,876 $70,126
60 cu. m. (2100 cu.ft.) with
M- 201 Doffing Roll Storage Bins 54 hp ea 304/316SS 2 INCLUDED
conveyors/scrapers
M- 202 Pin Drum Feeder 7.5 hp 316LSS 2 INCLUDED
M- 203 Plug Screw Feeder 1510 hp 316LSS 2 INCLUDED
M- 204 Prehydrolysis/Vertical 16' x 62' - 10 min. residence time Dup. 2205 1 INCLUDED
M- 205 Pin Drum Feeder 7.5 hp Incoloy 825- 2 INCLUDED
M- 206 Plug Screw Feeder 1510 hp Incoloy 825- 2 INCLUDED
M- 207 Pretreatment Reactor 3,000 mm x 10,300 mm each - scaled 135 hp Incoloy 825- 5 $16,400,000 2013 $31,200,000 SCIS.a200.211b 63166 kg/hr 0.60 1.5 62942 1.00 $31,133,709 $31,192,291 $46,788,436
P- 201 Sulfuric Acid Pump 9 gpm, 245 FT TDH 316SS 1 $8,000 2009 $8,000 strm.710 3720 kg/hr 0.80 2.3 2240 0.60 $5,332 $5,984 $13,763
P- 203 Blowdown Tank Discharge 1900 gpm, 150 FT TDH 125 316SS 1 $25,635 2010 $25,635 strm.a200.222 292407 kg/hr 0.80 2.3 226960 0.78 $20,932 $22,258 $51,193
P- 204 Flash Tank Discharge Pump 900 gpm, 150 FT TDH 75 316SS 1 $30,000 2009 $30,000 strm.a200.254 204390 kg/hr 0.80 2.3 216084 1.06 $31,366 $35,200 $80,960
P- 209 Hydrolyzate Pump 1771 gpm, 150 FT TDH 100 316SS 1 $22,500 2009 $22,500 strm.a200.228 402194 kg/hr 0.80 2.3 324669 0.81 $18,958 $21,275 $48,933
T- 201 Sulfuric Acid Tank 12,800 gal, 24hr residence time PLASTIC 1 $6,210 2010 $6,210 strm.710 1981 kg/hr 0.70 3.0 2240 1.13 $6,768 $7,197 $21,591
T- 203 Blowdown Tank 23' x 48' - 25 min. - 110,000 gal. SS316 1 INCLUDED
T- 204 Flash Tank 23' x 48' - 110,000 gal. SS316 1 $511,000 2009 $511,000 strm.a200.223 264116 kg/hr 0.70 2.0 226960 0.86 $459,546 $515,724 $1,031,447
T- 209 Ammonia Addition Tank 118,000 gal, 1hr residence time SS304 1 $236,000 2009 $236,000 strm.a200.228 410369 kg/hr 0.70 2.0 324669 0.79 $200,308 $224,795 $449,590
T- 225 Deacetylation Reactor 14' x 30' vessel, quad discharge screw 15hp ea SS316 3 $780,000 2013 $780,000 strm.a200.211a 277167 kg/hr 0.70 1.7 278573 1.01 $2,348,303 $2,352,721 $3,999,626
Deacetylation Reactor Feed and discharge drag conveyors 40 hp SS316 3 $110,000 2013 $110,000 strm.a200.211a 277167 kg/hr 0.80 1.7 278573 1.01 $331,338 $331,962 $564,335
Area 200 Totals $34,809,510 $34,986,375 $53,572,336

98
 NUM REQD

Scaling Exp
Scaling Val

Inst Factor

Size Ratio
EQPT NO

New Val
Year of
Quote
Purch Cost in Scaled Purch Purch Cost in Inst Cost in

Units
EQUIPMENT TITLE DESCRIPTION HP MATERIAL $ Scaling Variable
Base Yr Cost Proj year Proj year

Mechanical Equipment List Scaled Installed Costs

A- 300 Enzymatic Hydrolysis Reactor 30 hp SS304 1 $52,500 2009 $52,500 inumsscf 1 ea 1.00 1.5 6 6.00 $315,000 $353,507 $530,261
A- 306 Enzymatic Hydrolysis Storage 20 hp SS304 2 $68,300 2009 $68,300 strm.a300.306a 425878 kg/hr 0.50 1.5 331610 0.78 $60,269 $67,636 $101,455
A- 308 Enzyme-Hydrolysate Mixer Inline mixer 1673 gpm 100 hp SS316 1 $109,000 2009 $109,000 strm.a300.t310fd 379938 kg/hr 0.50 1.7 331610 0.87 $101,832 $114,281 $194,277
F- 300 Enzymatic Hydrolysis Reactor 1,000,000 gallon ea 304SS 12 $10,128,000 2009 $10,128,000 inumsscf 12 ea 1.00 1.5 6 0.50 $5,064,000 $5,683,052 $8,524,578
H- 301 Hydrolyzate Cooler Plate & frame 32.5 MMBTU/hr SS 304 1 $85,000 2010 $85,000 heat.a300.qc301 8 Gcal/hr 0.70 2.2 7 0.82 $74,086 $78,781 $173,317
P- 306 Hydrolyzate Storage Transfer 2152 gpm, 171 FT TDH 125 hp 316SS 1 $26,800 2009 $26,800 strm.a300.306a 488719 kg/hr 0.80 2.3 331610 0.68 $19,651 $22,054 $50,723
P- 310 Saccharification Transfer 352 gpm, 150 FT TDH 20 316SS 5 $47,200 2009 $47,200 strm.a300.306 421776 kg/hr 0.80 2.3 331610 0.79 $38,938 $43,699 $100,507
306 Enzymatic Hydrolysis Storage
T- 1,200,000 gallon 316SS 1 $1,317,325 2011 $1,317,325 strm.a300.306a 328984 kg/hr 0.70 1.8 331610 1.01 $1,324,678 $1,324,678 $2,384,421
A Tank
T- 310 Saccharification Tank 250,000 gal each - 19' dia. x 120' tall 304SS 8 $3,840,000 2009 $3,840,000 strm.a300.306 421776 kg/hr 0.70 2.0 331610 0.79 $3,244,992 $3,641,678 $7,283,356
Lignin Filter (after enzymatic 660 hp
S- 205 (4) 170 m2 horizontal belt filters 304SS 5 $2,152,500 2013 $2,152,500 scis.a200.solids 328984 kg/hr 1.00 1.7 15801 1.00 $10,762,500 $10,782,751 $18,330,677
hydrolysis) ea
E- 303 Sugar Concentration Mechanical vapor recompression 3600 KW 304SS 1 $6,370,000 2013 $6,370,000 strm.sugmix.301SUG 244084 kg/hr 0.70 2.0 360732 1.48 $8,373,199 $8,388,954 $16,777,909
F- 320 Polishing Filter Ceramic microfiltration 316L 4 $440,000 2014 $440,000 STRM.REFINE.101 53,204 kg/hr 0.90 1.8 55,664 1.05 $1,833,058 $1,867,169 $3,360,905
Polished Hydrolysate Storage
T- 321
Tank
8500 gal - 20 min residence time SS 1 $168,000 2011 $168,000 STRM.REFINE.101 76,712 kg/hr 0.70 1.8 111,930 1.46 $218,860 $218,860 $393,948

