U.S. Environmental Protection Agency
Office of Resource Conservation and Recovery
Evaluating the Environmental Impacts of Packaging
Fresh Tomatoes Using Life-Cycle Thinking &
Assessment:
A Sustainable Materials Management
Demonstration Project
FINAL REPORT
October 29, 2010
EPA530-R-11-005
-------
Evaluating the Environmental Impacts of Packaging
Fresh Tomatoes Using Life-Cycle Thinking &
Assessment:
A Sustainable Materials Management
Demonstration Project
FINAL REPORT
Prepared for:
Office of Resource Conservation and Recovery
U.S. Environmental Protection Agency
EPA Contract No. EP-W-07-003, Work Assignment 3-68
-------
Acknowledgments
The authors, Martha Stevenson, Christopher Evans, Julia
to thank the following individuals for their contributions
• Greg Norris, Sylvatica
• Cynthia Forsch, Eco-Logic Strategies
• Kevin Grogan & Terry Coyne, Pactiv
• Helene Roberts & Hugh Mowat,
Marks & Spencer
• Reggie Brown, Florida Tomato
Committee
• Chad Smith, Earthbound Farms
• Niels Jungbluth, ESU Services
• Bo Weidema, Ecoinvent
• Cashion East, University of Arkansas
• Peter Arbuckle & Glenn Schaible,
USDA
Forgie, and Louise Huttinger, would like
to this report:
• Jon Dettling, Quantis
• Bryan Silberman, Produce Marketing
Association
• John Bernardo, Sustainable
Innovations
• Peppy, Mastronardi Produce
• Clare Broadbent, World Steel
Association
• Dwight Schmidt, Fibre Box Association
• Ed Klein, TetraPak
• Gerri Walsh, Ball Corporation
• Bionaturae
We would also like to thank Deanna Lizas, Adam Brundage, and Nikhil Nadkarni at ICF
International for their support on this project, and Jennifer Stephenson and Sara Hartwell at EPA's
Office of Resource Conservation and Recovery for the input and guidance they provided for this
study.
-------
The U.S. Environmental Protection Agency (EPA)'s report, Sustainable Materials Management: the
Road Ahead, outlined a roadmap for shifting our focus from waste management to life cycle
materials management. Materials management is "an approach to serving human needs by using
or reusing resources most productively and sustainably throughout their life cycles" (EPA, 2009).
To promote the management of materials and products on a life-cycle basis, the report suggested
that EPA "select a few materials/products for an integrated life-cycle approach, and launch
demonstration projects." This demonstration project, conducted for the Office of Resource
Conservation and Recovery (ORCR), evaluates both the direct environmental impacts of packaging
options for fresh tomatoes, and the impact of tomato packaging decisions on the life-cycle
environmental impacts of the packaged product. We use packaging to deliver a product, so why
are the two analyzed separately from one another?
The goal is to demonstrate how life cycle thinking can be used as a tool to promote sustainable
materials management. It considers three packaging options for fresh tomatoes:
1. "Loose", or minimally-packaged tomatoes that are transported in a corrugated container
box with a General Purpose Polystyrene (GPPS) liner. Four tomatoes (2lbs) are purchased
at a time by the consumer in a Polyethylene (PE) produce bag;
2. "PS Tray", where four tomatoes (2 Ibs) are packaged in an Expanded Polystyrene (EPS) tray,
wrapped in a Polyethylene (PE) film, and transported in bulk in corrugated container boxes;
and
3. "PET Clamshell", where four (2 Ibs) tomatoes are packaged in a Polyethylene Terephthalate
(PET) clamshell container and transported in bulk within a corrugated container.
This report quantifies the environmental impacts associated with three different packaging
options for fresh tomatoes, but also assesses the effect of different packaging options on the life-
cycle impacts of growing, transporting, and retailing fresh tomatoes to consumers.
While this project uses the tools and approaches of LCA, it is not an ISO-compliant LCA and should
not be used to support any claims or make any definitive choices with regards to packaging or
product design. It is intended as a thought piece to expand both understanding of packaging and
its relationship to the product and current biases toward considering climate change as the only
impact category.
This report demonstrates how LCA can be used to answer two key questions posed by the Vision
2020 report:
-------
1. What are the significant environmental impacts associated with this
material or product?
By collecting data on the growing, packaging, storage and retail, and transportation of 100 pounds
of fresh tomatoes, we developed a preliminary life-cycle inventory (LCI) of inputs and outputs of
the system to assess the global warming, acidification, respiratory effects, and smog formation
environmental impacts (Exhibit ES 1).
Exhibit ES 1: Environmental impacts, by life-cycle stage, of three packaging options for producing
and delivering 100 pounds of fresh tomatoes from San Joaquin Valley, California, to Chicago,
Illinois
40
35
30
25
20
15
10
5
0
0.73
7.27
6.19
PET Clamshell
l Growing • Packaging
4.50
6.19
•__
PS Tray
Storage & Retail
0.73
3.61
6.19
• •__
Loose
• Transport
PET Clamshell
l Growing • Packaging
PS Tray
Storage & Retail
Loose
l Trans port
(a) global warming
(b) acidification
0.035
0.030
0.025
0.020
0.015
0.010
0.005
0.000
l.OE-02
4.0E-05
7.5E-0:
- J
4.0E-05
.4E-03
l.OE-02
4.0E-05
3.4E-03
1.1E-02
PETCIamshell PSTray Loose
I Growing • Packaging • Storage & Retail • Transport
PETCIamshell PSTray Loose
I Growing • Packaging • Storage & Retail • Transport
(c) respiratory effects
(d) smog
-------
We found that:
• Although the difference is modest across all three packaging scenarios, the PET clamshell
scenario is the most environmentally-intensive, primarily because it requires a greater
amount of plastic material and because PET manufacture uses more energy per pound
than the other two plastics (PE and PS).
• The contribution of transportation to global warming, acidification, respiratory effects, and
smog impacts are sizable, and may dominate the impacts from other life-cycle stages at
longer distances. A sensitivity analysis confirmed that the magnitude of transportation's
environmental impacts across these categories varies greatly, depending on total
transportation distance.
• The impacts associated with producing and transporting packaging for tomatoes are
surprisingly large relative to the impacts associated with growing tomatoes, especially
considering the relatively small amount of material required to package tomatoes.
We evaluated the amount of water consumed by growing, packaging, storing and retailing, and
transporting tomatoes. Since it may not be obvious, please note that the water used in transport is
associated with the production of diesel for the truck. The results (Exhibit ES 2) show that:
Irrigation of tomatoes during the growing phase dominates all three of the scenarios,
although manufacturing packaging is
Exhibit ES 2: Water consumption, by life-cycle
stage, of three packaging options for fresh
tomatoes delivered to Chicago, Illinois
still a significant source of water use.
This water use is associated with the
manufacturing processes of the
corrugated box and the generation of
hydroelectricity used as part of the
electrical grid mix.
Growing, packaging, and transporting
one pound of tomatoes to the
supermarket requires between 700 to
850 pounds (80-100 gallons) of water.
I Transport
I Storage & Retail
I Packaging
I Growing
PET
Clamshell
PS Tray Loose
-------
Exhibit ES 3: Comparison of the effect of
packaging and tomato spoilage on GHG
emissions
Loose Packed
PET Clamshell
PS Tray
Finally, we demonstrated how LCA can be used
to investigate the effect of packaging choices on
the life-cycle impacts of growing, transporting,
and storing and retailing tomatoes. Based on
estimates of how packaging reduced fresh
tomato spoilage rates, we showed how
different packaging options could either
increase or reduce life-cycle GHG emissions
relative to loose-packed fresh tomatoes,
depending on the emissions associated with
producing the packaging, and the effect that
packaging has on reducing spoilage before the
tomatoes are sold to consumers (Exhibit ES 3).
In a case where both PET clamshell and PS tray
packaging reduce the spoilage of tomatoes at retail, the GHG emissions from manufacturing PET
clamshell slightly increased overall GHG emissions relative to loose-packed tomatoes, while PS tray
packaging slightly reduced total emissions by reducing spoilage.
2. If all impacts are not being addressed, what more can be done?
This demonstration project established a framework that can be improved and extended by EPA
to:
• Gather further data on other impact categories, including eutrophication, carcinogens,
non-carcinogens, ozone depletion, ecotoxicity, land use, and social considerations;
• Support efforts to improve LCA methodologies or other life-cycle tools that evaluate hard-
to-quantify aspects of water and land use environmental impacts, and social impacts;
• Include other packaging options, such as processed tomato packaging in steel cans, glass
jars, or aseptic (pouch or carton) containers;
• Extending the analysis to include other vegetables. For example, assessing carrots (a
relatively hardy vegetable with a longer shelf life than tomatoes) or spinach (a vegetable
with a short shelf life and number of fresh and processed packaging options, similar to
tomatoes) alongside the tomato analysis. Other kinds of tomatoes, such as greenhouse-
grown tomatoes could also be included.
• Evaluating other packaging-product systems outside of produce from an integrated life-
cycle perspective.
-------
Since this demonstration project did not involve a full, ISO-compliant LCA of fresh tomatoes, it is
subject to several caveats and limitations that would need to be addressed before making any
claims or definitive choices with regards to packaging or fresh tomato production:
• Impacts associated with growing tomato transplants were not included due to data
limitations
• Impacts associated with ethylene spray were not included due to data limitations.
• Losses of tomatoes at the farm were not included due to data imitations. Although Kantor
et al. (1997) found evidence of losses in the growing and harvesting process, they were not
able to quantify the extent of these losses.
• Scenarios where tomatoes are repacked after harvest and before wholesale were not
included, even though it is understood as a standard practice. This decision was made to
limit the number of permutations of the study.
• Eutrophication, human toxicity and ecotoxicity impacts were not included as we were not
able to locate data on air and water emission from tomato growing process, other than
those associated with combustion of fuels used to run equipment.
• U.S. data sources were used where possible but also the majority of the background data
represented the European context, primarily from ecoinvent to fill data gaps.
• Environmental impacts associated with infrastructure or equipment were not included.
Overall, this demonstration project contributes to a shift toward sustainable materials
management by:
1. Evaluating the environmental impacts associated with different packaging options from an
integrated perspective of food production, packaging, and delivery.
2. Assessing environmental impacts from a life-cycle perspective.
3. Extending the analysis to a number of different environmental impact categories that
provide information relevant to EPA efforts to reduce GHGs, reduce air pollution, conserve
water, and reduce material use.
4. Applying LCA tools and thinking to characterize the material inputs and processes specific
to the life-cycle of fresh tomatoes, and the environmental impacts of these activities,
including both quantitative and qualitative discussions.
-------
5. Developing approaches for collecting and compiling LCI data from existing USDA databases.
By applying the tools of LCA to evaluate the impacts of packaging options for tomatoes, this study
illustrates the advantages of an LCA approach in evaluating a full range of environmental impacts
that can inform multiple EPA programs and priorities. It also demonstrates how LCA can be used to
assess trade-offs in environmental impacts along the life-cycle of materials and products, and to
identify hot spots and areas for further investigation. This report also identifies weaknesses in the
LCA tools and considers a full range of environmental issues that may not be supported by LCA
tools, but are valid considerations nonetheless.
-------
Table of Contents
Acknowledgments 3
Executive Summary 4
1 Introduction 13
1.1 Sustainable Materials Management: Vision 2020 Report 13
1.2 Impacts of Packaging Materials in the Vision Report 14
1.3 Designing the Packaging Demonstration Project 15
2 Methodology and Project Description 16
2.1 Goal of Study & Intended Application 16
2.2 Methodology and Functional Unit 16
Methodology 16
Functional Unit Definitions 17
Life-cycle Boundary Diagram 20
2.3 Fresh Tomato Process Description & Data Development 21
Process Description 21
Life-cycle Inventory Data Development 22
2.4 Packaging Manufacture Descriptions and Data Selection 23
Corrugated 23
Polystyrene Lining and Tray 25
Polyethylene Bag and Overwrap 25
PET Clamshell 26
2.5 Transportation Description and Data Selection 26
2.6 End of Life 27
10
-------
2.7 Life-Cycle Impact Assessment Methodology: TRACI 3.01 28
Global Warming 29
Acidification 29
Human Health - Criteria Air Pollutants (i.e., Respiratory Effects) 29
Smog Formation 30
2.8 Life-Cycle Impact Assessment Methodology: Water 30
2.9 Data Quality Assessment 30
2.10 Caveats and Limitations to the Model 31
3 Quantitative Results & Discussion 33
3.1 Results from TRACI 33
3.2 Results for Water Consumption 39
3.3 Sensitivity Analyses 40
Impact of Packaging on Spoilage and Shelf Life 40
Transportation 43
4 Qualitative Discussion 46
4.1 Water Use 46
4.2 Land Use 46
4.3 Other Environmental Impacts 47
4.4 Other Human and Social Impacts 48
4.5 Study in Context 50
5 Conclusions 52
6 References 57
Appendix A: Additional Analyses 62
11
-------
Processed Tomatoes 62
Processed Tomato Descriptions 62
BPA Migration 66
Other Vegetables 66
Appendix A References 69
12
-------
1.1 Sustainable Materials Management: Vision 2020 Report
In June 2009, the U.S. Environmental Protection Agency (EPA) released the report, Sustainable
Materials Management: the Road Ahead (aka The 2020 Vision). It outlined a roadmap for shifting
our focus from waste management to improved materials management, which involves
understanding and reducing life-cycle impacts, using less material, reducing use of toxic chemicals,
using more renewable materials, considering the substitution of services for products, and
recovering more materials. This shift will impact the way that our economy uses and manages
materials and products.
EPA's roadmap provides three broad recommendations:
1. Promote efforts to manage material and products on a life-cycle basis;
2. Build capacity and integrate materials management approaches in existing government
programs; and
3. Accelerate the broad, public dialogue necessary to start a generation-long shift in how we
mange materials and create a green, resilient, and competitive economy.
Under the first broad recommendation, the report suggests a specific course of action to "select a
few materials/products for an integrated life-cycle approach, and launch demonstration projects."
This report is one of two demonstration projects currently being conducted in the Office of
Resource Conservation and Recovery (ORCR) to begin implementation of these recommendations.
This demonstration project is focused on vegetable packaging and its relationship with the
product. A parallel demonstration project is being conducted by another group on construction
and demolition waste.
The Vision 2020 Report poses four questions to guide these demonstration projects:
1. What are the significant environmental impacts associated with this material or product?
2. What is currently being done to address the impacts associated with this material or
product?
3. If all impacts are not being addressed, what more can we (EPA) do?
4. What strategies for improvement are advised? How should progress be measured?
13
-------
This demonstration project will focus on the first and third questions proposed in the Vision
Report, as its subject matter seeks to look across various sectors identified in the Vision Report.
