MANAGEMENT OF ELECTRONIC WASTE IN THE UNITED STATES:
                    APPROACH TWO

                    Draft Final Report
                       April 2007
                    EPA530-R-07-004b

-------
                                ACKNOWLEDGEMENTS


This Report is based on analyses prepared under contract for the Office of Solid Waste by
Industrial Economics, Incorporated. The Office of Solid Waste would like to thank especially
Jason Price and Cho-Yi Kwan for their assistance in developing the model upon which this
report is based and in drafting the report.

-------
TABLE OF CONTENTS

CHAPTER 1 -Introduction	4
CHAPTER 2-Summary of Methodology	7
  Model Structure	7
  Data Inputs	9
CHAPTER 3-Summary of Data Inputs	10
  Sales Data	10
  Mass Data	19
  Product Lifespan	21
    Initial Service Life	21
    Initial Service Life Probability Distribution	25
    Second Service Life Duration and Probability Distribution	28
  End-of-Life Management	30
  End-of-Total-Life E-waste Management	31
  End-of-First-Life E-waste Management	38
CHAPTER 4 - Baseline E-waste Assessment Results	42
  Retirement Estimates	42
  Management of Retired Electronics	42
  Electronics Put Into Re-Use or Storage	46
  Management of Electronics Reaching the End of Its First or Second Life	47
APPENDIX A: Glossary of Terms	53
APPENDIX B: Estimating the Total Lifespan and Second Life of Electronic Devices	55
APPENDIX C: Estimating the Reuse and Storage of Electronic Devices	63
APPENDIX D: Material Composition of Select Electronic Devices	68
APPENDIX E: Material Composition of Retired Electronics	75

-------
CHAPTER 1  -  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 lifespans 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 Two."

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.
The scope of this report includes the following electronic products:

   •   Televisions,
   •   Personal  computers (desktops, laptops, and computer monitors),
   •   Printers3
   •   Computer mice
   •   Keyboards
   •   Cell phones.
 Approach One includes hard-copy peripherals, which is comprised of printers, scanners, and fax machines while Approach Two only includes
printers.

-------
The purpose of this report is to present EPA's baseline assessment, using the methodology
outlined in this report, on e-waste generation and management in the U.S. We present this
information in the following chapters:

  •  Summary of Methodology. In this chapter, we describe the methodology that we
     developed for estimating the amount of e-waste generated in the U.S. each year and for
     assessing how this waste is managed.
  •  Summary of Data Inputs. Following the discussion of our methodology, we summarize
     the main data sources and assumptions used to implement this methodology.
  •  Baseline E-waste Assessment Results. In the final chapter of this document, we present
     and discuss the results of our analysis.
Exhibit 1-1 summarizes the results of our analysis, averaged over the 2003-2005 period. As
indicated by the exhibit, CRT televisions and monitors made up nearly two-thirds of the subset
of e-waste analyzed in the report in the U.S. between 2003 and 2005. Thus, these electronic
products will continue to be an important part of the U.S. e-waste stream for years to come
despite the ongoing shift to flat-screen televisions and monitors. The results in Exhibit 1-1 also
indicate that the vast majority of U.S. e-waste is deposited in landfills, while approximately 20
percent is recycled.  By comparison, EPA estimates that approximately 32 percent of the
municipal solid waste generated in the U.S. in 2005 was recycled.4 We present a more detailed
accounting of our results in Chapter 4 below.
4 Estimate includes composting. U.S. EPA, Municipal Solid Waste in the United States: 2005 Facts and Figures, October 2006.

-------
       EXHIBIT 1-1   SUMMARY OF ANNUAL E-WASTE GENERATION AND MANAGEMENT:
                           2003-2005 (THOUSANDS OF TONS)1
PRODUCT

Desktop Computers
Laptop Computers
CRT Monitors
LCD Monitors
Televisions
CRT Televisions
Projection TVs
Cell Phones
Printers
Keyboards
Computer Mice
All Products
MANAGEMENT METHOD
RECYCLED2
65.7
7.0
97.5
0.6
112.5
98.5
14.0
1.9
62.0
14.1
0.9
362.2
LANDFILLED
180.9
19.3
291.6
1.8
709.1
620.7
88.3
7.9
170.6
38.6
2.4
1422.1
INCINERATED
5.1
0.5
8.3
0.0
20.2
17.7
2.5
0.2
4.8
1.1
0.1
40.4
TOTAL
251.7
26.8
397.4
2.4
841.7
736.9
104.8
10.0
237.5
53.8
3.4
1,824.8
Notes:
1 . The results presented in this exhibit represent average annual tonnages for the 2003-
2005 period.
2. As described in the main body of this report, the recycling estimates in this report
include amounts exported; however, insufficient data are available to distinguish
between recycled and exported waste.
CHAPTER 2 - SUMMARY OF METHODOLOGY
To assess the baseline generation and management of electronic waste in the U.S, we developed
a waste flow model that tracks e-waste generation and management over time for the electronic
products listed in Chapter 1. Based on several data inputs, the model estimates the number and
total mass of electronic products that enter the U.S. waste stream each year and apportions this
waste across various management options (e.g., recycling, landfill disposal, and incineration).
This chapter provides an overview of the methodology reflected in the model and summarizes
the data inputs necessary to implement this methodology.
MODEL STRUCTURE

The waste flow model developed for this baseline assessment simulates the generation and
management of electronic waste based on a series of sequential calculations, as outlined in
Exhibit 2-1.  First, based on sales (in units) and per unit mass information, the model calculates

-------
annual tonnages sold for each product. Following each cohort of products (e.g., laptop
computers sold in 1995) through time, the model then allocates a portion of each cohort to
storage/re-use when it reaches the end of its first life (i.e., when the original owner of a product
stops using it), allowing some products to move on to a second life.  The model simulates the
EOL management of the remainder, allocating the waste to landfills, incinerators, or recycling.

For products moving onto a second life5 (i.e., products stored or re-used at the end of their first
life), the model follows a similar procedure, allocating these items to each management option.
For any given year, the waste flow model estimates total retirements (by management method)
by summing devices reaching the end of their first life that year and devices reaching the end of
their second life.
                         EXHIBIT 2-1. WASTE  FLOW MODEL DESIGN

Sales
TV f f-111 T ,1^TT,
> Mass Sold Ltngtn ol rust
	 1 	 Life Span



> i
% Stored/Reused
End ol rnsl Lile '
i
1
% Landfilled
% Recycled
% Incineratec
% Exported
\< \< V T
Amount Landfilled/Recycled/
Incinerated
After First Life
i
1
Total Retire



Second Life
|



Length of Second
Life Span
1
V
End-of-Life Retirement
I I I

i i i
% Landfilled % Recycled % Incinerated % Exported
Amount Landfilled/Recycled/
Incinerated
at End-of-Life
i
1
jment



5 A product's second life could include multiple users and stages of use.

-------
DATA INPUTS

As suggested by the schematic presented in Exhibit 2-1, the waste flow model requires several
data inputs to assess the generation and management of electronic waste in the United States for
any given year.  These inputs include: sales, product mass, the initial service life of each product,
the second service life of each product, probability distributions for any first and second service
lives expressed as a range, and EOL management allocation weights.6  We describe each of these
inputs below:

  •   Sales: A key input for the model is the annual sales (in units) for each electronics product.
  •   Mass: To estimate the tonnage of electronic products sold and retired, the model requires
      per unit mass estimates for each product.  Since products produced in more recent years
      may have a different per unit mass than products sold in earlier years, the model requires
      product-specific mass estimates by year (e.g., for desktop computers sold in 1998).
  •   First Service Life: The first service life of a device is the length  of time the electronic
      product is used by its original or initial owner. The model uses this information to estimate
      the volume of electronics reaching the end of their first life each year.  To reflect the
      uncertainty associated with each product's initial service life, the model allows users to
      enter this input as a range.  For example, if desktops sold in 1990 have an average initial
      service life of two to four years, users can enter two years as the low end of the initial
      service life range and four years as the high end.
  •   First Service Life Probability Distribution: For first service life estimates expressed as a
      range, the model requires users to specify a service life probability distribution.  For
      example, if desktops sold in  1990 have an average initial lifespan of two to four years,
      model users can specify that 25 percent of desktops will be retired by their original users
      after two years of use, 50 percent in the third year, and the remaining 25 percent in the
      fourth year.
  •   Second Service Life: Similar to the first service life, the model requires information on the
      duration of the second service life of each electronics product. We define second service
      life as the length of time over which a product is reused or kept in storage after its first life.
  •   Second Service Life Probability Distribution: The model requires a probability
      distribution for each second service life estimate expressed as a range.
  •   Management Allocation Weights: The model requires users to specify how devices at the
      end of their first life are allocated across the following management options: recycling,
      landfilling, incineration, and re-use/storage.  Similarly, the model requires users to indicate
      how devices reaching the end of their second life are allocated across the following
      management options: landfilling, recycling, and incineration.7
6 For any product, an allocation weight is the percentage of units at the end of their first or second life that the waste flow model apportions to a
 specific management method (e.g., incineration).

7 Re-use/storage is not a management option at the end of a product's second service life.

-------
CHAPTER  3 - SUMMARY OF DATA  INPUTS
As explained in the previous chapter, the baseline e-waste assessment presented in this document
is based on a data-intensive waste flow model that tracks e-waste generation and management
over time.  This chapter summarizes the various information sources and assumptions supporting
the development of the model's inputs, which include product sales, product mass, first service
life durations and probability distributions, second service life durations and probability
distributions, and management method allocation weights.8
SALES DATA

As part of our effort to estimate the total volume of e-waste generated in the U.S., we collected
data on electronics sales (in units) from 1975 through 2004 based on publicly available data from
sources such as INFORM and the U.S. Census Bureau, and on limited data made publicly
available by the market research firms Gartner and Display Search.9 Exhibit 3-1 summarizes the
relevant data available from these sources.
                      EXHIBIT 3-1. SOURCES OF ELECTRONICS  SALES DATA
 SOURCE
                            REPORT(S) AND DATA AVAILABLE
 U.S. Census
 Bureau
 INFORM

 Gartner

 DisplaySearch
• Current Industrial Report series, "Computers and Office and Accounting Machines"
(annually): Domestic shipments, imports, and exports (units) of desktops, laptops, monitors,
keyboards and printers.
• Current Industrial Report series "Consumer Electronics" (annually): Domestic shipments,
imports, and exports (units) of CRT televisions and projection televisions.
• Waste in the Wireless World: The Challenge of Cell Phones: Presents and cites Gartner data
published in 2000 on U.S. cell phone sales (units) from 1995 through 2000, as well as forecasted
sales for 2001 through 2004.
• Periodic Press Releases: Aggregate PC shipments to U.S. retailers (in units) and global cell
phone unit sales.
• Periodic Press Releases: Global (and, in some cases,  North American) LCD monitor sales.
8 As part of EPA's effort to characterize e-waste generated in the U.S., we also collected information on the material composition of electronic
 products. Appendix D summarizes the composition of televisions, desktops, laptops, monitors, cell phones, and computer peripherals. We were
 unable to locate information on the composition of printers, mouse devices, and projection televisions.  We note that the material composition
 information available for these products does not in all cases account for 100 percent of the materials contained in these products.

9 Our sales estimates for some products do not go as far back as 1975. Some of the products included in this analysis were not yet on the market in
 1975 (e.g., cell phones), and for other products, the available data do not extend as far back as 1975.
                                                                                                  10

-------
Exhibit 3-2 presents our estimates of electronics sales for the 1975-2004 period.  These estimates
suggest that sales of personal computers (laptops and desktops) have been growing steadily since
1978, although desktop sales fell by nearly 14 percent during the economic slowdown of 2001.
Our sales estimates also suggest that sales of CRT monitors have been declining since hitting a
high of 40 million units in 1999, in all likelihood because of consumer substitution to LCD
monitors.  In contrast, as of 2004, sales of CRT televisions were increasing, which may reflect
relatively high prices at the time for substitutes, such as LCD and plasma televisions.  Our sales
estimates also show that cell phone sales increased dramatically in the late 1990s.

We discuss the available sales data for each product in further detail below.
DESKTOP COMPUTERS

To estimate 1978 through 2004 desktop sales (excluding "white boxes," or computers manufactured
by parties other than a branded manufacturer), we used the U.S. Census Bureau's "Computers and
Office and Accounting Machines" Current Industrial Report (CIR) series, which contains sales-
related data for brand name desktop computers. 10'u The CIR series includes separate data for
domestic shipments (i.e., domestic production), exports, and imports.  Therefore, we estimate
sales as the sum of domestic shipments and imports minus exports. For those years  in which no
CIR data were available, we generate sales estimates by interpolating between the years where
data were available.

Due to various limitations in the CIR data, we made several assumptions to ensure that our sales
estimates are methodologically consistent over the entire 1978-2004 period.  For example,
between 1978 and 1989, the CIR series includes four categories of computers: general-purpose
digital computers, general-purpose analog and hybrid computers, special-purpose digital
computers, and special-purpose analog and hybrid computers. We use the CIR data for general-
purpose digital computers as a proxy for non-white-box desktop computer data for these years
because later CIRs indicate that desktop computers are reflected only in the general-purpose
digital computers category.12 In addition, from 1989 through 2004, the CIR series does not
always provide the same level of product detail for imports and exports as it does for domestic
shipments.  As a result, to estimate sales for these years, we assume that the composition of
desktop exports and imports reflects the composition of domestic shipments.  For example,
although the CIR series reports domestic shipments of desktop and laptop computers separately,
it combines the two in the same category for imports and exports.  Therefore, we assume that
10 The domestic shipment data reported in the CIR for 1977 are drastically lower than subsequent years and appear to be inconsistent with the
 numbers reported for later years. As a result, we limit our estimates to desktops sold from no earlier than 1978.

11 White boxes are customized, non-branded computers that retailers assemble themselves from individual computer components.

12 Laptops may also be reflected in this category. However, because laptops were a relatively new technology in the late 1980s, we assume that a
 negligible number of laptops are reflected in the CIR general-purpose computers category.  In addition, large-scale processing computers may also
 be reflected in the CIR general-purpose digital computer data between 1978 and 1991. To the extent that such units were sold during this period,
 we may overestimate non-white-box desktop sales for these years.


                                                                                           11

-------
desktops' share of the combined desktop and laptop category for imports and exports is the same
as its share of domestic shipments.
WHITE BOX DESKTOP COMPUTERS

We used data from the Census CIR series to estimate sales of non-white-box desktop computers
and laptops. To estimate white box sales, we employed data on the number of computers in use
and the number of non-white box units sold on an annual basis.

To estimate white box sales for 1990 and earlier years, we used publicly available data reported
by the market research firm eTForecasts on the number of computers in use in 1980, 1985, and
1990, and our estimates of brand-name (i.e., non-white-box) desktop sales during this period.
Assuming that computers sold in the 1980s had a lifespan of four years, we used our estimates of
brand name computer sales to estimate the  number of brand name desktops in use annually.13
For example, we estimate brand-name computer use in 1990 as the sum of sales from  1987
through 1990.  We then estimate white box computer use for 1990 by calculating the difference
between the total number of desktops in use, as reported by eTForecasts, and our estimate of the
number of brand-name computers in use. Continuing with our 1990 example, the difference
between our estimate for brand-name desktop use in 1990 and  eTForecasts'  use estimate for all
desktops represents the estimated use of white box computers in 1990. To estimate white box
computer sales for 1990, we assume that the ratio of white box computer sales to total computer
sales in 1990 is the same as the ratio of white box computer use to total computer use.  We
followed a similar procedure for 1985, using eTForecasts' estimate of computer use in  1985 and
our estimates of 1982-1985 brand-name desktop sales.  The results of our analysis suggest that a
negligible number of white box units were  sold in 1985 or earlier. To estimate white box sales
for  1986 through 1989, we followed a two-step process. First,  we estimated the white box
market share for these years by assuming that the white boxes'  share of the desktop market grew
linearly between the market share  we estimated for 1985 (0 percent) and 1990 (28.1 percent).
We then applied these values to our estimates of brand-name desktop sales for these years to
generate white box sales estimates.14

To estimate white box sales between 1997  and 2004, we used publicly available data from
Gartner press releases on the size of the overall PC market (non-white-box desktops, laptops,
white box systems, and in some cases servers) and the non-white-box desktop and laptop sales
estimates were derived from the Census CIR data. Using these data,  we estimate white box sales as
follows:
13 This four-year estimate represents the high end of our estimate of a computer's first life, as indicated in Exhibit 3-5 below. Because computers
 were not as widely used in the 1970s and 1980s as in the 1990s, we assume that the high end of the lifespan range is more appropriate for use in
 estimating pre-1990 white box sales.

M More specifically, if M equals the white boxes share of the desktop market and 6 equals the brand-name desktop sales, we estimated white box
 sales as MxB/(1-M).
                                                                                       12

-------
              WS=PCS-SS-DS-LS

       where, Ws=White box sales

              PCS= Total PC shipments (non-white-box desktop computers, servers, white box desktop
              computers, and laptops)

              Ss= Server shipments
              Ds= Non-white-box desktop sales

              Ls= Laptop sales

Gartner's press releases provide estimates of annual PC sales from 1997 through 2004 and
annual server sales for 2000 through 2002. These data suggest that servers, on average,
accounted for approximately 3.8 percent of the PC market between 2000 and 2002.  To estimate
server sales for other years, we assume that this percentage applies to any year for which the
Gartner PC sales data include servers.15 Using these data in conjunction with our estimates of
non-white-box desktop and laptop sales, we estimate that white boxes made up 36.6 percent of
the PC market in 1997 and 22.6 percent of the market in 2004.

For 1991 through  1996, we were unable to identify data on white box sales or the white box
share of the desktop market.  Therefore, to estimate white box sales for this period, we assumed
the same interpolation approach outlined above for 1986 through 1989.
LAPTOP COMPUTERS

Similar to our analysis of desktop computer sales, we derive our estimates of domestic laptop
sales from the U.S. Census Bureau's "Computers and Office and Accounting Machines" CIR
series. Inadequate data are available from the CIR series to estimate laptop sales prior to 1993;
therefore, we do not present estimates of laptop sales for those years.  For 1997-1998 and 2000-
2004, our basic approach for estimating domestic laptop sales is similar to our approach for
desktops in that we estimate U.S. laptop sales as domestic shipments plus imports less exports.16
To estimate laptop sales in 1999, we interpolate between the CIR data for 1998 and 2000.17 For
1994 though 1996, the CIR series combines domestic laptop shipments with other devices with
attached displays in a category called "Portable Computers." Therefore, we do not use the CIR
data to estimate laptop sales for these years. Publicly available data from Gartner, however,
15 Gartner's PC sales data reflect server sales for every year, except for 1997 and 1998.

16 To estimate imports and exports of laptops for these years, we assume that the composition of computer exports and imports is the same as the
 composition of domestic shipments (i.e., if laptops represent 0.17 percent of total domestic computer shipments in 2004, then we assume laptops
 represent 0.17 percent of computer imports and exports in 2004).

17To estimate total laptop sales for 1999, we calculated the laptops' share of the personal computer market in 1999 based on 1998 and 2000 data,
 and apply the percentage to the 1999 total computer sales.


                                                                                          13

-------
indicates that global laptop sales increased by 57.7 percent between 1995 and 1998.18 Based on
this estimate and our CIR-derived estimate of U.S. laptop sales in 1998, we were able to estimate
U.S. laptop sales in 1995. To estimate 1996 sales, we interpolated between our 1995 sales
estimate and our CIR-based estimate for 1997, applying the same methodology used to estimate
1999 sales. To estimate sales in 1994, we assumed that the CIR growth rate for "Portable
Computer" sales (i.e., sales of laptops and other devices with attached displays combined)
between 1994 and 1995 was the same as the growth rate in laptop sales. Based on this growth
rate and our sales estimate for 1995, we projected backwards in time to estimate sales in 1994.
CRT MONITORS

To estimate annual sales of CRT-monitors, we relied on domestic shipment, import, and export
data available from the U.S. Census Bureau's CIR "Computer and Office and Accounting
Equipment" series, and the information described above for desktop computer sales.  More
specifically, for 1989-2004, we used the CIR data to estimate CRT sales by estimating total
shipments less exports plus imports.19'20 For 1978 through 1988, insufficient information is
available to estimate CRT monitor sales.  As a result, we assume that the growth rate in monitor
sales for this period is the same as the growth rate in desktop sales.
LCD MONITORS

We estimate U.S. sales of LCD monitors based on data released by the market research firm
DisplaySearch. Between 1998 and 2004, DisplaySearch issued quarterly press releases on the
global LCD market that in many cases included estimates of North American LCD monitor
sales.  Based on these data, we developed a time series of North American LCD sales for the
entire  1998-2004 period.  Although the DisplaySearch press releases do not report North
American LCD sales for each quarter of the 1998-2004 period, we generated sales estimates in
such cases through interpolation of the DisplaySearch data for other quarters.  To estimate U.S.
LCD monitor sales from these North American sales estimates, we assume that the ratio of U.S.
to North American LCD monitor sales is the same as the ratio of U.S. to North American GDP.21
18 Gartner, Inc. as cited in cnn.com, "Mobile workforce strains IT staff," January 18, 1999,
 http://www.cnn.com/TECH/computing/9901/18/roadwarriors.ent.idg/

19 For 1989 through 1991, the CIR series includes all monitors in a single category, "Monitor-like or graphic displays, excluding graphic terminals."
 Because other monitor technologies were not highly developed in the early 1990s, we assume that all monitors reflected in the 1989 through 1991
 CIRs are CRT monitors.

