ENVIRONMENTAL ASSESSMENT
OF WOOD DERIVED
HEMICELLULOSIC ETHANOL
USING ALTERNATIVE LCA
MODELS
Rachel Bowman, Anthony Halog
Forest Bioproducts Research Initiative
University of Maine
Massachusetts Institute of Technology
Western Kentucky University
ACLCA Conference
Presentation Format
Project Overview
 Why?
 Description of System
 Functional Unit
 Assumptions
 Graphs and Results for LCA Models
 Pros and Cons of LCA Models
 Conclusions

Project Overview
Use Lifecycle Assessment (LCA) to
evaluate the environmental impact of
wood derived bioethanol
 Complete LCA with commercial SimaPro
model, Eco-LCA , and openLCA
 Compare results of three LCAs and
determine which model best represents
the data
 Learn more about openLCA and EcoLCA

Why?
Energy crisis worldwide calls for new,
renewable energy sources like bioethanol
 Many LCAs have been performed of corn
derived bioethanol but not wood
derived– wood doesn’t compete with
food
 The near-neutral extraction process is
fairly new and knowing its full lifecycle
before widely implementing it is
important

Description of System
Schematic of Near-Neutral Extraction Process to produce bioethanol1
Description of System

Hemicellulose is extracted before pulping
◦ Reagents: anthraquinone and green liquor
(sodium hydroxide, sodium sulfide, sodium
carbonate)
◦ Green liquor- breaks apart hemicellulose
◦ Anthraquinone- works as catalyst to maintain
pulp quality/quantity
Functional Unit
9.8x109 Btu worth of ethanol (density
.789 g/mL and 75,700 Btu per gallon)
 39.5 tonnes ethanol, 45.1 tonnes acetic
acid, 1000 tonnes pulp produced daily
 From 2,166 tonnes of wood chips
 From Mao’s provisional data

Assumptions
Average moisture content of the wood is
46%
 All wood is assumed to be a product of
the Northeast Region of the United
States
 Wood chipped in the field using mobile
wood chipping technology
 Wood is composition 42.6% cellulose,
29.7% hemicellulose, 27.5%lignin, and .2%
ash on a dry wood weight basis

Allocation
SimaPro 7.1 model was allocated by the
mass of the products
 openLCA model would have been
allocated by mass of products
 Eco-LCA allows only for economic
allocation

◦ Allocated economically by value of wood
chips, Kraft pulp, acetic acid, and bioethanol
Data Collection
Primary data– consultation with Dr.
Genco in Chemical Engineering Dpt.
 Mao’s original WinGEMS model from
thesis for LCI input-output data
 Secondary data:

◦
◦
◦
◦
Wooley’s NREL Report
US Data (preferrably Northeast [Maine])
Relevant timeframe
Relevant processes
LCI Databases Used

SimaPro
◦ Ecoinvent
◦ Franklin
◦ USLCI

Eco-LCA
◦ Eco-LCA standard ecosystem goods and
services database

OpenLCA
◦ Would have used Ecoinvent, like SimaPro
SimaPro 7.1 Normalized Results
Fossil fuels, respiratory inorganics, climate change,
and acidification are most affected impact
categories
 Biggest contributing processes to these
environmental impacts are the liming/causticizing
process, acid hydrolysis, wood chipping/wood
production, and hog fuel boiler
 Heavy fuel oil consumption in the liming process
contributes most to the fossil fuel use

◦ large amounts of oil must be pumped into the system
for the liming process to take place