1.2 Impacts of Packaging Materials in the Vision Report
Packaging is not included as a stand-alone sector in the Vision Report, instead the materials
associated with packaging manufacture are included in their respective material-based sectors
(e.g., aluminum, paper) and the activities associated with packaging are included with their
respective processes (e.g., food processing and canning). For this reason, it is difficult to analyze
the associated impacts of packaging materials only using the input-output economic model, which
underlies the Vision Report. In order to gain a sense of the identified impacts by packaging
material sector, a summary table of three packaging base materials is provided in Exhibit 1-1.
Exhibit 1-1: Summary of Direct Impacts of Selected Packaging Material Sectors from Vision
Report (Rankings are out of 480 Industry Sectors)
Paper & Paperboard Making Plastic Materials & Resins
Rank Impact Area Rank Impact Area
Primary Aluminum
Rank Impact Area
1st
~nd
4th
5th
6th
gth
10th
12th
13th
17th
Marine Sedimental
Ecotoxicity Potential
Human Toxicity Potential
Energy Use
Freshwater Sedimental
Ecotoxicity Potential,
Acidification Potential
Marine Aquatic Ecotoxicity
Potential
Water Use
Terrestrial Ecotoxicity
Potential, Photochemical
Oxidation Potential
Freshwater Aquatic
Ecotoxicity
Global Warming Potential
Eutrophication Potential
4th
6th
10th
11th
19th
20th
Ozone Depletion
Marine Sedimental
Ecotoxicity Potential
Material Waste
Human Toxicity Potential,
Marine Aquatic Ecotoxicity
Potential, Freshwater
Sedimental Ecotoxicity
Terrestrial Ecotoxicity
Potential, Acidification
Potential
Freshwater Aquatic
Ecotoxicity
~nd
7th
gth
12th
14th
20th
Ozone Depletion, Marine
Aquatic Ecotoxicity Potential,
Freshwater Sedimental
Ecotoxicity Potential
Marine Sedimental
Ecotoxicity Potential
Human Toxicity Potential
Acidification Potential
Photochemical Oxidation
Potential
Terrestrial Ecotoxicity
Potential, Material Waste
14
-------
1.3 Designing the Packaging Demonstration Project
The first cited life-cycle assessment (LCA) study was conducted on packaging materials in the late
1960s (Baumann, 2004). Since that time, hundreds of LCAs have been conducted looking at various
packaging scenarios and their associated life cycles. These studies often do not include impacts of
the product as part of the analysis. Using the principles outlined in the Vision Report (i.e., to
reduce material and energy use across the whole supply chain) the project was approached from a
different angle: we need to assess packaging as part of the product supply chain to understand the
overall impacts and put them into context. Source reduction of packaging material has been a
long-standing practice in the field of waste/material management, but when this source reduction
leads to increased product damage (and typically increased overall environmental damage) it
results in a net detriment to human and environmental health. We use packaging to deliver a
product, so why are the two analyzed separately from one another?
In designing this demonstration project, the team wanted to analyze packaging's role in a product
supply chain that generated significant environmental impacts on its own, assuming that
delivering high-impact products safely to market was of utmost importance from an
environmental health perspective. The Vision Report reflected that those sectors housed in the
Food Products & Services division cause significant environmental burdens from production to
final consumption including meat production, dairy production, restaurants (eating and drinking
places), food preparation and others. Currently, several studies are being conducted in the United
States on meat and dairy products. These projects include primary data collection from their
specific industries and have significant budgets and longer timelines. Given that these areas were
already being analyzed in a more robust manner, the vegetable supply chain was selected for this
demonstration project. Please see Exhibit 1-2 for the Vegetable sector's identified impacts in the
Vision Report (EPA, 2009a).
Exhibit 1-2: Vegetable Sector Impacts Identified in Vision Report
Ranking
(of 480 sectors)
Impact Area
4tt,
6th
15th
16th
19th
Direct Water Use
Direct Freshwater Aquatic Ecotoxicity,
Ecotoxicity
Direct Terrestrial
Direct Land Use
Direct Global Warming Potential
Direct Material Use
15
-------
This section describes the goal of this study, key details of the LCA methodology, the processes
modeled, and data sources. It discusses the goal of this report within the context of the EPA's
2020 Vision Report, then describes the general LCA method that was applied to achieve this goal,
and explains key methodological details, such as how fresh tomatoes—the food product evaluated
in this study—were selected, definition of the functional unit for the study, and the packaging
comparisons, or scenarios, assessed by this study. This section then provides an overview of the
tomato growing, harvesting, packaging, and transportation processes that were modeled, and the
data sources that were used to assemble the life-cycle inventory (LCI) for fresh tomatoes.
2.1 Goal of Study & Intended Application
The goal of this demonstration project is to take the recommendations from the Vision 2020
Report and apply them toward deepening our understanding of packaging's environmental
impacts, positively and negatively, as related to delivery of produce. While this demonstration
project utilizes the tools of LCA, it is not an ISO-compliant LCA and should not be used to support
any claims or make any definitive choices with regards to packaging or product design. It is
intended as a thought piece to expand both understanding of packaging and its relationship to the
product and current biases toward considering climate change as the only impact category. As
outlined in the report, significant data gaps were encountered during this research including
limited data for some of the life cycle phases and limited emissions data for agricultural processes.
For those impact areas where these data gaps generate "low-confidence," we have used the
scientific literature to support a qualitative discussion rather than address them quantitatively.
2.2 Methodology and Functional Unit
Methodology
This study uses the tools of life cycle assessment (LCA), scientific literature, and expert interviews
to evaluate the environmental impacts of packaging and packaging's effects on the delivery of
produce. To evaluate the links between packaging and its life-cycle impacts on produce, this
analysis involves two main comparisons:
1. An assessment of the life-cycle environmental impacts of different produce packaging
choices compared to each other; and
2. An assessment of the environmental impacts of produce growing, transportation, and retail
compared to the life cycle impacts of packaging, and the effect of different packaging
options on the impacts of produce growing, transportation, and retail.
16
-------
To conduct this assessment, we developed a list of candidate produce for evaluation in this
analysis. In developing this list, we considered the number of packaging options available for
different fruits and vegetables, the existence of a "loose packaging" option (an option with no
packaging other than containers used to transport the product to the produce section in the
supermarket) that could serve as a baseline for comparison, and the availability of LCI data and
literature.
Based on an extensive survey of literature related to the production, packaging, and distribution of
produce, we selected tomatoes as the produce type for study using LCA. Of the candidates,
tomatoes offered a number of different packaging options, and a sufficient level of existing LCI
data and literature was available for tomatoes compared to the other vegetables.
Functional Unit Definitions
The functional unit is the reference unit against which the environmental impacts of a product
system are compared. It is necessary to select a functional unit in order to consistently compare
the impacts of the different tomato packaging options and to evaluate their influence on the life-
cycle impacts of producing and delivering tomatoes to the supermarket. There are a number of
different characteristics that could be used to define a functional unit for tomatoes. For example,
the functional unit could be defined in terms of:
• Mass or volume of tomatoes (e.g., one pound, or one cubic foot of tomatoes). This is the
most common functional unit chosen today, however it is increasingly coming into
question as product comparisons are being made (Schau, 2008).
• Nutritional value, a measure of the nutrient quality of different tomato products.
Nutritional value functional units have been used by other studies to make comparisons
across different types of foods or products (Carlsson-Kanyama et al., 2003). For example,
nutritional value could be used to compare fresh tomatoes and canned or processed
tomatoes.
• Quality of the tomatoes, such as their taste, color, or overall class or value. Quality is more
subjective, difficult to quantify, and may vary depending upon the form and use of the
product. Quality is often quantified and evaluated in studies that evaluate freshness or
shelf life (e.g., Parihar, 2007).
For the purposes of this study, we defined our functional units in terms of the mass of tomatoes.
Mass was selected as the basis for the functional units because we are making comparisons within
one specific type of tomato (i.e., fresh slicing tomatoes), and we are most interested in evaluating
the environmental impacts from the production, packaging, and distribution of tomatoes, rather
17
-------
than variations in quality or other characteristics. Consequently, we assume that the nutritional
value and quality of tomatoes is the same within a particular type of tomato, regardless of the
material used to package the tomato.
We performed two different analyses for this study, which required us to define two separate
functional units:
1. One hundred pounds (100 Ibs) of tomatoes delivered to supermarket. This functional unit
encompasses the life-cycle environmental impacts associated with growing, packaging, and
transporting tomatoes to a supermarket. We used this functional unit to compare different
tomato packaging options to each other (see sections 3.1 and 3.2), and also to investigate
the sensitivity of our results to the distance that tomatoes are transported for sale (see
section 3.3).
2. One hundred pounds (100 Ibs) of tomatoes delivered to consumer for consumption. This
functional unit encompasses the same processes as the first unit, but includes the
environmental impacts associated with spoilage of tomatoes at the supermarket prior to
purchase by the consumer. We used this functional unit to investigate the effect that
packaging can have on reducing environmental impacts from spoilage in the supermarket
(see section 3.3).
Drawing from the information we gathered on tomatoes, we defined the following three
packaging options for delivery of fresh tomatoes. These scenarios were used in both functional
unit comparisons:
1. "Loose", or minimally-packaged tomatoes that are transported in a corrugated container
box with a General Purpose Polystyrene (GPPS) liner. This box contains 25 pounds of
tomatoes, approximately 50 tomatoes1. Polyethylene (PE) bags used by the consumer to
transport tomatoes home were also included in this scenario. It was assumed that four
tomatoes (2 Ibs) were purchased at a time.
2. "PS Tray", or an Expanded Polystyrene (EPS) tray, where four tomatoes (2 Ibs) are
packaged in a PS tray, wrapped in a polyethylene film, and transported in bulk in
corrugated container boxes. This corrugated container holds 20 pounds of tomatoes and
associated packaging.
1 Fresh slicing tomatoes can range in weight from 4oz to 2lbs. It is assumed that the average weight of a slicing tomato is 8oz for
this study.
18
-------
3. "PET Clamshell", or a Polyethylene Terephthalate (PET) clamshell, where four (2 Ibs)
tomatoes are packaged in a PET clamshell container and transported in bulk within a
corrugated container that holds 20 pounds of tomatoes and associated packaging.
The material compositions of the three packaging options, or scenarios, that were developed for
this analysis are shown in Exhibit 2-1 below. According to industry expert, Cynthia Forsch, "Loose"
slicing tomatoes comprise 85 percent of the retail market in the United States, "PET Clamshells"
make up 10 percent of the market, "PS Tray" is two percent of the market, and the remaining 3%
are in other packaging formats (Forsch, 2010).
Exhibit 2-1: Composition of packaging materials in the three packaging scenarios for delivery of
100 pounds of fresh tomatoes
PET Clamshell
PS Tray
Loose
DPET
HPS
HPE
Corrugated
1234567
Pounds of packaging material per 100 pounds of tomatoes
19
-------
Life-cycle Boundary Diagram
Exhibit 2-2 provides a flow diagram of the overall process for tomatoes. The life-cycle boundaries
corresponding to each of the functional units defined for this analysis are shown by the dashed
lines. The outer box in blue defines the boundaries that were used to assess the qualitative
environmental impacts discussed in Sections 3.1 and 3.2; the inner box in red, which encompasses
"tomato spoilage" and "packaging disposal in landfill", defines the boundaries used in the
sensitivity analysis of packaging and spoilage rates provided in Section 3.3.
Exhibit 2-2: Process flow diagram of the tomato life-cycle and the life-cycle boundaries
established for this analysis
Life-cycle boundaries for quantitative analysis in Sections 3.1 and 3.2
Growing
transplants
Upstream production of fertilizer, pesticides, fuels and
electricity
Packaging
manufacture
iransportation
to warehouse
and distribution
to retail
Storage and
retail
spoilage &
landfill
disposal
Packaging
disposal in
landfill
Packaging
disposal
and
recycling
f
4.'
Transportation
to end use
Consumption
of tomatoes
Disposal of
Life-cycle boundaries for sensitivity analysis in Section 3.3
* Tomato growing includes: irrigation, plowing/tillage, application of fertilizer, but does not include other air, water, and soil emissions.
See Section 2.5 for details.
As shown in Exhibit 2-2, growing transplants, transportation to end use, consumption of tomatoes,
disposal of uneaten tomatoes (e.g., tomatoes cores, seeds, or uneaten leftovers) and the excretion
of tomatoes after consumption were not included in the assessment. Selection of the life-cycle
boundaries was informed by data availability and the primary goal of the analysis, which was to
focus on the environmental impacts associated with packaging and the effect of packaging on the
life-cycle impacts of tomatoes.
20
-------
2.3 Fresh Tomato Process Description & Data Development
Process Description
USDA statistics indicate that California and Florida are the leading states for growing tomatoes in
the United States. In 2007, California was responsible for 76% of the total harvested area for
tomato growing in the U.S. (USDA, 2010). Since data representative of U.S.-average tomato
growing practices were not available, California tomatoes, as the leader in U.S. tomato production,
were chosen to represent the data inputs for the growing process in this analysis.
California produces both fresh and processed tomatoes. Fresh market tomatoes are juicer and
harvested while immature, while processed tomatoes have a thicker skin and firmer consistency
limiting damage during harvesting and transportation (CFAIC, 2009). In 2006, fresh market
tomatoes had an average yield of 28,000 pounds per acre and are mostly grown in fields as bushes
from transplanted plants instead of seeds to ensure protection from weeds, disease, and pests.
Transplants are grown in commercial greenhouses and then transferred to the field as plug plants
using a mechanical planter. Prior to planting, farms prepare the ground by tillage, which includes
sub soiling, disking, rolling, land planning, and listing beds (Hartz et al., 2008). All tomato fields in
California are irrigated using subsurface drip, furrow, and sprinkler irrigation methods. Fertilizers
and pesticides are then mechanically applied to the crop to prevent disease and insect infestation.
Exhibit 2-3 shows the steps of the growing process for fresh tomatoes in California.
Exhibit 2-3: Diagram for fresh tomato growing
Planting using transplants
from a greenhouse
Cultivation, Application of
Fertilizers, and Irrigation
(mechanical trimming, fertilizer,
To harvesting
^^M f
application machinery, irrigation and
(mechanical trimming, fertilizer,
f pesticides, heribicides, and /
application machinery, irrigation and
ground water pumping machinery)
Fresh tomatoes are hand harvested at the mature green and pink stages and transported from the
field to the packing shed where they are rinsed and sorted. They are then put into a cool storage
area where they are sprayed with ethylene before being shipped out to market. Ethylene is a
naturally occurring gas produced by tomatoes that that is used to accelerate ripening and
promotes earlier coloration and maturing; spraying tomatoes with ethylene prior to shipping
accelerates the ripening process (LeStrange et al., 2008). Exhibit 2-4 shows the steps of the
harvesting process for fresh tomatoes in California.
