ELECTRONICS WASTE MANAGEMENT
IN THE UNITED STATES
APPROACH 1
April, 2007
Draft Final
EPA530-R-07-004a
Office of Solid Waste
U.S. Environmental Protection Agency
Washington, DC
-------
ACKNOWLEDGEMENTS
This Report is based on analyses prepared under contract for the Office of Solid Waste by
Eastern Research Group, Inc of Lexington, MA. The Office of Solid Waste would like to
thank especially Lynn Knight and Shelly Schneider for their assistance in developing the
model upon which this report is based and in drafting the report. This Office would also
like to thank Robin Ingenthron of American Retroworks Inc., Good Point Recycling and
the World Reuse, Repair and Recycling Association for his assistance on the end markets
discussion.
-------
TABLE OF CONTENTS
1.0 Introduction 1
1.1 Objectives and Scope 2
1.2 Overview of Methodology 3
1.3 Organization of the Report 5
2.0 Data and Assumptions in the Model 7
2.1 Historical Sales Data: Televisions, Cell Phones, and Personal Computer
Products 7
2.2 Assumptions Regarding the Life Span of Electronic Products 10
2.3 Average Weight Data: Televisions and Computer Equipment 17
3.0 Model Results 21
3.1 The Quantity of EOL Electronics Generated for Management Each Year 21
3.2 Estimating the Quantity of EOL Products Generated That Are Recycled Versus
Disposed 23
3.3 Estimating the Quantity in Storage 26
4.0 Quantity Available for Recycling Based on Observed Collection Rates .... 29
5.0 Examining End Markets of Material Collected Through Electronics
Collection Programs in the TJ.S 31
6.0 Summary and Conclusion 33
APPENDIX A AGE DISTRIBUTION OF ELECTRONIC PRODUCTS BASED ON
ANALYSES OF DATA FROM THE FLORIDA STATE DEPARTMENT
OF ENVIRONMENTAL PROTECTION A-l
APPENDIX B ESTIMATING THE QUANTITY OF EOL ELECTRONIC
PRODUCTS DISPOSED: ANALYSIS OF WASTE SORT
STUDIES B-l
APPENDIX C AVERAGE COLLECTION RATES FROM EXISTING ELECTRONICS
COLLECTION PROGRAMS C-l
-------
1.0 Introduction
Although electronics represent less than two percent of the municipal solid waste stream,
options have increased for reusing and recycling electronics in recent years. Over 800
communities have instituted electronics collection events to help manage obsolete
electronics from households1. Many manufacturers of personal computers now offer take
back programs on-line, at least seven states ban landfilling of certain electronics, and four
states have programs that institute state-wide recovery programs for used electronics.
Many other states are looking to pass similar legislation this year, and many are
interested in Federal action to harmonize electronics recovery laws.
Recycling end-of-life (EOL) electronics, rather than disposing of them, makes use of
valuable components and materials, thereby conserving natural resources and saving
energy. EPA has been active in promoting the recycling and reuse of EOL electronics
through various programs, including Plug-In To eCycling and the Federal Electronics
Challenge.
Policymakers at the Federal, state and local levels, as well as manufacturers, retailers,
recyclers, non-governmental organizations (NGOs) and many others are interested in
updated national estimates of how many TVs, PCs, cell phones and other common
electronic products are in storage, recycled, or disposed. In 1999, the National Safety
Council issued the first large-scale survey and analysis of electronic product recycling
and reuse in the United States2. However, since that time, consumption and disposal, as
well as reuse and recycling of electronics in the US has continued to mount along with
the need for updated data.
The International Association of Electronics Recyclers publishes a comprehensive
triennial report on the state of the electronics recycling industry in the US. This report
surveys "all electronics" that are recycled by the electronics recycling industry. Its
estimates of recycling include consumer electronics and electronic equipment from
industry and manufacturers (including medical equipment, robotics systems, movie
production equipment), and therefore do not highlight information specific to the
products that are the subject of our analysis.
In response to stakeholder requests for detailed examination of the sales and management
of the electronics most commonly addressed by community collection programs and state
recycling legislation, EPA looked at this issue from two different points of view. EPA
assembled two different data sets and used two different methodologies to estimate the
amounts of commonly handled electronics that are stored, reused, recycled and disposed.
Our results are detailed in two detailed reports, plus an Overview that summarizes both.
1 "817 cities and jurisdictions provide some type of electronics recycling services in the US." Gracestone
Inc. and E-Scrap News. "Public Sector Offering of E-Scrap Services: The Why and Why Not."
Presentation: E-Scrap Oct. 2006.
2 The NSC survey covered the years 1997 and 1998 and included the following electronic products:
desktop computers, mainframe computers, workstation computers, portable computers, CRT monitors,
computer peripherals, telecommunications equipment, and CRT TVs.
-------
Looking at both of the detailed reports together, it is evident that the results are quite
similar. We believe that the dual approaches lend credibility to the range of results
obtained and enable readers to view the results from several different and helpful angles.
The two detailed reports comprise the entire analysis:
o "Electronics Waste Management in the United States: Approach One." This
analysis relies primarily on market research data on sales of electronic
products. It then applies these sales data to some of the most comprehensive
collection information available to estimate product life spans and the
amounts of particular products that are ready for EOL management. From
these EOL estimates, we subtract the estimated quantity recycled to yield the
quantity disposed. This approach also provides information on the export of
CRT monitors and TVs, as well as the amount of selected electronics
cumulatively in storage.
o "Electronics Waste Management in the United States: Approach Two."
Approach Two relies primarily on government statistics on sales of electronic
products. It then uses the same lifespan data (with some modifications) as
Approach One to estimate EOL quantities. From these EOL estimates, we
subtract the quantity of selected electronics disposed to yield the quantity
recycled. This approach also provides information on the composition of
electronic products, as well as the number of select electronic devices entering
storage/reuse annually.
The report that follows is Electronics Waste Management in the United States: Approach
One."
Readers should consider that the information presented in both Approach One and
Approach Two provides a "snapshot" of electronics waste generation and management in
the United States in recent years. As products, usage patterns and EOL management
options change over time, purchase, storage, and end-of-life disposition patterns will also
change.
1.1 Objectives and Scope
In pursuing activities related to EOL electronics, information regarding the amount of
material potentially in need of EOL management needs to be up-dated periodically. This
report presents a compilation and assessment of data to establish a baseline of knowledge
that can be built upon as the nation moves forward in managing electronics. The scope of
products covered in this report includes:
• Personal computers (PCs), including desktops, portables, and computer monitors
• Televisions
-------
• Hard copy computer peripherals, including printers, scanners, and fax machines
• Computer mice and keyboards
• Cell phones
These products were chosen because they make up the majority of the electronic products
collected and have been the focus of electronics recycling initiatives at the federal, state,
and local level. This analysis includes products from all sectors of the economy (i.e.,
residential, commercial, and institutional).
The objectives of this study are to:
• Estimate the number and weight of products that will become obsolete and need
EOL management annually.
• Estimate what portion of EOL electronic products are recycled versus disposed.
• Estimate how much material that is ready for EOL management may be in
storage.
• Examine the collection rates experienced by existing electronics recycling
programs as an indicator of the amount of material that is, on a practical basis,
available for recycling.
• Examine the current situation regarding the end markets for TV and CRT
monitors collected for recycling.
It has been several years since a national study of electronics waste generation was
undertaken. The EPA cosponsored a study published by the National Safety Council in
1999 ("Electronic Product Recovery and Recycling Baseline Report"), which was the
first large-scale survey and analysis of EOL electronic product recycling and reuse in the
United States. The results of this study will update the available information on this topic
and provide a basis for supporting strategic and policy decisions aimed at providing
national, regional, or local solutions to a prominent issue.
1.2 Overview of Methodology
The estimates developed in this report are based on several sources of data. Sales data
are based primarily on industry data on product sales. In addition, this report relies on
data from the Florida Department of Environmental Protection and other state data, as
well as data bases developed by EPA for the publication of the report Municipal Solid
Waste in the United States: 2003 Facts and Figures.
The pattern of product use forms the methodological framework used in this study. This
pattern begins at the point the product is purchased and ends with its final disposition.
Figure 1.1 depicts the framework used in this analysis. As shown in the figure, the first
phase of a product's life begins with the purchaser or "first user" of the product. After the
first use is Phase 2, in which the product may be given or sold to someone else for reuse,
be stored (e.g., in a closet or basement) for a period of time, or undergo some
combination of reuse and storage. Phase 2 may include the transfer of the product from
one person to another, either as a gift or a sale, but only if this transfer is from individual
-------
to individual as opposed to involving a third party, such as an electronics recycler,
broker, or donation organization. Phase 3 is the point at which the last user is ready to
remove the product from a private home or business. This change can result from the
desire to replace or otherwise stop using the product or the desire to remove the product
from storage. It is at this point that we state that the product is ready for EOL
management and it is transferred to a third party, such as a recycler or donation
organization, or it is disposed. Once the product is in the hands of a recycler, the product
may be sold for reuse "as is" or after some refurbishment. The resale may occur
domestically or by firms outside the United States. Electronic devices that are not
candidates for resale are dismantled or shredded, and the resulting material is separated
into secondary material streams and recovered. Recovered materials from the recycling
process are used to make new products, and the residuals of the processing stage are
disposed of in a landfill or incinerator. Material recovery may occur domestically or
abroad.
This report quantifies the number and weight of products that correspond to each phase of
the products lifecycle as illustrated in Figure 1.1. For Phase 1, we assembled product
sales data, as well as data on the average weight of products by year. We then developed
assumptions regarding how long Phases 1 and 2 would last. Since the life spans of
different types of products vary, unique life span assumptions were made for each type of
product. For example, televisions are typically kept longer than computers. Combining
the product sales and weight data, and applying the life span assumptions, we used a
spreadsheet model to predict the number and corresponding weight of material that would
become ready for EOL management each year. The model considered product sales from
1980 through 2004, and predicted the annual quantity needing EOL management through
2006.
-------
Product
Purchase
k-
Reuse
(give / sell to
family /others)
and/or
Putin
Storage
t- .>
Resale
outside
of the US
Process
outside
of the US
Phase 1
Phase 2
Phase 3
First use
Second use / storage Reach EOL
management
Phase 4
Recycling processing/
Resale for reuse
Figure 1.1 Framework for Modeling the Product Lifecycle
Having estimated the annual quantities of EOL products needing management, we
examined how much material has been recycled in recent years by the electronics
recycling industry. We then calculated the amount potentially being disposed of by
finding the difference between what is generated for management and what is collected
for recycling on an annual basis. More detail on data and the assumptions used is
provided in Section 2.0. The organization of the report is described below.
1.3 Organization of the Report
Section 2.0 provides a description of the data and assumptions used to develop estimates
of the number of products ready for EOL management annually. We quantified the
number of products sold historically by collecting data on product sales. (See Section 2.1
for more detail.) We then developed assumptions regarding the time for which the
product is used before it reaches EOL management. (Section 2.2 describes this
methodology.) The methodology used to estimate average product weights is described in
Section 2.3.