Liming process also heavy in respiratory
inorganics
SimaPro 7.1 Normalized Results–
Mass Allocation
6.5
6
5.5
5
4.5
4
3.5
3
2.5
2
1.5
1
0.5
0
Carcinogens
Resp. organic
s
Resp. inorgan
ics
Wood Chips
Evaporation- Ethanol
Liming and Causticizing
Recovery Boiler
Climate
change
Radiation
Ozone layer
Wood Extraction
Acid Hydrolysis
Fermentation
Ecotoxicity
Acidification
/ Eutrophicati
Split
Lignin Filtration
Ethanol and Acetic Acid Separation
Analyzing 1 p 'Ethanol Production'; Method: Eco-indicator 99 (H) V2.06 / Europe EI 99 H/H / normalization
Normalized Results of SimaPro Model
Land use
Minerals
Flash Tanks
Liquid-Liquid Extraction
Hog Fuel Boiler2
Fossil fuels
SimaPro 7.1 Normalized Results–
Economic Allocation
46000
44000
42000
40000
38000
36000
34000
32000
30000
28000
26000
24000
22000
20000
18000
16000
14000
12000
10000
8000
6000
4000
2000
0
Carcinogens
Resp. organics
Resp. inorganics
Wood Chips
Acid Hydrolysis
Ethanol and Acetic Acid Separation
Climate change
Radiation
Wood Extraction
Lignin Filtration
Hog Fuel Boiler2
Analyzing 1 p 'Ethanol Production'; Method: Eco-indicator 99 (H) V2.06 / Europe EI 99 H/H / normalization
Ozone layer
Split
Liquid-Liquid Extraction
Recovery Boiler
Ecotoxicity
Acidification
/ Eutrophication
Flash Tanks
Liming and Causticizing
Land use
Minerals
Evaporation- Ethanol
Fermentation
Fossil fuels
SimaPro 7.1 Single Score Results–
Mass Allocation
2.6
2.4
2.2
2
1.8
kPt
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
Wood Chips
Wood Extra
ction
Carcinogens
Ozone layer
Fossil fuels
Split
Flash Tanks
Evaporatio
n- Ethanol
Resp. organics
Ecotoxicity
Acid Hydrol
ysis
Lignin Filtra
tion
Liquid-Liqui
d Extractio
Resp. inorganics
Acidification/ Eutrophication
Liming and
Causticizin
Fermentati
on
Ethanol and
Acetic Acid
Climate change
Land use
Analyzing 1 p 'Ethanol Production'; Method: Eco-indicator 99 (H) V2.06 / Europe EI 99 H/H / single score
Single Score Representation of SimaPro model
Hog Fuel
Boiler2
Radiation
Minerals
Recovery
Boiler
SimaPro 7.1 Single Score Results–
Economic Allocation
18
17
16
15
14
13
12
MPt
11
10
9
8
7
6
5
4
3
2
1
0
Wood Chips
Wood Extracti
on
Carcinogens
Ecotoxicity
Split
Flash Tanks
Evaporation
- Ethanol
Resp. organics
Acidification/ Eutrophication
Acid Hydrolysis Lignin Filtration
Resp. inorganics
Land use
Analyzing 1 p 'Ethanol Production'; Method: Eco-indicator 99 (H) V2.06 / Europe EI 99 H/H / single score
Liquid-Liquid
Extraction
Climate change
Minerals
Liming and
Causticizing
Fermentation
Radiation
Fossil fuels
Ethanol and
Acetic Acid
Hog Fuel
Boiler2
Ozone layer
Recovery
Boiler
SimaPro 7.1 Single Score Results
The liming and causticizing process
carries the heaviest environmental impact
of all the processes in producing
bioethanol
 Heavy fuel oil consumption in the liming
process
 Respiratory inorganic output of the
process,

◦ sulfate ions being released during the
causticizing process
SimaPro 7.1 Single Score Results
Other process that has a heavy impact is the
wood chip production.
 In line with the Bhander/Halog study that
showed that wood chip production had a
heavy environmental impact
 Includes everything from the planting of the
trees to the felling and chipping process,
 Heaviest in climate change and respiratory
inorganics

◦ The climate change category includes green
house gas emissions from gases such as CO2, SOx,
NOx, and CH4.
Eco-LCA Emissions Results
Emissions Normalization of Eco-LCA model
Eco-LCA Emissions Results
Showed that the pulp production process
is far heavier in green house gas emissions
than logging, acetic acid, or ethanol
production
 Heaviest impact categories were ethanol
and ammonia, in land, air, and water

◦ Ammonia accounted for is most likely the
respective inorganic material that is shown in
SimaPro model produced from the liming/
causticizing process and wood chipping
Eco-LCA Emissions Results
Emissions mostly from pulping process,
which makes sense since the model is
allocated economically
 Particulate matter (<10 µm) is shown,
probably some of the climate change
SimaPro is accounting for

Eco-LCA CO2 Emission Results
CO2 Normalization of Eco-LCA model
Eco-LCA CO2 Emissions Results

Show that more carbon is actually sequestered as
a result of the production of ethanol than is
emitted
◦ trees being used to feed the pulping plant sequester
more carbon during growth than is emitted in pulping
process

Kraft pulping process doesn’t use (much)
external energy
◦ Energy required is produced in the hog fuel and
recovery boilers
◦ Plant is not using any external electricity from fossil
fuels like coal or natural gas that would increase
emissions
Eco-LCA Energy Consumption
Results
Energy Normalization of Eco-LCA model
Eco-LCA Energy Consumption
Results