21
-------
Exhibit 2-4: Process diagram for harvesting fresh tomatoes
^^^
Hand harvest
^^^ ^^j
4
Transported to
packing shed
^^ ^^
4
^^ ^^
Rinsed, sorted and
bulk packed
^^ ^^
4
^^
Cool storage &
ethylene
^^ ^^
After harvest, the packaging step can take many different forms. Fresh slicing tomatoes can be
packed in boxes in the field and are not repacked prior to market delivery. Some fresh slicing
tomatoes are packaged in the field and transported to a packing facility where they are
repackaged before being distributed for sale. Fresh tomatoes are typically packaged in 20 to 25-
pound containers (corrugated or reusable plastic containers "RPCs") for delivery to market. If the
packaging scenario includes an interim step, typically 1,000-pound reusable plastic bins or 2,500-
pound "gondolas" are used to transport tomatoes from the field to packing facilities. Once
packaged, fresh tomatoes are typically sold to retail supermarkets, restaurants, or wholesale in 20-
25 pound boxes (Brown, 2010).
Life-cycle Inventory Data Development
Due to data limitations, this study focuses on developing LCI data for field grown, fresh tomatoes,
as opposed to processed tomatoes. The following section discusses the data that were acquired to
model the phases associated with fresh tomato production, as described in section 2.3. For a more
detailed description of the process steps associated with processed tomatoes, refer to Appendix A.
For this analysis, we used California-specific data where possible, and filled data gaps with generic
information from other data sources, primarily the ecoinvent database managed by the Swiss
Centre for Life Cycle Inventories (2008). Exhibit 2-5 includes all California-specific raw data and
inputs used for the analysis. The data are based on the total number of acres planted. Fertilizer
and pesticide use data for California fresh and processed tomatoes were taken from the 2006 U.S.
Agricultural Chemical Usage Survey (USDA 2007). The California Department of Water Resources
estimated annual land and water use estimates, which were used to quantify total irrigation water
use for both crops, were from 2001. The data in the table were scaled to provide inputs in terms
of producing one kilogram (i.e., 2.205 pounds) of fresh tomatoes.
Exhibit 2-5: California Fresh Tomato specific data used as inputs for the growing process
INPUT UNIT DATA SOURCE
Acres Planted
Acres Harvested
Yield
Production
acres
acres
short tons
short tons
41,400
41,000
14
574,000
USDA 2010
USDA 2010
USDA 2010
USDA 2010
FERTILIZERS- data for entire CA acres of fresh tomatoes planted
22
-------
Nitrogen
Phosphate
Potash
Sulfur
1,000 Ibs
1,000 Ibs
1,000 Ibs
1,000 Ibs
9682.8
5116.5
6432.7
4358.6
USDA2007
USDA2007
USDA2007
USDA2007
PESTICIDES- data for entire CA acres of fresh tomatoes planted
Herbicide
Insecticide
Fungicide
Other
1,000 Ibs
1,000 Ibs
1,000 Ibs
1,000 Ibs
24.1
40
310.4
589.7
USDA2007
USDA2007
USDA2007
USDA2007
IRRIGATION WATER
Irrigation Water
Use
m3 per acre
3071.37
California Department
of Water Resources
2001
We were unable to find California- or tomato-specific information on the mechanical inputs used
for tilling, fertilizing, applying pesticides, and irrigating. We assumed that these tractors and
agricultural machinery do not vary widely; therefore we used data inputs from Nemeck and Kagi
(2007). Based on the descriptions of tomato tillage practices, three processes were chosen from
the data set to represent ground preparation: plowing for sub soiling, harrowing for disking, and
rolling. No data were available for the process of listing fields; therefore listing tomato fields is not
included in this analysis.
Nemeck and Kagi (2007) take account for diesel fuel consumption in agricultural machinery on a
per-hectare basis (2.47 acres). The following activities were considered part of the work process:
preliminary work at the farm (e.g., attaching the adequate machine to the tractor); transfer to
field (with an assumed distance of 1 km); field work (for a parcel of land of 1 ha surface); transfer
to farm and concluding work (e.g., uncoupling the machine).
Data on the greenhouse growing of transplants and harvesting were not available. Therefore, this
analysis covers the growing process from when the transplants arrive at the farm from the
greenhouse up to the time right before harvest begins.
2.4 Packaging Manufacture Descriptions and Data Selection
Corrugated
Corrugated is a paper product most often made from hardwood and softwood wood chips,
recycled paper, water, starches, and sizing. The process chosen for this study was corrugated from
"mixed fibre" including both recycled and virgin material content. The wood chips are typically
chemically pulped through the Kraft process, using heat and chemicals to separate the lignin from
the fibers. After pulping, the water-laden mixture is uniformly applied to a screen and fed through
23
-------
a series of rollers contained in a papermaking machine. During this process, the pulp goes from
containing 97% water to 3% water and is finished by winding into large reels of containerboard,
the base papers for making corrugated. Single-walled corrugated, the type of corrugated typically
used to construct a 20-25-pound box of tomatoes, consists of a corrugated medium affixed to two
sheets of linerboard on either side. These three base papers are fed into a corrugating machine
where steam is applied to "wrinkle" the center medium and starch is applied to attach the three
layers together. The corrugated sheet is then printed, die-cut, folded, and glued according to box
design.
Exhibit 2-6: Photograph of a cross-section of a corrugated container consisting of two linerboard
sheets on either side of a corrugated linerboard medium (Oksay, 2008)
In our interviews with tomato companies and retailers, we found that some tomato growers are
fully integrated and make their own boxes on-site. Others purchase their boxes from a
manufacturer. There is no standard method for this industry. Exhibit 2-5 is an example of a tomato
box with a liner.
Exhibit 2-7: Photograph of tomato box with liner. (Photo by Martha Stevenson)
24
-------
LCI data for corrugated box manufacture and conversion were selected from the ecoinvent 2.0
database (Swiss Centre for Life Cycle Inventories, 2008) with modifications made to match
electricity inputs to the U.S. electrical grid average. A study is currently being conducted by the
AF&PA on U.S.-represented corrugated manufacture; however those results had not been
released to the public for use in this report. There are notable differences in paper manufacture
between the United States and Europe, including species and cultivation of fiber feedstocks.
Polystyrene Lining and Tray
Polystyrene is a polymer derived from natural gas and crude oil through production of ethylene
and benzene. These two fractions are alkylated using a catalyst to produce ethylbenzene. The
ethylbenzene is dehydrogenated to produce the styrene monomer, which is then polymerized to
produce polystyrene. In this packaging study, polystyrene is used in its expanded form to produce
the tray and in its general-purpose form to produce the liner in the loose tomato corrugated box
(see Exhibit 2-8 for an example). Data for PS and associated conversion were used from the
ecoinvent 2.0 database with modified electricity to the U.S. grid average, but originally developed
by Plastics Europe (Swiss Centre for Life Cycle Inventories, 2008).
Exhibit 2-8: Tomatoes packaged in PS Tray (Photo by Martha Stevenson)
Polyethylene Bag and Overwrap
Low-Density Polyethylene is a polymer in the polyolefin family derived from steam cracking crude
oil and natural gas to produce ethylene. LDPE is produced by polymerizing ethylene in high-
pressure reactors using a catalyst. LDPE is typically used as a film in packaging applications through
the conversion process of blown film extrusion, where the pelletized resin is heated and fed
through a thin die to form a tube, continuously inflating it to form a thin tubular sheet that can be
used directly, or slit to form a flat film. Data for LDPE were obtained from the U.S. LCI database
25
-------
and represent the U.S. context (NREL, 2009). These datasets were developed in 2007 through
broad surveys of the plastics industry (FAL, 2007). Data for film extrusion conversion were used
from the ecoinvent 2.0 database with modified electricity to the U.S. grid average, but originally
developed by Plastics Europe (Swiss Centre for Life Cycle Inventories, 2008).
PET Clamshell
Polyethylene Terephthalate (PET) is a petroleum-based co-polymer made from the monomers
ethylene glycol and terephthaltic acid, which are derived from crude oil and natural gas
respectively. This plastic resin is used in many packaging applications including soda bottles, jars,
and clamshells for produce. In order to produce a PET clamshell, the resin material is fed into a
mold as a plastic sheet, heated to a specific temperature, and shaped into the desired form. This
process is called "thermoforming." Data for PET were obtained from the U.S. LCI database and
represent the U.S. context (NREL, 2009). These datasets were developed in 2007 through broad
surveys of the plastics industry (FAL, 2007). Data for thermoforming were used from the ecoinvent
2.0 database with modified electricity to the U.S. grid average, but originally developed by Plastics
Europe (Swiss Centre for Life Cycle Inventories, 2008).
2.5 Transportation Description and Data Selection
The transportation emissions attributed to tomatoes are determined by: (i) the mode of
transportation (e.g., truck, train, ship), (ii) the distance traveled, (iii) the fuel consumption and load
carried by the vehicle, and (iv) whether the transportation vessel returns empty (i.e., whether the
backhaul distance needs to be attributed to the original cargo or not).
Tomatoes are typically shipped in trucks or container vans. The ideal temperature for
transportation of mature, green tomatoes (i.e., tomatoes harvested prior to ripening) is between
55° to 70°F to prevent chilling damage at lower temperatures, and decay at higher temperatures.
As a result, tomatoes are may be shipped in refrigerated trucks to protect tomatoes. When
tomatoes are transported through areas with temperatures below freezing, the tomatoes can be
protected by minimizing contact with the floors and walls of the truck and by circulating warmed
interior air around the load. (University of California, 2010; USDA, 2006)
Transportation of fresh tomatoes from the field to retail in supermarket was modeled using the
following assumptions:
• Tomatoes are packaged at or near the field where they are grown, then transported to a
distribution or wholesale facility before retail in local supermarkets;
26
-------
• Long-distance transportation in a combination truck (equivalent to a class 8b heavy-duty
truck); short-distance transportation from distribution facility to supermarket in a city
delivery truck (equivalent to a class 4 or 5 heavy-duty truck);
• Long-distance transportation from San Joaquin Valley, California to Chicago, Illinois, a
distance of 2,155 miles;
• Short-distance transportation from distribution facility in Chicago to local supermarket, a
distance of 20 miles;
• Trucks are fully-loaded with cargo;
• Backhauls are not included (i.e., we assumed that the trucks return carrying other loads, so
backhaul trips do not need to be included in the analysis); and
• Although tomatoes may be refrigerated during transport when exterior air temperatures
are at, or above, the recommend temperature range for transport, we did not include
refrigeration estimates in the transportation of tomatoes. Refrigerated transportation will
increase fuel consumption and emit small amounts of refrigerant through "fugitive" leaks
in the refrigerant during transportation, and the relative impacts would increase as
transportation distance increases.
We modeled truck fuel consumption and emissions using equivalent European heavy-duty truck
models available from ecoinvent (Swiss Centre for Life Cycle Inventories, 2008).
2.6 End of Life
The end of life of the tomatoes was not included as part of the life cycle model. Please see section
4.4 for a discussion of tomato waste and spoilage (i.e., tomatoes are not consumed, but instead
rejected as waste due to damage, spoiling, or as waste in the food preparation process) and
human excrement (i.e., tomatoes are consumed by the end user for nutrition).
The end of life for the various packaging scenarios was included in the life cycle model, by
including treatment of these materials through recycling, incineration with energy recovery or
landfilling, based on national statistics. The proportions of waste pathways were applied using
EPA's Municipal Solid Waste: Facts & Figures Report from 2008 (EPA, 2009b) and are reflected in
Exhibit 2-9. One modification was made to the listed recycling rates. Corrugated is recycled at a
relatively high rate in the United States including both residential and commercial combined at
77%. Because our study includes corrugated in a commercial context, we modified the average
recycling rate to 95% by weight, as most supermarkets have dumpsters dedicated solely to
corrugated recycling due to economic advantages.
27
-------
Exhibit 2-9: End-of-life management assumptions for packaging material (EPA, 2009b]
Packaging Material
% Recycled
% Incinerated
with Energy
Recovery
% Landfilled
_
Corrugated Box
95
PET Clamshell
PS Liner & Tray
PE Wrap & Bag
0
6.9
14
20
18.6
17.2
80
74.5
68.8
2.7 Life-Cycle Impact Assessment Methodology: TRACI 3.01
The Tool for the Reduction and Assessment of Chemical and Other Environmental Impacts (TRACI)
is a mid-point Life Cycle Impact Assessment methodology developed by EPA. The TRACI model
assesses multiple impacts and includes U.S.-focused models for its assessment parameters. TRACI
was selected due to the geographic scope of the project - tomato growing and delivery within the
United States. Exhibit 2-10 includes a list of the TRACI impact categories, the units each category is
measured in, the geographic boundary of the model and the source information for the model
development. TRACI version 3.01 was used for this study.
Exhibit 2-10: Description of TRACI environmental impact categories
Impact Category
Units
Site Specificity Source for Model Development
Global Warming
kgCO2eq
Global
Ozone depletion
kg CFC-11 eq
Global
Intergovernmental Panel on Climate Change
World Meteorological Organization's "Handbook for
the International Treaties for the Protection of the
Ozone Layer"
Human health cancer kg benzene eq
United States Used CalTox Version 2.2 to develop a multi-media
model including 23 pathways of exposure. Only 330
chemicals characterized (Note: those chemicals
represent 80% of weight of releases listed in TRI)
United States
Human health non-
cancer
kg toluene eq
Smog formation
Used CalTox Version 2.2 to develop a multi-media
model including 23 pathways of exposure. Only 330
chemicals characterized (Note: those chemicals
represent 80% of weight of releases listed in TRI)
kg NOX eq
U.S. east or west Used model developed for California Air Resources
of the Board to develop a U.S.-specific model for TRACI.
Mississippi River,
U.S. census
regions, states
Human health criteria
pollutants (i.e., kg PM2.5 eq
respiratory effects)
U.S. east or west Model developed by Harvard School of Public Health
of the based on emissions fate & transport, and
Mississippi River, epidemiological studies on concentration-response
U.S. census and translation to mortality and morbidity effects.
regions, states
28
-------
Acidification
H+ moles eq
Ecotoxicity
kg 2,4-D eq
Eutrophication
kg N eq
U.S. east or west
of the
Mississippi River,
U.S. census
regions, states
United States
U.S. east or west
of the
Mississippi River,
U.S. census
regions, states
U.S.-specific model developed forTRACI based on
acidification factors for each US state and calculated
into four regions. Used US National Acid Precipitation
Assessment Program model.
