LIFE CYCLE ASSESSMENT OF
SUBSURFACE WATER RETENTION TECHNOLOGY
By
Ning Gong

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Packaging-Master of Science
2014

ABSTRACT
LIFE CYCLE ASSESSMENT OF SUBSURFACE WATER RETENTION TECHNOLOGY
By
Ning Gong
Subsurface water retention technology (SWRT) is a system of contoured engineered
membranes made of a blend of linear low and low density polyethylene, and drip irrigation
installed below the root zone to increase water use efficiency. It is claimed that SWRT is
capable of increasing crop yields by 50 to 400% through extending the water retention
time in the root zone. More than four million hectares of sandy soil area in the U.S. creates a
great opportunity for using SWRT to improve the soil use efficiency; however, SWRT’s
environmental footprint (EFP) is still unknown. The focus of this LCA study was to
determine the EFP of SWRT compared to control methods and previous benchmark studies
in growing corn. The functional unit was selected as 1000 kg of corn grain grown in sandy
soil in Michigan, U.S. The study boundary and temporal condition was cradle-to-gate with
data obtained from 2000 to 2013. The ReCiPe 1.07 (H) midpoint impact assessment
methodology was used. Six different treatments (15” -row spacing- SWRT, 30” SWRT, 15”
Control, 30” Control, Irrigated SWRT, and Nonirrigated Control) planted on Sandhill farm,
Michigan State University were evaluated. A simulation of continuous corn production on
the Sandhill farm was also performed to estimate the pay-off time for the additional burden
of installing SWRT. The benchmark comparison confirmed that the study results were in a
reasonable range. The study identified climate change, fossil fuel depletion, and terrestrial
acidification as the largest impacts for SWRT. Contribution analysis indicated that
irrigation was a hotspot activity. The parameter of corn grain yield was highly sensitive.

Copyright by
NING GONG
2014

ACKNOWLEDGEMENTS

I received tremendous support and help with this project. I am indebted to many people
who contributed their work with my thesis. First of all, I would like express my sincere
thanks to my parents. Without their supports on both spirit and finance, I would not be
able to finish this thesis. Second, I cannot find words to express my gratitude to Dr. Auras,
my supervisor. Thank you for his guidance, support, and patient along the past three years.
It is my honor to be one of RAA member, being influenced largely from Dr. Auras and
everyone in the group. Thanks to Dr. Alvin J.M. Smucker, together with the SWRT project’s
members for providing the significant contributions and critical advices on this thesis.
Thanks to Dr. Selke for her valuable inputs and care to both my research and living. Also, I
wish to express my thanks to Dr. Bruno Basso. Without his support, I am afraid my thesis
could not be done in another one year. Also, I share the credit of my work with my
boyfriend Yao Li who helped me go over the fear and loss when started this project, and
brought positive energy to my life. Thanks to all my friends for their encouragement and
the good time we had. They are the parts I am bet to miss the most of life in MSU.

iv

TABLE OF CONTENTS

LIST OF TABLES................................................................................................................................................. viii
LIST OF FIGURES ................................................................................................................................................. xii
KEY TO ABBREVIATIONS ................................................................................................................................ xv
Introduction & Motivation................................................................................................................................. 1
REFERENCES .......................................................................................................................................................... 3
Literature Review ................................................................................................................................................. 6
2.1
SWRT background ...................................................................................................................... 6
2.2
Field management ................................................................................................................... 10
2.3
Water management................................................................................................................. 12
2.3.1
Goal of water management ......................................................................................... 12
2.3.2
Water demands in different corn grown phases ................................................ 13
2.3.3
Soil water............................................................................................................................ 15
2.3.4
Irrigation approaches .................................................................................................... 16
2.4
Chemical fertilizers ................................................................................................................. 17
2.4.1
Nitrogen fertilizer ........................................................................................................... 17
2.4.2
Phosphate fertilizer ........................................................................................................ 19
2.4.3
Potassium fertilizer ........................................................................................................ 19
2.5
Corn harvest............................................................................................................................... 20
2.6
Crop residue utilization ......................................................................................................... 20
2.7
Tillage ........................................................................................................................................... 22
2.8
Introduction of crop growth and management models ............................................ 23
2.9
Land use....................................................................................................................................... 25
2.10
Overview of life cycle assessment ..................................................................................... 27
2.11
Framework of LCA ................................................................................................................... 30
2.11.1 Goal and Scope Definition ............................................................................................ 31
2.11.2 Inventory Analysis .......................................................................................................... 33
2.11.3 Impact Assessment ......................................................................................................... 33
2.11.4 Interpretation ................................................................................................................... 34
2.12
Allocation and system expansion methods ................................................................... 34
2.13
Past corn studies ...................................................................................................................... 37
2.14
Identified reasons of large variations .............................................................................. 38
REFERENCES ....................................................................................................................................................... 42
Goal and Scope .................................................................................................................................................... 49
3.1
Intended use of study ............................................................................................................. 49
3.2
Target audience ........................................................................................................................ 50
3.3
Functional unit .......................................................................................................................... 51
3.4
System boundary ..................................................................................................................... 51
v

3.5
Cut-off rules................................................................................................................................ 52
3.6
Allocation rules ......................................................................................................................... 52
3.7
Temporal and technology representative ...................................................................... 53
3.8
Software and data collection ............................................................................................... 53
3.9
Life cycle impact assessment methodology and impact categories ..................... 53
REFERENCES ....................................................................................................................................................... 55
Life Cycle Inventory .......................................................................................................................................... 57
4.1
Project design ............................................................................................................................ 57
4.1.1
Experimental dataset..................................................................................................... 57
4.1.2
Simulation data ................................................................................................................ 58
4.2
Overview ..................................................................................................................................... 61
4.3
Machinery ................................................................................................................................... 63
4.3.1
Tillage .................................................................................................................................. 64
4.3.2
Sowing ................................................................................................................................. 67
4.3.3
Fertilizing ........................................................................................................................... 69
4.3.4
Combine harvesting ....................................................................................................... 71
4.4
Irrigation ..................................................................................................................................... 73
4.4.1
Drip irrigation, irrigation water and electricity .................................................. 73
4.4.2
Drip tape production ..................................................................................................... 74
4.4.3
End of life (EOL) of disposed drip tape ................................................................... 75
4.5
Chemical ...................................................................................................................................... 75
4.6
Seed .............................................................................................................................................. 78
4.7
SWRT ............................................................................................................................................ 81
4.8
Planting ........................................................................................................................................ 83
4.8.1
Field preparation............................................................................................................. 84
4.8.2
Harvest ................................................................................................................................ 85
4.9
Calculation procedure ............................................................................................................ 86
4.9.1
Seed flow calculation ..................................................................................................... 87
4.9.2
Chemical flow calculation ............................................................................................ 88
4.9.3
Machinery flow calculation ......................................................................................... 88
4.9.4
Irrigation flow calculation ........................................................................................... 89
4.9.5
SWRT flow calculation .................................................................................................. 90
4.9.6
Biogenic carbon flow calculation .............................................................................. 90
4.9.7
Land occupation flow calculation ............................................................................. 91
4.9.8
Fertilizer emission flow calculation ......................................................................... 92
4.10
Assumptions .............................................................................................................................. 93
4.10.1 Machinery .......................................................................................................................... 93
4.10.2 Irrigation ............................................................................................................................ 94
4.10.3 Chemical ............................................................................................................................. 95
4.10.4 Seed ...................................................................................................................................... 95
4.10.5 SWRT.................................................................................................................................... 95
4.10.6 Planting ............................................................................................................................... 96
4.10.7 Missing data ...................................................................................................................... 96
APPENDIX ............................................................................................................................................................. 97
REFERENCES ..................................................................................................................................................... 116
vi

Results and interpretation ........................................................................................................................... 119
5.1
Evaluation of result quality ................................................................................................ 119
5.1.1
Completeness check ..................................................................................................... 119
5.1.2
Consistency check ......................................................................................................... 122
5.1.3
Contribution analysis .................................................................................................. 126
5.2
LCIA Results ............................................................................................................................. 129
5.2.1
LCIA Results of experimental treatments ............................................................ 130
5.2.2
Benchmark of published studies ............................................................................. 138
5.3
Scenario comparisons .......................................................................................................... 141
5.3.1
Yield increase scenario ............................................................................................... 141
5.3.2
Drip tape lifetime scenario ........................................................................................ 143
5.3.3
Scenarios regarding multifunctionality methods of allocation ................... 147
5.3.4
LCA methodology scenario ........................................................................................ 151
5.4
Uncertainty analyses ............................................................................................................ 154
5.4.1
Data quality evaluation ............................................................................................... 154
5.4.2
Land use (LU) ................................................................................................................. 154
5.4.3
Monte Carlo simulations based on the pedigree matrix ................................ 163
APPENDICES ...................................................................................................................................................... 164
Appendix A5: Contribution analyses ........................................................................................................ 164
Appendix B5: Allocation scenarios............................................................................................................ 198
Appendix C5: Pedigree matrix .................................................................................................................... 201
REFERENCES ..................................................................................................................................................... 212
Normalization and weighting ...................................................................................................................... 215
6.1 Normalization ..................................................................................................................................... 216
6.2 Weighting.............................................................................................................................................. 219
REFERENCES ..................................................................................................................................................... 220
Conclusions and future work ...................................................................................................................... 222
7.1
Conclusions .............................................................................................................................. 222
7.2
Future work ............................................................................................................................. 226

vii

LIST OF TABLES

Table 4- 1 Yield and irrigation of year 2012 experiment ...................................................................58
Table 4- 2 Yield and irrigation of year 2013 experiment ...................................................................58
Table 4- 3 Unit processes grouping of corn production system .......................................................63
Table 4- 4 Representation of tillage activity ............................................................................................65
Table 4- 5 Input /Output flows of chisel tillage 1 ha farm .................................................................66
Table 4- 6 Input /Output flows of sowing corn seed per 1 ha farm with .....................................68
Table 4- 7 Representation of processes in sowing activity ................................................................69
Table 4- 8 Input /Output flows of fertilizing by broadcaster process ............................................70
Table 4- 9 Representation of processes in fertilizing activity ...........................................................71
Table 4- 10 Inputs and Outputs of combine harvesting process ......................................................72
Table 4- 11 Representation and key assumptions of harvesting activity .....................................73
Table 4- 12 Chemical processes description ............................................................................................77
Table 4- 13 LCI of seed process .....................................................................................................................79
Table 4- 14 Representation and data resource of SWRT processes ...............................................83
Table 4- 15 LCI of field preparation process in irrigated SWRT plan.............................................85
Table 4- 16 LCI of harvest process in irrigated SWRT plan ................................................................86
viii

Table A- 1 LCI of 2012 experiment ..............................................................................................................98
Table A- 2 LCI of 2013 experiment ........................................................................................................... 102
Table A- 3 LCI of 2004-2013 simulation ................................................................................................. 105
Table A- 4 Soil profile in SALUS .................................................................................................................. 108
Table A- 5 Crop management profile ....................................................................................................... 108
Table A- 6 Yields from SALUS simulation for year 2004 to 2013 ................................................. 109
Table A- 7 Difference between average yields and simulated aggregate NOSWRT ............... 109
Table A- 8 LUs summary of experiment treatments .......................................................................... 110
Table A- 9 Field preparation process flow calculations of experiment treatments ............... 111
Table A- 10 Field preparation process flow calculations of simulated treatments ............... 113
Table A- 11 LU [acre] to produce 1000 kg grain of simulated treatments ................................ 114
Table A- 12 Drip tape measurement record .......................................................................................... 115
Table 5- 1 Completeness check .................................................................................................................. 120
Table 5- 2 Consistency check ...................................................................................................................... 123
Table 5- 3 Published corn grain study result for reference ............................................................. 140
Table 5- 4 Comparison of LCIA from database and published studied results, and the 2012
and 2013 SWRT and Ctrl results ................................................................................................................ 140
Table 5- 5 Time [year] to pay-off SWRT burden if yield increase due to SWRT application
................................................................................................................................................................................. 143

ix

Table 5- 6 Mean and SD of LU ..................................................................................................................... 155
Table 5- 7 Water consumption comparisons ........................................................................................ 162
Table A5- 1 LCIA of 2012 15” SWRT for contribution analysis...................................................... 166
Table A5- 2 LCIA of 2012 30” SWRT for contribution analysis...................................................... 168
Table A5- 3 LCIA of 2012 15” Ctrl for contribution analysis .......................................................... 170
Table A5- 4 LCIA of 2012 30” Ctrl for contribution analysis .......................................................... 172
Table A5- 5 LCIA of 2013 Irrigated SWRT for contribution analysis........................................... 174
Table A5- 6 LCIA of 2013 Nonirrigated SWRT for contribution analysis .................................. 176
Table A5- 7 LCIA of simulated 2004 Ctrl for contribution analysis ............................................. 178
Table A5- 8 LCIA of simulated 2004 SWRT for contribution analysis ........................................ 180
Table A5- 9 Flow relative contributions (%) of 2012 15” SWRT .................................................. 182
Table A5- 10 Flow relative contributions (%) of 2012 30” SWRT ............................................... 184
Table A5- 11 Flow relative contributions (%) of 2012 15” Control ............................................. 186
Table A5- 12 Flow relative contributions (%) of 2012 30” Ctrl .................................................... 188

x

Table A5- 13 Flow relative contributions (%) of 2013 Irrigated SWRT .................................... 191
Table A5- 14 Flow relative contributions (%) of 2013 Nonirrigated SWRT ............................ 193
Table A5- 15 Flow relative contributions (%) of 2004 simulated Ctrl ....................................... 195
Table A5- 16 Flow relative contributions (%) of 2004 simulated SWRT .................................. 197
Table B5- 1-a Allocation scenarios on 15” SWRT and 30” SWRT ................................................. 199
Table B5- 1-b Allocation scenarios on 15” Ctrl and 30” Ctrl ........................................................... 200
Table B5- 1-c Allocation scenarios on Irrigated SWRT and Nonirrigated Ctrl ........................ 201
Table C5- 1 Pedigree matrix used to assess the data quality .......................................................... 203
Table C5- 2 Pedigree matrix used for uncertainty analysis ............................................................. 204
Table C5- 3 Monte Carlo simulation on SWRT machine diesel consumption rate ................. 210
Table C5- 4 Monte Carlo simulation on drip tape production ........................................................ 211
Table C5- 5 Monte Carlo simulation on seed ......................................................................................... 212
Table 6- 1 ReCiPe 1.07 (H) World midpoint normalization factor .............................................. 216
Table 6- 2 LCIA of SWRT .............................................................................................................................. 218
Table 6- 3 Normalized LCIA ........................................................................................................................ 219

xi

LIST OF FIGURES

Figure 2- 1 SWRT configuration ..................................................................................................................... 9
Figure 2- 2 Components of SALUS [39] ......................................................................................................25
Figure 2-3 Stages of an LCA [1] .....................................................................................................................31
Figure 3-1 System boundary ..........................................................................................................................52
Figure 4- 1 Average grain yield from year 1942-2012 in Ingham, MI [1] .....................................59
Figure 4- 2 Corn grain production systems ..............................................................................................62
Figure 4- 3 Tillage process model.................................................................................................................64
Figure 4- 4 Sowing plan model .....................................................................................................................67
Figure 4- 5 Drip irrigation process ..............................................................................................................74
Figure 5- 1 Contribution analysis of 2012 SWRT treatments: 15”SWRT (left top), 30’’SWRT
(right top), 15” Ctrl (left bottom), and 30” Ctrl (right bottom) ..................................................... 127
Figure 5- 2 Contribution analysis of 2013 Irrigated SWRT (left top), Nonirrigated Ctrl (right
top), 2004 simulated Ctrl (left bottom), and 2004 simulated SWRT (right bottom) ............ 128
Figure 5- 3 ReCiPe 1.07 (H) Midpoint of Agricultural Land Occupation ................................... 130
Figure 5- 4 ReCiPe 1.07 (H) Midpoint of Climate Change ............................................................... 131
Figure 5- 5 ReCiPe 1.07 (H) Midpoint of Fossil Depletion .............................................................. 132

xii

Figure 5- 6 ReCiPe 1.07 (H) Midpoint of Freshwater Ecotoxicity (top) and freshwater
eutrophication (bottom) ............................................................................................................................... 134
Figure 5- 7 ReCiPe 1.07 (H) Midpoint of Human Toxicity ............................................................... 135
Figure 5- 8 ReCiPe 1.07 (H) Midpoint of Terrestrial Acidification .............................................. 136
Figure 5- 9 ReCiPe 1.07 (H) Midpoint of Water Depletion ............................................................. 137
Figure 5- 10 Time to pay-off SWRT burden on climate change impact ..................................... 142
Figure 5- 11 Drip tape lifetime scenario on climate change .......................................................... 145
Figure 5- 12 Drip tape lifetime scenario on fossil depletion .......................................................... 145
Figure 5- 13 Drip tape lifetime scenario on human toxicity .......................................................... 146
Figure 5- 14 Drip tape lifetime scenario on terrestrial acidification .......................................... 146
Figure 5- 15 System expansion corn system ........................................................................................ 150
Figure 5- 16 Allocation scenarios on agricultural land occupation ............................................. 151
Figure 5- 17 Impact assessment methodology scenarios on climate change .......................... 152
Figure 5- 18 Agricultural land occ. [m2 * a], columns with the different lower case letters
are statistically significant different at 95% CI, with a indicating the lowest, using the
simulated LU mean and SD of simulated LU mean ............................................................................. 157
Figure 5- 19 LU Climate change [kg CO2 eqv.], columns with the different lower case letters
are statistically significant different at 95% CI, with a indicating the lowest, using the
simulated LU mean and SD of simulated LU mean ............................................................................. 157
Figure 5- 20 LU Fossil depletion [kg oil eqv.], columns with the different lower case letters
are statistically significant different at 95% CI, with a indicating the lowest, using the
simulated LU mean and SD of simulated LU mean ............................................................................. 158
xiii

Figure 5- 21 Freshwater ecotoxi. [Kg 1, 4 -DB eqv.], columns with the different lower case
letters are statistically significant different at 95% CI, with a indicating the lowest, using
the simulated LU mean and SD of simulated LU mean ..................................................................... 159
Figure 5- 22 Freshwater eutrophication [kg P eqv.], columns with the different lower case
letters are statistically significant different at 95% CI, with a indicating the lowest, using
the simulated LU mean and SD of simulated LU mean ...................................................................... 159
Figure 5- 23 Human toxicity. [Kg 1, 4 -DB eqv.], columns with the different lower case
letters are statistically significant different at 95% CI, with a indicating the lowest, using
the simulated LU mean and SD of simulated LU mean ..................................................................... 160
Figure 5- 24 Terrestrial acidification [kg SO2 eqv.], columns with the different lower case
letters are statistically significant different at 95% CI, with a indicating the lowest, using
the simulated LU mean and SD of simulated LU mean ...................................................................... 160
Figure 5- 25 Water depletion [m3], columns with the different lower case letters are
statistically significant different at 95% CI, with a indicating the lowest, using the simulated
LU mean and SD of simulated LU mean .................................................................................................. 162

xiv

KEY TO ABBREVIATIONS

agg

Aggregated unit process

aLCA

Attributional LCA

AMS

Ammonium Sulfate

AN

Ammonium nitrate

CI

Confident interval

cLCA

Consequential LCA

EC

Electronic conductivity

EFP

Environmental footprint

EOL

End of life

FU

Functional unit

GHG

Greenhouse gas

IPCC

Intergovernmental Panel in Climate Change

ISO

International Organization for Standardization

LCA

Life cycle assessment

LCI

life cycle inventory

LCIA

Life cycle impact assessment

LLDPE

Linear low density polyethylene

LU

Land use

MSW

Municipal solid waste

NMVOCs Non-methane volatile organic compounds
PE

Polyethylene

xv

SALUS

The system approach to land use sustainability model

SD

Standard deviation

SOM

Soil organic matter

SRR

Stover removal rates

SWRT

Subsurface water retention technology

USDA

United States Department of Agriculture

u-so

Single operation unit process

xvi

Introduction & Motivation
Irrigation withdrawal, about one third of the nation water use in 2005, was the
largest use of freshwater in U.S. [1]. Subsurface water retention technology (SWRT)
is a new generation of agricultural water saving technology. A contour shaped
plastic membrane based on a polyethylene blend and drip irrigation system are the
major components of SWRT. The membrane is installed below the root zone to
extend the water retention time in soil. A drip irrigation system is used to reduce
the water loss during the water amendment process. SWRT not only increase the
water use efficiency by extending retention time, but it also increases the yield by
enhancing the grain to shoot ratio [2]. Considering that 61% of the crop growing
area in Michigan is sandy soil [3], there is great potential for SWRT to increase
Michigan agricultural soil productivity.
The largest grain stock in the U.S. is corn, which account for 64.2% of the total
grain stock [4]. According to a USDA statistical service report, the U.S. corn grain
production was 353,698,734 metric tons in 2013 [5]. Many environmental
performance studies regarding corn grain and corn stover production have
previously been published [6-9]. Most of these studies were aimed to provide an
insight to the corn-ethanol and stover-ethanol production discussion. The main
driving reasons for the corn related biofuel life cycle assessment LCA studies were
the rising petroleum price and federal subsidies. Only a few studies explored factors
affecting the environmental footprint (EFP) of corn production such as irrigation
and row spacing. These might be driven by the small amount of irrigated corn in the
US [10]. However, studies showing the great economic return for irrigation begin to

1

raise the attraction of the irrigation method to a growing number of corn growers
[11]. Until now, none of the available studies evaluate the potential EFP of using
SWRT for crop production.
Thus, the objective of this study is using LCA to evaluate the EFP of growing corn
using SWRT in sandy soil with various row spacings and irrigation management
strategies. Specifically, the study will focus on two main parts: a) using experimental
data from growing corn in sandy soil during the years 2012 and 2013, and b)
creating simulated data for corn production from 2004 to 2013 to determine the
EFP of SWRT for long term corn production and the pay-off of using SWRT. The
overall motivation of this study is to understand the EFP of SWRT application and
use, which is necessary for policy makers, corn growers, and SWRT researchers for
evaluation and implementation of this technology.

2

REFERENCES

3

REFERENCES
1.

State, N.a.o.t.U. Water use in the United States. Jan 14, 2013 [cited 2014 Apr
11]; Available from:
http://nationalatlas.gov/articles/water/a_wateruse.html.

2.

Smucker, A.J.M, Corn Yield of 2012 SWRT experiment in Sandhill farm, in 2012
Experiment results. 2012, Michigan State University: East Lansing.

3.

Lyles, L., Sandy surface soils in the United States, in Agricul tural Research
Service. 1975, USDA: Manhattan,Kansas.

4.

NASS, U., Grain stocks, USDA-NASS, Editor. 2014, NASS; Available from
http://usda.mannlib.cornell.edu/usda/current/GraiStoc/GraiStoc-03-312014.pdf

5.

Service, U. N. A. S. (2013). 2013 US total corn grain production_ Survey.
USDA, USDA National Agricultural Statistic Service; available from:
http://quickstats.nass.usda.gov/results/D429F7A1-8B99-3054-AE71C106DA7C3DDB

6.

Kim, S., B.E. Dale, and R. Jenkins, Life cycle assessment of corn grain and corn
stover in the United States. The International Journal of Life Cycle Assessment,
2009. 14(2): p. 160-174.

7.

Kim, S. and B.E. Dale, Life cycle assessment of various cropping systems utilized
for producing biofuels: Bioethanol and biodiesel. Biomass and Bioenergy, 2005.
29(6): p. 426-439.

8.

Sheehan, J., et al., Energy and Environmental Aspects of Using Corn Stover for
Fuel Ethanol. Journal of Industrial Ecology, 2003. 7(3-4): p. 117-146.

9.

Pimentel, D. and T.W. Patzek, Ethanol Production Using Corn, Switchgrass, and
Wood; Biodiesel Production Using Soybean and Sunflower. Natural Resources
Research, 2005. 14(1): p. 65-76.

10.

(NASS), U.-N.A.S.S., USDA michigan specified crops harvested-yeild per acre
irrigated and nonirrigated:2007. 2007, United States department of
Agriculture.
4

11.

Market, F.t. Corn: Environmental Results. Environmental and Socioeconomic
Indicators for Measuring Outcomes of On-Farm Agricultural Production in
the United States: Second Report 2012 [cited 2014 Apr 10]; From the 2012
Environmental and Socioeconomic Indicators Report ]. Available from:
http://www.fieldtomarket.org/report/national-2/PNT_NatReport_Corn.pdf.

5

Literature Review
The current chapter describes the SWRT background and management of soil,
irrigation, and fertilizer in cultivating corn. For prescription agriculture and yield
projection purposes, crop growth and management models are introduced. The
basic framework of an LCA study and an overview of the published LCA corn studies
are reported.
2.1

SWRT background
Temperature, solar radiation, soil, management, and available water are major

factors in crop production. Temperature is a pivotal factor for germination,
respiration and evaporation. Solar radiation satisfies the basic need for
photosynthesis, which is proportional to plant growth when assuming that other
environmental factors are not under stress. Soil, the host to the crops, is a
composition of solid particles, liquids, and air. Different levels of the components in
combination make soil display a wide range of features in texture, water holding
capacity, organic compound level, pH, electronic conductivity (EC), etc. The soil type
has great impact on plant growth in ways that providing growing conditions, such as
acid or alkaline, anaerobic or aerobic environment in the root zone, fertility, and
drainage rate, which affect the health of the crops. Management is a human effect on
plant growth. Management approaches include crop species selection, tillage,
adjusting soil pH, designing row spacing, weeding, fertilizing/ fertigation, and
scheduling dates to plant, manage, and harvest.

6

Among the above crop production parameters, water availability is a critical
factor for crop yield and quality [2]. Soil water deficits rank among the highest
stress limitations to crop growth and productivity. Droughts, limits land
productivity and decreasing fertilizer efficiency, will lead to crop price increase and
aggravate the hunger issue. Irrigation water is responsible for 80% of total water
consumption in the U.S. Due to the wide-spread use of irrigation, underground
water body is experiencing unsustainable withdrawals. If current withdrawal rates
continue, the next worse scenarios would be extensive droughts. A greater demand
for irrigation has to compete with municipal, industrial, and commercial interests
[3], while fresh water declines. In this situation, new technologies that maintain
water and nutrient levels in the soil to achieve high productivity are highly sought,
such as water supplementation, prescription agriculture, water desalination, and
drought-tolerant plants [4-6].
Conventional methods to fight droughts are either to increase water
supplementation or to develop drought-tolerant plants. The former method is a
short-term practice, raising problems such as high labor and resource cost, as well
as increasing leaching of fertilizers, pesticides, manure, and other soil nutrients. The
latter method achieves the goal via gene modification to develop plants with larger
root systems that can absorb sufficient water and nutrients in deeper soil. This
method is time consuming, and requires costly investment for every cropping
system.

7

Another alternative developed by Michigan State University is subsurface water
retention technology (SWRT) [7]. SWRT consists of two components: contoured
engineered membranes made of a blend of linear low density polyethylene (LLDPE)
and low density polyethylene (LDPE), and drip irrigation systems. Membranes are
installed below the root zone with two installation depths- 56 cm and 40 cm from
the soil surface. Each installation depth is used in every other row, as shown in
Figure 2-1. In the long term, iDrip tapes are placed in the root zone above the
membranes. The current practice was to place the pipes on the soil surface parallel
to the linear direction of the SWRT membranes. In order to install the membranes
with the target contoured shape, a new type of tractor was developed for SWRT [7].

8

Figure 2- 1 SWRT configuration
SWRT water saving membranes are contoured engineered LLDPE films strategically
spaced below the plant root zone with space available for root growth and internal
drainage during excess rainfall. Note: (。) Installed drip pipes located 10 cm above
the SWRT water saving membranes.
Experimental plots have successfully been installed and field tested. It is claimed
that the SWRT method helps increase aboveground biomass and food production by
50 to 400%, while reducing irrigation water by 60% at seven experiments in MI and
TX [8]. It is expected that application of SWRT will increase soil water and nutrient
level, and will eventually lead to soil enrichment and conversion to productive soil
[9]. Preliminary studies show that SWRT could transform at least 160 million acres

9

of sandy soils across the USA (3 million acres in Michigan alone) into waterconserving and highly sustainable agricultural production landscapes with minimal
leaching [8, 10]. In addition to greater yields, the long-term use of SWRT has the
potential to increase soil water holding capacity. In response, less irrigation (60%
less in Michigan, U.S.) and nutrient supplements should be required [8].
Furthermore, it is claimed that using SWRT on permeable sandy soil would
effectively reduce global warming impact compared to planting on clay soil.
To date, the potential environmental impact of this technology has not been
studied. On one hand, yield records from the seven SWRT experimental sites suggest
that SWRT helps reduce irrigation, and nutrient and agrochemical run-off [8].
Meanwhile, there are potential benefits by planting on sandy soil to both reduce
greenhouse gas (GHG) emission and effectively use idle sandy soil. On the other
hand, SWRT introduces plastic membranes into the soil, which consumes
nonrewable resources, and requires extra energy for installation. Life cycle
assessment (LCA) is a systematic approach that could evaluate the environmental
footprint of installing and using SWRT, from raw material extraction to the end of
life of the membranes.
2.2

Field management
Corn, as a kind of C4 plant, is the major crop grown in the U.S. with large

harvested area and sales revenue [11]. The average corn yield [kg/ha] grown by
more than 300% since 1942 [12]. This tremendous yield increase should be

10

attributed to agricultural technology development and growing understanding of
field management skills.
Row spacing is a factor affecting yield. A continuous narrowing trend on row
spacing has occurred in modern agriculture. Current row widths employed by
producers typically vary from 38.1 to 96.5 cm, with most producers at 76.2 cm.
Narrow row width has increased appeal for many corn growers. Farnham et al [13]
at Iowa State University Extension conducted a six-year study and found that,
compared to 76.2 cm row spacing, 38.1 cm rows resulted in 0.3% higher yield,
while 96.5 cm rows led to 2.9% lower yield.
After the row spacing decision has been made and seeds have been broadcast,
optimal field managements requires different actions in relation to the corn
development stages. A most commonly used theory to identify corn plant
development stages is Iowa System [14]. Corn development is divided into two main
stages: the vegetative stage and the reproductive stage. Vegetative stage can be
subdivided into emergence, leaves development, and tasseling. The reproductive
stage includes silking, blister, milk, dough, dent, and physiological maturity [14].
Emergence requires soil temperatures above 10°C for germination. Most corn
growers choose the soil temperature as the critical factor to determine the planting
date. Early planting would allow a longer growing season, which may be beneficial
to higher yield. However, if planted too early (e.g., in March in Michigan), soil
temperature will not facilitate corn germination and will increase the seeding
mortality risk. Corn seeds start emergence in 7 to 10 days after broadcasting. Then

11

corn plants typically develop 20-21 leaves in the vegetative stage. About 130 days
after emergence, field corn becomes physically mature. Farmers decide to harvest
field corn based on tassel observation. A yellow or dark brown color silk is a signal
for harvest. A delayed harvest will not result in continuous accumulation of dry
matter. In the opposite, occurs the corn plant will itself consume starch and sugar
for respiration. Farmers decide the date to harvest corn by considering many factors.
The factors include but are not limited to corn production purposes (silage or grain),
corn variety, plant appearance, weather, availability of machines, etc.
2.3

Water management
Water management is part of field management. Despite the higher

infrastructure cost and labor investment required compared to purely relying on
rain feed, water amendment gains increasing popularity due to the significant yield
increases brought by irrigation [15].
2.3.1 Goal of water management
The goal of water management for crop producers is to maintain the field at
its near maximum soil water holding capacity (also called field capacity), to prevent
water stress on plant growth. Irrigated crops have greater economic revenue than
ones relying on rain only. Though only 18% of crop production area is irrigated in
Michigan, the irrigated area produces 23% of the total production value.
Agricultural producers tend to have higher irrigation rates on high value crops, such
as vegetables, potatoes, seed crops, and turf [16].

12

Since agricultural water use takes up to 90 percent of the nation’s water use in
many western states [17], as well as issues such as exhaustive fresh water
withdrawal situations, it is urgent to improve the irrigation water management to
enhance water use efficiency. This can be achieved by controlling the irrigation rate
and frequency, selecting the irrigation method, and retaining water in the soil for a
longer time. Irrigation management should take both economic and environmental
consequences into consideration. Irrigating corn during its water stress sensitive
periods helps to obtain the greatest benefits in the yield. On the other hand,
maximum irrigation rate should be adjusted based on average soil texture, slope,
and rate of residue. Over irrigation not only leads to unnecessary economic cost,
but also causes issues as excessive ponding, runoff, erosion, and an anaerobic
environment for roots.
2.3.2 Water demands in different corn grown phases
Crop water use is the sum of evaporation from the soil surface and plant
respiration [18]. For corn, soil surface evaporation and plant respiration are
responsible for 20 - 30% and 70 - 80% of total crop water use, respectively. Corn
daily water use depends on external conditions and internal factors. Major external
factors are air temperature, humidity, solar radiation, wind speed, and location.
Internal factors are associated with plant growth stages. Especially in summer,
when precipitation is smaller than evaporation, water deficit become a major threat
to crop growth. To eliminate this negative impact on yield, a number of crop

13

growers irrigate their crops. In Michigan, about 10% of corn is irrigated [19], and
this percentage is expected to increase.
Irrigation should be adjusted based on the corn growth phase. Before plant
emergence, surface irrigation is highly discouraged. It can lead to crust formation on
the soil surface and impede the seedlings [18]. In the early N leaf stages, when only a
few leaves have emerged and the root system is tiny, plant water demand is
relatively low due to the limited area to transpire water. Most water use is
attributed to soil surface evaporation during this period. A month and a half after
planting, when corn plants reach the 12-leaf stage, the average daily water use rate
increase to approximately 0.26 in/day, as much as double the amount in the 4-leaf
stage [18]. Normally at 60-80 days after seeding, when corn is in the stages of early
tassel, silking, and blister kernel, daily water use rates reach peak values. Severe
water stress occurring during these reproduction phases will have the greatest
threat to yield. Therefore, in dry years, irrigation at least once during this stage is
highly recommended to offset possible yield loss. Three months after planting, the
corn plant is in the beginning dent stage. Because air temperature begins to cool
down and the root system finishes development, both soil surface evaporation
water and plant respiration water demand decrease. In the next stages, full dent
and maturity stages, the lower leaf drop leads to a continuous shrinking water
demand. From an economic standpoint, most corn growers do not irrigate in the last
two stages, because grain has been developed and water stress will not create
severe consequence in the yield.

14

2.3.3 Soil water
After precipitation enter the soil, the amount of water that can be held by the soil
is determined by the soil water holding capacity. Soil water holding capacity varies
greatly by soil textures, soil organic matter (SOM), and the effective rooting depth of
the crop with the soil. The available water is the difference between the field
capacity and the permanent wilting point (the point at which plants cannot absorb
water from soil and thus die), which depends on pore size and the surface area. For
example, coarse-textured soil (fine sands, loamy sands, and fine sandy loams) is
characterized by a relatively small surface area and weak Vander Waals forces
attributed to large pore size. The water holding capacity of coarse-textured soil is as
low as 0.021 to 0.100 [cm water/ cm soil] (equivalent to 0.25 to 1.2 inches
water/foot of soil), while that of fine-textured soil can be as high as 0.21 [cm water/
cm soil] (equivalent to 2.5 inches water/foot of soil) [20]. At the same time, soil
texture, soil structure, and slope greatly determine the infiltration rate. Slope
determines the tendency for water movement by gravity. Soil texture and structure
determine the available water that can flow into the soil. In general, high clay
content in the soil indicates good water holding capacity. When making irrigation
plans, the effective root depth of the planted crop should also be considered. Some
crops, such as alfalfa and corn, have very strong branching root systems that
penetrate more than 122 cm deep, while a few crops, like soybeans, could not have
effective root depth up to 61 cm [16].