Section 3.1 presents the results of the modeling conducted and estimates when and what
volume of products are ready for EOL management on an annual basis (estimates for
1999 through 2006 are presented). The estimates regarding the portion that is collected
for recycling and disposal are described in Section 3.2. Estimates of the number and
weight of products that might be in storage at a given point in time are presented in
Section 3.3.
-------
In theory, all of the material that is in storage is ready for EOL management. In practice,
however, product users are ready for EOL management at different times. Some may
choose to hold onto products that have some perceived value to them. The distinction
between theoretical and practical EOL management is discussed in Section 4.0. Section
5.0 presents an analysis of the EOL management of CRTs to assess what portion
collected in the United States is managed domestically versus abroad. Finally, Section 6.0
summarizes the results and conclusions reached.
-------
2.0 Data and Assumptions in the Model
2.1 Historic Sales Data: Televisions, Cell Phones, and Personal Computer Products
The sales of televisions, cell phones, and personal computer products form the basis for
estimating the number and weight of products within the scope of this report requiring
EOL management at some point in the future. Historic sales data from industry sources
was the primary source (supplemented where necessary by government statistics from the
U.S. Census Bureau and the U.S. International Trade Commission). The following is a
discussion of data sources for each product type.
The market research firm, IDC, provided sales data on desktop and portable computers,
as well as hard copy peripherals. Hard copy peripherals include printers, multi-function
printers, faxes, and other devices. The availability of industry data was important,
especially for computer product sales. The sales estimates of personal computers based
solely on the Census and Trade Commission data would not have accounted for the sale
of "white box" products—generic computers with no brand names, manufactured by
vendors that purchase components. It is widely accepted that white box sales account for
a substantial portion of total U.S. consumption. In a 2004 press release, IDC stated that
the white box market share in the personal computer sector is about 28 to 30 percent in
the United States.3 In addition, Census and trade data were not available for faxes and
some other hard copy peripheral devices.
Sales of personal computer monitors, mouse devices, and keyboards were derived by
analyzing Federal government statistics. In this latter case, we developed sales estimates
by calculating what is referred to in this study as "apparent consumption," which
represents products sold in the United States for use in the United States. Apparent
consumption was estimated using the following formula:
Apparent consumption = U.S. shipments - domestic exports + imports for consumption
The U.S. Census Bureau's Current Industrial Reports (CIRs) show U.S. shipments, as
well as domestic exports and imports for consumption. However, the export and import
data are shown as combined categories, which would not allow us to develop totals by
product type. Therefore, to better account for the export and imports of single user
computers, we used the U.S. International Trade Commission (USITC) interactive
database.4 The government statistics used in this analysis are shown in Table 2.1.
3 Halperin, David, Mac News World. The Secret Market Contender: White-Box PCs. Technology Special
Report. May 1, 2004. www.TechNewsWorld.com.
4 The source cited by the Census Bureau for exports was the Harmonized Systembased Schedule B; for
imports the Harmonized Tariff Schedule (HTS) was cited. The USITC data are also based on the HTS.
-------
Table 2.1
Federal Government Statistics Used for Historical U.S. Sales Baseline
Document
Current
Industrial
Reports
Current
Industrial
Reports
USITC
Interactive
Tariff and Trade
Dataweb
Reference
U.S. Census
Bureau
MA36M and
MA334M
U.S. Census
Bureau
MAS 5R and
MA334R
U.S.
International
Trade
Commission
Data Type
(units)
U.S. shipments
of household
television
receivers
U.S. shipments
of single-user
computers and
office equipment
Domestic exports
and imports for
consumption (all
products)
* CIR reports withheld facsimile shipment data to avoid disclosing
data for individual companies.
Television sales data were supplied by the Consumer Electronics Association. Cell
phone sales data were based on a combination of Consumer Electronics Association data
on consumer sales and total cell phone sales as reported by Inform.5
TV and cell phone sales are shown in Table 2.2. Table 2.3 shows the U.S. sales data for
computer-related equipment by product type and year.
5 Inform, Inc., Waste in a Wireless World: The Challenge of Cell Phones, 2001. In this report, Inform
published total cell phone sales figures for 1995 through 2004. Sales prior to 1995 were interpolated based
on the annual growth rate in prior years as exhibited by the CEA data.
-------
Table 2.2
Historic Sales Data - Television and Cell Phone Products
(Million units)
Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Color CRT
<19"
5.4
5.6
5.7
7.0
8.0
8.4
9.1
9.7
10.1
10.9
10.4
9.4
9.7
10.6
11.7
10.8
10.1
9.6
10.3
11.2
12.2
9.8
11.7
8.3
6.9
Color CRT Flat Panel
>19" TVs
5.4
5.6
5.7
7.0
8.0
8.4
9.1
9.7
10.1
10.9
10.4
10.7
12.3
14.0
15.0
14.6
14.5
14.0
15.1
16.4 0.002
17.1 0.008
16.4 0.1
17.0 0.2
17.6 1.0
17.8 2.7
Color
Projection
0.2
0.3
0.3
0.3
0.3
0.3
0.4
0.4
0.4
0.5
0.6
0.8
0.9
0.9
1.1
1.3
1.7
2.0
2.5
2.7
3.5
Monochrome
6.7
5.7
5.7
5.7
5.1
3.7
4.0
3.5
2.6
1.7
1.4
0.8
0.6
0.6
0.5
0.5
0.4
0.4
0.3
0.3
0.3
0.3
0.2
0.2
0.2
Total TVs
17.6
16.8
17.1
19.7
21.3
20.8
22.5
23.2
23.1
23.6
22.6
21.3
23.0
25.6
27.9
26.7
25.9
24.9
26.8
29.3
31.3
28.4
31.6
29.7
31.2
Cell
Phones
0.04
0.11
0.40
0.80
1.3
2.1
2.6
3.4
5.4
7.9
12.4
14.5
16.6
22.2
30.6
49.3
72.9
100.1
122.3
140.0
151.9
Source: Data for TVs were obtained from Consumer Electronics Association Market Research, 2005 and 2006. Data for cell
phones was based on CEA data and sales data reported by Inform, Inc. (See text, footnote 5).
-------
Table 2.3
Historic Sales Data - Computer-Related Products
(Million units)
Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Note:
Source:
Hard Copy
Desktops Portables Peripherals
1.0
2.0
3.0
5.5
6.7
5.8
6.9
8.2
8.7
8.9
9.5
9.5
9.9
13.0
15.3
19.1
22 4
26.8
32.5
39.5
40.8
35.1
35.1
37.0
39.4
1.9
2.5
3.2
3.6
4.9
6.0
6.4
7.9
9.6
9.6
10.9
13.8
16.6
0.5
1.0
1.6
2.9
3.5
3.0
3.6
4.3
4.6
4.7
5.0
5.0
6.2
8.2
9.7
11.9
14.9
16.2
22.5
27.5
28.7
26.8
28.7
30.7
32.2
Hard copy peripherals (HCPs) include printers,
Data for desktops and portable PCs, as well as
Mice
1.0
2.0
3.0
5.5
6.7
5.8
6.9
8.2
8.7
8.9
9.5
14.3
20.9
31.3
39.7
19.1
22.4
24.9
27.9
39.5
56.2
53.0
57.5
37.0
39.4
Keyboards PC CRTs PC Flat Panel
1.0
2.0
3.0
5.5
6.7
5.8
6.9
8.2
8.7
17.5
21.7
27.0
37.6
36.1
41.4
47.6
53.8
55.6
65.0
63.7
51.7
43.8
48.6
51.3
47.2
1.0
2.0
3.0
5.5
6.7
5.8
6.9
8.2
8.7
8.4
9.4
10.5
13.4
17.3
18.1
22.2
23.1
26.6
32.6
36.9
37.5
27.2
23.3
15.8
15.9
1.1
0.9
1.5
1.7
1.8
2.8
3.0
2.3
1.2
1.8
11.2
12.8
14.0
23.5
34.3
44.2
, multifunction printers, digital copiers, and faxes.
hard copy peripherals (HCPs) were obtained from IDC WW Quarterly PC
Tracker in October 2005. Data for HCPs were estimated for 1980-1996 based on the ratio of HCPs to PCs. Data for flat panel
and CRT PC monitors, mice, and keyboards were based on ERG analysis of US Census data on shipments and Trade
Commission data on imports and exports. Data prior to 1990 for mice and 1988 for CRT monitors and keyboards were
estimated assuming one mouse, keyboard, or monitor per desktop PC.
10
-------
2.2 Assumptions Regarding the Life Span of Electronic Products
The life span—the time between the initial purchase of a product and the time it is ready
for EOL management—is one of the most critical assumptions in this type of
methodology. The total life span of any particular product will encompass several stages
of use. The "first use" is the time period that the product was considered functional to the
first purchaser. When the product ceases to be functional to the first user, the product
may be put in storage, discarded or recycled. If it is in working order, however, it will
most likely be reused by someone else. This is referred to as the "second use" stage.
There are many combinations of use, reuse, and storage underlying the second use stage
before the last user is ready for EOL management of the product.
Life Span of Televisions and Computer Products
In the past, researchers have modeled the flow of products from purchase to EOL
management by assuming a time period for each use stage (i.e., number of years for first
use, second use, etc.). However, if the ultimate goal is to model when products are ready
for EOL management, the pattern of use prior to EOL management is somewhat
immaterial. For example, one product could be used initially for 3 years, reused for 2, and
then put in a closet for 5, while a second product might be used initially for 5 years and
reused for another 5 years. However, both products enter the EOL management stream
after 10 years. Because our objective was to model when products will enter the EOL
management stream, we examined the age distribution of the products being collected by
electronics collection programs and used that as a proxy for both the first and second use
stage that occurs prior to the EOL management stage. By using this approach, we
assumed that the time and effort for any type of EOL management, either recycling or
disposal, are roughly the same—that is, a comparable action is required to remove an
unwanted product from a home or office whether it is recycled or disposed. We
acknowledge that this may not be true in all cases, but in the absence of better data, we
considered this a reasonable assumption.
The State of Florida has been providing grants to its counties for electronic collection
programs since 2002. In 2004, the Florida DEP conducted a study in which the
individual units in the loads from electronics collection programs were sorted and the
product type, brand, weight, and age were recorded. These loads represent collections
from residential and small business sources that are generally served by county recycling
or thrift store donation services. The Florida counties did not charge a fee for recycling
services, so fees were not an influence in people's decisions to participate in the
collection program.
At the time of this analysis, the data set from this project had information on 20,619 units
collected in a 12-month period beginning April 2004. It contained the date of
manufacture of each product if those data were easily identified. We analyzed these data
to investigate the age distribution of each type of product collected. Out of 20,619
products collected, the vintage of 12,801 (62 percent) units was recorded. We calculated
the age of the products by subtracting the date of collection from the date of manufacture.