Used CalTox Version 2.2 model and developed model
using concentration-to-source ratio for emissions and
impact-to-concentration ratio. Only 161 chemicals
characterized.
U.S.-specific model developed forTRACI
Due to data gaps in the tomato LCI, we do not have confidence in the analysis of several of the
impact categories and have excluded them from the study quantitatively to avoid "false positive"
results, i.e. although our model provides a number for a "low-confidence" impact area, we are
lacking data for the air and water emissions from the tomato growing process, so we could be
missing significant additional impacts. The impacts assessed quantitatively in the report include:
global warming potential, acidification, human health criteria pollutants (particulates, or
respiratory effects) and smog formation.
Global Warming
Global warming potential is an indicator for a product or system's contribution to climate change.
The ability of chemicals to retain heat on the earth (radiative forcing) is combined with the
expected lifetime of these chemicals in the atmosphere and expressed in C02 equivalents. TRACI
includes the Intergovernmental Panel on Climate Change's (IPCC's) 2001 Global Warming
Potentials with a 100-year time perspective.
Acidification
Acidification is the potential of a chemical emission to acidify ecosystems and thus disrupt the
chemical equilibrium of the ecosystem, including loss of species biodiversity and loss of soil
productivity. The main causes of acidification include coal-fired power plants, fuel combustion,
and livestock growing.
Human Health - Criteria Air Pollutants (i.e., Respiratory Effects)
Particulate matter is a complex mixture of organic and inorganic substances of varying dimensions,
which suspend in air. Given the complexity and variety in terms of chemical composition of
particulate matter, their characterization and quantification in air is typically performed on the
29
-------
basis of physical measures. The TRACI methodology normalizes particulate emissions to kg PM2.5
equivalents. Impacts to human health from particulates can include asthma, lung cancer,
cardiovascular issues, and premature death. Fuel combustion is a primary contributor to
particulate emissions.
Smog Formation
Smog Formation is the potential of ozone creation at ground level (i.e. tropospheric ozone)
through photochemical transformation of ozone precursor emissions. The main ozone precursor
compounds are nitrogen oxides (NOx) and non-methane volatile organic compounds (NMVOC).
Fuel combustion is a significant contributor to smog formation. Similar to criteria air pollutants,
smog can cause irritation to the respiratory system and induce asthma.
2.8 Life-Cycle Impact Assessment Methodology: Water
Water consumption was assessed using a single-issue impact assessment method provided in the
SimaPro software. The methodology is titled "Water vl.Ol" and was developed in 2008. It is
essentially a counting metric, meaning that it only counts inputs of water from different sources to
the modeled processes and does not characterize them toward specific impacts (e.g., removing
water from critical habitat or quality of water after use). The result is a total volume of "blue"
water (i.e., water removed from surface or groundwater) delivered to a process, where
information is present in the inventory data.
2.9 Data Quality Assessment
As described in the fresh tomato and packaging material process descriptions in sections 2.3 and
2.4, an LCI was compiled from public sources for use in this report. The data are specific to the U.S.
context and thus the geographic and temporal representation is considered high, however due to
significant gaps, the data are not complete. Other than the combustion of fuels in on-farm
machinery, no air or water emissions from the tomato growing stage were included in the dataset.
A data gap this significant would not be accepted for an ISO peer-reviewed LCA. Overall, while this
data development represents an innovative approach toward working in a constrained and data
poor U.S. environment, the quality is considered moderate to low for the eutrophication,
carcinogenic, non-carcinogenics, ecotoxicity, ozone depletion indicators, and GHG emissions from
nitrous oxide (N20) released by application of synthetic fertilizers to tomato fields.
Data for PET and LDPE were obtained from the U.S. LCI database and also represent the U.S.
context (NREL, 2009). These datasets were developed in 2007 through broad surveys of the
plastics industry. This data quality is considered to be high. (FAL, 2007)
30
-------
Data for PS and plastic conversion were used from the ecoinvent 2.0 database with modified
electricity to the U.S. grid average, but originally developed by Plastics Europe (Swiss Centre for
Life Cycle Inventories, 2008). The geographic representation of this data is low-quality; however
the technological representation is on par with U.S. operations.
Data for on-farm processing, corrugated manufacture, transportation, and storage were taken
from the ecoinvent 2.0 data and the DK food database (Nielsen et al., 2007; Swiss Centre for Life
Cycle Inventories, 2008). Data quality with regards to completeness is considered high, however
due to the geographic and technological differences the overall data quality is considered
moderate.
2.10 Caveats and Limitations to the Model
A list of various assumptions and limitations to the life cycle model for the various scenarios has
been included below as a reference for the reader. Additional work on this demonstration project
would provide an opportunity to address some of these limitations.
• Impacts associated with growing tomato transplants and infrastructure were not included
due to data limitations.
• Impacts associated with ethylene spray were not included due to data limitations.
• Losses of tomatoes at the farm were not included due to data imitations. Although Kantor
et al. (1997) found evidence of losses in the growing and harvesting process, they were not
able to quantify the extent of these losses.
• Scenarios where tomatoes are repacked after harvest and before wholesale were not
included, even though it is understood as a standard practice. This decision was made to
limit the number of permutations of the study.
• Although tomatoes may be refrigerated during transport when exterior air temperatures
are at, or above, the recommended temperature range for transport, we did not include
refrigeration estimates in the transportation of tomatoes. Refrigerated transportation will
increase fuel consumption and emit small amounts of refrigerant through "fugitive" leaks
in the refrigerant during transportation, and the relative impacts would increase as
transportation distance increases.
• Since the effects of packaging on environmental impacts from transportation to end use,
consumption, disposal of uneaten tomatoes (e.g., tomatoes cores, seeds, or uneaten
leftovers), and excretion are minimal, these were excluded from the study.
31
-------
• Eutrophication, human toxicity and ecotoxicity impacts were not included as we were not
able to locate data on air and water emission from tomato growing process, other than
those associated with combustion of fuels used to run equipment.
• Used U.S. data sources where possible but also incorporated European data, primarily from
ecoinvent to fill data gaps.
• Environmental impacts associated with infrastructure or equipment were not included.
32
-------
3
3.1 Results from TRACI
Exhibit 3-1 compares the environmental impacts of the three packaging options for fresh tomatoes
grown in San Joaquin Valley, California and delivered to retail in Chicago, Illinois. These results
include the environmental impacts from growing the tomatoes in the field, harvest, packaging,
transportation to wholesale in Chicago, and distribution to retail in a Chicago supermarket, but are
subject to the data limitations described in section 2.10. The packaging scenarios are the only
aspect that varies in the results shown in Exhibit 3-1. Data on the absolute environmental impacts
for the three packaging options are provided in Exhibit 3-2.
Packaging fresh tomatoes in PET clamshells has the greatest impact across the four environmental
impact categories (global warming, acidification, respiratory effects, and smog). The
environmental impacts from packaging fresh tomatoes in PS trays with PE wrapping are five to ten
percent lower, and eight to 15 percent lower for loose tomatoes transported in corrugated
containers, across the categories shown in Exhibit 3-1. The ranking of the packaging options is
similar across the four environmental impact categories in Exhibit 3-1 because these categories are
all influenced primarily by the combustion of fuels.
The PET clamshell scenario is the most intensive, primarily because it requires a greater amount of
plastic material and because PET manufacture uses more energy per pound than the other two
plastics (PE and PS). But it is important to note that the difference between all three packaging
options is somewhat modest.
33
-------
Exhibit 3-1: Relative comparison of environmental impacts of three packaging options for fresh
tomatoes delivered to from San Joaquin Valley, California to Chicago, Illinois
100%
90%
PET Clamshell
PS Tray
Loose
Global Warming Acidification Respiratory effects
Impact Category
Smog
Exhibit 3-2: Environmental impacts of three packaging options for fresh tomatoes delivered to
Chicago, Illinois
Impact category
Global Warming
Acidification
Respiratory effects
Smog
Unit per 100 Ibs. of
tomatoes
kg CO2 eq
H+ moles eq
kg PM2.5eq
kg NOx eq
PET Clamshell
34.5
7.62
0.0294
0.104
PS Tray
31.8
6.94
0.0262
0.0980
Loose
30.9
6.73
0.0252
0.0954
34
-------
Exhibit 3-3 summarizes the contribution of each life-cycle stage (i.e., growing, packaging, storage
and retail, and transportation) to global warming, acidification, respiratory effects, and smog
impacts across the three packaging scenarios. These graphs include environmental impacts from
growing the tomatoes in the field, harvest, packaging manufacture, transportation to wholesale in
Chicago, and distribution to retail in a Chicago supermarket.
Exhibit 3-3: Environmental impacts, by life-cycle stage, of three packaging options for 100
pounds of fresh tomatoes delivered to Chicago, Illinois
PET Clamshell
I Growing • Packaging
PS Tray
Storage & Retail
Loose
• Transport
PET Clamshell
I Growing • Packaging
PS Tray
Storage & Retail
Loose
I Trans port
(a) global warming
(b) acidification
0.035
0.000
PETCIamshell PSTray Loose
• Growing • Packaging Storage & Retail • Transport
(c) respiratory effects
Note change in units on the vertical axis of each graph.
PETCIamshell PSTray Loose
l Growing • Packaging • Storage & Retail • Transport
(d) smog
35
-------
The impact from packaging is surprisingly high across these four impact categories, given the
relatively small amount of packaging material used in the packaging of 100 pounds of tomatoes
(i.e. 4.5 to 7 pounds of packaging material per 100 pounds tomatoes, depending upon the
packaging scenario). This may be a result of the categories assessed, which are driven primarily by
emissions from the combustion of fuels used for process heat and electricity in the manufacturing
processes. Similar to the results shown in Exhibit 3-1 and Exhibit 3-2, the environmental impacts
associated with PET clamshell packaging are the largest of the three packaging scenarios.
Transportation of the tomatoes and packaging from San Joaquin Valley to Chicago contributes to
over 50% of the total impact in most categories across the three packaging scenarios. The impacts
associated with tomato growing are slightly larger than packaging. Impacts associated with cool
storage of the tomatoes after harvest and at the retail store contribute to a very small fraction of
the total environmental impact.
Exhibit 3-4 reflects the comparison between the impacts of growing, transport, and storage of the
tomatoes to manufacturing the packaging and transporting to retail. The embodied impacts
associated with the tomato are typically three times those impacts associated with the packaging.
These results are well-supported in the Industry Council for Packaging and the Environment
(INCPEN) report "Table for One: The Energy Costs to Feed One Person" where primary and
transport packaging comprise 10% of the energy burden for one person's weekly consumption of
food (INCPEN, 2009). These results suggest that we should prioritize activities toward decreasing
the impacts of food production, consider sustainable consumption of food, and holistically
approach the relationship between packaging and product.
Exhibit 3-6 presents the results for GHG emissions emitted from growing, packaging, storing,
transporting (from San Joaquin Valley, California to Chicago, Illinois), and disposing of or recycling
the packaging of 100 pounds of loose, fresh tomatoes. The results are presented for each stage of
the life-cycle, with each bar comprised by the GHG emissions of the activities within a given life-
cycle stage. Presenting the results in this way allows for a clear representation of the major
sources of emissions, primarily in the transportation stage, but also in the growing and packaging
stages.
36
-------
Exhibit 3-4: Comparison of environmental impacts of manufacturing and transportation of three
different packaging materials relative to growing and transporting fresh tomatoes from San
Joaquin Valley, California to Chicago, Illinois
Global Warming Acidification Respiratory effects Smog
Impact Category
H Tomato Growing & Transport PET Clamshell & Transport PS Tray & Transport Loose & Transport
Exhibit 3-5: Environmental impacts of three packaging options for fresh tomatoes delivered to
Chicago, Illinois
Impact category
Global Warming
Acidification
Respiratory
effects
Eutrophication
Unit per 100
Ibsof
tomatoes
kg CO2 eq
H+ moles eq
kg PM2.5eq
kg N eq
Tomatoes &
Transport
27.3
5.90
0.0218
0.0124
PET Clamshell &
Transport
7.27
1.71
0.00754
0.0158
PS Tray &
Transport
4.50
1.03
0.00435
0.01022
Loose &
Transport
3.61
0.828
0.00343
0.00734
37
-------
Exhibit 3-6: GHG emissions from growing, packaging, retail and storage, transportation from San Joaquin Valley, California to
Chicago Illinois, and packaging end of life for loose, fresh tomatoes
25 -,
to
CL>
O
-M
03
E
o
to
-------
3.2 Results for Water Consumption
Exhibit 3-7 reflects the total amount of water consumed for the three different packaging
scenarios of 100 pounds of tomatoes delivered to Chicago, Illinois. The water used during
irrigation of tomatoes during the growing phase dominates all three of the scenarios. However,
the packaging is still a significant source of water use. This water use is associated with the
manufacturing processes of the corrugated box and the hydroelectricity used as part of the
electrical grid mix. Transportation (fuel production) and storage contribute small amounts to the
overall water use.
Overall, producing one pound of tomatoes requires nearly 500 pounds (58.5 gallons) of water in
the growing cycle, 175 to 300 pounds (21-37 gallons) of water for packaging, and 45 pounds (5.3
gallons) of water in storage and transportation.
Exhibit 3-7: Water consumption, by life-cycle stage, of three packaging options for fresh
tomatoes delivered to Chicago, Illinois
900
100
I Transport
Storage& Retail
I Packaging
I Growing
PET Clamshell
PS Tray
Loose
39
-------
3.3 Sensitivity Analyses
Impact of Packaging on Spoilage and Shelf Life
Spoilage and shelf-life affect the life-cycle environmental impacts of producing and consuming
tomatoes. A high spoilage rate influences the growing phase life cycle impacts because a large
quantity of tomatoes must be produced to supply the consumer with a given quantity. For
example, Oliver Wyman (2008) states that up to one in seven truckloads of perishables delivered
to a store will be thrown out—equivalent to a spoilage rate of 15 percent. Pydynkowski et al.
(2008) estimate that the amount of "shrink", or spoilage, in a given store can be as high as 10
percent. In a recent study by Cuellar et al., (2010), the authors found that over 32% of fresh
vegetables are wasted in the United States, based on a 1997 USDA paper (Kantor, 1997). As a
result, they concluded that the embedded energy in wasted food accounts for 2% of annual
energy use in the United States. Given the high rates of spoilage and the reduction potential from
packaging, this impact should be incorporated into full vegetable LCAs.
It is impossible to completely avoid food spoilage and waste, but a goal of this study was to
investigate the role that packaging can play a role in reducing spoilage by protecting produce from
damage and by extending the shelf life of fruits and vegetables in retail and in consumers' homes.