15

The residue cover rate should also be taken into account when scheduling
irrigation. Residue surface coverage can minimize soil evaporation by reflecting a
portion of the incoming solar radiation [18]. Watts and Klocke et al [21]
demonstrated that after a single wetting event, the evaporation rate of bare soil is
higher initially, and decreases as a function of time (days) until the soil water
content is identical to that of residue-covered soil in about 8 days. In reality,
precipitation occurs more frequently than every 8 days for most places during the
growing season. Hence, residue has a barrier effect in reducing total evaporation. In
addition, residue layers are capable of slowing down the speed of surface runoff
[22].
2.3.4 Irrigation approaches
Traditional overhead sprinkler, drip, trickle, furrow, and sub-surface irrigation
systems are commonly used irrigation approaches in the U.S. Center pivot, one of
the overhead sprinkler methods, is the most often used water amendment approach
for corn. Compared to drip irrigation systems, overhead sprinkler irrigation systems
operate with relatively lower water use efficiency. The water use efficiency of
overhead sprinkler is as low as 45%, while that of drip irrigation system is 90%
[23]. A considerable portion of water loss occurs during the delivery of water to
plants for the overhead sprinkler methods. Unlike the drip irrigation method deliver
water directly to roots, reducing excessive evaporation. Less water is required to
deliver to the plants, so reducing electricity consumption and irrigation time. In
addition, drip irrigation water is a good carrier of liquid nutrients and fertilizer,

16

simplifying the application. Drip irrigation also has potential functional benefit in
eliminating crop burning, crop contamination, and plant diseases [17].
A drip irrigation system consists of pumps, filters, drip tubing, valves, regulators,
connectors, ground sensors, and data collectors. Pumps provide the driving force for
the system, either powered by fuel or electricity. In the U.S., electric pumps are
popular on large farms. SWRT employs drip irrigation systems, combined with
membranes to extend the soil water retention time.
2.4

Chemical fertilizers
Fertilizer use in Michigan has increased since commercial fertilizers became

available. At the same time, agricultural areas are continuously shrinking [24].
Investing more money on fertilizers for productive harvest has gained popularity in
Michigan. However, considerable environmental cost and risks should be concerns
associated with over-fertilization. Excessive fertilizer nutrients, especially N and P,
appearing in both surface and groundwater, has become non-point source of
contamination and is hard to control [25, 26].
2.4.1 Nitrogen fertilizer
The N recommendations for most crops grown on organic soils are 45 to 56
kg/ha (40 to 50 lbs. /acre). Little nitrogen replenishment is recommended for
legumes. However, corn, a kind of responsive crop, faces nitrogen limitation most
often, followed by phosphorus. For the purpose of a good benefit return, many corn
producers use more N and P fertilizer than necessary [24].

17

Best management of N fertilizer is estimated based on expected yield and N
credits. A realistic expected yield can be made based on average history under
favorable growing conditions. An unrealistically high yield goal will lead to over
fertilization, raising in both economic and environmental issues. A recommended N
application rate on corn is 112 to 213 kg/ha (100-190 lbs. /acre) [27] depending on
the goal, SOM and N credits. Nitrogen credit sources include NO3-N concentration in
the soil and irrigation water, soil organic matter concentration, manure application,
and legume credit. Detail equations in determine proper N application rate can be
found in corn production guidelines [28].
In addition to the N application rate, choices of nitrogen fertilizer form should be
under consideration based on consideration of soil physics, type of application
method, environment, and plant growth phase. Nitrate, ammonia, and urea are
common forms of N fertilizer. Nitrate N (calcium nitrate or ammonium nitrate) is
readily available for plants, while ammonia and urea require transformation
processes into nitrate N. Their high solubility means a high leaching issue at the
same time. Unless the plant is actively growing, a high nitrate N rate is strongly
discouraged for use on highly permeable soil. For a lower leaching rate, ammonium
is a good substitute. It can incorporate with clay and SOM to delay leaching. Only
high temperature and moist conditions favor the nitrification process to convert it
into nitrate. Urea N is sensitive to volatilization under warm temperatures, high soil
pH, and high humidity. Surface application should be avoided for urea N. Once urea
N becomes gaseous ammonia, plants hardly make use of it, which for crop producers

18

mean losses. To maximize N use efficiency, applying N only a few days ahead of
planting, instead of in early spring, and splitting N applications are encouraged.
2.4.2 Phosphate fertilizer
Phosphate is the second most often occurring nutrient stress for corn growing.
The inorganic forms of phosphorus that can be directly uptaken by plants from soil
are H2PO4- and HPO42-. Precipitation-dissolution and sorption-desorption processes
are the domainant inorganic reactions from P2O5 to phosphate in soil. Plant uptake
will promote both reactions. A common practice to replenish P2O5 fertilizer is via
either broadcasting or banding to incorporate fertilizers in the root zone prior to
planting. Choices of fertilizer, fertilizer spreaders, P concentrate in soil, and
additional manure application are factors to determine P fertilizer rates. For
example, the recommended application rate via the banded method is
approximately half of the broadcasting one. If manure has been applied, a lower P
fertilizing rate should be used. Because corn only responds to low to medium level P,
an excessive amount cannot help yield but incurs the risk of surface water
eutrophication. Depending on the P concentration in sandy soil, 45 to 90 kg/ha (40
– 80 lbs. /acre) P fertilizer is usually applied to corn [27].
2.4.3 Potassium fertilizer
Shortage of extractable K in the soil can cause issues such as leaf
yellowing/browning and lodging during corn two to eight leaves with visible leaf
collars stages. The K2O fertilizing rate usually takes soil K concentration, cation
exchange capacities (CEC), and goal of yield into consideration. For low CEC soil (< 7

19

meq /100 g) sandy soil, a suggested K application range is 80 to 140 lbs./acre [27]
for average corn yield in Michigan (about 150 bushels /acre).
2.5

Corn harvest
In the U.S., the vast majority of corn production is harvested by combine

machines and simultaneously shears the cobs to collect the kernels. At harvest, each
field corn plant typically has developed one cob, with kernels at 30 – 35% moisture
content. For long-term storage purposes, the harvested kernels need to be dried to
13 – 15% moisture content.
Traditionally, corn stover is left behind in the field to cover and protect the soil.
A small percentage of corn farms operate a second-pass run to collect a portion of
the stover for silage and ethanol production. Based on common experience, the total
mass of produced stover is assumed to be a 1:1 ratio of stover to grain. The stover
removal rate needs to be determined to achieve maximum economic benefit, while
keeping local soil erosion losses below the USDA’s tolerable soil-loss limit. Nelson
[29] established a model to estimate the constrained maximum stover removal rate
based on the tolerable soil loss. It considers the rainfall and wind erosion models to
calculate the minimum left behind residue. Dale et al [30] studied the effects of corn
stover removal on soil organic carbon and soil nitrogen dynamics aspects by
running residue removal rate scenarios from 0 to 70% in the DAYCENT model.
2.6

Crop residue utilization
Stover, the major residue left behind in corn harvesting, refers to the above

ground part of maize except grain. Stover takes up about 50% of the total biomass
20

yield. The majority components in stover are stalks and leaves. Stover can be
chopped as fodder for animal feed, which only takes about 5%, while a popular
approach to deal with the recovering of 90% stover is leaving it onsite without
utilization [31]. Even with cellulose ethanol conversion technology development,
less than 1% of corn stover in the U.S. is collected for industrial processes.
The top three chemical components of corn stover by weight are cellulose
(37.7%), hemicellulose (27.5% total, mainly from xylan (21.1%) and arabinan
(2.9%)), and lignin (18.0%) [32]. According to the National Renewable Laboratory
study [32], after corn stover is collected as large round bales and transported to a
nearby plant, it is pretreated with dilute acid to release hemicellulosic sugars. In the
next step, the cellulose polymer will hydrolyze to dissolve sugars into a liquid phase,
leaving mostly lignin in solid form to be removed. Next, the soluble sugar released
from the acid-pretreatment and hydrolysis steps are fermented to ethanol and
purified. Meanwhile, the lignin-rich solid residue is used for steam and electricity
generation via combustion. However, due to the low efficiency of cellulose-ethanol
conversion and high cost of stover baling, storage, and transportation, corn stover
utilization for bioenergy purpose still faces great challenges [32].
While the bioengineering industry is working on corn stover utilization, soil
scientists bring up their concerns on potential negative impacts on soil fertility and
structure from continuous stover continuously removal. Several studies [30, 32]
using CENTURY model-based simulated the stover removal effects on soil, and draw
a consistent conclusion that harvesting corn stover could lower the SOM

21

accumulation rate. But even at the maximum removal rate, which was constrained
to maintain tolerable soil erosion level, the SOM level climbed gently over a period
of decades.
In the opposite case, zero removal of corn stover, covering all corn residues on
the top of soil, there are negative consequences on agronomy. A thick stover layer
covering the surface can delay planting and retard plant development. Because most
of the incident sunshine is reflected, little heat is absorbed by soil to defrost the soil.
The stover layer actslike a barrier, locking most moist with low temperature, and
can retard seedling. An additional consequence is less uniform emergence.
2.7

Tillage
Tillage is an agricultural process for soil preparation. It can loosen the top soil,

mix residue and manure into the soil, and destroy weeds for planting crops. Based
on the tillage extent, tillage systems can be classified into three levels: no-till,
conservation till, and conventional till. A 2010 USDA report estimates that the U.S.
corn cultivation employed conventional till, conservation till and no till at 28.8%,
47.5%, and 23.5%, respectively, in 2005 [33].
Conventional till often involves multiple operations, beginning with implements
such as moldboard, disk, or chisel plows, and ending with harrows to prepare the
seed beds. Conventional till buries more than 80% of residue into deep soil, which is
harmful to the environment [34]. Because these techniques accelerate the
decomposition process of stover and facilitate most of the residue carbon
mineralizing to CO2 emissions. Meanwhile, the deep soil being turned over results in

22

nutrients being vulnerable to loss. In addition, several large machine-pass runs
leave considerable compaction. The negative effect of tilling was determined as
constraining of root development because of hard pan formation and soil aggregates
destruction. So, a growing number of agricultural growers in the U.S. began to
become aware of the economic and environmental benefits from reducing the
number of machine travel times, including fuel and labor saving.
Conservation tillage has less interruption to the soil, with less than 70% of
residue being incorporated into soil, usually by moldboard or disk plows. No till
management, as the name indicates, does not employ any tools to turnover the soil
and incorporate residue.
2.8

Introduction of crop growth and management models
Crop growth and management models can be used to simulate plant growth and

optimize agricultural management approaches for benefit optimization. Previous
biomass production LCA studies [30,32,35,36] employed agricultural models in
their studies mostly for the needs of simulating long-term land use effects and
projections of biomass yields. A group of models are designed to simulate SOM
change and GHG emission, such as CENTURY, DNDC, EPIC, etc. For plant growth
simulation, CERES [37] is a commonly used model. CERES is designed to estimate
the crop growth and development based on the crop species, available light and
temperature. The system approach to land use sustainability model (SALUS) [38] is
one of the integrity models, which can be used to simulate both soil conditions and
plant growth. It is a powerful decision support model, which has been developed

23

into both a PC version and web version. It is capable of helping crop growers to
prescribe their schedules for irrigation, fertilizing, tillage, and harvesting to avoid
unnecessary loss and achieve expected yields.
The SALUS biophysical model is composed of three main components as
illustrated in Figure 2-2: I) growth models for 19 major crop species; II) a SOM and
nutrient cycling model; III) a soil water balance and temperature model.
The crop growth models are derived from the CERES and the International
Benchmark Site Network for Agrotechnology Transfer project family of crop
production models. Therefore, the fundamental crop growth algorithm in SALUS is
identical to CERES. The growth algorithm is governed by variety-specific genetic
coefficients and environmental variables (e.g., degree days, photoperiod). Additional
effort was taken to link the crop growth model with soil water, nutrient and
management submodels in SALUS. Whenever stresses are reported from these
submodels, growth limitation will be posted on carbon assimilation and dry matter
production [38].
The SOM and nutrient cycling models in SALUS are derived from the Century
model with modifications. The Century model was initially designed on a monthly
time frame to simulate carbon pool dynamics. Additional effort was devoted into
daily step simulation in SALUS for an more precise projection. Carbon source were
physically divided into aboveground and belowground carbon in the model. Both
carbon streams can be further traced to structural and metabolic to represent
recalcitrant and easily decomposable residues, based on residue lignin and N
content. Eventually, the entire carbon source will either decompose in the three soil

24

organic matter (SOM) pools (active, slow, and passive) or mineralize to CO2,
depending on their turnover rates and characteristic C/N ratios.
The soil water balance model in SALUS is extracted from CERES and determines
infiltration, drainage, evaporation and runoff. An important modification in SALUS is
that a new concept named time-to-ponding is used to replace infiltration and runoff
calculations [38].

Figure 2- 2 Components of SALUS [39]
2.9

Land use
In field experiments, SWRT demonstrated a desirable effect on the soil because it

maintained soil moisture and promoted soil carbon sequestration. SWRT was

25

initially designed for sandy soil. Compared to clay, sandy soil has 10-times larger
particle size and significantly higher porosity and gas permeability. For this reason,
little agrochemical surface run-off occurs in sandy soil, as well as little chance of
anaerobic denitrification. The anaerobic denitrification process is the critical
process to release N2O. The Intergovernmental Panel on Climate Change (IPCC)
reported that 1 kg N2O is equivalent to 298 kg CO2 in climate change effect. It is
reasonable to propose that planting crops in sandy soil can be an effective solution
to mitigate climate change burdens from agriculture. In other words, SWRT not only
can preserve the biodiversity and value of the land, but also has potential to
improve the sustainability of the agriculture land for future production.
Land use in LCA has been addressed by different approaches [40-42]. For
instance, the LANCA® method has been incorporated into ReCiPe, which is a widely
recognized LCA impact assessment methodology. It covers a variety of aspects that
involve the land occupation area, the duration of occupancy, activities on the land,
land transformation, impact on biodiversity, and soil physics. These aspects are
expressed in erosion resistance, physicochemical filtration, mechanical filtration,
biotic production and ground water replenishment in the LANCA® method. In order
to calculate these values according to the LANCA® method, a few parameters
concerning the land environment and soil physics must be known. Environmental
parameters include average summer precipitation, mean annual temperature,
declination of the land, water and nutrient supply resource, and distance from the
surface to ground water. Soil physics parameters needed for the modeling include
texture of soil, soil organic matter content and soil pH. Texture of soil is defined by
26

the textural triangle, i.e., the percentage of clay (below 0.002 mm), silt (0.002-0.05
mm), and sand (0.05-2.0 mm) [43]. Different soil textures have their own distinct
pore size to capture water and organic compounds. Sandy soil, with approximately
0.1 mm diameter, much larger than loam and clay (below 0.001 mm), has very low
specific surface (no more than 1 or 2 m2g-1 ). Clay’s specific surface is about 100
times higher than sand’s [43]. The specific surface area of the soil is positively
correlated with important phenomena such as cation exchange, adsorption of
various chemicals, and retention of water.
2.10 Overview of life cycle assessment
LCA is a systematic way to evaluate the environmental footprint of products and
systems. It grew from mere energy analysis to a comprehensive environmental
burden analysis in the 1970s. It has been recognized that, for many products, a large
share of environmental burden did not occur during the use phase, but in the
production, transportation, and disposal phases[44]. To evaluate the comprehensive
environmental impact of a product, a full-fledged LCA is required. During 1970s and
1980s, there was a lack of international agreement on the theoretical framework of
LCA. LCAs were conducted with diverging methodologies and terminologies.
Therefore, less agreement in the LCA results presented during these periods can be
found [44], preventing LCA from emerging as a truly scientific subject at that time.
During 1990-2000, life cycle methodology standardization gained remarkable
momentum and growth. Well acknowledged basic steps, principles and
methodologies to estimate carbon footprint (global warming potential) were

27

established, and a number of LCA guides and handbooks were published. A few
standards were established at that time and are still being followed today, such as
the International Organization for Standardization (ISO) 14040 and 14044.
Currently, LCA practitioners and community are exploring broadening the scope of
LCA by developing regional scale impact indicators for land use (land occupation
and land transformation), expanding the LCA analysis beyond limited geographic
boundaries (typically dominated by Europe and North America at present)[44], and
completing water footprint indicators. The modelling phase of LCA has raised
scientific attention as well. Attributional LCA (aLCA), which is merely accountability
of the input and output processes in an LCA study, has dominated LCA studies until
now. With deeper understanding of LCA, the consequential LCA (cLCA) modeling
method, which identifies the displaced product and accounts for the relative
footprint differences due to the marginal change, has raised increasing attention. In
order to accommodate cLCA studies, LCA data providers are changing their database
structure accordingly. The Swiss Center for Life Cycle Inventories, one of the leading
LCA data providers, has recently published their Ecoinvent 3.0 version of the
database to facilitate cLCA modeling [45]. Additionally, LCA scientists are working
on enhancing transparency of LCAs and facilitating LCA collaboration with other
disciplines. LCA is still evolving into a robust methodology. Limitations of
conducting and applying LCA studies’ results still occur due to assumptions, data
gaps, representative technologies, spatial differences, etc., that affect the final result
significantly. Moreover, lengthy time, and huge effort in collecting reliable
inventories are also critical obstacles for popularizing LCA.

28

Today, most LCA studies are comparative aLCA case studies, in which alternative
products typically have a number of distinguishing features; these studies address
many fierce debates, such as in the packaging area: e.g. plastic bags versus paper
bags, or glass bottle versus alternative plastic bottle or aluminum cans. On the other
hand, there is a rising demand for cLCA studies. aLCA and cLCA are modeled in
different ways. In a cLCA, environmental consequences of a marginal change in
demand are assessed, whereas in an aLCA the environmental burdens of a product
are evaluated.
The corn grain and corn stover example is used to illustrate a simplified cLCA
calculation process. Since cLCA is developed from the system expansion method,
they are similar, but mainly differentiated in the products that are being substituted.
To estimate the EFP of 1000 kg corn grain for ethanol production, the system
expansion method is used to solve the co-product (stover) issue in aLCA, while the
cLCA method is applied to estimate the environmental consequence of an additional
1000 kg grain production. In the system expansion method, the stover EFP is
substituted by an equivalent switchgrass EFP, which will be further explained in
section 2.12. In cLCA, assuming the 1000 kg marginal corn grain that used to be
feeding animals is spent on ethanol production instead, to compensate for the
absent of 1000 kg animal-feed corn, 1097.6 kg sorghum is used to feed the animal to
replace corn. If the EFPs of producing 1000 kg corn grain, producing 1097.6 kg
sorghum, and fermented 1000 kg grain to ethanol are 400 kg CO2, 320 kg CO2, and
588 kg CO2 equivalent, respectively, then in aLCA, the EFP of ethanol production
from 1000 kg corn grain is calculated as (400 + 588) kg CO2, while in cLCA, the EFP
29

is equal to (588 + 320 –400) kg CO2. In this case, the EFP calculated from cLCA is
lower than that from aLCA. This is because the absent grain is substituted by a less
burden intensive product, sorghum.
As mentioned above, for many of these products/services, a large share of the
environmental impacts are not attributed to the use phase, but to the raw material
extraction, production, transportation, and the end of life scenarios (landfill,
incineration, recycling). By evaluating the potential environmental impact of the
product system throughout their life cycle, results can be convincing enough to be
accepted for making important decisions. To date, governments all over the world
encourage the use of LCA as a tool for scientific environmental decisions. This
method has become a core element in environmental policy or in voluntary actions
in the European Union, the U.S., Japan, Korea, Canada, Australia, and has been
initiated in booming economies like India and China [44].
2.11 Framework of LCA
A series of ISO standards such as ISO 14040, 14044 and the ILCD series
handbook detail principles and procedures for conducting an LCA. Fundamentally,
the LCA framework consists of a goal and scope definition, inventory analysis,
impact assessment, and interpretation. Conducting an LCA study is an iterative
process between parts. As illustrated in Figure 2-3, to achieve the required precision
with minimum effort, it is recommended to first define the goal, scope, functional
unit (FU) and collect inventory data in an iterative manner.

30

Life cycle assessment

Goal
and scope
definition

Interpretation

Inventory
analysis

Impact
assessment

Figure 2- 3 Stages of an LCA[1]

2.11.1 Goal and Scope Definition
The goal definition is the first phase of conducting an LCA study. A clear goal
definition should identify six aspects: reasons for conducting the project, intended
applications of the delivered results, limitations, target audiences, type of study
(public or internal use), and commissioners of the study [1].
The scope of the project should be defined according to the goal. It defines what
to analyze and how. In scope, qualitative and quantitative aspects of the FU and the

31

reference flows must be specifically documented. The FU is a measure unit of the
output performance for the studied product systems. The FU is the basis for
selecting one or more alternative product systems that provide equivalent services.
It enables fair comparison of different systems to be treated as functionally
equivalent and allows reference flows to be calculated [46]. The FU could be widely
diverging in different case studies. For example, the FU could be defined as 10,000
hours at 600 lumen of light intensity when comparing incandescent bulbs and
fluorescent bulbs [47]; the FU might be “20 m2 of wall covering with a colored
surface of 98% opacity” when comparing environmental impact of various types of
wall paint. Defining the FU requires a good understanding of the studied products or
system. In the previous wall paint example, if the paints’ thermal resistance
property is crucial for users living in cold regions, the original FU is unable to ensure
an equivalent service comparison, and specific thermal resistance criteria should be
included in the FU. Furthermore, if different paints do not stay on the wall for a
same duration, one more criteria should be added, such as service for 5 years. Once
the FU is properly defined, it is feasible to estimate input and output flows and
scaling in relation to the FU. In addition, temporal and geographical boundaries
should be included in the scope section. They define the technological and
geographical representativeness of the inventory. Allocation procedures, impact
categories and methodology selected, data requirements, assumptions, limitations,
and type of critical reviews should also be clear documented in the scope definition.

32

2.11.2 Inventory Analysis
The step of life cycle inventory (LCI) accounts for material and energy resources
consumption and the environmental emissions throughout the lifetime of the
product and/or system. During the LCI step, the data resources used should be
accurately documented. Any assumptions involved with calculation for each unit
process, and work flow of the system model and submodel should be recorded. This
documented information ensures the transparency and robustness of an LCA study.
Many LCA practitioners are aware that different data quality, lack of inventory
completeness, wrong assumptions and differing system boundaries are often
reasons for a lack of agreement between conclusions from LCA projects with the
same studied object.
2.11.3 Impact Assessment
Life cycle impact assessment (LCIA) is the stage where the LCI and the potential
environmental impacts for a product system are evaluated. Impact methodologies
translate inventory data to midpoint or endpoint environmental impact values for
the different areas of protection (i.e., environment, humans, and resources). For
example, 1 kg N2O emission has a climate change effect equivalent to 298 kg CO2. So
if 1 kg N2O emission is stocked in the inventory, the equivalent amount of CO2 is
calculated in LCIA phase. The characterization factor of 286 is obtained from the
impact methodology used. Different impact methodologies might have different
values for the same input of 1 kg N2O. In the previous example, the ReCiPe
Midpoint 1.07(H), CML 2010 and TRACI 2.1 adopt a value of 298, while

33

Impact2002+ sets this value as 156. Therefore, multiple methodologies are
recommended to be used in a study analysis to prevent methodology-resulted bias.
According to ISO 14040 and 14044, normalization, data quality analysis, and
weighting are optional. Due to lack of agreement on weighting methodology,
weighting is not recommended in academic LCA reports.
2.11.4 Interpretation
Interpretation is the last phase of an LCA study. Results calculated from the LCIA
phase are first evaluated for consistency and completeness. And then other analyses
are carried out during the interpretation phase. A series of analyses include:
sensitivity analysis, which estimates the effect of choices made on methods and data;
uncertainty analysis, that quantify the uncertainty due to cumulative effects of
model imprecision and data availability [48]; and contribution analysis, which
investigates the hotspots of impacts. At last, conclusions and recommendations are
formulated in relation to the initial goal and scope.
2.12 Allocation and system expansion methods
In an LCA study, only one product or system is involved out of several products
from the same activity to be quantified the environmental footprint. Partitioning
flows to estimate the impact of the studied product alone is often required in LCA
studies. Allocation and system expansion are traditional solutions to separate the
impacts between products produced by a common system, like corn-stover and corn.
Allocation is the action of partitioning the input or output flows of a process or a
product system between the studied product system and other products occurring
34

from the same activity based on a given ratio. According to ISO 14044, allocation
should be avoided whenever possible by either dividing the unit process into two or
more sub-processes to collect the related sub-process data, or system expansion.
System expansion is defined as subtracting the impacts of co-products from the total.
The ISO standard requires avoiding allocation mainly due to the controversial issues
concerning the partitioning ratio between study products or systems. The
partitioning ratio is often assigned based on a physical relation like energy content,
mass ratio and/or economic value. For example, for an LCA study to quantify the
environmental footprint of producing 1000 kg corn grain, 1000 kg corn stover is
generated as a co-product simultaneously. In an allocation method, the flows are
partitioned between the inputs and outputs going to grain and stover mainly based
on physical relationships, such as mass or economic value relationships between
them. Many LCA studies demonstrate that different choices of partitioning ratio
significantly influence and even alter the final results and can provide a misleading
assessment of the environmental footprint of a product.
Compared with arbitrary defined ratios for allocation, dividing unit processes
and substitution are more convincing in partitioning impacts based on facts.
However, dividing unit processes is not always possible, while the substitution
method is believed to be too complex, hard to determine the displacement of the
non-determining product, and sometimes involves with endless regressions [49].
In the above-discussed corn and stover example, to comply with ISO 14040 and
14044, allocation is not the top candidate method to isolate the impact of the

35

studied product system/ activity. To perform the system expansion method, the
displaced products due to the additional co-product in the market need to be
identified first. Then, an investigation of the quantity relationship between
displaced the product and the co-product is conducted. After that, the LCIA of the
displaced product is calculated. At last, the LCIA of displaced product is assumed to
be equivalent to the co-product, and is deducted from the total studied LCIA.
In the case of applying system expansion in the corn and co-product stover case:
a) producing 2000 kg corn plant (1000 kg corn grain plus 1000 kg corn stover)
produces 500 kg CO2 equivalent emissions was calculated first; b) then switchgrass
ethanol was identified as the displaced product of corn stover; c) after that the
stover-ethanol production efficiency, switchgrass-ethanol production efficiency, and
the displaced quantity relationship were found from the literature; d) the scaling
relation between stover and switchgrass was determined to be that the ethanol
produced from 1000 kg stover was equivalent to the ethanol fermented 375 kg
switchgrass; e) next, the LCIA of 375 kg switchgrass was estimated to be 100 kg CO2
equivalent; f) last, the LCIA of producing 1000 kg stover was substituted by the
switchgrass one (100 kg CO2 equivalent) and removed from the total 500 kg CO2
equivalent. Thus, in this example, the LCIA of producing 1000 kg corn grain is 400
kg CO2 equivalent.
B.P. Weidema, S. Suh et al, and others [49-51]have published a series of papers
to explain detailed procedures of system expansion and a matrix-based approach
towards endless regression issues. This paved the foundational techniques for cLCA.

36

2.13 Past corn studies
Many LCA corn studies [32, 36, 52, 53] have been developed to quantify the
environmental footprint of biofuel compared to petroleum. The initial interest that
drove these studies was: a) biofuel is a promising alternative fuel resource to
address with limited petroleum stock; b) replacing non-renewable resources by
biofuel is a potential way to mitigate the global warning effect; c) new stoverethanol conversion technology has been developed to promote a higher corn crop
utilization rate. Vote et al., [54]conducted a literature review of 67 biofuel LCA
studies published between 2005 and 2010. They found that almost all biofuel
studies involved global warming as an impact indicator. Although results had large
variations, most of the studies concluded that biofuel displayed GHG advantages
compared with fossil alternatives. Contrarily, in the eutrophication and toxicity
categories, biofuel showed worse performance. With respect to acidification and
photochemical smog formation, inconsistent conclusions were drawn in the
different studies. Moreover, other important indicators, such as specific land use
and water depletion, were hardly included in the studies. Only one study of
bioethanol explicitly discussed water consumption and resources [55].The
incomplete conclusions are insufficient to support public policy. The large variation
of different studies’ results is recognized as one of the main LCA limitations. The
divergence in results can derived from differences of system boundaries, studied
geographic locations, key assumptions, allocation methods, and agronomic
processes. Many of them are objective differences among studies, like geographic

37

locations, while a few variables are subjective issues, like allocation methods and
key assumptions.
2.14 Identified reasons of large variations
The FUs used in most corn LCA studies can be divided into two types: type I-the
amount of fuel (biofuel/petroleum) used to drive a unit distance for a certain type of
car; and type II-a unit mass of corn grain or/and corn stover. For example, the FU
can be either service-oriented, defined as the power to wheels for 1 km driving a
midsize car [53]; product-oriented as 1 kg of dry grain and 1 kg of dry stover [31],
or occupied area-oriented as 1 acre of farmland [30]. For studies focusing on the
environmental impact of the biofuel life cycle, the FU is often selected as serviceoriented; for studies aiming to evaluate the environmental performance of
agricultural activity towards the cropping system, the other two types of FUs are
favored. As a consequence of using different FUs, system boundaries are usually
different. For the first type of FU, the system boundary is usually identified from
cradle to grave, i.e., agricultural production of the feedstock, transportation and
conversion of stover-based ethanol, fuel use phase, and waste management. For the
product-oriented and occupied area-oriented types of FUs, their system boundary is
normally from cradle to farm gate. A key assumption is that after the farm gate,
products from different treatments have equivalent functions, so the following
phases after leaving the farm gate could be left out in comparative studies.
One of the reasons for large variations between the different LCA studies about
bioethanol derived from differences in the studied locations and temporal scopes.

38

Most geographic locations in corn LCA studies are either Europe or the U.S. (mainly
within the Corn Belt states which produce over 80% of the total corn in the U.S. [56],
especially within Iowa [30-32, 53]). Different geographic regions represent different
soil physics, climate, and favorite tillage methods. Due to data-availability
limitations, only a few studies allow investigating the effects of location [31].
Additionally, temporal scope of yield records varies significantly. Dale et al [31]
used 4-year average values between 2000 and 2003 for yield and agronomic input
data, while Sheehan et al [32] employed corn yield reports from 1995-1997 and
inconsistently used climate records of 1961-1991. The inconsistency in temporal
scope may involve inventory errors due to two reasons: (1) the average corn yield
continues to increase with the advancement of agricultural technology. According to
the USDA National Agricultural Statistics Services report [12], the average yield has
increased from 100 bushels/acre to 160 bushels/acre since 1983. (2) The weather
conditions vary significantly year-to year, which affect decision making about
agronomic inputs and yields dramatically. The yield increase dilutes the
environmental footprint of a cropping system largely by spreading them over more
outputs. For this reason, it is believed that yield and yield change are highly
sensitive parameters in crop related LCA studies [57]. In addition, the time and
geography representation of the yields and weather data should be matched and
well documented.
Another large source of variation is the great uncertainty of the inventory for
stover-based ethanol conversion. Most reported studies are constructed on a
projected inventory of advanced stover-ethanol conversion technology. To the best

39

of the author’s knowledge, few records of commercialized stover-conversion
technology were available when this work was published. DuPont launched one of
the first and largest commercial biorefinery in the world to produce fuel from
cellulose in 2012 [58]. Another key factor resulting in less agreement in different
studies is the stover removal rate without additional fertilizer supplement. The
stover removal rate (SRR) is a factor to determine the stover share of total corn crop
environmental impacts. It varies from 0 to 70% among corn LCA studies. Some
studies[30, 31, 53] defined SRR based on arbitrary assumptions or so-called a
dominant represented value, while one study [32] set up a constraint on tolerable
soil loss to model the maximal stover collection rate.
There are different treatments and blurred descriptions of biogenic carbon
among biofuel LCA studies. In theory, all relevant interventions must be included. In
biofuel LCAs, plants capture CO2 from atmosphere and release equivalent CO2 when
they burn. Therefore, a common simplified method is to feature the biofuel chain as
“carbon neutral” to avoid unnecessary errors by not accounting for capture of CO2
and burning emission CO2 interventions [54]. In the IPCC methodology, when
calculating GHG, both included biogenic CO2 and excluded biogenic CO2 approaches
were developed, while a few methodologies, such as ReCiPe 1.07, default to exclude
biogenic carbon. However, if the system boundary of studies is from cradle to farm
gate, which does not involve fuel burning, biogenic carbon should be included.
Furthermore, Luo et al [53] demonstrate that differences in results might be hidden
if a method that excludes biogenic carbon is used. Especially in cases of large
economic value differences between the determined product and co-products,

40

including the biogenic CO2 makes a large difference since credits for extracted CO2
are allocated to different parts of a multiproduct-system than the debits for CO2
emissions.

41

REFERENCES

42

REFERENCES
1.

Commission, E., International Reference Life Cycle Data System (ILCD)
Handbook—general guide for life cycle assessment—detailed guidance. Joint
Research Centre—Institute for Environment and Sustainability. Publications
Office of the European Union, Luxembourg, 2010.

2.

Thorne, D.W. and M.D. Thorne, Soil, water and crop production. . 1979: THE
AVI PUBLISHING COMPANY, ING. Westport, Connecticut. 355.

3.

Smucker, A.J.M, et al., Membrane Install That Double Soil Water Holding
Capacity in Highly Permeable Soils. 2012, Michigan State University, USDAARS, Lubbock, TX, Burnham Soil and Water Management, Ithaca, MI.

4.

Owens, W.A., Contrasting Yield, Irrigation Water Use Efficiency, and Economics
of Center Pivot and Subsurface Drip Irrigation Systems for Corn Production in
the Southeastern Coastal Plain. 2013.

5.

Basso, B., A.D. Kendall, and D.W. Hyndman, The future of agriculture over the
Ogallala Aquifer: Solutions to grow crops more efficiently with limited water.
Earth's Future, 2013.

6.

Cohen, Y., et al., Membrane desalination of agricultural drainage water, in
Salinity and Drainage in San Joaquin Valley, California. 2014, Springer. p. 303341.

7.

Smucker, A.J.M, Subsurface Barrier Retention System and Methods Related
Thereto. 2011, Google Patents.

8.

Smucker, A.J.M, Subsurface Water Retention Technologies (swrt) Convert
Marginal Sandy and Other Excessively Permeable Soils into Highly Sustainable
Food and Fiber Production Systems while Improving Soil Fertility,
Csequestration, Ecosystem Services and Biodegradation of Toxins Resulting in
Sustainable Economic Growth and Environmental Protection, in White paper
with CEPEM 9 26 12. 2012, Michigan State University: East Lansing.

9.

Smucker, A., Zeyuan Yang , Xuechen He, Fuhua Lai, Bruno Basso, SWRT is a
New Technology for Converting Arid Sands into Oases of Food and Fiber
Production. International Journal of Biobased Materials and Bioenergy, 2012.

43

10.

Smucker, A., G. Schultink, and K. Thelen, SWRT Enhancement of Water Use
Efficiency and Economics of Irrigated Low Water Holding Capacity Sandy Soils,
in 2009 Annual Meetings of the American Society of Agronomy, Crop Science
Society of America and Soil Science Society of America (ASA-CSSA-SSSA 2009).
2009.

11.

(NCGA), N.C.G.A., US Corn Production 2013. 2013, National Corn Growers
Association: Washinton DC. p. 20.

12.

NASS. U.S. corn yield statistics from year 1983 to year 2013 by USDA National
Agricultural Statistics Service. 2013 [cited 2013 Nov.].

13.

Lori Abendroth, R.E. What row spacing is best? 2006 [cited 2013 Sep 15];
Available from:
http://www.agronext.iastate.edu/corn/production/management/planting/r
ow.html.

14.

Heather Darby, J.L., Plant Physiology Critical Stages in the Life of a Corn Plant.
Diagnostic Section 3: p. 17.

15.

Perlman, H. Irrigation water use. 2005 Mar 17, 2014 [cited 2014 Apr 1];
Available from: http://water.usgs.gov/edu/wuir.html.

16.

Development, M.D.o.A.R., Generally Accepted Agricultural and Management
Practices for Irrigation Water Use, M.C.o.A.R. Development, Editor. 2013,
www.michigan.gov/mda: lansing.

17.

Glenn Schaible, M.A. Irrigation & Water Use. 2013 Friday, June 07, 2013
[cited 2013 Sep.24].

18.

Willian L. kranz, s.I., Simon J. van Donk, C.Dean Yonts, Derrel L. Martin,
Irrigation management for corn. 2008, Division of the Institute of Agriculture
and Natural Resources at the University of Nebraska: Lincoln. p. 4.

19.

(NASS), U.-N.A.S.S., USDA michigan specified crops harvested-yeild per acre
irrigated and nonirrigated:2007. 2007, United States department of
Agriculture.

44

20.