W- 330 Ion Exchange Strong acid cation/weak base anion 1 $5,250,000 2014 $5,250,000 STRM.REFINE.101 53,204 kg/hr 0.90 1.8 55,664 1.05 $5,467,930 $5,569,681 $10,025,426
T- 331 Deionized Sugar Storage Tank 8500 gal - 20 min residence time SS 1 $168,000 2011 $168,000 STRM.REFINE.101 76,712 kg/hr 0.70 1.8 111,930 1.46 $218,860 $218,860 $393,948
T- 303 Concentrated Sugar Storage 5,500 gallons - 20 min residence time 20 hp SS 1 $168,000 2011 $168,000 scis.a200.solids 76712 kg/hr 0.70 1.8 15801 0.21 $55,588 $55,588 $100,059
T- 305 Filtered Hydrolysate Storage 30,000 gallons - 20 min residence 20 hp SS 1 $51,600 2011 $51,600 STRM.a200.S205.S205LI 348690 kg/hr 0.70 1.8 360732 1.03 $52,841 $52,841 $95,114
Area 300 Totals $51,476,284 $52,734,070 $68,820,879

Size Ratio
EQPT NO

New Val
Year of

Scaling

Scaling

Factor
Quote
Purch Cost in Scaled Purch Purch Cost in Inst Cost in
REQD
NUM

Units

Inst
Exp
Val
EQUIPMENT TITLE DESCRIPTION HP MATERIAL $ Scaling Variable
Base Yr Cost Proj year Proj year

Mechanical Equipment List Scaled Installed Costs

A- 400 Cellulase Fermentor Agitators 800.0 SS316 $580,000 2009 $580,000 CLVESSEL 1 ea 1.00 1.5 5 5.00 $2,900,000 $3,254,512 $4,881,769
A- 401 Cellulase Fermentor Agitators 0.75hp SS316 $3,420 2009 $3,420 ICLSEED 1 ea 1.00 1.5 4 4.00 $13,680 $15,352 $23,028
A- 402 Cellulase Fermentor Agitators 8 hp SS316 $63,000 2009 $63,000 ICLSEED 1 ea 1.00 1.5 4 4.00 $252,000 $282,806 $424,209
A- 403 Cellulase Fermentor Agitators 80 hp SS316 $11,000 2009 $11,000 ICLSEED 1 ea 1.00 1.5 4 4.00 $44,000 $49,379 $74,068
A- 406 Cellulase Nutrient Mix Tank 3 hp CS 1 $4,800 2009 $4,800 strm.a400.416 174 kg/hr 0.50 1.6 112 0.65 $3,857 $4,329 $6,926
A- 410 Cellulase Hold Tank Agitator 10 hp SS316 1 $26,900 2009 $26,900 strm.422 10930 kg/hr 0.50 1.5 6941 0.64 $21,437 $24,058 $36,087
F- 400 Cellulase Fermentor 80,000 gal, 1 atm, 28 C, internal coil SS316 $400,500 2009 $400,500 CLVESSEL 1 ea 1.00 2.0 5 5.00 $2,002,500 $2,247,297 $4,494,594
F- 401 1st Cellulase Seed Fermentor 80 gallon skid complete - $46,000 ea 304SS $46,000 2009 $46,000 ICLSEED 1 ea 1.00 1.8 4 4.00 $184,000 $206,493 $371,688
F- 402 2nd Cellulase Seed Fermentor 800 gallon skid complete - $57,500 ea 304SS $57,500 2009 $57,500 ICLSEED 1 ea 1.00 1.8 4 4.00 $230,000 $258,116 $464,610
F- 403 3rd Cellulase Seed Fermentor 8,000 gallon skid complete - $95,400 304SS $95,400 2009 $95,400 ICLSEED 1 ea 1.00 1.8 4 4.00 $381,600 $428,249 $770,848
H- 400 Cellulase Fermentation Cooler Cooling coil included with cellulase 304SS INCLUDED
Cooling coil included with media prep
H- 405 Media Prep Tank Cooler 304SS 1 INCLUDED
tank
M- 401 Fermenter Air Compressor 8000 SCFM @ 16 psig CS 2 $350,000 2009 $350,000 strm.a400.450 33168 kg/hr 0.60 1.6 16346 0.49 $228,920 $256,904 $411,047
P- 400 Cellulase Transfer Pump 59 gpm, 100 FT, TDH SIZE 2X1-10C 3 316SS 1 $7,357 2010 $7,357 strm.a400.420 13399 kg/hr 0.80 2.3 6941 0.52 $4,347 $4,623 $10,632
P- 401 Cellulase Seed Pump 3 gpm, 100 FT TDH SIZE 2X1-10 2 316SS 4 $29,972 2010 $29,972 strm.a400.409 681 kg/hr 0.80 2.3 386 0.57 $19,030 $20,236 $46,543
P- 406 Cellulase Nutrient Transfer Gear Pump 2 gpm, 100 FT 1 316SS 1 $1,500 2009 $1,500 strm.a400.416 454 kg/hr 0.80 2.3 112 0.25 $491 $551 $1,267
P- 410 Cellulase Feed Pump Gear Pump 1 316SS 1 $1,500 2009 $5,700 strm.a400.422 18168 kg/hr 0.80 2.3 6941 0.38 $2,640 $2,963 $6,814
P- 420 Anti-foam Pump Gear Pump 2 gpm, 100 FT 1 316SS 1 $1,500 2009 $1,500 strm.a400.444 11 kg/hr 0.80 2.3 6.7 0.63 $1,039 $1,166 $2,683
T- 406 Cellulase Nutrient Mix Tank HDPE, 8,000 gal HDPE 1 $9,000 2010 $9,000 strm.a400.416 224 kg/hr 0.70 3.0 112 0.50 $5,553 $5,905 $17,714
T- 410 Cellulase Hold Tank 80,000 gal 304SS 1 $248,070 2009 $248,070 strm.a400.422 10930 kg/hr 0.70 1.8 6941 0.64 $180,532 $202,602 $364,683
Area 400 Totals $6,475,627 $7,265,541 $12,409,209

99
 NUM REQD

Scaling Exp
Scaling Val

Inst Factor

Size Ratio
EQPT NO

New Val
Year of
Quote
Purch Cost in Scaled Purch Purch Cost in Inst Cost in

Units
EQUIPMENT TITLE DESCRIPTION HP MATERIAL $ Scaling Variable
Base Yr Cost Proj year Proj year

Mechanical Equipment List Scaled Installed Costs

R- 510 APR-1 Reactor System Note 1 317L Clad $7,960,000 2014 $7,960,000 STRM.REFINE.101c 53,204 kg/hr 0.37 2.2 48,981 0.92 $7,720,111 $7,863,773 $18,338,318
R- 520 APR-2 Reactor System Note 1 317L Clad $2,044,000 2014 $2,044,000 STRM.REFINE.101c 53,204 kg/hr 0.37 2.2 48,981 0.92 $1,982,400 $2,019,290 $4,708,985
R- 530 Condensation Reactor System Note 1 317L Clad $2,645,000 2014 $2,645,000 STRM.REFINE.101c 53,204 kg/hr 0.37 2.2 48,981 0.92 $2,565,288 $2,613,025 $6,093,574
Conversion FE HX Shell and tube 317 SS 1 $948,600 2011 $948,600 heat.REFINE.17 79.2 Gcal/hr 0.70 2.2 74 0.94 $906,592 $906,592 $1,994,503
Condensation Feed Heater Plate and shell Duplex SS 1 $908,250 2014 $908,250 heat.REFINE.QHB20 17 Gcal/hr 0.70 2.2 15 0.85 $811,106 $826,199 $1,817,639
Condensation Effluent Economizer Plate and shell Duplex SS 1 $1,668,850 2014 $1,668,850 heat.REFINE.DHOGEFC 4 Gcal/hr 0.70 2.2 4 0.96 $1,625,343 $1,655,589 $3,642,295
Condensation Effluent Air Cooler Air cooler 317 SS 1 $1,028,500 2011 $1,028,500 heat.REFINE.QACCATCL 45 Gcal/hr 0.70 2.8 42 0.94 $982,809 $982,809 $2,751,865
Condensation Effluent Water
Cooler
Shell and tube 317 SS 1 $1,405,000 2011 $1,405,000 heat.REFINE.25 13 Gcal/hr 0.70 2.2 12 0.93 $1,339,247 $1,339,247 $2,946,342