For example, Exhibit 3-8 below summarizes estimates of the shelf life of loose tomatoes relative to
different types of packaging, including Modified Atmosphere Packaging (MAP) filled with nitrogen
(N2) or carbon dioxide (C02), and canned tomatoes. Depending on the estimate, MAP packaging
can double or triple the shelf life of tomatoes; canning tomatoes allows for much longer storage
times, although direct comparisons between processed tomatoes and fresh tomatoes are difficult
due to the different functions of these food products. Note that we did not examine the effect of
different gas mixture atmospheres for the PS tray and PE wrap packaging scenario in this analysis.
However, the PET Clamshell and PS Tray examples would be considered "passive" MAP. Since
fruits and vegetables continue to respire after they are packaged, the C02 to 02 ratio changes
within the wrapped packaging to create a modified atmosphere, limiting presence of 02, without
actively changing the gaseous mixture.
Although comprehensive information on the effect of packaging on spoilage rates was not
available, we did find evidence that packaging can reduce spoilage in fresh tomatoes. For instance,
Marks & Spencer, a grocery store chain based in the UK, indicated that loose tomatoes suffer from
a 5.5 percent spoilage rate, while clamshell tomatoes have only 4.4 percent spoilage (Marks &
Spencer, 2010).
However, estimates of tomato spoilage as well as shelf life rates vary widely. U.S.-based estimates
suggest much higher rates of spoilage in retail: in a study of food loss rates across multiple food
40
-------
products, the USDA found that loose tomatoes have an average spoilage rate of 13.2 percent in
retail stores. Food waste in the home is even higher. According to the study, an additional 29
percent of tomatoes brought home from the supermarket are discarded as waste (Buzby, et al.,
2009). Exhibit 3-9 presents these varying spoilage rates.
Exhibit 3-8: Estimated shelf life of tomatoes across different packaging options
Material
Plastic
Plastic
Plastic
None
Metal
Metal
Type
MAP
MAP
MAP
Fresh
Canned
Canned
Description
Not specified
N2 atmosphere
CO2 atmosphere
Ambient air
Refrigerator storage
Opened, refrigerator storage
Unopened, pantry storage
Shelf life
2 weeks
3 weeks
1 week
1 week
1 week
3-5 days
1 year
Source
Hui et al. (2004)
Parihar, 2007
Parihar, 2007
Parihar, 2007
Boyer, 2009
Boyer, 2009; Northampton County
Cooperative Extension (2009)
Boyer, 2009; Northampton County
Cooperative Extension (2009)
Exhibit 3-9: Spoilage rates for loose tomatoes and PET clamshell-packaged tomatoes
Source Loose PET Clamshell
Retail Home Retail Home
Marks & Spencer Ts% ~ 4^4%-
USDA 13.2% 29% -- --
Spoilage rates have been incorporated to some extent in other tomato LCAs. For example,
Andersson and Ohlsson (1999) assumed a five percent product loss in the consumer-use phase for
tomato ketchup. To investigate the impact that changes in spoilage rates would have on the
overall environmental impact of providing tomatoes to the consumer, we conducted a sensitivity
analysis of the global-warming impacts of providing 100 pounds of tomatoes to the consumer at
the store.
For this analysis, we assumed that, on average 13.2 percent of fresh, loose tomatoes are discarded
at the supermarket as waste, based on data from Buzby et al. (2009). As a result, in order to
provide a consumer with 100 pounds of fresh tomatoes, the supermarket must stock 115 pounds
of tomatoes, since 13.2 percent of the 115 pounds will spoil (i.e., approximately 15 pounds of
tomatoes).
Next, we compared the global warming impacts of growing, packaging, and transporting tomatoes
in either PS trays or PET clamshell, assuming that these packaging options reduce the amount of
spoilage at retail by 2%, based on the estimates provided by Marks & Spencer (2010). As a result,
supermarkets will need to stock 113 pounds of tomatoes to provide the consumer with 100
pounds.
In all cases, we assumed that spoiled tomatoes and their associated packaging are sent to landfill
at end of life. We included estimates of methane generation in landfills (Barlaz, 1998), but we
41
-------
assumed that carbon dioxide emissions from tomatoes are carbon-neutral (i.e., carbon dioxide
emissions from tomatoes are end-of-life are balanced by their uptake of carbon dioxide as they are
grown). We included landfill gas capture at the U.S. national average, based on data from the U.S.
GHG Inventory (EPA, 2010), but assumed that this gas was flared and did not include an electricity
generation offset for energy recovery from captured landfill gas.
As shown in Exhibit 3-10, while packaging with PET clamshell reduces spoilage, the additional GHG
emissions associated with producing and transporting the additional packaging are higher than for
loose packed tomatoes. In contrast, packaging tomatoes in PS trays decreases spoilage, and the
overall GHG emissions—including the production and transportation of the additional packaging-
are slightly reduced relative to loose tomatoes.
Exhibit 3-10: Comparison of the effect of packaging and tomato spoilage on GHG emissions.
Results include GHG emissions from: growing tomatoes, manufacturing packaging, transporting
packaged tomatoes from San Joaquin Valley, California to Chicago, Illinois, supermarket retail,
and end-of-life of spoiled tomatoes and packaging
Loose Packed
PET Clamshell
PS Tray
This sensitivity analysis shows that the effect of different packaging options on tomatoes is
relatively modest from a life-cycle perspective. Even so, the high rates of spoilage and food waste
42
-------
in the United States suggests that a large portion of the environmental impacts associated with
producing and distributing tomatoes may be lost as waste in retail and in the home. As a result,
further investigation into options to reducing spoilage and food waste—including the role that
packaging can play in reducing spoilage by protecting produce and extending shelf life—is
warranted.
Additionally, it is important to note that the results of this analysis were based on very limited data
on the rates of spoilage in supermarkets, and the factors that contribute to spoilage of fresh
produce. This analysis does provide, however, an example of how adopting a life-cycle perspective
can be used to consider the upstream and downstream impacts of packaging decisions on fresh
food products. Further research and data development on similar effects of packaging on
tomatoes or other types of food products could be used to extend this analysis more broadly.
Transportation
Since emissions from the combustion of fuels in transporting tomatoes from San Joaquin Valley to
Chicago accounted for a majority of the environmental impacts of growing, packaging, and
delivering fresh tomatoes to the super market, we conducted a sensitivity analysis of
transportation distance to determine the effect that distance has on overall environmental
impacts.
The quantitative results described in sections 3.1 and 3.2 assume transportation from San Joaquin
Valley in California to Chicago, Illinois (a total distance of 2,175 miles). We developed a separate
scenario for transportation from San Joaquin Valley to San Francisco (total distance of 60 miles) to
evaluate the effect of transportation distance on overall environmental impacts. Both
transportation scenarios are summarized in Exhibit 3-11 below.
Exhibit 3-11: Transportation scenarios investigated in sensitivity analysis
Scenario
A
(default)
B
Trip Leg
1
2
1
2
Departure
San Joaquin Valley
Chicago, Wholesale
San Joaquin Valley
San Francisco, Wholesale
Destination
Chicago, Wholesale
Chicago, Supermarket
San Francisco, Wholesale
San Francisco, Supermarket
Mode
Combination truck
City delivery truck
Combination truck
City delivery truck
Distance
(miles)
2,155
20
40
20
Exhibit 3-12 summarizes the overall results of the sensitivity analysis. The exhibit shows that the
global warming, acidification, respiratory effects, and smog impacts are very sensitive to the
distance over which tomatoes are transported. Tomatoes that are grown in San Joaquin Valley and
delivered to San Francisco have roughly two-fifths of the smog-forming impact, half of the global
43
-------
warming and acidification impacts, and two-thirds of the respiratory effect impacts of tomatoes
delivered to Chicago.
Exhibit 3-12: Relative (i.e., normalized) environmental impacts of growing, packaging, storing,
and transporting 100 Ibs of tomatoes from San Joaquin Valley to Chicago (Scenario A) or San
Francisco (Scenario B) in PET clamshell packaging
• Chicago
IS San Francisco
kg CO2 eq
Global Warmine
Acidification
Respiratory effects
Impact Category
The results in Exhibit 3-12 are for tomatoes in PET clamshell packaging, but the relative results are
the same across the other fresh tomato packaging options included in this analysis (i.e., tomatoes
packaged in PS trays with PE film, and loose tomatoes).
The results of this sensitivity analysis can be compared across the three packaging types and two
transportation scenarios to yield insights into the relative sensitivity of the analysis to
transportation distance and packaging type. Our analysis, however, is also subject to the following
caveats and limitations:
• Although tomatoes may be refrigerated during transport when exterior air temperatures
are at, or above, the recommended temperature range for transport, we did not include
refrigeration or cooling in transportation of tomatoes. Refrigerated transportation will
increase fuel consumption and emit small amounts of refrigerant through "fugitive" leaks
in the refrigerant during transportation. Including refrigerated transportation would
44
-------
increase the impacts across the four categories assessed in this analysis, and the relative
impacts would increase as transportation distance increases.
• We assumed that tomatoes would be packaged on-site after harvest. To simplify the data
requirements and complexity of the analysis, we assumed that tomatoes were packaged
on-site, or close to the field where the tomatoes are grown. We did not include
transportation for the packaging materials to the tomato growing site. Transportation of
packaging materials to the farm or packaging site will not have a major impact on relative
comparisons between the three different packaging options as long as the transportation
modes and distances are roughly the same for each of the packaging types.
• We modeled truck fuel consumption and emissions using equivalent European heavy-duty
truck models available from ecoinvent (Swiss Centre for Life Cycle Inventories, 2008).
Equivalently sized European heavy-duty trucks have different rates of fuel consumption
and air pollutant emissions than trucks in the United States. Information on U.S. heavy-
duty trucks could be used to improve the accuracy of our estimates, but the relative
comparisons across packaging types and transportation scenarios are valid since they are
calculated consistently across the same set of data.
These results are well supported from other fresh produce studies including a case study on
apples, runner beans and watercress which reflects that transportation, and especially air-freight
transport, dominate the life cycles of over-seas production of produce. This study also reflects that
the electricity requiring phases including grading, packing storage, agro-chemical production and
transport dominate local production of produce (Sim, 2007). Broadly this can be interpreted to
suggest that when specific crops are in season, the environmental burden is lower by purchasing
locally. When crops are out of season, the energy-consuming activities of production and shipping
should be done where the best energy profile is achieved either through grid mix or fuel
mix/intensity of transport.
45
-------
4 Oii«si;:::a:fwo!Jn^€",3:;:;^.0:j)
While life cycle assessment is a sound methodology for comparisons, it is limited in application by
data availability and impact assessment method accuracy. In the spirit of the Sustainable Materials
Management report, the project team wanted to assess more than those impact categories
supported by a quantitative analysis. Both scientific journal articles and personal interviews were
conducted to support the following sections, which discuss other environmental and social
considerations that are not measurable due to data gaps and methodological limitations.
4.1 Water Use
The previous section discussed the water used primarily in irrigation and electricity generation.
This water is tracked through metering on pipes delivering water to the field or to the
manufacturing facility. Information on water including source, quantity, and quality is often
missing from LCI data (Mila I Canals, 2009). Life Cycle Impact Assessment (LCIA) methods for water
consumption are under development, but are not in wide use. LCA currently does not have the
capability to capture regional water issues because LCI data tend to be "site-independent,"
meaning data sets are not fixed to a geographic location. Water flows associated with the natural
water cycle, including precipitation and evapotranspiration, are not included in LCA, but are critical
to agricultural processes. As an example, the vegetable sector ranked 4th in water use in the Vision
2020 Report (EPA, 2009a). This is seen as a significant data gap and any robust assessment of
water issues with regard to agricultural processes would need to include an in-depth study of the
impacts of the process on the natural water cycle and site-dependent linkage of water from
specified sources with local water issues.
4.2 Land Use
EPA's report on Sustainable Materials Management (EPA, 2009a) found that the vegetables sector
ranked as the 15th-highest sector in land use impacts, distinguishing it as an important impact
category associated with the production of vegetables, including tomatoes. In this analysis, we
have considered the land area planted with tomatoes—both nationally, and in California—and
yield data from fresh and processed tomatoes, but we have not quantitatively evaluated the
environmental impacts associated with land use, competition for land use, or land use changes.
A thorough evaluation of land use impacts is difficult due to both methodological and data-
availability limitations. A number of methods for quantifying the environmental impacts of land
use in LCA have been proposed (see Finnveden et al., 2009 for a list of recent publications), but a
common framework has not been established in the field of LCA. A key methodological issue is
that there are a number of different land-use aspects associated with land-use activities, inputs
46
-------
from land, outputs to land, and the associated effects on the natural environment, resources, and
human society (Udo de Haes, 2006; Finnveden et al., 2009). While certain aspects, such as the
surface area required for agricultural activities, are compatible with current LCA methods, other
aspects are harder to integrate, or may be appropriately addressed through other life-cycle
approaches (Udo de Haes, 2006). Additionally, land use impacts will depend upon site-specific
conditions and use characteristics, making it difficult to generalize environmental impacts from
land use changes in LCA studies.
In addition to the methodological issues associated with quantifying land use impacts, we were
unable to locate sufficient data that would allow a more thorough evaluation of land use impacts
associated with tomato production. Further research and methodological development is needed
in order to extend the EPA's identification of land use as an important impact category for
agricultural products, such as tomatoes and other vegetables.
4.3 Other Environmental Impacts
As data on emissions to air, water and soil were not readily available for tomato production, with
the exception of fuels combusted in on-site equipment, several environmental impacts were not
analyzed. Eutrophication, the nutrient loading of ecosystems changing species balance and
viability, is a significant environmental issue that was not addressed in the quantitative analysis.
One of the significant causes of eutrophication is the application of fertilizer in agricultural
systems. According to estimates used for modeling nitrogen emissions from agricultural systems,
agricultural processes (both crop and animal production combined) are responsible for up to 87%
of ammonia (NH3) emissions globally and 47 percent of nitrous oxide (N20) emissions globally
(Brentrup, 2000). These emissions are difficult to track due to their variability influenced by soil
type, climatic conditions, and agricultural management practices. Accurate measurements would
require both considerable time and financial resources, and so models are typically used for
estimation rather than actual measurements.