20 The CIR series distinguishes between CRT and other monitors from 1992 through 2004, but reports sufficient data to estimate sales only for 1992-
 93 and 1997-2004. To estimate CRT monitor sales for 1994 through 1996, we interpolated from the 1993 and 1997 estimates we derived from the
 CIR data. In addition, the CIR series contains CRT monitor import and export data for 2003, but lacks complete information on shipments from
 domestic manufacturers for this year. To estimate 2003 sales, we used the CIR import and export data for 2003 and the average of the CIR
 domestic shipment estimates for 2002 and 2004.

21 Consistent with DisplaySearch, we define North America as the U.S. and Canada for the purposes of this analysis.


                                                                                             14

-------
EXHIBIT 3-2. VOLUME OF ELECTRONICS SOLD: 1975-2004 (MILLIONS OF UNITS)
YEAR
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
TOTAL
DESKTOPS
N.E.
N.E.
N.E.
0.7
1.1
1.5
1.9
3.5
5.5
8.2
7.0
7.8
9.8
10.4
11.8
14.4
16.7
18.0
19.4
20.9
23.2
22.9
27.7
33.1
37.8
40.9
NON-WHITE
BOXES
N.E.
N.E.
N.E.
0.7
1.1
1.5
1.9
3.5
5.5
8.2
7.0
7.4
8.7
8.7
9.2
10.4
11.8
12.5
13.3
14.0
15.3
14.8
17.5
18.0
27.7
30.1
WHITE-
BOXES
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
0.4
1.1
1.8
2.7
4.1
4.9
5.5
6.2
6.9
7.9
8.1
10.1
15.1
10.1
10.8
LAPTOPS
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
1.4
1.9
2.0
2.7
2.9
5.3
6.6
CRT
MONITORS
N.E.
N.E.
N.E.
0.6
0.9
1.2
1.5
2.8
4.4
6.5
5.6
6.2
7.8
8.3
9.4
10.3
12.0
13.4
17.3
20.3
23.3
26.3
29.2
35.1
40.0
37.5
LCD
MONITORS
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
0.2
0.8
1.4
TOTAL TVS
11.5
14.4
14.7
17.4
16.2
16.8
18.2
17.4
21.3
24.2
28.1
27.7
27.7
22.9
25.6
23.2
21.4
28.1
27.9
26.6
25.3
31.1
34.1
35.9
37.6
37.4
CRT
TVS
11.5
14.4
14.7
17.4
16.2
16.8
18.2
17.4
21.3
24.2
28.1
27.7
27.7
22.4
24.9
22.5
20.6
27.3
27.1
25.4
23.6
28.7
31.4
32.9
34.7
33.2
PROJECTION
TVS
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
0.6
0.7
0.7
0.7
0.8
0.9
1.3
1.7
2.4
2.7
3.0
2.9
4.2
CELL
PHONES
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
14.5
16.6
22.2
30.6
49.3
71.2
PRINTERS
N.E.
N.E.
N.E.
1.2
1.7
1.8
2.8
3.2
4.7
5.8
4.6
5.5
6.8
7.8
10.7
9.6
9.8
12.6
16.4
22.4
28.5
34.5
26.2
31.9
43.4
46.4
KEY-
BOARDS
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
15.5
19.4
20.7
23.4
28.6
33.1
35.8
38.6
41.4
47.6
53.8
55.6
65.0
63.7
60.6
COMPUTER
MICE
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
N.E.
1.1
3.4
5.7
8.0
10.4
11.8
12.5
13.3
14.0
15.3
14.8
17.5
18.0
27.7
30.1

-------
YEAR
2001
2002
2003
2004
TOTAL
DESKTOPS
35.2
40.8
44.6
47.5
NON-WHITE
BOXES
26.0
29.4
34.4
36.8
WHITE-
BOXES
9.3
11.4
10.3
10.7
LAPTOPS
6.9
8.6
10.9
12.4
CRT
MONITORS
27.2
24.3
16.4
15.9
LCD
MONITORS
4.4
10.6
14.9
21.1
TOTAL TVS
32.1
38.3
36.8
46.9
CRT
TVS
28.4
33.4
31.1
38.4
PROJECTION
TVS
3.7
4.9
5.7
8.6
CELL
PHONES
68.9
74.4
89.6
116.2
PRINTERS
43.1
43.4
50.9
63.6
KEY-
BOARDS
57.5
54.4
51.3
47.2
COMPUTER
MICE
26.0
29.4
34.4
36.8
Notes:
1 . N.E. denotes that sales were not estimated.
2. Totals may not match due to rounding.
16

-------
CRT-BASED AND PROJECTION TELEVISIONS

To estimate annual sales of televisions between 1975 and 2004, we used shipment, import, and
export data from the CIR "Consumer Electronics" series.  The CIR series reports separate
domestic shipment estimates for CRT-based televisions and projection televisions, but combines
these categories in its reporting of television exports and imports.  To address this issue, we
assume that the composition of television exports and imports is the same as the composition of
television shipments (e.g., if projection televisions represent 13 percent of shipments in 2002, we
assume that projection televisions also represent 13 percent of projection television exports and
imports in 2002).

Due to limitations in the CIR television data, we made several assumptions in deriving our
television sales estimates from these data. First, for 1975 through 1987, the CIR series reports
only "Table and portable models," and "Console and consolette models." We assume that both of
these categories represent CRT-based televisions.  Second, for  1988 through 1990, the CIR series
separates televisions into "Table and portable models," "High definition television (HDTV),"
and "Projection televisions." To make our pre-1990 CRT television sales estimates consistent
with our estimates for later years  (i.e., post-1990), we consider all "Table and portable models"
to be CRT-based televisions, all "Projection televisions" to be projection televisions, and we
distribute high-definition televisions proportionately across the table and portable models and
projection televisions (i.e., if table and portable models represent 97 percent of total  table and
portable models, and projection televisions in 1990, then we assume 97 percent of high definition
televisions in 1990 are table and portable models).  Lastly, we estimated 1994 sales by
interpolating between our estimates for 1993 and 1995 because the CIR series does not contain
sufficient data on 1994 sales.22
COMPUTER PERIPHERALS (PRINTERS, KEYBOARDS, AND MICE)

The CIR "Computer and Office and Accounting Equipment" series contains limited sales-related
data for a number of peripheral products, including printers, keyboards, and computer mice.23
For printers, the CIR series presents no import/export data prior to 1978, and no domestic
shipment data prior to  1976; therefore, we limit our analysis of printers to units sold no earlier
than 1978.  Due to limitations in the CIR printer data for the 1978-2004 period, we use  several
approaches to generate printer sales estimates for these years.  For example, because the CIR
series contains no import data for printers between 1978 and 1985, we used the ratio of printer
imports to total domestic shipments of printers in 1986 and apply this ratio to the 1978-1985 CIR
printer shipment data to estimate printer sales for these years.  In addition, the CIR import and
22 Although the CIR series contains shipment, import, and export figures for televisions in 1994, the import estimate for this year (1.9 million units)
 is significantly less than the 16.3 million units and 13.2 million units imported in 1993 and 1995 respectively. Because of the magnitude of this
 discrepancy, we suspect that the 1994 import estimate in the CIR series is incorrect.

23 The Consumer Electronics Association also has printer sales estimates as far back as 1981.  Because the CEA data are proprietary and cannot be
 released to the public by EPA, we do not use the CEA estimates in this analysis.

-------
export data for 1989 through 1991 combine printers and plotters into a single category. To
estimate imports and exports of printers, excluding plotters, we multiply the CIR import and
export estimates for printers/plotters by the ratio of domestic printer shipments to total domestic
printer and plotter shipments.

The CIR series provides very limited information on keyboard sales. Sufficient data to estimate
keyboard sales are available from CIR only for the years 1994, 1996 through 1999, 2003, and
2004.  To estimate  1995 sales, we interpolate between our estimates for 1994 and 1996.
Similarly, we interpolate from 1999 and 2003 sales to estimate sales for 2000-2002. In the
absence of better data for keyboard sales, we estimate pre-1994 keyboard sales by assuming that
the growth rate for keyboard sales is the same as the growth rate for total desktop sales during
this period. We expand our estimates only back to 1986 because estimates for earlier years are
not necessary to generate complete keyboard retirement estimates for 2003 through 2005 (i.e.,
the years for which we assess e-waste generation and management in Chapter 4).

The CIR series contains even less information for computer mice, reporting U.S. shipments only
for 1997 and 1998. In the absence of better data for mouse sales, we use non-white-box desktop
sales as a lower-bound proxy for mouse sales between 1990 and 2004.24 For computer mice
sales prior to 1990, we do not use non-white-box computer sales as a proxy because many
computers sold during this period were not equipped with a mouse.  Instead, we assume that
mice sales were approximately the same as Apple Macintosh sales in the mid-1980s and grew
steadily through the rest of the decade as other computer makers began releasing computers with
graphical user interface (GUI) operating systems (e.g., AMIGA).25  Although we did not identify
estimates of Macintosh sales during the 1980s, an article published by Time magazine indicates
that Apple had a 14.6 percent share of the U.S. personal computer market in 1986.26 Based on
this information and our estimate of desktop sales in 1986, we estimate mouse sales of 1.1
million units for 1986.  To estimate 1987 to 1989 mouse sales, we interpolated between this
estimate and our 1990 estimate.
CELL PHONES

The INFORM report "Waste in the Wireless World: The Challenge of Cell Phones" presents
estimates of U.S. cell phone sales for 1995 through 1999 developed by Gartner.  Unlike the data
available from the Census Bureau's CIR series, these data represent sales, rather than domestic
24 The extent to which mouse sales exceed non-white-box sales is uncertain, but we believe non-white-box desktop sales represent a reasonable
 lower bound estimate because mice are packaged with most brand name computer systems.

25 Our research suggests that the first commercially viable computer that required a mouse was the Apple Macintosh released in the mid-1980s.
 (Sources: "OK, Mac, Make a Wish," Newsweek, February 4, 2004, 143(6): 41; "Apple Turnover," Time, October 2, 1995, 146(14): 56.); Mice did
 not become necessary on Microsoft-based systems until 1990 when Windows 3.0 was introduced, and companies began to adopt it as their main
 operating system.

26 "If you can't beat them...," Time, August 18, 1997, 150(7): 35-7.


                                                                                          18

-------
shipments, exports, or imports. Therefore, we directly use the INFORM data for 1995-1999 cell
phone sales.

To estimate cell phone sales for 2000 through 2004, we rely on a series of press releases issued
by Gartner between 2001 and 2005.  One of these releases issued in March of 2001 indicates that
Gartner projected 2001 North American cell phone sales of "90.1 million units, an 18 percent
decline over 2000."27  Based on this information, we estimated North American cell phone sales
of 76.4 million units in 2000 (90.1 million/1.18= 76.4 million). To estimate U.S. sales in 2000,
we assume that the ratio of U.S. to North American cell phone sales is the same as the ratio of
U.S. to North American  GDP.  For 2001 through 2004, Gartner's press releases report global cell
phone sales, but provide no information on North America's share of the global cell phone
market.  To estimate U.S. cell phone sales for these years, we assume that the growth rate of the
U.S. cell phone market between 2000 and 2004 mirrored that of the global cell phone market,
applying the global growth rate implied by the Gartner data to our estimate of U.S. cell phone
sales in 2000.
MASS DATA

In addition to sales data, we collected information on the average mass of each of the electronic
products included in this analysis. Together with the sales estimates presented above, this
information allows the waste flow model to estimate the total mass of electronics sold by
product. Exhibit 3-3  summarizes the product mass estimates we developed for the analysis.  We
derived these estimates based on data contained in publications from the Northeast Recycling
Council (NERC), the National Safety Council (NSC), the Cascadia Consulting Group, PCWorld,
EnviroSIS, RIS International, and ICF Consulting, as well as electronics collection data
compiled by the Florida Department of Environmental Protection (FLDEP) as part of its 2004-
2005 electronics sorting study.
27 Gartner, Inc. "Gartner Dataquest Says Worldwide Mobile Phone Sales to Surpass a Half Billion Units in 2001," press release March 20, 2001.

                                                                                       19

-------
EXHIBIT 3-3. PRODUCT MASS ESTIMATES
PRODUCT

Desktop Computers


Laptop Computers



CRT Monitors

LCD Monitors

Keyboards
Computer Mice

Desktop Computer Printers
Projection Televisions

CRT Televisions



Cell Phones


YEAR SOLD
1975-19893
1990-2005b
1990-2000C
2001 d
2002 d
2003-20046
2005f
1975-19893

1990-20058
1990-2005h
1975-19893
1990-2005'
1975-2005j
1975-19893
1990-2005k
1975-20051
1975-19793
1980-1 989 a
1990-20053
1990-1993m
1994-1996"

1997-1999m
2000-2005°
AVERAGE UNIT MASS (LBS.)
22.0
26.8
12.1
10.6
9.1
7.7
6.1
29.8

34.3
12.7
3.0
2.2
0.3
18.0
18.7
149.1
55.9
59.6
63.4
0.66
0.57

0.49
0.30
                                                        20

-------
PRODUCT
YEAR SOLD
AVERAGE UNIT MASS (LBS.)
Note:
* Although most of these sources provide estimates of the average mass of specific products, the source we used from the Florida
DEP presents the total number and mass of each device collected as part of its 2004-2005 electronics sorting study (e.g., 64
laptops with a total mass of 911 pounds).  Therefore, in those cases where we incorporate the Florida data into our average mass
calculations, we use the average mass per unit of each collected device (i.e., total mass collected divided by total number
collected).  Florida DEP, Florida Electronic Product Brand Distribution Project,
http://www.dep.state.fl.us/waste/categories/electronics/pages/  FloridaElectronicProductBrandDistributionProject.htm.
Sources:
a. Florida Department of Environmental Protection  (FLDEP), Florida Electronic Product Brand Distribution Project, 2004-2005.
  (Data accessed  on 19 December 2005.)
b. Average of mass values presented in Minnesota Office of Env. Assistance, "Brand and vintage analysis research project: 2004,"
  Aug 2004; Cascadia Consulting Group, "E-Waste Generation in NW Washington," 11/21/03; Northeast Recycling Council, Inc.
  (NERC), "Used Electronics Market Study Survey Analysis," August 2003; NSC, "Electronic Product Recovery and Recycling
  Baseline Report," May 1999; FLDEP data, as of 9/20/2005.
c. Average of mass values presented in National Safety Council (NSC), op.  cit.; FLDEP data, as of 9/20/2005.
d. Interpolated from 2000 and 2003-2004 estimates.
e. Average of mass values presented in Cascadia Consulting Group,  op. cit.; Minnesota Office of Env. Assistance, op. cit.; NERC,
  op. cit.
f. "A Walk on the Wide Side," PCWorld, May 2005.
g. Average of mass values presented in Caplan, Richard A., "Expanding and Developing Markets for Used and End-of-Life
  Electronics," Spring 2002; Cascadia Consulting Group,  op. cit.; U.S. EPA, Desktop Computer Displays: A Life-Cycle Assessment,
  EPA-744-R-01-004a,  December 2001; Franklin Associates, "Energy and Greenhouse Gas Factors for Personal Computers," 2002;
  Minnesota Office of  Env. Assistance, op. cit.; NERC, op. cit.; NSC, op. cit.;  FLDEP data, as of 9/20/2005.
h. Average of mass values presented in Cascadia Consulting Group,  op. cit.; U.S. EPA, op. cit.; RIS International, Ltd.,
  Information Technology (IT) and Telecommunications (Telecom)  Waste in Canada - 2003 Update, October 16, 2003.
i. Average of mass values presented in Minnesota Office of Env. Assistance, op. cit.; NERC, op. cit.; FLDEP data, as of 9/20/2005;
  PHA Consulting Associates,  Electronic Waste Recovery Study, 1 October 2004.
j. Average mass of the following five mouse devices, as indicated on manufacturer websites: Creative  Labs Creative Mouse
  Classic, Creative Labs Creative Mouse Wireless Optical, Microsoft Wireless Optical Desktop Pro Mouse, Sony PCGAWMS5S VAIO
  Wireless Optical Mouse, and Sony Optical USB Mousespacer SMU-CL2/L.
k. FLDEP data, as of 9/20/2005.
I. Average mass of 25 projection televisions from electronics  retailers and 12  projection televisions collected as part of the
  Florida Electronic Product Brand Distribution Project.
m. Average of mass values presented in Environment Canada, IT and Telecom Waste in Canada, Enviros RIS, October 2000.
n. Average of 1990-93  and 1997-99 estimates.
o. Average of mass values presented in Environment Canada, op. cit.;  "The Ultimate Wireless Buyers Guide," PC World, October
  2000.
PRODUCT LIFESPAN

As indicated in Chapter 2, the waste flow model that we developed for this baseline assessment
uses information on the lifespan of each product to estimate the volume of electronic products
retired each year.  This section summarizes our estimates of each product's first life (i.e., the
period of time that a product's first owner uses the product) and second life (i.e., the period of
time a product is reused or kept in storage after its first life, and prior to its EOL).


INITIAL  SERVICE LIFE

To estimate the initial service life for each product included in this analysis, we consulted
publications from INFORM,  Business Week, PC World, the National Safety Council (NSC),  and

                                                                                                            21

-------
EPA.  Although the NSC provides relevant information for desktop computers, laptops, CRT
monitors, and CRT televisions, we relied on more recent data sources where possible.  Exhibit 3-
4 presents our estimates of each product's initial service life.  We discuss our estimates for each
product in further detail below.
            EXHIBIT 3-4.  INITIAL  SERVICE  LIFE OF SELECT ELECTRONICS PRODUCTS
          PRODUCT
DURATION OF INITIAL SERVICE LIFE
            (YEARS)
          Desktop Computers3
          Laptop Computers'5
          CRT Monitors0
          LCD Monitors"
          Keyboards
          Computer Mice6
          Desktop Computer Printers'
          Projection Televisions8
          CRT Televisions8
          Cell Phonesh
        1978-1998: 2 to 4
       1999-2004: 3.3 to 4
             2 to 3
               4
             3 to 8
   Same as Desktop Computers
             1 to 5
             3 to 6
            7 to 13
            7 to 13
           1.5 to 2.5
          Sources:
          National Safety Council (NSC), Electronic Product Recovery and Recycling Baseline Report, May 1999;
          "Dell, the Conqueror," BusinessWeek, 24 September 2001; Gartner Press Release, "Gartner Says
          Extending the Life Cycle of Desktop PCs Won't Necessarily Save Money on Total Ownership," 15
          September 2003.
          "The PC Replacement Decision," Information Week, 20 June 2005; Texas Department of Information
          Resources, PC Life Cycles: Guidelines for Establishing Life Cycles for Personal Computers, January 2003.
          c. NSC, op. at.
          d. "Shedding Some Light on LCDs," PCWeek, 5 October 1998; "LCD Monitors: Light, Slight, and Stylish,"
          PCWorld, August 1999; "Is It Time to go LCD?," Home Office Computing, November 2000.
          e. "Mouse Scurries Toward Future," Dell "Browser" Magazine, Spring 2000, as cited by lightglove.com.
          f. U.S. EPA, Municipal Solid Waste in the United States: 2001 Facts and Figures, October 2003; Cardwell,
          Annette, "The Paperless Office?" Ziff Davis Smart Business, 14 (Dec2001/Jan2002).
          g. U.S. EPA, op. cit.; Oregon Advisory Committee on Electronic Product Stewardship, Report to the 2005
          Oregon Legislature, January 2005.
          h. U.S. EPA, op. cit.; "When Phones Go Bad," Washington Post, 10/31/2004; and "Handsets: Catching
          Customers with Color,"  Wireless Week, 1/1/2003.
                                                                                                       22

-------
Desktop computers

In its 1999 report, the NSC estimates that the initial service life of desktop computers (e.g., 386,
486, Pentium I, and Pentium II) ranged from two to four years,28 and we used this estimate for
all desktop computers sold between 1978 and  1998.  However, a different estimate is provided
on the initial service life of desktop computers sold between  1999 and 2005 based on three
service life estimates generated by Gartner.  A 2001 BusinessWeek article cites a Gartner study in
which the lifespan of a corporate PC in 1999 was estimated to be 3.3 years.29 In addition, a 2003
Gartner press release recommends a desktop life cycle of four years, while an article published in
InformationWeek in 2005 references a Gartner study estimating that businesses replace desktop
computers every 43 months (3.6 years).30'31  Because all three of these sources refer to
replacement cycles for desktop computers' original users, they are reasonable data sources to use
in estimating the initial service life of a desktop computer. Based on these three studies, we
assume that the initial service life of a desktop sold between  1999 and 2005 is between 3.3 and
four years.