MU Compressor Reciprocating compressor 3162 hp 1 $2,930,000 2014 $2,930,000 work.REFINE.10 2,359 KW 0.87 1.9 2620 1.11 $3,210,187 $3,269,925 $6,212,857
Recycle Compressor Centrifugal compressor 6300 hp 1 $2,900,000 2014 $2,900,000 work.REFINE.16 4,700 KW 0.78 1.4 5060 1.08 $3,071,665 $3,128,825 $4,380,355
S- 570 Hydrotreating Facility 610 hp 1 $11,500,000 2013 $ 11,500,000 strm.REFINE.HRT.OUT- 13528 kg/hr 0.50 2.0 22874 1.69 $14,953,898 $14,982,036 $29,964,072
REFINE Totals $39,168,648 $39,587,309 $82,850,805

NUM REQD

Scaling Exp
Scaling Val

Inst Factor

Size Ratio
EQPT NO

New Val
Year of
Quote
Purch Cost in Scaled Purch Purch Cost in Inst Cost in

Units
EQUIPMENT TITLE DESCRIPTION HP MATERIAL $ Scaling Variable
Base Yr Cost Proj year Proj year

Mechanical Equipment List Scaled Installed Costs

T- 606 Anaerobic System (4) 27 MG reactors, 8 recycle pumps, 795 hp Concrete 4 $25,800,000 2012 $25,800,000 COD 27211 kg/hr 0.60 1.1 19955 0.73 $21,419,106 $21,459,408 $23,819,943
Biogas Emergency Flare 4 INCLUDED 2012 INCLUDED
T- 608 Aeration Basin Concrete and steel, not installed cost Concrete 3 $4,908,054 2012 $4,908,054 Hydraulic flow 2.7 MGD 0.60 2.1 1.3 0.47 $3,108,842 $3,114,692 $6,447,412
Pump - Submersible,
2500 gpm ea 15 hp ea CS 2 INCLUDED 2012 INCLUDED
Anaerobic Feed
Pump - Centrifugal, Aeration 852 gpm ea 45 hp CS 4 INCLUDED 2012 INCLUDED
Aeration Grid Full floor aeration grid CS 1 INCLUDED 2012 INCLUDED
Caustic Feed System 1.5 hp CS 4 $20,000 2012 $20,000 COD 27211 kg/hr 0.60 3.0 19955 0.73 $16,604 $16,635 $49,906
Blowers 15000 SCFM @ 10.3 psig ea 1000 hp CS 9 $2,070,000 2012 $2,070,000 COD 27211 kg/hr 0.60 2.0 19955 0.73 $1,718,510 $1,721,743 $3,443,486
Surface Aerators 100 hp CS 3 $150,000 2012 $150,000 Hydraulic flow 2.7 MGD 0.60 2.7 1.3 0.47 $95,012 $95,191 $254,161
R- 609 Membrane Bioreactor Includes membrane, CIP, scour 85 hp ea CS 1 $4,898,500 2012 $4,898,500 Hydraulic flow 2.7 MGD 1.00 1.6 1.3 0.47 $2,288,471 $2,292,777 $3,760,154
Pump, Centrifugal , MBR, RAS 160 hp CS 6 INCLUDED 2012 INCLUDED
S- 612 Gravity Belt Thickeners 2m presses 48hp CS 3 $750,000 2012 $750,000 COD 27211 kg/hr 0.60 1.6 19955 0.73 $622,648 $623,820 $991,874
S- 611 Centrifuge 165 hp CS 1 $686,800 2012 $686,800 COD 27211 kg/hr 0.60 2.7 19955 0.73 $570,180 $571,253 $1,536,670
Pump, Centrifugal, Centrifuge 105 gpm 15hp CS 2 INCLUDED 2012 INCLUDED
Pump, Submersible, Centrate 100 gpm 10 hp ea CS 2 INCLUDED 2012 INCLUDED
Dewatering Polymer Addition 9.8 gph neat polymer 1 hp ea CS 2 INCLUDED 2012 INCLUDED
Conveyor 10 hp ea CS 1 $7,000 2012 $7,000 COD 27211 kg/hr 0.60 2.9 19955 0.73 $5,811 $5,822 $16,652
S- 610 Reverse Osmosis CS 7 $2,450,000 2012 $2,450,000 Hydraulic flow 2.7 MGD 1.00 1.8 1.3 0.47 $1,144,586 $1,146,739 $2,006,794
M- 640 Evaporator 368 gpm 1480 hp Titanium 1 $5,000,000 2012 $5,000,000 Hydraulic flow 2.7 MGD 0.60 1.6 1.3 0.47 $3,167,082 $3,173,041 $5,108,597
Ammonia Addition System 0.63 gpm 4.5 hp CS 4 $195,200 2012 $195,200 COD 27211 kg/hr 0.60 1.5 19955 0.73 $162,055 $162,360 $250,034
Area 600 Totals $34,318,906 $34,383,482 $47,685,681

100
 NUM REQD

Scaling Exp
Scaling Val

Inst Factor

Size Ratio
EQPT NO

New Val
Year of
Quote
Purch Cost in Scaled Purch Purch Cost in Inst Cost in

Units
EQUIPMENT TITLE DESCRIPTION HP MATERIAL $ Scaling Variable
Base Yr Cost Proj year Proj year

Mechanical Equipment List Scaled Installed Costs

A- 720 CSL Storage Tank Agitator 10 hp SS304 1 $21,200 2009 $21,200 strm.735 1393 kg/hr 0.50 1.5 83 0.06 $5,160 $5,791 $8,686
A- 760 DAP Makeup Tank Agitator 5.5 hp SS304 1 $9,800 2009 $9,800 strm.755 163 kg/hr 0.50 1.5 0 0.00 $0 $0 $0
C- 755 DAP Bulk Bag Unloader Super sack unloader 1 $30,000 2009 $30,000 strm.755 163 kg/hr 0.60 1.7 0 0.00 $0 $0 $0
P- 701 Product Pump 150 gpm, 112 FT TDH 5.0 CS 2 $9,200 2009 $9,200 strm.703 22681 kg/hr 0.80 3.1 16226 0.72 $7,038 $7,898 $24,484
P- 703 Sulfuric Acid Pump 5 gpm, 150 FT TDH SIZE 2X1-10 0.5 SS316 1 $7,493 2010 $7,493 strm.710 1981 kg/hr 0.80 2.3 2240 1.13 $8,268 $8,792 $20,221
P- 704 Firewater Pump 2500 gpm, 150 FT TDH 125.0 CS 1 $15,000 2009 $15,000 strm.a700.713 8343 kg/hr 0.80 3.1 5968 0.72 $11,474 $12,877 $39,917
P- 720 CSL Pump 8 gpm, 80 FT TDH 0.5 CS 1 $3,000 2009 $3,000 strm.735 1393 kg/hr 0.80 3.1 83 0.06 $313 $351 $1,088
P- 760 DAP Pump 2 gpm, 100 FT TDH 0.5 CS 1 $3,000 2009 $3,000 strm.755 163 kg/hr 0.80 3.1 0 0.00 $0 $0 $0
T- 701 Product Storage Tank 1.2 M gal., 7 day storage, floating roof A285C 1 $1,146,219 2009 $1,146,219 strm.703 18145 kg/hr 0.70 1.7 16226 0.89 $1,059,967 $1,189,543 $2,022,223
300,000 gal., 7 day storage, floating
T- 702 Product Storage Tank A285C 1 $352,790 2009 $352,790 strm.REFINE.C8-MIX 4536 kg/hr 0.70 1.7 3821 0.84 $312,840 $351,083 $596,841
roof
T- 703 Sulfuric Acid Storage Tank 12,600 gal, 12' dia x15' H SS 1 $96,000 2010 $96,000 strm.710 1981 kg/hr 0.70 1.5 2240 1.13 $104,630 $111,260 $166,890
Caustic Storage Tank 12,600 gal, 12' dia x15' H SS 1 $96,000 2011 $96,000 strm.a200.2 1981 kg/hr 0.70 1.5 2046 1.03 $98,187 $98,187 $147,281
Caustic Storage Tank 12,600 gal, 12' dia x15' H SS 1 $96,000 2011 $96,000 strm.a200.2 1981 kg/hr 0.70 1.5 1122 0.57 $64,478 $64,478 $96,717
T- 704 Firewater Storage Tank 600,000 gal - 4 hrs @ 2500 gpm Glass lined 1 $803,000 2009 $803,000 strm.a700.713 8343 kg/hr 0.70 1.7 5968 0.72 $635,160 $712,805 $1,211,769
T- 706 Ammonia Storage Tank 28,000 gal SA- 516-70 2 $196,000 2010 $196,000 strm.717 1171 kg/hr 0.70 2.0 368 0.31 $87,090 $92,608 $185,217
T- 720 CSL Storage Tank 70,000 gal Glass lined 1 $70,000 2009 $70,000 strm.735 1393 kg/hr 0.70 2.6 83 0.06 $9,681 $10,865 $28,248
Glucose Storage Tank 70,000 gal Glass lined 1 $70,000 2009 $70,000 strm.a400.401 1393 kg/hr 0.70 2.6 1427 1.02 $71,195 $79,898 $207,736
T- 755 DAP Bulk Bag Holder Super sack holder 1 INCLUDED
T- 760 DAP Make-up Tank 12,800 gal SS304 1 $102,000 2009 $102,000 strm.717 1615 kg/hr 0.70 1.8 368 0.23 $36,189 $40,613 $73,104
Area 700 Totals $2,511,670 $2,787,049 $4,830,422