11
-------
Although we cannot represent with certainty that the Florida results are representative of
the nation as a whole, it is the largest available data set that accounts for the age of the
electronic products collected. The State of Minnesota also conducted a study of age,
brand, and weight of about 1,000 electronic products collected at one 3-day collection
event in September 2004. Statistical tests that were conducted (the Kolmogrov-Smirnov
two-sample test) showed that the age distribution of the laptops and desktop computers
collected in these two locations were not significantly different. This result does not
allow us to conclude that the Florida data are representative of the nation, but at least we
know that two states in different parts of the country did not exhibit significant
differences in this regard.
Table 2.4 presents the general statistics for each product type from the Florida
brand/vintage sort project at the time of the data analysis. The age distribution of selected
product types is depicted graphically in Appendix A. These data give us insight into how
long consumers and small businesses hold on to products before they bring them to
available collection sites for EOL management. As already stated, this is a multiple-phase
life span, which could include first use, second use, and storage. The Florida data show
that the large majority of computer products that enter EOL management are over 5 years
old and the average first use of computers is often cited as about 4 years.6 Noting that the
oldest TVs, desktop computers, and CRT monitors are around 30 years old, it is obvious
that reuse/storage periods range widely from what might be considered short to very long
term.
To develop the assumptions on the life span of products to use in the modeling effort, we
examined the distributions and decided on an approach that would best represent the data
for each individual product type. For most products, the life span assumptions are based
on the medians of each of the quartiles. (For example, the "youngest" 25 percent of
desktop computers are used for the length of time represented by the median of the 1st
quartile; the next-youngest 25 percent are assumed to be used for the length of time
represented by the median of the 2nd quartile; etc.) Thus, for most electronic products,
there are four life span assumptions, one for each 25 percent of products sold in any
given year.
For hard copy peripherals, we conducted statistical tests to determine whether or not the
individual age distributions for printers, fax machines, multifunction devices, and
scanners were similar. Using the Kruskal-Wallis Test, we found that there was no
statistically significant difference in the age distribution among the four product types.
Therefore, we combined the data for all hard copy peripherals and used the median of the
four quartiles to represent the life spans of all these products.
For some products, such as laptop computers, keyboards, and LCD monitors, we took a
different approach to best represent the data because there were relatively few
6 Matthew, H. Scott, Deanna H. Matthews. 2003. "Information Technology Products and the
Environment." In: Kuehr, R. and Williams, E., Eds. 2003. Computers and the Environment. Dordrecht, The
Netherlands: Kluwer Academic Publishers, pp. 41-72.
12
-------
observations in these data sets. In these cases, we used the mean or mode, or chose
specific quartiles to represent the distribution. Table 2.5 lists the life span assumptions
and basis for each product type.
We recognize that the Florida data represent product usage patterns today, and that 5 or
10 years ago, people may have used products for longer or shorter periods of time. Usage
patterns could have varied in the past based on changes in purchasing behaviors or
changes in technology that may have influenced purchasing behavior. Due to data
limitations, we assumed that usage patterns were not significantly different in the past.
Differences in Life Span Between the Residential and Commercial Sectors
The Florida electronics collections data represents computer and TV products mainly
from the residential sector. We assumed that the life spans exhibited for TVs in this data
set were also appropriate for the commercial and institutional sectors. In other words, for
each type of television technology (e.g., flat panel, projection, etc.), we assumed there is
not a significant difference in patterns of use or upgrade cycles between residential and
commercial TV users.7
For computers, however, we do believe there is a difference in usage patterns between
sectors. We, therefore, developed separate lifespan assumptions for products used in the
commercial/institutional sector. Due to data limitations on usage patterns in the non-
residential sector, we relied on industry expert opinion. We assumed that 40 percent of
businesses remove and manage their computers after 3 years (the midpoint between a 2 to
4 year replacement cycle), while 60 percent are on a 5-year cycle (the midpoint between a
3 to 7 year replacement cycle).8'9 The 40 percent assumption is based on IAER reporting
that about a third of the output of recycling operations is equipment for reuse.10 In other
words, 33 percent is suitable for resale "as is" or with some refurbishment. One would
expect computer products to meet the criteria for resale "as is" only if they are relatively
new, such as under 3 years old. A small portion of equipment collected by recyclers is
from the residential sector, which tends to be older and unsuitable for reuse. To account
for this, using professional judgment, we adjusted the 33 percent to 40 percent to account
for the fact that the percentage suitable for reuse for just commercial equipment would
likely be higher than the percentage for all sectors.
The share of computer product sales that are residential versus commercial is based on
an analysis of IDC PC sales data for both sectors. An average of 48 percent residential
and 52 percent commercial based on data from 1992 to 2004 was assumed. This
assumption most likely overestimates the residential share of computer products in the
Based on a phone conversation with Shawn DuBravac of the Consumer Electronics Association,
September 11, 2006.
8 Lynch, Jim. 2004. "Islands in the Wastestream: Baseline Study of Noncommercial Computer Reuse in the
United States," CompuMentor, Fall 2004.
9 Based on a phone conversation with John Powers of the International Association of Electronics
Recyclers, June 2, 2006.
10 International Association of Electronics Recyclers, IAER Electronics Recycling Industry Report, 2006.
13
-------
early years of PC use, however, due to lack of additional data, we assumed a constant
rate.
14
-------
Table 2.4. Florida Electronics Sort Project: Age Distribution By Product Type
Product Type
Desktop
Laptop
Keyboard
Monitor (CRT)
Monitor (LCD)
Multifunction device
Printer
Fax
Scanner
TV<19'
TV>19'
TV - console
TV - projection
Total
Observations
With Data
(Number)
1
4
1
1
3
,912
20
41
,515
3
30
,032
52
23
,914
,196
56
7
12,801
Observations
Missing Data
(Number)
2,222
44
1,844
897
1
46
1,286
168
204
355
691
55
5
7,818
Percentile (%)
25th
8
5
4
7
6
7
6
7
5
11
10
9
7
50th
12
6
5
9
10
9
9
9
7
15
13
11.5
8
75th
16
8
8
12
12
11
11
12.5
12
20
18
17
9
100th
27
14
20
30
12
17
29
18
14
34
33
27
10
Mean
12.2
6.5
6.7
9.3
9.3
9.5
8.8
9.8
8.3
15.1
13.5
12.7
8.0
Mode
8
8
4
8
N/A
11
4
9
7
12
12
8
8
Min.
1
4
3
1
6
5
0
4
4
1
0
2
6
Max.
27
14
20
30
12
17
29
18
14
34
33
27
10
Source: Florida Department of Environmental Protection. Database accessed 8/22/05.
www.dep.state.fl.us/waste/categories/electronics/pages/FloridaElectronicProductBrandDistributionProject.htm
15
-------
Table 2.5
Life Span Assumptions By Product Type-Residential Sector
(Number of Years Before Collection)
Product Type
PCs — desktop
PCs — portable
PC Monitors— CRT
PC Monitors — flat panel
PC Hard copy peripherals
PC Keyboards
TVs— CRT < 19"
TVs— CRT > 19"
TVs — Projection
TVs— Flat Panel
Assumption
25% — 7 years
25%— 10 years
25%— 14 years
25%— 18 years
20%— 4 years
15% — 5 years
20%— 6 years
45% — 7 years
25% — 5 years
25%— 8 years
25%— 10 years
25%— 13 years
100%— 9 years
25% — 4 years
25% — 7 years
25%— 10 years
25%— 14 years
100%— 5 years
25%— 8 years
25%— 13 years
25%— 17 years
25%— 23 years
25% — 7 years
25%— 12 years
25%— 15 years
25%— 20 years
100%— 8 years
100%— 9 years
Basis
Median of each quartile
1st through 4th quartiles
Median of each quartile
Mean of all observations
Median of each quartile
Median of all observations
Median of each quartile
Median of each quartile
Mean of all observations
No data- Assumed the same
as PC flat panel monitors
Note: Assumptions were based on statistical analyses of data from the Florida DEP.
Life Span of Cell Phones
Unfortunately, we do not have access to any data on the age distribution of cell
phones when they are collected for recycling. Here again, we relied on industry expert
opinion. We assumed that 65 percent of cell phones were 2 years old based on industry
reports that about that percentage of cell phones collected were suitable for resale "as is"
16
-------
or after refurbishment.11 We also assumed that the remaining 35 percent of phones
collected were 5 years old and were only suitable for materials recovery.12 Due to lack of
data, we did not assume any difference in cell phone use patterns between the residential
and commercial sectors.
2.3 Average Weight Data: Televisions and Computer Equipment
The average weight of products is an important input to this methodology for modeling
the use stages of electronic products. We developed product weight estimates for each
product type, for each year covered in the analysis. Within some product types, such as
TVs, weights vary depending on the size and type of screen. Product weights also can
vary over time as technology, style, and features change.
To develop estimates of average product weights, we reviewed two data sets. The first
data set was developed from electronics collection data obtained from the Florida DEP
(described in Section 2.2 above). At the time of this analysis, the Florida data set had
weight and year of manufacture data for 12,801 units. The average weights were
calculated for each product and each year of manufacture. Again, we cannot represent
with certainty that the Florida results are representative of the nation as a whole, but it is
a large, available data set with product weight and age information.
In addition, we used the data set on product weights developed over past years for the
EPA municipal solid waste (MSW) characterization report series.13 These data were
gathered from Consumer Reports Annual and Monthly Buying Guides from 1984
through 1999. For the most recent years studied (2000 to 2004), data on product weights
were collected from product specification listed by large consumer electronic retailers.
Table 2.6 identifies which data set was used for each product type in this analysis.
1: Based on a phone conversation between Jennifer Chambers, Recellular, Inc. and Lynn Knight, ERG July,
2006.
12 See footnote 10.
13 U.S. EPA. Municipal Solid Waste in The United States: 2003 Facts and Figures and previous years'
editions of the same report, www.epa.gov/epaoswer/non-hw/muncpl/msw99.htm.
17
-------
Data
Table 2.6
Source, by Product Type
Product
Source of Product Weight Data
TVs
<19 inch
>19 inch
Projection
Flat screen
Florida DEP collection study
Florida DEP collection study
Consumer product publications
and retailer product specifications
Consumer product publications
and retailer product specifications
Computers
Desktop
Laptop
Florida DEP collection study
Consumer product publications
and retailer product specifications
Monitors
CRT
Flat panel
Keyboards
Mouse devices
Printers and other
hard copy devices
Cell Phones
Consumer product publications
and retailer product specifications
Florida DEP collection study and
ERG in-house data
Florida DEP collection study
Consumer product publications
and retailer product specifications
Florida DEP collection study and
consumer product publications and
retailer product specifications
Retailer product specifications
The average weights for televisions and cell phones, and computer products are shown in
Tables 2.7 and 2.8, respectively.