Covert, D.M.N. Plant & soil sciences eLibrary. Soils - Part 2: Physical Properties
of Soil and Soil Water 2011 [cited 2013 Dec.10]; soil water holding capacity
resource]. Available from:
http://passel.unl.edu/pages/index.php?contactus=1.

21.

S. J. van Donk, D.L.M., S. Irmak, S. R. Melvin, J. L. Petersen, D. R. Davison, Crop
residue cover effects on evaporation, soil water content, and yield of deficitirrigated corn in west-central Nebraska. American Society of Agricultural and
Biological Engineers, 2010. 53(6): p. 10.

22.

van Donk, S., et al., Crop Residue Cover Effects On Evaporation, Soil Water
Content, And Yield Of Deficit‐Irrigated Corn In West‐Central Nebraska. 2010.

23.

Capra, A. and B. Scicolone, Water Quality and Distribution Uniformity in
Drip/Trickle Irrigation Systems. Journal of Agricultural Engineering Research,
1998. 70(4): p. 355-365.

24.

Development, M.C.o.A.a.R., Generally Accepted Agricultural for nutrient
management, M.C.o.A.R. development, Editor. 2013, www.michigan.gov/mda:
Lansing.

25.

Power, J. and J. Schepers, Nitrate contamination of groundwater in North
America. Agriculture, ecosystems & environment, 1989. 26(3): p. 165-187.

26.

Carpenter, S.R., et al., Nonpoint pollution of surface waters with phosphorus
and nitrogen. Ecological applications, 1998. 8(3): p. 559-568.

27.

M.L. Vitosh, J.W.J., D.B. Mengel. Tri-State Fertilizer Recommendations for Corn,
Soybeans, Wheat and Alfalfa. [cited 2014 Feb 4]; Available from:
http://ohioline.osu.edu/e2567/.

28.

J.G. Davis, D.G.W. Fertilizing Corn. 2014 [cited 2014 Jan 20]; Available from:
http://www.ext.colostate.edu/pubs/crops/00538.html.

29.

Nelson, R.G., Resource assessment and removal analysis for corn stover and
wheat straw in the Eastern and Midwestern United States—rainfall and windinduced soil erosion methodology. Biomass and Bioenergy, 2002. 22(5): p.
349-363.

45

30.

Kim, S. and B.E. Dale, Life cycle assessment of various cropping systems utilized
for producing biofuels: Bioethanol and biodiesel. Biomass and Bioenergy, 2005.
29(6): p. 426-439.

31.

Kim, S., B.E. Dale, and R. Jenkins, Life cycle assessment of corn grain and corn
stover in the United States. The International Journal of Life Cycle Assessment,
2009. 14(2): p. 160-174.

32.

Sheehan, J., et al., Energy and Environmental Aspects of Using Corn Stover for
Fuel Ethanol. Journal of Industrial Ecology, 2003. 7(3-4): p. 117-146.

33.

Horowitz, J., No-till farming is a growing practice. 2010: DIANE Publishing.

34.

Basso, F., M. Pisante, and B. Basso, 25 Soil Erosion and Land Degradation.
Mediterranean desertification: A mosaic of processes and responses, 2003: p.
347.

35.

Feng, H., O.D. Rubin, and B.A. Babcock, Greenhouse gas impacts of ethanol
from Iowa corn: Life cycle assessment versus system wide approach. Biomass
and Bioenergy, 2010. 34(6): p. 912-921.

36.

Seungdo Kim, B.E.D., Life Cycle Assessment to Improve the Sustainability and
Competitive Position of Biobased Chemicals: A Local Approach. Jan 2009,
Michigan State University: Lansing. p. 1-160.

37.

Ritchie, J.T., et al., A User's Guide to CERES Maize, V2. 10. 1992: International
Fertilizer Development Center.

38.

Basso, B., et al., Simulation of tillage systems impact on soil biophysical
properties using the SALUS model. Italian Journal of Agronomy, 2010. 1(4): p.
677-688.

39.

Ritchie, J.T. SALUS MODEL. [cited 2013 Nov.6]; Available from:
http://nowlin.css.msu.edu/salus/overview.html#N_1_.

40.

Weidema, B.P. and E. Lindeijer, Physical impacts of land use in product life
cycle assessment. Technical University of Denmark, 2001.

46

41.

Mattila, T., T. Helin, and R. Antikainen, Land use indicators in life cycle
assessment. The International Journal of Life Cycle Assessment, 2012. 17(3):
p. 277-286.

42.

Dipl.-Geoökol. Tabea Beck, D.-G.U.B., Dipl.-Ing. Bastian Wittstock, Dr.-Ing.
Martin Baitz, Dipl.-Ing. Matthias Fischer, Prof. Dr.-Ing. Klaus Sedlbauer,
LANCA-Land Use Indicator Value Calculation in LCA-Method Report,
Fraunhofer Institute for Building Physics Department Life Cycle Engineering:
Stuttgart.

43.

Daniel, H., Environmental soil physics. 1998: Academic Press.

44.

Jeroen B. Guinee, R.H., Gjalt Huppes, Life Cycle Assessment: Past, Present, and
Future. Environ. Sci. Technol., 2010. 2011(45): p. 90-96.

45.

(SCLCI), S.C.f.L.C.I. Introduction to ecoinvent Version 3 for Existing Users. 2014
[cited 2014 Mar 6th]; Available from:
http://www.ecoinvent.org/database/ecoinvent-version-3/introduction/.

46.

Guinée, J.B., et al., Handbook on life cycle assessment. Operational guide to the
ISO standards, 2002: p. 1-708.

47.

Auras, R.A., PKG891-LCA: Background, Principles, Calculations, and
Applications, in Charpter III-Beginning an LCA. 2012, Michigan State
University: East Lansing. p. 28.

48.

Standardization, I.O.f., International standards Stadard 14040 only, in
Environmental management- Life cycle assessment-Principles and framework.
2006: Switzerland.

49.

Weidema, B.P., Market information in life cycle assessment. Vol. 863. 2003:
Miljøstyrelsen.

50.

Weidema, B., Avoiding Co‐product Allocation in Life‐Cycle Assessment. Journal
of Industrial Ecology, 2000. 4(3): p. 11-33.

51.

Suh, S., et al., Generalized Make and Use Framework for Allocation in Life Cycle
Assessment. Journal of Industrial Ecology, 2010. 14(2): p. 335-353.

47

52.

Miller, S.A., A.E. Landis, and T.L. Theis, Feature: Environmental trade-offs of
biobased production. Environmental Science & Technology, 2007. 41(15): p.
5176-5182.

53.

Luo, L., et al., Allocation issues in LCA methodology: a case study of corn stoverbased fuel ethanol. The International Journal of Life Cycle Assessment, 2009.
14(6): p. 529-539.

54.

van der Voet, E., R.J. Lifset, and L. Luo, Life-cycle assessment of biofuels,
convergence and divergence. Biofuels, 2010. 1(3): p. 435-449.

55.

Mishra, G.S. and S. Yeh, Life Cycle Water Consumption and Withdrawal
Requirements of Ethanol from Corn Grain and Residues. Environmental
Science & Technology, 2011. 45(10): p. 4563-4569.

56.

NASS. U.S. Corn production by county in year 2012, prepared by National
Agricultural Statistics Service. 2012 [cited 2013 Nov]; corn 2012 production
statistics by county].

57.

Fukushima, Y. and S.-P. Chen, A decision support tool for modifications in crop
cultivation method based on life cycle assessment: a case study on greenhouse
gas emission reduction in Taiwanese sugarcane cultivation. The International
Journal of Life Cycle Assessment, 2009. 14(7): p. 639-655.

58.

Koninckx, J. Biorefinery Beckons: Plant will produce ethanol from corn stover.
2013 [cited 2013 Oct]; DuPont will estabalish one of the first and largest
commercial biorefineries in the world making fuel from cellulose]. Available
from: http://www.chemicalprocessing.com/articles/2013/biorefinerybeckons-plant-will-produce-ethanol-from-corn-stover/.

48

Goal and Scope
This chapter defines the purpose of the study, intended use of the delivered
results, and target audience. To enable environmental comparisons, the scope is to
define the boundary of this attributional LCA study. The motivation of the studied
system is to investigate the environmental performance of producing corn grain on
Sandhill farm using SWRT. Sandhill farm is an experimental plot located south to
Michigan State University main campus (East Lansing, Michigan, US).
Purpose of StudyThe primary purpose of this study is to evaluate the environmental
footprint of using SWRT and irrigation to grow corn in Michigan. The four
combination agronomic scenarios were: i) SWRT with irrigation, ii) SWRT without
irrigation, iii) non-SWRT with irrigation, and iv) non-SWRT without irrigation. The
study aims compare the environmental footprints between SWRT experimental
treatments and control treatments. Another purpose of this study is to provide
recommendations on planning agronomic management based on the studied results.
3.1

Intended use of study
The primary use of this study is to establish a baseline for the SWRT

environmental performance. The study is a purely methodological study without
relationship to decision support and accounting/monitoring of the study’s object.
The findings can help SWRT researchers identify hotspots of environmental burden
associated with corn growth on sandy soil. In addition, the study can provide
guidance for agricultural researchers and offer advises to farmers about planning
corn growing with a preferable environmental performance.

49

3.2

Target audience
The primary target audiences for this study are SWRT researchers, corn growers,

and policy makers. The interest parties will be mainly the agricultural industry,
plastic industry, soil scientists, agricultural hydraulic engineers, and governments
responsible for agricultural land management. The study itself is mainly intend for
academic purposes, not aimed at a public comparative assertion.

50

3.3

Functional unit
The functional unit is 1000 kg of corn grain grown in sandy soil in MI, US. (dry

weight, with 15.5% water content).
3.4

System boundary
This study focuses on the comparison of the corn production process involved

with SWRT and irrigation technologies. A cradle to farm gate system boundary was
selected for corn production comparison, reducing the chance to incorporate
potential errors and reducing noise by avoiding the steps after leaving the farm gate.
A cradle to grave boundary was used for the SWRT membrane.
For the corn production life cycle, the cradle here refers to raw material
extraction, such as mining for further fertilizer production. The farm gate refers to
the farm where the corn is harvested. For the membrane life cycle, the cradle is
petroleum extraction for resin production, and the grave means the end of life (EOL)
of the membrane. The whole studied system includes 14 major processes, as shown
in Figure 4-1. These processes are membrane manufacture, membrane installation
by tractor, twice conservation tillage (before planting and after harvest), seed
production, sowing, irrigation, combine harvesting, stover removal, and five
agrochemical application processes (nitrogen, phosphate, potassium fertilizers,
glyphosate and ammonium sulfate).

51

[1] Membrane
manufacture

[7] N fertilizer
[8] N fertilizer

[2] Membrane
installation
[3] Tillage

[14] Stover removal

[9] K fertilizer

Field production

Harvest

[6] Irrigation

[5] Sowing
[5] Seed
production

[10] Rdup
[11] AMS

[12] Combine
harvest
[13] Tillage

Figure 3-1 System boundary

3.5

Cut-off rules
Due to the extensive number of inputs and outputs in this LCA study, flows that

contribute less than 1% of the cumulative mass and the total environmental
footprint were excluded. In some cases, small amounts of certain components have
great impacts on certain impact categories (such as sulphur hexafluoride and N2O
on climate change). Therefore, the initial identification of inputs and outputs were
selected primarily based on mass, but also keep cautious to a group of components
with high environmental impacts.
3.6

Allocation rules
In accord with the functional unit ( 1000 kg corn grain in this study), the

environmental impacts of grain production should be isolated from the whole plant

52

growing process. System expansion was conducted to partition impacts to grain and
stover productions. For a comprehensive understanding, allocation methods based
on mass, economic values, and energy contents were used as well. Mass allocation is
the default allocation method to solve the co-products or multifunctionality issues.
3.7

Temporal and technology representative
The time representative for most technologies is the most recent 10 years,

except for SWRT, a pilot technique, where it is the post 5 years. The study complies
with the ISO 14040 and 14044 standards [1, 2].
3.8

Software and data collection
This LCA study was mainly conducted using GaBi 6 professional version from PE

International Company (Stuttgart, Germany). A part of the system expansion work
was conducted in Matlab Version R2010b (Mathworks, Natick, MA, U.S.). The study
mainly consists of two parts: experimental treatments and model-based simulations.
The data used in evaluation of SWRT experiments were collected from primary
results. And the model-based simulation part partially employed primary data such
as fuel cost of membrane installation, and irrigation construction, while it used
simulated data for corn yield and water amendment from simulation using the
SALUS model Version 1.0 (Alpha). Detailed information regarding data sources can
be found in chapter 4 Life cycle inventory.
3.9

Life cycle impact assessment methodology and impact categories
To quantify the environmental impact from inventories, the midpoints for

ReCiPe 1.07(H) were mainly used. A mixture of impact assessment methodologies

53

including CML2001 - Nov. 2010IPCC, TRACI 2.1, and IPCC were used for benchmark
comparisons with published corn LCA studies. The following six impact categories
were mainly assessed: a) climate change, b) water depletion, c) non-renewable
resource depletion, d) land use, e) aquatic eutrophication, f) acidification, g)
ecotoxicity, and h) human toxicity. They were selected by following LCA guidelines
for grains[3] that discusses the important impact categories for LCA grain studies.

54

REFERENCES

55

REFERENCES
1.

Standardization, I.O.f., International standards Stadard 14040 only, in
Environmental management- Life cycle assessment-Principles and framework.
2006: Switzerland.

2.

Standarization, I.O.f., International standars Standard 14044 only, in
Environmental management- Life cycle assessment-Requirements and
guidelines. 2006: Switzerland.

3.

Mia Lafontaine, François Charron-Doucet, and E. Clément, Production of
Pulses, Grains and Oilseeds - A guide for LCA Practicioners, in LCA guidelines
for pulses, grains and oilseeds, Quantis, Editor. 2013, Pulse Canada. p. 42.

56

Life Cycle Inventory
4.1

Project design
The Life Cycle Inventory (LCI) was created as two dataset to be analyzed: a) an

experimental dataset and b) a simulated dataset. The experimental dataset was
created from the data obtained from the inputs and outputs of the unit processes of
growing corn in the experiment run in the Sandhill farm. The simulated dataset was
created from running corn growth scenarios in the SALUS simulation software [2].
Detailed LCI inventories of both the experimental and simulation dataset are
provided in Appendix -LCI section 1 LCI of the Sandhill farm experiment.
4.1.1 Experimental dataset
Field corn was grown in the Sandhill farm located in East Lansing, MI in 2012
and 2013. The soil texture on Sandhill farm is 96.1% sand and 3.9% silt and clay.
The total experimental site area was 65.580 m * 27.432 m (N-S * E-W) = 1,881.287
m2. The total area was divided into five plots. Treatments were randomly assigned
to each plot.
In 2012, two factors were studied: using SWRT and row spacing. Each factor had
two levels: SWRT, without SWRT, 0.381 m row spacing, and 0.762 m row spacing.
Due to the unusually low precipitation that year, to keep the corn plants alive,
additional water supplement was applied to all four treatments at different levels,
which is described in Table 5-7. In 2013, two factors were studied using SWRT and
irrigation. Theoretically, each factor had two levels: SWRT, without SWRT, and
irrigated, nonirrigated. The 2013 experiment was an incomplete design experiment,

57

so the treatment irrigated SWRT and nonirrigated without SWRT were the only
ones conducted. The yields and irrigation water consumption for 2012 and 2013 are
provided in Tables 4-1 and 4-2, respectively.
Table 4- 1 Yield and irrigation for 2012 experiment
SWRT
Irrigation

Row spacing
Yield
m
[kg/ha]
m3/acre
0.381
16830±1256
1989
0.762
13376±1319
1070
* Yield [kg/ha]: mean ± stand error

No SWRT
Irrigation

Yield
[kg/ha]
9671±2135
9921±1444

m3/acre
583
269

Table 4- 2 Yield and irrigation for 2013 experiment
SWRT
Yield
[kg/ha]
11981±359

Irrigated
Nonirrigated
* Yield [kg/ha]: mean ± stand error

No SWRT
Irrigation Yield
m3/acre
[kg/ha]
765
3544±957

Irrigation
m3/acre
0

4.1.2 Simulation data
Since the studied factors are different between years 2012 and 2013, and the
studied factors were monitored for only a single year, drawing conclusions based
simply on year 2012 and 2013 experiment datasets has its limitations.
Yield is a data with large year-to-year variations because of weather changes (i.e.,
temperature, precipitation, radiation, wind speed, etc.). So, a short study duration
decreases the representativeness of experiment results to some extent.
Furthermore, year 2012 was a special dry year, which decreased its

58

representativeness of average results. An obvious drop in corn grain average yield
in Ingham county, MI was observed in 2012 as illustrated in Figure 4-1, which
further ratifies the assumption about being a dry year. Treatments were irrigated at
different levels. So, it is difficult to isolate the water irrigation effect from the studied
factors. In addition, the incomplete experiment design of 2013 caused trouble in
drawing conclusions towards two of the studied factors: irrigation and SWRT.

180

grain yield (Bu/a)

160
140
120
100
80
60
40
20
0

1942

1952

1962

1972
1982
year

1992

2002

2012

Figure 4- 1 Average grain yield from 1942-2012 in Ingham County, MI [1]
Based on the above reasons, the need for data simulation emerged. The
simulated dataset can overcome the limitation of short study periods in agricultural
experiments. At the same time, long-period data would eliminate the disturbance of
years with extreme weather conditions.

59

The SALUS model was used to simulate corn yields in Sandhill farm from 2004 to
2013. Four specific input files were created for the SALUS simulation scenarios: crop
file, field management file, soil texture file, and weather file 2004 to 2013.
A crop file was created based on a pre-created library with more than 18
common crops in the U.S. From the SALUS simulation library, the studied crop
species was selected.
The soil texture file required information regarding the lower limit of soil water,
drainage in the upper limit of soil, the saturated soil water content, soil hospitality,
saturated hydraulic conditions, bulk density, organic carbon content, clay content,
and silt content. The soil texture file was created based on the Sandhill farm soil
texture, as presented in Appendix-LCI Table A-4.
The weather file for the SALUS simulation was obtained from daily weather
records of the MSU Horticulture Teaching and Research Center (1.8 miles from
Sandhill farm), which was available in the Michigan Automated Weather Network
[3]. The weather records of air temperature, precipitation, solar flux density, and
wind speed from 2004 to 2013 were used to create the weather file.
The crop management parameter settings in the SALUS model are mostly
derived from experimental records. A series of typical inputs are plant population,
row spacing, dates and type of tillage, planting and harvesting. Fertilizer types and
application levels, important inputs for both the SALUS model and LCA studies, are
decided based on expert opinion for Michigan [4]. The crop management settings
can be found in Appendix-LCI Table A-5.

60

To select the plant growth and water balance model in SALUS, the output yields
and irrigation water were used for the LCA study of the simulation treatments. Two
levels of irrigation factor, irrigated and non-irrigated, were simulated in the SALUS
model. Because only about 10% of corn is irrigated in Michigan, the aggregated
dataset used in the second part of the LCA evaluations to represent corn grown in
Sandhill farm was modelled as 10% irrigated and 90% non-irrigated. Appendix-LCI
Table A-7 provides the irrigated yield, non-irrigated yield, and aggregated yield
from the SALUS simulation, and average yields of the U.S., Michigan, and Ingham
County retrieved from the United States Department of Agriculture (USDA) National
Agricultural Statistics Survey.
4.2

Overview

The life cycle inventory (LCI) step of an LCA study documents the data collection
and calculation processes, and keeps records of inputs and outputs of the whole
system. Input and output quantities are adjusted according to the functional unit

61

(FU) of the study. The corn grain production system is shown in Figure 4- 2.
Seed

(SWRT)
Energy

Nutrien

Water

Corn Grain
Production System

Grain production

Agricultural machinery

Emissions Emissions Solid
to water
to air
wast
e systems
Figure 4- 2 Corn grain production
There are 28 unit processes within the whole system. To simplify the analysis,
they were grouped into 6 groups as shown in Table 4-3. The following section
introduces the LCI by groups.

62

Table 4- 3 Unit process grouping for corn production system
Group
Machinery

Chemical

Irrigation

Seed
SWRT

Planting

4.3

Unit processes
Harvesting
Fertilizing
Sowing
Tillage
Gypsum
Roundup
AMS
K2O
P2O5
Urea
AN (liquid)
KCl (liquid)
irrigation water
drip irrigation
Electricity
drip tape
pipe landfill
pipe incineration
pipe recycle
Seed
PE membrane
membrane install
tillage machine production
tractor production
Diesel
Shed
Field Preparation
Harvest

Machinery
The machinery group includes plans of tillage, sowing, broadcast fertilizing, and

combine harvesting. Chisel tillage machinery was used twice per growing season.
The first use was several days prior to sowing seeds, and the other time was applied

63

after harvesting. The remaining three machinery activities were modeled on the
field once per growing season.
In general, each machinery plan was controlled by a central unit process. It fixed
the reference flow, 10000 m2 machinery work flow output in this case, to manage
the quantity of inputs and outputs of the connected processes. The input processes
that were connected to the reference process can be categorized into three types: i)
machine production (including tractor and agricultural machinery), ii) shed, the
solid construction of building for machinery production, iii) diesel, the primary
energy for operating the machine.
4.3.1 Tillage
The tillage unit process was a cradle to farm gate plan. The output of the unit
process was 1 ha farm being chisel tilled. The model of the chisel tillage process is
illustrated in Figure 4-3 below.
CH: shed

US: diesel, at refinery
1 ha farm being
chisel tilled

CH: tractor production
CH: agricultural machinery,
tillage, production

CH: tillage, cultivating, chiseling

Figure 4-3 Tillage process model

The tillage plan was developed from the Ecoinvent Data v2.2 (Swiss Centre for
LCI) unit process named “tillage, cultivating, chiselling, single operation unit process
(u-so)”, whose detailed LCI is listed in Table 4- 5. The primary energy supply,

64

Switzerland diesel, was substituted by U.S. diesel. The purpose of primary energy
substitution is to localize the original Switzerland tillage process to the U.S. The
remaining three processes were obtained from Ecoinvent Data v2.2 aggregated unit
process (agg) without further change. The temporal representation, geography
representation, and key assumptions are listed in the Table 4- 4.
Table 4- 4 Representation of tillage activity with primary US energy substitution
Process
Temporal Geography
Key assumptions
agricultural machinery,
1995800 kg machine, 800 hours
tillage, production
2002
Switzerland
useful life, 0.72 repair factor
19953000 kg machine, 7000 hours
tractor, production
2002
Switzerland
useful life, 0.74 repair factor
dataset was built on one
example of a typical
1994agricultural building, 50 years
shed
2002
Switzerland
lifetime
diesel, at refinery
2009
U.S.
US LCI dataset
tillage, cultivating,
1991chiselling
2002
Switzerland
working width 2.5 m

65

Table 4- 5 Input /Output flows of chisel tillage 1 ha farm with primary US energy
substitution
Standard
Flow
Amount
Unit deviation
Input
CH: agricultural machinery, tillage, production
[Machines]
1.48 kg
111%
US: diesel, at refinery[fuels]
15.5 kg
111%
CH: shed [buildings]
0.00573 m2
301%
CH: tractor, production [Machines]
0.883 kg
111%
Output
CH: tillage, cultivating, chiselling
Waste heat [Other emissions to air]
Carbon dioxide [Inorganic emissions to air]
Nitrogen oxides [Inorganic emissions to air]
Carbon monoxide [Inorganic emissions to air]
Dust (PM2.5) [Particles to air]
NMVOC (unspecified) [Group NMVOC to air]
Sulphur dioxide [Inorganic emissions to air]
Methane [Organic emissions to air (group VOC)]
Nitrous oxide (laughing gas) [Inorganic emissions to
air]
Zinc (+II) [Heavy metals to agricultural soil]
Ammonia [Inorganic emissions to air]
Benzene [Group NMVOC to air]
Polycyclic aromatic hydrocarbons (PAH) [Group
PAH to air]
Copper (+II) [Heavy metals to air]
Zinc (+II) [Heavy metals to air]
Lead (+II) [Heavy metals to agricultural soil]
Nickel (+II) [Heavy metals to air]
Chromium (unspecified) [Heavy metals to air]
Benzo [5]pyrene [Group PAH to air]
Cadmium (+II) [Heavy metals to agricultural soil]
Cadmium (+II) [Heavy metals to air]
Selenium [Heavy metals to air]

66

10000
705
48.3
6.08E-01
1.30E-01
8.28E-02
0.0302

m2
MJ
kg
kg
kg
kg
kg

0%
111%
121%
152%
501%
305%
152%

0.0156 kg
2.00E-03 kg

121%
156%

1.86E-03
0.00112
3.10E-04
0.000113

kg
kg
kg
kg

156%
152%
156%
156%

5.11E-05
2.64E-05
1.55E-05
1.85E-06
1.09E-06
7.76E-07
4.66E-07
4.24E-07
1.55E-07
1.55E-07

kg
kg
kg
kg
kg
kg
kg
kg
kg
kg

305%
505%
505%
152%
505%
505%
505%
152%
505%
156%

4.3.2 Sowing
The sowing plan included the machine and energy cost of corn seed sowing.
Similarly to the tillage process, it was built based on the Ecoinvent Data v2.2 u-so
type unit process named “CH: sowing”. For primary energy, Switzerland diesel was
substituted by U.S. diesel. The sowing plan model is illustrated in Figure 4- 4. The
input and output flows for sowing per hectare are presented in Table 4-6. The
temporal, geography and key assumptions of the processes in this plan are
documented in Table 4-7.
CH: shed

US: diesel, at refinery

CH: tractor production
CH: agricultural machinery,
tillage, production

CH: sowing

Figure 4- 4 Sowing plan model

67

1 ha farm being
chisel tilled

Table 4- 6 Input /Output flows of sowing corn seed per 1 ha farm with primary US
energy substitution
Standard
Flow
Amount
Unit deviation
Input
CH: agricultural machinery, general, production
[Machines]
0.966 kg
111%
CH: diesel, at regional storage [fuels]
3.82 kg
111%
2
CH: shed [buildings]
0.00546 m
301%
CH: tractor, production [Machines]
0.596 kg
111%
Output
CH: sowing [work processes]
10000 m2
0%
Waste heat [Other emissions to air]
Carbon dioxide [Inorganic emissions to air]
Nitrogen oxides [Inorganic emissions to air]
Carbon monoxide [Inorganic emissions to air]
Dust (PM2.5) [Particles to air]
NMVOC (unspecified) [Group NMVOC to air]
Sulphur dioxide [Inorganic emissions to air]
Zinc (+II) [Heavy metals to agricultural soil]
Methane [Organic emissions to air (group VOC)]
Nitrous oxide (laughing gas) [Inorganic emissions
to air]
Ammonia [Inorganic emissions to air]
Benzene [Group NMVOC to air]
Polycyclic aromatic hydrocarbons (PAH) [Group
PAH to air]
Copper (+II) [Heavy metals to air]
Zinc (+II) [Heavy metals to air]
Lead (+II) [Heavy metals to agricultural soil]
Cadmium (+II) [Heavy metals to agricultural soil]
Nickel (+II) [Heavy metals to air]
Chromium (unspecified) [Heavy metals to air]
Benzo pyrene [Group PAH to air]
Cadmium (+II) [Heavy metals to air]
Selenium [Heavy metals to air]

68

174
11.9
0.17
0.0143
0.0128
0.0125
0.00385
0.00085
0.00049

MJ
kg
kg
kg
kg
kg
kg
kg
kg

111%
121%
152%
501%
305%
152%
121%
152%
156%

0.00046 kg
7.64E-05 kg
2.79E-05 kg

156%
156%
156%

1.26E-05
6.50E-06
3.82E-06
1.44E-06
3.25E-07
2.68E-07
1.91E-07
1.15E-07
3.82E-08
3.82E-08

305%
505%
505%
152%
152%
505%
505%
505%
505%
156%

kg
kg
kg
kg
kg
kg
kg
kg
kg
kg

Table 4- 7 Representation of processes in sowing activity with primary US
energy substitution
Process
Temporal Geography
Key assumptions
seeder: 1000 kg machine,
agricultural machinery,
19951000 hours useful life, 0.54
general, production
2002
Switzerland
repair factor
19953000 kg machine, 7000 hours
tractor, production
2002
Switzerland
useful life, 0.74 repair factor
dataset was built on one
example of a typical
1994agricultural building, 50 years
shed
2002
Switzerland
lifetime
diesel, at refinery
2009
U.S.
US LCI dataset
1991Sowing
2002
Switzerland
working width 3 m

4.3.3 Fertilizing
The fertilizing activity refers to a process that applies granular fertilizers using a
broadcaster. The fertilizer plan inventories the machine cost and primary energy
consumption to broadcasting fertilizer per one-hectare farm. The NPK fertilizer
consumption is not included in the fertilizing plan.
Similar to the tillage process, the fertilizing plan consisted of five processes: i)
general agricultural machinery production, ii) tractor production, iii) shed, iv) U.S.
diesel, and v) fertilizing by broadcaster. Process iv was obtained from the US LCI
database, and the other four processes are derived from Ecoinvent Data v2.2. Detail
inputs and outputs of the fertilizing by broadcaster unit process (u-so) are
inventoried in Table 4-8. The temporal, geographical representations and key
assumptions of each process in this plan are listed in Table 4-9.

69

Table 4- 8 Input /Output flows of fertilizing by broadcaster process with primary US
energy substitution
Standard
Flow
Amount
Unit deviation
Input
CH: agricultural machinery, general, production
[Machines]
0.241 kg
111%
CH: diesel, at regional storage [fuels]
5.29 kg
111%
CH: shed [buildings]
0.00171 m2
301%
CH: tractor, production [Machines]
0.687 kg
111%
Output
CH: fertilizing, by broadcaster [work
processes]
10000 m2
0%
Ammonia [Inorganic emissions to air]
0.00011 kg
156%
Benzene [Group NMVOC to air]
3.86E-05 kg
156%
Benzo pyrene [Group PAH to air]
1.59E-07 kg
505%
Cadmium (+II) [Heavy metals to air]
5.29E-08 kg
505%
Cadmium (+II) [Heavy metals to agricultural soil]
3.40E-07 kg
152%
Carbon dioxide [Inorganic emissions to air]
16.5 kg
121%
Carbon monoxide [Inorganic emissions to air]
0.021 kg
501%
Chromium (unspecified) [Heavy metals to air]
2.65E-07 kg
505%
Copper (+II) [Heavy metals to air]
9.00E-06 kg
505%
Dust (PM2.5) [Particles to air]
0.0208 kg
305%
Lead (+II) [Heavy metals to agricultural soil]
1.49E-06 kg
152%
Methane [Organic emissions to air (group VOC)]
0.00068 kg
156%
Nickel (+II) [Heavy metals to air]
3.70E-07 kg
505%
Nitrogen oxides [Inorganic emissions to air]
0.231 kg
152%
Nitrous oxide (laughing gas) [Inorganic emissions
to air]
0.00064 kg
156%
NMVOC (unspecified) [Group NMVOC to air]
0.0143 kg
152%
Polycyclic aromatic hydrocarbons (PAH) [Group
PAH to air]
1.74E-05 kg
305%
Selenium [Heavy metals to air]
5.29E-08 kg
156%
Sulphur dioxide [Inorganic emissions to air]
0.00533 kg
121%
Waste heat [Other emissions to air]
240 MJ
111%
Zinc (+II) [Heavy metals to air]
5.29E-06 kg
505%
Zinc (+II) [Heavy metals to agricultural soil]
0.0009 kg
152%

70

Table 4- 9 Representation of processes in fertilizing activity with primary US energy
substitution
Process
Temporal Geography Key assumptions
agricultural machinery, 1995Switzerland seeder: 1000 kg machine, 1000
general, production
2002
hours useful life, 0.54 repair factor
tractor, production
1995Switzerland 3000 kg machine, 7000 hours
2002
useful life, 0.74 repair factor
shed
1994Switzerland dataset was built on one example
2002
of a typical agricultural building,
50 years lifetime
diesel, at refinery
fertilizing, by
broadcaster

2009
19912002

U.S.
US LCI dataset
Switzerland
500 fertilizer carrying capacity

4.3.4 Combine harvesting
A Combine harvester machine is one of the most often used machine to harvest
grain crops. A combine harvester can finish reaping, threshing, and winnowing corn
simultaneously in a single pass. Depending on the needs, the specific model, and
crop row spacing, the combine can harvest about three to ten rows at the same time.
Thus, the combine harvester is one of the most important labor saving machines for
corn harvesting.
There are inevitable corn losses in corn harvesting. Prior to machine harvesting,
about one bushel per acre ear is lost due to early drop [6]. Machine-loss is
comprised of failure in corn collection by the combine and processing loss in the
combine machine. The Sandhill farm experiment employed a hand-pick method to
harvest corn due to the small experimental area and the need for several batches in
harvest sampling. Therefore, a zero machine loss rate was ideally assumed in the
combine-harvesting plan for these scenarios.

71

The combine harvesting plan consisted of four unit processes: i) harvester
production, ii) shed, iii) diesel, and iv) combine harvesting. The inputs and outputs
to harvest a one-hectare farm were inventoried in Table 4- 10, and the
representation and key assumptions of processes in this plan were recorded in
Table 4- 11.
Table 4- 10 Inputs and Outputs of the combine harvesting process
Flow
Input
CH: diesel, at regional storage [fuels]
CH: harvester, production [Machines]
CH: shed [buildings]
Output
CH: combine harvesting [work processes]
Ammonia [Inorganic emissions to air]
Benzene [Group NMVOC to air]
Benzo pyrene [Group PAH to air]
Cadmium (+II) [Heavy metals to air]
Cadmium (+II) [Heavy metals to agricultural soil]
Carbon dioxide [Inorganic emissions to air]
Carbon monoxide [Inorganic emissions to air]
Chromium (unspecified) [Heavy metals to air]
Copper (+II) [Heavy metals to air]
Dust (PM2.5) [Particles to air]
Lead (+II) [Heavy metals to agricultural soil]
Methane [Organic emissions to air (group VOC)]
Nickel (+II) [Heavy metals to air]
Nitrogen oxides [Inorganic emissions to air]
Nitrous oxide (laughing gas) [Inorganic emissions
to air]
NMVOC (unspecified) [Group NMVOC to air]
Polycyclic aromatic hydrocarbons (PAH) [Group
PAH to air]
Selenium [Heavy metals to air]
Sulphur dioxide [Inorganic emissions to air]

72

Amount

Unit

Standard
deviation

33.3 kg
6.3 kg
0.00858 m2

111%
111%
301%

10000
0.00067
0.00024
1E-06
3.3E-07
8.8E-07
103
0.32
1.7E-06
5.7E-05
0.149
3.8E-06
0.0043
2.3E-06
1.7

m2
kg
kg
kg
kg
kg
kg
kg
kg
kg
kg
kg
kg
kg
kg

0
156%
156%
505%
5.05
152%
121%
501%
5.05
505%
305%
152%
1.56
505%
152%

0.004 kg
0.145 kg

156%
1.52

0.00011 kg
3.3E-07 kg
0.0336 kg

305%
156%
121%

Table 4- 10 (cont’d)
Flow
Waste heat [Other emissions to air]
Zinc (+II) [Heavy metals to air]
Zinc (+II) [Heavy metals to agricultural soil]

Amount
1510
3.3E-05
0.00238

Unit
MJ
kg
kg

Standard
deviation
1.11
505%
152%

Table 4- 11 Representation and key assumptions of harvesting activity with primary
US energy substitution
Process
harvester
production

Temporal Geography Key assumptions
1995Switzerland 1000 kg machine, 1300 hours
2002
useful life, 0.55 repair factor

shed

19942002

diesel, at refinery
combine
harvesting

2009
19912002

4.4

Switzerland dataset was built on one example of
a typical agricultural building, 50
years lifetime
U.S.
US LCI dataset
Switzerland working width 4.5 m, straw
treatment is not included

Irrigation
The most widely used irrigation approach to supplement water for corn is center

pivot irrigation [7]. To enhance the water use efficiency, a drip irrigation method
was employed to irrigate corn on the Sandhill farm experiment, combined with
SWRT.
4.4.1 Drip irrigation, irrigation water and electricity
The drip irrigation process manages the quantity relationship between
electricity consumption per unit of irrigation water, as illustrated in Figure 4- 5.