101
 NUM REQD

Scaling Exp
Scaling Val

Inst Factor

Size Ratio
EQPT NO

New Val
Year of
Quote
Purch Cost in Scaled Purch Purch Cost in Inst Cost in

Units
EQUIPMENT TITLE DESCRIPTION HP MATERIAL $ Scaling Variable
Base Yr Cost Proj year Proj year

Mechanical Equipment List Scaled Installed Costs

H- 801 Burner Combustion Air INCLUDED 1 INCLUDED
H- 811 BFW Preheater INCLUDED 1 INCLUDED
Pretreatment/BFW Heat
H- 812 9.4 MM BTU/hr SS304 1 $41,000 2009 $41,000 heat.QH812 -2 Gcal/hr 0.70 2.2 -2 0.86 $36,776 $41,272 $90,797
Recovery
M- 802 Air Intake Fan INCLUDED INCLUDED
M- 803 Boiler 525,000 lb/hr @ 900 psig CS 1 $28,550,000 2010 $28,550,000 strm.a800.a810.813C 238203 kg/hr 0.60 1.8 246134 1.03 $29,116,607 $30,961,505 $55,730,709
Baghouse, spray dryer scrubber,
M- 804 Combustion Gas Baghouse 1 $11,000,000 2013 $11,000,000 strm.a800.a810.813C 238203 kg/hr 0.60 1.8 0 0.00 $0 $0 $0
flues/ducting
M- 811 Turbine/Generator 23.6 KW, 2 extractions 1 $9,500,000 2010 $9,500,000 work.wtotal -42200 KW 0.60 1.8 -46446 1.10 $10,062,505 $10,700,089 $19,260,161
M- 820 Hot Process Water Softener 1 $78,000 2010 $78,000 strm.a800.a810.813C 235803 kg/hr 0.60 1.8 246134 1.04 $80,033 $85,104 $153,187
M- 830 Amine Addition Pkg. 1 $40,000 2010 $40,000 strm.a800.a810.813C 235803 kg/hr 0.00 1.8 246134 1.04 $40,000 $42,534 $76,562
M- 832 Ammonia Addition Pkg. 1 INCLUDED
M- 834 Phosphate Addition Pkg. 1 INCLUDED
P- 804 Condensate Pump SS316 2 INCLUDED
P- 811 Turbine Condensate Pump SS304 2 INCLUDED
P- 824 Deaerator Feed Pump SS304 2 INCLUDED
P- 826 BFW Pump SS316 5 INCLUDED
P- 828 Blowdown Pump CS 2 INCLUDED
P- 830 Amine Transfer Pump CS 1 INCLUDED
T- 804 Condensate Collection Tank A285C 1 INCLUDED
T- 824 Condensate Surge Drum SS304 1 INCLUDED
T- 826 Deaerator Tray type CS;SS316 1 $305,000 2010 $305,000 strm.a800.a810.813C 235803 kg/hr 0.60 3.0 246134 1.04 $312,949 $332,778 $998,334
T- 828 Blowdown Flash Drum CS 1 INCLUDED
T- 830 Amine Drum SS316 1 INCLUDED
NUM REQD Area 800 Totals $39,648,870 $42,163,282 $76,309,750

Scaling Exp
Scaling Val

Inst Factor

Size Ratio
EQPT NO

New Val
Year of
Quote
Purch Cost in Scaled Purch Purch Cost in Inst Cost in

Units
EQUIPMENT TITLE DESCRIPTION HP MATERIAL $ Scaling Variable
Base Yr Cost Proj year Proj year

Mechanical Equipment List Scaled Installed Costs

M- 902 Cooling Tower System 44,200 gpm 750 hp FIBERGLASS 1 $1,375,000 2010 $1,375,000 strm.a900.945 10037820 kg/hr 0.60 1.5 13741966 1.37 $1,660,154 $1,765,345 $2,648,018
M- 904 Plant Air Compressor 400 SCFM@125 psig 150 hp 1 $28,000 2010 $28,000 DRY101 83333 kg/hr 0.60 1.6 83333 1.00 $28,000 $29,774 $47,639
M- 908 Chilled Water Package 2 x 2350 tons (14.2 MM kcal/hr) 3400 hp 1 $1,275,750 2010 $1,275,750 heat.a900.qchwop 14 Gcal/hr 0.60 1.6 4 0.28 $589,025 $626,347 $1,002,155
M- 910 CIP System 100,000 GAL SS304/SS31 1 $421,000 2009 $421,000 strm.a900.914 63 kg/hr 0.60 1.8 145 2.30 $694,222 $779,088 $1,402,358
P- 902 Cooling Water Pump 16,120 gpm, 100 FT TDH SIZE 20X20-28 500.0 CS 3 $283,671 2010 $283,671 strm.a900.945 10982556 kg/hr 0.80 3.1 13741966 1.25 $339,384 $360,888 $1,118,753
P- 912 Makeup Water Pump 685 gpm, 75 FT TDH SIZE 6X4-13 20.0 CS 1 $6,864 2010 $6,864 strm.a900.904 155564 kg/hr 0.80 3.1 341548 2.20 $12,877 $13,693 $42,448
P- 914 Process Water Circulating 2285 gpm, 75 FT TDH SIZE 8X6-13 75.0 CS 1 $15,292 2010 $15,292 strm.a900.905 518924 kg/hr 0.80 3.1 590356 1.14 $16,954 $18,028 $55,888
S- 904 Instrument Air Dryer 670 SCFM - cycling type CS 1 $15,000 2009 $15,000 DRY101 83333 kg/hr 0.60 1.8 83333 1.00 $15,000 $16,834 $30,301
T- 904 Plant Air Receiver 3800 gal - 72" x 228" vertical CS 1 $16,000 2009 $16,000 DRY101 83333 kg/hr 0.60 3.1 83333 1.00 $16,000 $17,956 $55,664
T- 914 Process Water Tank No. 1 250,000 gal CS 1 $250,000 2009 $250,000 strm.a900.905 451555 kg/hr 0.70 1.7 590356 1.31 $301,594 $338,462 $575,386
Area 900 Totals $3,673,210 $3,966,415 $6,978,608