Logging and chipping processes consume the
most energy in process.
◦ heaviest in sunlight and wood energy
 Imply that trees absorb sunlight as part of their growth
process
 Some of the wood waste is later used to produce steam in
the hog fuel boiler

The other energy consumed in the process was
generally low in all categories, especially in that of
fossil fuels.
◦ Any fossil fuel use can be attributed to the equipment
used for logging and chipping, such as mobile wood
chippers or the trucks that transport the chips to the
Kraft Pulp Mill
Eco-LCA Water Consumption
Results
Water Normalization of Eco-LCA model
Eco-LCA Water Consumption
Results
The majority of the water consumed in
pulping process
 Can be attributed to economic allocation
 Consumption of water for logging is the
highest with respect to agriculture and
feedstock water,
 Powerplant water had the highest total
consumption in the process

Eco-LCA Water Consumption
Results
Can likely be attributed to the fact that
water coming to the Kraft Pulp Mill would
be coming from a supply other than the
public water source and being filtered and
treated separately from the public drinking
water.
 Water supply in any chemical plants would
also be processed and filtered separately
from the public water supply
 Water supply for the ethanol and acetic acid
production is assumed to be the same as for
the Kraft Pulp Mill

OpenLCA (The Abbreviated
Version)
Issues getting openLCA database to load
 Designed to simply import complete
SimaPro 7.1 model into openLCA in
EcoSpold format

◦ EcoSpold formatting only available with
SimaPro Developer version
 FBRI only has Analyst version
Bottom Line: openLCA model was never
completed due to software issues
Pros and Cons- SimaPro

Cons:
◦ Cost-- ~$14,000 (US) for permanent license
for developer version
◦ Doesn’t account for ecosystem goods or
services
◦ Doesn’t show graphically what specifically is
contributing to each impact category
◦ Minimal US specific data
 Should improve in future versions as LCA becomes
more popular in North America
Pros and Cons- SimaPro
Ecoinvent database was the most
comprehensive database of models
 Allows simple loading of outside data and
processes into the program
 Extensive tutorials, user guides, etc. for
LCA beginners
 Many options for allocating, characterizing,
and normalizing data (methods)

◦ Specifically used Ecoindicator in this study
Pros and Cons- Eco-LCA

Cons
◦
◦
◦
◦
Not as many options in general as SimaPro
Only set up to handle economic allocation
Hard to enter outside data
Only very simple tutorial available, a little
difficult for an inexperienced user to figure
out
◦ Difficult to transfer results out of flashplayer
for viewing in papers, presentations, etc.
Pros and Cons- Eco-LCA

Pros
◦ Includes ecosystem goods and services–
allows user to get bigger and better view of
lifecycle
◦ While no user guide, developers are easily
accessible for questions
◦ Free
◦ Provides graphical representation of different
substances that are contributing to an impact
category
Pros and Cons- OpenLCA (In a
theoretical world where SimaPro
Developer was available)

Cons
◦ Individual database didn’t load– what would
someone do without another imported
model?
◦ User manual is difficult to access but does
exist

http://www.openlca.org/uploads/media/Guide_to_testing_the_openLCA_framework_Alphaversion.pdf
◦ Still being developed– still some issues
◦ A little difficult to learn how to use
Pros and Cons- OpenLCA (In a
theoretical world where SimaPro
Developer was available)

Pros
◦ Includes file format converter to directly
import data from another LCA model in
SimaPro or Excel
◦ Flows set up allows for the production of a
system model as well as an LCA
◦ Free and publicly available
◦ Good graphical representation once LCA is
complete
Conclusions
Mass and economic allocation have about
same results in SimaPro, so economic
allocation is probably accurate
 Eco-LCA provides best representation of
this data set despite allocation only
economically
 Results of openLCA would have looked a
lot like the SimaPro results, but provided
an input-output model as well

Conclusions
The real advantage with Eco-LCA is the
inclusion of ecosystem goods and
services– gives much better ‘big picture’
outlook of system
 Using a free program makes it so that one
can report their conclusions to anyone,
whether they have a license for the
program or not

Conclusions
It appears that the actual results from the
Eco-LCA and SimaPro models are pretty
similar
 Eco-LCA has different impact categories,
but one can see how the high impact
substances contribute to the more
general impact categories in SimaPro

Acknowledgements
Funding-- provided by The National
Science Foundation grants EEC-06 48793,
EPS-05 54545
 REU program/FBRI
 Dr. Halog
 Dr. Genco
 Shashi Dhungel and Binod Neopane,
graduate students