18
-------
Table 2.7
Average Weight of Television and Cell Phone Units (pounds)
Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
color UK I
(<19
inches)
42.0
42.0
42.0
42.0
42.0
40.6
41.1
40.8
41.2
41.0
40.5
41.1
40.9
40.7
41.1
40.9
41.3
40.7
41.6
41.2
39.8
41.1
40.4
41.0
41.0
uoioruKi
(>19
inches)
73.0
73.0
73.0
73.0
73.0
72.6
73.0
73.0
72.9
71.7
74.8
73.9
73.5
75.4
73.3
73.5
72.8
73.8
74.1
73.0
74.5
72.2
72.8
73.0
73.0
Flat Panel Projection
219.0
221.0
223.0
225.0
227.0
229.0
231.0
233.0
235.0
237.0
239.0
241.0
243.0
245.0
247.0
29.0 249.0
29.0 251.0
29.0 251.0
29.0 223.3
29.0 195.7
29.0 168.0
Monochrome
42.0
42.0
42.0
42.0
42.0
40.6
41.1
40.8
41.2
41.0
40.5
41.1
40.9
40.7
41.1
40.9
41.3
40.7
41.6
41.2
39.8
41.1
40.4
41.0
41.0
Cell
Phones
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.4
0.4
0.3
0.3
0.3
19
-------
Table 2.8
Average Weight of Personal Computer-Related Units (pounds)
Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Desktop
Computers
22.0
22.0
22.0
22.0
22.0
22.0
22.0
22.0
22.0
21.9
21.8
21.8
22.2
21.9
21.7
23.0
22.1
22.6
22.7
22.0
22.1
22.0
24.1
22.0
22.0
Portable CRT
Computers Monitors
9.0
8.7
8.5
8.2
7.9
7.7
7.4
7.1
7.1
7.0
6.8
6.6
6.4
24.50
24.50
24.50
24.50
24.50
24.50
24.50
24.50
24.50
24.50
24.63
24.75
24.88
25.00
28.86
32.71
36.57
40.43
44.29
48.14
52.00
51.62
51.25
50.87
50.50
Flat Panel
Monitors
24.6
24.6
24.6
24.6
24.6
24.6
24.6
24.6
24.6
24.6
24.6
24.6
24.6
24.6
24.6
24.6
Keyboards
2.9
2.9
2.9
2.9
2.9
2.9
2.9
2.9
2.9
2.9
2.9
2.9
2.9
2.9
2.9
2.9
2.9
2.9
2.9
2.9
2.9
2.9
2.9
2.9
2.9
Mouse Hard Copy
Devices Peripherals
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
18.00
18.00
18.00
18.00
18.00
18.00
18.00
18.00
18.00
17.86
19.62
18.36
17.43
17.76
17.81
16.83
15.37
16.74
16.27
16.40
18.46
16.93
16.42
16.59
17.40
Note: Average weight for hard copy peripherals is based on a weighted average of printers and
scanners each year.
20
-------
3.0 Model Results
3.1 The Quantity ofEOL Electronics Generated for Management Each Year
This section presents estimates of the quantity of EOL electronic products generated for
management each year. As described earlier, we developed the estimates by starting with
product sales data and assuming specific life spans for each product type to represent the
time between product purchase and the need for EOL management. These estimates
would correspond to Phase 3 in Figure 1.1. As explained earlier, ready for EOL
management means that the product has gone through a first use and possibly a second
use stage (which could include reuse and storage) and the last user is ready to give the
product to a recycler or dispose of it. The quantities of products generated for EOL
management each year are presented for personal computer products, televisions, and cell
phones in Tables 3.1 and 3.2.
According to the estimates presented in Table 3.1, 28.4 million desktop computers, 22.9
million hard copy peripherals, and 9.0 million portable computers were ready for EOL
management in 2005. Almost half of the weight (47 percent) of computer-related
equipment generated in 2005 was from CRT computer monitors. About 23 percent of the
weight was attributed to desktops and 14 percent was from hard copy devices. Portable
computers accounted for 2 percent of the weight of EOL products generated that year.
As shown in Table 3.2, in 2005, 24 million TVs were ready for EOL management. TVs
with CRTs accounted for the majority of the TV units as many of the newer projection
and flat panel units had not yet reached EOL. Over time, the proportion of computer-
related products reaching the EOL stage has increased dramatically relative to TVs. For
example, in 1999, the weight of EOL TVs was very close to the weight of all EOL
computer equipment generated that year. In 2005, however, TVs account for only a little
over half (55 percent) of the weight of computer equipment. It is estimated that 116
million cell phones will be ready for EOL management in 2005 as well.
21
-------
Table 3.1
Estimated Annual Personal Computer Products Ready for EOL Management, By Year
Year Desktops
Units(mill)
1999 12.6
2000 15.4
2001 18.4
2002 21.9
2003 24.7
2004 26.6
2005 28.4
2006 28.3
Tons(OOO)
138.3
174.3
204.4
244.8
275.0
293.6
322.6
311.6
Portables
Units(mill) Tons(OOO)
3.2 13.5
3.9 16.0
4.8 19.0
5.8 22.0
6.9 25.4
7.8 28.2
9.0 31.8
10.2 35.2
Hard Copy Devices
Units(mill)
9.2
10.9
13.6
16.2
19.6
21.3
22.9
24.0
Tons(OOO)
77.6
91.8
110.7
134.7
166.7
181.7
198.3
199.1
Mice/Keyboards
Units(mill)
81.2
66.7
76.2
80.5
92.8
103.2
107.9
96.8
Tons(OOO)
64.1
70.9
80.2
83.1
97.0
96.3
80.6
68.8
CRT Monitors
Units(mill)
15.7
18.9
21.1
23.9
27.7
27.8
28.5
23.8
Tons(OOO)
238.3
314.8
386.6
480.7
597.8
627.8
673.1
550.3
Flat Panel
Units(mill)
1.8
1.9
1.9
3.6
4.6
7.8
10.0
12.1
Monitors
Tons(OOO)
21.7
23.4
23.6
44.2
56.4
96.2
122.6
148.5
Total
Units(mill)
123.7
117.8
136.0
151.9
176.3
194.6
206.6
195.2
Tons(OOO)
553.5
691.1
824.5
1,009.5
1,218.2
1,323.8
1,429.1
1,313.4
Source: ERG estimates based on modeling results.
Table 3.2
Estimated Annual Television and Cell Phone Products Ready for EOL Management
Year
Color CRT <1 9"
Units(mill) Tons(000)
1999
2000
2001
2002
2003
2004
2005
2006
6.1
6.6
7.2
7.7
9.0
8.7
8.8
9.7
125.0
135.8
148.3
158.4
183.6
179.1
180.3
200.0
Color CRT >1 9"
Units(mill) Tons(000)
7.5
9.5
10.1
10.1
10.6
11.3
12.0
12.8
274.2
350.3
369.2
371.4
386.6
412.8
445.0
470.0
Flat Panel
Units(mill) Tons(000)
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Projection
Units(mill) Tons(000)
0.4
0.4
0.5
0.6
0.8
0.9
0.9
1.1
44.3
47.5
55.1
76.0
98.8
107.8
112.3
133.6
Monochrome
Units(mill) Tons(000)
2.6
2.5
2.0
1.5
3.1
2.6
2.3
2.1
54.2
51.4
42.5
30.0
65.0
54.0
48.4
43.2
Total TVs
Units(mill)
16.5
19.0
19.8
19.9
23.5
23.5
24.0
25.7
Tons(000)
497.7
585.0
615.1
635.8
734.1
753.6
786.0
846.8
Cell Phones
Units(mill)
18.8
25.0
37.9
55.2
75.8
96.8
116.5
133.8
Tons(000)
3.0
3.4
3.4
3.5
7.5
7.5
8.2
10.4
Source: ERG estimates based on modeling results.
22
-------
The share of computer products sold to the commercial versus the residential sector is
based on recent sales data and is assumed constant for all the study years. This
assumption could have the effect of overestimating the residential computer share in
earlier years when household penetration of personal computers was lower than it is
today. Therefore, the life spans could be too long for the portion of residential products
in the 1980's that may have actually been in the commercial sector.
To test the sensitivity of this assumption, we ran a scenario in which we assumed that 25
percent of desktop computer sales were attributable to the residential sector from 1980
thru 1985, 40 percent from 1986 thru 1990, and 48 percent (the constant rate assumed in
the original analysis) from 1991 thru 2004. Testing this scenario, the resulting estimates
of units ready for EOL management in 2005 were less than 1 percent lower than the
original estimates. In 2000, the sensitivity test resulted in about a 3 percent lower
estimate. Therefore, we do not believe our lack of data regarding the split between
computer product sales in the residential versus the commercial sector in previous years
has a significant effect on the estimates of EOL products ready for management in recent
years.
3.2 Estimating the Quantity of EOL Products Generated That Are Recycled Versus
Disposed
The modeling effort resulted in estimates of the quantity of products that are generated
annually for EOL management. EOL management consists of recycling or disposal.
This corresponds to the two options in Phase 3 of Figure 1.1: "Dispose" or "Bring to
Recycling Collection." We estimated the amount of EOL electronics recycled by
gathering data from the recycling industry. Disposal was estimated as the difference
between what was generated for EOL management and what was recycled. The
following sections discuss the details of this part of the analysis.
Estimating the Portion of EOL Electronics Recycled
Recycling of consumer electronics includes the recovery of products by municipal and
other collection programs for materials separation and recovery, as well as reuse in both
domestic and foreign end markets. It also includes businesses and institutions contracting
directly with electronic recyclers for recycling services of their EOL equipment.
Donation organizations also collect EOL electronic equipment for reuse or recycling. In
this report, we do not distinguish between a for-profit electronics recycler and a donation
organization that collects EOL equipment. The term "reuse" in the EOL management
stage refers to products entering the recycling materials management system that are in
working order and can be resold "as is" or refurbished for resale by electronics recyclers
and dismantlers. The reuse of consumer electronics before they enter the management
system (i.e., products that pass between individual users) is assumed to occur prior to
EOL management.
23
-------
The recycling estimates for 1999 through 2004 were developed previously for EPA.14
They were based on extrapolations of recovery and market share data from a few
electronics recyclers. The amount recycled in 2005 was projected based on the same
recovery rate exhibited in 2004. Estimated quantities of EOL consumer electronics
recycled from 1999 through 2005 are shown in Table 3.3 below.
As a check on the 2005 recycling estimate, we turned to industry data. In its 2006
Industry Report, the IAER estimated that the electronics recycling industry processed an
annual total volume of 1.4 million tons of electronic equipment in 2005.15 This estimate
was based on a survey of over 500 electronics recyclers, OEMs and non-profit
organizations. However, it included a much broader scope of products, as well as several
source sectors that were not included in this analysis. For example, the IAER estimate
would include servers, main frames, copiers, DVDs, VCRs, etc. It also includes
equipment collected from industrial, medical, and other sectors that are not addressed in
this study. To compare the IAER estimates with our estimate, we did the following
analysis. The IAER reported that 74 percent of the equipment processed was computer
and consumer equipment. Further, they reported that 39 percent of equipment was from
the residential, commercial, manufacturing, industrial, or institutional sectors. If we
assume that the proportion of equipment types collected is equal in each of the source
sectors, we could estimate that roughly 404,000 tons of EOL consumer and computer
equipment from the residential and commercial/institutional sectors was processed by
recyclers in 2005. Our adjusted IAER estimate of 404,000 is higher than our estimate of
345,000 tons for 2005. However, given that the IAER data reflects certain types of
consumer and computer equipment that we were not addressed in this report (e.g., VCRs,
DVDs, servers, main frames, copiers), and further, we cannot relate the product types
within each source/sector category, our estimate does not appear to be significantly
different. In any case, we conducted a sensitivity analysis to test the effect of a possible
underestimate of the amount recycled in Section 6 below.