102
Appendix B. Discounted Cash Flow Rate of Return Worksheet
Year -2 -1 0 1 2 3 4 5 6
Fixed Capital Investment $20,047,207 $150,354,051 $80,188,827
Land $1,848,000
Working Capital $31,323,761
Loan Payment $56,017,968 $56,017,968 $56,017,968 $56,017,968 $56,017,968 $56,017,968
Loan Interest Payment $2,405,665 $20,448,151 $30,070,810 $30,070,810 $27,995,037 $25,753,203 $23,332,022 $20,717,146 $17,893,080
Loan Principal $30,070,810 $255,601,886 $375,885,127 $349,937,969 $321,915,038 $291,650,273 $258,964,326 $223,663,504 $185,538,616
Fuel Sales $172,159,887 $229,546,516 $229,546,516 $229,546,516 $229,546,516 $229,546,516
Byproduct Credit $4,024,206 $5,365,608 $5,365,608 $5,365,608 $5,365,608 $5,365,608
Total Annual Sales $176,184,093 $234,912,124 $234,912,124 $234,912,124 $234,912,124 $234,912,124
Annual Manufacturing Cost
Feedstock $43,460,550 $57,947,400 $57,947,400 $57,947,400 $57,947,400 $57,947,400
Baghouse Bags $519,687 $519,687
Initial Catalyst Charge $7,466,161
Other Variable Costs $61,482,169 $70,265,336 $70,265,336 $70,265,336 $70,265,336 $70,265,336
Fixed Operating Costs $16,080,969 $16,080,969 $16,080,969 $16,080,969 $16,080,969 $16,080,969
Total Product Cost $129,009,536 $144,293,705 $144,293,705 $144,293,705 $144,293,705 $144,813,392
Annual Depreciation
General Plant Writedown 14% 24.49% 17.49% 12.49% 8.93% 8.92%
Depreciation Charge $78,618,644 $134,735,521 $96,223,939 $68,715,666 $49,129,776 $49,074,759
Remaining Value $471,546,817 $336,811,295 $240,587,356 $171,871,690 $122,741,914 $73,667,155
Steam Plant Writedown 3.75% 7.22% 6.68% 6.18% 5.71% 5.29%
Depreciation Charge $2,861,616 $5,508,801 $5,095,202 $4,713,653 $4,359,576 $4,032,970
Remaining Value $73,448,134 $67,939,333 $62,844,131 $58,130,478 $53,770,902 $49,737,932
Net Revenue ($64,376,514) ($77,620,941) ($36,453,926) ($6,142,923) $16,411,921 $19,097,922
Losses Forward ($64,376,514) ($141,997,455) ($178,451,381) ($184,594,304) ($168,182,383)
Taxable Income ($64,376,514) ($141,997,455) ($178,451,381) ($184,594,304) ($168,182,383) ($149,084,461)
Income Tax $0 $0 $0 $0 $0 $0
Annual Cash Income -$8,843,412 $34,600,450 $34,600,450 $34,600,450 $34,600,450 $34,080,764
Discount Factor 1.2100 1.1000 1.0000 0.9091 0.8264 0.7513 0.6830 0.6209 0.5645
Annual Present Value $356,968,848 -$8,039,465 $28,595,413 $25,995,830 $23,632,573 $21,484,157 $19,237,703
Total Capital Investment + Interest $29,404,055 $187,882,422 $141,583,398
Net Present Worth $0

Year 7 8 9 10 11 12 13 14 15
Fixed Capital Investment
Land
Working Capital
Loan Payment $56,017,968 $56,017,968 $56,017,968 $56,017,968 $0 $0 $0 $0 $0
Loan Interest Payment $14,843,089 $11,549,099 $7,991,589 $4,149,479 $0 $0 $0 $0 $0
Loan Principal $144,363,737 $99,894,868 $51,868,489 $0 $0 $0 $0 $0 $0
Fuel Sales $229,546,516 $229,546,516 $229,546,516 $229,546,516 $229,546,516 $229,546,516 $229,546,516 $229,546,516 $229,546,516
By-Product Credit $5,365,608 $5,365,608 $5,365,608 $5,365,608 $5,365,608 $5,365,608 $5,365,608 $5,365,608 $5,365,608
Total Annual Sales $234,912,124 $234,912,124 $234,912,124 $234,912,124 $234,912,124 $234,912,124 $234,912,124 $234,912,124 $234,912,124
Annual Manufacturing Cost
Feedstock $57,947,400 $57,947,400 $57,947,400 $57,947,400 $57,947,400 $57,947,400 $57,947,400 $57,947,400 $57,947,400
$519,687
Baghouse Bags
Other Variable Costs $70,265,336 $70,265,336 $70,265,336 $70,265,336 $70,265,336 $70,265,336 $70,265,336 $70,265,336 $70,265,336
Fixed Operating Costs $16,080,969 $16,080,969 $16,080,969 $16,080,969 $16,080,969 $16,080,969 $16,080,969 $16,080,969 $16,080,969
Total Product Cost $144,293,705 $144,293,705 $144,293,705 $144,293,705 $144,813,392 $144,293,705 $144,293,705 $144,293,705 $144,293,705
Annual Depreciation
General Plant Writedown 8.93% 4.46%
Depreciation Charge $49,129,776 $24,537,380
Remaining Value $24,537,380 $0
Steam Plant Writedown 4.89% 4.52% 4.46% 4.46% 4.46% 4.46% 4.46% 4.46% 4.46%
Depreciation Charge $3,730,021 $3,450,727 $3,404,941 $3,404,178 $3,404,941 $3,404,178 $3,404,941 $3,404,178 $3,404,941
Remaining Value $46,007,911 $42,557,184 $39,152,243 $35,748,065 $32,343,124 $28,938,946 $25,534,005 $22,129,827 $18,724,886
Net Revenue $22,915,533 $51,081,213 $79,221,888 $83,064,761 $86,693,791 $87,214,240 $87,213,477 $87,214,240 $87,213,477
Losses Forward ($149,084,461) ($126,168,928) ($75,087,715) $0 $0 $0 $0 $0 $0
Taxable Income ($126,168,928) ($75,087,715) $4,134,173 $83,064,761 $86,693,791 $87,214,240 $87,213,477 $87,214,240 $87,213,477
Income Tax $0 $0 $1,446,960 $29,072,666 $30,342,827 $30,524,984 $30,524,717 $30,524,984 $30,524,717
Annual Cash Income $34,600,450 $34,600,450 $33,153,490 $5,527,784 $59,755,905 $60,093,434 $60,093,701 $60,093,434 $60,093,701
Discount Factor 0.5132 0.4665 0.4241 0.3855 0.3505 0.3186 0.2897 0.2633 0.2394
Annual Present Value $17,755,502 $16,141,365 $14,060,316 $2,131,200 $20,944,080 $19,147,620 $17,407,005 $15,824,479 $14,385,954
Total Capital Investment + Interest
Net Present Worth