Laptop Computers

We estimate the duration of a laptop's initial service life based on data from Gartner and the
Texas Department of Information Resources.32  Based on a survey of large businesses
conducted by Gartner, Information Week reports that mobile  PCs are replaced every 36 months.
Similarly, the Texas Department of Information Resources indicates that the industry standard
for replacing a laptop computer is two to three years, citing a 2001  Gartner Research Note.
Based on these data sources, we assume that the duration of a laptop's initial service life is two
to three years.

CRT  Monitors

In its 1999 report on electronics recycling, the NSC estimated that the duration of a CRT
monitor's initial service life is four years.33  We use this estimate for all CRT monitors sold
between 1978 and 2005.
 National Safety Council, Electronic Product Recovery and Recycling Baseline Report, May 1999.

29 "Dell, the Conqueror," BusinessWeek, September 24, 2001.

30 "The PC Replacement Decision," Information Week, June 20, 2005.

31 Gartner Press Release, "Gartner Says Extending the Life Cycle of Desktop PCs Won't Necessarily Save Money on Total Ownership," September 15,
 2003.

32 Information Week, Ibid.; Texas Department of Information Resources, PC Life Cycles: Guidelines for Establishing Life Cycles for Personal
 Computers, January 2003.

33 National Safety Council, op. cit.


                                                                                           23

-------
LCD Monitors

Our survey of the relevant literature revealed three sources containing information on the
average service life of an LCD monitor.34'35 These sources present service life estimates between
three and eight years. Two of the three lifespan estimates reflected in this range are based on the
expected  useful life of an LCD display.  In the absence of data describing how long the first user
of an LCD monitor keeps the display, we assume that the first user of an LCD monitor keeps it
for its entire useful life.36 Therefore, we assume that the initial service life of all LCD monitors
sold between 1990 and 2005 is three to eight years.

Keyboards

We located no service life information for keyboards. Absent such information, we assume that
the initial service life of a keyboard is the same as that of a desktop computer, with which it is
likely sold.

Computer Mice

The only  source we identified with information on the service life of a computer mouse was an
article published in a 2000 issue of Dell Browser Magazine.37 This article indicates that the
typical computer mouse has a useful life of one to five years, but provided no explanation as to
how this estimate was developed. Because we located no additional sources with service life
estimates for mice, we use this one-to-five year range as our estimate for the length of a mouse's
first life.

Desktop Computer Printers

We identified two sources of information on the service life  of a computer printer. EPA's report
on municipal solid waste (MSW) generation and management in 2001 states that a printer's total
life expectancy (i.e., including both primary and secondary (or reuse) life) is approximately three
to five years.38  We assume that the low end of this range represents the length of a printer's
initial service life. In addition, an article published in Ziff Davis Smart Business in 2001 states
that the average  lifespans of inkjet and laser printers are three years and six years, respectively,
34 The LCD monitors that we refer to include stand-alone monitors, but not laptop computer screens.

35 The three articles that mention LCD monitor lifespan are: "Shedding Some Light on LCDs," PCWeek, 10/5/1998; "LCD Monitors: Light, Slight, and
 Stylish," PCWorld, August 1999; "Is It Time to go LCD?," Home Office Computing, November 2000. PCWeek (1998) provides a LCD lifespan of 5 to
 8 years. PCWorld (1999) and Home Office Computing (2000) report service lives of 3 to 6 years and 6 years, respectively.

36 Because we assume that the first owner of an LCD monitor keeps it for its entire useful life, we also assume that all LCD monitors not retired
 after their first life spend their second life in storage.

37 "Mouse Scurries Toward Future," Dell "Browser" Magazine, Spring 2000, as cited by lightglove.com.

38 U. S. EPA, Municipal Solid Waste in the United States: 2001 Facts and Figures, October 2003.


                                                                                              24

-------
according to information released by Gartner and Hewlett-Packard.39  Based on these data
points, we assume that the initial service life of a printer is between three and six years.

Cell Phones

We identified three data sources with information on the service life of a cell phone.40 EPA's
2001 MSW report states that the life expectancy of a wireless telephone is between two and four
years, including primary and secondary use. We assume that the low end of this range reflects
the duration of a cell phone's first life. A 2004 Washington Post article cites research from the
Yankee Group indicating that the average life cycle of a phone was 25 months in 2001, but that
by 2004, this average had fallen to  19.4 months.  Although this study suggests that the average
first life of a cell phone has fallen in recent years, Nokia has estimated that, on average, cell
phones are replaced every 2.5 years.41 Based on these data sources, we estimate an initial
lifespan range between 1.5 and 2.5  years.

Televisions

The research revealed two sources of service life data for televisions.42 EPA's report on MSW
generation and management in 2001 states that a television's life (primary and secondary lives
combined) ranges from 13 to 15 years for direct view color TVs, projection TVs, and LCD color
TVs. We assume the low end of this range represents the initial  service life of a television.  In
addition, the Oregon Advisory Committee on Electronic Product Stewardship (2005) estimated
that the lifespan of a television is seven years.  Our own experience suggests that most
televisions are functional for much longer than seven years; therefore, we assume that this
estimate reflects the length of a television's initial service life. Based  on these two sources, we
estimate that the duration of a television's initial  service life is seven to thirteen years. In the
absence of data distinguishing between CRT and projection models, we assume that this
represents the initial service life range for both CRT and projection televisions.
INITIAL SERVICE LIFE PROBABILITY DISTRIBUTION

As indicated in Chapter 2, the waste flow model requires information on the distribution of the
initial service life expressed as a range. In such cases, we assume a discrete approximation to a
triangular distribution (i.e., a bell-shaped distribution) for initial service life ranges spanning
more than two years. For an initial service lifespanning two years, we assume a uniform
39 Cardwell, Annette, "The Paperless Office?" Ziff Davis Smart Business, 14 (Dec2001/Jan2002).

40 The three studies are U. S. EPA, op. cit.; "When Phones Go Bad," Washington Post, 10/3172004; and "Handsets: Catching Customers with Color,"
 Wireless Week, 1/1/2003.

41 Wireless Week, op. cit.

42 U. S. EPA, op. cit.; Oregon Advisory Committee on Electronic Product Stewardship, Report to the 2005 Oregon Legislature, January 2005.


                                                                                          25

-------
probability distribution (i.e.,  50 percent each year).  Exhibit 3-5 summarizes the probability
distributions used in the model for this baseline assessment for each product's initial service
life.43
  For products with an initial service life not expressed as integers, we developed service life probability distributions such that the expected value
 for each product's initial service life equals the average of the high- and low-ends of its service life range. For example, the initial service life for
 desktops sold in 1999 is 3.3 to 4.0 years. Therefore, we assume that 35 percent of desktop computers will be retired after three years and that
 the remaining 65 percent will be retired after four years. Based on these probability values, the expected value of the initial service life of a
 computer sold in 1999 is  3.65 years (0.35x3 + 0.65x4 = 1.05 + 2.6 = 3.65 years).
                                                                                                                     26

-------
CO
CO
<
CO
O
a:
a.
a:
LU
l/l
CO

E
X



_i
LU
O




Lrt
PHONE:


LU
O
Z
Q.


Z
o

in
PROJEC


to
I-
o:
o


o

o





Q_
g
0.
5




SO
LU

Q.
O
Q.
O
£
to
LU
0


z
i
<
LU







S







MONITORS

MONITORS




^5
UJ
O
O






.Q
2
2


z
O
P
3
y
O


3j"
° S? S? S? 	
• in o in ! ! ! !
g ts in ts
^


^.
o
(S| xp S? S? S? xp
)i K o o o go
<£ ^ ts m ts ""



s
Q
^ 1 1 ^~ ^~ ^~ ^~ 1 1
i ' ' "" ' •



•^-


N 	 S? r§
in ' ' ' ' ' ' in ^i


o
S 	 55 S?
in ' ' ' ' ' ' in ^



0
S xp s? s? s? s? xp
6 : : in £ ° ° £ Lo

s
s , , , g , , , ,
Q ' ' ' O ' ' ' '





T
o
° , s? s? 	
i ' °> °> 	




s
° s? s?
1 ! ! in in ! ! ! !
I ro so
CO
j! ! LO o LO ! ! ! !
« ts in ts


Q
O
Op
UJ


to

8
(N


to
S
8
ro



to
JS
	 >,
8
ro


to

(0
0)
°1 ro °1 ° ^ g
O


to
(0
0)
^i /^i ^i ^^ m f~^
CM ro CM -r- 0
to


(0
; ; ; ; ; >.
O
in
in

to
	 >,
8





(0
0
in
(N


to
(0
01

in
ro
to
%
01
8
ro

(0
"a ^

^~ ^~ ^~ ^~ Q_ U
LU C
*
ro M

ro j] so
^ .S2 ti (N
U (Q +
s t?"§g
O x *•"" "
u-^T
1 ~° -Q x
-n « S K2
lt*s
8.1 § A
S "O £1 x
O> j_ gj -^
(u o Q. in
!= i- ro
« 0)
1o = on't'
fJs C <_'
^ .y ^E «
D (u E m
to w 0) ^O
s-s S«
.2 43 ^ .to
J-J '^ ^^
^ _ i ' O^
_Q ., (0 O^
•i"" UJ (— O\
2 " -a c
T3 0) C '^
>--Q. *2
5 S S '-
^ QJ ^* 4J
"o i-  QJ
J-J Q ^ QJ i- >^
z| Sf "s
-S .-9 '5 -B «
s°l'E
_i2 ••-' QJ
e c as
01 01 E „_
Efi 8 °
4J 0 U 0)
C — Q. D
a "2S g
« •«-^
Jj -c -a S

S !E ^ oi
Q. 0) •" x"
X ^ S 0)
0) 4-< KU
^^5^1
C QJ Q. ^
(D 2P LO }^
^^ Qj j-J •
Qj >• (0 (0
y ns _c >

QJ 5 C ^

Is "« 1.2
if2&
C  .^ v*~
3 «||
D — H ^
"O (0 (Q
g B jo OQ
CL.E (0 .
££o i
(Q D -^ >>
ES =
Q. ^4-
Q.

^H
^
to
-8
(0
V4—
O
1
4-1
same as
omputer
0) U
-^ Q.
"^ O
T3 "«
L_ QJ
(0 "O
f'S
to «
<4— J-<
0 to
QJ (0



1 i
C 0)
Oi £
to 4J
-- 0)
<0 ^2
•"t^ ~~^
OI T— I
^ (5
-M o
ns _Q
£ S^
0) -*
E «
D s_
:ause we ass
stribution fo
(->
keyboards 1
probability
i- 0)
£ =
C a,
.2 ."
j-> ^
D C
_Q QJ
'C w
j-j ^
to (0
^ 5
c c
ns "~
Q. (U
to -C

'•= oT
J-l S-
to O

c£
s^
£ H
S-b
« g.
1 °
Q.

'o
0
0)

^

-------
SECOND SERVICE LIFE DURATION AND  PROBABILITY DISTRIBUTION
To capture the storage and reuse of electronic products, the waste flow model simulates waste
management decisions for such products at two points in time: the end of a product's first life
(i.e., the period of time during which its original owner uses it on a regular basis) and the end of
its second life (i.e., the period of time between the end of its first life and the end of its total life).
Our approach for estimating the probability that a device will reach a second life is presented
below. To determine the duration of each product's second life, data was used on the age of
electronics collected for recycling as part of the Florida Electronic Product Brand Distribution
Project (the Florida Project) during the 12-month period beginning in April 2004. Based on
these data, a distribution was developed of the length  of time each product remains in circulation
before it is disposed of or recycled.  Adjusting these distributions based on the procedure
outlined in Appendix B, a distribution was developed for the second life of each product
included in the analysis, as  shown in Exhibit 3-6.

Because the Florida data do not include information for computer mice and cell phones, the
second lifespan distributions was used for keyboards as a proxy for computer mice and of
laptops as a proxy for cell phones. In addition, as the results in Exhibit 3-6 suggest, we do not
estimate a distribution for the second life of projection televisions. The Florida collection data
and our estimate of a projection television's first life suggest that  almost no projection televisions
go into storage or are re-used.  Therefore, all projection televisions are assumed to be recycled or
disposed of at the end of their first life.  Exhibit 3-7 combines the first life information in Exhibit
3-4 and the second lifespan ranges included in Exhibit 3-6.

-------
EXHIBIT 3-6. DISTRIBUTION OF THE SECOND LIFESPAN OF SELECT ELECTRONICS PRODUCTS1
YEARS IN
CIRCULATION
AFTER FIRST LIFE
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
TOTAL
DESKTOP
COMPUTERS
0.5%
1 .3%
2.0%
3.4%
4.8%
2.2%
3.3%
6.6%
2.8%
5.3%
4.0%
5.8%
4.2%
7.0%
4.4%
2.1%
4.0%
6.6%
2.9%
9.6%
9.6%
5.5%
2.1%
100%
LAPTOP
COMPUTERS
9.8%
9.2%
22.1%
18.4%
40.5%
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%
0.0%
0.0%
100%
CRT
MONITORS
3.7%
5.4%
7.4%
9.3%
7.1%
8.0%
10.0%
10.0%
9.1%
4.9%
4.7%
2.8%
4.0%
2.3%
1 .4%
2.0%
1.1%
1 .5%
2.7%
0.9%
0.6%
1 .0%
0.0%
100%
LCD
MONITORS
16.7%
16.7%
16.7%
16.7%
16.7%
16.7%
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%
0.0%
100%
CRT TVS
9.2%
6.8%
9.4%
7.3%
9.0%
6.7%
6.8%
5.9%
3.8%
5.7%
6.4%
4.6%
4.7%
3.9%
2.6%
2.3%
2.5%
1 .4%
1 .0%
0.0%
0.0%
0.0%
0.0%
100%
PROJECTION
TVS
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%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
PRINTERS
5.0%
6.8%
8.0%
6.4%
8.6%
13.3%
7.7%
10.2%
9.3%
7.1%
4.5%
3.8%
3.1%
2.7%
1.3%
1.1%
0.0%
1 .0%
0.0%
0.0%
0.0%
0.0%
0.0%
100%
KEYBOARDS
38.9%
16.5%
4.1%
0.0%
2.5%
2.8%
3.2%
0.0%
1 1 .5%
0.0%
14.6%
6.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
100%
COMPUTER
MICE2
38.9%
16.5%
4.1%
0.0%
2.5%
2.8%
3.2%
0.0%
11.5%
0.0%
14.6%
6.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
100%
CELL
PHONES3
9.8%
9.2%
22.1%
18.4%
40.5%
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%
0.0%
0.0%
100%
Notes:
The distributions presented in this exhibit apply only to products not disposed of after their first life.
We use the distribution of the duration of a keyboard's second life as a proxy for mice.
We use the distribution of the duration of a laptop computer's second life as a proxy for cell phones.

-------
               EXHIBIT 3-7. PRODUCT LIFESPANS USED IN WASTE FLOW MODEL
PRODUCT
Desktop Computers
Laptop Computers
CRT Monitors
LCD Monitors
CRT Televisions
Projection
Televisions
Desktop Computer
Printers
Keyboards
Computer Mice
Cell Phones
LENGTH OF FIRST LIFE
1978-1998: 2 to 4 years
1999-2005: 3 to 4 years
2 to 3 years
4 years
3 to 8 years
7 to 1 3 years
7 to 1 3 years
3 to 6 years
1978-1998: 2 to 4 years
1999-2005: 3 to 4 years
1 to 5 years
1 to 3 years
LENGTH OF
SECOND LIFE
Up to 23 years
Up to 5 years
Up to 22 years
Up to 6 years
Up to 19 years
0 years*
Up to 18 years
Up to 12 years
Up to 12 years
Up to 5 years
LENGTH OF TOTAL LIFE
1978-1998: 2 to 27 years
1999-2005: 3 to 27 years
2 to 8 years
4 to 26 years
3 to 1 4 years
7 to 32 years
7 to 1 3 years
3 to 24 years
2 to 1 6 years
1 to 1 7 years
1 to 8 years
Note:
* Products disposed of after their first life have a second life of zero years.
END-OF-LIFE  MANAGEMENT

Based on the sales, mass and lifespan data presented above, we estimate the quantity of e-waste
generated in the U.S. each year.  To inform EPA program and policy development, it is equally
important to characterize how this waste is managed at the end of its life. The management
options for e-waste include the following:44

Landfilling: Electronic waste may be placed in a landfill for final disposal.

Incineration: Electronic waste may be burned at an incineration or waste-to-energy
facility.

Recycling: Electronic products may be recovered for the purpose of dismantling, parts and/or
materials recovery, and/or resale (resale that occurs by a recycler and not by the user of the
product).
44 Open burning is another potential waste management option for electronic waste, but we Open burning may occur in relatively low population
 density areas of the United States; open burning of waste is usually banned in high population density areas. We assume that the quantity of
 electronic waste managed through open burning is minimal.  (Outside the United States, open burning of electronic waste may occur.)

-------
Exportation: Transport of electronic products outside U.S. borders for re-use, refurbishing,
recycling or disposal.

Re-use/Storage: When a product reaches the end of its first life, it is not necessarily
disposed or recycled at that time. The product's original owner may donate or sell the
product or keep the product in storage, even though he or she no longer uses it.45

All five of these management methods apply to devices at the end of their first life (i.e., after its
first owner no longer uses it on a regular basis).  When devices reach the end of their second (and
final) life, only the first four of these methods apply.

The waste flow model that was developed for this baseline assessment allocates e-waste to the
various management options outlined above based on information from several waste
characterization studies and  e-waste management surveys. In this section, we describe the
approach for using these data to distribute the e-waste generation estimates across the various
management options outlined above.
END-OF-TOTAL-LIFE E-WASTE MANAGEMENT

In this section, we describe the approach for apportioning electronic products reaching the end of
their total life (i.e., electronic products at the end of their second life and devices at the end of
their first life that are not re-used or placed in storage) to the various management methods
identified above. For each device, we estimate a series of allocation weights—one for each
management method—that indicate how the retirement of a given product is distributed across the
various management methods. For example, the desktop  computer allocation weight for
landfilling is 71.8 percent, which indicates that we allocate 71.8 percent of retired desktops to
landfilling.

As indicated above, the allocation weights presented in this section correspond to the
management of electronic products at the end of their total life.  Although it is also important to
develop allocation weights for such products at the end of their first service life, we begin with
the allocation weights specific to electronic products at the end of their total life because our end-
of-first-life allocation weights are based on the allocation weights presented in this section.

To estimate the allocation weights for each electronic product, we rely heavily on the results of
five state-specific waste characterization studies that included detailed information on e-waste
discards.46 Based on the per capita e-waste discard rates derived from these studies, the
45 A second or third user could store the product as well.