To investigate the sensitivity of our GHG emission results at the growing stage, we conducted a
screening analysis using the Carbon Trust's Footprint Expert Crop Calculator (Carbon Trust
Footprinting Company, 2008); a tool that calculates the GHG footprint of agricultural crops in
accordance with IPCC guidelines for national GHG inventories (IPCC, 2006). The analysis showed
that N20 emissions from applying synthetic fertilizer to tomato fields could increase GHG
emissions at the fresh tomato growing stage by 30 percent, or a 1.8 kgC02e per 100 pounds of
fresh tomatoes (in addition to our current estimate of 6.2 kgC02e per 100 pounds of fresh
tomatoes). This is obviously a significant data gap that would need to be addressed before the
results of this study could be used to support decision-making with regards to packaging or tomato
production.
47
-------
Another missing impact, commonly included in LCA studies, is human and eco-toxicity. Again, this
assessment was removed from the analysis due to limits in the emissions to air and water
inventory data. Pesticides, herbicides, and fungicides are all used in the tomato production
processes. Given the generality of the USDA data, specific chemicals were not disclosed in the
available data. Another complicating factor is that toxicants tend to be used in smaller quantities
and are not reported well in LCI data, which typically consider mass material and energy flows of
systems. If specific data were obtained on the pesticides, herbicides, and fungicides used in the
tomato growing process, a risk assessment could be conducted to estimate exposure levels for the
farm worker or average consumer.
4.4 Other Human and Social Impacts
In addition to the direct life-cycle impacts from the packaging process and packaging's impact on
spoilage and shelf-life of tomatoes, packaging may have social impacts or be dependent on social
norms that indirectly affect the full life cycle of the tomato.
Packaging options may affect the frequency of shopping trips and/or the consumer's use of
secondary or tertiary packaging to transport tomatoes from the stores to the home. Consumer
transport requirements may be reduced if a given type of packaging significantly increases the
shelf-life of tomatoes. For instance, if one form of packaging doubles shelf-life, then a consumer
could buy twice the quantity at one time and eliminate every other trip to the store. This change in
local transport at the consumption stage is likely to be small compared to the life cycle impacts
from the tomato growing, packaging, and distribution phases, and is not considered in the
quantitative analysis.
Various types of packaging may also induce consumers to use multiple packaging. This study
examines the packaging used to make tomatoes ready to sell at the grocery store. However,
consumers may be more likely to place certain types of packaged tomatoes in a secondary
package such as plastic produce bags. The consumption of these additional packaging items is not
considered in this analysis2, as it is expected to vary widely by store, community, and larger region,
and no data clearly indicate the frequency and extent of consumer use of this additional
packaging.
Just as packaging types may impact consumer behavior, consumer satisfaction and demand rather
than environmental benefits may drive a retail outlet's choice of packaging. For instance,
according to Pactiv representative, Kevin Grogan, grapes sold in markets historically were not
2 The polyethylene bags used to take home loose tomatoes was considered in the "loose" tomato scenario, however the possibility
that a consumer would put a Polystyrene tray of tomatoes into a second bag was not considered.
48
-------
bunched together in plastic bags. However, consumers began pulling grapes off, squishing them
on the floor, and then suing grocery stores after pretending to slip on the grapes. After a number
of copycat lawsuits in the 1990s, suppliers began bagging grapes to avoid these legal battles
(Grogan, 2010).
Further, consumer preferences may depend on various non-environmental cultural indicators such
as hygiene or visual appeal. The former sustainability manager for Albertson's Grocery store chain,
Cynthia Forsch, suggested that in some areas such as NY and Florida, people consider loose fruit as
"dirty" and would therefore not buy it. Instead, people opt for tomatoes placed in Expanded PS
trays which are then shrink-wrapped. In these regions, the perception of hygiene is more
important than spoilage or shelf-life, although we did find evidence that packaging—specifically
PET clamshell packaging rather than PS trays—may also reduce tomato spoilage rates (Forsch,
2010). Reggie Brown, manager of the Florida Tomato Committee also noted that the way in which
produce is merchandised to the consumer can play a more important role in packaging decisions
than concerns about preservation (Brown, 2010). Companies may also choose certain types of
packaging to boost their advertising efforts. Plastic clamshells, for instance, provide a convenient
opportunity for marketing directly on the packaging.
This study does not evaluate the extent to which these social considerations impact the overall
lifecycle of tomatoes. However, previous LCA literature has highlighted the potentially significant
environmental impact at the consumption phase from individual behavior (Jungbluth et al., 2000).
The point of tomato spoilage after the growing phase is a contentious economic issue within the
life cycle of the tomato that could influence future choices of packaging and the overall life cycle
impacts. Specifically, Reggie Brown of the Florida Tomato Committee indicated that the amount of
tomato spoilage in the transport phase is likely very low, since there is an economic incentive to
provide packaging that protects the product (Brown, 2010). In contrast, retail store
representatives indicated that there is no financial mechanism to control in-transit spoilage.
Retailers can claim damage only on boxes worth at least $25, which is higher than the value of
tomato or other produce boxes. As a result, retailers essentially pay for any transit damages
themselves (Forsch, 2010). It is unclear who is actually responsible for any spoilage of tomatoes in
transit, or exactly how significant this shrink may be. These uncertainties are not considered in the
quantitative analysis.
An additional and potentially significant life cycle impact not considered in the quantitative
assessment is human excrement. Munoz et al. (2010) examined the impact of human excrement in
a typical Spanish diet, which includes tomatoes. Results of this full life-cycle study indicate that,
although food production is the major source of emissions, human excretion along with further
wastewater treatment is not a negligible process in eutrophication or global warming potential
(GWP) impact categories. In fact, human excretion contributes 17% of the overall emissions in
49
-------
these sectors. Since the environmental impacts from human excretion of consumed food is
separate from food packaging decisions, these impacts were not included in this study, although it
is important to note that impacts at this life-cycle can be significant.
Finally, social indicators, such as worker health, wage rates and hours, company behavior and
treatment of workers, and other related concerns associated with tomato growing and packaging
are not considered in the quantitative analysis. However, there is increasing discussion within the
LCA community about incorporating social indicators into LCAs either quantitatively or on a
qualitative basis (Andrews, et al., 2009). These issues may be relevant to this tomato analysis at
the packaging stage, since worker conditions may vary by type of packaging.
4.5 Study in Context
The analysis conducted as part of this study found the GHG footprint of producing, transporting,
and retailing a 4-ounce serving of a tomato to be 0.17 Ib C02e. As mentioned in the introduction to
the report, several meat and dairy studies are currently being conducted within the United States.
Preliminary results from a LCA on pork products indicate that 2.2 Ib of C02e are produced for every
4-ounce serving of pork. This study includes inputs and emissions from the following life cycle
phases: nursery to finish of the pig, sow barn (including feed and manure handling), processing,
packaging, retail (electricity and refrigerants), and consumer (refrigeration and cooking) (Smith,
2010). Another set of preliminary results from a recent study on the dairy industry reflects that 1.1
Ib C02e are generated for an 8-ounce serving of milk in the U.S (University of Arkansas, 2010).
While the boundaries, methods, and data of these studies differ from those utilized in this
analysis, this comparison supports the assertion that animal-derived food products cause greater
environmental impacts than solely plant-based food.
A study completed by the Department for Environment, Food and Rural Affairs (Defra) in the
United Kingdom calculated the greenhouse gas emissions associated with a number of agricultural
and horticultural commodities. In this study, a 4-ounce service of tomatoes was found to have a
GHG footprint of 2.35 Ib C02e (Williams 2006). It should be noted that these tomatoes were grown
in a greenhouse for their entire life, a more energy-intensive process than field growing tomatoes.
A study conducted in Sweden in 2003, reflected a twelve-fold difference in the energy inputs to
field grown tomatoes versus greenhouse grown tomatoes (Carlsson-Kanyama 2003). Using this
multiplier toward our results of 0.17 Ib C02e per 4-ounce serving would suggest a 2.04 Ib C02e per
4-ounce serving of tomatoes grown in a greenhouse in the United States.
Finally, we compared our results at the tomato growing stage against the Carbon Trust
Footprinting Company's Crop Calculator (2008). As described in section 4.3, the Crop Calculator
calculates the GHG footprint of agricultural crop production based on IPCC good practice
guidelines for national GHG inventories (IPCC, 2006). Using the same data inputs outlined in
50
-------
section 2.3, the Footprint Expert Crop Calculator estimated that growing 100 pounds of fresh
tomatoes would emit 7.06 kgC02—a 14 percent increase compared to this study's estimate of 6.20
kgC02per 100 pounds of fresh tomatoes. This difference is likely driven by two differences: first,
the Crop Calculator does not include an input for the irrigation of crops, which are included in this
analysis; second, the Crop Calculator includes N20 emissions from synthetic fertilizer applications,
which account or 25 percent of the final footprint. As explained in section 2.9, apart from the
combustion of fuels in on-farm machinery, no air or water emissions from the tomato growing
stage were included in the dataset.
Another interesting study that highlights sustainable materials management and identifies food
production as a significant focal area was recently conducted by Green Seal as part of their
development of a certification for the Restaurant Industry known as GS-46. "Eating and Drinking
Places" ranked high in the Vision Report analysis for several of the final consumption impact
categories including Eutrophication Potential (2nd), Terrestrial Ecotoxicity Potential (2nd), Land Use
Change (2nd) and Global Warming Potential (4th). In Green Seal's study they found that food
procurement was the most significant contributor to the restaurant industries' environmental
profile across several impact categories. This led to the development of a standard that focused
guidance on food procurement and source reduction as a hotspot area, including in-restaurant
waste audits to reduce the amount of food left on a consumer's plate and thus less upstream
purchase of food. This study is a good example of how life-cycle thinking can be applied to focus
areas of activity toward the greatest reduction in environmental impacts (Baldwin, 2010).
51
-------
This study applies the recommendations from the Vision 2020 Report (EPA, 2009) to food
packaging in order to deepen our understanding of the environmental impacts of packaging
related to the delivery of produce. Following on the Report's recommendation to "select a few
materials/products for an integrated life-cycle approach, and launch demonstration projects", this
study evaluates the environmental impacts of fresh tomatoes packaging options from a life-cycle
perspective, including the stages of growing, packaging , transportation, storage, and retail.
Although this study applies LCA as a tool, it is not an ISO-compliant LCA and should not be used to
support any claims or to make definitive choices with regards to packaging or product design.
Instead, this study provides a framework for the evaluation of packaging from an integrated
perspective that considers not just the impacts associated with the production of packaging itself,
but also the effects on packaged product as well.
Consequently, the results of this study inform four broad categories of conclusions. First, that the
method and framework established by this study promote a sustainable materials management
perspective of the environmental impacts associated with food packaging. Second, the results
enable us to identify potential areas of significant environmental impacts associated with fresh
tomatoes and three packaging options. Third, this study illustrates several advantages to adopting
a life-cycle perspective to evaluate sustainable materials management options. Based on these
conclusions, and the data gaps and limitations that we identified in this assessment, we
recommend specific areas for improving upon and extending this analysis. These four conclusions
are discussed in further detail as follows:
1. This study contributes to a shift towards sustainable materials management by:
a. Evaluating the environmental impacts associated with different packaging options
from an integrated perspective of food production, packaging, and delivery. This
involves evaluating not just the direct impacts from manufacturing different types
of packaging, but also the effect of different packaging options on the final
packaged product, its use, and disposal. For example, evaluating packaging's effect
on spoilage rates and how spoilage influences life-cycle environmental effects from
growing and delivering food.
b. Assessing environmental impacts from a life-cycle perspective. Instead of
addressing impacts from "siloed" economic sectors such as agriculture,
transportation, manufacturing, and retail/buildings, this study evaluates the
impacts from a cohesive product life-cycle starting at fresh tomato growing to point
of sale to consumers, and including the end-of-life disposal of packaging.
52
-------
c. Extending the analysis to a number of different environmental impact categories
that provide information relevant to EPA efforts to reduce GHGs, reduce air
pollution, conserve water, and reduce material use. If the budget and time to
collect primary data were provided, this study's framework could be extended to
analyze toxic chemical impacts, eutrophication and nutrient management, and land
use considerations. In addition, there are currently methodological issues in
assessing the full range of water and land use impacts from a LCA perspective.
Practitioners are actively involved in developing approaches to evaluate these
impacts more comprehensively.
d. Applying LCA tools and thinking to characterize the materials inputs and processes
specific to the life-cycle of fresh tomatoes, and the environmental impacts of these
activities.
e. Developing approaches for collecting and compiling LCI data from existing USDA
databases. As was demonstrated in this study for certain inputs to the tomato
process (e.g., fertilizers, pesticides, water use), it would be possible to synthesize
existing USDA and state-level data to quantitatively evaluate environmental impacts
for agricultural crops, although a thorough review and assessment of the data
availability for other impacts was beyond the scope of this analysis. This provides
an example of how existing information that is not yet integrated into LCI databases
can be used to develop or augment LCI data. This is particularly useful in a U.S.-
context, where there is currently a lack of LCI data, particularly with respect to
agricultural products.
2. This study finds that the following are significant impacts associated with the production
and packaging of tomatoes:
a. The contribution of transportation to global warming, acidification, respiratory
effects, and smog impacts are sizable, and may dominate the impacts from other
life-cycle stages at longer distances. A sensitivity analysis confirmed that the
magnitude of transportation's environmental impacts across these categories varies
greatly, depending on total transportation distance.
b. The impacts associated with producing and transporting packaging for tomatoes are
surprisingly large relative to the impacts associated with growing tomatoes,
especially considering the relatively small amount of material required to package
tomatoes. This result is assumed to be different if the tomatoes were greenhouse
grown instead of field grown. We have also identified a number of data gaps in this
53
-------
analysis, primarily at the growing stage, which would need to be addressed in order
to verify this finding.
c. The changes in impacts across the packaging options considered in this study were
relatively modest. Of the three options, the impacts from producing PET clamshell
packaging were greater than PS trays and loose tomatoes in terms of global
warming, acidification, respiratory effects, and smog impacts.
d. Producing, packaging, and distributing tomatoes require large inputs of water,
particularly at the growing and packaging stages. Growing, packaging, and
transporting one pound of tomatoes to the supermarket requires between 700 to
850 pounds (80-100 gallons) of water. The largest sources of water consumption
were irrigation at the growing stage, and the use of water in generating electricity
used in the production of corrugated containers and other packaging materials.
e. There is evidence that packaging can influence the quantity of tomatoes discarded
as waste due to spoilage before sale to the consumer. We located a number of
estimates suggesting that plastic MAP packaging can increase the shelf life of
tomatoes, potentially reducing the amount of fresh produces that spoils in the retail
store. It was less clear whether the type of packaging influences the quantity of
tomatoes damaged in transport, as we received conflicting accounts from industry
experts.
f. A sensitivity analysis of the effect of packaging on the life-cycle GHG emissions of
tomatoes demonstrated that different packaging options could increase or reduce
life-cycle GHG emissions, depending on the emissions associated with producing
the packaging, and the effect that packaging has on reducing spoilage before the
tomatoes are sold to consumers.
g. Other considerations that are relevant to packaging decision-making include:
product marketing and merchandising, hygienic or visual appeal of the product, the
influence of packaging and shelf life extension on consumers' trips to the store, and
in inducing consumers to use additional packaging.
h. The impacts associated with cool storage of the tomatoes after harvest and at retail
are minor compared to the other life-cycle stages.