103
Year 16 17 18 19 20 21 22 23 24
Fixed Capital Investment
Land
Working Capital
Loan Payment $0 $0 $0 $0 $0 $0 $0 $0 $0
Loan Interest Payment $0 $0 $0 $0 $0 $0 $0 $0 $0
Loan Principal $0 $0 $0 $0 $0 $0 $0 $0 $0
Fuel Sales $229,546,516 $229,546,516 $229,546,516 $229,546,516 $229,546,516 $229,546,516 $229,546,516 $229,546,516 $229,546,516
By-Product Credit $5,365,608 $5,365,608 $5,365,608 $5,365,608 $5,365,608 $5,365,608 $5,365,608 $5,365,608 $5,365,608
Total Annual Sales $234,912,124 $234,912,124 $234,912,124 $234,912,124 $234,912,124 $234,912,124 $234,912,124 $234,912,124 $234,912,124
Annual Manufacturing Cost
Feedstock $57,947,400 $57,947,400 $57,947,400 $57,947,400 $57,947,400 $57,947,400 $57,947,400 $57,947,400 $57,947,400
$519,687 $519,687
Baghouse Bags
Other Variable Costs $70,265,336 $70,265,336 $70,265,336 $70,265,336 $70,265,336 $70,265,336 $70,265,336 $70,265,336 $70,265,336
Fixed Operating Costs $16,080,969 $16,080,969 $16,080,969 $16,080,969 $16,080,969 $16,080,969 $16,080,969 $16,080,969 $16,080,969
Total Product Cost $144,813,392 $144,293,705 $144,293,705 $144,293,705 $144,293,705 $144,813,392 $144,293,705 $144,293,705 $144,293,705
Annual Depreciation
General Plant Writedown
Depreciation Charge
Remaining Value
Steam Plant Writedown 4.46% 4.46% 4.46% 4.46% 4.46% 2.23%
Depreciation Charge $3,404,178 $3,404,941 $3,404,178 $3,404,941 $3,404,178 $1,702,471
Remaining Value $15,320,708 $11,915,767 $8,511,590 $5,106,648 $1,702,471 $0
Net Revenue $86,694,554 $87,213,477 $87,214,240 $87,213,477 $87,214,240 $88,396,261 $90,618,418 $90,618,418 $90,618,418
Losses Forward $0 $0 $0 $0 $0 $0 $0 $0 $0
Taxable Income $86,694,554 $87,213,477 $87,214,240 $87,213,477 $87,214,240 $88,396,261 $90,618,418 $90,618,418 $90,618,418
Income Tax $30,343,094 $30,524,717 $30,524,984 $30,524,717 $30,524,984 $30,938,691 $31,716,446 $31,716,446 $31,716,446
Annual Cash Income $59,755,638 $60,093,701 $60,093,434 $60,093,701 $60,093,434 $59,160,040 $58,901,972 $58,901,972 $58,901,972
Discount Factor 0.2176 0.1978 0.1799 0.1635 0.1486 0.1351 0.1228 0.1117 0.1015
Annual Present Value $13,004,568 $11,889,218 $10,808,332 $9,825,800 $8,932,506 $7,994,330 $7,235,870 $6,578,064 $5,980,058
Total Capital Investment + Interest
Net Present Worth

Year 25 26 27 28 29 30
Fixed Capital Investment
Land ($1,848,000)
Working Capital ($31,323,761)
Loan Payment $0 $0 $0 $0 $0 $0
Loan Interest Payment $0 $0 $0 $0 $0 $0
Loan Principal $0 $0 $0 $0 $0 $0
Fuel Sales $229,546,516 $229,546,516 $229,546,516 $229,546,516 $229,546,516 $229,546,516
By-Product Credit $5,365,608 $5,365,608 $5,365,608 $5,365,608 $5,365,608 $5,365,608
Total Annual Sales $234,912,124 $234,912,124 $234,912,124 $234,912,124 $234,912,124 $234,912,124
Annual Manufacturing Cost
Feedstock $57,947,400 $57,947,400 $57,947,400 $57,947,400 $57,947,400 $57,947,400
$519,687
Baghouse Bags
Other Variable Costs $70,265,336 $70,265,336 $70,265,336 $70,265,336 $70,265,336 $70,265,336
Fixed Operating Costs $16,080,969 $16,080,969 $16,080,969 $16,080,969 $16,080,969 $16,080,969
Total Product Cost $144,293,705 $144,813,392 $144,293,705 $144,293,705 $144,293,705 $144,293,705
Annual Depreciation
General Plant Writedown
Depreciation Charge
Remaining Value
Steam Plant Writedown
Depreciation Charge
Remaining Value
Net Revenue $90,618,418 $90,098,732 $90,618,418 $90,618,418 $90,618,418 $90,618,418
Losses Forward $0 $0 $0 $0 $0 $0
Taxable Income $90,618,418 $90,098,732 $90,618,418 $90,618,418 $90,618,418 $90,618,418
Income Tax $31,716,446 $31,534,556 $31,716,446 $31,716,446 $31,716,446 $31,716,446
Annual Cash Income $58,901,972 $58,564,176 $58,901,972 $58,901,972 $58,901,972 $58,901,972
Discount Factor 0.0923 0.0839 0.0763 0.0693 0.0630 0.0573
Annual Present Value $5,436,416 $4,913,854 $4,492,906 $4,084,460 $3,713,145 $3,375,587
Total Capital Investment + Interest ($1,901,026)
Net Present Worth

104
Appendix C. Process Parameters/Operating Summary
Feedstock Conditioning Boiler/Turbogenerator
Moisture Content 20% Ammonia Loading (g/L hydrolyzate) 1.8 WWT Sludge Moisture % 60.0%
Pretreatment Sugar Losses: First Effect Lignin Moisture % 62.5%
Acid Conc (wt %) 0.43 Xylose 0.0% Combined Boiler Feed Moisture 61.9%
Acid Loading (mg/g dry biomass) 9.0 Arabinose 1.0% Boiler Feed Combustion Energy MMkcal/hr -182
Total Solids (wt %) 30.0% Glucose 0.0% % from Lignin Cake 36%
Insoluble Solids in (wt %) 26.0% Galactose 1.0% % from Biogas 44%
Insoluble Solids out (wt %) 18.7% Mannose 1.0% Boiler Feed Heating Value (kcal/kg) -445
Temperature (C) 158 Cellobiose 1.0% Boiler Feed Heating Value (BTU/lb) -800
Pressure (atm) 5.7 Enzyme Production
Conversions: Overall Glucose Feed (kg/hr) 1,213 Boiler Efficiency (feed/steam) 67%
Cellulose to Glucolig 0.3% Protein Produced (kg/hr) 290
Cellulose to Cellobiose 0.0% Protein Yield (kg/kg substrate) 0.24 Steam Loop Flow (kg/hr) 202,855
Cellulose to Glucose 9.9% Protein Harvest Titer (g/L) 40 Total Generator Output (KW) -46,446
Cellulose to HMF 0.0% to A100 Feedstock Handling 859
Xylan to Oligomer 2.4% Compressor Electricity (KW) 706 to A200 Pretreatment 7,975
Xylan to Xylose 90.0% Agitator Electricity (KW) 1050 to A300 Hydrolysis/Conditioning/Purifica 4,834
Xylan to Furfural 5.0% Chiller System Electricity (KW) 0 to A400 Enzyme Production 1,891
Xylan to Tar 0.0% A400 Total Electricity (KW) 1891 to A500 Conversion + Upgrading 8,574
Mannan to Oligomer 2.4% to A600 Wastewater 6,198
Mannan to Mannose 90.0% Electricity Demand (KWh/kg protn) 7 to A700 Storage 5
Mannan to HMF 5.0% Production Selectivity: to A800 Boiler/Turbogen 1,502
Galactan to Oligomer 2.4% Selectivity to Protein 31.3% to A900 Utilities 3,563
Galactan to Galactose 90.0% Selectivity to Cell Mass 4.1% Excess Electricity to Grid 11,045
Galactan to HMF 5.0% Selectivity to CO2 67.6% Electricity Used (KW) 35,401
Arabinan to Oligomer 2.4% Electricity Used (%) 76%
Arabinan to Arabinose 90.0% Catalytic Conversion and Upgrading Wastewater Treatment
Arabinan to Furfural 5.0% Total Feed Rate to APR 99,620 NH3-N in Combined WW mg/L 543
Arabinan to Tar 0.0% Hydrogen Feed Molar Ratio 9.80 Protein-N 0
Acetate to Oligomer 0.0% Cellulase-N 3
Acetate to Acetic Acid 100.0% Hydrogenation Temperature (C) 253 Hydraulic Load (L/hr) 198,951
Furfural to Tar 100.0% Hydrogenation Pressure (atm) 72 Hydraulic Load (gpm) 876
HMF to Tar 100.0% Hydrogenation WHSV (hr-1) 1.2 Hydraulic Load (MMgal/day) 1.3
Lignin to Soluble Lignin 5.0% Total COD (kg/hr) 19,955
Xylose Olig to Xylose APR Temperature (C) 300 Total COD (g/L) 100
Mannose Olig to Mannose APR Pressure (atm) 62 Total Insoluble Solids (g/L) 17
Arabinose Olig to Arabinose APR WHSV (hr-1) 1.0 Total Solids (g/L) 120
Galactose Olig to Galactose Biogas Composition (dry molar%) CH4 51%
Enzymatic Hydrolysis Hydrotreating Temperature (C) 367 CO2 49%
Enzyme Loading (mg/g cell) 10.0 Hydrotreating Pressure (atm) 37 Methane Produced (kg/kg COD) 0.281
-1
Total Solids (wt %) 20.0% Hydrotreating WHSV (hr ) 1.2 Cell Mass 0.058
Insoluble Solids (wt %) 12.3%
Temperature (C) 0 Conversions:
Pressure (atm) 0.0 C Yield to RDB vs APR Feed Components 69.4%
Residence Time (days) 3.5 C Yield to Naphtha vs APR Feed Component 16.2%
Conversions: Overall Overall C Yield 85.6%
Cellulose to Glucolig 4.0% Total Hydrogen Consumptions wt % 6.5%
Cellulose to Cellobiose 1.2%
Cellulose to Glucose 90.0%
Glucolig to Cellobiose 0.0%
Glucolig to Glucose 0.0%
Cellobiose to Glucose 0.0%