* These studies are as follows: Oregon Department of Environmental Quality, 2002 Oregon Solid Waste Characterization and Composition, April 20,
 2004; Cascadia Consulting Group, Statewide Waste Characterization Study, prepared for California Integrated Waste Management Board,
 December 2004; Cascadia Consulting Group, Wisconsin Statewide Waste Characterization Study, prepared for the Wisconsin Department of
 Natural Resources, May 2003; Georgia Department of Community Affairs, Georgia Statewide Waste Characterization Study, June 22, 2005; and

                                                                                           31

-------
electronics retirement estimates generated by the waste flow model, and information from other
sources, we estimate device-specific allocation weights for each end-of-life management option.
The calculations we performed to generate these estimates are as follows:

1.   Adjust the waste characterization results to account for interstate import and export of
    electronic waste.  The results of most of the waste characterization studies that were
    identified reflect movement of electronic wastes between states (interstate import and export;
    as distinguished from import and  export to and from abroad).  Each state's exports are
    excluded from its results, but its imports are included. Therefore, per capita e-waste discard
    rates from these studies may not be representative of the e-waste discard rate for the entire
    U.S. (i.e., if a state is a major net importer of e-waste, its per capita discard rate would not
    accurately reflect the volume of e-waste discarded per capita in the U.S.). To address this
    issue, the results  of these studies were adjusted to include each state's exports of e-waste to
    other states and exclude its e-waste imports from other states.47

2.   Remove electronic products not included in our analysis from the waste characterization
    results.  The waste characterization studies that were identified provide discard estimates for
    a limited number of broad product categories. Many of these categories contain devices not
    included in our analysis (e.g., the  computer-related electronics  category in one of these
    studies includes modems and fax  machines).  To adjust the waste characterization results to
    reflect only those computer devices included in our analysis, we assumed that the
    composition of each state's electronic waste is consistent with the composition of the
    electronic waste collected for recycling through the Florida Project.48 Unlike the state waste
    characterization studies, the Florida Project compiled  product-specific data on the mass of
    the electronic products collected.  Therefore, if scanners represent 5 percent  of the total mass
    of computer products collected through the Florida Project, we subtract five  percent of the
    total computer product mass estimated in the state waste characterization studies. Based on
    these adjustments, the volume of e-waste discards were  estimated for three categories of
    products: computer equipment (i.e., desktop computers,  laptop computers, printers,
    keyboards, and mice), CRT monitors, and CRT televisions.49'50

 Minnesota Pollution Control Agency and Minnesota Office of Environmental Assistance, Statewide MSW Composition Study, March 2000. The
 California, Oregon, and Minnesota studies are based on  a representative sample of disposal facilities in each state, while the Georgia and
 Wisconsin studies are based on a sample of the largest landfills in each of these two states. Therefore, the results of the California, Oregon, and
 Minnesota studies may more accurately reflect the composition of waste collected for landfilling or incineration.
47 We obtained interstate export and import data from James E. McCarthy and Anne L. Hardenbergh,  Congressional Research Service, Interstate
 Shipment of Municipal Solid Waste: 2002 Update, November 26, 2002.
48 Florida Department of Environmental Protection, Florida Brand Distribution Project,
 http://www.dep.state.fl.us/waste/categories/electronics/pages/FloridaElectronicProductBrandDistributionProject.htm.
49 Some of the waste characterization studies include CRT monitors as a separate category, while others include a category for monitors in general.
 Because the waste flow model estimates that the mass  of LCD monitors retired in recent years has been less than 1.3 percent of the total mass of
 retired monitors, we assume that all of the monitors reflected in the waste characterization studies are CRT monitors.
50 Some of the waste characterization studies include a general television category, while others include a category specific to CRT televisions. To
 estimate CRT discards from the studies that do not distinguish between CRT TVs and non-CRT TVs, we multiplied the general television discard
 estimates by the ratio of CRT retirements to total TV retirements (by weight), as estimated by the waste flow model.
                                                                                                 32

-------
    Estimate the e-waste discards per capita for each electronic product category. Based on the
    e-waste discard estimates generated in step 2 and the population data from the Census Bureau
    for the five states that conducted waste characterization studies , state-specific estimates were
    generated for the e-waste discards per capita for each of the three product categories listed
    above, as presented in Exhibit 3-8.
         EXHIBIT 3-8.  ELECTRONIC WASTE  DISCARDS  PER CAPITA PER YEAR  BY STATE
       STATE
COMPUTER EQUIPMENT"
(TONS DISPOSED OF PER
       CAPITA)
CRT MONITORS
(TONS DISPOSED
OF PER CAPITA)
                                                                  CRT TELEVISIONS
                                                                (TONS DISPOSED OF PER
                                                                      CAPITA)
       Minnesota
       Wisconsin
       Oregon
       California
       Georgia
       Weighted Average*5
     2.271 x10
     4.983x10'
     1.565x10:
     2.765 x 10:
     4.122x10'
     1.396x10
  2.839x10'
  1.859x10:
  1.187x10:
  2.401 x10:
  4.716x10'
  1.036x10
                                                                     2.585 x 10 4
                                                                     4.028 x 103
                                                                     2.135 x103
                                                                     4.108 x 103
                                                                     3.417 x 10 4
                                                                Low End: 1.820x103
                                                                High End: 3.204 x 103
Notes:
a. Computer equipment includes desktop computers, laptop computers, desktop computer
printers, keyboards, and computer mice.
b. Calculations to estimate the weighted averages exclude the highest and lowest values in each
column. For CRT televisions, two weighted averages were estimated because of uncertainty in
the television disposal data. The lower of the weighted averages for televisions is based on data
for Wisconsin, Oregon, and Georgia, while the higher estimate is based on data for California,
Oregon, and Georgia.
4.  Estimate the population-weighted e-waste discards per capita.  Using the state-specific per
    capita e-waste discard estimates generated in step 3 and the Census Bureau's population
    estimates for each state, population-weighted e-waste discards per capita were calculated.
    Because we do not know whether the five states listed in Exhibit 3-8 are representative of the
    entire U.S. with respect to e-waste disposal, the states with the highest and lowest discards
    per capita were excluded from our weighted average calculations.51  For example, California
    and Georgia were excluded from our calculations for the weighted average of computer
    equipment discards per capita. Exhibit 3-8 presents weighted average discards per capita for
    each of the product categories defined above.
51 For CRT televisions, discards per capita are slightly higher in Wisconsin than in California. If we exclude Wisconsin from our analysis, our
 population-weighted estimate of television discards per capita implies that total television discards are greater than the television retirement
 estimates generated by the waste flow model. However, if we exclude California instead of Wisconsin, the resulting estimate of television
 discards per capita yields total discard estimates significantly less than the retirements estimated by the model. To address this inconsistency,
 parallel analyses were conducted of television discards per capita-one excluding the Wisconsin data and one excluding the California data. We
 combine the results of these two analyses to estimate allocation weights for televisions.
                                                                                                 33

-------
5.   Estimate the Discard Rate for Retired Electronic Products. Multiplying the population-
    weighted annual discards per capita estimates generated in step 4 by the U.S. population,
    total discards were estimated for each product category for 2003 and 2004.52 For example,
    based on the annual  estimate of computer equipment discards per capita presented in Exhibit
    3-8, we estimate that 406,100 tons of computer equipment were discarded in 2003 and
    410,100 tons in 2004, or 816,200 tons for both years combined as indicated in Exhibit 3-9.
    After estimating the 2003 and 2004  (combined) discards for each product category, these
    estimates were divided by the 2003  and 2004 retirement estimates generated by the waste
    flow model, which yields discard rates for each product category (i.e., the percent of retired
    electronics that are either landfilled  or incinerated), as shown in Exhibit 3-9. Continuing
    with our example for computer equipment, the waste flow model estimates that
    approximately 1.1 million tons of computer equipment were retired in 2003 and 2004.
    Dividing the 816,200 tons in computer equipment discards in 2003 and 2004 by this value,
    we estimate that approximately  73.9 percent of retired computer equipment is  discarded (i.e.,
    deposited in a landfill or incinerated) rather than recycled. Because  discards include waste
    that is either landfilled  or incinerated, the discard rates presented in Exhibit 3-9 represent the
    sum of our end-of-total-life allocation weights for landfilling and incineration.
52 We chose to estimate discards for these two years because the waste flow model's retirement estimates for these two years are more reliable
 than for earlier years (i.e., we lack sufficient sales data to generate complete e-waste retirement estimates for earlier years). We obtained this
 information for Oregon, California, Wisconsin, and Georgia: Georgia Department of Community Affairs, Georgia Solid Waste Management Report
 2004; Oregon Department of Environmental Quality, "2003/2004 Disposal Status, State of Oregon: Oregon DEQ 2004 Solid Waste Report to the
 Legislature," 2004; Wisconsin Department of Natural Resources, A Study of the Future of Solid Waste Management: A Report to the Wisconsin
 Legislature, January 2001; Enviros, Recycling Achievement in North America, 2000. We were not able to obtain comparable data for Minnesota.
 U.S. Population data is from: Table 1. Annual Estimates of the Population for the United States and States, and for Puerto Rico: April 1, 2000 to
 July 1, 2004 (NST-EST2004-01) Source: Population Division, U.S. Census Bureau.  Release Date: December 22, 2004. We excluded the data for
 Puerto Rico from the analysis.


                                                                                                 34

-------
           EXHIBIT 3-9.  ESTIMATION OF DISCARD RATE FOR RETIRED ELECTRONICS3

DEVICE CATEGORY

Computer Equipment15
CRT Monitors
CRT Televisions0
TOTAL DISCARDS
2003 AND 2004
(THOUSANDS OF
TONS)
816.2
605.5
Low End: 1,063.4
High End: 1,872.4
TOTAL RETIREMENT
2003 AND 2004
(THOUSANDS OF
TONS)
1,104.6
802.5
1,451.4

ESTIMATED DISCARD RATE

73.9%
75.5%
Low End: 73.3%
High End: 100%
(assumed discard rate: 86.6%)
    Notes:
    a. Although this exhibit presents estimated discard rates for 2003 and 2004, these estimates are based on
    estimates for the years 1999 through 2004.
    b. Computer equipment includes desktop computers, laptop computers, printers, keyboards, and mice.
    c. For CRT televisions, two estimates were presented for total discards and the discard rate to reflect
    uncertainty in the television discard data.  The lower of total discards and the discount rate is based on
    data for Wisconsin, Oregon, and Georgia, while the higher estimates are based on data for California,
    Oregon, and Georgia.  For the purposes of our analysis, we use the average of the two discard rates (86.6
    percent).
    d. The high end CRT television discard estimate exceeds the retirement estimate, and thus subtracting
    the discards from the retirement would result in a negative number.  See the discussion in the text.
    Estimate the End-of-Total-Life Allocation Weights for Landfilling and Incineration. To
    divide discarded e-waste between landfilling and incineration, estimates of the percentage of
    municipal solid waste landfilled and the percent incinerated for four of the five states with
    waste characterization data were obtained.53 Applying these percentages to the e-waste
    disposal estimates for these states, derived from the states' waste characterization studies, the
    volume of electronic waste incinerated and landfilled in each state were estimated.  Summing
    these results across all four states, we estimate that approximately 97.2 percent of discarded
    (i.e., not recycled) electronic waste is landfilled, while the remaining 2.8 percent is
    incinerated. Multiplying these values by the discard rates in Exhibit 3-9, we estimate the
    landfilling and incineration allocation weights presented in Exhibit 3-10 for computer
    equipment, CRT monitors, and  CRT televisions.
53 We obtained this information for Oregon, California, Wisconsin, and Georgia: Georgia Department of Community Affairs, Georgia Solid Waste
 Management Report 2004; Oregon Department of Environmental Quality, "2003/2004 Disposal Status, State of Oregon: Oregon DEQ 2004 Solid
 Waste Report to the Legislature," 2004; Wisconsin Department of Natural Resources, A Study of the Future of Solid Waste Management: A Report
 to the Wisconsin Legislature, January 2001; Enviros, Recycling Achievement in North America, 2000. We were not able to obtain comparable
 data for Minnesota.
                                                                                                  35

-------
7.  Estimate the End-of-Total-Life Allocation Weights for Recycling. Any electronic waste
   that is not landfilled or incinerated at the end of its total life is likely recycled (either in the
   US or by export to another country for reuse, refurbishment or recycling, which may include
   some disposal of residuals). Therefore, the end-of-total-life recycling allocation weights for
   each device were estimated as follows.

       R/EP= I-ZP-/P

where, R/Ep = The end-of-total-life recycling/export allocation weight for product P;

       LP = The landfill allocation weight for product P, and

       /P = The incineration allocation weight for product P.
Following this methodology, we estimate the recycling allocation weights presented in Exhibit 3-
10. Although estimating the volume of electronic waste recycled in the US separately from the
amount exported would be useful for policy planning and program development purposes, we
were unable to identify adequate data to estimate the two separately. Therefore, we combine all
waste not discarded (i.e., landfilled or incinerated) into a single recycling category (which
necessarily includes some export which we have not quantified in this analysis).
   EXHIBIT 3-10. END-OF-TOTAL-LIFE ALLOCATION WEIGHTS FOR COMPUTER EQUIPMENT,
                            MONITORS, AND CRT TELEVISIONS
PRODUCT CATEGORY
Computer Equipment
CRT Monitors
CRT Televisions
LANDFILL
ALLOCATION WEIGHT
71.8%
73.4%
84.2%
INCINERATION
ALLOCATION WEIGHT
2.0%
2.1%
2.4%
RECYCLING
ALLOCATION WEIGHT
26.1%
24.5%
13.4%
As indicated above, the computer equipment product category listed in Exhibits 3-8 through 3-10
includes desktop computers, laptop computers, printers, keyboards, and mice. Therefore, the
computer equipment allocation weights presented in Exhibit 3-10 were applied to all five of
these products.  In addition, because the waste characterization studies identified do not contain
information on LCD monitors or projection televisions, the allocation weights for CRT monitors
were used as a proxy for LCD monitors and the allocation weights of CRT televisions as a proxy
for projection televisions. Exhibit 3-11 presents the estimated allocation weights for each
product. These results suggest that consumers and businesses are more likely to recycle their
computers, monitors, and peripheral equipment  than their televisions. Consequently, we
                                                                                      36

-------
estimate that the landfill and incineration rate for televisions is higher than it is for computers
and related products.

The waste characterization studies that were used to estimate allocation weights for computer
equipment, monitors, and televisions do not contain detailed data on discards of cell phones.
Therefore, to estimate the allocation weights for cell phones presented in Exhibit 3-11, the results
of the Consumer Electronics Association's (CEA's) household e-waste management survey and
data from INFORM's Calling All Cell Phones report published in 2004 were used.54  The results
of these two sources,  however,  reflect the high level of uncertainty on the percentage  of cell
phones that are recycled.  Based on the INFORM report and our cell phone retirement estimates,
the recycling rate for  cell phones is relatively low (approximately 2.3 percent in 2002).  The
results of the CEA survey, on the other hand, indicate that the recycling rate may be much higher
(26 percent). As a result, to estimate a recycling rate for inclusion in the waste flow model, we
calculated the recycling rate in  2004 inferred by the INFORM- and CEA-derived recycling rates
for 2002 and 2005, assuming that the recycling rate grew at a constant rate between these two
years.55

Based  on this  approach, a cell phone recycling rate of 19.2 percent was estimated, as indicated in
Exhibit 3-11.  We assume that the remaining 80.8 percent of cell phones are discarded.  As
indicated above, we estimate that 97.2 percent of discarded electronic waste is landfilled, while
the remaining 2.8 percent is incinerated.   Applying these values to our estimate of the
percentage of cell phones discarded, 78.5  percent of cell phones are estimated to be landfilled
and 2.2 percent are incinerated.
M Consumer Electronics Association, Consumer Electronics Reuse and Recycling, October 2005, and Eric Most, Calling All Cell Phones, INFORM
 Report, 2004.

55 We use the recycling rate for 2004 inferred by the INFORM and CEA estimates because this is the second of the three years for which we present
 model results in Chapter 4.


                                                                                         37

-------
    EXHIBIT  3-11.  PRODUCT-SPECIFIC  END-OF-TOTAL LIFE ALLOCATION WEIGHTS FOR THE
                               MANAGEMENT OF ELECTRONIC WASTE


PRODUCT


Desktop Computers
Laptop Computers
CRT Monitors
LCD Monitors
CRT Televisions
Projection Televisions
Cell Phones
Printers
Keyboards
Computer Mice
LANDFILL
ALLOCATION WEIGHT

(% OF DISPOSED
PRODUCTS)
71.8%
71.8%
73.4%
73.4%
84.2%
84.2%
78.5%
71.8%
71.8%
71.8%
INCINERATION
ALLOCATION WEIGHT

(% OF DISPOSED
PRODUCTS)
2.0%
2.0%
2.1%
2.1%
2.4%
2.4%
2.2%
2.0%
2.0%
2.0%
RECYCLING
ALLOCATION WEIGHT

(% OF DISPOSED
PRODUCTS)
26.1%
26.1%
24.5%
24.5%
13.4%
13.4%
19.2%
26.1%
26.1%
26.1%
END-OF-FIRST-LIFE E-WASTE MANAGEMENT

The allocation weights presented in Exhibits 3-10 and 3-11 reflect how electronic waste is
managed at the end of its second life (i.e., after it will no longer be re-used or held in storage).
Therefore, these allocation weights do not apply to products at the end of their first life because
such products may be placed in storage or re-used.  To generate allocation weights specific to
each product's first life, we followed the two-step process outlined below:

1.  Estimate the percentage of devices that are stored or re-used. Based on the results of the e-
    waste management surveys summarized in Exhibit 3-12, allocation weights that indicate the
    extent to which devices reaching the end of their first life are re-used or placed in storage
    were estimated.56 For example, the allocation weight of 70.1 percent for desktop computers
56 Although the e-waste management surveys include a great deal information with respect to household and private sector management of
 electronic waste, we do not use these results to estimate the volume of electronic waste landfilled, incinerated, recycled, or exported. Many
 households and businesses know whether they send their waste to recycling companies, but they are unlikely to know how much of their
 electronic waste is actually recycled, as recycling companies do not necessarily recycle every item they receive. The decision of whether to
 recycle an electronic product is based largely on the potential profit to be earned from recycling it. If a product does not contain enough
 valuable material for a recycling facility to cover the cost of recycling it, the facility will most likely discard it, in which case it is landfilled,
 incinerated, or exported.  Because of the uncertainty associated with these decisions, we do not use the results of the e-waste management
 surveys to estimate the volume of electronic waste associated with each end-of-life management option. (End-of-life management options include
 options for the final disposition of a product, including landfilling, incineration, recycling, and exporting.  Storage and re-use are not end-of-life
 management options.)
                                                                                                       38

-------
    indicates that 70.1 percent of them are assumed to be placed in storage or re-used at the end
    of their first life.57
                       EXHIBIT  3-12.  E-WASTE MANAGEMENT SURVEY DATA
         SURVEY
                         DESCRIPTION
         Consumer Electronics
         Association Re-use and
         Recycling Survey3


         MetaFacts, Inc.
         Technology User Profile
         Survey*5

         IBM Survey of Senior IT
         Executives at U.S.
         Companies0
         Massachusetts
         Department of
         Environmental
         Protection (DEP)
         Household Surveyd
         California Integrated
         Waste Management
         Board E-waste Diversion
         Study6
In September 2005, the Consumer Electronics Association (CEA)
conducted a nationwide household survey in which respondents
were asked to indicate the number of televisions VCRs, cell phones,
desktop computers, laptop computers, and monitors they had
thrown away, recycled, sold, or given away in the past 12 months.
For its 2004 Technology User Profile, MetaFacts surveyed 7,527
households and 2,500 workplaces, asking respondents to specify
how they managed computers they had replaced during the
previous year.
IBM surveyed 176 senior-level IT executives at companies with more
than 5,000 computers to determine which e-waste management
methods were most prevalent among larger businesses

In 1999, the Massachusetts DEP surveyed 450 Massachusetts
residents asking them to indicate how they managed televisions
that stopped working and computers that they no longer used.