Estimating the Portion of EOL Electronics Disposed
To estimate the portion of the estimated EOL electronics generated every year that is
disposed, we subtracted the amount estimated to be recycled from the estimated amount
generated for EOL management. Table 3.3 includes the disposal estimates for 1999
through 2005.
14 U.S. EPA. Municipal Solid Waste In The United States: 2003 Facts and Figures and previous years'
editions and updates of the same report.
15 International Association of Electronics Recyclers, IAER Electronics Recycling Industry Report, 2006.
24
-------
Table 3.3
Distribution of EOL Products By Management Method
Total EOL
Year Units(mill)
1999
2000
2001
2002
2003
2004
2005
159.0
161.7
193.7
227.0
275.6
314.8
347.1
Products
Total Recycled
Total Disposed
Tons(OOO) Units(mill)*Tons(000) Tons (%) Units(mill)*
1,054.2
1,279.5
1,443.1
1,648.8
1,959.8
2,085.0
2,223.3
23.7
24.0
28.2
34.4
40.8
48.3
53.9
157.0
190.0
210.0
250.0
290.0
320.0
345.0
15%
15%
15%
15%
15%
15%
16%
135.4
137.7
165.5
192.6
234.8
266.5
293.2
Tons(OOO)
897.2
1,089.5
1,233.1
1,398.8
1,669.8
1,765.0
1,878.3
Tons (%)
85%
85%
85%
85%
85%
85%
84%
*Number of units estimated based on the units/ton factor exhibited by EOL products.
According to this analysis, about 15 percent, by weight, of the EOL electronics generated
in the study years were collected for recycling. During the time period examined, even
though the amount of material being recycled has increased, the amount of EOL products
generated has kept pace such that the percentage of material being recycled has remained
relatively constant.
The majority of EOL material that is not being recycled is probably mostly going into
landfills. According to EPA data, only about 14 percent of all MSW goes to the waste-to-
energy process.16 Within that 14 percent, furthermore, it is possible that computer
monitors or televisions with CRTs are not being combusted, but rather are being removed
on the tipping floor and sent to landfills (unless there is a ban on CRT disposal in
landfills). Waste-to-energy operators would be inclined to remove these items because
the glass is not combustible and because of concerns about the resulting lead in the ash
from the CRT glass. Non-CRT computer equipment may not be removed because it
contains plastic, which is combustible. Without further research, we cannot predict how
EOL electronic products disposed of in communities with waste-to-energy are actually
managed, but based on national averages, we can say that no more than 14 percent that is
disposed of is combusted, which we believe is a very conservative scenario, and at least
86 percent goes to a landfill.
In order to check the accuracy of the disposal estimate, we explored a second way to
estimate the quantity of EOL electronics disposed. We reviewed and analyzed seven
waste sort/sampling studies, all of which delineated consumer electronics as a separate
category. This methodology resulted in an average number of pounds of consumer
electronics disposed per person. By applying the pounds per person result to the U.S.
population we estimated that 1.4 million tons of waste electronics was disposed of in
2003. This is within a reasonable range of the estimate of 1.7 million tons we derived
from the model results. The details of this methodology (based on the waste sort
analysis) are described in Appendix B.
16 U.S. EPA. Municipal Solid Waste in the United States: 2003 Facts and Figures. 2005.
25
-------
Limitations of the Analysis
Our estimates of the amount of EOL material generated annually and the proportion that
is subsequently managed through recycling versus disposal rely heavily on many
assumptions and analyses as discussed in Section 2. Some assumptions regarding the
estimates of the amount of material generated annually for EOL management have the
following limitations:
• The age distribution of the electronic products collected for recycling in Florida is
assumed to be representative of the age distribution of products collected for EOL
management nationally. This assumption may misrepresent product usage
patterns, which could have the effect of over- or under-estimating the volume of
products ready for EOL management in any given year.
• The age distribution of electronic products exhibited in Florida collections in
recent years represents a product usage pattern that is held constant in all years of
the analysis. This assumption could have overestimated product life span in
specific years past when, for example, significant changes in product
technologies, such as computer processor improvements, temporarily spurred a
faster product replacement rate. This limitation would have the most effect on
computer products.
• The share of computer products sold to the commercial versus the residential
sector is based on recent sales data and is assumed constant for all the study years.
This limitation could have overestimated the residential computer share in earlier
years, which might have increased the product life spans assumed for a portion of
the products.
As already discussed in Section 3.1, we conducted a sensitivity analysis to explore the
potential effect of the last limitation and concluded that the lack of data regarding the
split between computer product sales in the residential versus the commercial sector in
previous years did not have a significant effect on the estimates of EOL products ready
for management in recent years.
3.3 Estimating the Quantity in Storage
We have presented estimates of the number of products ready for EOL management on
an annual basis. During the time period prior to EOL management, however, they may
have been used, reused and/or stored for a period of time. Therefore, in 2005 for
example, our model predicts that 37 million personal computers will be ready for EOL
management; however, there are units that are not being actively used anymore and are
being stored because their owners are not ready to bring them for EOL management. To
try to gain an understanding of the possible number of units being stored in this manner,
we calculated the total number of units sold from 1980 through 2004, and then removed
those units that we estimate have already been brought for EOL management. Further, we
26
-------
removed those products that are likely to still be in their first or second use stage.17 The
remaining amount is what might be in storage.
Table 3.6 presents estimates of the number of units in each use stage in 2005. We
estimate that, of the products considered in this study that were sold between 1980 and
2004, approximately 180 million units could be in storage as of 2005. Televisions
account for the largest portion of units being stored, while desktop PCs account for the
next largest portion. Overall, 9 percent of the products sold in this time period are in
storage, while almost half are still in use and 42 percent have already been collected for
EOL management (i.e., have already been recycled or disposed). At the time of this
study, we did not have enough information to make similar estimates for cell phones.
Table 3.6
Estimated Number of Units in Various Stages of Use as of 2005 (million units)
Collected for EOL Management
Product Type
Desktop PCs
PC monitors
Portable PCs
Hard copy peripherals
Televisions
Total
Number
188.1
250.4
36.4
137.1
229.7
841.8
% of Total
42.6%
46.1%
37.6%
45.1%
37.5%
42.1%
In Storage
Number
43.3
28.0
1.4
13.8
93.9
180.3
% of Total
9.8%
5.1%
1.4%
4.6%
15.3%
9.0%
Still in Use
Number
210.0
265.3
59.0
152.9
288.3
975.7
% of Total
47.6%
48.8%
60.9%
50.3%
47.1%
48.8%
Total Sold
1980-2004
Number
441
543.7
96.9
303.8
611.9
1997.8
% of Total
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
Source: ERG estimates based on model results.
For the residential sector, we examined the use stages in a little more detail by trying to
separate the first use from the reuse stage for computer equipment. Estimates of the
number of units in the first use and reuse stage in 2005 were based on the assumption that
computer equipment is used for 4 years initially and reused for an additional 3 years.18'19
Figures 3.1 through 3.4 illustrate the results for 2005 based on these assumptions. Of the
computer-related products sold between 1980 and 2004, 37 percent were in their first-use
stage and 22 percent were in the reuse stage in 2005.
When considering the amount of material possibly being stored, one has to keep in mind
that different product owners will choose to keep items for different periods of time.
Therefore, even if there were a convenient, free outlet for EOL management of this
material, many people would not necessarily change their pattern and the products
lifecycle. The next section examines the question of how residential product owners are
likely to make use of electronics recycling opportunities.
17 Assumptions regarding the number of years products are probably still in their first or second use was
based on Consumer Electronics Association survey data on perceived life expectancy of electronic
products: 11 years for a color TV; 6 years for a notebook PC; seven years for a desktop PC and PC
monitor. Assumed the same 7 years as desktop PC for hard copy devices. Presentation at 2005 E-Scrap
North American Electronics Recycling Conference: "From Here to There: Facts on Product Life Cycles
and Recycling. " By Shawn G. DuBravac, Consumer Electronics Association.
18 Matthews, H. Scott and Deanna H. Matthews. 2003. "Information Technology Products and the
Environment." In: Kuehr, R. and Williams, E., Eds. 2003. Computers and the Environment. Dordrecht,
The Netherlands: Kluwer Academic Publishers, pp. 17-39.
19 International Association of Electronics Recyclers, IAER Electronics Recycling Industry Report, 2006.
27
-------
Figure 3.1
Figure 3.2
Stages of Use in 2005 Households:
Desktop PCs
Reuse Mana9ed
26% ,-p. 21%
W Storage
20%
1st Use
33%
Stages of Use in 2005 Households: Hard
Copy Peripherals
Reuse
1 go/0 Managed
X"P\ 32%
V^XT
1st Use — Storage
40% 9%
Figure 3.3
Stages of Use in 2005 Households: PC
Monitors
Reuse
21%
1st Use
36%
Managed
32%
Storage
11%
Figure 3.4
Stages of Use in 2005 Households: Portable
PCs
1st Use
53%
Managed
31%
Storage
3%
28
-------
4.0 Quantity Available for Recycling Based on Observed Collection Rates
Recovery rates experienced by existing state and local electronics collection programs
allow us to begin to understand how residents respond to available recycling
opportunities. This is important when municipalities or other entities are planning to
establish electronics recycling opportunities to a given population. We examined
recovery rate information on a per capita basis from a handful of programs to get a sense
of the quantities of material that might be anticipated.
Each of the recycling programs selected for this analysis met the following criteria:
• The program served a discrete and quantifiable population. Examples include
state programs available to all state residents, collections run out of county or
municipal waste management facilities, and curbside collections within a specific
city or county.
• The program involved regular daily, weekly, or monthly collection—not a one-
time or annual collection event.
We obtained data from:
• Hennepin County, Minnesota
• Waukesha County, Wisconsin
• Alberta, Canada
• State of Massachusetts
• Citrus County, Florida
• Broward County, Florida
• Charlotte County, Florida
• Lee County, Florida
• Miami-Bade County, Florida
• Polk County, Florida
• Sarasota County, Florida
We found that overall results can vary widely from one program to the next. The variance
in recovery rates experienced by these programs can reflect a number of factors, such as
the program type (drop-off, special event, curbside pickup, etc.), accessibility and
frequency of collection, whether or not there is a recycling fee, extent and effectiveness
of public outreach and education, availability of alternative disposal options, and
characteristics of the target population (e.g., attitudes toward recycling). The programs in
this analysis represent somewhat of a cross-section of electronics recycling programs not
only in terms of scale (state, county, and municipal programs), but also in terms of
program type and maturity.
Also, the results from Massachusetts should be viewed with the perspective that there is a
statewide ban on the disposal of CRTs.