73

According to the irrigation process in Ecoinvent report no.15 [8], 2.64 MJ electricity
is spent to pump 1000 kg water with a delivery pressure of 700,000 to 800,000 Pa.
Electricity, at grid, western US

2.64 MJ
Drip irrigation

Water

1000 kg

Figure 4- 5 Drip irrigation process
4.4.2 Drip tape production
Drip tapeused in the Sandhill farm experiment is a type of black color HDPE tube.
It is a semi-rigid textured tube manufactured by extrusion. The data for the
manufacturing process dataset to produce the irrigation tube was not available.
Therefore, the process of general purpose HDPE pipe production was used as drip
tape production. This was a cradle to plant gate dataset, which included petroleum
extraction, oil fractionation for ethylene, polymerization for HDPE, and pipe
extrusion. The aggregated HDPE pipe production unit process was developed based
on Europe. The geography and technology differences should be noted.
The mass of drip tape used per unit area was measured from sampling (as
shown in Appendix-LCI Table A-12). The total length of the drip tape was calculated
based on irrigation maps; the mass to length ratio of the drip tape was obtained
from measurement; the average drip tape consumption (kg/m2) was calculated
from the total mass of irrigation tubes [kg] divided by the total irrigated area [m2].
The average drip tape consumption rate was used for drip tape infrastructure flow

74

calculations in the field preparation process. A detailed calculation of irrigation tube
consumption is provided in Section 4.9.4.
4.4.3 End of life (EOL) of disposed drip tape
Drip tapes are disposed every year after crop harvesting. Because the soil
contamination is higher than general accepted contamination levels for recycling,
and the non-PVC irrigation pipe-recycling program is immature, only a small
fraction of the disposed irrigation pipes are recycled. A majority of them are sent to
incineration and landfill. Because agricultural plastic is not included in municipal
solid waste (MSW) report that is generated annually by the Environmental
Protection Agency, little solid data can be found regarding the drip tape EOL. The
result of the New York State agricultural plastic disposal survey in 2004 [9] was
used as the benchmark for the EOL fraction assumption: 66% wt. incineration, 27%
wt. landfill, and 7% wt. recycle. These EOL fractions are used in the harvesting unit
process.
4.5

Chemical
Urea, K2O, and P2O5 are granular fertilizers applied ahead of planting via

broadcasting. The liquid fertilizers (KCl and Ammonium Nitrate (AN)) and
herbicides (Ammonium Sulfate (AMS) and glyphosate) are delivered in solution.
They are applied to the crops via the irrigation system during the growing phases.
Since the amount of electricity consumed to apply these liquid chemicals is
negligible, the electricity consumption is not accounted. Gypsum (CaSO4 ·2H2O) is a
common used sulfuric acid form soil-pH adjuster. The ideal pH for most crop

75

growth is around 6.5. To lower soil pH, gypsum is a commonly used amendment.
Lowering pH is a slow process; it typically takes more than a year to be effective.
Since the latest gypsum amendment in 2009, no more gypsum was added to the
Sandhill farm during the 2012-2013 experiments. According to soil experts, gypsum
on NOSWRT treatment should be added by the year 2015, while they were
expecting less frequent amendment of gypsum on SWRT treatment. The gypsum
amendment rate [kg/ha] is assumed the same for both SWRT and NOSWRT
treatments.
In the LCIs for the 2012 and 2013 experiments, as described above, there were
eight unit processes in the chemical group. In the LCIs of 2004-2013 simulation,
there were six unit processes, with liquid fertilizers (KCl and AN) being left out.
Leaving out liquid fertilizers was due to a) the small irrigation fraction (10%) of the
simulated dataset leads to tiny liquid fertilizer consumption, and b) a lack of
confident predictions concerning the application rates.
All of the chemical unit processes were cradle to gate production processes,
which include material, energy used, and transportation. They represented the
standard technology level in their corresponding temporal scope. The tracked
elementary output flow was 1 kg of chemical production for all eight chemical
production processes. The geography, temporal, technology representations, and
data resource are summarized in Table 4- 12.

76

Table 4- 12 Chemical processes description
Process
Urea

Geography Temporal Database
U.S.

20112014

PE

P2O5

U.S.

20112014

PE

K2O

Germany

2000

Ecoinvent
v2.2

AN (liq.)

U.S.

20112014

KCl (liq.)

U.S.

Glyphosate Europe
AMS

Gypsum

U.S.

Germany

20112014
20002010
20112014
20112014

PE
PE
Ecoinvent
v2.2
PE

PE

Note
46% N content, represent technology that ammonia and CO2
transformation into ammonium carbonate, which then being cracked
into water and urea by heating; country specific energy supply,
transportation included
45% P content, technology representation: rock phosphate and
phosphoric acid are transferred by energy input to triple super
phosphate (TSP); after that product is pelletized. Country specific
energy supply, transportation included
60% K2O content, a mixture of technologies in salt concentration
(solution in hot water, flotation, and electrostatic separation)
52% N content, represent technology of neutralization of nitric acid
with gaseous ammonia; country specific energy supply, transportation
included
60% potassium compound content, market value allocation, country
specific data
Including materials, energy uses, infrastructure and emissions; from
literature data, modeled for Europe
AMS and acrylonitrile are produced as co-products of a combine
reaction; country specific data, transportation included
Represent production technology via opening pit mining, then gypsum
stone being crushed, grinded, dried, and purified. Country specific
energy grid, transportation included.

77

4.6

Seed
The corn seeds planted in the Sandhill farm experiment were Roundup Ready

DeKalb DKC 46-61 hybrid for both the 2012 and 2013 experiments. They were
purchased from DeKalb Genetics Corporation. Thus, the seed process was modeled
as an open-loop seed resource.
Field corn seeds planted on Sandhill farm were directly applied via machine
sowing without pretreatment (such as seedling in greenhouse). The seed unit
process in this LCA study referred to 1 kg maize seed production (fresh weight with
12% humidity). Unit process data from Ecoinvent v2.2 named “maize seed IP, at
farm” was used. This was a cradle to farm gate corn seed production, including seed,
chemicals, machinery, energy, biogenic carbon, transportation, and land occupation
and transformation. The seed process represented the situation in Switzerland, with
3000 kg/ha grain yield in 2000.
Table 4- 13 illustrates the input and output inventories for the seed process.
From the table, it can be speculated that the corn seed was cultivated without
artificial water supplement. In addition, it should be acknowledged that the
Switzerland seed process was less representative of greenhouse U.S. situation
because the grain yield (3000 kg/ha) was much lower than the U.S. average

78

Table 4- 13 LCI of seed process
Group
Seed

Chemical

Machinery

Energy
Biogenic
carbon

Transport

Flows
Inputs
CH: maize seed IP, at regional storehouse [seed]
RER: ammonium nitrate, as N, at regional
storehouse [mineral fertilizer]
RER: triple superphosphate, as P2O5, at regional
storehouse [mineral fertilizer]
RER: potassium chloride, as K2O, at regional
storehouse [mineral fertilizer]
CH: green manure IP, until April [plant production]
RER: acetamide-anillide-compounds, at regional
storehouse [Pesticide]
RER: organophosphorus-compounds, at regional
storehouse [Pesticide]
RER: triazine-compounds, at regional storehouse
[Pesticide]
CH: tillage, harrowing, by spring tine harrow [work
processes]
CH: tillage, ploughing [work processes]
CH: mowing, by rotary mower [work processes]
CH: sowing [work processes]
CH: fertilizing, by broadcaster [work processes]
CH: application of plant protection products, by
field sprayer [work processes]
CH: combine harvesting [work processes]
CH: grain drying, low temperature [work processes]
CH: electricity, low voltage, at grid [supply mix]
Energy, calorific value, in organic substance
[Renewable energy resources]

Amount

Unit

0.005

kg

0.03536

kg

0.02178

kg

0.01502

kg

3.3333

m2

0.0002

kg

8.33E05

kg

0.00022

kg

10

m2

3.3333
20
11.667
10

m2
m2
m2
m2

5.6667

m2

3.3333
0.35385
0.0468

m2
kg
MJ

16.216

MJ

Carbon dioxide [Renewable resources]

1.401

kg

CH: transport, freight, rail [Railway]
CH: transport, lorry 20-28t, fleet average [Street]
CH: transport, tractor and trailer [work processes]

0.01714
0.01714
0.01
9.00E05
0.1543

t*km
t*km
t*km

CH: transport, van <3.5t [Street]
RER: transport, barge [Water]

79

t*km
t*km

Table 4- 13 (cont’d)
Group

Land occ. &
transform.

Flows
Occupation, arable, non-irrigated [Hemerobie
ecoinvent]
Transformation, from arable, non-irrigated
[Hemerobie ecoinvent]
Transformation, from pasture and meadow
[Hemerobie ecoinvent]
Transformation, to arable, non-irrigated
[Hemerobie ecoinvent]
Outputs
CH: maize seed IP, at farm [seed]
Ammonia [Inorganic emissions to air]
Atrazine [Pesticides to agricultural soil]
Cadmium (+II) [Heavy metals to fresh water]
Cadmium (+II) [Heavy metals to agricultural soil]
Chromium (+VI) [Heavy metals to fresh water]
Chromium (unspecified) [Heavy metals to
agricultural soil]
Copper (+II) [Heavy metals to fresh water]
Copper (+II) [Heavy metals to agricultural soil]
Glyphosate [Pesticides to agricultural soil]
Lead (+II) [Heavy metals to fresh water]
Lead (+II) [Heavy metals to agricultural soil]
Metolachlor [Pesticides to agricultural soil]
Nickel (+II) [Heavy metals to fresh water]
Nickel (+II) [Heavy metals to agricultural soil]
Nitrate [Inorganic emissions to fresh water]
Nitrogen oxides [Inorganic emissions to air]
Nitrous oxide (laughing gas) [Inorganic emissions
to air]
Phosphate [Inorganic emissions to fresh water]

80

Amount

Unit

1.6667

m2*a

2.3667

m2

0.96667

m2

3.3333

m2

1
0.00086
0.00022
2.63E08
2.42E06
7.57E06
5.10E06
1.77E06
-3.16E07
8.33E05
9.57E08
4.32E07
0.0002
7.83E07
2.35E06
0.09769
0.00042

kg
kg
kg

0.00199

kg

8.58E05

kg

kg
kg
kg
kg
kg
kg
kg
kg
kg
kg
kg
kg
kg
kg

Table 4- 13 (cont’d)
Group

Flows
Phosphorus [Inorganic emissions to fresh water]
Waste heat [Other emissions to air]
Zinc (+II) [Heavy metals to fresh water]
Zinc (+II) [Heavy metals to agricultural soil]

4.7

Amount
4.78E05
0.0468
5.67E06
8.39E06

Unit
kg
MJ
kg
kg

SWRT
One of the intentions of using SWRT is to extend the time that the soil water

stays in the root zone. To achieve this goal, plastic membrane strips were installed
below the root zone with a contoured shape. The inventories of using SWRT should
include: membrane production, special machine production for the SWRT
membrane installation (referred as the SWRT machine in the following chapters),
and energy used for installation. The data sources, and representation of temporal,
geography, and technology are presented in Table4- 14.
Because SWRT is a pilot technology, with limited information available, a
number of key assumptions were necessary to conduct this LCA study.
Currently, the membrane used in SWRT is made of linear low-density
polyethylene (LLDPE). This is a commercial thermoplastic without degradability.
Theoretically, it is not expected that the SWRT membrane will fail due to
degradation for at least a century. In practice, the membrane might fail to work
because of animal attack, contour shape deformation, and accidental human
breakage. However, any of the above failures would simply be repaired locally and

81

do not affect the life of the total membranes across any field. Thus, a very
conservative 10-year lifetime assumption for SWRT lifetime was made.
The second key assumption was that the SWRT machine production had
equivalent environmental impact to tillage machine production. There was a lack of
SWRT machine production data; however, the SWRT machine has similar tilling
effects as the tillage machine.
The third assumption was that the diesel consumption rate was assumed to be a
function of the SWRT installation area; in other words, installation depth was
irrelevant. It took about six hours to install one-acre of land with John Deere 8520
tractor (Moline, IL, US) at the very beginning tractor development phase in the
experiments.. The diesel consumption rate was estimated according to the engine
test report [10]. Because the installation rate is positive related to the number of
chisels on the tractor, the advanced tractor with more chisels are expected to have a
higher installation efficiency.

82

Table 4- 14 Representation and data resource of SWRT processes
Process

Source

Polyethylene Plastic
film
Europe

Temporal Geography

Technology

20052012

Cradle to gate production, from
oil extraction, to obtain ethylene
through oil fractionation to
produce LDPE resin, and then
film extrusion

Europe

Tillage
machine
production

Ecoinvent 1995v2.2
2002

Represent plough class machine,
Switzerland with 800 kg, 800 hours lifetime,
and 0.72 repair factor

Tractor,
production

Ecoinvent 1995v2.2
2002

Switzerland

Shed

Ecoinvent 1994v2.2
2002

Dataset was built on one
Switzerland example of a typical agricultural
building, 50 years lifetime

3000 kg machine, 7000 hours
useful life, 0.74 repair factor

Diesel, at
refinery

US LCI

2009

U.S.

Represent conventional diesel
production from well drilling,
crude oil production and
processing as well as
transportation of crude oil via
pipeline resp. vessel to the
refinery.

Membrane
installation

modified
from
tillage,
chiselling

20122013

U.S.

Pilot SWRT installation
technology, machine working
width 1 m

4.8

Planting
The planting group covers two unit processes: field preparation and harvest. The

field preparation process defines the inputs for planting corn and outputs of
fertilizer emissions. The harvesting process is connected right after the field
preparation process. It describes the machinery to harvest the corn.

83

4.8.1 Field preparation
The field preparation process was the only fixed process in the LCA plan. This
means the field preparation process was the reference process, and every other
process was scaled in relation to this fixed process. The inputs were the tracked
elementary output flows of the seed, machinery, chemicals, irrigation, and SWRT
processes. Its tracked output was 1000 kg corn grain, with the scaling factor 1 being
fixed. The LCI of the field preparation process for year 2013 irrigated SWRT is
presented in Table 4- 15. The other LCI tables for the four 2012 experiments, 2013
non-irrigated, and 2004-2013 simulated NOSWRT are listed in Appendix-LCI
Section 4 Field preparation process LCI tables.

84

Table 4- 15 LCI of field preparation process in irrigated SWRT plan
Group
Seed

Chemical

Machinery
Irrigation
SWRT
Biogenic
carbon
Land
occupation

Flow
Inputs
Maize seed IP, at farm [seed]
Urea (agrarian) [Agro chemicals]
Triple superphosphate (agrarian, 45% P2O5)

Amount Unit
8.3684
3.3456
1.3038

kg
kg
kg

Potassium chloride, as K2O, at regional storehouse
Ammonium nitrate (solution 52%, agrarian)
Potassium chloride (agrarian, 60% K2O)
Ammonium sulfate, as N, at regional storehouse
Glyphosate
Gypsum
Tillage, cultivating, chiselling [work processes]
Sowing [work processes]
Fertilizing, by broadcaster [work processes]
Pipe
Irrigating [work processes]
Polyethylene-film (PE) [Plastic parts]
Membrane installation [work processes]

3.3456
3.1734
2.091
0.633
0.1148
79.934
630.52
331.85
331.85
10.254
62.73
3.526
33.185

kg
kg
kg
kg
kg
kg
m2
m2
m2
kg
m3
kg
m2

Carbon dioxide [Renewable resources]

760

kg

Occupation, permanent crop, fruit, intensive

331.85

m2*a

1000
4.1174
0.0602
0.0235

kg
kg
kg
kg

Outputs
Elementary
flow
Fertilizer
emissions

US: corn, at farm [plant production]
Nitrate [Inorganic emissions to fresh water]
Nitrogen oxides [Inorganic emissions to air]
Phosphorus [Inorganic emissions to fresh water]

4.8.2 Harvest
The harvest process refers to harvesting corn grain at the farm and disposal of
the drip tape. As discussed in section 5.3.4 combine harvesting, an ideal assumption
was made that 100% of the produced corn can be harvested. There was no loss

85

between the 1000 kg corn grain output flow in the field preparation process and the
1000 kg corn grain output flow in the harvest process. Drying, stover treatment, and
transportation of goods were not included. The LCI of the harvest process is
presented in Table 4- 16.
Table 4- 16 LCI of harvest process in irrigated SWRT plan
Flow
Inputs
Combine harvesting [work processes]
Corn, at farm [plant production]
Outputs
Corn, grains [Renewable primary products]
Incineration good [Waste for disposal]
Municipal solid waste deposition [landfill]
Recycling goods [Waste for recovery]

4.9

Amount Unit
331.85
1000

sqm
kg

1000
6.7677
2.7686
0.7178

kg
kg
kg
kg

Calculation procedure
This section explains the calculation procedures by which input and output

flows in the field preparation processes were calculated. They are introduced group
by group as divided in Table 4- 15. Also, flows of irrigated SWRT are used as an
example to demonstrate the calculation process.
Before exploring calculation details, an important parameter land use (LU)
should be introduced. LU is extensively used in most flow quantity calculations.
Yield, a common recorded indicator in agricultural activity, was expressed in
harvested mass per unit area, e.g., bushel/acre, kg/ha. Most of the agricultural
consumptions are recorded as mass/ volume/ energy per unit area. The FU in this

86

LCA study is producing 1000 kg corn grain. Here, LU is the intermediate parameter
to convert the consumption information from kg/acre (Acre Cost) into mass/
volume/ energy per 1000 kg corn grain production (FU Cost). The LUs of each
treatment are first calculated based on Eq. 4- 1. The calculated LUs of total six
treatments in 2012 and 2013 experiments are list in Appendix-LCI Table A-8.
𝑌𝑖𝑒𝑙𝑑 [𝑘𝑔]
𝑈𝑛𝑖𝑡 𝑎𝑟𝑒𝑎 [𝑎𝑐𝑟𝑒]

1000 [𝑘𝑔]

= 𝐿𝑈 [𝑎𝑐𝑟𝑒]

(Eq. 4- 1)

Another extensively used parameter is the allocation factor (AF). Corn grain is
one of the output products of the planting process. A portion of responsibility of the
whole cultivation activity loads to grain production. The AF defines the portion,
which varies from 0 to 1. If the corn grain production activity is charged with the
whole corn cultivation duty, AF is 1. If the duty allocated on corn grain is based on
its economic value [11-14], energy content [15], or mass ratio [16], the AF equals
0.79, 0.63 or 0.5, respectively. The mass allocation method (AF=0.5) is the default
method in the calculation.
4.9.1 Seed flow calculation
The seed flow in the field preparation process for the 2013 irrigated SWRT
treatment was calculated based on Eq. 4-2:
𝐴𝑐𝑟𝑒 𝑐𝑜𝑠𝑡 [𝑘𝑔 /𝑎𝑐𝑟𝑒 ] ∗ 𝐿𝑈 [𝑎𝑐𝑟𝑒 ] ∗ 𝐴𝐹 = 𝐹𝑈 𝑐𝑜𝑠𝑡 [𝑘𝑔 /𝐹𝑈 𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑜𝑛 ]

(Eq. 42)

The acre cost was 102 kg/acre; it took 0.164 acre to produce 1000 kg corn grain.
Thus, the FU cost of seed flow was calculated as:

87

102 [(𝑘𝑔 𝑠𝑒𝑒𝑑)/𝑎𝑐𝑟𝑒] ∗ 0.164 [𝑎𝑐𝑟𝑒] ∗ 0.5 = 8.364 [(kg seed)/(FU production)]
4.9.2 Chemical flow calculation
Similar to seed flow calculation, the chemical flow calculation use Eq. 4- 2 to
calculate the FU cost of each chemical flow quantity in the field preparation process.
The Acre cost of each type of chemicals is inventoried in Appendix-LCI Table A-9
and Table A-10. The LU values can be found in Appendix-LCI Table A-11. The urea
flow in field preparation process of 2013 irrigated SWRT is presented as a
calculation example:
By checking Appendix-LCI Table A-2 and Appendix-LCI Table A-11, the urea acre
cost =40.8 [kg/acre], and LU irrigated SWRT = 0.164 acre were known.
Urea FU cost=40.8 [kg/acre] *0.164 [acre]*0.5 = 3.346 [kg urea/FU production]
4.9.3 Machinery flow calculation
Machine flows are calculated following Eq. 4-2. Different from the chemical flow,
the LU is the FU cost of machinery flows. The output elementary flows of the
machinery group are 1000 m2. A unit conversion process was required in flow
calculations. The sowing flow and fertilizing flow in field preparation process of
2013 irrigated SWRT treatment were calculated as: (1 acre = 4047 m2)
0.164 acre *4047 [m2/acre] * 0.5 = 331.85 m2
The tillage flow was a little different from other machinery flows in the
calculation. For NOSWRT treatments (2012 15” Ctrl & 30” Ctrl, 2013 nonirrigated
Ctrl, 2004-2013 simulated Ctrl), there were two tillages per growing season.
Therefore, the quantity was calculated as: 2* FU *4047 [m2/acre] * 0.5. For SWRT

88

treatments (2012 15”SWRT, 30” SWRT, and 2013 irrigated SWRT), the initial
membrane installation completed the tillage simultaneously. Given the key
assumption that the SWRT lifetime was 10 years, the annual tillage flow was
calculated as:
(1+2* t [a]) * FU [acre] * 4047 [m2/acre] *AF/ t = tillage flow

(Eq. 4- 3)

where t is SWRT lifetime
For 2013 irrigated SWRT: (1+2*9)*FU/10 * 4047 [m2/acre] * 0.5 = 630.52 [m2 ]
4.9.4 Irrigation flow calculation
Irrigation and drip tape are the two flows in the irrigation group of field
preparation processes. Irrigation flow (irrigation water [m3]) connects the output of
the drip irrigation process to the input of the field preparation process. Irrigating
flow was calculated followed Eq. 4-2.
To perform the drip tape flow calculation, the drip tape acre cost calculation
need to be done first. Each irrigated plot used 14 irrigation tubes, and each tube was
9.144 m (30 feet) long. Each plot area was 41.8 m2 (450 ft2). Thus, the tube used per
unit area was 14 * 9.144 m / 41.8 m2 = 3.063 [m/ m2]. According to the sampling
measurement, the drip tape was calculated as 9.904 *10-3 [kg/m]. Thus, the mass of
drip tape per irrigation unit area was 9.904 *10-3 [kg/m] * 3.063 [m/ m2] = 3.033 *
10-2 [kg/ m2] (equivalent to 125 [kg/acre]). The remaining calculation procedure for
drip tape FU cost was followed by Eq. 4- 2.

89

4.9.5 SWRT flow calculation
Polyethylene-film flow and membrane installation flow were the two SWRT
flows in the field preparation processes. The polyethylene film flow describes the
weight of LLDPE used to produce 1000 FU annually. The membrane installation flow
depicted the annual membrane installation area per FU.
The membrane total surface area used per 10000 m2 (1 hectare) farmland was
15152 m2. The membrane was maintained in a contour shape with width to depth in
the ratio of 2:1. Based on measurement result, the membrane thickness was 7.6E5±3E-6 m (3±0.1 mil). The density of LLDPE was about 0.92 [g/cm3 ] (eqv. to 920
[kg/m3]). Thus, the average membrane mass per hectare was 15152 [m2/ha] * 7.6 E5 [m] * 920 [kg/m3] = 1059 [kg/ha] (eqv. to 430 [kg/acre]).
Given that the SWRT membrane lifetime was 10 years, based on Eq. 4- 3, the
annual membrane consumption per FU is:
430 [kg/acre] * 0.164 [acre] *0.5 /10 [a] = 3.526 [kg /a]
Because there were no gaps between each membrane strips in the horizontal
direction, the installation area is the farm area that employed SWRT. For irrigated
SWRT treatment, the average annual membrane installation area per FU was:
0.164 [acre]* 4047 [m2/acre] * 0.5/10 [year] = 33.185 [m2/a]
4.9.6 Biogenic carbon flow calculation
CO2 in the air is naturally fixed by plants to be used for producing carbohydrates
via photosynthsis process. At this point, CO2 in the atmosphere is sequestered by
plants. The CO2 fixed and released by biomass production is called biogenic carbon.

90

From cradle to farm gate perspective, plants could receive credits for reducing the
GHG in the atmosphere. But from the whole life cycle perspective, growing plants
should not receive biogenic carbon dioxide credits. Plants will release the carbon
eventually when they are burned or decomposed. A classic view of point is that the
captured and released CO2 amount should be equal, which is called carbon neutral.
In this study, since the scope is cradle to farm gate, the corn production receives a
carbon credit.
The amount of CO2 fixed per 1000 kg corn production is estimated based on
literature [17-19], which reported 1.494, 1.33, and 1.75 kg CO2 fixed per 1 kg corn
plant production. The estimated CO2 captured per 1000 kg corn grain production is
illustrated below, where 0.5 is the default allocation factor by mass.
0.5*1000 [kg]*(1.494+1.33+1.75) [kg CO2/1 kg plant] /3=760 [kg CO2/FU]
4.9.7 Land occupation flow calculation
Crop land was occupied by corn plants for grain production. In the field
preparation process, the land occupation flow refers to the direct land used to
produce 1000 kg corn grain. Even though corn plants only physically occupy the
crop land less than six months (from May to Oct), the whole year of land occupation
was counted. One reason is that no other crops were planted during winter; the
other reason is that the fallow winter benefits the land for the next year’s corn
production. The land occupation flow was calculated using Eq. 4-2:
0.164 [acre/year] * 4047 [m2/acre]*0.5 = 331.85 [m2/year]

91

4.9.8 Fertilizer emission flow calculation
The fertilizer emission flows take nitrogen, phosphorous, and potassium
fertilizer emission to soil, water and air into consideration. The main impacts to air
are generated from N2O, NH3, and NOx. The major impacts to water are from nitrate
and phosphorus leaching. Impacts to soil are mainly due to heavy metal adhere to
soil and soil pH change due to fertilizer.
The NOx emission was the main fertilizer emission to air in Sandhill farm
condition. Based on the 2006 Intergovernmental Panel on Climate Change (IPCC)
guidelines for National Greenhouse Gas Inventories, direct N2O emissions from soil
due to nitrogen fertilizer application are about 1% of all applied N [20]. The N2O gas
generation requires anaerobic conditions in the soil. Considering the little anaerobic
environment in sandy soil driven by the large pore size compared to clay, direct N2O
emissions from N application were negligible. NH3 gas formation normally occurs on
the soil surface. But, the high permeability of sandy soil speeds up the N fertilizer
penetration into the soil, and this greatly lessens NH3 formation. The NOx gas
emission rate was calculated based on a reference [8].
Nitrate and phosphate leaching were the main contamination flows to water.
Nitrate is the form of N that can be directly used by plant growth. Nitrate is
negatively charged and highly water-soluble. In other words, nitrate is less likely to
attach to sandy soil that contains less soil organic matter soil organic matter (SOM)
(positive charge), and easily drains with water. The N drainage rate was about 63%
of the total applied N based on a study [21]. Phosphate leaching to water leached 1.8%
of the total applied P based on a reference [8].

92

Because of the negligible quantity and lack of reliable data, the emission to soil
flows were left out.
The emission flows for fertilizer emission were calculated based on Equation 44:
E= F*R*LU*AF

(Eq. 4- 4)

where
E is fertilizer emissions; the unit is kg/FU;
F is the fertilizer application rate, expressed in the unit of kg/acre;
R is the fertilizer release rate, expressed in the unit of kg/kg;
LU is the land use to produce 1 FU, in the unit of acre;
AF is the allocation factor, which defaults to 0.5 based on mass allocation.

4.10 Assumptions
This section lists the key assumptions in this study. These assumptions are
significant to this LCA study, either helping to simplify or support the LCI
calculations.
4.10.1 Machinery


The machine production was included as machine cost in the machinery work
flow. They were proportionally loaded to the machinery work flow based on the
operation time to finish 1 ha surface area over total machine lifetime.



The assumed machine lifetimes were: general tractor 7000 h, general
agricultural machine (machine class: seeder, hoe, fertilizer spreader, rotary

93

mower, self-loading-trailer) 1000 h, tillage machine 800 h, combine harvester
1300 h. The basis for the assumption is the agricultural machine weights
reported in ecoinvent report no.15- Life Cycle Inventories of Agricultural
Production Systems.


Operation time to finish 1 ha surface area for machinery processes were:
combine harvesting 1.3 h, fertilizing by broadcaster 1.3 h, sowing 1.3 h, and
chisel tillage 1.2 h.



The agricultural machines (general tractor, general agricultural machine, tillage
machine, combine harvester) used in Sandhill farm were manufactured in
Switzerland, and used in the U.S. Thus, the energy grid to produce the machines
was Swiss energy, and the primary energy (diesel) was U.S. diesel.



Transport machine from factory to the farm was included, same as the
assumptions of agricultural field work processes in Ecoinvent 2.2.

4.10.2 Irrigation


Water used for irrigation was well water.



Irrigation cost came from drip tape production, irrigation water use, and
electricity consumption for pumping water.



The water pumped from reservoir was 100% used on plant, without losees
during water transportation.



The pump engine was 22 kW with 30 m3 volume capacity.



The drip tape was disposed annually after harvest.



The EOL of the drip tape was 66% wt. to incineration, 27% wt. to landfill, and 7%
wt. to recycle.
94

4.10.3 Chemical


The consumption of carrier water and electricity to apply the liquid fertilizers
(AN and KCl) were negligible.



The 2004-2013 simulated experiment did not use liquid fertilizers.



The NPK granular fertilizer application levels for 2004-2013 simulated
treatments were assumed to be 134-36-90 kg/ha. The assumed application level
was a little higher than the 2013 experiment level because of the absence of
liquid fertilizer. (The NPK granular fertilizer application rates were 135-39-112
kg/ha in the 2012 experiment, and 80.2-31.2-80.2 kg/ha in the 2013 experiment)



The AMS and roundup application rate for 2004-2013 simulated treatments
were assumed to be the same as level used in 2012 and 2013 experiments.

4.10.4 Seed


The corn seed used in the experiment had the same environmental footprint
(EFP) as the Switzerland seed that the Ecoinvent v2.2 dataset refers to.

4.10.5 SWRT


The EFP of SWRT machine production was the same as the EFP of the
Switzerland tillage machine production.



It took 14.8 hours for the SWRT machine to install membrane in a 1 ha surface
area farm. The diesel consumption of the SWRT machine was 46.9 kg/h.



The SWRT membrane was LLDPE material with uniform 7.6E-5 m (3 mil)
thickness. The density of the membrane was 0.92 g/cm3.

95



The EFP of LLDPE membrane production was equal to EFP of PE membrane
production.

4.10.6 Planting


There was little surface run-off.



The amounts of N2O and NH3 direct emission from soil due to N fertilizer
amendment were negligible.



The NPK drainage rates were the same in SWRT and NOSWRT treatments.



The NO3 leaching rate was 63% of total N amendment. The NOx gas emission to
air was 1.8% of total N amendment. The phosphate leaching was 1.8% of total P
addition.



There was no machine lose in the combine harvest process.



Mass (corn grain): Mass (stover) = 1:1



The harvest corn grain had moisture content of 15.5%.



Drying and stover treatment were not included in the harvest process.

4.10.7 Missing data


Pump production for irrigation.



Fertilizer emission to soil.



Transportation of disposal drip tapes to EOL facilities, chemicals from
manufacture to farm, corn seed from the seed farm to Sandhill farm, SWRT
machine and membrane from factories to farm, and harvest corn grain to storage
barn.

96

APPENDIX

97

APPENDIX
The LCI tables in this section records the entire acre cost for flow quantity calculation and some information to assist in the
LCA calculation.