In 2001, the California Integrated Waste Management Board
surveyed 1,003 California households asking them to specify how
they managed the televisions and computer monitors they had most
recently stopped using.
         Sources:
           a. Consumer Electronics Association, Consumer Electronics Reuse and Recycling, October 2005.
           b. Metafacts Inc., Technology User Profile 2004 as cited in Karl Schoenberger, "Many Old Computers Put
             to Use Again, Study Finds," San Jose Mercury News, April 27, 2005.
           c. IBM Global Financing, survey of senior IT executives, cited in John G. Spooner, "Weighing the results
             of PC recycling," CNET News.com, April 16, 2004.
           d. Massachusetts Department of Environmental Protection, Massachusetts DEP Electronic Equipment and
             Household Chemicals Disposal Research, July 1999.
           e. California Integrated Waste Management Board, Selected E-waste Diversion in California: A Baseline
             Study, November 2001.
    Estimate the proportion of devices that are not stored/re-used at the end of their first life.
    The proportion of devices that are not stored/re-used at the end of their first life is
    represented by the expression (1 - ^EFL,?) below.  The proportion not stored/re-used is the
    proportion of devices to which we need to apply an allocation weight for incineration,
    landfilling, and recycling.
57 Appendix C provides a more detailed description of our approach for estimating the storage/re-use allocation weight for devices reaching the end
 of their first life.
                                                                                                     39

-------
3. Estimate the allocation weights for incineration, landfilling, and recycling. To estimate
   end-of-first-life allocation weights for incineration, landfilling, and recycling, the allocation
   weights presented in Exhibit 3-11 were multiplied by the percentage of products that are not
   stored or re-used following the end of their first life (i.e., for those products for which the end
   of their first life is also the end of their total life—the estimate obtained in step 2).  The
   following equation illustrates this approach for incineration:

       /EFL,P = (1 - -^EFL,p) X /EOL,P

where /EFL,P = The end-of-first-life allocation weight for the incineration of product P;

      ^EFL,P = The end-of-first-life allocation weight for the storage/re-use of product P.

       /EOL,P = The end-of-total-life allocation weight for the incineration of product P.

Based on this approach, we estimate the end-of-first-life allocation weights presented in Exhibit
3-13.

As indicated by the allocation weights presented in Exhibit 3-13, at least half of all products,
except for projection televisions,  were estimated to be stored or re-used after they reach the end
of their first life. In addition, the results in Exhibit 3-13  suggest that CRT televisions are less
likely to  be recycled at the end of their first life as compared to CRT monitors.
                                                                                          40

-------
                                 EXHIBIT 3-13. END-OF-FIRST-LIFE MANAGEMENT ALLOCATION WEIGHTS
        PRODUCT
    STORAGE/RE-USE
  ALLOCATION WEIGHT
(% OF PRODUCTS AT THE
END OF THEIR FIRST LIFE)
       LANDFILL
  ALLOCATION WEIGHT
(% OF PRODUCTS AT THE
END OF THEIR FIRST LIFE)
     INCINERATION
  ALLOCATION WEIGHT
(% OF PRODUCTS AT THE
END OF THEIR FIRST LIFE)
 RECYCLING ALLOCATION
       WEIGHT
 (% OF PRODUCTS AT THE
END OF THEIR FIRST LIFE)
TOTAL
Desktop Computers                    70.1%                     21.5%                      0.6%                       7.8%                100%
Laptop Computers                     70.1%                     21.5%                      0.6%                       7.8%                100%
CRT Monitors                         64.7%                     25.9%                      0.7%                       8.7%                100%
LCD Monitors                         64.7%                     25.9%                      0.7%                       8.7%                100%
CRT Televisions                       65.1%                     29.4%                      0.8%                       4.7%                100%
Projection Televisions                 0.0%                     84.2%                      2.4%                      13.4%               100%
Cell Phones                           52.9%                     37.0%                      1.1%                       9.1%                100%
Printers                             70.1%                     21.5%                      0.6%                       7.8%                100%
Keyboards                           70.1%                     21.5%                      0.6%                       7.8%                100%
Computer Mice                       70.1%                     21.5%                      0.6%                       7.8%                100%
Note:
The allocation weights presented in this exhibit reflect the portion of each product that the waste flow model allocates to each management method at the end of
the product's first life (i.e., when its original owner stops using it on a regular basis).

-------
CHAPTER 4  - BASELINE E-WASTE ASSESSMENT  RESULTS
This chapter presents the results of our baseline e-waste assessment, based on the data and
methods described in the previous chapters. We first present retirement estimates that reflect
electronic products reaching the end of their total life (i.e., devices at the end of their second life
and products disposed of at the end of their first life). We then present product-specific
estimates of the volume of electronic waste landfilled, incinerated, or recycled.58

RETIREMENT ESTIMATES

Exhibit 4-1 summarizes the product retirement estimates for 2003 through 2005.  As retirement
estimates, these results reflect electronic products that are disposed of, but not products that are
put into storage or re-use, since stored and re-used products have not yet reached their
retirement. The results in Exhibit 4-1  suggest that retired tonnages are highest for CRT
televisions and CRT monitors, with the tonnage of CRT televisions retired each year higher than
the tonnage retired for any other product. Exhibit 4-1 also suggests that retirement volumes (in
tons and units) for CRT monitors are declining, while retirement volumes (by tons  and units) for
LCD monitors appear to be increasing. This shift reflects the decline in CRT monitor  sales in
recent years as more consumers switch from CRT to LCD monitors.  In addition, Exhibit 4-1
reveals that the most significant electronic products in terms of the number of units retired are
keyboards and cell phones.  However, the tonnages retired for these devices are fairly low due to
their low per unit mass relative to other such devices.

MANAGEMENT OF RETIRED ELECTRONICS

Based on the management allocation weights discussed in the previous chapter, we estimate
retirement by management method for each product included in the analysis.  Exhibits 4-2
through 4-4 summarize our estimates.  Key results included in these exhibits are as follows:

Exhibit 4-2 and 4-4 suggests that, on average,  75 percent of all electronic products  retired (by
number of units) are landfilled each year, while approximately 23 percent are recycled and the
remaining 2 percent incinerated.

Exhibit 4-3 and 4-4 suggests that, on average,  78 percent of all electronic products  retired (by
number of tons) are landfilled each year, while approximately 20 percent are recycled  and the
remaining 2 percent incinerated.
58 As indicated in the previous chapter, adequate data were not available to estimate e-waste exports. However, because our estimates of the
 landfill and incineration rates are based on product discard data, we assume that electronic waste exported is reflected in our estimates of total
 e-waste recycled.

-------
          EXHIBIT 4-1.  AGGREGATE E-WASTE RETIREMENT ESTIMATES BY YEARA'B
PRODUCT

Desktop Computers
Laptop Computers
CRT Monitors
LCD Monitors
Televisions
CRT Televisions0
Projection
Televisions
Cell Phones
Printers
Keyboards0
Computer Micec
All Products
2003
UNITS
18.5
4.0
24.5
0.1
24.4
23.3
1.1
49.0
22.3
43.6
20.8
207.1
TONS
242.1
23.3
418.6
0.6
795.4
716.1
79.3
8.6
207.7
48.5
3.1
1,747.8
2004C
UNITS
19.4
4.8
22.5
0.3
25.2
23.8
1.4
57.0
25.8
50.0
22.7
227.8
TONS
253.6
26.4
383.9
1.8
837.8
735.4
102.4
9.8
241.0
55.5
3.4
1,813.1
2005C
UNITS
19.8
6.1
22.8
0.8
26.3
24.5
1.8
70.6
28.2
52.2
24.2
251.0
TONS
259.5
30.8
389.8
4.9
891.9
759.1
132.8
11.7
263.8
57.5
3.6
1,913.6
TOTALd
UNITS
57.6
15.0
69.8
1.1
75.9
71.6
4.2
176.6
76.3
145.7
67.7
685.9
TONS
755.2
80.5
1,192.3
7.3
2,525.1
2,210.6
314.5
30.1
712.5
161.5
10.1
5,474.5
Notes:
a. Units are in millions and tons are in thousands. Totals may not match due to rounding.
b. Because products in storage or re-use have not yet reached their final retirement, units put into storage or
retirement are not reflected in this exhibit.
c. Due to the long total service lives of desktop computers, CRT monitors, and CRT televisions, the amounts shown
here provide incomplete estimates of aggregate e-waste volumes for these products. Based on the sales information
used in the model, the earliest years for which complete estimates could be generated for desktop computers, CRT
monitors, and CRT televisions are 2005, 2004, and 2007, respectively.
d. Totals may not be the sum of the individual years shown due to rounding.
At least 60 percent of the total tonnage landfilled each year represents the disposal of CRT
televisions and monitors.

Although keyboards and cell phones account for more than 40 percent of the total units landfilled
each year, they make up only 3 to 4 percent of the total tonnage landfilled.

More than 60 percent of the total tonnage incinerated is CRT television and monitor waste.

Desktops and CRT monitors account for more than 40 percent of the tonnage of e-waste recycled
each year.
                                                                                    43

-------
EXHIBIT 4-2. AGGREGATE E-WASTE RETIREMENT ESTIMATES BY MANAGEMENT METHOD (IN MILLIONS OF UNITS)*'
PRODUCT

Desktop Computers
Laptop Computers
CRT Monitors
LCD Monitors
Televisions
CRT Televisions*
Projection TVs
Cell Phones
Printers
Keyboards*
Computer Mice*
All Products
Note:
RECYCLED
2003
4.8
1.0
6.0
0.0
3.3
3.1
0.1
9.4
5.8
11.4
5.4
47.2

2004
5.1
1.3
5.5
0.1
3.4
3.2
0.2
11.0
6.7
13.1
5.9
52.0

2005
5.2
1.6
5.6
0.2
3.5
3.3
0.2
13.6
7.4
13.6
6.3
57.0

LANDFILLED
2003
13.3
2.9
18.0
0.1
20.5
19.6
0.9
38.5
16.0
31.3
14.9
155.5

2004
13.9
3.5
16.5
0.2
21.2
20.1
1.2
44.8
18.5
35.9
16.3
170.9

2005
14.2
4.4
16.7
0.6
22.2
20.6
1.5
55.4
20.3
37.5
17.4
188.7

INCINERATED
2003
0.4
0.1
0.5
0.0
0.6
0.6
0.0
1.1
0.5
0.9
0.4
4.4

2004
0.4
0.1
0.5
0.0
0.6
0.6
0.0
1.3
0.5
1.0
0.5
4.9

2005
0.4
0.1
0.5
0.0
0.6
0.6
0.0
1.6
0.6
1.1
0.5
5.4

TOTAL
2003
18.5
4.0
24.5
0.1
24.4
23.3
1.1
49.0
22.3
43.6
20.8
207.1

2004
19.4
4.8
22.5
0.3
25.2
23.8
1.4
57.0
25.8
50.0
22.7
227.8

2005
19.8
6.1
22.8
0.8
26.3
24.5
1.8
70.6
28.2
52.2
24.2
251.0

* Due to the long total service lives of CRT televisions, keyboards, and mice, the amounts shown here provide incomplete estimates of aggregate e-waste volumes for these
products. Based on the sales information used in the model, the earliest years for which complete estimates could be generated for CRT televisions, keyboards, and mice are
2004, 2005, and 2004, respectively.
** Total may not sum due to rounding.

-------
                 EXHIBIT 4-3.  AGGREGATE  E-WASTE RETIREMENT ESTIMATES BY MANAGEMENT METHOD (IN THOUSANDS OF TONS)
       PRODUCT
RECYCLED
LANDFILLED
INCINERATED
                                                                                                                                         TOTAL
                          2003       2004       2005        2003         2004         2005       2003      2004       2005         2003          2004         2005
Desktop Computers       63.2       66.2       67.8       173.9       182.2        186.5        4.9        5.2        5.3         242.1        253.6        259.5
Laptop Computers        6.1        6.9         8.0        16-8        19.0        22.1         °-5        0.5        0.6         23-3         26.4         30.8
CRT Monitors            102.7     94.2       95.7       307.1       281.7        286.0        8-7        8.0        8.1         418.6        383.9        389.8
LCD Monitors            0.1        Q.4         1.2        °-4         1.3         3.6         0.0        o.O        0.1         °-6          1.8          4.9
Televisions              106.3     112.0      119.3      670.0       705.8        751.4        19.1       20.1       21.4        795.4        837.8        891.9
   CRT Televisions*       95.7       93.3       101.5      603.2       619.5        639.5        17-2       17.6       18.2        716-1        735.4        759.1
   Projection TVs        10.6       13.7       17.8       66.8        86.3        111.9        1-9        2.5        3.2         79-3         102.4        132.8
Cell Phones              1-7        1.9         2.2        6.8         7.7         9.2         0.2        Q.2        0.3         8-6          9.8          11.7
Printers                 54.2       62.9       68.9       149.2       173.1        189.5        4.2        4.9        5.4         207.7        241.0        263.8
Keyboards*              12.7       14.5       15.0       34.8        39.8        41.3         1-0        1.1        1.2         48.5         55.5         57.5
Computer Mice*          0.8        Q.9         0.9        2-2         2.5         2.6         0.1        Q.1        0.1         3-1          3.4          3.6
All Products              347.8     359.9      379.0      1,361.2      1,413.0      1,492.2      38.7       40.2       42.4        1,747.8      1,813.1      1,913.6
Note:
* Due to the  long total service lives of CRT televisions, keyboards, and mice, the amounts shown here provide incomplete estimates of aggregate e-waste volumes for these
products.  Based on the sales information used in the model, the earliest years for which complete estimates could be generated for CRT televisions, keyboards, and mice are
2004, 2005, and 2004, respectively.
** Total may not sum due to rounding.
                                                                                                                                                45

-------
EXHIBIT 4-4. RETIREMENT MANAGEMENT METHOD  FOR SELECT ELECTRONIC PRODUCTS: ZOOS-
                                            ZOOS*
        PRODUCT
RECYCLED
                                   26%
                                   26%
                                   25%
                                   25%
                                   13%
                                   13%
                                   19%
                                   26%
                                   26%
                                   26%
                                   23%
                                   20%
                                        LANDFILLED
                  72%
                  72%
                  73%
                  73%
Desktop Computers
Laptop Computers
CRT Monitors
LCD Monitors
CRT Televisions
Projection Televisions
Cell Phones
Printers
Keyboards
Computer Mice
   All Products (units):
    All Products (tons):
Note:
* The percentages are the same for total volumes based on units or tons, except for All Products,
which are shown separately for units and tons.
                  79%
                  72%
                  72%
                  72%
                  75%
INCINERATED
    2%
    2%
    2%
    2%
    2%
    2%
TOTAL
 100%
 100%
 100%
 100%
 100%
 100%
 100%
 100%
 100%
 100%
 100%
 100%
 ELECTRONICS  PUT INTO  RE-USE OR STORAGE

 As indicated above, the retirement estimates presented in Exhibits 4-2 through 4-4 do not include
 products put into storage or re-use. Nevertheless, the waste flow model tracks the storage and re-
 use of electronics to estimate the volume (in units) and tonnage of electronics retired each year,
 as retired electronics include units that leave storage or re-use. Exhibit 4-5 presents the waste
 flow model's estimates for the quantity of electronics put into storage or re-use for the years
 2003 through 2005. The results in this exhibit reflect a number of trends in the electronics
 industry.  First, consistent with the retirement estimates presented above, the results in Exhibit 4-
 5 suggest that the number of CRT monitors put into storage or re-use each year is declining,
 while the number of LCD monitors put into storage or re-use is increasing. This trend is
 consistent with the ongoing shift in sales from CRT to LCD monitor technology.59 Similarly, the
 storage and re-use of desktop computers appears to be leveling off or falling, while laptop
 computers in storage/re-use is on the rise, which reflects laptops' growing share of the personal
 computer market. The waste flow model estimates that no projection televisions are put into
 storage or re-use because the data identified on the lifespan of projection televisions suggests that
 they are disposed of at the end of their first (and  only) life.
 59 Because of this shift, the tonnage of electronic products in re-use or storage between 2003 and 2005 fell, while the number of products (units) in
  re-use or storage increased.

-------
           EXHIBIT 4-5. ELECTRONICS PUT INTO  STORAGE OR RE-USE:  2003-2005
                       (MILLIONS OF UNITS AND THOUSANDS  OF TONS)
PRODUCT

Desktop Computers
Laptop Computers
CRT Monitors
LCD Monitors
Televisions
CRT Televisions
Projection TVs
Cell Phones
Printers
Keyboards
Computer Mice
All Products
2003
UNITS
27.2
4.7
24.3
0.2
16.7
16.7
0.0
37.5
26.2
43.8
19.7
200.3
TONS
365.0
26.9
416.5
1.0
529.0
529.0
0.0
5.6
244.7
48.2
3.0
1,639.8
2004
UNITS
27.2
5.5
17.6
0.5
16.9
16.9
0.0
40.6
30.4
41.7
19.6
200.0
TONS
365.1
26.6
302.3
2.9
535.4
535.4
0.0
6.1
284.4
45.8
2.9
1,571.7
2005
UNITS
26.0
6.8
15.7
1.2
17.4
17.4
0.0
48.9
31.2
39.5
21.1
207.9
TONS
349.0
28.4
269.4
7.7
552.1
552.1
0.0
7.3
291.5
43.4
3.2
1,552.0
MANAGEMENT OF ELECTRONICS REACHING THE END  OF ITS  FIRST  OR SECOND LIFE

The retirement estimates presented in Exhibits 4-2 through  4-4 combined with the storage/re-use
estimates in Exhibit 4-5 represent the universe of electronics products reaching the end of either
their first or second lives between 2003 and 2005. Exhibits 4-6 through 4-9 summarize the
management of these products. The results in Exhibits 4-6  through 4-9 suggest that most
electronics reaching the end of their first or second life are either put into storage/re-use or are
deposited in a landfill.  This is consistent with the allocation weights in Exhibits 3-11 and 3-13,
which indicate that most electronic products are put into storage or re-use at the end of their first
life and landfilled at the end of their second life.60
50 The allocation weights in Exhibits 3-11 and 3-13 are different than the percentages in Exhibits 4-9 because Exhibits 3-11 and 3-13 apply
 exclusively to electronic products at the end of their second life and first life, respectively, while Exhibit 4-9 combines these two classes of
 products.
                                                                                            47

-------
EXHIBIT 4-6.  SUMMARY OF ELECTRONIC PRODUCTS REACHING THE END OF THEIR FIRST OR SECOND LIVES - 2003
                            (MILLIONS OF UNITS ANDTHOUSANDS OF TONS)
PRODUCT

Desktop Computers
Laptop Computers
CRT Monitors
LCD Monitors
CRT Televisions*
Projection Televisions
Cell Phones
Printers
Keyboards*
Computer Mice*
All Products
PUT INTO STORAGE/RE-USE
UNITS
27.2
4.7
24.3
0.2
16.7
0.0
37.5
26.2
43.8
19.7
200.3
TONS
365.0
26.9
416.5
1.0
529.0
0.0
5.6
244.7
48.2
3.0
1,639.8
RECYCLED
UNITS
4.8
1.0
6.0
0.0
3.1
0.1
9.4
5.8
11.4
5.4
47.2
TONS
63.2
6.1
102.7
0.1
95.7
10.6
1.7
54.2
12.7
0.8
347.8
LANDFILLED
UNITS
13.3
2.9
18.0
0.1
19.6
0.9
38.5
16.0
31.3
14.9
155.5
TONS
173.9
16.8
307.1
0.4
603.2
66.8
6.8
149.2
34.8
2.2
1,361.2
INCINERATED
UNITS
0.4
0.1
0.5
0.0
0.6
0.0
1.1
0.5
0.9
0.4
4.4
TONS
4.9
0.5
8.7
0.0
17.2
1.9
0.2
4.2
1.0
0.1
38.7
TOTAL DISPOSED OR
PUT INTO
STORAGE/RE-USE
UNITS
45.7
8.7
48.8
0.2
40.0
1.1
86.5
48.4
87.4
40.5
407.4
TONS
607.1
50.2
835.1
1.6
1,245.1
79.3
14.2
452.3
96.7
6.1
3,387.6
Note:
* Due to the long total service lives of CRT televisions, keyboards, and mice, the amounts shown here provide incomplete estimates of aggregate e-waste
volumes for these products. Based on the sales information used in the model, the earliest years for which complete estimates could be generated for CRT
televisions, keyboards, and mice are 2004, 2005, and 2004, respectively.