29
-------
The per capita recovery rates exhibited by the specific programs we examined varied
widely. Table 4.1 summarizes the range of results for 2004, broken down by product
type. Without further analysis, it is difficult to say what one might expect as far as a per
capita recovery rate from residential programs on a national basis. Combining the product
types shown in the table, a range of about 1 to 3 pounds per capita has been experienced.
The midpoint of that range is 2 pounds per capita. Using this recovery rate as a basis and
extrapolating to the entire U.S. population, we could estimate that if all residents had
access to electronics recycling programs, 294,000 tons of material could have been
collected in 2004 from the residential population.
The recovery rates in both Hennepin County, as well as Massachusetts were on the high
end of the range. The recovery rate of 3 pounds per capita experienced in Hennepin
County is a good indicator of the quantity of material available from a residential
population when there is a well-established, well-publicized program. In Massachusetts,
there is a state-wide ban on CRT disposal and the recovery rates for just CRTs was about
3 pounds per person as compared to 2.6 pounds per person for just CRTs in Hennepin
County in 2004. From this limited analysis, it may be that jurisdictions with disposal
bans on some or all EOL electronic products can experience a somewhat higher recovery
rate.
Table 4.1
Range of Per Capita Per Year Recovery Rates Among Selected Programs, 2004
Product Type
Monitors
CPUs
TVs
Printers
Laptops20
Total
Recovery Range (Pounds per Capita per
Year)
Minimum
0.23
0.13
0.4
0.07
0.002
0.83
Maximum
0.60
0.35
1.99
0.17
0.007
3.12
' Only Waukesha County, Alberta, and Hennepin County reported laptops as a distinct category in 2004.
30
-------
5.0 Examining the End Markets of Products Collected Through Electronics
Collection Programs in the United States
The purpose of this section is to examine the end markets, both domestic and abroad, of
EOL electronic products that are collected for recycling in the U.S. by electronics
recyclers or donation organizations. This is a difficult task since data on the end markets
of these materials is not publicly available. In order to arrive at a rough estimate, we
relied on an industry expert to develop a best estimate for CRT-containing devices.21
These estimates were derived based on industry data and specific knowledge of the
domestic and export end markets for EOL electronics. Only CRT-containing devices—
televisions and computer monitors—were considered because there are reasonably
feasible means of tracing these material flows. This section presents estimates of the
quantity of collected CRT-containing devices that are sent to domestic and foreign end
markets.
Table 5.1 presents estimates of the quantity of CRT-containing products collected by
U.S. recyclers going to the various end markets. These estimates were developed by an
industry expert who researched the U.S. resale outlets typically used by recyclers, as well
as interviewed individuals from firms conducting CRT refurbishing, glass processing,
and lead smelting. Data on glass-to-glass recycling were obtained via EPA's Office of
Solid Waste research. 22
Table 5.1
End Markets for EOL TVs and CRT Monitors Collected for Recycling in the U.S. in 2005
End Market
Resale "as is" or after some repair/upgrade in the U.S.
Resale "as is" or after some repair/upgrade abroad
Refurbishing or remanufacturing into specialty monitors in the U.S.
Refurbishing or remanufacturing into new TVs or specialty monitors
abroad*
CRT glass-to-glass factories in the U.S.
CRT glass-to-glass factories abroad
CRT glass to smelters in North America for lead recovery **
Plastic, metal, and other material recovery from demanufacturing***
Total
Tons/Year
3,000
3,500
2,500
107,500
4,000
24,000
10,000
20,500
175,000
% of
Total
2%
2%
1%
61%
2%
14%
6%
12%
100%
Source: World Reuse, Repair and Recycling Association, 2005. Figures for CRT glass-to-glass factories are based on EPA research.
*Industry experts interviewed by Robin Ingenthron report that about 30% of material destined for remanufacturing abroad is not
technically suitable for remanufacturing and has to be recycled or disposed. The recycling or disposal of unsuitable units occurs
abroad.
"Includes units shipped to one smelter in each of the U.S. and Canada.
***End markets for these materials are both domestic and abroad.
21 Robin Ingenthron of the World Reuse, Repair and Recycling Association developed these estimates after
gathering knowledge from Association members, industry contacts, and published data sources.
22 Research conducted by Bob Tonetti, U.S. EPA, Office of Solid Waste, 2006.
31
-------
In total, about 175 thousand tons of CRT-containing products coming from U.S.
electronics recyclers in 2005 was accounted for in this analysis. The estimates presented
in Table 5.1 show that the largest portion of this material (about 61 percent) was sent to
markets abroad for the purpose of refurbishing or remanufacturing CRTs into new
televisions or specialty monitors. The next largest portion (about 14 percent) was CRT
glass sold to markets abroad for glass-to-glass processing, while lead recovery in North
America accounts for about 6 percent of the material. Resale for reuse in the U.S. and
abroad, as well as remanufacturing in the U.S. are other relatively small end markets.
There may be additional end markets that were not identified or quantified in these
estimates, however, it is believed that the major end markets were covered and therefore
these estimates should assist us in understanding where used and EOL CRT devices and
components are going after they are collected and processed by U.S. electronics
recyclers. The estimates in the table were not intended to be comprehensive estimates of
the actual tonnage of EOL CRT-containing products collected and managed. They were
developed to explore the distribution of these EOL products among the major end
markets.
Further, these estimates reflect the state of the recycling industry in 2005. Changing
industry trends have had a significant impact on these estimates since then and will
continue to alter the distribution in the near future. For example, the domestic market for
CRT glass has changed since U.S. CRT glass-to-glass factories have closed. Further into
the future, as more flat panel monitors and TVs replace CRT-containing products, end
markets will shift again when these products are ready for EOL management.
32
-------
6.0 Summary and Conclusion
The purpose of this report is to establish a baseline regarding the management of EOL
electronic products. These data can be referenced in the future to characterize changes
and trends with respect to the generation and handling of EOL electronic products.
The baseline data addresses televisions, cell phones, personal computers (including
desktops, laptops, monitors, keyboards, and mice), and hard copy devices (e.g., printers,
scanners, faxes) sold between 1980 and 2004. This timeframe is probably adequate to
account for all but a very small portion of the quantity of computer products used in the
United States. It does not, however, account for televisions sold before 1980 that might
still be in storage, which is probably a relatively small quantity at this point.
Annual Quantity of Obsolete Electronic Products Generated for EOL
Management
As described in Section 3.1, the analysis presents the quantity of EOL electronic products
generated for management annually. Table 6.1 presents these estimates for the period of
1999 through 2006. The table also puts the totals in terms of pounds per capita based on
U.S. Census population estimates.23 According to the modeling results, 2.2 million tons
of EOL TVs, cell phones, and personal computer products were ready for EOL
management in 2005. This is equivalent to about 15 pounds per capita.
Table 6.1
Estimated Annual Products Ready for EOL Management
Year
TVs
Units(mill) Tons(OOO)
1999
2000
2001
2002
2003
2004
2005
2006
16.5
19.0
19.8
19.9
23.5
23.5
24.0
25.7
497.7
585.0
615.1
635.8
734.1
753.6
786.0
846.8
Computer Products
Units(mill)
123.7
117.8
136.0
151.9
176.3
194.6
206.6
195.2
Cell Phones
Tons(OOO) Units(mill) Tons(OOO)
553.5
691.1
824.5
1,009.5
1,218.2
1,323.8
1,429.1
1,313.4
18.8
25.0
37.9
55.2
75.8
96.8
116.5
133.8
3.0
3.4
3.4
3.5
7.5
7.5
8.2
10.4
Units(mill)
159.0
161.7
193.7
227.0
275.6
314.8
347.1
354.6
Total
Tons(OOO) Lbs/Cap
1,054.2
1,279.5
1,443.1
1,648.8
1,959.8
2,085.0
2,223.3
2,170.6
7.7
9.1
10.1
11.5
13.5
14.2
14.9
14.5
Source: ERG estimates based on modeling results.
Annual Quantity of Electronic Products Ready for EOL Management That Is
Recycled
Section 3.2 presents estimates of the quantity of EOL electronic products collected that is
recycled in the United States. These estimates are duplicated in Table 6.2. As shown in
the table, an estimated 345,000 tons of EOL TVs, cell phones, and personal computer
products were available for recycling in 2005. This is equivalent to about 2.2 pounds per
capita, which accounts for about 15 percent of the 15 pounds per capita estimated to be
'U.S. Census figures for beyond 2006 were interpolated based on Census projections for 2005 and 2010.
33
-------
ready for EOL management in 2005. If the recovery rate for recycling in 2005 were
increased from 15 to 35 percent, 5.2 pounds per capita would be recycled, increasing the
total weight of EOL electronics recycled to 778,000 tons.
Table 6.2
Estimated Annual Products Collected for Recycling
Year
TVs
Units(mill) Tons(OOO)
1999
2000
2001
2002
2003
2004
2005
2.5
2.8
2.9
3.0
3.5
3.6
3.7
74.1
86.9
89.5
96.4
108.6
115.7
122.0
Computer Products
Units(mill)
18.4
17.5
19.8
23.0
26.1
29.9
32.1
Cell Phones
Tons(OOO) Units(mill)
82.4
102.6
120.0
153.1
180.3
203.2
221.8
2.8
3.7
5.5
8.4
11.2
14.8
18.1
Tons(OOO) Units(mill)
0.008
0.010
0.013
0.018
0.043
0.054
0.067
23.7
24.0
28.2
34.4
40.8
48.3
53.9
Total
Tons(OOO) Lbs/Cap
157.0
190.0
210.0
250.0
290.0
320.0
345.0
1.2
1.4
1.5
1.7
2.0
2.2
2.3
Source: ERG estimates based on modeling results.
According to our estimates, for the study period 1999 to 2005, about 15 percent of the
electronic products ready for EOL management are recycled and the remainder is
disposed. The fact that the percent recycled has remained constant over this period
should not lead the reader to conclude that electronics recycling rates have remained
constant. The opposite is true. In 1999, 157,000 tons of EOL electronic products within
the scope of this study were recycled as compared to 345,000 tons in 2005. However, at
the same time, the products generated annually for EOL management has been
increasing. As shown in Table 6.2, the per capita recovery rates have exhibited steady
increases, from an estimated 1.2 to 2.3 pounds per capita in 1999 to 2005.
An important factor in this analysis is the estimate of the quantity of EOL electronic
products recycled annually. The scope of this study did not allow for an in-depth analysis
of this factor. However, given that this estimate directly affects the proportion of EOL
products that are recycled versus disposed, we conducted a sensitivity analysis to test the
effect on the results by assuming that our estimate of the amount recycled was
underestimated by 20 percent. When assuming that 412,000 tons of EOL electronic
products were recycled in 2005 (a 20 percent increase over 345,000 tons), the estimated
portion of the total EOL products generated for management that are recycled increases
from 15 to 19 percent and the portion that is disposed of decreases from 85 to 81 percent.
Based on our analysis and industry estimates of electronic products recycled (See Section
3.2.), we believe that 15 to 20 percent is a reasonable range to represent the portion of
electronic products generated for EOL management that are recycled.