3. By applying tools of LCA to evaluate the impacts of packaging options for tomatoes, this
study has illustrated the following advantages to a life-cycle approach:
54
-------
a. Evaluating as full a range of environmental impacts as possible allows the
assessment to inform multiple EPA programs and priorities. For example, the results
of this demonstration project are relevant to EPA's work in areas such as Design for
the Environment (DfE, http://www.epa.gov/dfe/). Green Chemistry
(http://www.epa.gov/gcc/). Resource Conservation
(http://www.epa.gov/epawaste/conserve/rrr/index.htm). Lean Manufacturing
(http://www.epa.gov/lean/), EPA's Sustainable Products Network (SPIN), as well as
EPA partnerships through groups such as the Sustainable Packaging Coalition (SPC).
b. The life-cycle perspective enables policy makers to assess trade-offs in
environmental impacts along the life-cycle. This helps ensure that environmental
impacts are assessed holistically and reduces the risk of missing important impacts
or inadvertently shifting impacts from one stage or sector to another. For example,
in evaluating the effect of packaging on spoilage rates we were able to identify
which packaging scenario reduced overall environmental impacts relative to loose
fresh tomatoes in cardboard containers.
c. Finally, LCA can identify hot spots and areas for further investigation. For example,
we found that transportation contributed significantly to life-cycle impacts across
several categories. In a sensitivity analysis, we were able to show the extent to
which these impacts could be mitigated by reducing transportation distance.
4. This study established a framework that can be improved and extended:
a. The data gaps and limitations included in this study can be improved upon by:
i. Gathering further data on other impact categories, including eutrophication,
carcinogenics, non-carcinogenics, ozone depletion, ecotoxicity, land use, and
social LCA considerations. Quantifying these environmental impacts may
require developing models to accurately assess the inputs and outputs from
the system; for example, modeling the flow of nutrient inputs to tomatoes
at the growing stage to evaluate eutrophication or N20 emissions from the
application of fertilizers.
ii. Supporting efforts to improve LCA methodologies or other life-cycle tools
that evaluate hard-to-quantify aspects of water and land use environmental
impacts, and social impacts.
b. The framework and results from this study can also be extended to evaluate:
55
-------
i. Other packaging options, such as processed tomato packaging in steel cans,
glass jars, or aseptic containers. We have included additional data on
processed tomatoes (as opposed to fresh tomatoes) in Appendix A to serve
as a starting point for evaluating these options in further work.
ii. Identifying additional data sources and using information from literature,
industry experts, or other resources to include greenhouse tomatoes, or
tomatoes grown in other parts of the United States (e.g., Florida, another
major producer in the United States).
iii. Extending the analysis to include other vegetables. For example, assessing
carrots (a relatively hardy vegetable with a longer shelf life than tomatoes)
or spinach (a vegetable with a short shelf life and number of fresh and
processed packaging options, similar o tomatoes) alongside the tomato
analysis. We have included information from the literature survey we
conducted on these three vegetable types in Appendix A.
iv. Evaluating other packaging-product systems outside of produce from an
integrated life-cycle perspective.
56
-------
Andersson, K., & Ohlsson, T. (1999). Including Environmental Aspects in Production Development:
A Case Study of Tomato Ketchup. Lebensmittel-Wissenschaft und-Technologie, 32(3), 134-
141. doi:10.1006/fstl.1998.0513
Andrews, E., P. Lesage, C. Benoit, J. Parent, G. Norris, and J. Reveret. (2009). Life Cycle Attribute
Assessment. Journal of Industrial Ecology 2009:13 (4): 565-578.
Baldwin, Cheryl, Wilberforce, Nana, and Kapur, Amit. (2010). "Restaurant and food service life
cycle assessment and development of a sustainability standard." International Journal of
Life Cycle Assessment. Published Online: 23 September 2010.
Bare, Jane C., Norris, Gregory A., Pennington, David W., and McKone, Thomas. (2003) "TRACI: The
Tool for the Reduction and Assessment of Chemical and Other Environmental Impacts."
Journal of Industrial Ecology 6.3-4 (2003): 49-78.
Barlaz, M. A. (1998). Carbon storage during biodegradation of municipal solid waste components
in laboratory-scale landfills. Global biogeochemical cycles, 12(2), 373-380. Retrieved from
http://cat.inist.fr/?aModele=afficheN&cpsidt=2339715
Baumann, Henrikke and Tillman, Anne-Marie. (2004). The Hitch Hiker's Guide to LCA. Lund,
Sweden: Studentlitteratur.
Boyer, R., and J. McKinney. (2009) "Food Storage Guidelines for Consumers." Virginia Cooperative
Extension (2009): n. pag. Web. 7 Dec 2009.
Brentrup, Frank, Kusters, Jurgen, Lammel, Joachim, and Kulmann, Hermann. (2000). "Methods to
Estimate On-Field Nitrogen Emissions from Crop Production as an Input to LCA Studies in
the Agricultural Sector." International Journal of Life Cycle Assessment 5 (6) 349-357.
Brown, R. (2010) Manger, Florida Tomato Committee. Personal communication with Christopher
Evans in September, 2010.
Buzby, J.C., H.F. Wells, B. Axtman, and J. Mickey. (2009). Supermarket Loss Estimates for Fresh
Fruit, Vegetables, Meat, Poultry, and Seafood and Their Use in the ERS Loss-Adjusted Food
Availability Data. United States Department of Agriculture Economic Information Bulletin:
44. March 2009.
57
-------
California Department of Water Resources. (2001). Annual Land and Water Use Estimates,
Agiricultural Land & Water Use Retrieved from:
http://www.water.ca.gov/landwateruse/anaglwu.cfm
Carbon Trust Footprinting Company. (2008). Footprint Expert Crop Calculator version 3.0.
Carlsson-Kanyama, A., Ekstrom, M. P., & Shanahan, H. (2003). Food and life cycle energy inputs:
consequences of diet and ways to increase efficiency. Ecological Economics, Ecological
Economics, 44(2-3), 293-307. Retrieved from
http://ideas.repec.0rg/a/eee/ecolec/v44y2003i2-3p293-307.html
CFAIC. (2009). California Foundation for Agriculture in the Classroom. Commodity Fact Sheet:
Processing Tomatoes. Retrieved from
http://www.cfaitc.org/factsheets/pdf/ProcessingTomato.pdf
Cuellar, A., M.E. Webber. (2010). Wasted Food, Wasted Energy: The Embedded Energy in Food
Waste in the United States. Environmental Science and Technology 2010: 44(16): 6464-
6469.
EPA. (2008). Municipal Solid Waste Generation, Recycling, and Disposal in the United States: Facts
& Figures. U.S. Environmental Protection Agency (EPA). Retrieved from
http://www.epa.gov/epawaste/nonhaz/municipal/msw99.htm
EPA. (2009a). Sustainable Materials Management: The Road Ahead. U.S. Environmental Protection
Agency (EPA). Retrieved from http://www.epa.gov/wastes/inforesources/pubs/vision.htm.
EPA. (2009b). Municipal Solid Waste in the United States: 2008 Facts and Figures, Data Tables. U.S.
Environmental Protection Agency (EPA). Retrieved from
http://www.epa.gov/epawaste/nonhaz/municipal/pubs/msw2008data.pdf
EPA. (2010). Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 - 2008. U.S.
Environmental Protection Agency (EPA). Retrieved from
http://epa.gov/climatechange/emissions/usinventoryreport.html
Finnveden, Goran, Hauschild, Michael, Ekvall, Tomas, Guinee, Jeroen, Heijungs, Reinout, Hellweg,
Stefanie, Koehler, Annette, Pennington, David and Suh, Sangwon. 2009. "Recent
developments in Life Cycle Assessment." Journal of Environmental Management 91 (2009)
1-21.
Forsch, C. (2010). Principle, Ecologic Strategies. Personal communication with Martha Stevenson,
ICF International on September 23, 2010.
58
-------
Grogan, K. (2010). Representative, Pactiv. Personal communication with Martha Stevenson in
September, 2010.
Hartz, T., et al. (2008). Processing Tomato Production in California. University of California,
Division of Agriculture and Natural Resources, Publication 7228.
Hui, Y. H., Ghazala, S., & Graham, D. M. (2004). Handbook of vegetable preservation and
processing. Marcel Dekker.
INCPEN. (July 2009). Table for One: The Energy Costs to Feed One Person. Retrieved from
http://www.incpen.org/displayarticle.asp?a=49&c=3.
Jungbluth, N., 0. Tietje, R.W. Scholz. (2000). Food Purchases: Impacts from the Consumers' Point
of View Investigated with a Modular LCA. International Journal of Life Cycle Assessment
2000: 5(3):134-142.
Kantor, L. S., Lipton, K., Manchester, A., & Oliveira, V. (1997). Estimating and Addressing America's
Food Losses. Food Review, 2-12. Retrieved from
http://www.ers.usda.gov/publications/foodreview/ianl997/
LeStrange, M., Schrader, W., and T. Hartz. (2000). Fresh-Market Tomato Production in California.
University of California, Division of Agriculture and Natural Resources, Publication 8017.
Marks & Spencer. (2010). Email communication between Hugh Mowat, Marks & Spencer, and
Martha Stevenson. September 24, 2010.
Mila I Canals, Llorenc, Chenoweth, Jonathan, Chapagain, Ashok, Orr, Stuart, Anton, Assumpcio,
and Clift, Roland. (2009). "Assessing Freshwater Use Impacts in LCA: Part I - Inventory
Modeling and Characterization Factors for the Main Impact Pathways." International
Journal of Life Cycle Assessment 14 (2009): 28-42.
Mourad, Anna Lucia, Coltro, Leda, Oliviera, Paula A.P.L.V., Kletecke, Rojane M., and Baddini, Jose
Paulo O.A. (2007). "A Simple Methodlogy for Elaborating the Life cycle Inventory of
Agricultural Products." International Journal of Life Cycle Assessment 12 (6) 408-413
(2007).
Munoz, I., L.Mila i Canals.,A.R. Fernandez-Alba. (2010). Life cycle assessment of the average
Spanish diet including human excretion. International Journal of Life Cycle Assessment
2010 (15):794-805.
59
-------
Northampton County Cooperative Extension. (2009). Is It Safe To Eat? Penn State Cooperative
Extension. Retrieved from
http://northampton.extension.psu.edu/Nutrition/safe to eat.html
Nielsen, P. H., Nielsen, A. M., Weidema, B. P., Dalgaard, R., & Halberg, N. (2007). LCA Food
Database. Retrieved November 10, 2009, from http://www.lcafood.dk/
Nemecek, T., & Kagi, T. (2007). Life Cycle Inventories of Swiss and European Agricultural
Production Systems. Final report ecoinvent v2.0. Swiss Centre for Life Cycle Inventories.
Retrieved from http://www.ecoinvent.ch/
NREL. (2009). U.S. Life-Cycle Inventory Database. U.S. Department of Energy, National Renewable
Energy Laboratory (NREL). Retrieved from http://www.nrel.gov/lci/
FAL. (2007). Cradle-to-gate life cycle inventory of nine plastic resins and two polyurethane
precursors. The American Chemistry Council. Retrieved from
http://www.americanchemistry.com/s_plastics/sec_pfpg.asp?CID=1439&DID=5336
Oliver Wyman. (2008). A Retailer's Recipe for Fresher Food and Far Less Shrink.
Oksay, W. (2008). "Glued in supports". Photograph of corrugated cardboard. Retrieved from
http://www.flickr.com/photos/oskay/2716373580/.
Parihar, P. (2007). Effect of packaging methods on the shelf life of tomato. Haryana Journal of
Horticultural Sciences, 36(1&2), 173-174.
Pydynkowski, K., A. Herchek, D. Drennan, and W. Chen. (2008). A Life Cycle Analysis for Tomatoes
in NH. Report prepared for ENGS 171 course. Spring 2008.
Schau, Erwin Meissner, and Fet, Annik Magerholm. (2008). "LCA Studies of Food Products as
Backgrouind for Enviornmental Product Declarations." International Journal for Life Cycle
Assessment 13 (3) 255-264.
Sim, Sarah, Barry, Mike, Clift, Roland, and Cowell, Sarah. 2007. "The Relative Importance of
Transport in Determining an Appropriate Sustainability Strategy for Food Sourcing."
International Journal of Life Cycle Assessment 12 (6) 422-431.
Smith, Rod. (January 29, 2010). "Pork Producers study footprint" Feedstuffs FoodLink: Connecting
Farm to Fork." Retrieved from
http://www.feedstuffsfoodlink.com/ME2/dirmod.asp?sid=&nm=&type=news&mod=News
&mid=9A02E3B96F2A415ABC72CB5F516B4C10&tier=3&nid=FA5F860E289A41F0808E7343
6DD1B169.
60
-------
Swiss Centre for Life Cycle Inventories. (2008). ecoinvent Database v2.0. Swiss Centre for Life Cycle
Inventories. Retrieved from http://www.ecoinvent.ch/
University of Arkansas. (2010). Executive Summary of National-Level Fuild Milk Carbon Footprint
Study. Full version not yet available.
University of California (2010) Postharvest Handling Systems: Fruit vegetables. Vegetable Research
and Information Center, University of California Cooperative Extension, UC Davis
Department of Plant Sciences. Retrieved from
http://vric.ucdavis.edu/postharvest/fruitveg.htm.
USDA. (2006). Protecting Perishable Foods During Transport by Truck. Handbook number 669.
Agricultural Marketing Service, Transportation and Marketing Programs. Originally
published in 1995, reprinted in 2006. Retrieved from:
http://www.ams.usda.gov/AMSvl.O/getfile?dDocName=STELDEV3021003&acct=atpub.
USDA. (2007). Agricultural Chemical Usage 2006 Vegetables Summary. U.S. Department of
Agriculture (USDA). Retrieved from
http://usda. ma nnlib.cornell.edu/MannUsda/viewDocument I nfo.do?documentlD=1561
USDA. (2010). U.S. Tomato Statistics. U.S. Department of Agriculture (USDA), Economic Research
Service (ERS). Retrieved from
http://usda. ma nnlib.cornell.edu/MannUsda/viewDocu me nt I nfo.do?documentlD=1210.