-------
EXHIBIT 4-7.  SUMMARY OF ELECTRONIC PRODUCTS REACHING THE END OF THEIR FIRST OR SECOND LIVES - 2004
                            (MILLIONS OF UNITS ANDTHOUSANDS OF TONS)
PRODUCT

Desktop Computers
Laptop Computers
CRT Monitors
LCD Monitors
CRT Televisions*
Projection Televisions
Cell Phones
Printers
Keyboards*
Computer Mice*
All Products
PUT INTO STORAGE/RE-
USE
UNITS
27.2
5.5
17.6
0.5
16.9
0.0
40.6
30.4
41.7
19.6
200.0
TONS
365.1
26.6
302.3
2.9
535.4
0.0
6.1
284.4
45.8
2.9
1,571.7
RECYCLED
UNITS
5.1
1.3
5.5
0.1
3.2
0.2
11.0
6.7
13.1
5.9
52.0
TONS
66.2
6.9
94.2
0.4
98.3
13.7
1.9
62.9
14.5
0.9
359.9
LANDFILLED
UNITS
13.9
3.5
16.5
0.2
20.1
1.2
44.8
18.5
35.9
16.3
170.9
TONS
182.2
19.0
281.7
1.3
619.5
86.3
7.7
173.1
39.8
2.5
1,413.0
INCINERATED
UNITS
0.4
0.1
0.5
0.0
0.6
0.0
1.3
0.5
1.0
0.5
4.9
TONS
5.2
0.5
8.0
0.0
17.6
2.5
0.2
4.9
1.1
0.1
40.2
TOTAL DISPOSED OR PUT
INTO STORAGE/RE-USE
UNITS
46.6
10.3
40.1
0.7
40.7
1.4
97.7
56.2
91.7
42.3
427.8
TONS
618.8
53.0
686.2
4.8
1,270.8
102.4
15.9
525.3
101.3
6.3
3,384.9
Note:
* Due to the long total service lives of CRT televisions, keyboards, and mice, the amounts shown here provide incomplete estimates of aggregate e-waste
volumes for these products. Based on the sales information used in the model, the earliest years for which complete estimates could be generated for CRT
televisions, keyboards, and mice are 2004, 2005, and 2004, respectively.
                                                                                                   49

-------
EXHIBIT 4-8. SUMMARY OF ELECTRONIC PRODUCTS REACHING THE END OF THEIR FIRST OR SECOND LIVES - 2005
                            (MILLIONS OF UNITS AND THOUSANDS OF TONS)
PRODUCT

Desktop Computers
Laptop Computers
CRT Monitors
LCD Monitors
CRT Televisions*
Projection Televisions
Cell Phones
Printers
Keyboards*
Computer Mice*
All Products
PUT INTO STORAGE/RE-USE
UNITS
26.0
6.8
15.7
1.2
17.4
0.0
48.9
31.2
39.5
21.1
207.9
TONS
349.0
28.4
269.4
7.7
552.1
0.0
7.3
291.5
43.4
3.2
1,552.0
RECYCLED
UNITS
5.2
1.6
5.6
0.2
3.3
0.2
13.6
7.4
13.6
6.3
57.0
TONS
67.8
8.0
95.7
1.2
101.5
17.8
2.2
68.9
15.0
0.9
379.0
LANDFILLED
UNITS
14.2
4.4
16.7
0.6
20.6
1.5
55.4
20.3
37.5
17.4
188.7
TONS
186.5
22.1
286.0
3.6
639.5
111.9
9.2
189.5
41.3
2.6
1,492.2
INCINERATED
UNITS
0.4
0.1
0.5
0.0
0.6
0.0
1.6
0.6
1.1
0.5
5.4
TONS
5.3
0.6
8.1
0.1
18.2
3.2
0.3
5.4
1.2
0.1
42.4
TOTAL DISPOSED OR PUT
INTO STORAGE/RE-USE
UNITS
45.8
13.0
38.5
2.0
41.9
1.8
119.5
59.4
91.7
45.3
458.9
TONS
608.5
59.3
659.2
12.6
1,311.2
132.8
19.0
555.3
100.9
6.8
3,465.6
Note:
* Due to the long total service lives of CRT televisions, keyboards, and mice, the amounts shown here provide incomplete estimates of aggregate e-waste
volumes for these products. Based on the sales information used in the model, the earliest years for which complete estimates could be generated for CRT
televisions, keyboards, and mice are 2004, 2005, and 2004, respectively.
                                                                                                    50

-------
EXHIBIT 4-9. MANAGEMENT OF ELECTRONIC PRODUCTS REACHING THE END OF THEIR FIRST OR SECOND LIVES
                                         2003 - 2005*
PRODUCT


Desktop Computers
Laptop Computers
CRT Monitors
LCD Monitors
CRT Televisions
Projection Televisions
Cell Phones
Printers
Keyboards
Computer Mice
All Products
PUT INTO STORAGE/RE-USE

2003
60.1%
53.5%
49.9%
62.7%
42.5%
0.0%
39.5%
54.1%
49.9%
48.7%
48.4%

2004
59.0%
50.2%
44.1%
61.8%
42.1%
0.0%
38.4%
54.1%
45.3%
46.3%
46.4%

2005
57.3%
48.0%
40.9%
61.1%
42.1%
0.0%
38.6%
52.5%
43.1%
46.5%
44.8%
RECYCLED

2003
10.4%
12.1%
12.3%
9.2%
7.7%
13.4%
1 1 .6%
12.0%
13.1%
13.4%
10.3%

2004
10.7%
13.0%
13.7%
9.4%
7.7%
13.4%
11.8%
12.0%
14.3%
14.0%
10.6%

2005
11.1%
13.6%
14.5%
9.5%
7.7%
13.4%
1 1 .8%
12.4%
14.9%
14.0%
10.9%
LANDFILLED

2003
28.6%
33.4%
36.8%
27.4%
48.4%
84.2%
47.5%
33.0%
36.0%
36.9%
40.2%

2004
29.5%
35.8%
41 .0%
28.0%
48.7%
84.2%
48.4%
33.0%
39.3%
38.6%
41.7%

2005
30.6%
37.4%
43.4%
28.5%
48.8%
84.2%
48.3%
34.1%
40.9%
38.4%
43.1%
INCINERATED

2003
0.8%
1 .0%
1 .0%
0.8%
1 .4%
2.4%
1 .4%
0.9%
1 .0%
1 .0%
1.1%

2004
0.8%
1 .0%
1 .2%
0.8%
1 .4%
2.4%
1 .4%
0.9%
1.1%
1.1%
1 .2%

2005
0.9%
1.1%
1 .2%
0.8%
1 .4%
2.4%
1 .4%
1 .0%
1 .2%
1.1%
1 .2%
TOTAL DISPOSED
OR PUT INTO
STORAGE/RE-USE
2003, 2004, AND
2005
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
Note:
* Percent of e-waste tonnage associated with each method.
                                                                                               51

-------
CONCLUSION  AND NEXT STEPS

The purpose of this report is to establish a set of baseline data, using the methodology outlined in
this report, 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
electronic products.

The results should provide interested parties with useful information for developing choices on
how to manage electronic products at the end of their useful lives.  For example, Exhibit 4-1
indicates that, of the products studied, CRT monitors and televisions make up more than half of
the electronic waste generated in recent years. Although sales of these products are falling as
consumers convert to newer technologies, the results of our analysis suggest that they will
continue to make up a significant portion of the electronic waste stream for years to come,
particularly if households continue to keep their televisions for as long as they have in the past.
Even as new technologies, such as plasma televisions and LCD monitors increase in sales, there
will still be millions of CRTs that will be disposed of during the next several years.

The existing information indicates that, generally speaking, U.S. landfills are the primary
repository of discarded electronic products, with recycling facilities playing a secondary role.
This finding raises important questions about resource conservation, since such electronics
deposited in landfills represent lost energy and material resources.

Material composition of electronic devices is important to examine. Information on material
composition is important to identify products with  recycling potential. For example, although
cell phones make up a small fraction of the e-waste deposited in landfills each year, the
information in  Appendix E suggests that they may  contribute a larger proportion of silver to U.S.
landfills than other devices included in this study.61 Silver and other valuable materials found in
electronic products represent important resources to recover from electronic waste.
61 Exhibit E presents the material composition of retired electronic products by management method and by product for the following devices:
 desktop computers, laptop computers, CRT monitors, LCD monitors, CRT televisions, cell phones, and keyboards. Material composition data were
 not available for projection televisions, printers, and mouse devices.

-------
APPENDIX A: GLOSSARY OF TERMS
Allocation Weight: For a given quantity of devices, the percentage of those devices at the end of
their first or second life that the waste flow model allocates to a specific management method
(e.g., incineration).

Age Distribution:  A distribution describing the various ages at which a particular product is
made available for end-of -life management and the frequency at which products are made
available for such management at a given age.  The age of a product is based on the number of
years between its original sale and the end of its life.

Cohort: A group of products that are sold by the original manufacturer or retailer in the same
year.

Discard: Disposal  of waste through landfilling or incineration.

Discard Rate: The rate at which waste is landfilled or incinerated (i.e., the percent of waste
incinerated or landfilled).

Disposal: Management of a product at the end of its useful life through landfilling or
incineration.

End-of-First-Life  Management: The method used to manage a product when it reaches the end
of its first life. Management options for the end of a product's first life include landfilling,
incineration, recycling (including export), reuse, and storage.

End-of-Total-Life Management: The method used to manage a product when it reaches the end
of its total life. Management options for the end of a product's total life include landfilling,
incineration, recycling (including export) (end-of-total-life is the sum of products reaching  end-
of-first-life and the end-of-second-life).

Export/exportation: Transport of devices outside U.S. borders for re-use, refurbishing,
recycling or disposal.

First Life: The length of time a product is used by its original or initial owner.

Gross State Product (GSP): The value of goods and services produced by the labor and
property located in a state.

Incineration: Electronic waste may be burned at an incineration or waste-to-energy
facility.
                                                                                      53

-------
Initial Retirement: The point in time at which the original or initial owner of a product stops
using it.

Initial Service Life: See First Life.

Landfilling: Electronic waste may be deposited in a landfill for final disposal.

Lifespan Distribution: Probabilities assigned across a range of years indicating the likelihood of
a product reaching the end of its lifespan in any given year.

Recycling: Electronic products may be recovered for the purpose of dismantling, parts and/or
materials recovery, and/or resale (resale that occurs by a recycler and not by the user of the
product).

Retirement: Retirement is when a product reaches the end of its life.

Reuse:  Occurs when the first user gives up a product by informal sale or donation (other than
making it available for end of life management) and a subsequent user uses the product for its
intended purpose.

Second Life: The length of time over which a product is reused or kept in storage after its first
life.  See First Life and Initial Retirement.

Storage: Holding or storing a product for a temporary period by the first owner of the product or
any other owner, at the end of which it is reused, resold, recycled, or disposed.

Total Life or Total Lifespan: The period of time between when a product is initially purchased
and when it reaches the end of its life. The length of time for a product's total lifespan is the sum
of its first life and its second life. See End-of-Life Management,
First Life, and Second Life.

Total Service Life: See Total Life.

White boxes: computers that are distributed without a well-recognized brand name.
                                                                                      54

-------
APPENDIX B: ESTIMATING THE TOTAL LIFESPAN
AND  SECOND LIFE OF ELECTRONIC PRODUCTS
As described in the main body of this report, the amount of electronic waste generated in the
U.S. each year was estimated based on the annual sales of individual electronic products and the
typical lifespan of them (i.e., the amount of time that elapses from their initial purchase to final
disposition/recycling). To estimate the total lifespan of each product included in our baseline
assessment, data was used on the age of the electronic products collected for recycling as part of
the Florida Electronic Product Brand Distribution Project (the Florida Project) during the 12-
month period beginning in April 2004. Based on these data, a distribution was developed of the
length of time each product remains in circulation before it is recycled or disposed.  Exhibit B-l
summarizes these data for several of the electronic products collected as part of the Florida
Project. Although the data presented in this exhibit represent the age distribution of electronic
products collected through the Florida Project, they do not represent the distribution of each
product's lifespan. For example, although 7.1 percent of the CRT televisions collected were 15
years old, this does not imply that 7.1 percent of televisions remain in circulation for 15 years
after they are sold. The number of 15-year-old televisions included in the Florida data reflects
the amount of time that a television remains in circulation and the number of televisions sold 15
years ago. Therefore, if no televisions had been sold 15 years prior to the Florida collection, the
Florida data would include no 15-year-old televisions, even though a certain (non-zero)
percentage of televisions are retired when they reach this age.

As the example above illustrates, the age data presented in Exhibit B-l reflects both the amount
of time a product remains in use (or in storage) and the product's annual sales prior to the Florida
collection.  Therefore, because sales of each product vary from year to year, the age distributions
in Exhibit B-l do not accurately represent the distribution of each product's lifespan. To address
this issue, we standardized the Florida collection data to account for differences in electronic
product sales from year to year based on the following procedure:
                                                                                      55

-------
EXHIBIT B-1: AGE DISTRIBUTION OF PRODUCTS
   COLLECTED FOR RECYCLING IN FLORIDA
AGE
(YEARS)
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
DESKTOP
COMPUTERS
0.1%
0.1%
0.4%
1 .0%
1.7%
4.6%
6.0%
8.5%
10.0%
4.6%
6.1%
11.3%
4.4%
7.7%
4.9%
5.9%
3.7%
5.8%
2.8%
1 .2%
2.5%
2.8%
0.7%
1.3%
1 .0%
0.4%
0.1%
0.2%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
LAPTOP
COMPUTERS
0.0%
0.0%
0.0%
0.0%
20.0%
15.0%
20.0%
15.0%
25.0%
0.0%
0.0%
0.0%
0.0%
0.0%
5.0%
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%
0.0%
0.0%
0.0%
0.0%
0.0%
CRT
MONITORS
0.0%
0.1%
1 .2%
2.3%
4.4%
8.1%
10.3%
11.6%
13.0%
8.8%
8.6%
9.1%
6.9%
5.6%
2.6%
2.3%
1 .2%
1.6%
0.7%
0.4%
0.6%
0.2%
0.2%
0.2%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
LCD
MONITORS
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
33.3%
0.0%
0.0%
0.0%
33.3%
0.0%
33.3%
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%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
CRT TVS
0.1%
0.6%
1 .2%
1.6%
2.4%
2.9%
3.8%
3.8%
4.0%
4.1%
5.7%
8.2%
6.0%
6.3%
5.2%
7.1%
4.7%
5.9%
5.0%
3.2%
4.0%
3.9%
2.2%
2.3%
1.7%
1.1%
1 .0%
0.9%
0.5%
0.3%
0.0%
0.3%
0.0%
0.1%
0.0%
PROJECTION
TVS
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
14.3%
14.3%
42.9%
14.3%
14.3%
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%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
PRINTERS
0.4%
1.7%
2.6%
2.4%
12.2%
7.5%
7.2%
7.9%
12.3%
8.1%
8.5%
9.5%
4.2%
4.3%
3.8%
3.2%
1.5%
1.1%
0.7%
0.5%
0.3%
0.2%
0.0%
0.1%
0.0%
0.0%
0.0%
0.0%
0.0%
0.1%
0.0%
0.0%
0.0%
0.0%
0.0%
COMPUTER
KEYBOARDS
0.0%
0.0%
0.0%
4.9%
43.9%
19.5%
4.9%
0.0%
2.4%
2.4%
2.4%
0.0%
7.3%
0.0%
7.3%
2.4%
0.0%
0.0%
0.0%
0.0%
2.4%
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%
Source: Florida Department of Environmental Protection, Florida Brand Distribution Project,
http://www.dep.state.fl.us/waste/categories/electronics/pages/FloridaElectronicProductBrandDistributionProje
ct.htm.
                                                          56

-------
1.   First, Florida sales of each product were estimated based on the assumption that Florida's share
    of the national electronics sales is proportional to its share of U.S. economic output, as indicated
    by its gross state product.62 Equation B-l summarizes this calculation.
    (B-l)
                       F   - N
                       -1 P,T   JV P,Y
GSPFJ
                                     50
                                    I GSPSJ
           where, FPjY = Florida sales of product P in year Y;
                  NPJ = National sales of product P in year Y;
                  GSPpj= Florida's gross state product in year Y, and
                  GSPsj = The gross state product of State S in year Y.

    For example, as shown in Exhibit B-2, Florida generated 4.8 percent of the U.S. economic output
    in 2000, and U.S. sales of computer keyboards were approximately 60.6 million units that year.
    Therefore, based on Equation B-l, 2.9 million computer keyboards were estimated to be sold in
    Florida in 2000.

2.  Based on the Florida sales estimates generated in step 1 and the 2004 collection data from the
    Florida Project, the percentage of each cohort (e.g., electronics sold in 1992) collected in 2004
    was estimated. For example,  as indicated in Exhibit B-2, 18 keyboards sold in 2000 were
    collected through the Florida Project in 2004; these keyboards represent 0.0006 percent of the
    keyboards sold in Florida in 2000.63

3.  Using the percentage values calculated in step 2, the number of electronic units was estimated for
    each cohort that would have been collected through the Florida Project if annual electronic sales
    had been constant over time.  For example, if annual  keyboard sales had been constant  at
    1,000,000 units per year between 1989 and 2004, approximately 6 keyboards sold in 2000 would
    have been collected in 2004 as part of the Florida Project, as shown in Exhibit B-2.

4.  After performing the calculations outlined in step 3, it is possible to estimate an age distribution
    for electronic products that would have been collected if sales were constant over time.  For
    example, as indicated in Exhibit B-2, keyboards sold in 2000 would represent the keyboards
    collected in 2004 if keyboard  sales were the same each year. As indicated above, the age
     Gross state product (GSP) as defined by the Bureau of Economic Analysis is "the value added in production by the labor and property located in a
     state." U.S. Bureau of Economic Analysis, http://www.bea.gov/bea/regional/definitions/nextpage.cfm?key=Gross%20state%20product%20(GSP).

    63 This step in our calculations relies on the availability of sales data for the cohorts of each product reflected in the Florida collection data. In
     most cases, we were able to obtain these sales data. In the few cases where we were unable to obtain sales estimates, we exclude the
     corresponding collection data from our analysis. For example, the Florida data indicate that one of the 20 laptops collected through the Florida
     Project was sold in 1990.  Because we lack laptop sales data for 1990, we do not include the collection data for 1990 vintage laptops in our
     analysis. Overall, such exclusions do not significantly affect our lifespan estimates because we exclude very few collected items from our analysis
     (i.e., one of 20 laptops included in the Florida data, four of 1,912 desktops, one of 4,517 CRT monitors, and one of 1,028 printers).


                                                                                                   57

-------
approximately 37 percent of the keyboards collected in 2004 if keyboard sales were the same
each year.  As indicated above, the age distribution of products collected for recycling reflects
the lifespan of the collected products only if sales do not vary over time. Because we held
annual sales constant to generate the keyboard age distribution presented in Exhibit B-2, this
distribution represents the lifespan distribution of keyboards collected for recycling through the
Florida Project.

Exhibit B-3 presents the lifespan distributions developed, applying the method outlined above to
each product included in our analysis. The results presented in Exhibit B-3 suggest that
televisions have the longest lifespan of the products included in our analysis, while laptop
computers have the shortest.
                                                                                        58

-------
EXHIBIT B-2: ESTIMATION OF KEYBOARD LIFESPAN DISTRIBUTION
VINATAGE
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
TOTAL
AGE
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
-
U.S. KEYBOARD
SALES (1000S
OF UNITS)1
23,434
28,603
33,099
35,784
38,563
41,442
47,599
53,756
55,605
64,957
63,675
60,571
57,468
54,364
51,260
47,209
757,388
FLORIDA SHARE
OF U.S.
ECONOMIC
ACTIVITY2
4.5%
4.5%
4.6%
4.6%
4.7%
4.7%
4.7%
4.7%
4.8%
4.8%
4.8%
4.8%
4.9%
5.0%
5.1%
5.1%
-
FLORIDA
KEYBOARD SALES
(1000S OF UNITS)
1,058.7
1,296.8
1,513.6
1,644.8
1,805.0
1,944.1
2,240.9
2,547.2
2,642.2
3,115.0
3,062.1
2,920.9
2,838.8
2,727.2
2,598.3
2,409.2
36,364.7
Notes:
Derived from U.S. Census Bureau, Current Industrial Report sen
Based on gross state product data from the Bureau of Economic
Florida Department of Environmental Protection, Florida Electn
The annual values do not sum to 41 because the Florida collecti
exclude this keyboard from our analysis because we were unable to obta
KEYBOARDS
COLLECTED IN
2004 FLORIDA
STUDY3
1
3
0
3
0
1
1
1
0
2
8
18
2
0
0
0
414
PERCENT OF COHORT
COLLECTED BY FLORIDA
PROJECT IN 2004
0.00009%
0.00023%
0.0%
0.00018%
0.0%
0.00005%
0.00004%
0.00004%
0.00000%
0.00006%
0.00026%
0.00062%
0.00007%
0.0%
0.0%
0.0%
-
SIMULATED NUMBER OF
KEYBOARDS COLLECTED IN
2004, BY COHORT,
ASSUMING 1 MILLION
KEYBOARDS SOLD PER YEAR
0.9
2.3
0.0
1.8
0.0
0.5
0.4
0.4
0.0
0.6
2.6
6.2
0.7
0.0
0.0
0.0
16.6
AGE AND
LIFESPAN
DISTRIBUTION
5.7%
14.0%
0.0%
11.0%
0.0%
3.1%
2.7%
2.4%
0.0%
3.9%
15.8%
37.2%
4.3%
0.0%
0.0%
0.0%
100.0%
es, "Computers and Office and Accounting Machines, " 1989 through 2004.
Analysis, www.bea.gov.
inic Product Brand Distribution Project, Collection data for 2004-2005.
Dn data include one keyboard sold in 1984 that is not included in our calculations. We
n keyboard sales values for 1984.