Electronic Products in Storage and Other Stages of the Lifecycle
In Section 3.3 (Table 3.6), we estimated that in 2005, about 180 million units of EOL
electronic products could be in storage. This represents the quantity of material that may
be in home closets, attics, or basements. Televisions comprise about one-half and desktop
computers about one-quarter of the stored products in that timeframe.
34
-------
According to our modeling results, of all the electronic products sold (excluding cell
phones) from 1980 to 2004, by 2005 almost half are still in use either by the first or
subsequent users. (See Figure 6.1.) Of the remainder, about 42 percent would have
already been managed via recycling or disposal, and about 9 percent are likely to be
ready for management, but still are in storage.
. Already
Still in Use f \ Managed
CD
In Storage
Figure 6.1 Stages of Lifecycle: Electronic Products 2005
For computer products, we estimated the number of products that might be in their "first
use" and "second use" or reuse stage. According to our modeling results, in 2005, of the
computer-related products sold from 1980 through 2004, 37 percent are in their "first
use" stage (i.e., the first owner is still using) and 22 percent are being reused (second
owner is using). Intuitively, we believe that the reuse of electronic products is high, since
these products are relatively expensive, and even though product turnover rates may be
short compared to other products, there is still substantial intrinsic value to the product.
In fact, according to Consumer Electronics Association survey data, consumer-to-
consumer reuse of electronic products through donations, give-away, or sales ranges from
65 percent of consumers for televisions to 83 percent of consumers for laptops.24
Quantity ofEOL Electronic Products Collected by Existing Collection Programs
In Section 4.0, we examined a number of mostly residential electronics collection
programs. The programs studied were not a statistical sample; rather, they were selected
based on the suitability and availability of the data. In this review of program data, we
found that, for municipal electronics collection programs the mid-range of 2 pounds per
capita (based on a range of 0.8 to 3.1 pounds per capita) of EOL TVs and personal
computer equipment was recovered for recycling. (See Table 4.1.) This wide range is
indicative of the maturity and type of program, as well as the extent of outreach and the
characteristics of the population; they all play an important role in determining the
recovery rates. When there is a disposal ban, higher recovery rates may be experienced.
Electronic Products Collected for Recycling, Managed Domestically Versus Abroad
In Section 5, we developed estimates regarding where and how CRTs from computer
monitors and TVs were being handled after collection by electronics recyclers in the U.S.
24 DuBravac, Shawn G., From Here to There: Facts on Product Life Cycles and Recycling, Consumer
Electronics Association. 2005.
35
-------
Of the CRT devices collected in the U.S. by recyclers, more than 80 percent25 is sent to
end markets abroad for resale, refurbishment, remanufacturing, or materials recovery.
Most of the used and EOL electronic products containing CRTs (about 60 percent) are
sent to end markets in Asia, South America, and elsewhere, where CRTs that are in
working order are refurbished or remanufactured into new TVs or specialty equipment
(e.g., monitors for medical or other technical applications). However, the end market
distribution is changing rapidly as glass factories close and the increasing popularity of
flat panel displays changes the nature of the EOL stream from TVs and computer
monitors.
The handling of CPUs, laptops, and hard copy peripherals in the U.S. versus abroad was
not examined in this analysis. These products at EOL have different technical and
economic characteristics and therefore, the resale and recycling end markets are not at all
similar. Non-CRT-containing products may be addressed in a subsequent analysis.
25 The "more than 80%" estimate is derived by adding the following categories from Table 4: 1) Resale
"as is" or repair abroad; 2) Remanufacturing abroad; and 3) CRT glass abroad. These three categories
yield about 77%. We also assume that a portion of two other categories in Table 4 should also be included,
bringing the total up to at least 80%: 1) CRT glass to lead smelters in North America; and 2) Plastic, metal,
etc. from demanufacturing.
36
-------
APPENDIX A
AGE DISTRIBUTION OF ELECTRONIC PRODUCTS
BASED ON ANALYSES OF DATA FROM
THE FLORIDA STATE DEPARTMENT OF ENVIRONMENTAL PROTECTION
-------
Age Distribution fa Desktops
20. cr
17.5-
15. cr
12.5-
c 10. cr
e
n
t
7.5'
5. O-
2.5-
o. cr
O
I
12
I
15
18
21
24
27
age
A-l
-------
Age Distribution for Monitors (CRT)
20.0"
17. 5"
15.0'
12. 5"
c 1O. O'
e
n
t
7.5'
5.0-
2.5"
O. O
1.2 4. 8
' I I i
8. 4 12. O 15. 6
age
19. 2
I '
22. 8
26. 4 3O. O
A-2
-------
Age Pstribution fa Laptops
50'
40"
P30-
e
r
c
e
n
1 20-
10-
3. 75
6. 25
I
8. 75
age
11. 25
13. 75
A-3
-------
Age Distribution for TV < 19 in
25'
2O~
P 151
e
r
c
e
n
* 10-
12
I
16
2O
24
28
I
32
36
A-4
-------
Age Distribution for TV > 19 in
14"
12'
1CT
r
c
e
n
t
\^
O
\
3
\
6
\^
9
12
15
I
18
21
24
27
r
3O
33
age
A-5
-------
Age uistri button for Scanners, Printers, hax Machines, and Muitl-unction uevices
22. 5
20. 0
17. 5
15. O
12. 5
c
e
n 1O. O
t
7. 5
5. O
2. 5
O. O
I
3
I
5
I
9
I
11
I
13
I
15
I
17
I
19
I
21
I I
23 25
I
27
29
age
A-6
-------
APPENDIX B
ESTIMATING THE QUANTITY OF EOL ELECTRONIC PRODUCTS DISPOSED:
ANALYSIS OF WASTE SORT STUDIES
-------
Waste sort studies characterize the waste streams sampled from both residential and
commercial disposal loads. These characterizations describe the volume of different
categories of materials disposed as a percentage of the solid waste disposed. These
figures can be used to estimate the pounds per person per year for each category. The
major material categories examined in the waste sorts are paper, plastic, metal, and glass,
etc.; each major category is further divided into specific products.
Although all the waste sort studies included a category for consumer electronics, the
contents of this category—i.e., its subdivision into individual product types, such as
televisions, monitors, and computers—varied from study to study. The annual data from
the waste sorts also varied, by the year the data were collected, but each sort presented
numbers that represented a full 1-year period. Six of the studies were statewide (covering
California, Georgia, Minnesota, Oregon, Pennsylvania, and Wisconsin) and one was local
(covering Seattle).26
Data from each study were analyzed in detail to make sure that the waste categories
matched each other as closely as possible. For example, major categories such as
construction and demolition debris, ash, sewage sludge, and other special wastes that
were included in some, but not all of the waste sort studies were removed. The yard waste
category was also removed, due to regional differences associated with this type of waste.
The percentages of the remaining categories were then adjusted to reflect the adjusted
composition.
A second series of adjustments were made to the broad waste categories to make the
studies more comparable. For example, if there were insufficient data in the waste sort
studies to match the specific items included in these broad categories between studies,
these categories were removed and replaced with national average data to fill in the data
gaps. An example of when such adjustment might be made is in the category of "durable
goods." Durable goods include such items as appliances, tires, carpet, and furniture and
furnishings. In some of the studies, these items were specifically cited, while in others,
they were included in other broad categories, such as "other metals," "other plastics," and
"wood and wood products." Another example is that some broad categories may also
have included non-MSW products such as lumber, windows, doors, fixtures, tubing, and
26 California Integrated Waste Management Board. Contractor's Report to the Board: Statewide Waste
Characterization Study. Publication Number 340-04-005. December 2004.
Department of Community Affairs. Georgia Statewide Waste Characterization Study. June 22, 2005.
Solid Waste Management Coordinating Board et al. Statewide MSW Composition Study: A Study of
Discards in the State of Minnesota. March 2000.
Oregon Department of Environmental Quality et al. 2002 Oregon Solid Waste Characterization and
Composition. 2004.
Pennsylvania Department of Environmental Protection et al. Final Report: Statewide Waste Composition
Study. 2003.
Cascadia Consulting Group, Inc., et al. Wisconsin Statewide Waste Characterization Study: Final Report.
May 2003.
Seattle Public Utilities. 2002 Residential Waste Stream Composition Study. August 2003. 2004 Commercial
and Self-Haul Waste Streams Composition Study. September 2005.
B-l
-------
wiring from home or commercial renovation projects; also, automobile parts are
commonly included. Adjustments were made to correct inconsistencies between studies.
Table B.I shows the results of the waste sort analysis on a pounds per person per year
basis. Detailed data from the waste sorts were segregated to the extent possible to reflect
the consumer electronics that are the focus of this study. There is some uncertainty due to
the lack of information in the individual studies as to whether scanners and facsimile
machines are included in all of the factors shown.
The average of the waste sort studies, on a per person per year basis, is shown in the
following table. Although seven waste sort studies were analyzed, only six were judged
to have sufficient information to allow for the separation of the specific consumer
electronic product categories.
Table B.I
Discards of Electronic Products, in Average Pounds per Capita
Average of
Waste
Sorts California
Georgia Minnesota Wisconsin Oregon Seattle
Consumer Electronic Discards (Ibs/cap/yr)
Computer-related electronics
CRTs (monitors and televisions)
Total
Data year
3.6
5.8
9.4
(lbs/cap/yr)(lbs/cap/yr" (Ibs/cap/yr)
6.9
13.0
19.8
2002
2.0
0.9
2.9
2003/2004
2.5
1.2
3.7
1998
(Ibs/cap/yr)
1.0
12.6
13.6
2001
^bs/cap/yf
4.9
6.6
11.6
2002
(Ibs/cap/yr)
4.2
0.7
4.8
2002
Source: ERG estimates based on analysis of six waste sort studies. See footnote #26.
Applying the resulting average factor of 9.4 pounds per person per year to the U.S.
resident population, results in the discard estimates for the EOL electronic products
presented in Table B.2.
Table B.2
Estimated Discards of EOL Electronics Based on Waste Sort Averages
Year
1999
2000
2001
2002
2003
U.S. Population
(millions)
272.7
281.4
284.8
288.0
290.8
Estimated Discards
(Million Tons per
Year)
1.28
1.32
1.34
1.35
1.37
Comparing the results for 2003, both methods of estimating the amount disposed are
reasonably close. The estimate in Table 3.3, which was based on the modeling results of
EOL electronic products ready for EOL management less the amount recycled, yielded
B-2
-------
1.7 million tons, whereas the estimate developed based on the average results from six
waste sort studies estimated 1.4 million tons.
B-3
-------
APPENDIX C
AVERAGE COLLECTION RATES FROM EXISTING ELECTRONICS COLLECTION
PROGRAMS
-------
Recovery rates experienced by existing state and local electronics collection programs
allow us to begin to understand how residents respond to recycling opportunities when
they are available. This is important when municipalities or other entities are planning to
establish electronics recycling opportunities to a given population. We examined
recovery rate information on a per capita basis from a number of programs to get a sense
of the quantities of material that might be anticipated.