Williams, A.G., Audsley, E. and Sandars, D.L. (2006) Determining the environmental burdens and
resource use in the production of agricultural and horticultural commodities. Main Report.
Defra Research Project IS0205. Bedford: Cranfield University and Defra. Available on
www.silsoe.cranfield.ac.uk, and www.defra.gov.uk
61
-------
£\iry,r;,,{''.--'iir ~ -•? & - £•. ;Hi-r iisr riir :;• ii &r;"vi
-•; • ik^ ik^ -^ • ". --U.'. :•' -I • :: /•'.-L.!! i:.--. ..!—l^ .'1^.--•:.-. !!. -I -M*. ^.:IM ^ Jl^3
Due to time, data, and budget constraints, this project was only able to address one vegetable
type (i.e., tomatoes), and focused only on fresh tomatoes and their associated packaging options.
There are two directions that could be explored in further work that would leverage and extend
the existing work completed to date. One direction would be to analyze processed tomatoes and
the associated packaging, like diced tomatoes in steel cans, aseptic (pouch or carton) containers,
and glass jars. The second direction would be to analyze other types of vegetables and their
associated packaging; for example, examining vegetables with a longer shelf-life, such as carrots,
as a counter-point to tomatoes.
The next two sections provide an outline of work developed in these two directions and list data
sources identified during the initial research of this project. This information is provided to support
further exploration on the relationship of vegetables and packaging in sustainable materials
management.
Processed Tomatoes
While primary and secondary data on the packaging types associated with diced tomatoes
(processed) was available, the life cycle data for the actual processing of the tomatoes was not
readily available. Information on processing descriptions, possible life-cycle data sources, and
packaging data were identified and included below. Also, cursory research on the issue of
Bisphenol-A (BPA) in can linings was conducted and articles purchased to support this section.
Processed Tomato Descriptions
After tomatoes reach a processing plant, they are typically canned whole, diced, or pureed, or they
are frozen. Tomatoes that will be canned are first graded on color, firmness, defects, and size.
They are then washed thoroughly in order to remove contaminants. Most often tomatoes are
soaked for several minutes in large tanks with paddles or aeration that agitates the tomatoes and
loosens any dirt. A final rinse removes remaining debris.
Next the tomatoes are machine sorted, typically using a photoelectric sorter that removes green
tomatoes before peeling, and pink tomatoes after peeling. In the past, tomatoes were cored by
machine after being sorted. However, since tomato varieties are now bred with very small cores,
this step is no longer needed. Instead, after sorting, the tomatoes are peeled using a steam or lye
process. In California, most peeling is done by steam: fruit are placed on a moving belt and passed
through a steam box under high temperatures. Waste peels that are produced can be used as
fertilizer or animal feed or processed into other products. Tomatoes in the Midwest are typically
peeled by passing them under jets of hot lye (sodium hydroxide) or through a lye tank. The lye
62
-------
effectively breaks down the skin cells by dissolving the cuticular wax and hydrolyzing the pectin, at
which point the skins fall off. Potassium hydroxide is sometimes used instead of lye. Steam peeling
results in a higher tomato yield, but removes much less of the peel than lye. After each of these
processes, the tomatoes pass through a series of rubber disks or a rotating drum under high-
pressure water sprays to remove any adhering peel.
Before filling into cans, tomatoes are manually sorted to remove any rotten parts, and diced and
inspected for green or blemished dices if appropriate. They are then heated and packed in
enameled cans and lids. FDA standards of identity require that some form of tomato juice or puree
be used as the packing medium in the container. In addition, a small quantity of calcium (not to
exceed 0.045% by weight), organic acids (such as citric acid), sugar, and/or salt may be added. As
the can is sealed, steam is injected into the top. Once sealed, the cans are cooled by chlorinated
water or air to 100 degrees Fahrenheit before being shipped to stores. Canned tomatoes typically
have a shelf-life of 18-24 months.
The LCI data collected for processed tomatoes are provided in Exhibit A-l. Exhibit A-2 provides a
process flow chart for the steps involved in tomato processes.
63
-------
Exhibit A-l: Annual inputs for processed tomatoes grown in California
INPUT ^pir °ATA S°URCE
Acres Planted
Acres Harvested
Yield
Production
acres
acres
short tons
short tons
283,000
282,000
36
10,104,000
USDA 2010
USDA2010
USDA 2010
USDA 2010
FERTILIZERS- data for entire CA acres planted
Nitrogen
Phosphate
Potash
Sulfur
1,000 Ibs
1,000 Ibs
1,000 Ibs
1,000 Ibs
52014.6
20407.1
9445.2
1807.7
USDA 2007
USDA 2007
USDA 2007
USDA 2007
PESTICIDES- data for entire CA acres planted
Herbicide
Insecticide
Fungicide
Other
1,000 Ibs
1,000 Ibs
1,000 Ibs
1,000 Ibs
362.1
426.3
6669.2
1453.6
USDA 2007
USDA 2007
USDA 2007
USDA 2007
IRRIGATION WATER
Irrigation Water
Use
m3 per acre
3675.78
California
Department of
Water
Resources
2001
64
-------
Exhibit A-2: Process diagram for processing processed tomatoes after growing
Coring and Trimming (not
commonly used)
Sorting
• Energy
•Water (hydrosorters)
Peeling (steam or lye)
• Energy use (steam and lye)
•Water use (steam)
• Lye (sodium hydroxide)
• Peels (steam and lye)
•Wastewater (mainly lye)
Dicing (diced tomatoes only)
• Energyfdicing process)
•Calcium added to container
•Juice or puree input
•Calcium (cannot exceed 0.045%
weight)
•Organic acids, sugar, salt
Manual Sorting
Exhausting and Sealing
65
-------
BPA Migration
Bisphenol A (BPA) is a widely-used compound used in the manufacture of polycarbonate plastics
and epoxy resins. Studies have shown that BPA, a systemic toxicant and endocrine disrupter, has
reproductive and developmental health implications (EPA 2010). Since the public is exposed to
BPA through consumer food product packaging, and especially given its presence in children's
formula bottles and canned foods, the potential health and environmental impacts of BPA have
become a serious concern.
BPA is present in the tomato life cycle through the lining of steel cans used to package processed
tomatoes. Processed food from cans is a source of BPA exposure, as the substance can leach from
the lining into the ingested food product (FDA 2010). Studies have investigated whether BPA
migration is caused by storage conditions such as shelf time, heat, or can damage. Goodson et al.
(2004) and Cao et al. (2009) both concluded that storage conditions do not change BPA levels
significantly, indicating that most migration (80-100% of BPA from coating) occurs during the
canning processing stage under high heat sterilization. Few alternative can lining materials are
available for processed tomatoes due to their high acidity which breaks down vegetable-based
resins.
Given the health and environmental concerns on the effects BPA exposure, federal and state
governments have begun investigate the issue. The U.S. Environmental Protection Agency (EPA)
has identified BPA on the Concern List under the Toxic Substances Control Act (TSCA). While the
EPA has taken steps to address then environmental concerns of BPA, the Food and Drug
Administration (FDA) has addressed human health effects through re-assessing established safety
levels, pursuing further research in scientific findings and supporting efforts to replace or minimize
BPA levels in food can linings (FDA 2010). State governments have also taken action on their own
to impose legislation regulating BPA in consumer products. Connecticut, Minnesota, Wisconsin,
Washington, Chicago, and Suffolk County, N.Y., have all banned the sale of polycarbonate baby
bottles, food containers, and cups that contain BPA in an effort to reduce infant exposure levels
(EPA 2010).
Other Vegetables
The original proposal indicated that one or more vegetables could be studied for this
demonstration project. The team identified three possible vegetables during the preliminary
literature search and based on a multiplicity of packaging types: tomatoes, spinach, and carrots.
Because of the lack of life cycle inventory (LCI) data on agricultural processes, entire data sets had
to be developed for this project. This limited the amount of resources available for data
development of other vegetables. The following section outlines information found on spinach and
carrots in the preliminary literature search which could be used toward further analysis.
66
-------
Initially, we contacted industry experts and conducted a preliminary review of available literature
to identify the three candidate vegetables for this analysis: tomatoes, carrots, and spinach. These
vegetables were selected because each: (i) provides a number of packaging options that allow for
a wide range of potential packaging alternatives to be assessed in the comparative assessment; (ii)
has a "no packaging" option (i.e., sold as fresh produce); and (iii) a preliminary assessment
indicated that it was likely that sufficient LCI data could be compiled to conduct an analysis of food
packaging impacts, although data gaps were identified for the vegetable options.
Next, we conducted a detailed literature survey using several bibliographic databases, including:
AGRICOLA, CAB Abstracts, Biosis Previews, CA SEARCH - Chemical Abstracts, Food Science and
Technology Abstracts, and EMBASE (a comprehensive biomedical database). Each database was
searched using three different search categories:
1. [vegetable name] and packaging and shelf life impact keywords
2. [vegetable name] and packaging and LCA environmental impact keywords
3. [vegetable name] and LCA environmental impact keywords
Where [vegetable name] was replaced with each of the three candidate vegetables: tomatoes,
carrots, and spinach. This detailed survey produced a number of academic articles on packaging
and vegetable shelf life, LCAs of vegetables and packaging, and research on BPA. Titles of the
most-relevant articles located are summarized in Exhibit A-3.
Exhibit A-3: Summary of the titles of the most-relevant articles located in the detailed literature
survey, sorted by search term topic and vegetable
Search Term
Topic
Spinach
Tomatoes
Carrots
Packaging
effects on
food quality,
shelf life
• Microbial and quality
changes in minimally
processed baby spinach
leaves stored under super
atmospheric oxygen and
modified atmosphere
conditions
• Retention offolate,
carotenoids, and other
quality characteristics in
commercially packaged fresh
spinach
• Shelf life of fresh-cut spinach
as affected by chemical
treatment and type of
packaging film
• The antioxidant activity and
• Quality changes in fresh cut
tomato as affected by
modified atmosphere
packaging
• Maintaining quality of fresh-
cut tomato slices through
modified atmosphere
packaging and low
temperature storage
• Storage studies of tomato
and bell pepper using eco-
friendly films
• Effect of packaging methods
on the shelf life of tomato
(Lycopersicon esculentum
Mill.).
• Handbook of vegetable
• Improving the health-
promoting properties of fruit
and vegetable products
• Effect of modified
atmosphere packaging on
the quality and shelf life of
minimally processed carrots.
• Post harvest technology of
vegetables
67
-------
composition of fresh, frozen,
jarred and canned
vegetables
preservation and processing
LCAs of
environmental
impacts
associated
with food
production
• A new method for assessing
the sustainability of land-use
systems (II): Evaluating
impact indicators.
• An improved water footprint
methodology linking global
consumption to local water
resources: a case of Spanish
tomatoes
• Identification of the main
factors affecting the
environmental impact of
passive greenhouses.
• Including environmental
aspects in production
development: a case study of
tomato ketchup
• Environmental impact of
greenhouse tomato
production strategies using
life cycle assessment
approach
• Life-cycle assessment of
carrot puree
• Environmental life-cycle
assessment of agricultural
food production.
BPA
• Migration of bisphenol A
from can coatings-effects of
damage, storage conditions
and heating
The detailed literature survey provided a number of useful insights and findings on the available
literature:
1. In general, the detailed literature survey uncovered many articles on the effects of
packaging on food quality. In particular, articles focused on plastic films and Modified
Atmosphere Packaging (MAP), but we located less information on canning and frozen
foods.
2. There were a few (predominantly older) studies on canning and freezing spinach.
3. Most importantly, only a few articles summarized the life-cycle environmental impacts
associated with packaging or vegetable production. No comprehensive source of U.S.-
specific LCI data was located for production of any of the three vegetable options. Out of
the three vegetables, we found that a majority of LCA studies have been conducted on
tomatoes.
68
-------
The detailed literature review and project scoping uncovered an encouraging amount of data;
however, there were a number of important data gaps that we identified. Extending the analysis
conducted for tomatoes in this study to other vegetables such as carrots or spinach will need to
address these data gaps:
• U.S.-specific LCI data availability on food production. Although we uncovered useful U.S.-
specific LCI data sources, and contacted several U.S.-based LCI practitioners, we did not
locate a comprehensive data source for information on the targeted list of vegetables. We
anticipate that it will be necessary to develop information on the environment impacts of
each vegetable based on a number of secondary data sources, followed by review with
LCA practitioners and industry experts.
• Environmental impact data availability. Based on our detailed review of literature, we
anticipate that impact assessment data for environmental toxicity, human toxicity, and
eutrophication impacts will be harder to locate or develop than for land use (i.e., surface
area used), water use (i.e., blue water consumption), energy use, and GHG emission
impact categories. For impact categories where there is less data available, it may be
possible to use the more aggregate environmental impact data from the Vision 2020
report (EPA, 2009) to discuss potential impacts based on specific insights gained from the
comparative assessment of the packaging scenarios and their effect on food production
processes.
Appendix A References
California Department of Water Resources. (2001). Annual Land and Water Use Estimates,
Agiricultural Land & Water Use Retrieved from:
http://www.water.ca.gov/landwateruse/anaglwu.cfm
Cao, X., Corriveau, J., & Popovic, S. (2009). Migration of bisphenol A from can coatings to liquid
infant formula during storage at room temperature. Journal of Food Protection, 72(12),
2571-2574. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/20003741
EPA. (2010). U.S. Environmental Protection agency Bisphenol A Action Plan. Retrieved from
http://www.epa.gov/opptintr/existingchemicals/pubs/actionplans/bpa_action_plan.pdf
FDA. (2010). U.S. Food and Drug Administration: Update on Bisphenol A for Use in Food Contact
Application. Retrieved from
http://www.fda.gov/downloads/NewsEvents/PublicHealthFocus/UCM197778.pdf
69
-------
Goodson, A., Robin, H., Summerfield, W., & Cooper, I. (2004). Migration of bisphenol A from can
coatings—effects of damage, storage conditions and heating. Food Additives and
Contaminants, 21(10), 1015. doi:10.1080/02652030400011387
Hui, Y. H., Ghazala, S., & Graham, D. M. (2004). Handbook of vegetable preservation and
processing. Marcel Dekker.
USDA. (2007). Agricultural Chemical Usage 2006 Vegetables Summary. U.S. Department of
Agriculture (USDA). Retrieved from
http://usda. ma nnlib.cornell.edu/MannUsda/viewDocu me nt I nfo.do?documentlD=1561
USDA. (2010). U.S. Tomato Statistics. U.S. Department of Agriculture (USDA), Economic Research
Service (ERS). Retrieved from
http://usda. ma nnlib.cornell.edu/MannUsda/viewDocu me nt I nfo.do?documentlD=1210.
70
------- |