-------
         EXHIBIT B-3: LIFESPAN  DISTRIBUTION OF SELECT ELECTRONIC PRODUCTS
LIFESPAN
(YEARS)
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
TOTAL
DESKTOP
COMPUTERS
0.0%
0.0%
0.1%
0.3%
0.5%
1 .3%
2.0%
3.4%
4.8%
2.2%
3.3%
6.5%
2.8%
5.3%
3.9%
5.8%
4.2%
7.0%
4.4%
2.1%
3.9%
6.5%
2.8%
9.5%
9.6%
5.5%
2.1%
0.0%
0.0%
0.0%
100%
LAPTOP
COMPUTERS
0.0%
0.0%
0.0%
0.0%
9.8%
9.2%
22.1%
18.4%
40.5%
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%
0.0%
0.0%
0.0%
0.0%
0.0%
100%
CRT
MONITORS
0.0%
0.1%
0.8%
1.5%
2.1%
3.6%
5.2%
7.1%
8.9%
6.8%
7.7%
9.5%
9.5%
8.7%
4.7%
4.5%
2.7%
3.9%
2.2%
1 .4%
1 .9%
1.1%
1 .4%
2.6%
0.9%
0.6%
0.9%
0.0%
0.0%
0.0%
100%
LCD
MONITORS
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
14.3%
14.3%
14.3%
14.3%
14.3%
14.3%
14.3%
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%
0.0%
100%
CRT TVS
0.1%
0.4%
0.8%
1 .2%
1.6%
1 .9%
2.6%
2.8%
3.2%
4.0%
5.2%
7.0%
5.2%
7.2%
5.6%
6.9%
5.1%
5.2%
4.5%
2.9%
4.4%
4.9%
3.5%
3.6%
3.0%
2.0%
1.7%
1 .9%
1.1%
0.8%
100%
PROJECTION
TVS
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
9.7%
10.8%
37.2%
18.2%
24.1%
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%
0.0%
0.0%
0.0%
100%
PRINTERS
0.1%
0.6%
1.1%
1.1%
5.1%
3.4%
4.5%
6.0%
7.1%
5.7%
7.6%
11.8%
6.8%
9.1%
8.3%
6.3%
3.9%
3.3%
2.7%
2.4%
1 .2%
1 .0%
0.0%
0.9%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
100%
KEYBOARDS
0.0%
0.0%
0.0%
4.3%
37.2%
15.8%
3.9%
0.0%
2.4%
2.7%
3.1%
0.0%
11.0%
0.0%
14.0%
5.7%
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%
100%
MICE1
0.0%
0.0%
0.0%
4.3%
37.2%
15.8%
3.9%
0.0%
2.4%
2.7%
3.1%
0.0%
1 1 .0%
0.0%
14.0%
5.7%
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%
100%
CELL
PHONES2
0.0%
0.0%
0.0%
0.0%
9.8%
9.2%
22.1%
18.4%
40.5%
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%
0.0%
0.0%
0.0%
0.0%
0.0%
100%
Notes:
We use the lifespan distribution for keyboards as a proxy for mice.
We use the lifespan distribution for laptop computers as a proxy for cell phones.
Although Exhibit B-3 presents lifespan information for mice and cell phones, the Florida data do
not include these two products. Therefore, the keyboard lifespan distribution was used as a
proxy for mice and the laptop distribution was used for cell phones.  We believe that keyboards
are a reasonable proxy for mice because both are relatively inexpensive input devices necessary
to use a computer.  Although laptops differ from cell phones in several important ways, we

-------
believe they represent the best proxy for cell phones among the various products included in our
analysis because both are portable technologies that have advanced significantly during the past
ten years. In addition, because only three LCD monitors are reflected in the Florida data, the
methodology outlined above is not applied to estimate their lifespan. Instead, LCD monitors
were assumed to have a uniform lifespan distribution, with the low end of the distribution
representing the age of the youngest LCD monitor included in the Florida data and the high end
reflecting the oldest.


DURATION OF SECOND LIFE

The distributions in Exhibit B-3 represent the total lifespan of each product. As outlined in the
methodology chapter in the main body of this report, however, the management decisions for
individual products were simulated at two points in time: the end of their first life (i.e., the point
in time when the original owner of a product stops using it) and the end of their second life (i.e.,
the period of time over which a product is reused or in storage after its first life). To develop a
distribution of the second life of each product included in Exhibit B-3, the steps outlined above
were followed for each product's total life, but we limited the analysis to that data collected that
corresponded to the second life of each product. For example, the  main body of this report
indicates that a keyboard's first life lasts for approximately three years. Therefore, to estimate a
lifespan distribution for the second life of a computer keyboard, the steps outlined above were
followed, but we did not use the data collected for keyboards three years  old and younger at the
time of their collection.

Applying this approach to each product,  lifespan distributions were generated for the second
lives of the products included in our analysis, as presented in Exhibit B-4. The distributions
presented in this exhibit suggest that desktop computers have the longest second life among the
products included in our analysis, while keyboards and mice have the shortest.

Similar to the total lifespan distributions presented in Exhibit B-3,  the second lifespan
distributions of keyboards was used as a proxy for mice and laptops was used as a proxy for cell
phones. In addition, as the results in Exhibit B-4 suggest, a distribution for the second life of
projection televisions  was not estimated.  The Florida collection data and our estimate of a
projection television's first life suggest that almost no projection televisions go into  storage or
are re-used.  Therefore, all projection televisions were assumed to be disposed of or  recycled at
the end of their first life.
                                                                                       61

-------
EXHIBIT B-4: DISTRIBUTION OF THE SECOND LIFESPAN OF SELECT ELECTRONIC PRODUCTS
YEARS IN
CIRCULATION
AFTER FIRST LIFE
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
TOTAL
DESKTOP
COMPUTERS
0.5%
1.3%
2.0%
3.4%
4.8%
2.2%
3.3%
6.6%
2.8%
5.3%
4.0%
5.8%
4.2%
7.0%
4.4%
2.1%
4.0%
6.6%
2.9%
9.6%
9.6%
5.5%
2.1%
100%
LAPTOP
COMPUTERS
9.8%
9.2%
22.1%
18.4%
40.5%
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%
0.0%
0.0%
100%
CRT MONITORS
3.7%
5.4%
7.4%
9.3%
7.1%
8.0%
10.0%
10.0%
9.1%
4.9%
4.7%
2.8%
4.0%
2.3%
1.4%
2.0%
1.1%
1.5%
2.7%
0.9%
0.6%
1.0%
0.0%
100%
LCD
MONITORS
16.7%
16.7%
16.7%
16.7%
16.7%
16.7%
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%
0.0%
100%
CRT TVS
9.2%
6.8%
9.4%
7.3%
9.0%
6.7%
6.8%
5.9%
3.8%
5.7%
6.4%
4.6%
4.7%
3.9%
2.6%
2.3%
2.5%
1.4%
1.0%
0.0%
0.0%
0.0%
0.0%
100%
PROJECTION
TVS
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%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0%
PRINTERS
5.0%
6.8%
8.0%
6.4%
8.6%
13.3%
7.7%
10.2%
9.3%
7.1%
4.5%
3.8%
3.1%
2.7%
1.3%
1.1%
0.0%
1.0%
0.0%
0.0%
0.0%
0.0%
0.0%
100%
KEYBOARDS
38.9%
16.5%
4.1%
0.0%
2.5%
2.8%
3.2%
0.0%
11.5%
0.0%
14.6%
6.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
100%
MICE1
38.9%
16.5%
4.1%
0.0%
2.5%
2.8%
3.2%
0.0%
11.5%
0.0%
14.6%
6.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
100%
CELL PHONES2
9.8%
9.2%
22.1%
18.4%
40.5%
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%
0.0%
0.0%
100%
Notes:
We use the distribution of the duration of a keyboard's second life as a proxy for mice.
We use the distribution of the duration of a laptop computer's second life as a proxy for cell phones.

-------
APPENDIX C:  ESTIMATING THE REUSE AND STORAGE OF ELECTRONIC
PRODUCTS
As part of our assessment of the baseline generation and management of electronic waste, the
number of products that are reused or placed in storage was estimated after their first life (i.e.,
the number of products that move on to a second life after their original owners stop using them).
To estimate the percentage of electronic products (by product) that are reused or stored, data was
used from surveys conducted by the Consumer Electronics Association, MetaFacts, IBM, the
Massachusetts  Department of Environmental Protection, and the California Integrated Waste
Management Board.64 In the following sections, we summarize how we used these data for each
product included in our baseline assessment.

STORAGE AND  RE-USE OF COMPUTERS AND COMPUTER PERIPHERALS

To estimate the storage/re-use rate for computers (desktops and laptops), separate storage/re-use
rates were estimated for computers used by households and computers used by businesses. The
weighted average of these two values was then estimated and combined to a single storage/re-use
rate for all computers. This rate was applied to both desktop and laptop computers, as well as to
the peripheral devices included in the baseline assessment (i.e., printers, keyboards, and mouse
devices).

Based on data from surveys conducted by MetaFacts and the Massachusetts Department of
Environmental Protection, a household storage/re-use rate of 82.7 percent was estimated.65  This
value represents the mean of the storage/re-use rates implied by these two studies,  both of which
asked respondents about their management of computers that they replaced or that they no longer
used.

To estimate the storage/re-use rate for workplace computers, publicly available results were used
from an IBM survey of 176 IT executives.  This survey asked executives to list the various ways
in which their companies disposed of their computers, but it did not ask about the extent to which
companies relied on each method or whether companies stored old computers before disposing
of them. Therefore, the results of this survey at best provide a rough approximation of the
storage/re-use rate for workplace computers. The results of the survey are presented in Exhibit
C-l.
  Consumer Electronics Association, Consumer Electronics Reuse and Recycling, October 2005; Metafacts Inc., Technology User Profile 2004 as
 cited in Karl Schoenberger, "Many Old Computers Put to Use Again, Study Finds," San Jose Mercury News.  April 27, 2005; IBM Global Financing,
 survey of senior IT executives, cited in John G. Spooner, "Weighing the results of PC recycling," CNET News.com, April 16, 2004; Massachusetts
 Department of Environmental Protection, Massachusetts DEP Electronic Equipment and Household Chemicals Disposal Research, July 1999; and
 California Integrated Waste Management Board, Selected E-waste Diversion in California: A Baseline Study, November 2001.

55 In developing this estimate, we assume that computers repaired or sold by households are re-used.

-------
                      EXHIBIT C-1: IBM IT EXECUTIVE SURVEY RESULTS
MANAGEMENT OPTION
Use Outside Disposal
Company
Donate
Discard On Site
TOTAL
PERCENT OF COMPANIES THAT USE
MANAGEMENT PRACTICE FOR AT LEAST
SOME OF THEIR UNWANTED COMPUTERS1
50.0%
40.0%
23.0%
113%
SURVEY RESULTS
SCALED TO 100
PERCENT TOTAL2
44.2%
35.4%
20.4%
100%
Notes:
Column does not total to 100 percent because the survey asked respondents to
list all of the management practices they use. Therefore, respondents were able to list
multiple management methods.
This column represents the previous column scaled to add up to 100 percent.
Source: IBM Global Financing, survey of senior IT executives, cited in John G. Spooner,
"Weighing the results of PC recycling," CNET News.com, April 16, 2004.
To estimate the storage/re-use rate for workplace computers based on the IBM data, the IBM
survey results were assumed to represent the proportional relationship between the number of
workplace computers managed by disposal companies, donated to charity, and discarded on site
(e.g., depositing them in on-site trash containers).66 Based on this assumption, approximately 44
percent of workplace computers were  estimated to be managed by an outside disposal company,
while 35 percent were donated to  charity.  Due to resource constraints, we were unable to
identify any studies of the electronics disposal industry indicating how many computers are re-
sold by disposal companies on an  annual basis.  An article published by the Kansas City Star in
May of 2000, however, suggests that these companies re-sell approximately 50 percent of the
electronic products they collect.67 Based on this percentage, approximately 22 percent of
workplace computers (i.e., half of 44 percent) were estimated to be managed by disposal
companies and subsequently re-sold.  Adding this to the 35 percent of workplace computers
donated to charity, a storage/re-use rate of 58 percent was estimated for workplace computers.68

To estimate the storage/re-use rate for all computers, the weighted average of the household and
workplace storage/re-use rates presented above was estimated, using the number of computers in
 Mathematically, this entails scaling the survey results to sum to 100 percent.

67 Bergstrom, Bill. "New breed of recyclers handles castoff computers," Kansas City Star, May 9, 2000. In this article, Neil Peters-Michaud, chief
 executive of Cascade Asset Management, is quoted as saying that the company re-sells 50 percent of the electronics it collects by weight.

68 This estimate is rounded from 57.5 percent.
                                                                                          64

-------
                                                             69
residential and commercial buildings in 1999 as weights.    According to the Energy Information
Administration (EIA), approximately 58 million computers were in commercial buildings in
      70
1999.   Based on EIA and Census data for 1997 and 2001 respectively, approximately 57.6
                                                                           71
million computers were estimated to be in U.S. households as of 1999.    Therefore, calculating
the weighted average of the household and workplace storage/re-use rates presented above, the
overall storage/re-use rate for computers (desktops and laptops) was estimated to be 70.1
percent, as indicated in Exhibit C-2.  Because we identified no survey data specific to the storage
or re-use of computer peripherals, the computer storage/re-use rate was applied to printers,
keyboards, and mice, as well.
           EXHIBIT C-2: STORAGE/RE-USE RATE OF SELECT ELECTRONICS PRODUCTS
               PRODUCT
STORAGE/RE-USE RATE
               Desktop Computers                                 70.1%
               Laptop (Portable) Computers                        70.1%
               CRT Monitors                                       64.7%
               LCD Monitors                                       64.7%
               CRT Televisions                                    65.1%
               Projection Televisions                               0.0%
               Desktop Computer Printers                          52.9%
               Keyboards                                         70.1%
               Computer Mice                                     70.1%
               Cell Phones                                        70.1%
               Note:
               We apply these percentages only to products at the end of their first life. By
               definition, products at the end of their total life are disposed of rather than
               stored or re-used.
69 We use estimates for 1999 because this is the most recent year for which data were available for both households and commercial buildings.

70 U.S. DOE, EIA, "Computers and Photocopiers in Commercial Buildings,"

 http://www.eia.doe.gov/emeu/cbecs/pc_copier/pccopier99.html, accessed November 10, 2005.

71 EIA estimates that approximately 43 million computers were in U.S. households in 1997 (U.S. DOE, EIA, "U.S. Households Usage of Appliances in
 1997" http://eia.doe.gov/emeu/recs/recs97/appusage.html, accessed November 10, 2005). In addition, the U.S. Census Bureau estimates that
 approximately 45.8 million U.S. households had one computer in 2001 and that approximately 15.7 million households had more than one
 computer (U.S. Census Bureau, Current Population Survey, September 2001). Assuming that all households with more than one computer had two
 computers, we estimate that approximately 77.1 million computers were in U.S. households in 2001. Assuming a constant growth rate between
 1997 and 2001, we estimate that approximately 57.6 million computers were in U.S. households in 1999.
                                                                                                     65

-------
STORAGE AND  RE-USE OF COMPUTER MONITORS
The estimate of the storage/re-use rate for household computer monitors is based on the results
of the California Integrated Waste Management Board's e-waste diversion survey.72 Based on
the results of the California survey, a household storage/re-use rate of 72 percent was estimated.

None of the information sources identified contained data specific to the storage or re-use of
computer monitors in workplaces.  Therefore, the storage/re-use rate estimated for workplace
computers (57.6 percent) was used as a proxy for the storage and re-use of monitors in the
workplace.  Similarly, the proportional relationship between the number of monitors in
households and the number of monitors in commercial buildings was assumed to be the same as
the corresponding relationship for computers. Calculating the weighted average of the
household and workplace storage/re-use rates, the overall  storage/re-use rate for monitors was
estimated to be 64.7 percent

STORAGE AND RE-USE OF  TELEVISIONS

To  estimate the storage/re-use rate for CRT televisions, household storage/re-use was assumed to
be representative of all storage and re-use of televisions.  Several of the surveys identified
contain questions about televisions, but only the California Integrated Waste Management Board
e-waste diversion survey included questions that capture both the storage and re-use of
televisions.73  According to the results of this survey, 65.1 percent of respondents gave away,
stored, sold, or traded in televisions that they stopped using. Therefore, the  storage/re-use rate
for CRT televisions was assumed to be 65.1 percent, as indicated in Exhibit  C-2.

Although projection televisions and CRT televisions are similar in several ways, the results of
the California survey were not used to estimate the storage/re-use rate for projection televisions.
As  indicated in the main body of this report, no projection televisions were assumed to be re-
used or put into storage at the end of their first life.

STORAGE AND RE-USE OF  CELL PHONES

The only data source identified with information related to the storage and re-use of cell phones
is the Consumer Electronics Association's (CEA's) re-use and recycling survey. Based on the
results of this  survey, consumers donate or sell approximately  52.9 percent of their unwanted cell
phones. Because the CEA survey asks respondents about cell  phones they removed from their
homes rather than cell phones they no longer use, the results of the survey do not reflect the
72 The California Integrated Waste Management Board's survey asks respondents about monitors in general with no specific mention of CRT or LCD
 monitors. Therefore, we assume that the results of the survey apply to both technologies.

73 The Massachusetts DEP survey asked individuals what they did with their broken televisions; however, because several of the respondents had
 their televisions repaired, the survey includes several televisions that are not at the end of their total life. Therefore, we do not use the results
 of the survey in our analysis.
                                                                                          66

-------
storage of cell phones. Absent data from other sources, the CEA results were used as a proxy for
cell phone storage and re-use.
                                                                                      67

-------
APPENDIX  D: MATERIAL  COMPOSITION OF
SELECT ELECTRONIC PRODUCTS

As indicated in the main body of this report, information on the material composition of several
electronic products was collected to inform our baseline assessment of the generation and
management of electronic waste in the U.S.74 Exhibits D-l through D-7 summarize this
information.  As these exhibits indicate, we were able to obtain detailed constituent data for
desktop and laptop computers,  CRT monitors, flat-panel LCD monitors, CRT televisions, and
cell phones.  We also obtained  composition data for keyboards, although these data are not as
detailed as the information we have compiled for other products.75 We were unable to locate
information on the composition of printers, mice, and projection televisions.

The constituent estimates presented in Exhibits D-l through D-7 indicate that most of the
electronic products included in our analysis include lead. Lead is  most highly concentrated in
CRT products, but it is also found in desktop and laptop computers, LCD  monitors, and cell
phones. The constituent data included in Exhibits D-l through D-7 also show that a number of
valuable metals are included in most electronic products. For example,  CRT monitors, desktops,
and cell phones all contain silver, while desktops and CRT monitors both  contain gold.
74 The data sources from which we obtained these data include Bhuie, A. K., et at., "Environmental and Economic Trade-Offs in Consumer
 Electronic Products Recycling," Proceedings of IEEE 2004 conference on electronic waste; California Department of Toxic Substances Control,
 Hazardous Materials Laboratory, Determination of regulated elements in discarded laptop computers, LCD monitors, Plasma TVs and LCD TVs,
 SB20 Report, December 2004; Five Winds International, Toxic and Hazardous Materials in Electronics, October 2001; Microelectronics and
 Computer Technology Corporation, Electronic Industry Environmental Roadmap, 1996, as cited in Silicon Valley Toxic Coalition (SVTC), Poison PCs
 and Toxic TVs, February 2004; Matsuto, T, et al., "Material and heavy metal balance in a recycling facility for home electrical appliances," Waste
 Management, 24, (2004); RIS International, Ltd., Baseline Study of End-of-Life Electrical and Electronic Equipment in Canada, June 2003; U.S.
 EPA, Desktop Computer Displays: A Life-Cycle Assessment, December 2001; and Williams, Eric. "Environmental Impacts in the Production of
 Personal Computers," Computers and the Environment, edited by Ruediger Kuehr and Eric Williams, Kluwer Academic Publishers, 2003.