Group
Land

subcategory
direct land
use
soil

Seed

Irrigation

parameter
land occupation to
produce 1Mg grain
farm location
soil texture
seed
irrigation water
rainfall water
drip tape
pump electricity

Table A- 1 LCI of 2012 experiment
38.1 cm
76.2 cm
unit
NOSWRT
NOSWRT
m2

1032

degree
kg/ha
m3/ha
m3/ha
kg/ha
MJ/m3
water

98

252
1441

1008

38.1 cm
SWRT

76.2 cm
SWRT

595

749

42.6805, -84.4669
96.1% sand 3.9% silt& clay
126
252
665
4915
1824
309

2.64

126
2644

Group

subcategory
film

SWRT

tractor

thickness
surface area

Table A- 1 (cont’d)
38.1 cm
unit
NOSWRT
m
m2 /ha

density

g/cm3

LLDPE mass

kg/ha
Kg
diesel/h
hour /ha

parameter

fuel rate
working hours
tractor life time
employed hour

Chemical

granular

liquid

herbicide
pH-adjust
CO2
binding

machine weight
Urea (32% N)
phosphorus pentoxide
(P2O5) (43.7% P)
potassium chloride
(KCl) (47.1% K)
AN[ammonium
nitrate]
P[concentrated
superphosphate]
K[potassium chloride]
AMS
Rdup-glyphosate
gypsum / lime
CO2 capture from
atmosphere

76.2 cm
38.1 cm
NOSWRT
SWRT
7.6E-5±3E-6
15152
0.92
1062
45.2-48.6
14.8

hour

15000

kg
kg/ha

3000
135

kg/ha

39

kg/ha

112

kg/ha

52

kg/ha

30

kg/ha
kg/ha
kg/ha
kg/(ha*a)
kg/kg
plant

99

18
19
3.5
2409
1.52

76.2 cm
SWRT

Group

subcategory
by-product
wt. content
wt. content

fertilizer
emission

root
emission to
air

emission to
ground water

allocation
method

emission to
surface water

parameter
stover ± standard
error
corn grain commercial
moisture content
stover oven dried
moisture content
root depth

Table A- 1 (cont’d)
38.1 cm
unit
NOSWRT
kg/ha

9671±2135

76.2 cm
NOSWRT

38.1 cm
SWRT

9922±1444

16830±1256

%

15.5

%

4

cm

40-55
0

ammonia (NH3)

%

N2O
NOx

%
%

0
1.8

nitrate leaching

%

63

phosphorus leaching

%

1.8

phosphorus run-off

%

0

mass

mass (grain) : mass
(stover)

energy
content

grain

MJ/kg

53.4

stover
energy (grain) : energy
(stover)

MJ/kg

31.4

grain

$/ton

214

stover
price (grain) : price
(stover)

$/ton

50-66.7

economic
value

1

1

1

100

0.5:0.5

0.63:0.37

0.79:0.21

76.2 cm
SWRT
13376±13
19

Table A- 1 (cont’d)
38.1 cm
unit
NOSWRT

Group

subcategory

parameter

76.2 cm
NOSWRT

38.1 cm
SWRT

system
expansion

substitute
stover
environment
al impact
from total

stover collected rate

%

50

ethanol
convention
efficiency

stover ethanol

L/kg

0.3

Switchgrass ethanol
Switchgrass ethanol by
product- electricity
Switchgrass ethanol
net electricity
consumption for 150L
eqv. ethanol
production
yield of switchgrass

L/kg

0.4

KWh/ kg

0.206

kWh

21.75

ton/ha/a

101

10

76.2 cm
SWRT

LCI of 2013 experiment
Table A- 2 LCI of 2013 experiment
Group

sub-category

Parameter

unit

land

direct land
use

land occ. to produce 1000 kg grain

m2

land location
row spacing
soil composition
seed
irrigation water
rainfall water
total direct water
HDPE drip tape

degree
cm

seed

soil

water

surface area

kg/ha
m3/ha
m3/ha
m3/ha
kg/ha
MJ/m3
water
m2 /ha

thickness

m

density
LLDPE mass
fuel rate
working hours

g/cm3
kg/ha
kg diesel/h
hour /ha
hour
kg
kg/ha

pump electricity
SWRT

membrane

tractor

tractor life time employed hour

Chemical

granular

machine weight
urea
phosphorus pentoxide (P2O5)
(43.7% P)
K2O
102

kg/ha
kg/ha

IrrigatedSWRT

NonirrigatedNOSWRT

835
2820
42.6805, -84.4669
38.1
96.1% sand 3.9% silt& clay
252
1890
0
3998
3998
5889
3998
309
NA
2.64

15152
7.6E-5±3E6
2.54
1063
45.2-48.6
14.8
15000
3000
80.2
31.2
80.2

NA
NA
NA
NA
NA
NA
NA
NA
NA

Table A- 2 (cont’d)
Group

sub-category

0
62.9
19
3.46
2408
1.52

0
0

kg/ha

15679

3227

15.5
4
48

122

ammonia (NH3)

%

N2O
NOx

%
%

emission to
ground water

nitrate leaching

%

emission to
surface water

phosphorus emission through soil
erosion by water+ phosphorus
leaching

%

mass

mass (grain) : mass (stover)

1 (cont’d)

energy
content

grain

MJ/kg

stover

MJ/kg

main product
by-product
wt. content

allocation
method

kg/ha
kg/ha
kg/ha
kg/ha
kg/(ha*a)
kg/kg plant
kg/ha

P[concentrated superphosphate]
K[potassium chloride]
AMS
Rdup-glyphosate
gypsum / lime
CO2 capture from atmosphere
corn grain
harvest index
stover
corn grain commercial moisture
content
stover oven dried moisture content
root depth

pH-adjust

fertilizer
emission

NonirrigatedNOSWRT

unit

herbicide

CO2 binding
Yield

IrrigatedSWRT

Parameter

wt. content
root
emission to
air

103

%
%
cm

11981±359 3544±957
0.49
0.58

0
0
1.80
63

1.80
0.5:0.5
53.4
31.4

Table A- 2 (cont’d)
Group

Parameter

unit

IrrigatedSWRT

energy (grain) : energy (stover)

1

0.63:0.37

grain

$/ton

stover
price (grain) : price (stover)

$/ton
1

214
50-66.7
0.79:0.21

substitute
stover
environmental stover collected rate
impact from
total

%

50

ethanol
convention
efficiency

stover ethanol

L/kg

Switchgrass ethanol
Switchgrass ethanol by productelectricity
Switchgrass ethanol net electricity
consumption for 150L eqv. ethanol
production
yield of switchgrass

L/kg

sub-category
economic
value (2011
USDA)

system
expansion

104

KWh/ kg

0.3
0.4
0.206

KWh
ton/ha/a

21.75
10

NonirrigatedNOSWRT

LCI of 2004-2013 experiment
Table A- 3 LCI of 2004-2013 simulation
Group
land
seed
Irrigation

sub-category

parameter

unit

direct land use land location
row spacing
soil
soil composition
seed
irrigation water
HDPE drip tape

degree
cm

SWRT

tractor

SWRT

42.6805, -84.4669
38.1
96.1% sand 3.9% silt& clay
252
493
NA
49
309
NA
31

surface area

kg/ha
m3/ha
kg/ha
MJ/m3
water
m2/ha

thickness

m

NA

NA

density
LLDPE mass

g/cm3
kg/ha
kg
diesel/h
hour /ha
hour
kg

NA
NA

NA
NA

15152
7.6E-5±3E6
2.54
1063

NA

NA

45.2-48.6

NA
NA
NA

NA
NA
NA

15
15000
3000

pump electricity
membrane

NOSWRT NOSWRT
irrigated nonirrigated

fuel rate
working hours
tractor life time employed hour

machine weight

105

2.64

NA

2.64

NA

NA

Table A- 3 (cont’d)
Group

Chemical

Yield

sub-category

parameter

unit

granular

pH-adjust

urea
P2O5
K2 O
AMS
Rdup-glyphosate
gypsum / lime

CO2 binding

CO2 capture from atmosphere

main product
by-product

corn grain
stover
corn grain commercial moisture
content
stover oven dried moisture content
Harvest index-grain percentage over
total
root depth

kg/ha
kg/ha
kg/ha
kg/ha
kg/ha
kg/(ha*a)
kg/kg
plant
kg/ha
kg/ha

herbicide

wt. content
wt. content
Harvest Index
root
emission to
air

Fertilizer
emission

emission to
fresh water

NOSWRT NOSWRT
irrigated nonirrigated
134
36
80
19
3.5
5950
1.52
appendix 2-3
appendix 2-3

%

15.5

%

4

%

50

cm

122

NH3

%

0

N2O
NOx

%
%

0
1.8

nitrate leaching

%

63

phosphorus leaching

%

1.8

106

SWRT

Table A- 3 (cont’d)
Group

Allocation
method

System
expansion

NOSWRT NOSWRT
irrigated nonirrigated

sub-category

parameter

unit

mass
energy
content

mass (grain) : mass (stover)

1

grain

MJ/kg

53.4

stover
energy (grain) : energy (stover)

MJ/kg
1

31.4
0.63:0.37

grain

$/ton

stover
price (grain) : price (stover)

$/ton
1

substitute
stover
environmenta
l impact from
total

stover collected rate

%

50

ethanol
convention
efficiency

stover ethanol

L/kg

0.3

Switchgrass ethanol
Switchgrass ethanol by productelectricity
Switchgrass ethanol net electricity
consumption for 150L eqv. ethanol
production
yield of switchgrass

L/kg

0.4

economic
value (2011
USDA)

107

0.5:0.5

(cont’d)

214
50-66.7
0.79:0.63

KWh/ kg

0.206

KWh

21.75

ton/ha/a

10

SWRT

SALUS settings and simulation results
Soil profiles in SALUS
Table A- 4 Soil profile in SALUS
Layer
cm
Lower Limit of
soil water
m3/m3
Drained upper
limit of soil
m3/m3
Saturated soil
water content
m3/m3
Soil hospitality factor
Sat. hydraulic
cond.
cm/h
Bulk density,
Mg/m3
Organic C
%
Clay content
%
Silt content
%

0-5

6-15

16-25

26-34

35-43

44-53

54-90

0.1

0.1

0.1

0.1

0.1

0.1

0.1

0.21

0.21

0.21

0.21

0.21

0.21

0.21

0.34 0.34
1 0.874

0.34
0.874

0.34
0.351

0.34
0.351

0.34
0.31

0.34
0.31

10
1.67
0.28
4.5
5

10
1.67
0.27
4.5
5

10
1.67
0.24
4.5
5

10
1.67
0.23
4.5
5

10
1.67
0.2
5
5

10
1.67
0.29
5
5

10
1.67
0.29
4.5
5

Crop management profile
Table A- 5 Crop management profile
Category
Planting

Irrigation
Fertilizer
Tillage
Harvest

Item
date
plant population
method
distribution
row spacing
threshold for automatic apply
threshold for automatic apply
urea
date of 1st one-way disk date
date of 2nd one-way disk date
Harvest 99% product

108

Value
May 15
18.06
seed
row
38.1
30% of max. available
water
95% N stress factor
May 13
Oct 10
Harvest at maturity

Unit/
comment
plant/m2

cm
at 25cm
depth
incorporate
at 2 cm depth

Yield and irrigation water used of simulation results
Table A- 6 Yields from SALUS simulation for year 2004 to 2013

year
unit
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013

Irrigated NOSWRT
yield
Irrigation
[kg/acre] [m3/acre]
5057.6
79
3511.3
157
3872.1
159
3306.6
319
3658.1
318
4073.2
160
3431.4
240
4389.5
163
3439.2
320
4388.5
79

Nonirrigated
NOSWRT
yield
Irrigation
[kg/acre] [m3/acre]
4877.3
0
2637.3
0
3628.8
0
2425.3
0
2128.2
0
3907.2
0
2858.3
0
4304.6
0
2787.5
0
3858.7
0

Aggregate:
10% irrigated+90%
nonirrigated
yield
Irrigation
[kg/acre] [m3/acre]
4895.3
7.9
2724.7
15.7
3653.1
15.9
2513.4
31.9
2281.2
31.8
3923.8
16.0
2915.6
24.0
4313.1
16.3
2852.6
32.0
3911.7
7.9

Comparisons between simulations and institute records
Table A- 7 Percentage difference between average yields and simulated aggregate
NOSWRT
Year
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013

USDA record [18]
US
MI
Ingham
10062 8411 9980
9283
8976 9101
9359
9227 9478
9459
7720 7407
9660
8662 8097
10338 9290 9478
9591
9415 9967
9239
9603 9315
7746
8348 7614
9967
9729 NA

SALUS
simulation
Aggregate
12097
6733
9027
6211
5637
9696
7205
10658
7049
9666

109

Data Verification
US
MI
Ingham
-17% -30% -17%
38%
33%
35%
4%
2%
5%
52%
24%
19%
71%
54%
44%
7%
-4%
-2%
33%
31%
38%
-13% -10% -13%
10%
18%
8%
3%
1%
NA

Land uses (LUs) of experiments to produce 1000 kg corn grain
Table A- 8 LUs summary of experiment treatments
Treatment

2012

2013

38.1 cm SWRT
76.2 cm SWRT
38.1 cm NOSWRT
76.2 cm NOSWRT
Irrigated-SWRT
Non-irrigated-NOSWRT

LU
acre
0.147
0.185
0.255
0.249
0.164
0.555

m2
595
749
1032
1008
835
2820

Field preparation process LCI tables
This part of the document provides the field preparation process, LCI table of 2012
experiments, 2013 nonirrigated NOSWRT, and 2004-2013 simulated data. These LCIs are
created based on the same principles with different input levels. Here, the equations to
calculate each flow quantity are summarized. The input/output levels are available to look
up in Appendix-LCI Table A-1, 1A-2, and A-3.

110

Field preparation process LCIs of experiment treatments
Table A- 9 Field preparation process flow calculations of experiment treatments
Group
Seed

Chemical

Machinery
Irrigation
SWRT
Biogenic
carbon
Land
occupation

Flow
Inputs
seed
Urea
P2O5
K2O
AN
KCl
AMS
Glyphosate
Gypsum
Tillage
Sowing
Fertilizing
Pipe
Irrigating
PE film
Install

Flow equation

Unit

seed[kg/acre]*LU[acre]*AF
urea[kg/acre]*LU[acre]*AF
P2O5[kg/acre]*LU[acre]*AF
K2O[kg/acre]*LU[acre]*AF
I*AN[kg/acre]*LU[acre]*AF
I*KCI[kg/acre]*LU[acre]*AF
AMS[kg/acre]*LU[acre]*AF
glyphosate[kg/acre]*LU[acre]*AF
gypsum[kg/acre]*LU[acre]*AF
(t1[a]*2-S)*4047[m2/acre]*LU[acre]*AF/t1[a]
4047[m2/acre]*LU[acre]*AF
4047[m2/acre]*LU[acre]*AF
I*pipe[kg/acre]*LU[acre]*AF/t2[a]
I*water[m3/acre]*LU[acre]*AF
S*PE film[kg/acre]*LU*AF/t1[a]
S*4047[m2/acre]*LU[acre]*AF/t1[a]

kg
kg
kg
kg
kg
kg
kg
kg
kg
m2
m2
m2
kg
m3
kg
m2

CO2

1520[kg]*AF

kg

Land occ.
Outputs

4047[m2/acre]*LU[acre]*AF*1 year

m2*a

1000
(urea[kg/acre]+AN[kg/acre])*0.6316*LU[acre]*
AF
(urea[kg/acre]+AN[kg/acre])*0.018*LU[acre] *
AF
P2O5[kg/acre]*0.018*LU[acre]*AF

kg

Elementary
flow
corn grain
Fertilizer
emissions

Nitrate
NOx
Phosphorus

kg
kg
kg

where
S is SWRT factor, when treatment is SWRT, S=1; treatment is NOSWRT, S=0;
I is irrigation factor, when treatment is irrigated, I=1; treatment is nonirrigated, I=0;
t1 is SWRT membrane lifetime, a default value t1=10 years;
111

t2 is drip tape lifetime, a default value t2=1 year;
AF is allocation factor, a default value AF=0.5;
LU is the land use to produce one unit FU; the specific values of different treatments are list
in Appendix Table A-11.
Field preparation process LCIs of simulated treatments
Different from experimental treatments with single year data, simulated treatments
have ten continuous year data. In experimental treatments, SWRT burdens are evenly
loaded to every year, while in simulated treatments, SWRT burdens are charged totally to
the initial installation year 2004.

112

Table A- 10 Field preparation process flow calculations of simulated treatments
Group
Seed
Chemical

Chemical
Machinery

Machinery
Irrigation
Irrigation
SWRT
SWRT
Biogenic
carbon
Land
occupation

Flow
Inputs
seed
Urea
P2O5
K2O
AMS
Glyphosate
Gypsum
Tillage
Sowing
Fertilizing
Pipe
Irrigating
PE film
Install

Flow equation

Unit

seed[kg/acre]*LU[acre]*AF
urea[kg/acre]*LU[acre]*AF
P2O5 [kg/acre]*LU[acre]*AF
K2O[kg/acre]*LU[acre]*AF
AMS[kg/acre]*LU[acre]*AF
glyphosate[kg/acre]*LU[acre]*AF
gypsum[kg/acre]*LU[acre]*AF
(2-S)*4047[m2/acre]*LU[acre]*AF
4047[m2/acre]*LU[acre]*AF
4047[m2/acre]*LU[acre]*AF
I*pipe[kg/acre]*LU[acre]*AF/t2[a]
I*water[m3/acre]*LU[acre]*AF
S*PE film[kg/acre]*LU*AF
S*4047[m2/acre]*LU[acre]*AF

kg
kg
kg
kg
kg
kg
kg
m2
m2
m2
kg
m3
kg
m2

CO2

1520[kg]*AF

kg

Land occ.
Outputs

4047[m2/acre]*LU[acre]*AF*1 [a]

m2*a

Elementary
flow
corn grain
Fertilizer
emissions
Nitrate
Fertilizer
emissions

1000
(urea[kg/acre]+AN[kg/acre])*0.6316*LU[acre]
*AF
(urea[kg/acre]+AN[kg/acre])*0.018*LU[acre]
NOx
* AF
Phosphorus P2O5[kg/acre]*0.018*LU[acre]*AF

kg
kg
kg
kg

where,
S is SWRT factor, when it is the first year to install SWRT, S=1; otherwise, S=0;
I is irrigation factor, when treatment is irrigated, I=1; treatment is nonirrigated, I=0;
t2 is drip tape lifetime, a default value t2=1 year;
AF is allocation factor, a default value AF=0.5;

113

LU is land use to produce one unit FU; the specific values of different treatments are list in
appendix 5 Table 11.

Yield increase scenario on 2004-2013 simulated treatments
Table A- 11 LU [acre] to produce 1000 kg grain of simulated treatments
Year
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013

Original
0.204
0.370
0.274
0.401
0.450
0.255
0.344
0.232
0.352
0.256

20%+
0.170
0.308
0.228
0.334
0.375
0.212
0.287
0.193
0.293
0.213

30%+
0.157
0.284
0.211
0.309
0.346
0.196
0.265
0.178
0.271
0.197

50%+
0.136
0.246
0.183
0.268
0.300
0.170
0.229
0.155
0.235
0.171

80%+
0.114
0.205
0.152
0.223
0.250
0.142
0.191
0.129
0.196
0.142

100%+
0.102
0.185
0.137
0.201
0.225
0.127
0.172
0.116
0.176
0.128

200%+
0.051
0.092
0.068
0.100
0.113
0.064
0.086
0.058
0.088
0.064

300%+
0.041
0.076
0.055
0.082
0.093
0.051
0.070
0.046
0.072
0.052

where
Original means the LU by NOSWRT without yield increase assumption;
20+ means if SWRT will lead to yield increase by 20% compared to NOSWRT.

114

Drip tape sampling records
Drip tapes were collected and measured in 5 replicates to estimate the mass per unit
length. Each replicate was 0.25 m in length. The mass was recorded in the following table.
Table A- 12 drip tape measurement record
Replicate #
#1
#2
#3
#4
#5
average
stdev

mass [g]
2.61
2.36
2.43
2.48
2.50
2.48
0.09

115

REFERENCES

116

REFERENCES
1.

NASS. U.S. Corn Yield Statistics from Year 1983 to Year 2013 by USDA National
Agricultural Statistics Service. 2013 [cited 2013 Nov.].

2.

Ritchie, J.T. SALUS MODEL. [cited 2013 Nov.6]; Available from:
http://nowlin.css.msu.edu/salus/overview.html#N_1_.

3.

(MUS-AWO);, M.A.W.O., M.S.C.s. office;, and M.D.o. Geography, Enviro-weather. 2012,
Michigan State University: East Lansing.

4.

Farnham, D., Corn Planting Guide, D. Marks, Editor. 2001, Department of Agronomy,
Iowa State University. p. 8.

5.

Basso, F., M. Pisante, and B. Basso, 25 Soil Erosion and Land Degradation.
Mediterranean desertification: A mosaic of processes and responses, 2003: p. 347.

6.

UWEX, C.E.o. Grain Harvesting. Corn Agronomy 2014 [cited 2014 Mar 19, 2014];
Available from: http://corn.agronomy.wisc.edu/Management/L032.aspx.

7.

Perlman, H. Irrigation water use. 2005 Mar 17, 2014 [cited 2014 Apr 1]; Available
from: http://water.usgs.gov/edu/wuir.html.

8.

Nemecek, T. and t. Kagi, Life Cycle Inventories of Agricultural Production Systems.
2007, Agroscope Reckenholz-Tanikon Research station ART. p. 360.

9.

Levitan, l., Use and Disposal of Agricultural Plastics in Prohibition of Open Burning of
Solid Waste in New York State, e.r.a. Program, editor. 2004, Cornell University: Ithaca.

10.

Nelson, D., Nebraska OECD Tractor Test 1801–Summary 367 John Deere 8520 Diesel
16 Speed. 2002, Agricultural Research Division Institute of Agriculture and Natural
Resources: Lincoln.

11.

USDA. Agricultural Prices. 2011 Dec 30, 2011 [cited 2013 Sep 1st]; Available from:
http://usda01.library.cornell.edu/usda/nass/AgriPric/2010s/2011/AgriPric-1230-2011.pdf.

117

12.

Jena Thompson and W.E. Tyner. Corn Stover for Bioenergy Production: Cost Estimates
and Farmer Supply Response. Renewable Energy [cited 2013 Sep 1]; Available from:
https://www.extension.purdue.edu/extmedia/EC/RE-3-W.pdf.

13.

Edwards, W. Estimating a Value for Corn Stover. Ag Decision Maker 2011 Dec 2011
[cited 2013 Sep 1st]; Available from:
http://www.extension.iastate.edu/agdm/crops/pdf/a1-70.pdf.

14.

Koundinya, V. Corn Stover. renewable energy 2009 Nov 2009 [cited 2013 Sep 1];
Available from: http://www.agmrc.org/renewable_energy/corn-stover.

15.

Gelfand, I., S.S. Snapp, and G.P. Robertson, Energy efficiency of conventional, organic,
and alternative cropping systems for food and fuel at a site in the US Midwest.
Environmental science & technology, 2010. 44(10): p. 4006-4011.

16.

Kim, S., B.E. Dale, and R. Jenkins, Life cycle assessment of corn grain and corn stover in
the United States. The International Journal of Life Cycle Assessment, 2009. 14(2): p.
160-174.

17.

PE-GaBi, Corn, whole plant, at field, U.S.L. Database, Editor. 2009, PE-Gabi.

18.

PE International, L.-G., Corn grain (field border), L.-G. PE International, Editor. 2012,
PE-GaBi.

19.

MONSANTO. Everything You Ever Wanted to Know About Corn. 2013 [cited 2013
Sep 30]; Available from:
http://www.americasfarmers.com/2014/01/29/everything-you-ever-wanted-toknow-about-corn/.

20.

(IPCC), I.P.O.C.C., 2006 IPCC Guidelines for National Greenhouse Gas Inventories.
Institutefor Global Environmental Strategies, Hayama, Kanagawa, Japan, 2006.

21.

Guber, A., Nitrate emissions on Sandhill farm using 2-D hydrus model, in 2-D hydrus
modling SWRT. 2014, Michigan State University: East Lansing.

118

Results and interpretation
5.1

Evaluation of result quality
This chapter describes the results and interpretation of the SWRT LCA study. It is

divided into the following sections: a) completeness check, b) consistency check, and 3)
contribution analysis. The evaluation of the result quality was performed to identify
significant issues such as data gaps, data inconsistencies, and wrong use of data with the
LCI and to determine the reliability and robustness of the results.
5.1.1 Completeness check
A completeness check was performed on the inventory to evaluate the degree of
completeness. This step helps to investigate the completeness level concerning processes
in the LCA framework. Table 5-1 provides the basic completeness check of the data and
indicates that the inventories of the studied treatments were completeness at a satisfactory
level. A minor number of flow values were missing, so they were logically assumed.

119

Table 5- 1 Completeness check
2012

Machinery

Chemical

Irrigation

Seed

2013

unit process
harvest
fertilizing
sowing
tillage
gypsum
roundup
AMS
KCl
P2O5
urea
AN
K2O
water
drip irrigation
electricity

15''
SWRT
×
×
×
×
*
×
×
×
×
×
×
×
×
×
×

30''
SWRT
×
×
×
×
*
×
×
×
×
×
×
×
×
×
×

15''
Ctrl
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×

30''
Ctrl
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×

Irrigated
SWRT
×
×
×
×
*
×
×
×
×
×
×
×
×
×
×

Nonirrigated Ctrl
×
×
×
×
*
×
×
×
×
×
×
×
×
×
×

drip tape
pipe incineration
pipe landfill
pipe recycle
seed

×
∆
∆
∆
×

×
∆
∆
∆
×

×
∆
∆
∆
×

×
∆
∆
∆
×

×
∆
∆
∆
×

×
∆
∆
∆
×

120

Simulation
SWRT
×
×
×
×
*!
!
!
Ω
Ω
Ω

Ctrl
×
×
×
×
*!
!
!
Ω
Ω
Ω

×
×
×

×
×
×

×
∆
∆
∆
×

×
∆
∆
∆
×

Table 5- 1 (cont’d)
2012
unit process
SWRT

Planting

PE membrane
membrane install
tillage machine
production
tractor production
diesel
shed
Field Preparation
Harvest

2013

Simulation

15''
SWRT

30''
SWRT

15''
Ctrl

30''
Ctrl

Irrigated NonSWRT
irrigated Ctrl

SWRT

Ctrl

×
×

×
×

×
×

×
×

×
×

×
×

×
×

×
×

×
×
×
×
×
×

×
×
×
×
×
×

×
×
×
×
×
×

×
×
×
×
×
×

×
×
×
×
×
×

×
×
×
×
×
×

×
×
×
×
×
×

×
×
×
×
×
×

where
×-completeness
*- missing gypsum applied level for SWRT, assuming SWRT use the same level as Ctrl;
!- missing herbicide level for Simulation, assuming the same level used as experiments;
Ω- missing value, assuming levels based on corn fertilizer guidelines;
∆- incineration, landfill, and recycle percentage of HDPE drip tape not found, assumed values based on literature.

121

5.1.2 Consistency check
A consistency check was performed to investigate if the assumptions, methods, and
data were applied consistently throughout the LCA study [1]. The consistent applications of
the LCA methodology and the LCI inventory data were two of the major aspects evaluated
in the consistency check.
The methodological issues were first evaluated according to several aspects, such as the
LCI modelling framework, approaches, setting of system boundaries, the consistency in the
impact assessment, and other assumptions. The LCA study was conducted as an
attributional LCA study for every group and treatment. The mass allocation approach was
used as a default to solve the multifunction issues; the system boundaries of the inputs and
system were cradle to gate; and comparisons between treatments were under the same
impact assessment methodology- ReCiPe midpoint 1.07 (H); and the key assumptions were
consistently used on the different comparison scenario treatments.
The second part was to investigate the consistency of the LCI. The data accuracy, data
time-related issue, technological representation, and geography representation were areas
to evaluate the representativeness of the LCI. Table 5-2 presents the results of the
consistency check of the LCI.

122

Check
harvesting
fertilizing
Machinery
sowing
tillage
gypsum ₁
roundup
AMS
K2O
Chemical
P2O5
urea
AN
KCl (liquid)
irrigation water
drip irrigation
electricity
Irrigation
drip tape
₂
pipe landfill
pipe incineration
pipe recycle
Seed

seed

Source
database
database
database
database
database
database
database
database
database
database
database
database
database
literature
database
database
database
database
database

Table 5- 2 Consistency check
Data
Accuracy age Technology coverage
good
14
commercial level
good
14
commercial level
good
14
commercial level
good
14
commercial level
good
<3
commercial level
good
<3
commercial level
good
<3
commercial level
good
15
commercial level
good
<3
commercial level
good
<3
commercial level
good
<3
commercial level
good
<3
commercial level
good
<3
commercial level
caution
13
commercial level
good
14
commercial level
good
10
commercial level
good
<3
commercial level
good
<3
commercial level
good
<3
commercial level

database

Significant lower yield
than MI average

good

15

123

Time-related
coverage
1991-2001
1999-2001
1999-2001
1991-2002
2011-2014
2000-2010
2011-2014
2000
2011-2014
2011-2014
2011-2014
2011-2014
2011-2014
1991-2002
1999-2001
2005-2012
2011-2015
2011-2014
2011-2014

Geographical
coverage
Switzerland ₅
Switzerland ₅
Switzerland ₅
Switzerland ₅
Germany
Europe
US
Europe
US
US
US
US
Europe
Switzerland
US
Europe
US
Europe
US

2000

Switzerland

Table 5- 2 (cont’d)
Check
PE membrane

SWRT ₃

Source
database
measure,
literature caution

<3

pilot (estimation)

2012

US

database

good

13

commercial level

1995-2002

Switzerland

good
good
good

13
<3
13

commercial level
commercial level
commercial level

1995-2003
2009-2014
1994-2002

Switzerland
US
Switzerland

Field Preparation

database
database
database
measure,
literature

caution

<2

2012-2013

US

Harvest

literature weak

pilot (estimation)
commercial
(estimation)

2012-2013

US

membrane install
tillage machine
production
tractor production
diesel
shed

Planting ₄

Data
Time-related Geographical
Accuracy age Technology coverage coverage
coverage
good
10
commercial level
2005-2012
Europe

<2

Where, footnotes (1-5) are consistency notes
1- Gypsum applied level of SWRT treatments was consistently assumed to be identical as the Ctrl treatments;
2- Processes on irrigation groups were modelled only in treatments receiving irrigation;
3- Processes on SWRT groups were modelled only in SWRT treatments;
4- The yields of 2012 15”SWRT and 30”SWRT were significantly higher than the average corn yield in Ingham County (MI,
U.S.), while yield of 2013 nonirrigated Ctrl were reported significant lower than the Ingham average;

124

Table 5- 2 (cont’d)
5- The primary energy supplies to the machinery group were substituted from Switzerland diesel to US diesel, the
remaining secondary energy supply was unchanged. That is to say, the energy used to power the agricultural machine
was US diesel, and the energy used to support machine production and transportation from factory to farm was still
Switzerland.

125

5.1.3 Contribution analysis
The contribution of each group of processes to the total LCIA is investigated in the
contribution analysis. The contribution analysis provides a comprehensive view of the LCA
study to identify the major contributors and reveals insights about where to concentrate
additional energy and time to improve the robustness of the study. A summary of the
contribution analysis results is presented in Figure 5-1 and 5-2. The contributions of each
group process are evaluated using ReCiPe 1.07 (H) Midpoint indicators. The figures
indicate that, for treatments using irrigation, the irrigation group takes up a significant
weight on the water depletion, terrestrial acidification, human toxicity, fossil depletion, and
agricultural land occupation impact categories; for treatments without irrigation, the
impacts from machinery, chemical, and seed are magnified to considerable levels due to the
irrigation absence of irrigation. The planting group displays great contributions in the
freshwater eutrophication and freshwater ecotoxicity impact categories. The fertilizer
emissions are responsible for that.
The LCIA values and relative contributions of each flow are presented in Appendix A5Contribution analysis Section A5-1 and A5-2, respectively.

126

Machinery

Chemical

Irrigation

Seed

SWRT

Planting

Water depletion
Terrestrial acid.
Human toxicity
Freshwater eutro.
Freshwater…
Fossil depletion

Climate change
Agricultural…

0%

20%

40%

60%

80%

100% 0%

20%

40%

60%

80%

100%

Water depletion
Terrestrial acid.
Human toxicity
Freshwater eutro.
Freshwater ecotoxi.
Fossil depletion
Climate change
Agricultural land occ.

20%
40%
60%
80%
100%
0%
20%
40%
60%
80%
100% 0%
Figure 5- 1 Contribution analysis of 2012 SWRT treatments: 15”SWRT (left top), 30’’SWRT (right
top), 15” Ctrl (left bottom), and 30” Ctrl (right bottom)
127

Machinery

Chemical

Irrigation

Seed

SWRT

Planting

Water depletion
Terrestrial acid.
Human toxicity

Freshwater eutro.
Freshwater ecotoxi.
Fossil depletion
Climate change
Agricultural land occ.

0%

20%

40%

60%

20%

40%

60%

80%

100% 0%

20%

40%

60%

80%

100%

20%

40%

60%

80%

100%

Water depletion
Terrestrial acid.
Human toxicity
Freshwater eutro.
Freshwater ecotoxi.
Fossil depletion
Climate change
Agricultural land occ.

0%

80%

100%

0%

Figure 5- 2 Contribution analysis of 2013 Irrigated SWRT (left top), Nonirrigated Ctrl (right top), 2004
simulated Ctrl (left bottom), and 2004 simulated SWRT (right bottom)
128

5.2

LCIA Results
The LCIA results of six experimental treatments were analyzed using the ReCiPe 1.07 (H)

midpoint methodology. This methodology covers 18 impact categories, and 8 of them were
selected to investigate the analyzed treatments. The chosen impact categories were
agricultural land occupation, climate change, fossil depletion, freshwater ecotoxicity,
human toxicity, terrestrial acidification, and water depletion. To perform the comparisons,
the LCIA results of the experimental treatments were summarized by impact categories. In
addition, the contribution analysis concluded that the irrigation group had large
contributions to 6 of the 8 impact categories. So, the irrigation group was individually
presented at the unit process level. In the second part of the LCIA result section, the LCIA
results of a few published corn LCA studies are summarized and compared as benchmark
studies. Since the results of these studies were reported using mixed impact assessment
methodologies, the results of this study were transformed for comparison.

129

5.2.1 LCIA Results of experimental treatments

Figure 5- 3 ReCiPe 1.07 (H) Midpoint of Agricultural Land Occupation
Figure 5-3 indicates that the agricultural land occupation impacts are mainly from the
planting and seed groups. The planting group accounts for the direct land use to produce
1000 kg of corn grain. The order of impacts from the planting group for the six experiment
treatments is 15’SWRT < Irrigated SWRT < 30”SWRT < 30” Ctrl < 15” Ctrl < Nonirrigated
Ctrl, which is reversed to the yield order. This explanation is confirmed by the reverse
order of yield. The second highest contributor group is seed. The corn seed unit process
takes land occupation into account. Therefore, the land occupation burden from corn seed
should be accounted for when corn seeds are used for corn grain production.

130

Figure 5- 4 ReCiPe 1.07 (H) Midpoint of Climate Change
Figure 5-4 indicates the contribution of every group to the total climate change. The
irrigation group (including drip tape EOL, drip tape production, irrigation electricity, and
irrigation water) is the biggest producer, taking up 65%-73% of the total climate change
impact. When further analyzed, about 50% of the burden in the irrigation group was
derived from the water amendment activity. On average, every extra cubic meter of water
incurred burden from water and pump electricity consumption. The remaining about 50%
of the burdens for the irrigation group came from the irrigation system infrastructure (i.e.,
drip tape production and pipe EOL). Since there was no artificial water amendment in the
2013 nonirrigated Ctrl, though its yield is the lowest, the nonirrigated Ctrl treatment
displays the lowest climate change impact.

131

Figure 5- 5 ReCiPe 1.07 (H) Midpoint of Fossil Depletion
Figure 5-5 shows the fossil depletion midpoint impact, which has a similar pattern to
climate change. A significant difference between them is that the portion of drip tape EOL
vanishes while the portion of drip tape expands.
Figure 5-6 shows the main impact sources of freshwater ecotoxicity and freshwater
eutrophication. More than 80% of the impacts were due to fertilizer emissions embedded
in the planting group. From the soil electrical conductivity experiments which were
discussed in the planting group calculation procedure in the LCI chapter, and Hydrus 2D
modelling results [2], there was a lack of evidence to model a significant difference in
fertilizer emissions between the SWRT and the Ctrl treatments. Thus, the fertilizer
emission impacts linearly correlate to the fertilizer application level, and are inverse

132

correlated to the yields. Therefore, a natural recommendation will be avoiding excessive
fertilizer applications, which is one of the study recommendations.

133

Figure 5- 6 ReCiPe 1.07 (H) Midpoint of Freshwater Ecotoxicity (top) and freshwater
eutrophication (bottom)
134

Figure 5- 7 ReCiPe 1.07 (H) Midpoint of Human Toxicity
The machinery group is the greatest source of the LCIA human toxicity. The diesel
consumption to power the machinery was responsible for approximately 70% of the
machinery impacts, while machine production was responsible for about 25%. The
remaining 3-5% burden of the machinery group was mainly caused by the non-methane
volatile organic compounds (NMVOCs) emitted during vehicle maintenance. Irrigation
burdens from water and electricity consumption rank second and third, respectively.
The irrigation group takes up the greatest amount of fossil depletion and human
toxicity impact, but there was an obvious pattern difference between the two impacts. In
the fossil depletion category, the absent of irrigation reduces the total kg of oil equivalent
consumption, so the Nonirrigated Ctrl has the lowest impacts. In contrast, due to the lowest
yield of the Nonirrigated Ctrl, the human toxicity impact was significantly magnified by the
135

use of machinery and the seed group impacts. So, the nonirrigated Ctrl was the one with
highest impacts.

Figure 5- 8 ReCiPe 1.07 (H) Midpoint of Terrestrial Acidification
Figure 5-8 reveals that the irrigation electricity impacts the most to the terrestrial
acidification impacts. The contributions from machinery, irrigation water, drip tape
production, SWRT and seed groups are all at similar levels. A common trigger reason to
produce terrestrial acidification is fuel burning. When a petroleum based fuel is burned, a
lot of S and N gases are emitted, such as NO2, NOx and SO2. Thus, processes involving power
use incurred larger terrestrial acidification impact scores. The terrestrial acidification
impact from the planting group is different from the others since most of the impact
derives from fertilizer emissions.
136

Figure 5- 9 ReCiPe 1.07 (H) Midpoint of Water Depletion
The water depletion impact category takes both direct and indirect water use into
consideration. Irrigation water, the greatest one, is direct water use. The impacts from the
chemical, machinery and seed group are indirect water use.
Figure 5-9 shows that the SWRT treatments consumed more water to produce 1000 kg
corn grain than the Ctrl treatments in year 2012.
Most of the indirect water used is spent on water used for turbine cooling for
production of all chemical, machinery and seed groups. This is a characteristic of how
water depletion for ReCiPe 1.07 (H) is implemented in GaBi 6.0, which considers water for
cooling as part of the water depletion indicator.

137

5.2.2 Benchmark of published studies
To evaluate whether the LCIA results obtained in this study were in the acceptable
range, the results were compared with published LCA corn grain studies.
The LCA of corn grain and corn stover in the United States published by Kim and Dale [3]
was one of the main published studies used for comparison since this study’s results are
representative of the US Midwest production. The sampling datasets for this study were
from multiple locations, so it was possible to overcome the soil, climate, and management
variations across US. Corn production datasets used in this study were taken from eight
counties in seven different states producing the majority of US corn. Another important
published data source used for comparisons was available LCA databases. The LCI of
several corn studies are available in the PE international database, Ecoinvent 2.2v, and the
USLCI datasets, which have undergone different levels of review. An additional advantage
of comparing results of this study with corn grain datasets from databases is the flexibility
in adapting the assessment methodology. In contrast, not all published corn studies are 100%
LCI transparent about their assumptions, geographical and boundary conditions to afford
reproducing the studies. Different published studies reported their results using different
methodologies. Results calculated from different methodologies on the same impact area
cannot easily be compared without extensive reverse data engineering of the involved
process. Thus, the benchmark comparisons of this study with previous results are listed in
separate tables.
Table 5- 3 shows the results conducted by Kim and Dale [3]. Table 5- 4 shows the LCIA
from database and results of this study. By comparing Table 5- 3 and Table 5- 4, on average,

138

the air acidification and climate change impact values of this study results are 66% and 33%
respectively lower than the Kim and Dale published study. This might be owing to the less
gas emissions from fertilizer in sandy soil compared to loamy and clay soil. As discussed
before, the NH3, N2O are less often produced in sandy soil due to its high soil permeability.

139

Table 5- 3 Published corn grain study result for reference [3]
Hardin Fulton Tuscola Morrison Freeborn Macon Hamilton Codington
(IA)
(IL)
(MI)
(MN)
(MN)
(MO)
(NE)
(SD)
Air Acid.