105
Appendix D. Aspen Properties
The table below is a list of the components used in the Aspen model. Previous versions of the
model used custom property databanks created at NREL. Where possible in the new model, these
components have been replaced with components from Aspens native databanks. Property
definitions for the few remaining custom components were moved into the model itself (i.e.,
inside the simulation file) and are discussed here.

The component CSL (corn steep liquor) used in previous models was removed. CSL streams are
now modeled as 50% water, 25% protein, and 25% lactic acid [110].

106
Component Property Quantity Units Reference
H2O - - - Native Aspen component
GLUCOSE - - - Native Aspen component (dextrose)
GALACTOS - - - Duplicate of GLUCOSE
MANNOSE - - - Duplicate of GLUCOSE
XYLOSE DHFORM -216752.65 cal/mol Native Aspen component (d-xylose) with DHFORM specified (5/6 of GLUCOSE DHFORM)

ARABINOS - - - Duplicate of XYLOSE

CELLOB - - - Cellobiose. Used native Aspen component sucrose
SUCROSE - - - Native Aspen component
GLUCOLIG MW 162.1424 Glucose oligomers. Most properties from GLUCOSE; MW is GLUCOSE minus H2O
DHFORM -192875.34 cal/mol Back-calculated to match Hc of CELLULOS
GALAOLIG - - - Galactose oligomers. Duplicate of GLUCOLIG
MANOLIG - - - Mannose oligomers. Duplicate of GLUCOLIG
XYLOLIG MW 132.11612 Xylose oligomers. Most properties from XYLOSE; MW is XYLOSE minus H2O
DHFORM -149412.58 cal/mol Back-calculated to match Hc of XYLAN
ARABOLIG - - - Arabinose oligomers. Duplicate of XYLOLIG
EXTRACT - - - Organic extractives. Duplicate of GLUCOSE
LGNSOL - - - Solubilized lignin. Native Aspen component vanillin (see note at LIGNIN)

107
Component Property Quantity Units Reference
HMF MW 126.11 (5-hydroxymethylfurfural) Properties for HMF were estimated within Aspen using NIST
TB 532.7 K TDE routines. Specify molecular structure, MW, TB and DHFORM. From [90].
DHFORM -79774.53 cal/mol
DHVLWT- 80550000 J/kmol
1
TC 731.012 K NIST TDE
PC 5235810 Pa
OMEGA 0.99364671
VC 0.3425 m3/kmol
RKTZRA 0.198177974
FURFURAL - - - Native Aspen component
AACID - - - Native Aspen component (acetic acid)
LACID - - - Native Aspen component (lactic acid)
XYLITOL - - - Native Aspen component
GLYCEROL - - - Native Aspen component
SUCCACID - - - Native Aspen component (Succinic acid)
NH3 - - - Native Aspen component
H2SO4 - - - Native Aspen component
NH4SO4 - - - Native Aspen component (ammonium sulfate)
NH4ACET PLXANT/1 -1.00E+20 atm Native Aspen component (ammonium acetate) forced non-volatile
DAP - - - Native Aspen component (diammonium phosphate)
HNO3 - - - Native Aspen component
NANO3 - - - Native Aspen component
NAOH - - - Native Aspen component
CNUTR - - - Cellulase nutrient mix. Duplicate of glucose
WNUTR - - - WWT nutrient mix. Duplicate of glucose
DENAT - - - Denaturant. Native Aspen component (n-heptane)

108
Component Property Quantity Units Reference
OIL - - - Corn oil antifoam. Native Aspen component (oleic acid)
O2 - - - Native Aspen component
N2 - - - Native Aspen component
NO - - - Native Aspen component
NO2 - - - Native Aspen component
CO - - - Native Aspen component
CO2 - - - Native Aspen component
CH4 - - - Native Aspen component
H2S - - - Native Aspen component
SO2 - - - Native Aspen component
CELLULOS DHSFRM -233200.06 cal/mol Native Aspen component with specified heat of formation; back-calculated
GALACTAN - - - Duplicate of CELLULOS
MANNAN - - - Duplicate of CELLULOS
XYLAN Formula C5H8O4 (monomer)
MW 132.117
DHSFRM -182099.93 cal/mol From [22]. Assumes the ratio of Hc of glucose:xylose is the same for cellulose:xylan.
ARABINAN - - - Duplicate of xylan
LIGNIN - - - Used native Aspen component vanillin (C8H8O3). The HHV of this compound (-23,906
BTU/kg) is very close to what we previously assumed for lignin as a custom component (-
24,206)

ACETATE - - - Used native Aspen component acetic acid

PROTEIN Formula CH1.57O0.31N0.29S0.007 Wheat gliadin [91]
MW 22.8396
DHSFRM -17618 cal/mol From literature value of gliadin Hc [92]
PLXANT/1 -1.00E+20 atm Forces non-volatility

109
Component Property Quantity Units Reference
ASH - - - Native Aspen component CaO
ENZYME Formula CH1.59O0.42N0.24S0.01 Provided by Novozymes [55]
MW 24.0156
DHSFRM -17618 cal/mol copied from PROTEIN
DENZ - - - Denatured enzyme. Duplicate of enzyme
ZYMO Formula CH1.8O0.5N0.2 Z. mobilis cell mass. Average composition of several microorganisms outlined
in [93]
MW 24.6264
DHSFRM -31169.39 cal/mol From [22] after [94]
TRICHO Formula CH1.645O0.445N0.205S0.005 T. reesei cell mass. Average of generic cell mass [56] and the ENZYME composition
MW 23.8204
DHSFRM -23200.01 cal/mol Copied BIOMASS
BIOMASS Formula CH1.64O0.39N0.23S0.0035 WWT sludge.
MW 23.238
DHSFRM -23200.01 cal/mol From [22] after [94]
TAR - - - Modeled as solid xylose
LIME - - - Aspen native component (calcium hydroxide)
CASO4 - - - Aspen native component
C - - - Aspen native component (graphite carbon)
NAPHTHA - C4-C9 Group C4-C9 components
DIESEL - C10-C30+ Group C10-C30+ components
HYDROGEN Formula H2 Aspen native component

110
Appendix E. Process Flow Diagrams
High-level stream table information from Aspen Plus modeling output follows, for key streams
associated with each process operation area for the base case model (external purchased
hydrogen). This is followed by high-level PFDs for the associated process areas. Space for
stream tables was limited; below is a key to lumped components. As the stream table information
focuses primarily on the high-level overall process and does not include every individual
modeled stream within each process area, mass balance closure around a given unit area may not
be 100%.