Each of the recycling programs selected for this analysis met the following criteria:
• The program served a discrete and quantifiable population. Examples include
state programs available to all state residents, collections run out of county or
municipal waste management facilities, and curbside collections within a specific
city or county.
• The program involved regular daily, weekly, or monthly collection—not a one-
time or annual collection event.
Hennepin County, Minnesota, which includes Minneapolis and several surrounding
communities, has one of the longest-running electronic waste recovery programs in the
nation. As of 2004, the county collected 54 percent of its e-waste at two drop-off centers,
29 percent through curbside pickup in the city of Minneapolis, and the remainder through
various hazardous waste collection events.27 Figure C.I shows annual per capita recovery
by product type from 1992 to 2004, based on data provided by the Minnesota Office of
Environmental Assistance. (These data were originally collected in units; they were
converted to pounds based on a standard set of mass factors that were used for all
program data presented in this section.)28 Per capita recovery was calculated using
county-wide population figures from the U.S. Census. Annual population figures between
1990 and 2000 were estimated by linear regression between the 1990 and 2000 Census
totals. Post-2000 population estimates were based on a linear regression between the
2000 Census total and the 2003 Census estimate.
Waukesha County, Wisconsin, is located just west of Milwaukee. The county collects
EOL computer products, but not TVs, at several drop-off locations, each open at least one
day per week to all county residents. Businesses recycle computers through a separate
program, although some businesses could have dropped electronics off at the residential
collection sites. Recovery data were provided by the Waukesha County Recycling and
Solid Waste Office, with units converted to pounds using our standard conversion factors.
Figure C.2 shows the per capita recovery data, which were calculated using the county-
wide population from the 2000 Census.
27 www.electronicsrecycling.com/NCER/UserDocuments/121541970Hennepin_Brand_Tally.pdf
28 Standard factors assumed for all per capita calculations were: CPUs, 23 pounds; monitors, 30 pounds;
Printers/peripherals, 21 pounds; laptops, 9 pounds; and TVs, 63 pounds.
C-l
-------
Figure C.I. Per Capita Electronics Recovery in Hennepin County, MN, 1992-2004
3.5
D Printers
• TVs
nCPUs
• Monitors
year
Notes: The program began in fall 1992. Minneapolis curbside electronics recycling began in 1997.
The program began tracking laptops as a separate category in 2003. Per capita laptop recovery in
2003 and 2004 was 0.002 pounds and 0.005 pounds, respectively.
Source: Minnesota Office of Environmental Assistance, September 2005.
Figure C.2. Per Capita Recovery of Computer Products Only in Waukesha County,
WI, 2000-2004
D Printers
nCPUs
• Monitors
2004
Notes: The program began in 2000 with two drop-off locations. A third drop-off location was added
in 2001; unlike the other two, it was open daily. A fourth drop-off site was added in May 2002.
Source: Waukesha County Recycling and Solid Waste Office, September 2005.
C-2
-------
The Canadian province of Alberta started an e-waste collection program in October 2004,
with collection locations throughout the province. Recycling is free to all residents, and is
funded by an Advance Recovery Fee (ARF) levied on certain electronic goods at the
point of sale. The Alberta Recycling Management Authority provided unit recovery data
through May 2005 by product type. Table C.I presents the recovery data in annual
pounds per capita, with the original 8 months of data multiplied by 1.5 (for an annual
estimate). Pounds per capita were calculated using the standard set of product mass
factors and the provincial population figure from Canada's 2001 Census. Note that, in
other programs that have been in operation several years, the recovery rate is usually low
in the first year and increases substantially in subsequent years.
Table C.I
Per Capita Electronics Recovery in Alberta
Product
Monitors
CPUs
TVs
Printers
Laptops
TOTAL
Pounds per Capita
0.43
0.29
0.40
0.17
0.004
1.30
Notes: Annual recovery rate is based on prorated data from October 2004 through May 2005. The province-wide
program began in October 2004. As of May 2005, more than 100 collection sites and e-roundups had been established.
Source: Alberta Recycling Management Authority, September 2005.
Citrus County, Florida, is a rural county located north of Tampa along the Gulf Coast.
The county operates an e-waste drop-off site at the county landfill. Businesses must pay a
per-item fee, while recycling for residents is free.29 Product recovery data for 2002 were
available from the National Electronics Product Stewardship Initiative (NEPSI), which
maintained a database of e-waste recovery data for a short period of time. The 2002
figures represent a 9-month pilot, but were extrapolated to annual rates for comparison.
Data for 2003 and fiscal year 2005 for Citrus County were obtained from the Florida
DEP. Table C.2 shows Citrus County's e-waste recovery in pounds per capita, calculated
using standard mass factors and the county's 2004 population estimate (U.S. Census).
Table C.2
Per Capita Electronics Recovery in Citrus County, FL, 2002-2005
Product
Monitors
CPUs
TVs
Printers30
TOTAL
Pounds per Capita (Annual Rate)
2002
0.43
0.19
0.72
1.33
2003
0.20
0.06
1.65
0.08
1.99
Oct 2004-Sep 2005
0.32
0.13
1.93
0.10
2.48
Notes: The program began as a pilot in 2002 with a monthly drop-off. After the pilot stage, the program was expanded
to a daily drop-off
Source: NEPSI On-Line Data Base, 2003. Florida Department of Environmental Protection. September 2005.
29 www.bocc.citrus.fl.us/swm/electronics_recycling_tipping.htm
30 In 2002, Citrus County reported printers as part of "peripheral devices," not as a separate category.
C-3
-------
Several other counties in Florida have operated pilot or permanent e-waste collection
programs. Product recovery data from these counties were available from the NEPSI
database. Programs with regular collection and at least 8 months of data (mostly from
2002) include:
• Broward County (1 weekly drop-off location; pilot program)
• Charlotte County (1 daily drop-off location; pilot program)
• Lee County (1 weekly drop-off location; permanent)
• Miami-Bade County (1 daily drop-off location; permanent)
• Polk County (1 daily drop-off location; permanent)
• Sarasota County (daily curbside pickup and 1 daily drop-off location; pilot
program)
Table C.3 shows the per capita recovery from these counties, by product type. Unit totals
from the NEPSI database were converted to pounds using the standard conversion factors
and compared with county populations from the 2000 Census.
Table C.3
Per Capita Electronics Recovery in Several Florida Counties
Product
Monitors
CPUs
TVs
Printers31
Laptops32
TOTAL
Pounds per Capita, by County
Broward
0.07
0.04
0.05
0.02
0.18
Charlotte
0.08
0.07
0.07
0.22
Lee
0.09
0.05
0.05
0.03
0.0002
0.21
Miami-
Dade
0.03
0.01
0.13
0.01
0.19
Polk
0.03
0.01
0.27
0.01
0.32
Sarasota
0.13
0.05
0.24
0.03
0.0006
0.45
Source: NEPSI On-Line Data Base, 2003. Florida Department of Environmental Protection. September
2005.
Another state with extensive e-recycling infrastructure is Massachusetts. Since the early
1990s, the state Department of Environmental Protection (MA DEP) has provided grants
to help municipalities establish CRT collection programs.33 In April 2000, the state
banned CRTs from all solid waste disposal facilities, and since then, CRT collection
programs have gradually expanded to cover a large percentage of the state's 6.4 million
residents. Programs are operated by individual municipalities, and range from curbside
pickups to daily drop-off locations, special collection events, collaborations with
charities, or a combination of methods. Some programs charge a per-unit fee.
Each municipality reports annual collection totals to MA DEP. Reporting methods vary,
with some programs reporting recovery in pounds as measured by their vendor, others
reporting pounds calculated from the number of units recovered, and the rest reporting
31 Charlotte County reported totals for peripheral devices, but did not break these down by type of device.
32 Many programs did not report laptops as a separate category.
'www.mass.gov/dep/recycle/reduce/crtqanda.htm
C-4
-------
unit totals, which MA DEP converts to pounds using a set of mass factors. These data
only include products with CRTs and are generally not broken down by product type, so
we were unable to determine what portion was monitors as opposed to TVs. Figure C.3
shows the per capita recovery of total CRTs from 2000 to 2004, based on data provided
by MA DEP. Each year's per capita figure reflects the population served by the CRT
recycling programs reporting data to MA DEP for that year (municipal population figures
provided by MA DEP). This denominator increased overall from 2000 to 2004, as more
and more communities established recycling programs.
Figure C.3. Per Capita CRT Recovery in Massachusetts, 2000-2004
$
a.
s
o
a.
•o
O
Q.
2000
2004
Notes: Data covers only computer monitors and televisions with a CRT.
Source: Massachusetts Department of Environmental Protection, September 2005.
Table C.4 summarizes the results from the individual programs examined.
Table C.4
Range of Per Capita Recovery Rates Among Selected Programs, 2004
Product Type
Monitors
CPUs
TVs
Printers
Laptops34
Total
Recovery Range (Pounds per Capita)
Minimum
0.23
0.13
0.4
0.07
0.002
0.83
Maximum
0.60
0.35
1.99
0.17
0.007
3.12
It was not in the scope of this study to delve more deeply into what factors most affect
collection rates. We did, however, briefly analyze the data from Massachusetts to see if
there was a statistically significant difference in the per capita recovery rates of different
types of programs (e.g., drop-off, curbside pickup, special event).
34
Only Waukesha County, Alberta, and Hennepin County reported laptops as a distinct category in 2004.
C-5
-------
As stated earlier, Massachusetts has had a ban on CRT disposal since 2000. Each town
chooses what type, if any, of collection program it will offer residents. The state
maintains a database of what is collected by the towns that access the statewide recycling
contracts to support their programs. We analyzed the effect of program type on the
pounds product per capita recovered. We used a nonparametric one-way anova procedure
for a simple standard analysis of variance test on the raw data. The multiple sample
Kruskal-Wallis test was used to detect a significant difference in the distributions of the
program type variable. The analysis showed that program type is, in fact, statistically
significant in determining the pounds per capita recovery rate.
Table C.5 presents the analysis of the variance for the recovery rate variable. As shown in
the table, drop-off programs in Massachusetts collected the most material on a pounds per
capita basis. Massachusetts state recycling officials informed us that many of the curbside
pickup programs require residents to make an appointment for a pickup and in some
cases there is a monthly limit constraining the number of pickups that can be scheduled.
This factor could partially explain why the curbside programs had a lower recovery rate
than the drop-off programs. Of the three main types of programs (curbside, drop-off, and
special event), special events had the lowest recovery rates. The data for the combination
programs are difficult to interpret because we have no information regarding the
predominant method of collection in these communities.
Table C.5
Analysis of Variance for the Pounds per Capita Variable
Classified by Program Type
Program Type
Drop-off
Curbside pickup
Special event
Drop-off/special event
Drop-off/curb side
Curbside/special event
Charity /curb side
Drop-off/curbside/special event
Number in Sample
524
71
67
25
60
7
5
9
Mean Pounds per
Capita
2.82
1.95
1.18
1.36
1.30
1.15
0.55
0.61
C-6
------- |