RAINS Model
for site
specific factor

Eutrophic. CML
GHG

IPCC2001

Fossil
energy

Sum up fossil
fuel * lower
heating value

g SO2 eq
/kg grain
g PO4 3- eq
/kg grain
g CO2 eq/
kg grain
kJ /kg grain

4.1

6.5

5.9

6.2

3.2

7.4

4.5

4.7

1.0

1.4

1.3

1.3

0.7

1.8

0.9

0.9

344

457

451

470

474

785

430

342

473

330

605

581

581

439

1202

294

Table 5- 4 Comparison of LCIA from database and published studied results, and the 2012 and 2013 SWRT and Ctrl results

g SO2 eq/
kg grain
g PO4 3- eq/
kg grain

15"
30"
15"
SWRT SWRT Ctrl

30"
Ctrl

2.4

3.5

10.4

2.2

1.9

2.0

1.5

4.25

2.26

1.63

1.82

2.66 2.34 1.44

3.23

433

580

374

344

368

286

285

242

443

402

600

616

624

534

514

348

TRACI 2.1

Eutrophic.
Potential

CML2001 0.90
Nov. 2010
IPCC global
warming,
g CO2 eq/kg
excluding
240
grain
biogenic
carbon
TRACI 2.1
kJ / kg
225

Fossil

2013
NonIrrigated
irrigated
SWRT
Ctrl

Corn,
Corn,
whole
at farm
plant, at
[5]
field [6]

Air Acid.

Global
Warming

2012

US Corn
grain (field
border) [4]

140

1.5

1.7

5.3

Scenario comparisons
Several scenario comparisons were performed to predict the consequence of

changing a number of parameters according to different assumptions/ methods.
5.3.1 Yield increase scenario
Yield increase scenarios analysis was conducted on simulated Ctrl treatment. It
was assumed that in 2004, SWRT was initially installed on the Sandhill farm. The
total environmental burden for using SWRT was loaded to 2004. Then, it was
assumed that SWRT increased corn grain yield by 20%, 30%, 50%, 80%, 100%,
200%, and 300% compared to the Ctrl treatment yield, with the remaining the rest
input and output levels the same. The extra yield diluted the environmental burden
of corn grain production and reduced the impact. The impact differences between
SWRT and Ctrl are the credit for using SWRT. The time required to pay off the SWRT
initial burden was calculated. When the SWRT burden was no larger than the sum of
the SWRT benefits, the SWRT burdens were paid off. Figure 5-10 uses the climate
change impact as an example to explain the calculation process. The SWRT burden
was the difference between the original 2004 SWRT column and the original 2004
Ctrl column, which is the distance marked by the black arrow in Figure 5-10. If
SWRT increased yield by 20% every year from 2004 to 2013, the SWRT benefits
were the differences between the original column and the 20%+ column, which
were marked as the red color right parenthesis. Based on the calculations, until the
end of the growing season in 2013, the SWRT burden was still larger than the sum of
SWRT benefits. Then, the conclusion was drawn that if SWRT increased yield by
141

20%, it took more than 10 years to pay off SWRT burden when considering the
climate change impact.
A summary of time to pay off the SWRT burden according to the yield increasing
scenarios on selected eight impact categories using ReCiPe 1.07 (H) midpoint
indicator is shown in Table 5-5. Climate change, fossil depletion, and terrestrial
acidification are the impact categories that take relatively long times to pay off the
SWRT burden. If SWRT utilization increased yield by 100%, the SWRT burden can
be paid within 10 years. A 10-year lifetime is a conservative assumption towards
SWRT effective lifetime since polyethylene membranes are not impacted due to
biodegradation [7].

Climate Change [kg CO2 eqv]

250

original
80%+

200

150

20%+
100%+

30%+
200%+

50%+
300%+

Burden

Benefits

100

50

0

2004 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Ctrl SWRT

Figure 5- 10 Time to payoff SWRT burden on climate change impact

142

Table 5- 5 Time [year] to pay-off SWRT burden if yield increase due to SWRT
application
20%+ 30%+ 50%+ 80%+ 100%+ 200%+ 300%+
Agricultural land
occupation
Climate change
Fossil depletion
Freshwater
ecotoxicity
Freshwater
eutrophication
Human toxicity
Terrestrial
acidification
Water depletion

1
>10
>10

1
>10
>10

1
7
>10

1
5
>10

1
5
9

1
4
6

1
4
6

1

1

1

1

1

1

1

1
1

1
1

1
1

1
1

1
1

1
1

1
1

>10
1

9
1

7
1

5
1

5
1

4
1

3
1

5.3.2 Drip tape lifetime scenario
In the LCI chapter it was mentioned that the drip tapes were disposed every year
after harvest. After field observations and disposal pipe sample collection were
performed, it was observed that there were no issues of either broken tubes or
blocked pores after 1year of use.
As discussed in section 6.2.1- LCIA analysis of experiment treatments, drip tape
production and drip tape EOL processes retained a significant weight on the climate
change, fossil depletion, human toxicity, and terrestrial acidification impact
categories.
Therefore, an drip tape lifetime scenario is worthy of evaluation. Agricultural
experts, who were consulted, suggested that usually drip tape could be used for up
to 4 years. So, the drip tape lifetime was considered as a scenario for comparison
under the different experimental treatments. Scenario I is a 1 year lifetime of the
143

pipe, and scenario II is a 4 year lifetime. In scenario II, the main reductions are
observed in the climate change, fossil depletion, human toxicity, and terrestrial
acidification categories (as presented in Figure 5-11 to 5-14), while the other four
categories had little change.

144

kg CO2 eqv.

200
180
160
140
120
100
80
60
40
20
0

Machinery
Irrigation water
Irrigation pipe production
Seed

Chemical
Irrigation electricity
irrigation pipe EOL
SWRT

Figure 5- 11 Drip tape lifetime scenario on climate change

MJ Nonrewable resource

Machinery
Irrigation water
Irrigation pipe production

Chemical
Irrigation electricity
irrigation pipe EOL

60
50
40
30

20
10
0

Figure 5- 12 Drip tape lifetime scenario on fossil depletion
145

kg 1,4-DB eqv.

45

Machinery

Chemical

40

Irrigation water

Irrigation electricity

35

Irrigation pipe production

irrigation pipe EOL

30

Seed

SWRT

25
20
15
10
5

0

Figure 5- 13 Drip tape lifetime scenario on human toxicity

1.2

kg SO2 eqv.

1.0

Machinery
Irrigation water
Irrigation pipe production
Seed
Planting

Chemical
Irrigation electricity
irrigation pipe EOL
SWRT

0.8
0.6
0.4
0.2
0.0

Figure 5- 14 Drip tape lifetime scenario on terrestrial acidification
146

5.3.3 Scenarios regarding multifunctionality methods of allocation
Corn production activity produced corn grain and corn stover simultaneously.
Methods were needed to assign the environmental footprint of corn grain and
stover. The ISO 14040 and ISO 14044 [8, 9] provided guidelines that, whenever
possible, system expansion should be used to estimate the studied activity/ process
by a co-product being substituted. However, due to the overwhelming workload and
sometimes impossibility of dividing subprocesses by system expansion, the
allocation method is preferable. Allocation was used in this study was aimed to
partition and distribute the EFP of corn grain over the corn stover. In general,
allocation methods are performed based on physical relationships, such as mass,
energy, and economic value.
Allocation is one of the most controversial issues in the LCA methodology. There
are two aspects that lead an allocation method to be less convincing: the allocation
factor is an arbitrary value that is unable to represent the real EFP distributions; the
allocation fails to consider the displacement effects and additional treatment of the
co-products before displacement takes place [10]. Several studies of corn [11, 12]
reported that allocation methods are a highly sensitive parameter affecting the
results. The choice of allocation method can significantly influence and/or even shift
the results.
The mass allocation method was used to partition the EFP of corn grain from
whole corn plant production. To ensure the robustness of this study, scenarios of
solving the multifunctionality method were evaluated. The scenarios of methods

147

were: allocation all on grain, allocation by economic value, allocation by energy
content, allocation by mass, and system expansion.
The reason for the allocation all on grain scenario was based on the initial
purpose of corn planting. The field corn was planted for corn grain production. And
none of the stover was collected but left behind to protect the soil fertility for corn
growing in the coming years. Thus, it was reasonable to charge the total EFP on corn
grain, even though stover was produced. In the allocation all on grain scenario, the
allocation factor [AF] was 1.
Allocations by economic value, by energy content, and by mass are commonly
used physical relationships in allocation methods. The justification of not charging
the total EFP on grain was if was unnecessary to abandon the stover totally on the
field to maintain the soil sustainability [13]. In other words, the unnecessary stover
waste should not being accounted on the corn grain, and corn stover was
responsible for part of the EPF. As explained in detail in the section 5.9 calculation
procedures, the AF of corn grain are 0.79 for economic value, 0.63 for energy
content, and 0.5 for mass relation.
System expansion was performed to substitute the stover from the whole corn
plant. The EFP of collected stover was substituted by an equivalent amount of
switchgrass biomass production. Then, the substituted EFP was deducted from the
EPF of corn biomass production. The EFP left behind was expected to be the EFP of
corn grain.
As illustrated below in Figure 5-15, every 1,000 kg of corn grain produced
means approximately 1,000 kg of corn stover is produced at the same time [3]. A
148

general level of maximum stover collection is 50%, which is constrained by the
tolerable soil loss that stover collection can be allowed without causing adverse
effects on soil and water resources [13]. Therefore, 500 kg stover was collected for
stover ethanol production. With the stover to ethanol conversion rate being 0.3 L
ethanol / kg stover[14], about 150 L stover- ethanol was produced from 500 kg
stover. Assuming the function and quantity were identical between stover ethanol
and switchgrass ethanol, the same volume (150 L) ethanol required 375 kg
switchgrass biomass, in which the switchgrass to ethanol efficiency is 0.4 L ethanol/
kg feedstock [15]. Meanwhile, 77.25 kWh electricity was produced as a co-product
[15]. The co-produced electricity was less than the total electricity consumption
during the switchgrass ethanol conversion process. The assumption was made that
the co-produced electricity was fully utilized for ethanol. The inputs/outputs stock
of switchgrass production were derived from references [16, 17]. The inventory of
375 kg switchgrass biomass production was computed in Matlab Version R2010b
(Mathworks, Natick, MA). Then, the LCIA of 375 kg switchgrass biomass production,
named as EFP stover, was calculated from the stover inventory. The last step was to
follow Equation 5-1 to calculate the EFP of 1000 kg of corn grain by subtracting the
EFP stover from the EFP of 2000 kg corn biomass.
EFP 1000 kg corn grain = EFP 2000 kg corn biomass - EFP 375 kg switchgrass biomass

149

(Equation 5-1)

Corn biomass
2000 kg

Corn grain
1000 kg

Corn stover
1000 kg

50% left
500 kg

Switchgrass biomass
375 kg

50% collect
500 kg

Stover ethanol
150 L

Switchgrass
ethanol
150 L

Electricity
77.25 kWh

Figure 5-15 System expansion corn system
The results of different scenarios in solving multifunctionality on agricultural
land use impact are presented in Figure 5-16. The figure implies that using different
allocation methods cannot change the relative EFP order among treatments, but the
absolutely value is highly sensitive to the allocation method. Full results of the
allocation scenarios on the eight selected impact categories are in Appendix B5Allocation scenarios. The results indicate that the relative ranking conclusion of this
study was not affected by the allocation method.

150

Agricultural land occ. [m2*a]

2500

all on grain (100%)
economic (79%)

2000

energy (63%)
mass (50%)

1500

system expansion (whole corn plant - stover)

1000

500

0

15'' SWRT

30'' SWRT

15" Cotrol

30" Cotrol

Irrigated
SWRT

Non-irrigated
Ctrl

Figure 5- 16 Allocation scenarios on agricultural land occupation

5.3.4 LCA methodology scenario
The ReCiPe 1.07 (H) midpoint is the default assessment methodology in this
study. As suggested by the LCA guidelines for grains and oilseeds [18], the following
impact categories should be covered in the LCA grain studies: global warming, water
consumption, non-renewable resource, aquatic eutrophication, acidification,
ecotoxicity, and human toxicity. The ReCiPe 1.07 (H) Midpoint Methodology was
selected because it covered most of the recommended impact categories. As
discussed in section 2.4 Impact Assessment, the characterization factors are not
always identical in different impact assessment methodologies. Thus, employing
different assessment methodologies might result in divergent conclusions. To
151

address this concern, scenarios on climate change (kg CO2 equivalent) in five impact
assessment methodologies (ReCiPe 1.07 (H) Midpoint, Impact 2002+ v2.1, TRACI
2.1, CML 2001-Nov.2010, IPCC global warming include biogenic carbon ) were

Climate Change [kg CO2 equivlent]

evaluated.

2012 15'' SWRT

2012 15" Control

2012 30'' SWRT

2012 30" Control

2013 Irrigated SWRT

2013 Nonirrigated Ctrl

200
150
100

50
0
ReCiPe 1.07
I02+ v2.1 TRACI 2.1, CML2001 - Nov. IPCC global
Midpoint (H) - Global warming Global Warming 2010, Global
warming, incl
Climate change
500yr Air
Warming
biogenic carbon
Midpoint
Potential (GWP
100 years), incl
biogenic carbon

Figure 5- 17 Impact assessment methodology scenarios on climate change
Figure 5-17 shows the relative order among the treatments is slightly different
under different methodologies. For ReCiPe 1.07 (H) midpoint and TRACI 2.1
midpoint, in which modelling of biogenic CO2 is excluded, the differences among
treatments are relatively smaller than when using the Impact 2002+, CML 2001, and
IPCC methodologies that include credits for biogenic CO2 , in Gabi 6.0. Due to the
152

increased differences, the order of the climate change impact values is slightly
different. The order is shifted between irrigated SWRT and 30” Ctrl. This implies
that the choice of impact assessment methodology has an effect on the results. Thus,
it is important to clearly state the employed impact assessment methodology and
the version when drawing conclusions.

153

5.4

Uncertainty analyses
Uncertainty analysis was conducted in this study due to the number of

estimations and assumptions causing variations from the experiments and to
understand if they will affect the conclusions.
5.4.1 Data quality evaluation
To understand which datasets have high uncertainty, data quality evaluation
was performed first. A pedigree matrix [19] was used to assess dataset standard
deviation (SD). Then, Monte Carlo simulation was performed on the datasets with
high SD. The outputs of the Monte Carlo simulations were considered as mean of
LCIA and the SD of the simulated mean. The number of Monte Carlo simulation runs
was determined depending on whether the simulation results converged or not. If
fluctuations were observed, the number of simulation runs was increased.
5.4.2 Land use (LU)
The main results obtained in this study were calculated by using the mean value
of the experimental yields, which were used to calculate the LU values (parameters)
for each studied treatment. Since large uncertainty was found in yield values, these
were propagated in the LU parameter. As explained in the LCI calculation procedure
section, the LU was shown to be highly sensitive to the results. Therefore, Monte
Carlo simulations were performed on the LU values.
Instead of using a pedigree matrix method, SD values of yields were derived
from experiments. However, LUs directly participated in the flow calculations. Since

154

the SD of the LUs was unknown, a Monte Carlo simulation was performed to
estimate the SD of the LUs first.
It was assumed that the yield of the treatments followed a normal distribution
with the SD of the experiments. Based on Equation 5-1, the LUs of the treatments
were calculated from the randomly taken yields in the 95% confidence interval (CI).
In other words, the yields of each treatment were simulated 100,000 times. Then
100,000 LU values were computed from the simulated yields, together with the SD
of the LUs. Table 5-6 presents the LU means directly calculated from the yields, LU
means from the Monte Carlo simulation, and the SD of the simulated LU mean.
Table 5- 6 Mean and SD of LU
Experiment LU
mean [acre]
2012 15'' SWRT
0.1468
2012 30'' SWRT
0.1847
2012 15''Ctrl
0.2554
2012 30''Ctrl
0.2489
2013 Irrigated SWRT
0.1640
2013 Nonirrigated Ctrl 0.5546

Simulated LU
mean [acre]
0.1477
0.1867
0.2720
0.2552
0.1648
0.6073

SD of simulated
LU mean
0.0117
0.0198
0.103
0.0420
0.0110
0.0520

After the SD values of the LUs were estimated, another Monte Carlo simulation
was performed, using the simulated LU mean and the SD of the simulated LU mean.
The goal of this Monte Carlo simulation was to evaluate the LU uncertainty effect on
the LCIA results.
Figures 5-18 to 5- 25 show the Monte Carlo simulation results of the LU
uncertainty effect on LCIA values. A two tail Z-test was performed at the 95% CI. In
the same figure, columns containing the same letter indicate that there was no
155

significant difference between two compared mean values. Letter “a” indicates the
lowest.
Figure 5-18 implies that the experimental means and Monte Carlo means are
very close, which indicates that the simulations reproduced the results very well.
From the two tail comparisons, the 15” SWRT and Irrigated SWRT do not have
significant differences; there was no significant difference among 30” SWRT, 30” Ctrl,
and Irrigated SWRT; the highest and second highest land occupation treatment were
nonirrigated Ctrl and 15” Ctrl.

156

1600

1400
1200

Monte Carlo Mean

1000

c

800
600
400

d

Experiment Mean

a

b
ab

b

200
0
15'' SWRT

30'' SWRT 15" Control 30" Control

Irrigated Nonirrigated
SWRT
Ctrl

Figure 5- 18 Agricultural land occ. [m2 * a], columns with the different lower case
letters are statistically significant different at 95% CI, with a indicating the lowest,
using the simulated LU mean and SD of simulated LU mean
Experiment Mean
250
200

Monte Carlo Mean

bc
b

b
abc

ac

150

a

100
50
0
15'' SWRT

30'' SWRT

15" Control 30" Control

Irrigated
SWRT

Nonirrigated
Ctrl

Figure 5- 19 Climate change [kg CO2 eqv.], columns with the different lower case
letters are statistically significant different at 95% CI, with a indicating the lowest,
using the simulated LU mean and SD of simulated LU mean

157

Figure 5-19 shows that according to the Monte Carlo simulation runs there were
no differences among the 2012 treatment means and among the 2013 treatment
means. This indicates that SWRT and irrigation application did not increase the
climate change burden significantly. The relatively large SD of the 15” Ctrl and
Nonirrigated Ctrl were due to the large SD of their yields.
Figure 5-20 implies that the 2012 Ctrl treatments had relatively lower fossil
depletion impact than the 2012 SWRT treatments. The Nonirrigated Ctrl being
significantly lower than Irrigated SWRT suggested that SWRT combined with
irrigation treatment significantly increased the fossil depletion burden over the
treatments without SWRT and irrigation.
80.0
70.0

c

bc

Experiment Mean

c

Monte Carlo Mean

b

60.0

b

50.0

a

40.0
30.0
20.0
10.0
0.0

15'' SWRT

30'' SWRT

15" Control

30" Control

Irrigated
SWRT

Nonirrigated
Ctrl

Figure 5- 20 Fossil depletion [kg oil eqv.], columns with the different lower case
letters are statistically significant different at 95% CI, with a indicating the lowest,
using the simulated LU mean and SD of simulated LU mean

158

14.00
12.00

Monte Carlo Mean

10.00
8.00
6.00

d

Experiment Mean
cd

d

c

b

a

4.00
2.00
0.00

15'' SWRT

30'' SWRT 15" Control 30" Control

Irrigated
SWRT

Nonirrigated
Ctrl

Figure 5- 21 Freshwater ecotoxi. [Kg 1, 4 -DB eqv.], columns with the different lower
case letters are statistically significant different at 95% CI, with a indicating the
lowest, using the simulated LU mean and SD of simulated LU mean

0.14
0.12

d

Experiment Mean
Monte Carlo Mean
d

0.10
0.08
0.06

d

c
b
a

0.04
0.02
0.00
15'' SWRT

30'' SWRT

15" Control

30" Control

Irrigated
SWRT

Nonirrigated
Ctrl

Figure 5- 22 Freshwater eutrophication [kg P eqv.], columns with the different
lower case letters are statistically significant different at 95% CI, with a indicating
the lowest, using the simulated LU mean and SD of simulated LU mean

159

60
50
40
30

c

Experiment Mean
Monte Carlo Mean
bc
b

ab

ab

a

20
10
0

15'' SWRT

30'' SWRT 15" Control 30" Control

Irrigated Nonirrigated
SWRT
Ctrl

Figure 5- 23 Human toxicity. [Kg 1, 4 -DB eqv.], columns with the different lower
case letters are statistically significant different at 95% CI, with a indicating the
lowest, using the simulated LU mean and SD of simulated LU mean

1.2

Experiment Mean
Monte Carlo Mean
ab

ab

b
b

1.0

a

0.8

a

0.6
0.4
0.2
0.0

15'' SWRT

30'' SWRT

15" Control 30" Control

Irrigated
SWRT

Nonirrigated
Ctrl

Figure 5- 24 Terrestrial acidification [kg SO2 eqv.], columns with the different lower
case letters are statistically significant different at 95% CI, with a indicating the
lowest, using the simulated LU mean and SD of simulated LU mean
160

Figure 5-21 and 5-22 show a similar pattern. Irrigated SWRT is the lowest, and
Nonirrigated Ctrl, 2012 Ctrl treatments are the highest. 2012 Ctrl treatments are
slightly higher than SWRT treatments. The driving reason for these results is the
yield.
Figure 5-23 indicates that there is no significant difference among 2012
treatments on human toxicity, while in 2013 Nonirrigated Ctrl was significant
higher than Irrigated SWRT.
Limited difference among treatments is shown on the terrestrial acidification
impact indicator. Despite an obvious mean difference that can be seen between
15”Ctrl and 30” Ctrl, when considering the large SD of 15” Ctrl, they are statistically
the same. This changed the conclusions that was simply drawn from the
experimental mean. The comparisons between Irrigated SWRT and Nonirrigated
Ctrl are similar to comparisons between 15”Ctrl and 30” Ctrl, in which conclusions
were changed when considering the uncertainty of LU.

161

350

Experiment Mean
Monte Carlo Mean

cd

300
250

d

bcd
c

200

ab

a

150
100
50
0
15'' SWRT

30'' SWRT

15" Control 30" Control

Irrigated
SWRT

Nonirrigated
Ctrl

Figure 5- 25 Water depletion [m3], columns with the different lower case letters are
statistically significant different at 95% CI, with a indicating the lowest, using the
simulated LU mean and SD of simulated LU mean

Table 5- 7 Water consumption comparisons

2012 15'' SWRT
2012 30'' SWRT
2012 15''Ctrl
2012 30''Ctrl
2013 Irrigated SWRT
2013 Nonirrigated Ctrl

Direct water
Application level
[m3/ha]
4915 (6)
2644 (5)
1441 (3)
665 (2)
1890 (4)
0 (1)

Direct water
consumption
[m3/FU]
166 (6)
112 (5)
85 (4)
38 (2)
71 (3)
0 (1)

(Direct + Indirect)
water consumption
comparisons
d
c
bcd
a
ab
cd

Table 5-7 compares the water consumption of the experimental treatments. The
direct water application level is the irrigation water given to the plant based on
observations and experience. The number in brackets right after each number is the
order of the value in that column, with (1) indicating the lowest. The direct water
162

consumptions in the unit of [m3/FU] represent the water consumption per FU. This
parameter not only considers the absolute water used, but also indicates a relatively
water spent when considering the yield effect. The last column summarizes the
statistic difference according to two-tail Z-test comparisons, which has the identical
letters in Figure 5-25. Since the total water consumption includes both direct and
indirect water use, the yield effect is intensified.
By comparing the yields and the irrigation applied to the 2012 SWRT treatments
and Ctrl treatments, it can be observed that over-irrigation is not recommended. In
contrast, no irrigation cut off the direct water use at the beginning, but the low yield
led to high cost in the indirect water consumption. The Nonirrigated Ctrl is a good
example, which ranks the lowest in direct water comparisons while ranking the
highest in total water comparisons. Thus, a suitable irrigation level is both
economical and environmentally preferable, such as irrigated SWRT and 30” Ctrl.
5.4.3 Monte Carlo simulations based on the pedigree matrix
According to SD estimated from the pedigree matrix method (as shown in
Appendix 5C-Pedigree matrix Table 5C-1), the diesel consumption rate of SWRT
installation, drip tape production, and seed were selected to be evaluated by the
Monte Carlo method. The results from the Monte Carlo simulations implied that
despite the great uncertainty with these parameters, they all had limited effect on
the LCIA results (as detailed in Appendix B5 Pedigree Matrix Table B5-1-a, B5-1-b,
and B5-1-c)

163

APPENDICES

164

Appendix A5- Contribution analyses
The contribution analyses were conducted using the ReCiPe 1.07 (H) Midpoint
indicator methodology. Eight impact indicators were chosen: agricultural land
occupation, climate change, fossil depletion, freshwater ecotoxicity, human toxicity,
terrestrial acidification, and water depletion. The LCIA values for each flow are
presented in Appendix A5-1, and the relatively contributions (%) are presented in
appendix A5-2 in total 8 treatments.

165

A5-1 LCIA value of each flow
LCIA values using ReCiPe 1.07 (H) midpoint methodology for each flow of the four treatments conducted in 2012, two
treatments in 2013, and two treatments simulated between 2004 and 2013 are documented in this section.

Indicator
Unit
total
harvest
fertilizing
Machine
sowing
tillage
gypsum
roundup
AMS
KCl
Chemical
P2O5
urea
AN
K2O

Table A5- 1 LCIA of 2012 15” SWRT for contribution analysis
Agri.
Climate Fossil
Freshwater Freshwater Human
land occ. change
depletion ecotoxi.
eutro.
toxicity
kg CO2kg 1,4-DB
kg 1,42
m *year Equiv.
kg oil eq
eq
kg P eq
DB eq
328.6
187.3
66.0
5.12
4.63E-02
23.6
0.163
4.59
1.51
0.0626
6.22E-04
3.01
0.0323
0.753
0.252
9.64E-03
8.56E-05
0.459
0.0964
0.676
0.222
8.72E-03
1.26E-04
0.406
0.198
4.02
1.32
0.0535
3.70E-04
2.33
7.42E-03 0.17
0.0465
1.44E-05
2.10E-07
3.82E-03
0.0201
1.05
0.373
0.0136
5.33E-04
0.691
2.20E-04 0.179
0.0939
3.51E-05
2.77E-07
1.48E-03
0.0734
1.66
0.589
0.0138
5.76E-04
0.816
5.84E-03 1.78
0.735
1.43E-03
1.38E-05
0.0736
0.0119
6.13
2.36
7.89E-04
5.78E-06
0.0565
1.63E-03 1.47
0.303
1.39E-04
1.16E-06
5.58E-03
2.13E-04 0.0991
0.0438
3.40E-06
1.59E-08
8.26E-04

166

Terrestrial Water
acid.
deplete.
kg SO2 eq
0.994
0.0396
5.65E-03
4.58E-03
0.0286
9.27E-04
0.0113
4.72E-04
5.51E-03
0.0105
0.0141
1.67E-03
2.32E-04

m3
229.7
8.12
1.17
1.72
5.72
0.187
6.2
0.0168
4.13
0.558
1.63
0.268
0.0158

Seed
SWRT

Planting

Indicator
drip
irrigation
electricity
drip tape
pipe landfill
pipe
incineration
pipe recycle
seed
PE
membrane
membrane
installation
Field
Preparation
Harvest

Agri.
Climate
land occ. change

Table A5- 1 (cont’d)
Fossil
Freshwater Freshwater Human
depletion ecotoxi.
eutro.
toxicity

Terrestrial Water
acid.
deplete.

0
0
0
8.51E-04

0
60.9
22.8
0.124

0
21
16.2
0.0392

0
0.272
4.93E-04
9.22E-03

0
0
4.38E-06
2.38E-04

0
6.31
0.028
1.54

0
0.495
0.0804
4.04E-04

0
0
0.533
0.0536

5.20E-04
1.67E-04
30.8

19
0.121
13.5

0.0485
0.0597
1.86

2.67E-05
2.25E-04
0.164

6.39E-08
1.96E-06
2.74E-03

5.47E-03
5.92E-03
3.43

1.41E-03
6.28E-04
0.0642

0.275
0.106
32.3

0

7.5

5.58

4.24E-04

3.80E-06

0.0109

0.0332

0.159

0.0785

6.61

2.19

7.84E-04

2.46E-05

0.114

0.0461

0.488

297
0

0
0

0
0

4.09
0

0.0372
0

0
0

0.0715
0

0
0

167

Indicator
Unit
total
Machine

harvesting
fertilizing
sowing
tillage
gypsum
roundup

Chemical

AMS
KCl
P2O5
urea
AN
K2O

Table A5- 2 LCIA of 2012 30” SWRT for contribution analysis
Agri.
Climate Fossil
Freshwater Freshwater Human
land occ. change depletion ecotoxi.
eutro.
toxicity
kg CO2kg 1,4-DB
kg 1,4m2*year Equiv.
kg oil eq
eq
kg P eq
DB eq
394.4
172.0
63.2
5.94
5.44E-02
21.3
0.205
5.78
1.91
0.0787
7.82E-04
3.79
0.0406
0.947
0.317
1.21E-02
1.08E-04
0.578
0.121
0.85
0.279
1.10E-02
1.58E-04
0.511
0.249
5.06
1.66
0.0673
4.65E-04
2.93
4.80E9.34E-03 0.214
0.0585
1.81E-05
2.64E-07
03
0.0253
1.32
0.469
0.0171
6.71E-04
0.869
1.87E2.77E-04 0.226
0.118
4.42E-05
3.49E-07
03
0.0924
2.09
0.742
0.0173
7.25E-04
1.03
7.35E-03 2.24
0.925
1.80E-03
1.74E-05
0.0926
0.015
7.72
2.97
9.93E-04
7.27E-06
0.0711
7.02E2.05E-03 1.85
0.381
1.75E-04
1.46E-06
03
1.04E2.68E-04 0.125
0.0551
4.27E-06
2.00E-08
03

168

Terrestrial Water
acid.
depletion
kg SO2 eq
0.877
0.0498
7.11E-03
5.76E-03
0.036

m3
171.7
10.2
1.47
2.17
7.2

1.17E-03
0.0142

0.235
7.8

5.94E-04
6.94E-03
0.0132
0.0177

0.0212
5.2
0.702
2.05

2.11E-03

0.337

2.92E-04

0.0199

Indicator
Unit
irrigation
water
drip
irrigation
electricity
Irrigation drip tape
pipe landfill
pipe
incineration

Seed
SWRT

Planting

pipe recycle
seed
PE
membrane
membrane
installation
Field
Preparation
Harvest

Table A5- 2 (cont’d)
Agri.
Climate Fossil
Freshwater Freshwater Human Terrestrial Water
land occ. change depletion ecotoxi.
eutro.
toxicity acid.
depletion
kg CO2kg 1,4-DB
kg 1,4m2*year Equiv.
kg oil eq
eq
kg P eq
DB eq
kg SO2 eq
m3

0.0906

23.1

7.56

0.283

2.55E-03

2.88

0.0526

112

0
0
0
1.07E-03

0
41.3
28.7
0.156

0
14.2
20.4
0.0494

0
0.184
6.21E-04
1.16E-02

0
0
5.52E-06
3.00E-04

0
0.335
0.101
5.08E-04

0
0
0.671
0.0675

6.54E-04

23.9

0.061

3.36E-05

8.04E-08

1.77E-03

0.346

2.10E-04
19.4

0.153
8.47

0.0751
1.17

2.83E-04
0.103

2.46E-06
1.73E-03

0
4.27
0.0352
1.94
6.89E03
7.44E03
2.16

7.90E-04
0.0404

0.133
20.3

0

9.44

7.02

5.34E-04

4.78E-06

0.0138

0.0417

0.2

0.0988

8.32

2.75

9.86E-04

3.10E-05

0.143

0.0581

0.614

374
0

0
0

0
0

5.15
0

0.0468
0

0
0

0.09
0

0
0

169

Indicator
Unit
total
harvesting
fertilizing
Machine
sowing
tillage
gypsum
roundup
AMS
KCl
Chem.
P2O5
urea
AN
K2O

Table A5- 3 LCIA of 2012 15” Ctrl for contribution analysis
Agri.
Climate Fossil
Freshwater Freshwater Human
land occ. change depletion ecotoxi.
eutro.
toxicity
kg CO2kg 1,4-DB
kg 1,4-DB
m2*year
Equiv.
kg oil eq
eq
kg P eq
eq
570.5
184.2
61.7
8.03
7.57E-02
27.9
0.283
7.97
2.63
0.109
1.08E-03
5.22
0.056
1.31
0.437
0.0167
1.48E-04
0.797
0.167
1.17
0.384
0.0151
2.18E-04
0.705
0.362
7.35
2.42
0.0977
6.75E-04
4.26
0.0129
0.295
0.0807
2.49E-05
3.64E-07
6.62E-03
0.0349
1.82
0.647
0.0236
9.25E-04
1.20
3.81E-04 0.311
0.163
6.09E-05
4.81E-07
2.57E-03
0.127
2.88
1.02
0.0239
0.001
1.42
0.0101
3.08
1.27
2.48E-03
2.39E-05
0.128
0.0206
10.6
4.10
1.37E-03
1.00E-05
0.098
2.83E-03 2.54
0.526
2.41E-04
2.01E-06
9.68E-03
3.70E-04 0.172
0.0759
5.89E-06
2.75E-08
1.43E-03

170

Terrestrial Water
acid.
deplete.
kg SO2 eq
0.886
0.0687
9.80E-03
7.94E-03
0.0523
1.61E-03
0.0196
8.18E-04
9.56E-03
0.0181
0.0244
2.90E-03
4.02E-04

m3
194.3
14.1
2.03
2.99
10.4
0.324
10.8
0.0292
7.16
0.968
2.82
0.465
0.0274

Indicator
Unit
irrigation
water
drip
irrigation
electricity
Irrigate
drip tape
pipe landfill
pipe
incineration
pipe recycle
Seed
seed
PE
membrane
SWRT
membrane
installation
Field
Planting Preparation
Harvest

m2*year

Table A5- 3 (cont’d)
Climate Fossil
Freshwater Freshwater Human
change depletion ecotoxi.
eutro.
toxicity
kg CO2kg 1,4-DB
kg 1,4-DB
Equiv.
kg oil eq
eq
kg P eq
eq

0.068

17.4

5.68

0.213

1.92E-03

0
0
0
1.48E-03

0
31.0
39.6
0.215

0
10.7
28.1
0.068

0
0.139
8.56E-04
0.016

9.02E-04
2.90E-04
53.4

33.0
0.210
23.3

0.0841
0.104
3.22

0

0

0
516
0

Agri.
land occ.

Terrestrial Water
acid.
deplete.
kg SO2 eq

m3

2.16

3.95E-02

84.5

0
0
7.61E-06
4.13E-04

0
3.21
0.0485
2.67

0
0.252
0.139
7.00E-04

0
0
0.924
0.093

4.63E-05
3.90E-04
0.284

1.11E-07
3.40E-06
4.76E-03

0.0095
0.0103
5.95

2.44E-03
1.09E-03
0.111

0.476
0.184
56.0

0

0

0

0

0

0

0

0

0

0

0

0

0

0
0

0
0

7.09
0

0.0645
0

0
0

0.124
0

0
0

171

Indicator
Unit
total
harvesting
fertilizing
Machine
sowing
tillage
gypsum
roundup
AMS
KCl
Chem.
P2O5
urea
AN
K2O

Table A5- 4 LCIA of 2012 30” Ctrl for contribution analysis
Agri.
Climate Fossil
Freshwater Freshwater Human
land occ. change depletion ecotoxi.
eutro.
toxicity
kg CO2kg 1,4-DB
kg 1,4m2*year
Equiv.
kg oil eq
eq
kg P eq
DB eq
531.2
143.1
50.1
7.53
7.06E-02
21.5
0.276
7.78
2.56
0.106
1.05E-03
5.10
0.0547
1.28
0.426
0.0163
1.45E-04
0.778
0.163
1.14
0.375
0.0148
2.13E-04
0.688
0.353
7.18
2.36
0.0954
6.59E-04
4.16
0.0126
0.288
0.0788
2.43E-05
3.55E-07
6.47E-03
0.0341
1.78
0.632
0.0231
9.04E-04
1.17
3.72E-04 0.304
0.159
5.95E-05
4.69E-07
2.51E-03
0.124
2.81
0.998
0.0233
9.76E-04
1.38
9.89E-03 3.01
1.24
2.42E-03
2.34E-05
0.125
0.0202
10.4
4.00
1.34E-03
9.78E-06
0.0957
2.76E-03 2.48
0.513
2.35E-04
1.96E-06
9.45E-03
3.61E-04 0.168
0.0741
5.75E-06
2.69E-08
1.40E-03

172

Terrestrial Water
acid.
deplete.
kg SO2 eq
0.657
0.0671
9.57E-03
7.75E-03
0.0511
0.00157
0.0191
7.99E-04
9.34E-03
0.0177
0.0239
2.83E-03
3.93E-04

m3
118.1
13.8
1.98
2.92
10.2
0.316
10.5
0.0285
6.99
0.945
2.76
0.454
0.0268

Indicator

Irrigate

Seed
SWRT

Plant

Unit
irrigation
water
drip
irrigation
electricity
drip tape
pipe landfill
pipe
incineration
pipe recycle
seed
PE
membrane
membrane
installation
Field
Preparation
Harvest

m2*year

Table A5- 4 (cont’d)
Climate Fossil
Freshwater Freshwater Human
change depletion ecotoxi.
eutro.
toxicity
kg CO2kg 1,4-DB
kg 1,4Equiv.
kg oil eq
eq
kg P eq
DB eq

0.0306

7.83

2.56

0.0959

8.64E-04

0
0
0
0.00144

0
14.0
38.6
0.21

0
4.81
27.5
0.0664

0
6.24E-02
8.36E-04
0.0156

8.81E-04
2.83E-04
26.1

32.2
0.205
11.4

0.0821
0.101
1.57

0

0

0
504
0

Agri.
land occ.