Other sugars (SS) arabinose, mannose, galactose, cellobiose, sucrose

Sugar oligomers (SS) oligomers of glucose, xylose, and all minor sugars
Organic soluble solids (SS) ammonium acetate, solubilized lignin, organic extractives, lactic
acid, cellulase and WWT chemicals
Inorganic soluble solids (SS) ammonium sulfate, diammonium phosphate, sodium hydroxide,
nitric acid, sodium nitrate
Furfurals furfural, HMF
Other organics glycerol, corn oil, sorbitol, succinic acid, xylitol, other organic
components; including alcohols, ketones, aldehydes, furans,
diols, triols, cyclic mono-oxygenates, polyoxygenates, organic
acids, propane
CO/SOX/NOX/H2S NO, NO2, SO2, CO, H2S
Other struct. carbohydr. (IS) arabinan, mannan, galactan
Protein (IS) corn protein, cellulase, denatured cellulase
Cell mass (IS) WWT sludge, T. reesei from cellulase production
Other insoluble solids (IS) tar, ash, carbon, lime, gypsum

111
 AMMONIA 210 GLUCOSE 400
440 VENT
A400
STEAM 220
NUTRIENTS 410 ENZYME
PRODUCTION
AIR 420
ACID 230

CELLULASE
RECYCLE
340
PROCESS WATER

330
430
CAUSTIC 240 WATER

A300
A200 DILUTE HYDROLYSATE
RAW
100
A100 DEACETYLATION, 300
200 350
FEEDSTOCK
FEED HANDLING PRETREATMENT PRETREATED
ENZYMATIC CONCENTRATION,
FEEDSTOCK
& CONDITIONING SLURRY HYDROLYSIS, RECYCLE FROM FILTER FILTRATION, ION
LIGNIN FILTRATION 360 EXCHANGE
FILTER PRESS
310
AIR IN
PROCESS
250
WATER
PROCESS
320
WATER

CONCENTRATED
HYDROLYSATE

PURIFIED,
650

500
AIR 630 FILTER PURGE &
WASTE
800
WWT
640 600 LIGNIN
CHEMICALS
BLACK LIQUOR,
810
FLASH CONDENSATE
FILTER PRESS
HYDROGEN 510
A500
AIR OUT
A600 PROCESS
CATALYTIC 530
FLUE
850 GAS
EVAPORATION 660 WASTEWATER WATER CONVERSION &
FURNACE AIR 520
RECYCLE
TREATMENT UPGRADING
670 FGD LIME 860
WATER
OFF-GAS

HYDROCARBON
AIR 870 820

PRODUCT
700
AUX FUEL 880

STACK FLUE
BIOGAS
A800 890A
GAS
830 BURNER/BOILER
SLUDGE
TURBOGENERATOR 890B ASH

A700
840 BLOWDOWN
890C
VENT STORAGE
BOILER BLOWDOWN
220 STEAM TO A200
610
A900
UTILITIES PRODUCED WATER
OVERALL PROCESS:
620 BIOMASS TO HYDROCARBON
NOV. 2014 PFD-P120-A000

Figure 22. Overall process schematic for Aspen stream tables

112
Aspen Plus mass balance information for key stream tables:

113
114
115
116
117
118
 CONDENSATE TO WATER
RECYCLE POOL

HYDROLYSATE FROM
FILTERED HYDROLYSATE
STORAGE TANK PFD-301

E-303
HYDROLYSATE
CONCENTRATION

T-303
CONCENTRATED HYDROLYSATE
STORAGE TANK

POLISHING FILTER
RETENTATE TO
LIGNIN FILTER PFD-301

F-320
POLISHING FILTER

RECYCLE PURGE TO WWT

PFD-601
T-321
POLISHED HYDROLYSATE
STORAGE TANK

W-330
ION EXCHANGE

HYDROLYSATE TO CATALYTIC
CONVERSION PFD-501

T-331
DEIONIZED SUGAR
STORAGE TANK

CATALYTIC CONVERSION OF LIGNOCELLULOSIC

HYDROLYSATE TO HYDROCARBON BIOFUELS

PROCESS FLOW DIAGRAM

AREA 300: HYDROLYSATE CONDITIONING
AND PURIFICATION
0 6/30/14 JAR ISSUED FOR FINAL REPORT Project No: Drawing:

30495.00 PFD-302
Rev:

Rev. Date By Description

0

119
120
121
 MAKEUP PURGE GAS TO BOILER
HYDROGEN PFD-801

MAKEUP HYDROGEN
COMPRESSOR RECYCLE
HYDROGEN
COMPRESSOR

PURGE

NOTE 1
MAKEUP NATURAL GAS

PREHEATER

HYDROTREATING
REACTOR
SEPARATOR

PRODUCED WATER TO
WASTEWATER TREATMENT
PFD-601

ORGANIC PHASE PARTIAL

FROM CATALYTIC CONDENSER
CONVERSION PFD-501 FEED-
EFFLUENT OFFGAS
HEAT
EXCHANGER
WATER NAPHTHA PRODUCT
COOLER
SEPARATOR AIR COOLER

PRODUCT
DISTILLATION
COLUMN

REBOILER

DIESEL PRODUCT

S-570
HYDROTREATING UNIT

NOTES:
1. NATURAL GAS FOR STARTUP AND AS NEEDED CATALYTIC CONVERSION OF LIGNOCELLULOSIC
HYDROLYSATE TO HYDROCARBON BIOFUELS

PROCESS FLOW DIAGRAM

AREA 500: UPGRADING
0 6/30/14 JAR ISSUED FOR FINAL REPORT Project No: Drawing:

30495.00 PFD-502
Rev:

Rev. Date By Description

0

122
 BIOGAS TO BOILER
PFD-801

BOILER BLOWDOWN

PRETREATMENT FLASH AND

BLACK LIQUOR R-609 S-610 TREATED WASTEWATER
MEMBRANE REVERSE OSMOSIS TO RECYCLE
BIOREACTOR SYSTEM
WASTEWATER FROM T-608
HYDROLYSATE
PURIFICATION AND
AERATION DIGESTER
CONVERSION
T-606
ANAEROBIC BASIN

MICROFILTER RECYCLE
PURGE

M-640
EVAPORATOR SYSTEM

WASTEWATER BRINE
TO DISPOSAL

POLYMER

SLUDGE TO BOILER
PFD-801
E-644
S-611
GRAVITY BELT
S-611
THICKENER
CENTRIFUGE

CATALYTIC CONVERSION OF LIGNOCELLULOSIC

HYDROLYSATE TO HYDROCARBON BIOFUELS

PROCESS FLOW DIAGRAM

AREA 600: WASTEWATER TREATMENT
0 6/30/14 JAR ISSUED FOR FINAL REPORT Project No: Drawing:

30495.00 PFD-601
Rev:

Rev. Date By Description

0

123
124