Terrestrial Water
acid.
deplete.
kg SO2 eq

m3

0.975

0.0178

38.1

0
0
7.43E-06
4.03E-04

0
1.44
0.0473
2.61

0
0.113
1.36E-01
6.84E-04

0
0
0.902
0.0908

4.52E-05
3.81E-04
0.139

1.08E-07
3.32E-06
2.32E-03

9.27E-03
0.01
2.90

2.38E-03
1.06E-03
0.0544

0.465
0.18
27.4

0

0

0

0

0

0

0

0

0

0

0

0

0

0
0

0
0

6.93
0

0.063
0

0
0

0.121
0

0
0

173

Indicator
Unit
total
Machine

Chem.

harvesting
fertilizing
sowing
tillage
gypsum
roundup
AMS
KCl
P2O5
urea
AN
K2O

Table A5- 5 LCIA of 2013 Irrigated SWRT for contribution analysis
Agri.
Climate Fossil
Freshwater Freshwater Human
land occ. change depletion ecotoxi.
eutro.
toxicity
kg CO2kg 1,4-DB
kg 1,4-DB
m2*year
Equiv.
kg oil eq
eq
kg P eq
eq
367.1
142.6
50.9
3.25
3.10E-02
18.9
0.179
5.06
1.69
0.0698
6.94E-04
3.35749
0.0356
0.83
0.281
1.08E-02
9.55E-05
0.51238
0.107
0.747
0.247
9.73E-03
1.40E-04
0.45331
0.219
4.43
1.48
0.0597
4.13E-04
2.60029
8.28E-03 0.19
0.0519
1.60E-05
2.34E-07
4.26E-03
0.0224
1.17
0.416
0.0152
5.95E-04
0.771
2.45E-04 0.2
0.105
3.92E-05
3.09E-07
1.65E-03
0.0736
1.66
0.591
0.0138
5.78E-04
0.818
3.69E-03 1.12
0.464
9.01E-04
8.71E-06
0.0465
0.00994
5.12
1.97
6.59E-04
4.82E-06
0.0472
3.35E-03 3.01
0.623
2.86E-04
2.38E-06
1.15E-02
8.42E-04 0.391
0.173
1.34E-05
6.27E-08
3.26E-03

174

Terrestrial Water
acid.
deplete
kg SO2 eq
0.661
0.0442
6.30E-03
5.11E-03
0.0319
1.03E-03
0.0126
5.26E-04
5.53E-03
0.0066
0.0118
3.44E-03
9.17E-04

m3
141.4
9.06
1.3
1.92
6.38
0.208
6.92
0.0188
4.14
0.352
1.36
0.551
0.0625

Irrigate

Seed
SWRT

Plant

Indicator

Agri.
land occ.

Unit

m2*year

Table A5- 5 (cont’d)
Climate Fossil
Freshwater Freshwater Human
change depletion ecotoxi.
eutro.
toxicity
kg CO2kg 1,4-DB
kg 1,4-DB
Equiv.
kg oil eq
eq
kg P eq
eq

0.0574

14.7

4.79

0.18

1.62E-03

1.83

0
0
0
8.24E-04

0
26.2
25.4
0.12

0
9.01
18.1
0.0438

0
0.117
5.50E-04
1.03E-02

0
0
4.89E-06
2.66E-04

0
0
2.71
0.212
0.0312
0.0897
(cont’d)
1.72
4.50E-04

0
0
0.594
0.0598

5.80E-04
4.50E-04
34.3

21.2
0.327
15

0.0541
0.0666
2.07

2.98E-05
2.51E-04
0.183

7.13E-08
2.18E-06
3.06E-03

6.11E-03
6.60E-03
3.83

1.57E-03
7.01E-04
0.0716

0.306
0.118
36

0

8.37

6.22

4.73E-04

4.23E-06

0.0122

0.037

0.177

0.0876

7.38

2.44

8.74E-04

2.75E-05

0.127

0.0515

0.544

332
0

0
0

0
0

2.58
0

0.0235
0

0
0

0.0337
0

0
0

irrigation
water
drip
irrigation
electricity
drip tape
pipe landfill
pipe
incineration
pipe recycle
seed
PE
membrane
membrane
installation
Field
Preparation
Harvest

175

Terrestrial Water
acid.
deplete
kg SO2 eq

m3

0.0333

71.3

Indicator
Unit
total
harvesting
fertilizing
Machine
sowing
tillage
gypsum
roundup
AMS
KCl
Chem.
P2O5
urea
AN
K2O

Table A5- 6 LCIA of 2013 Nonirrigated SWRT for contribution analysis
Agri.
Climate Fossil
Freshwater Freshwater Human
land occ. change
deplete.
ecotoxi.
eutro.
toxicity
kg CO2kg 1,4-DB
kg 1,4-DB
m2*year
Equiv.
kg oil eq
eq
kg P eq
eq
1238.3
121.1
32.0
9.98
9.84E-02
42.5
0.607
17.1
5.72
0.236
2.35E-03
11.40
0.121
2.81
0.95
0.0364
3.23E-04
1.73
0.363
2.53
0.836
0.0329
4.74E-04
1.53
0.78
15.8
5.26
0.213
1.47E-03
9.26
0.028
0.641
0.176
5.42E-05
7.91E-07
0.0144
0.0759
3.97
1.41
0.0514
2.01E-03
2.61
8.30E-04 0.677
0.355
1.33E-04
1.05E-06
0.0056
0.249
5.63
2.00
0.0467
1.96E-03
2.77
0.0125
3.80
1.57
3.05E-03
2.95E-05
0.157
0.0336
17.3
6.68
2.23E-03
1.63E-05
0.160
0
0
0
0
0
0
0
0
0
0
0
0

176

Terrestrial Water
acid.
deplete
kg SO2 eq
0.787
0.15
0.0213
0.0173
0.114
0.0035
0.0427
0.00178
0.0187
0.0223
0.0398
0
0

m3
230.3
30.7
4.41
6.50
22.7
0.705
23.4
0.0636
14.0
1.19
4.6
0
0

Indicator

Irrigate

Seed
SWRT

Plant

Unit
irrigation
water
drip
irrigation
electricity
drip tape
pipe landfill
pipe
incineration
pipe recycle
seed
PE
membrane
membrane
installation
Field
Preparation
Harvest

m2*year

Climate
change
kg CO2Equiv.

Table A5- 6 (cont’d)
Fossil
Freshwater Freshwater Human
deplete.
ecotoxi.
eutro.
toxicity
kg 1,4-DB
kg 1,4-DB
kg oil eq
eq
kg P eq
eq

0

0

0

0

0

0
0
0
0

0
0
0
0

0
0
0
0

0
0
0
0

0
0
116

0
0
50.8

0
0
7.01

0

0

0
1120
0

Agri.
land occ.

Terrestrial Water
acid.
deplete
kg SO2 eq

m3

0

0

0

0
0
0
0

0
0
0
0

0
0
0
0

0
0
0
0

0
0
0.618

0
0
0.0104

0
0
12.90

0
0
0.242

0
0
122.0

0

0

0

0

0

0

0

0

0

0

0

0

0

0
0

0
0

8.74
0

0.0794
0

0
0

0.114
0

0
0

177

Indicator
Unit
total
harvesting
fertilizing
Machine
sowing
tillage
gypsum
roundup
AMS
Chem.
K2O
P2O5
urea
irrigation
water
drip
irrigation
electricity
Irrigate drip tape

Table A5- 7 LCIA of simulated 2004 Ctrl for contribution analysis
Agri.
Climate Fossil
Freshwater Freshwater Human
land occ. change
deplete
ecotoxi.
eutro.
toxicity
kg CO2kg 1,4-DB
kg 1,4-DB
m2*year
Equiv.
kg oil eq
eq
kg P eq
eq
456.6
53.4
15.3
4.09
3.94E-02
15.0
0.223
6.31
2.10
0.0869
8.64E-04
4.18
0.0444
1.03
0.35
0.0134
1.19E-04
0.638
0.134
0.93
0.308
0.0121
1.75E-04
0.565
0.287
5.81
1.94
0.0783
5.41E-04
3.41
0.0103
0.236
0.0646
1.99E-05
2.91E-07
5.31E-03
0.0279
1.46
0.518
0.0189
7.41E-04
0.960
3.06E-04 0.249
0.131
4.88E-05
3.85E-07
2.06E-03
7.41E-04 0.344
0.152
1.18E-05
5.51E-08
2.87E-03
5.20E-03 1.58
0.654
1.27E-03
1.23E-05
0.0655
0.0206
10.6
4.10
1.37E-03
1.00E-05
0.098

Terrestrial Water
acid.
deplete
kg SO2 eq
0.337
0.055
7.85E-03
6.36E-03
0.0419
0.00129
0.0157
6.56E-04
8.06E-04
9.31E-03
0.0244

m3
81.9
11.3
1.62
2.39
8.37
0.259
8.61
0.0234
0.0549
0.497
2.82

7.16E-04

0.183

0.0597

2.24E-03

2.02E-05

0.0228

4.15E-04

0.889

0
0
0

0
0.326
3.07

0
0.112
2.19

0
1.46E-03
6.64E-05

0
0
5.91E-07

0
0.0337
3.77E-03

0
2.65E-03
0.0108

0.0145

5.28E-03

1.24E-03

3.21E-05

0.208

5.44E-05

0
0
0.0718
7.22E03

2.56
0.0394

6.53E-03
8.04E-03

3.60E-06
3.03E-05

8.61E-09
2.64E-07

7.37E-04
7.97E-04

1.90E-04
8.46E-05

0.037
0.0143

pipe landfill 9.95E-05
pipe
incineration 7.00E-05
pipe recycle 5.44E-05

178

Seed
SWRT

Plant

m2*year

Climate
change
kg CO2Equiv.

Table A5- 7 (cont’d)
Fossil
Freshwater Freshwater Human
deplete
ecotoxi.
eutro.
toxicity
kg 1,4-DB
kg 1,4-DB
kg oil eq
eq
kg P eq
eq

42.8

18.7

2.58

0.227

3.81E-03

0

0

0

0

0

0

0

413
0

0
0

0
0

Indicator

Agri.
land occ.

Unit
seed
PE
membrane
membrane
installation
Field
Preparation
Harvest

Terrestrial Water
acid.
deplete
kg SO2 eq

m3

4.77

0.0892

44.9

0

0

0

0

0

0

0

0

0

3.64
0

0.0331
0

0
0

0.070
0

0
0

179

Indicator
Unit
total
harvesting
fertilizing
Machine
sowing
tillage
gypsum
roundup
AMS
Chemical
K2O
P2O5
urea
irrigation
water
drip
irrigation
electricity
Irrigate
drip tape

Table A5- 7 LCIA of simulated 2004 SWRT for contribution analysis
Agri.
Climate Fossil
Freshwater Freshwater Human
land occ. change deplete
ecotoxi.
eutro.
toxicity
kg CO2kg 1,4-DB
kg 1,4-DB
m2*year
Equiv.
kg oil eq
eq
kg P eq
eq
457.5
246.6
122.2
4.06
3.95E-02
15.0
0.223
6.38
2.10
0.0869
8.64E-04
4.18
0.0444
1.05
0.35
0.0134
1.19E-04
0.638
0.134
0.939
0.308
0.0121
1.75E-04
0.565
0.144
2.94
0.968
0.0391
2.70E-04
1.70
0.0103
0.236
0.0646
1.99E-05
2.91E-07
5.31E-03
0.0279
1.46
0.518
0.0189
7.41E-04
0.960
3.06E-04 0.249
0.131
4.88E-05
3.85E-07
2.06E-03
7.41E-04 0.344
0.152
1.18E-05
5.51E-08
2.87E-03
5.20E-03 1.58
0.654
1.27E-03
1.23E-05
0.0655
0.0206
10.6
4.10
1.37E-03
1.00E-05
0.098

Terrestrial Water
acid.
deplete
kg SO2 eq
1.418
0.055
7.85E-03
6.36E-03
0.0209
0.00129
0.0157
6.56E-04
8.06E-04
9.31E-03
0.0244

m3
86.7
11.3
1.62
2.39
4.18
0.259
8.61
0.0234
0.0549
0.497
2.82

7.16E-04

0.183

0.0597

2.24E-03

2.02E-05

0.0228

4.15E-04

0.889

0
0
0

0
0.326
3.07

0
0.112
2.19

0
1.46E-03
6.64E-05

0
0
5.91E-07

0
0.0337
3.77E-03

0
2.65E-03
0.0108

0.0167

5.28E-03

1.24E-03

3.21E-05

0.208

5.44E-05

0
0
0.0718
7.22E03

2.56
0.0163

6.53E-03
8.04E-03

3.60E-06
3.03E-05

8.61E-09
2.64E-07

7.37E-04
7.97E-04

1.90E-04
8.46E-05

0.037
0.0143

pipe landfill 9.95E-05
pipe
incineration 7.00E-05
pipe recycle 5.44E-05

180

Seed
SWRT

Plant

Indicator

Agri.
land occ.

Unit

m2*year

seed
PE
membrane
membrane
installation
Field
Preparation
Harvest

Climate
change
kg CO2Equiv.

Table A5- 8 (cont’d)
Fossil
Freshwater Freshwater Human
deplete
ecotoxi.
eutro.
toxicity
kg 1,4-DB
kg 1,4-DB
kg oil eq
eq
kg P eq
eq

Terrestrial Water
acid.
deplete
kg SO2 eq

m3

42.8

18.7

2.58

0.227

3.81E-03

4.77

0.0892

44.9

0

104

77.5

5.89E-03

5.27E-05

0.152

0.461

2.21

1.09

91.9

30.4

0.0109

3.42E-04

1.58

0.641

6.78

413
0

0
0

0
0

3.64
0

0.0331
0

0
0

0.070
0

0
0

181

B-2

Relative contribution of each flow

Indicator

Machine

Chem.

harvest
fertilizing
sowing
tillage
gypsum
roundup
AMS
KCl
P2O5
urea
AN
K2O

Table A5- 9 Flow relative contributions (%) of 2012 15” SWRT
Agri.
Climate Fossil
Freshwater Freshwater Human
land
change
deplete ecotoxi.
eutro.
toxicity
occ.
%
%
%
%
%
%
0.05
2.45
2.29
1.22
1.34
12.78
0.01
0.40
0.38
0.19
0.18
1.95
0.03
0.36
0.34
0.17
0.27
1.72
0.06
2.15
2.00
1.04
0.80
9.89
0.00
0.09
0.07
0.00
0.00
0.02
0.01
0.56
0.56
0.27
1.15
2.93
0.00
0.10
0.14
0.00
0.00
0.01
0.02
0.89
0.89
0.27
1.24
3.46
0.00
0.95
1.11
0.03
0.03
0.31
0.00
3.27
3.57
0.02
0.01
0.24
0.00
0.78
0.46
0.00
0.00
0.02
0.00
0.05
0.07
0.00
0.00
0.00

182

Terrestrial Water
acid.
deplete
%
3.99
0.57
0.46
2.88
0.09
1.14
0.05
0.55
1.06
1.42
0.17
0.02

%
3.54
0.51
0.75
2.49
0.08
2.70
0.01
1.80
0.24
0.71
0.12
0.01

Table A5- 9 (cont'd)
Indicator

Irrigate

Seed
SWRT

Plant

irrigation
water
drip
irrigation
electricity
drip tape
pipe landfill
pipe
incineration
pipe recycle
seed
PE
membrane
membrane
installation
Field
Preparation
Harvest

Agri.
land
occ.
%

Climate
change

Fossil
Freshwater Freshwater Human
deplete ecotoxi.
eutro.
toxicity

Terrestrial Water
acid.
deplete

%

%

%

%

%

%

%

0.04

18.26

16.96

8.18

8.14

18.08

7.81

72.28

0.00
0.00
0.00
0.00

0.00
32.51
12.17
0.07

0.00
31.81
24.54
0.06

0.00
5.31
0.01
0.18

0.00
0.00
0.01
0.51

0.00
26.78
0.12
6.54

0.00
49.82
8.09
0.04

0.00
0.00
0.23
0.02

0.00
0.00
9.37

10.14
0.06
7.21

0.07
0.09
2.82

0.00
0.00
3.20

0.00
0.00
5.92

0.02
0.03
14.56

0.14
0.06
6.46

0.12
0.05
14.06

0.00

4.00

8.45

0.01

0.01

0.05

3.34

0.07

0.02

3.53

3.32

0.02

0.05

0.48

4.64

0.21

90.38
0.00

0.00
0.00

0.00
0.00

79.88
0.00

80.32
0.00

0.00
0.00

7.20
0.00

0.00
0.00

183

Indicator
harvest
fertilizing
Machine
sowing
tillage
gypsum
roundup
AMS
KCl
Chem.
P2O5
urea
AN
K2O

Table A5- 10 Flow relative contributions (%) of 2012 30” SWRT
Agri.
Climate Fossil
Freshwater Freshwater Human
land occ. change
deplete ecotoxi.
eutro.
toxicity
%
%
%
%
%
%
0.05
3.36
3.02
1.32
1.44
17.76
0.01
0.55
0.50
0.20
0.20
2.71
0.03
0.49
0.44
0.19
0.29
2.39
0.06
2.94
2.63
1.13
0.86
13.73
0.00
0.12
0.09
0.00
0.00
0.02
0.01
0.77
0.74
0.29
1.23
4.07
0.00
0.13
0.19
0.00
0.00
0.01
0.02
1.22
1.17
0.29
1.33
4.83
0.00
1.30
1.46
0.03
0.03
0.43
0.00
4.49
4.70
0.02
0.01
0.33
0.00
1.08
0.60
0.00
0.00
0.03
0.00
0.07
0.09
0.00
0.00
0.00

184

Terrestrial
acid.
%
5.68
0.81
0.66
4.11
0.13
1.62
0.07
0.79
1.51
2.02
0.24
0.03

Water
deplete
%
5.94
0.86
1.26
4.19
0.14
4.54
0.01
3.03
0.41
1.19
0.20
0.01

Table A5-10 (cont'd)
Indicator

Irrigate

Seed
SWRT

Plant

irrigation
water
drip
irrigation
electricity
drip tape
pipe landfill
pipe
incineration
pipe recycle
seed
PE
membrane
membrane
installation
Field
Preparation
Harvest

Agri.
land
occ.
%

Climate
change

Fossil
deplete

Freshwate
r ecotoxi.

Freshwate
r eutro.

Human
toxicity

Terrestria
l acid.

%

%

%

%

%

%
6.00

Water
deplet
e
%

0.02

13.43

11.97

4.76

4.69

13.49

0.00
0.00
0.00
0.00

0.00
24.02
16.69
0.09

0.00
22.48
32.29
0.08

0.00
3.10
0.01
0.20

0.00
0.00
0.01
0.55

0.00 (cont’d)
0.00
20.01
38.21
0.16
11.52
9.09
0.06

0.00
0.00
0.39
0.04

0.00
0.00
4.92

13.90
0.09
4.93

0.10
0.12
1.85

0.00
0.00
1.73

0.00
0.00
3.18

0.03
0.03
10.12

0.20
0.09
4.61

0.20
0.08
11.82

0.00

5.49

11.11

0.01

0.01

0.06

4.76

0.12

0.03

4.84

4.35

0.02

0.06

0.67

6.63

0.36

94.84
0.00

0.00
0.00

0.00
0.00

86.69
0.00

86.09
0.00

0.00
0.00

10.27
0.00

0.00
0.00

185

65.22

Indicator

Machine

Chem.

Irrigate

Seed

harvest
fertilizing
sowing
tillage
gypsum
roundup
AMS
KCl
P2O5
urea
AN
K2O
irrigation
water
drip
irrigation
electricity
drip tape
pipe landfill
pipe
incineration
pipe recycle
seed

Table A5- 11 Flow relative contributions (%) of 2012 15” Ctrl
Agri.
Climate Fossil
Freshwater Freshwater Human
land occ. change
deplete
ecotoxi.
eutro.
toxicity
%
%
%
%
%
%

Terrestrial Water
acid.
deplete
%
%

0.05
0.01
0.03
0.06
0.00
0.01
0.00
0.02
0.00
0.00
0.00
0.00

4.33
0.71
0.64
3.99
0.16
0.99
0.17
1.56
1.67
5.75
1.38
0.09

4.26
0.71
0.62
3.92
0.13
1.05
0.26
1.65
2.06
6.64
0.85
0.12

1.36
0.21
0.19
1.22
0.00
0.29
0.00
0.30
0.03
0.02
0.00
0.00

1.43
0.20
0.29
0.89
0.00
1.22
0.00
1.32
0.03
0.01
0.00
0.00

18.71
2.86
2.53
15.27
0.02
4.30
0.01
5.09
0.46
0.35
0.03
0.01

7.76
1.11
0.90
5.90
0.18
2.21
0.09
1.08
2.04
2.75
0.33
0.05

7.26
1.04
1.54
5.35
0.17
5.56
0.02
3.69
0.50
1.45
0.24
0.01

0.01

9.45

9.20

2.65

2.54

7.74

4.46

43.49

0.00
0.00
0.00
0.00

0.00
16.83
21.50
0.12

0.00
17.34
45.54
0.11

0.00
1.73
0.01
0.20

0.00
0.00
0.01
0.55

0.00
11.50
0.17
9.57

0.00
28.45
15.69
0.08

0.00
0.00
0.48
0.05

0.00
0.00
9.36

17.91
0.11
12.65

0.14
0.17
5.22

0.00
0.00
3.54

0.00
0.00
6.29

0.03
0.04
21.32

0.28
0.12
12.53

0.24
0.09
28.82

186

Table A5-11 (cont'd)
Indicator

SWRT

Plant

PE
membrane
membrane
installation
Field
Preparation
Harvest

Agri.
land
occ.
%

Water
deplet
e
%

Climate
change

Fossil
deplete

Freshwate
r ecotoxi.

Freshwate
r eutro.

Human
toxicity

Terrestria
l acid.

%

%

%

%

%

%

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

90.44
0.00

0.00
0.00

0.00
0.00

88.26
0.00

85.22
0.00

0.00
0.00

14.00
0.00

0.00
0.00

187

Indicator
harvest
fertilizing
Machine
sowing
tillage
gypsum
roundup
AMS
KCl
Chem.
P2O5
urea
AN
K2O
irrigation
water
drip
irrigation
electricity
Irrigate
drip tape
pipe landfill
pipe
incineration
pipe recycle

Table A5- 12 Flow relative contributions (%) of 2012 30” Ctrl
Agri.
Climate Fossil
Freshwater Freshwater Human
land occ. change
deplete ecotoxi.
eutro.
toxicity
%
%
%
%
%
%
0.05
5.44
5.11
1.41
1.49
23.71
0.01
0.89
0.85
0.22
0.21
3.62
0.03
0.80
0.75
0.20
0.30
3.20
0.07
5.02
4.71
1.27
0.93
19.34
0.00
0.20
0.16
0.00
0.00
0.03
0.01
1.24
1.26
0.31
1.28
5.44
0.00
0.21
0.32
0.00
0.00
0.01
0.02
1.96
1.99
0.31
1.38
6.42
0.00
2.10
2.47
0.03
0.03
0.58
0.00
7.27
7.98
0.02
0.01
0.44
0.00
1.73
1.02
0.00
0.00
0.04
0.00
0.12
0.15
0.00
0.00
0.01

Terrestrial
acid.
%
10.21
1.46
1.18
7.77
0.24
2.91
0.12
1.42
2.69
3.64
0.43
0.06

Water
deplete
%
11.69
1.68
2.47
8.64
0.27
8.89
0.02
5.92
0.80
2.34
0.38
0.02

0.01

5.47

5.11

1.27

1.22

4.53

2.71

32.27

0.00
0.00
0.00
0.00

0.00
9.79
26.98
0.15

0.00
9.60
54.88
0.13

0.00
0.83
0.01
0.21

0.00
0.00
0.01
0.57

0.00
6.70
0.22
12.13

0.00
17.19
20.69
0.10

0.00
0.00
0.76
0.08

0.00
0.00

22.51
0.14

0.16
0.20

0.00
0.01

0.00
0.00

0.04
0.05

0.36
0.16

0.39
0.15

188

Table A5-12 (cont'd)
Indicator

Seed
SWRT

Plant

seed
PE
membrane
membrane
installation
Field
Preparation
Harvest

Agri.
land
occ.
%
4.91

Fossil
deplete

Freshwate
r ecotoxi.

Freshwate
r eutro.

Human
toxicity

Terrestria
l acid.

%
7.97

%
3.13

%
1.85

%
3.29

%
13.48

%
8.27

Water
deplet
e
%
23.21

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

94.88
0.00

0.00
0.00

0.00
0.00

92.07
0.00

89.26
0.00

0.00
0.00

18.40
0.00

0.00
0.00

Climate
change

189

Indicator

Machine

Chem.

Irrigate

harvest
fertilizing
sowing
tillage
gypsum
roundup
AMS
KCl
P2O5
urea
AN
K2O
irrigation
water
drip
irrigation
electricity
drip tape
pipe landfill
pipe
incineration
pipe recycle

Table A5- 13 Flow relative contributions (%) of 2013 Irrigated SWRT
Agri. land
Climate Fossil
Freshwater Freshwater Human
occ.
change
deplete ecotoxi.
eutro.
toxicity
%
%
%
%
%
%
0.05
3.55
3.32
2.15
2.24
17.76
0.01
0.58
0.55
0.33
0.31
2.71
0.03
0.52
0.49
0.30
0.45
2.40
0.06
3.11
2.91
1.83
1.33
13.76
0.00
0.13
0.10
0.00
0.00
0.02
0.01
0.82
0.82
0.47
1.92
4.08
0.00
0.14
0.21
0.00
0.00
0.01
0.02
1.16
1.16
0.42
1.86
4.33
0.00
0.79
0.91
0.03
0.03
0.25
0.00
3.59
3.87
0.02
0.02
0.25
0.00
2.11
1.22
0.01
0.01
0.06
0.00
0.27
0.34
0.00
0.00
0.02

Terrestrial
acid.
%
6.68
0.95
0.77
4.82
0.16
1.90
0.08
0.84
1.00
1.78
0.52
0.14

Water
deplete
%
6.41
0.92
1.36
4.51
0.15
4.89
0.01
2.93
0.25
0.96
0.39
0.04

0.02

10.31

9.41

5.53

5.22

9.68

5.03

50.43

0.00
0.00
0.00
0.00

0.00
18.37
17.81
0.08

0.00
17.71
35.57
0.09

0.00
3.60
0.02
0.32

0.00
0.00
0.02
0.86

0.00
14.34
0.17
9.10

0.00
32.05
13.56
0.07

0.00
0.00
0.42
0.04

0.00
0.00

14.86
0.23

0.11
0.13

0.00
0.01

0.00
0.01

0.03
0.03

0.24
0.11

0.22
0.08

190

Indicator
Seed
SWRT

Plant

seed
PE
membrane
membrane
installation
Field
Preparation
Harvest

Agri.
Climate
land occ. change
%
%
9.34
10.52

Table A5-13 (cont'd)
Fossil
Freshwater Freshwater Human
deplete ecotoxi.
eutro.
toxicity
%
%
%
%
4.07
5.62
9.87
20.26

Terrestrial
acid.
%
10.82

Water
deplete
%
25.46

0.00

5.87

12.22

0.01

0.01

0.06

5.59

0.13

0.02

5.17

4.79

0.03

0.09

0.67

7.79

0.38

90.44
0.00

0.00
0.00

0.00
0.00

79.30
0.00

75.77
0.00

0.00
0.00

5.09
0.00

0.00
0.00

191

Indicator

Machine

Chem.

Irrigate

harvest
fertilizing
sowing
tillage
gypsum
roundup
AMS
KCl
P2O5
urea
AN
K2O
irrigation
water
drip
irrigation
electricity
drip tape
pipe landfill
pipe
incineration
pipe recycle

Table A5- 14
Agri.
land occ.
%
0.05
0.01
0.03
0.06
0.00
0.01
0.00
0.02
0.00
0.00
0.00
0.00

Flow relative contributions (%) of 2013 Nonirrigated SWRT
Climate Fossil
Freshwater Freshwater Human
change
deplete ecotoxi.
eutro.
toxicity
%
%
%
%
%
14.13
17.89
2.36
2.39
26.80
2.32
2.97
0.36
0.33
4.07
2.09
2.62
0.33
0.48
3.60
13.05
16.45
2.13
1.49
21.77
0.53
0.55
0.00
0.00
0.03
3.28
4.41
0.52
2.04
6.14
0.56
1.11
0.00
0.00
0.01
4.65
6.26
0.47
1.99
6.51
3.14
4.91
0.03
0.03
0.37
14.29
20.90
0.02
0.02
0.38
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00

Terrestrial
acid.
%
19.05
2.71
2.20
14.48
0.44
5.42
0.23
2.37
2.83
5.05
0.00
0.00

Water
deplete
%
13.33
1.92
2.82
9.86
0.31
10.16
0.03
6.08
0.52
2.00
0.00
0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00
0.00
0.00
0.00

0.00
0.00
0.00
0.00

0.00
0.00
0.00
0.00

0.00
0.00
0.00
0.00

0.00
0.00
0.00
0.00

0.00
0.00
0.00
0.00

0.00
0.00
0.00
0.00

0.00
0.00
0.00
0.00

0.00
0.00

0.00
0.00

0.00
0.00

0.00
0.00

0.00
0.00

0.00
0.00

0.00
0.00

0.00
0.00

192

Table A5-14 (cont'd)
Indicator

Seed
SWRT

Plant

seed
PE
membrane
membrane
installation
Field
Preparation
Harvest

Agri.
land
occ.
%
9.37

Climate
change

Fossil
deplete

Freshwate
r ecotoxi.

Freshwate
r eutro.

Human
toxicity

Terrestria
l acid.

Water
deplete

%
41.96

%
21.93

%
6.19

%

%

%

%

0.00

0.00

0.00

0.00

0.00

90.45
0.00

0.00
0.00

10.57

30.33

30.73

52.98

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00
0.00

87.58
0.00

80.66
0.00

0.00
0.00

14.48
0.00

0.00
0.00

193

Indicator

Machine

Chem.

Irrigate

Seed

harvest
fertilizing
sowing
tillage
gypsum
roundup
AMS
KCl
P2O5
urea
irrigation
water
drip
irrigation
electricity
drip tape
pipe landfill
pipe
incineration
pipe recycle
seed

Table A5- 8 Flow relative contributions (%) of 2004 simulated Ctrl
Agri.
Climate Fossil
Freshwater Freshwater Human
land occ. change
deplete ecotoxi.
eutro.
toxicity
%
%
%
%
%
%

Terrestrial Water
acid.
deplete
%
%

0.05
0.01
0.03
0.06
0.00
0.01
0.00
0.00
0.00
0.00

11.81
1.93
1.74
10.87
0.44
2.73
0.47
0.64
2.96
19.83

13.74
2.29
2.02
12.70
0.42
3.39
0.86
0.99
4.28
26.83

2.13
0.33
0.30
1.92
0.00
0.46
0.00
0.00
0.03
0.03

2.19
0.30
0.44
1.37
0.00
1.88
0.00
0.00
0.03
0.03

27.93
4.26
3.78
22.78
0.04
6.41
0.01
0.02
0.44
0.65

16.34
2.33
1.89
12.45
0.38
4.66
0.19
0.24
2.77
7.25

13.80
1.98
2.92
10.22
0.32
10.52
0.03
0.07
0.61
3.44

0.00

0.34

0.39

0.05

0.05

0.15

0.12

1.09

0.00
0.00
0.00
0.00

0.00
0.61
5.74
0.03

0.00
0.73
14.33
0.03

0.00
0.04
0.00
0.03

0.00
0.00
0.00
0.08

0.00
0.23
0.03
1.39

0.00
0.79
3.21
0.02

0.00
0.00
0.09
0.01

0.00
0.00
9.37

4.79
0.07
34.99

0.04
0.05
16.89

0.00
0.00
5.57

0.00
0.00
9.66

0.00
0.01
31.87

0.06
0.03
26.50

0.05
0.02
54.85

194

Table A5-15 (cont'd)
Indicator

SWRT

Plant

PE
membrane
membrane
installation
Field
Preparation
Harvest

Agri.
land
occ.
%

Climate
change

Fossil
deplete

Freshwater Freshwater Human
ecotoxi.
eutro.
toxicity

Terrestrial Water
acid.
deplete

%

%

%

%

%

%

%

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

90.46
0.00

0.00
0.00

0.00
0.00

89.11
0.00

83.95
0.00

0.00
0.00

20.79
0.00

0.00
0.00

195

Indicator

Machine

Chem.

Irrigate

harvest
fertilizing
sowing
tillage
gypsum
roundup
AMS
KCl
P2O5
urea
irrigation
water
drip
irrigation
electricity
drip tape
pipe landfill
pipe
incineration
pipe recycle

Table A5- 9 Flow relative contributions (%) of 2004 simulated SWRT
Agri.
Climate Fossil
Freshwater Freshwater Human
land occ. change
deplete ecotoxi.
eutro.
toxicity
%
%
%
%
%
%
0.05
2.59
1.72
2.14
2.18
27.89
0.01
0.43
0.29
0.33
0.30
4.26
0.03
0.38
0.25
0.30
0.44
3.77
0.03
1.19
0.79
0.96
0.68
11.34
0.00
0.10
0.05
0.00
0.00
0.04
0.01
0.59
0.42
0.47
1.87
6.40
0.00
0.10
0.11
0.00
0.00
0.01
0.00
0.14
0.12
0.00
0.00
0.02
0.00
0.64
0.54
0.03
0.03
0.44
0.00
4.30
3.35
0.03
0.03
0.65

Terrestrial
acid.
%
3.88
0.55
0.45
1.47
0.09
1.11
0.05
0.06
0.66
1.72

Water
deplete
%
13.04
1.87
2.76
4.82
0.30
9.93
0.03
0.06
0.57
3.25

0.00

0.07

0.05

0.06

0.05

0.15

0.03

1.03

0.00
0.00
0.00
0.00

0.00
0.13
1.25
0.01

0.00
0.09
1.79
0.00

0.00
0.04
0.00
0.03

0.00
0.00
0.00
0.08

0.00
0.22
0.03
1.39

0.00
0.19
0.76
0.00

0.00
0.00
0.08
0.01

0.00
0.00

1.04
0.01

0.01
0.01

0.00
0.00

0.00
0.00

0.00
0.01

0.01
0.01

0.04
0.02

196

Table A5-16 (cont'd)
Indicator

Seed
SWRT

Plant

seed
PE
membrane
membrane
installation
Field
Preparation
Harvest

Agri.
land
occ.
%
9.36

Climate
change

Fossil
deplete

Freshwater Freshwater Human
ecotoxi.
eutro.
toxicity

Terrestrial Water
acid.
deplete

%
7.58

%
2.11

%
5.59

%
9.63

%
6.29

%
31.82

%
51.81

0.00

42.18

63.42

0.15

0.13

1.01

32.52

2.55

0.24

37.27

24.88

0.27

0.86

10.54

45.22

7.82

90.27
0.00

0.00
0.00

0.00
0.00

89.61
0.00

83.69
0.00

0.00
0.00

4.94
0.00

0.00
0.00

197

Appendix B5 Allocation scenarios
Table B5- 1-a Allocation scenarios on 15” SWRT and 30” SWRT
2012 15'' SWRT
Impact
category
Agricultural
land occ.
Climate
change
Fossil
depletion
Freshwater
ecotoxi.
Freshwater
eutro.
Human
toxicity
Terrestrial
acid.
Water
depletion

2012 30'' SWRT

All on
grain

Economic Energy

Mass

System
All on
Expansion grain

Economic Energy Mass

System
Expansion

658

520

415

329

645

789

624

497

395

776

375

296

236

187

-655

344

272

217

172

-686

132

104

83.2

66

109

126

99.9

79.6

63.2

103

10.2

8.1

6.5

5.1

10.1

11.9

9.4

7.5

5.9

11.8

0.093

0.073

0.058

0.046 0.091

0.109

0.086

0.069

0.054 0.108

47.1

37.2

29.7

23.6

42.7

33.7

26.9

21.3

1.990

1.570

1.250

0.993 1.650

1.75

1.39

1.1

0.877 1.41

161

127

101

80.4

142

112

89.6

71.1

44.3

161

198

39.9

142

Table B5- 1-b Allocation scenarios on 15” Ctrl and 30” Ctrl
2012 15'' Ctrl
Impact
category
Agricultural
land occ.
Climate
change
Fossil
depletion
Freshwater
ecotoxi.
Freshwater
eutro.
Human
toxicity
Terrestrial
acid.
Water
depletion

All on
grain

Economic

Energy

1140

901

368

2012 30'' Ctrl
Mass

System
All on
Expansion grain

Economic Energy Mass

System
Expansion

719

571

1127

1060

839

669

531

1047

291

232

184

-662

286

226

180

143

-744

123

97.5

77.8

61.7

100

100

79.2

63.1

50.1

77

16.1

12.7

10.1

8.0

16.0

15.0

11.9

9.5

7.5

14.9

0.151

0.120

0.095

0.076 0.150

0.141

0.111

0.089

0.071 0.140

55.8

44.1

35.2

27.9

43.0

34.0

27.1

21.5

1.770

1.400

1.120

0.886 1.430

1.32

1.04

0.829

0.658 0.98

237

187

149

118

167

132

105

83.7

53.0

237

199

40.2

167

Table B5- 1-c Allocation scenarios on Irrigated SWRT and Nonirrigated Ctrl
2013 Irrigated SWRT
Impact
category
Agricultural
land occ.
Climate
change
Fossil
depletion
Freshwater
ecotoxi.
Freshwater
eutro.
Human
toxicity
Terrestrial
acid.
Water
depletion

All on
grain

Economic

Energy

734

580

285

2013 Nonirrigated Ctrl

Mass

System
All on
Expansion grain

Economic Energy Mass

System
Expansion

462

367

721

2480

1960

1560

1240

2467

225

180

143

-745

242

191

152

121

-788

102

80.4

64.1

50.9

79

63.8

50.4

40.2

31.9

40.8

6.5

5.1

4.1

3.3

6.5

19.9

15.7

12.5

10.0

19.8

0.062

0.049

0.039

0.031 0.061

0.196

0.155

0.124

0.098 0.195

37.8

29.8

23.8

18.9

85.0

67.1

53.5

42.5

1.320

1.050

0.834

0.662 0.980

1.57

1.24

0.99

0.786 1.23

283

223

178

141

460

363

290

230

35.0

283

200

82.2

460

Appendix C5 Pedigree matrix
The pedigree matrix was used to assess the standard deviation of the data
quality. The value of the standard deviation was calculated based on the following
equation. The corresponding scores are assessed according in Table C5-1.
SD = exp √[ln(𝑈1)2 ] + [ln(𝑈2)2 ] + [ln(𝑈3)2 ] + [ln(𝑈4)2 ] + [ln(𝑈5)2 ] + [ln(𝑈6)2 ]
where,
U1= the corresponding score of reliability;
U2= the corresponding score of completeness;
U3= the corresponding score of temporal correlation;
U4= the corresponding score of geographical correlation;
U5= the corresponding score of technological correlation;
U6= the corresponding score of sample size;

201

Indicator
Reliability
U1

Completeness

U2
Temporal
correlation
U3
Geographical
correlation

Table C5- 1 Pedigree matrix used to assess the data quality
1
2
3
4
Verified data partly
Non-verified
based on
Qualified Estimate
Verified data based
data partly
assumptions or non(e.g. by industrial
on measurement
based on
verified data based
expert)
assumptions
on measure
1.00
1.05
1.10
1.20
Representative data
Representative data
Representative
but from a smaller
from a sufficient
Representative data data from
number of sites and
sample of sites over from a smaller
adequate
shorter periods or
an adequate period number of sites but
number of sites
incomplete data from
to even out normal
for adequate periods but from shorter
an adequate number
fluctuations
periods
of sites and periods
1.00
< 3 years difference
to year of study
1.00
Data from area
under study

1.02

1.05
< 10 years
difference to
year of study

< 6 years difference
to year of study
1.03

1.10

Average data from
larger area in which
the area under study
is included

Data from area
with similar
production
conditions

202

1.10
< 15 years difference
to year of study
1.20
Data from area with
slightly similar
production
conditions

5
Non-qualified
estimate
1.50
Representativene
ss unknown or
incomplete data
from a smaller
number of sites
and/or from
shorter periods
1.20
Age of data
unknown or > 15
years difference
to year of study
1.50
Data from
unknown area or
area with very
different
production
conditions

Indicator
U4
Technology
correlation
U5
Sample size

Indicator
1.00
Data from
enterprises,
processes, and
materials under
study
1.00
> 100, continuous
measurement,
balance of
purchased products

Table C5- 1 (cont’d)
Indicator
Indicator
1.01
1.02
Data from
Data for processes
processes and
and materials under materials under
study but from
study but from
different enterprises different
technology
NA
1.20
> 20

> 10

203

Indicator
NA
Data from related
processes or
materials but same
technology

Indicator
1.10
Data on related
processes or
materials but
different
technology

1.50

2.00

>=3

unknown

Group

Process

Table C5- 2 Pedigree matrix used for uncertainty analysis
In/Out Flow
U1
U2
U3

harvesting
Inputs

Fertilizing
Inputs

Machinery

sowing
Inputs

tillage
Inputs

Diesel
harvester production
shed
tillage machine
production
Diesel
shed
tractor production
machinery
production
Diesel
shed
tractor production
machinery
production
Diesel
shed
tractor production

204

U4

U5

SD

1.00
1.00
1.00

1.02
1.10
1.10

1.00
1.00
1.00

1.01
1.02
1.02

1.00
1.00
1.00

1.27
1.57
1.57

1.00
1.00
1.00
1.00

1.10
1.02
1.10
1.10

1.00
1.00
1.00
1.00

1.02
1.01
1.02
1.02

1.00
1.00
1.00
1.00

1.57
1.27
1.57
1.57

1.00
1.00
1.00
1.00

1.10
1.02
1.10
1.10

1.00
1.00
1.00
1.00

1.02
1.01
1.02
1.02

1.00
1.00
1.00
1.00

1.57
1.27
1.57
1.57

1.00
1.00
1.00
1.00

1.10
1.02
1.10
1.10

1.00
1.00
1.00
1.00

1.02
1.01
1.02
1.02

1.00
1.00
1.00
1.00

1.57
1.27
1.57
1.57

Group

Chemical

Irrigation

Seed

Process
Gypsum stone
Glyphosate
AMS
KCl
P2O5
Urea
AN
K2O
Water
drip
irrigation

In/Out

Table C5- 2 (cont’d)
Flow
U1

Electricity, at grid
Water
Outputs irrigating
Inputs

Electricity, at
grid
pipe (PE-HD)
Pipe landfill
pipe
incineration
pipe recycling
corn seed

205

U2

U3

U4

U5

SD

1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00

1.00
1.00
1.00
1.05
1.00
1.00
1.00
1.00
1.00

1.00
1.00
1.00
1.50
1.00
1.00
1.00
1.00
1.00

1.02
1.02
1.01
1.01
1.01
1.01
1.01
1.02
1.02

1.20
1.00
1.20
1.00
1.00
1.00
1.00
1.00
1.00

1.76
1.15
1.69
2.60
1.10
1.10
1.10
1.15
1.15

1.05
1.00
1.00

1.05
1.00
1.00

1.20
1.00
1.00

1.02
1.00
1.00

1.00
1.00
1.00

2.74
1.00
1.00

1.05
1.00
1.00

1.05
1.00
1.00

1.20
1.20
1.00

1.00
1.02
1.01

1.00
1.20
1.00

2.38
2.70
1.10

1.00
1.00
1.05

1.00
1.00
1.10

1.00
1.00
1.20

1.02
1.01
1.10

1.00
1.00
1.00

1.15
1.10
3.54

Group

Process

In/Out

Table C5- 2 (cont’d)
Flow
U1

SWRT film (PE-LD)
SWRT

Planting
(simulated)

Installation

Inputs

Field
preparation

Inputs

SWRT machine
production
Diesel
shed
tractor production
AMS
fertilizing
corn seed
sowing
tillage
SWRT installation
Glyphosate
Gypsum
Pipe
SWRT film
K2O
P2O5
Urea
irrigating
CO2
crop land occ.

206

U2

U3

U4

U5

SD

1.00

1.00

1.10

1.02

1.00

1.57

1.00
1.00
1.00
1.00
1.05
1.00
1.05
1.00
1.00
1.00
1.05
1.10
1.05
1.05
1.20
1.20
1.20
1.10
1.05
1.00

1.10
1.02
1.10
1.10
1.10
1.00
1.02
1.00
1.00
1.00
1.02
1.10
1.02
1.10
1.00
1.00
1.00
1.00
1.00
1.00

1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00

1.02
1.01
1.02
1.02
1.01
1.02
1.10
1.02
1.02
1.02
1.10
1.10
1.10
1.10
1.10
1.01
1.01
1.01
1.01
1.00

1.50
1.00
1.00
1.50
1.20
1.00
1.00
1.00
1.00
1.20
1.00
1.00
1.20
1.00
1.00
1.00
1.00
1.00
1.00
1.00

2.96
1.27
1.57
2.96
2.88
1.15
1.95
1.15
1.15
1.76
1.95
2.52
3.00
2.31
2.09
1.69
1.69
1.50
1.38
1.00

Group

Process
Field
preparation

Planting
(simulated)
Harvest

Planting
(experiment)

Field
Preparation
2012-2013

In/Out

Table C5- 2 (cont’d)
Flow
U1

corn
nitrate
Outputs
NOx
Phosphorus
Inputs
harvesting
corn
Outputs Corn grains
pipe incineration
pipe landfill
pipe recycling good
AMS
fertilizing
corn seed
sowing
tillage
SWRT installation
Inputs
Glyphosate
Gypsum
Pipe
SWRT film
K2O
P2O5

207

1.00
1.10
1.10
1.10
1.00
1.00
1.00
1.00
1.00
1.10
1.05
1.05
1.05
1.05
1.05
1.05
1.05
1.10
1.05
1.05
1.05
1.05

U2

U3

U4

U5

SD

1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.10
1.10
1.20
1.10
1.10
1.10
1.10
1.10
1.10
1.10
1.20
1.10
1.10
1.10
1.10

1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.20
1.20
1.20
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00

1.00
1.00
1.00
1.00
1.02
1.00
1.00
1.02
1.02
1.10
1.01
1.02
1.10
1.02
1.02
1.02
1.10
1.10
1.10
1.10
1.10
1.01

1.00
1.00
1.00
1.20
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.20
1.00
1.00
1.20
1.00
1.00
1.00

1.00
1.36
1.36
2.09
1.15
1.00
1.00
2.40
2.40
4.36
1.88
1.95
2.31
1.95
1.95
3.00
2.31
2.84
3.54
2.31
2.31
1.88

Group

Process

Field
Preparation
2012-2013
Planting
(experiment)

Harvest
2012 & 2013

In/Out

Table C5- 2 (cont’d)
Flow
U1

Urea
KCl
AN
Inputs
irrigating
CO2
crop land occ.
corn
nitrate
Outputs
NOx
Phosphorus
harvesting
Inputs
corn
Corn grains
pipe incineration
Outputs
pipe landfill
pipe recycling good

208

1.05
1.05
1.05
1.05
1.05
1.05
1.00
1.10
1.10
1.10
1.05
1.00
1.00
1.10
1.10
1.10

U2

U3

U4

U5

SD

1.10
1.10
1.10
1.10
1.00
1.10
1.00
1.10
1.10
1.10
1.10
1.00
1.00
1.20
1.20
1.20

1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.20
1.20
1.50

1.01
1.01
1.01
1.01
1.01
1.00
1.00
1.00
1.00
1.00
1.02
1.00
1.00
1.02
1.02
1.10

1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.20
1.00
1.00
1.00
1.50
1.50
1.50

1.88
1.88
1.88
1.88
1.38
1.70
1.00
1.85
1.85
2.84
1.95
1.00
1.00
6.96
6.96
10.15

Table C5- 3 Monte Carlo simulation on SWRT machine diesel consumption rate
2012 15'' SWRT
Sample
MC.
mean
mean SD
Agri. land occ.
329
329
1.16E-03
Climate change
187
187
1.86E-02
Fossil depletion
66.0
66
3.70E-02
Freshwater ecotoxi. 5.12
5.12
2.73E-06
Freshwater eutro.
4.63E-02 0.0463 8.80E-08
Human toxicity
23.6
23.6
4.98E-04
Terrestrial acid.
0.994
0.993 1.02E-04
Water depletion
230
230
3.17E-03
Where MC. mean is the mean from Monte Carlo simulation.

2012 30'' SWRT
Sample
MC.
mean
mean SD
394
395
1.47E-03
172
172
2.36E-02
63.2
63.2
4.70E-02
5.94
5.94
3.46E-06
5.44E-02 0.0543 1.11E-07
21.3
21.3
6.30E-04
0.877
0.877 1.31E-04
172
172
4.02E-03

209

2013 Irrigated SWRT
Sample
MC.
mean
mean SD
367
367
1.32E-03
143
143
2.13E-02
50.9
50.9
4.21E-02
3.25
3.25
3.10E-06
3.10E-02 0.031 1.00E-07
18.9
18.9
5.67E-04
0.661
0.662 1.17E-04
141
141
3.61E-03

Table C5- 4 Monte Carlo simulation on drip tape production
2012 15'' SWRT
mean SD
Agri. land occ.
329
4.64E-05
Climate change 187
1.18E+00
Fossil depletion 66
4.57E-01
Freshwater
ecotoxi.
5.12
2.78E-04
Freshwater
eutro.
0.0463 6.81E-06
Human toxicity 23.6
4.41E-02
Terrestrial
acid.
0.993
2.31E-03
Water
depletion
230
2.69E-02

2012 30'' SWRT
mean SD
395
5.73E-05
172
1.46E+00
63.2
5.65E-01

2012 15" Control
mean SD
571
7.99E-05
184
2.02E+00
61.7
7.90E-01

2012 30" Control
mean SD
531
7.54E-05
143
1.92E+00
50.1
7.41E-01

Irrigated SWRT
mean SD
367
4.84E-12
143
2.06E-12
50.9
9.98E-14

5.94

3.44E-04 8.04

4.79E-04 7.52

4.52E-04 3.25

4.26E-14

0.0543
21.3

8.42E-06 0.0757
5.45E-02 27.9

1.17E-05 0.0706
7.59E-02 21.5

1.11E-05 0.031
7.16E-02 18.9

3.06E-16
1.60E-13

0.877

2.86E-03 0.886

3.98E-03 0.658

3.76E-03 0.662

9.53E-15

172

3.34E-02 194

4.64E-02 118

4.39E-02 141

2.83E-13

210

Table C5- 5 Monte Carlo simulation on seed

A
B
C
D
E
F
G
H

15'' SWRT
mean SD
329
1.082
187
0.473
66
0.065
5.12
0.006
9.7E0.0463
05
23.5
0.121
0.993 0.002
230
1.139

30'' SWRT
mean SD
395
0.703
172
0.308
63.2
0.042
5.94
0.004
6.2E0.0543
05
21.3
0.078
0.877 0.001
172
0.738

15" Control
mean
SD
570
1.938
184
0.848
61.7
0.117
8.04
0.010
1.7E0.0757
04
27.9
0.217
0.886
0.004
194
2.037

30" Control
mean
SD
531
0.881
143
0.386
50.1
0.054
7.52
0.005
7.9E0.0706
05
21.5
0.098
0.658
0.002
118
0.929

Irrigated SWRT
mean
SD
367
1.215
143
0.533
50.9
0.073
3.25
0.006
1.1E0.031
04
18.9
0.136
0.662
0.003
141
1.273

Nonirrigated Ctrl
mean
SD
1240
3.41E-12
121
1.63E-12
31.9
1.67E-13
9.96
1.12E-13
0.0982
42.5
0.786
230

1.01E-15
6.12E-13
4.44E-15
4.83E-12

Where A-H represent impact categories in ReCiPe 1.07 (H) Midpoint,
A- Agricultural land occupation; B- Climate change; C-Fossil depletion; D- Freshwater ecotoxicity; E- Freshwater
eutrophication; F- Human toxicity; G- Terrestrial acidification; H-Water depletion.

211

REFERENCES

212

REFERENCES
1.

IES, I., Handbook: General Guide for Life Cycle Assessment–Detailed Guidance.
JRC, IES, 2010.

2.

Hanson, B.R., J. Šimůnek, and J.W. Hopmans, Evaluation of Urea–ammonium–
nitrate Fertigation with Drip Irrigation using Numerical Modeling.
Agricultural water management, 2006. 86(1): p. 102-113.

3.

Kim, S., B.E. Dale, and R. Jenkins, Life Cycle Assessment of Corn Grain and Corn
Stover in the United States. The International Journal of Life Cycle Assessment,
2009. 14(2): p. 160-174.

4.

PE International, L.G., Corn grain (field border), L. G. PE International, Editor.
2012, PE-GaBi.

5.

Ecoinvent, US: corn, at farm, E. 2.2, Editor. 2011, GaBi 6.

6.

PE-GaBi, Corn, whole plant, at field, U.S.L. Database, Editor. 2009, PE-Gabi.

7.

Albertsson, A.C., Degradable polymers. Journal of Macromolecular Science,
Part A: Pure and Applied Chemistry, 1993. 30(9-10): p. 757-765.

8.

Standardization, I.O.f., International standards Stadard 14040 only, in
Environmental management- Life cycle assessment-Principles and framework.
2006: Switzerland.

9.

Standarization, I.O.f., International standars Standard 14044 only, in
Environmental management- Life Cycle Assessment-Requirements and
guidelines. 2006: Switzerland.

10.

Weidema, B., Avoiding Co‐Product Allocation in Life‐Cycle Assessment.
Journal of Industrial Ecology, 2000. 4(3): p. 11-33.

11.

Luo, L., et al., Allocation issues in LCA methodology: a case study of corn stoverbased fuel ethanol. The International Journal of Life Cycle Assessment, 2009.
14(6): p. 529-539.
213

12.

Van der Voet, E., R.J. Lifset, and L. Luo, Life-cycle Assessment of Biofuels,
convergence and divergence. Biofuels, 2010. 1(3): p. 435-449.

13.

Sheehan, J., et al., Energy and Environmental Aspects of Using Corn Stover for
Fuel Ethanol. Journal of Industrial Ecology, 2003. 7(3-4): p. 117-146.

14.

Kim, S. and B.E. Dale, Life cycle Assessment of Various Cropping Systems
Utilized for Producing Biofuels: Bioethanol and Biodiesel. Biomass and
Bioenergy, 2005. 29(6): p. 426-439.

15.

Laser, M., et al., Coproduction of Ethanol and Power from Switchgrass. Biofuels,
Bioproducts and Biorefining, 2009. 3(2): p. 195-218.

16.

Bai, Y., L. Luo, and E. Voet, Life Cycle Assessment of Switchgrass-derived
Ethanol as Transport Fuel. The International Journal of Life Cycle Assessment,
2010. 15(5): p. 468-477.

17.

Wu, M., et al., Consumptive Water Use in the Production of Bioethanol and
Petroleum Gasoline. Center for Transportation Research. Energy Systems
Division, Argonne National Laboratory, 2008.

18.

Mia Lafontaine, François Charron-Doucet, and E. Clément, Production of
Pulses, Grains and Oilseeds - A guide for LCA Practicioners, in LCA guidelines
for pulses, grains and oilseeds, Quantis, Editor. 2013, Pulse Canada. p. 42.

19.

Weidema, B.P. and M.S. Wesnæs, Data Quality Management for Life Cycle
Inventories—an example of using data quality indicators. Journal of Cleaner
Production, 1996. 4(3): p. 167-174.

214

Normalization and weighting
LCA is a useful tool to integrate the environmental impacts for product development
and policy making. In order to conduct a comprehensive life cycle study, scientific methods
must be used to evaluate life cycle thinking. A comparison of environmental impacts
elaborates the relative environmental performance of alternatives products in the context
of analyzed impact topics. Parallel to this, there is a need for simple, easy accessible
methods to interpret the LCA results for non-LCA practitioners, such as policy makers and
product designers. This audience is interested in the relative importance among studied
impact topics to reduce the product impact effectively. Normalization and weighting are
methods to reveal the relative importance of the analyzed impacts.
Normalization is an optional step according to ISO 14044 [1]. It helps to interpret the
environmental impact profile of a product and/or system. Also, normalization is an initial
step of fully aggregated results that needs an additional weighting step across indicators
[2].
Normalization is mostly used to interpret how the results of a LCA impact an average
citizen, a country, and/or globally, etc. When displaying the normalized LCIA results of
different impact categories next to each other, the relatively contributions of each impact
indicator can be seen. Normalization can be processed either on midpoint (such as
acidification, fossil depletion, etc.) or endpoint (such as human health, natural resource,
and natural environment) impact levels.
To note that, after normalization, the relatively contributions cannot be compared
directly across categories. To judge the relative importance among impact results,
215

additional weighting step is needed. Weighting makes the LCA result more deliverable,
while it introduces more uncertainties. Weighting is an optional step and it is not
recommended for academic LCA under ISO 14044 [1].

6.1 Normalization
Normalization is a supportive method to interpret the environmental impact profile of a
product and/or system. It is also the first step towards a fully aggregated result.
Normalized LCIA impacts will indicate the relatively contribution of the impact category of
the studied system in the total impact category per average citizen, per country, per GDP,
etc. It is important to note that normalized LCIA values across different impact cannot be
compared to interpret the absolute relevance (e.g., judge one impact category is more
concerning than the other based on the result). Because the normalized LCIA results
describe different impacts with different units, their value cannot be summed up as well.
Calculation of the normalized impact potential is performed as follows:
Norm IPIC = LCIAIC * Normref

(Equation 6-1)

where:
Norm IPIC is the normalized impact potential for a studied category;
LCIAIC is the LCIA result of a specific impact category;
Normref IC is the normalization reference of the corresponding studied category. In
this study, Normref equals to the total impact of all substances of the specific
category per person per year.

216

Equation 6-1 implies that the choice of normalization reference can greater affect/shift
the normalized LCIA results. Commonly used normalization references are EDIP97 and
EI99, which are originally designed for Europe. With LCA studies and applications
becoming globally, there is a need to establish the normalization and weighting references
for different Word regions. Unfortunately, this will take a long time and lots of effort to
collect sufficient information and establish references target to various regions.
In this paper, ReCiPe 1.07 (H) midpoint- world normalization factors were used, which
has two sets of parameters - Europe and World. At this point, we have to admit that
applying the World normalization reference to the U.S. studied results will introduce
certain levels of uncertainty to the results. Environmental impacts can be approximately
classified into two groups: global impacts and regional/ local impacts. In terms of global
impacts, such as global warming and ozone depletion, the uncertainties to the normalized
LCIAs in this study are relatively lower than regional and local impacts such as acidification,
eutrophication, ecotoxicity, and human toxicity.
Table 6.1 lists the World’s normalization references of ReCiPe 1.07 (H) midpoint. It is an
update from the 1994 version, which reflects the world’s population of 6,122,770,220 in
2000 [3].
Table 6 -1 ReCiPe 1.07 (H) World midpoint normalization factor [4]
Impact categories
Normalization factor
Unit
Climate change
1.45E-04
kg CO2 eq /p/year
Terrestrial acidification
2.62E-02
kg SO2 eq/p/Year
Freshwater eutrophication 3.45E+00
kg P eq/p/Year
Human toxicity
8.52E-03
kg 1,4-DB eq/p/Year
Freshwater ecotoxicity
2.31E-01
kg 1,4-DB eq/p/Year
Agri. land occupation.
1.84E-04
kg P eq/p/Year
Water depletion
0.00E+00
m3/p/Year
Fossil depletion
7.75E-04
kg oil eq/p/Year
217

Table 6 -2 LCIA of SWRT
Indicator

Agri.
land occ.

Unit

m2*year

2012 15” SWRT
2012 30” SWRT
2012 15” Ctrl
2012 30” Ctrl
2013 Irrigated SWRT
2013 Nonirri. SWRT
simulated 2004 Ctrl
simulated 2004 SWRT

328.6
394.4
570.5
531.2
367.1
1238.3
456.6
457.5

Climate
change
kg CO2eq.
187.3
172
184.2
143.1
142.6
121.1
53.4
246.6

Fossil
Freshwater Freshwater
depletion ecotoxi.
eutro.
kg 1,4-DB
kg oil eq.
kg P eq.
eq.
66
5.12
4.63E-02
63.2
5.94
5.44E-02
61.7
8.03
7.57E-02
50.1
7.53
7.06E-02
50.9
3.25
3.10E-02
32
9.98
9.84E-02
15.3
4.09
3.94E-02
122.2
4.06
3.95E-02

Human
toxicity
kg 1,4DB eq.
23.6
21.3
27.9
21.5
18.9
42.5
15
15

Terrestrial Water
acid.
depleti.
kg SO2 eq.

m3

0.994
0.877
0.886
0.657
0.661
0.787
0.337
1.418

229.7
171.7
194.3
118.1
141.4
230.3
81.9
86.7

Table 6 -3 Normalized LCIA
Indicator
2012 15” SWRT
2012 30” SWRT
2012 15” Ctrl
2012 30” Ctrl
2013 Irrigated SWRT
2013 Nonirri. SWRT
simulated 2004 Ctrl
simulated 2004 SWRT

Agri.
land occ.
0.06
0.07
0.10
0.10
0.07
0.23
0.08
0.08

Climate
change
0.03
0.02
0.03
0.02
0.02
0.02
0.01
0.04

Fossil
depletion
0.05
0.05
0.05
0.04
0.04
0.02
0.01
0.09

Freshwater
ecotoxi.
1.18
1.37
1.85
1.74
0.75
2.31
0.94
0.94

218

Freshwater
eutro.
0.16
0.19
0.26
0.24
0.11
0.34
0.14
0.14

Human
toxicity
0.20
0.18
0.24
0.18
0.16
0.36
0.13
0.13

Terrestrial
acid.
0.03
0.02
0.02
0.02
0.02
0.02
0.01
0.04

Water depleti.
NA
NA
NA
NA
NA
NA
NA
NA

Table 6.3 indicates that freshwater ecotoxicity contributes relatively more than
other impact categories. It has been discussed in chapter 5.2 LCIA results that more
than 80% of the freshwater ecotoxicity is from fertilizer’s inorganic phosphate
emissions to fresh water.

6.2 Weighting
Weighting is the additional step required to interpret an aggregated result after
normalization. As in normalization, weighting is an optional step under ISO14044.
In the weighting step, a specific weighting factor is used to multiply with an impact
category. By doing that, weighted LCIAs will be able to sum up or compare across
different impact categories. Note that under ISO 14044 [1] weighting shall not be
used in studies leading to comparative assertions intended to be disclosed to the
public.
Calculation of the weighted impact potential is performed as follows:
𝑊𝑒𝑖𝑔ℎ𝑡𝑒𝑑 𝐼𝑃𝐼𝐶 = WF𝐼𝐶 ∗ 𝑁𝑜𝑟𝑚 IP 𝐼𝐶
Where:

(Equation 6-2)

Norm IPIC is the normalized impact potential for a studied category,
which is the result from the normalization process;
WF IC is the weighting factor for a specific impact category;
Weighted IC is the weighted impact of the corresponding studied
category.

The weighting reference factors for ReCiPe 1.07 (H) midpoint impact assessment
has yet not been published [5]. Therefore, weighting didn’t not being performed in
this study.
219

REFERENCES

220

REFERENCES
1.

Standarization, I.O.f., International standars Standard 14044 only, in
Environmental management- Life cycle assessment-Requirements and
guidelines. 2006: Switzerland.

2.

Commission, E., International Reference Life Cycle Data System (ILCD)
Handbook—general guide for life cycle assessment—detailed guidance. Joint
Research Centre—Institute for Environment and Sustainability. Publications
Office of the European Union, Luxembourg, 2010.

3.

Census, U.S. (2014). International programs data base; Available from:
http://www.census.gov/population/international/data/idb/informationGat
eway.php.

4.

Sleeswijk, A.W., et al., Normalisation in product life cycle assessment: An LCA of
the global and European economic systems in the year 2000. Science of The
Total Environment, 2008. 390(1): p. 227-240.

5.

ReCiPe. Quick introduction into ReCiPe LCIA Methodology. 2008 [cited 2014;
Available from: http://www.lcia-recipe.net/project-definition.

221

Conclusions and future work
7.1

Conclusions
In this study, the environmental impacts of producing 1000 kg of corn grain

were evaluated considering 6 different treatments (15” SWRT, 30” SWRT, 15” Ctrl,
30” Ctrl, Irrigated SWRT, and Nonirrigated Ctrl) planted on Sandhill farm, East
Lansing, MI, US. To compensate for the limitations of the experimental data, a 10year continuous corn production on Sandhill farm dataset was simulated using the
SALUS program. The environmental impact of the simulated treatment was
conducted to estimate the time to pay off the additional environmental burdens for
using SWRT.
The LCIs were mainly assessed by eight impact indicators from the ReCiPe 1.07
(H) midpoint assessment methodology. Mass allocation was the default method to
solve the co-product issues.
Benchmark comparisons:
Several studies concerning the environmental performance of corn stover and corn
grain production have been published, and some of these results were used to
compare with the results of this study. Results from this study show that most
impact categories except climate change and air acidification were in good
agreement with previous results. The published results have as great as 126%
difference among them on air acidification impact. The result of this study is lower
than benchmark about 35 to 622% depending on the study used for comparison.
The result of this study is about 18% lower than the benchmark study on climate
222

change. The relatively lower impacts of this study might be owing to less gas
emissions generated from fertilizers in sandy soil conditions.
Hotspots identification:
The contribution of each group was evaluated to identify the environmental impact
hotspots. Among the groups, irrigation contributed the greatest impact. The
machinery, chemical, and seed groups were also intensive groups on most impacts.
Fertilizer emissions in the planting group were significant on freshwater ecotoxicity
and freshwater eutrophication impacts. This implies that these two impact
categories are very vulnerable to over fertilizing.
Agricultural land occupation:
Agricultural land is regarded as a limited resource. High land use efficiency (high
yield) can effectively reduce the land occupation. The LCIA result of agricultural land
occupation illustrates that the planting group largely affect this impact category.
Climate change and fossil depletion:
From the LCIA results, irrigation activity (water and electricity consumption for
pumping water) takes up 1/3 to 1/2 of the total indicators. Avoiding over irrigation
is suggested to be both economic and environmentally friendly according to the
results.
Eutrophication and acidification:
According to the soil electrical conductivity experiments and Hydrus-2D simulation
work performed on the Sandhill farm experiments, there is a lack of evidence
223

indicating a significant difference in fertilizer emissions between SWRT and Control
treatments. Thus, the differences in the LCIA results of freshwater eutrophication
and acidification were mostly driven by the yields.
Human toxicity:
The LCIA results on the human toxicity impact category illustrates that machinery
was the greatest contributing group. Diesel consumption generated 70% of the
machinery burden.
Terrestrial acidification:
The LCIA result for terrestrial acidification suggests that electricity used for
irrigation, machinery, irrigation water, pipe production, SWRT and seed affected
this impact category. A further analysis explained that fuel burning was the
underlying cause in these groups.
Allocation method and impact assessment methodology:
Bias in results due to the choices of allocation methods were reported in previous
published LCA studies. So, the allocation method scenario analysis showed that the
allocation method- allocated by mass, might introduce some biased results in the
absolute values, but not in the relative ranking of the EFP of the treatments. Since
different impact assessment methodologies employ different characterization
factors for the same items, similar to the allocation method, the impact assessment
methodology scenarios showed that different impact methodologies could result in
diverging conclusions. So, a clear definition of the impact assessment methodology
and the version should be provided.
224

Time to pay off additional burden from SWRT:
A yield increase scenario analysis due to the presence of SWRT was conducted on
the simulated Ctrl dataset to estimate the time to pay off the additional burden from
using SWRT. The scenario analysis indicates that if SWRT can increase the yield by
100%, the additional environmental burden incurred for installing SWRT in GHG
and fossil fuel depletion can be paid off in 5 and 9 years, respectively. Other
indicators were lower. Several approaches could decrease the SWRT environmental
burdens such as reducing the SWRT membrane thickness, extending SWRT
membrane lifetime; and enhancing the membrane installation efficiency to reduce
the machinery and diesel cost.
Impact of high sensitivity parameter standard deviations on conclusions:
An uncertainty analysis using Monte Carlo simulations conducted on land use (LU)
indicated that yield was a highly sensitive parameter in this study. High yield would
effectively dilute the EFP. Conclusions from LCIA comparisons between treatments
could be changed when accounting for yield variations. A representative case
showing the yield uncertainty effect on the conclusion is the climate change impact
result: the LCIA mean values are different among treatments, but they are
considered statistically the same for 2012 treatments and 2013 treatments when
accounting for yield uncertainty. Similar cases are human toxicity and terrestrial
acidification.

225

7.2

Future work

This study was a preliminary study towards the environmental performance of
SWRT. Many modeled processes were less representative of current farmer
situations. There were missing data/ data gaps that should be filled to complete a
full evaluation of SWRT technology. The following aspects could be researched to
improve the assessment of the environmental footprint of SWRT.
Develop method to simulated SWRT yields:
There are always schedules and cost issues in expecting enough repeating years for
agricultural experiments. Reproducing experiments by using simulation methods is
a good approach to overcome these experimental limitations. Currently, corn
production on Sandhill farm without SWRT has been successfully simulated.
However, successful modeling of corn production using SWRT was not possible.
Thus, modeling corn production using SWRT will provide a meaningful
breakthrough to estimate the EFP of SWRT in sandy soil and other type of soils.
Develop SWRT processes using primary data:
Creating unit process for SWRT using primary datasets will be helpful to describe
the environmental performance of SWRT. These processes include but are not
limited to: SWRT membrane production process, SWRT machine production process,
and consumptions and emissions of SWRT installation processes.

226

Complete fertilizer emission inventory in sandy soil:
Collection of information about fertilizer emissions in sandy soil will enable
completion of the fertilizer emissions profile. The environmental footprint
differences between fertilizer emissions in sandy soil and loamy soil are indirect
consequences of SWRT application. Accounting for the indirect consequences of
SWRT application will help for better quantification of the EFP.
Continue yield collections:
Yield is a highly sensitive parameter in corn production studies. Continuing
collecting SWRT yield records could enhance the temporal representative of the
whole study.
Modified less representative processes:
A few processes in this study lack representation for U.S. conditions. So if possible,
improving the following processes will enhance the geography and technology
representation. These processes include but are not limited to: seed production
process, agricultural machine productions, machinery work flow processes (tillage,
fertilizing, sowing, harvesting).
Collecting missing data:
The missing data include but are not limited to irrigation pump production and the
drip tape installation process. To avoid unnecessary errors, transportation of inputs
and products was not included. When transportation information becomes available,
transportation should be included.
227

Improve experimental design:
The experimental designs for the year 2012 and 2013 were not supportive of
drawing conclusions regarding SWRT application. A well-plan experimental design
in the coming few years should be considered to obtain robust comparison data to
determine the EFP

228