75 Lee, C. H., et al., "An overview of recycling and treatment of scrap computers," Journal of Hazardous Materials B114 (2004) 93-100.


                                                                                                   68

-------
  EXHIBIT D-1: MATERIAL COMPOSITION OF A DESKTOP COMPUTER
MATERIAL
AMOUNT CONTAINED IN A DESKTOP (GRAMS)
Metals
   Steel1
   Copper1
   Aluminum1
   Tin1
   Lead1
   Silver1
   Gold1
   Nickel1
   Germanium2
   Gallium2
   Indium2
   Europium2
   Ruthenium2
   Bismuth2
Non-metallic Elements
   Selenium2
   Arsenic2
Plastics1
Epoxy Resin1
Sources:
1. Williams, Eric. "Environmental Impacts in the Production of Personal
Computers," Computers and the Environment, edited by Ruediger Kuehr and Eric
Williams, Kluwer Academic Publishers, 2003.
2. Microelectronics and Computer Technology Corporation, Electronic Industry
Environmental Roadmap, 1996, as cited in Silicon Valley Toxic Coalition (SVTC),
Poison PCs and Toxic TVs, February 2004. This study combines desktops and CRT
monitors into one product category but provides some composition information
specific  to desktops.
             7,254 - 7,524
                6,050
                 670
                 440
                  47
                  27
                 1.4
                 0.36
                  18
                 <45
                 <45
                 <45
                 <45
                 <45
                 <45
              0.44 - 45.44
                 0.44
                 <45
                 650
                1,040
                                                                                    69

-------
EXHIBIT D-2: MATERIAL COMPOSITION OF A CRT MONITOR
PRODUCT


Metals
Steel1
Copper2
Aluminum3
Lead1
Ferrite4
Tin4
Silver4
Gold4

Nickel5
Barium6
Vanadium6
Yttrium6
Glass7
Plastics5
Epoxy Resin4
AVERAGE AMOUNT CONTAINED IN A
CRT MONITOR

GRAMS*
5,683-5,818
3,322
952
242
464
483
20
1.25
0.31

199
<45
<45
<45
6,845
2,481
141
PERCENT OF TOTAL
MASS
36.5°/o-37.4°/o
21 .4%
6.1%
1.6%
3.0%
3.1%
0.13%
0.01%
0.002%

1 .28%
<0.29%
<0.29%
<0.29%
44.0%
16.0%
0.91%
RANGE OF MASS VALUES
IDENTIFIED IN THE LITERATURE

GRAMS*
4,788-7,145
2,850-3,794
705-1,198
199-717
331-597




PERCENT OF TOTAL
MASS
30.8%-45.9%
18.3%-24.4%
4.5%-7.7%
1.3%-4.6%
2.1%-3.8%




Only one value identified.




5,982-6,865
2,235-3,555




38.5%-44.1%
14.4%-22.9%
Only one value identified.
Note:

* The composition estimates presented in grams are scaled from the sources we consulted to be
consistent with an average CRT monitor mass of 34.3 pounds, which is consistent with the
estimate for the 1990-2005 period presented in Exhibit 3-3.
Sources:


1. Based on average of percent of total mass from Williams, Eric. "Environmental Impacts in the
Production of Personal Computers," Computers and the Environment, edited by Ruediger Kuehr
and Eric Williams, Kluwer Academic Publishers, 2003 and U.S. EPA, Desktop Computer Displays:
A Life-Cycle Assessment, December 2001 .
2. Based on average of percent of total mass from Williams (2003) and Menad
tube recycling," Resources, Conservation, and Recycling, Vol. 26, 1999.

N. "Cathode ray

3. Average value based on percent of total mass as derived from Williams (2003). Low-end and
high-end estimates based on U.S. EPA (2001) and Menad (1999), respectively.
4. Based on percent of total mass as derived from Williams (2003).
5. Based on percent of total mass as presented in U.S. EPA (2001 ).



6. Microelectronics and Computer Technology Corporation, Electronic Industry Environmental
Roadmap, 1996, as cited in Silicon Valley Toxic Coalition (SVTC), Poison PCs and Toxic TVs,
February 2004. This study combines desktops and CRT monitors into one product category but
provides some composition information specific to CRT monitors.

7. Average value based on percent of total mass as presented in U.S. EPA (2001 ). Low-end and
high-end estimates based on Menad (1999) and Williams (2003), respectively.

8. Average value based on percent of total mass as derived from Menad (1999). Low -end and
high-end estimates based on U.S. EPA (2001) and Williams (2003), respectively.
                                                                 70

-------
EXHIBIT D-3: PRIMARY MATERIAL INPUTS FOR A 15-INCH LCD MONITOR
MATERIAL
Metals
Steel
Aluminum (heat sink)
Solder (60% tin, 40% lead)
Transition metals, other (Mo, Ti, W)
Mercury
Plastics
Glass
Miscellaneous Compounds
Polyvinyl alcohol
Indium tin oxide (ITO) (electrode)
Polyimide alignment layer
Other
Color filter pigment
Liquid crystals
Backlight lamp (CCFL)
AMOUNT CONTAINED
(GRAMS)
2,702
2,530
130
40
1.9
0.00399
1,780
590
11
10
0.5
0.5
44.2
40
2.3
1.9
PERCENT OF TOTAL
MASS
47.15%
44.12%
2.34%
0.66%
0.03%
0.0001%
30.98%
10.31%
0.17%
0.15%
0.01%
0.01%
0.72%
0.65%
0.04%
0.03%
Source:
EPA, "Desktop Computer Displays: A Life-Cycle Assessment," EPA-744-R-01-004a, December
2001.
                                                                      71

-------
EXHIBIT D-4: MATERIAL COMPOSITION OF A CELL PHONE
MATERIAL
Metals
Lead
Aluminum
Iron
Tin
Copper
Nickel
Zinc
Silver
Mercury
Plastics
Silica
Note:
AMOUNT CONTAINED
(GRAMS)*
58
1
12
11
1
26
1
4
1
1
63
5
PERCENT OF TOTAL
MASS
43.8%
0.9%
9%
8%
1%
19%
1%
3%
0.9%
1%
46%
4%

* Assuming a cell phone average weight of 1 36 g, consistent with the average mass of a cell
phone during the 2000-2005 period, as presented in Exhibit 3-3.
Source:
Bhuie, A. K., et al., "Environmental and Economic Trade-Offs in Consumer Electronic Products
Recycling: A case study of cell phones and computers," IEEE, 2004, 75.
EXHIBIT D-5: COMPONENTS IN A TYPICAL PC KEYBOARD
COMPONENT
Shell
Top plate
Bottom plate
1C Board
Button
Wire
MATERIAL
Plastic
Plastic
Plastic
1C, resin, copper, iron
Plastic
Copper, plastic
Total
Source:
AMOUNT CONTAINED
(GRAMS)
348
118
230
384
116
70
918
PERCENT OF TOTAL
MASS
37.91%
12.85%
25.06%
41.83%
12.63%
7.63%
100.00%

Lee, C. H., et al., "An overview of recycling and treatment of scrap computers," Journal of
Hazardous Materials B114 (2004) 93-100.
                                                              72

-------
EXHIBIT D-6: MATERIALS IN A CRT TELEVISION
MATERIAL
Metals
Steel/Iron1
Lead1-2
Copper1'2
Aluminum1
Zinc2
Tin2
Cadmium2
Chromium2
Antimony2
G/ass1
P/ast/c1
Note:
AMOUNT CONTAINED
(GRAMS)*
4,107-4,262
2,088
1,291-1,347
606-705
67
8.6
31.6
0.2
0.03
14.4
15,760
8,755
PERCENT OF TOTAL
MASS
14.3°/o-14.8°/o
7.3%
4.5°/o-4.7°/o
2.1°/o-2.5°/o
0.23%
0.03%
0.11%
0.001%
0.0001%
0.05%
54.8%
30.4%

* We scale the mass values presented in the sources listed below to be consistent with our
estimate of the average mass of a CRT TV during the 1990-2005 period as reported in Exhibit 3-3
(28.8 kg or 63.4 pounds).
Sources:

1. RIS International, Ltd., Baseline Study of End-of-Life Electrical and Electronic Equipment in
Canada, June 2003, Table 4.2. This table presents the composition of a 42.7-kg 28-inch CRT
TV.
2. Matsuto, T., et al., "Material and heavy metal balance in a recycling facility for home
electrical appliances," Waste Management 24 (2004), 434. This study presents grams of
material (e.g., cadmium) per kg of CRT TV mass.
                                                             73

-------
EXHIBIT D-7: MATERIALS IN A LAPTOP LCD PANEL AND PC BOARD
MATERIAL
Antimony
Arsenic
Barium
Chromium
Copper
Lead
Mercury2
Molybdenum
Nickel
Silver
Zinc
Notes:
GRAMS PER KG OF PRODUCT1
0.22
0.003
0.70
0.02
38.7
1.5
-
0.01
1.03
0.07
0.001
GRAMS PER LAPTOP, ASSUMING
A LAPTOP MASS OF 3.5 KG
0.77
0.01
2.45
0.07
135.45
5.25
0.00012-0.0005
0.04
3.61
0.25
0.004

1 . Except for mercury, these values represent the average composition of four laptops analyzed
in California Department of Toxic Substances Control, Hazardous Materials Laboratory,
Determination of regulated elements in discarded laptop computers, LCD monitors, Plasma TVs
and LCD TVs, SB20 Report, December 2004.
2. Based on various sources cited in Five Winds International, Toxic and Hazardous Materials in
Electronics, October 2001 .
                                                                  74

-------
APPENDIX E:  MATERIAL COMPOSITION OF RETIRED ELECTRONIC PRODUCTS

   This appendix presents estimates of the total quantity (i.e., mass) of individual constituents
included in the electronic waste stream in 2003, 2004, and 2005, by management method. These
estimates were developed by multiplying the product retirement estimates generated by the waste
flow model described in the main body of this report by the per unit constituent information
presented in Appendix D.  The results are presented  both in aggregate and for individual
products. It is important to note, however, that the aggregate estimates only reflect constituents
included in those products for which we were able to identify material composition information
(i.e., desktops, laptops, CRT monitors, LCD monitors, CRT televisions, cell phones, and
keyboards).  Constituents included in other products (e.g., mice) are not reflected in the results
presented in this appendix.

-------
EXHIBIT E-1: MATERIAL COMPOSITION OF AGGREGATE E-WASTE VOLUME FOR ELECTRONIC PRODUCTS BY MANAGEMENT
                                        METHOD (IN POUNDS)*

MATERIAL
Aluminum
Antimony
Arsenic
Backlight lamp (CCFL)
Barium
Bismuth
Cadmium
Chromium
Color filter pigment
Copper
Epoxy Resin
Europium
Ferrite
Gallium
Germanium
Glass
Gold
Indium
Indium tin oxide (ITO)
Iron
Lead
Liquid crystals
Mercury
Molybdenum
Nickel
Plastics
Polyimide alignment layer
LANDFILLED
2003
26,431,609
610,576
1,287,810
256
1,804,774
1,287,709
12,064
1,877
5,543
88,258,064
35,349,969
1,287,709
19,041,421
1,287,709
1,287,709
931,457,141
22,586
1,287,709
85
1,082,829
74,866,062
341
135,357
335
8,547,251
525,330,685
85
2004
26,498,975
627,803
1,349,293
800
1,660,163
1,349,179
12,389
1,997
17,331
87,330,456
36,307,212
1,349,179
17,462,811
1,349,179
1,349,179
927,062,970
22,060
1,349,179
267
1,228,467
74,894,463
1,067
153,564
379
7,942,740
534,412,179
267
2005
27,413,819
649,233
1,380,765
2,152
1,689,663
1,380,633
12,790
2,164
46,630
90,066,155
37,112,743
1,380,633
17,730,607
1,380,633
1,380,633
953,283,703
22,484
1,380,633
717
1,469,765
77,052,109
2,870
183,732
443
8,102,602
552,676,242
717
RECYCLED
2003
8,608,041
98,382
467,980
86
604,378
467,943
1,914
435
1,854
25,004,088
12,684,407
467,943
6,369,382
467,943
467,943
195,320,967
7,853
467,943
29
265,136
15,297,215
114
33,143
122
2,862,813
112,229,943
29
2004
8,620,096
101,312
490,322
268
556,094
490,281
1,966
472
5,797
24,562,575
13,045,649
490,281
5,841,335
490,281
490,281
190,716,659
7,691
490,281
89
300,796
15,043,390
357
37,602
138
2,659,812
113,625,481
89
2005
8,894,458
104,998
501,759
720
566,090
501,711
2,029
525
15,598
25,284,892
13,336,101
501,711
5,930,913
501,711
501,711
195,626,819
7,840
501,711
240
359,879
15,439,439
960
44,989
161
2,711,134
117,524,174
240
INCINERATED
2003
751,906
17,369
36,635
7
51,341
36,632
343
53
158
2,510,697
1 ,005,609
36,632
541,676
36,632
36,632
26,497,375
643
36,632
2
30,803
2,129,732
10
3,851
10
243,146
14,944,203
2
2004
753,822
17,859
38,384
23
47,227
38,380
352
57
493
2,484,310
1,032,840
38,380
496,769
38,380
38,380
26,372,373
628
38,380
8
34,946
2,130,540
30
4,368
11
225,949
15,202,546
8
2005
779,847
1 8,469
39,279
61
48,066
39,275
364
62
1,326
2,562,133
1,055,755
39,275
504,387
39,275
39,275
27,118,281
640
39,275
20
41,811
2,191,919
82
5,227
13
230,497
15,722,108
20

-------

MATERIAL
Polyvinyl alcohol
Ruthenium
Selenium
Silica
Silver
Solder (60% tin; 40% lead)
Steel
Steel/Iron
Tin
Transition Metals, other
(Mo, Ti, W)
Vanadium
Yttrium
Zinc
Total
LANDFILLED
2003
1,279
1,287,709
12,591
541,414
225,652
5,629
304,948,841
88,066,758
3,605,839
256
1,781,294
1,781,294
768,013
2,123,711,840
2004
4,000
1,349,179
13,192
614,233
239,163
17,598
303,115,768
90,441,248
3,657,826
800
1,633,618
1,633,618
832,389
2,127,286,180
2005
10,761
1,380,633
13,500
734,882
268,596
47,347
311,181,864
93,365,913
3,776,140
2,152
1,658,670
1,658,670
934,902
2,190,839,299
RECYCLED
2003
428
467,943
4,575
132,568
65,786
1,883
107,007,493
13,972,230
999,526
86
595,845
595,845
156,858
506,199,091
2004
1,338
490,281
4,794
150,398
68,900
5,886
106,633,087
14,348,954
1,010,847
268
546,447
546,447
171,781
502,048,822
2005
3,599
501,711
4,906
179,940
76,353
15,838
109,453,378
14,812,967
1,040,919
720
554,827
554,827
195,846
516,258,342
INCINERATED
2003
36
36,632
358
15,402
6,419
160
8,674,950
2,505,255
102,576
7
50,673
50,673
21,848
60,413,719
2004
114
38,380
375
17,473
6,804
501
8,622,804
2,572,803
104,055
23
46,472
46,472
23,679
60,515,399
2005
306
39,275
384
20,905
7,641
1,347
8,852,262
2,656,002
107,421
61
47,185
47,185
26,595
62,323,309
* Based on available material composition information for desktops, laptops, CRT monitors, LCD monitors, CRT televisions, cell phones, and keyboards.
77

-------
EXHIBIT E-2: MATERIAL COMPOSITION OF AGGREGATE E-WASTE VOLUME FOR DESKTOPS BY MANAGEMENT METHOD
                                            (IN POUNDS)

MATERIAL
Aluminum
Arsenic
Bismuth
Copper
Epoxy Resin
Europium
Gallium
Germanium
Gold
Indium
Lead
Nickel
Plastics
Ruthenium
Selenium
Silver
Steel
Tin
Total
LANDFILLED
2003
12,590,934
1,287,709
1,287,709
19,172,559
29,760,390
1,287,709
1,287,709
1,287,709
10,302
1,287,709
772,626
515,084
18,600,244
1,287,709
12,591
40,062
173,125,347
1,344,941
264,959,044
2004
13,191,975
1,349,179
1,349,179
20,087,780
31,181,032
1,349,179
1,349,179
1,349,179
10,793
1,349,179
809,508
539,672
19,488,145
1,349,179
13,192
41,974
181,389,655
1,409,143
277,607,123
2005
13,499,518
1,380,633
1,380,633
20,556,084
31,907,952
1,380,633
1,380,633
1,380,633
11,045
1,380,633
828,380
552,253
19,942,470
1,380,633
13,500
42,953
185,618,374
1 ,441 ,994
284,078,950
RECYCLED
2003
4,575,444
467,943
467,943
6,967,153
10,814,685
467,943
467,943
467,943
3,744
467,943
280,766
187,177
6,759,178
467,943
4,575
14,558
62,912,349
488,741
96,283,970
2004
4,793,857
490,281
490,281
7,299,737
11,330,934
490,281
490,281
490,281
3,922
490,281
294,168
196,112
7,081,834
490,281
4,794
15,253
65,915,531
512,071
100,880,180
2005
4,905,616
501,711
501,711
7,469,915
11,595,091
501,711
501,711
501,711
4,014
501,711
301,026
200,684
7,246,932
501,711
4,906
15,609
67,452,214
524,009
103,231,989
INCINERATED
2003
358,177
36,632
36,632
545,406
846,601
36,632
36,632
36,632
293
36,632
21,979
14,653
529,125
36,632
358
1,140
4,924,937
38,260
7,537,351
2004
375,275
38,380
38,380
571,442
887,014
38,380
38,380
38,380
307
38,380
23,028
15,352
554,384
38,380
375
1,194
5,160,033
40,086
7,897,154
2005
384,024
39,275
39,275
584,764
907,693
39,275
39,275
39,275
314
39,275
23,565
15,710
567,308
39,275
384
1,222
5,280,329
41,021
8,081,259
                                                                                                   78

-------
EXHIBIT E-3: MATERIAL COMPOSITION OF AGGREGATE E-WASTE VOLUME FOR LAPTOPS BY MANAGEMENT METHOD
                                           (IN POUNDS)

MATERIAL
Antimony
Arsenic
Barium
Chromium
Copper
Lead
Mercury
Molybdenum
Nickel
Silver
Zinc
Total
LANDFILLED
2003
7,380
101
23,480
671
1,298,123
50,315
3
335
34,550
2,348
34
1,417,339
2004
8,343
114
26,545
758
1,467,575
56,883
3
379
39,059
2,655
38
1,602,352
2005
9,741
133
30,993
886
1,713,462
66,413
4
443
45,604
3,099
44
1,870,821
RECYCLED
2003
2,682
37
8,533
244
471,728
18,284
1
122
12,555
853
12
515,050
2004
3,032
41
9,646
276
533,305
20,671
1
138
14,194
965
14
582,282
2005
3,540
48
11,263
322
622,658
24,134
1
161
16,572
1,126
16
679,841
INCINERATED
2003
210
3
668
19
36,928
1,431
0
10
983
67
1
40,319
2004
237
3
755
22
41,748
1,618
0
11
1,111
76
1
45,582
2005
277
4
882
25
48,743
1,889
0
13
1,297
88
1
53,220
                                                                                                  79

-------
EXHIBIT E-4: MATERIAL COMPOSITION OF AGGREGATE E-WASTE VOLUME FOR CRT MONITORS BY MANAGEMENT METHOD
                                             (IN POUNDS)

MATERIAL
Aluminum
Barium
Copper
Epoxy Resin
Ferrite
Glass
Gold
Lead
Nickel
Plastics
Silver
Steel
Tin
Vanadium
Yttrium
Total
LANDFILLED
2003
9,827,830
1,781,294
37,468,603
5,589,578
19,041,421
270,265,331
12,285
18,427,182
7,862,264
98,278,302
61 ,424
131,447,229
798,511
1,781,294
1,781,294
604,423,843
2004
9,013,064
1,633,618
34,362,305
5,126,180
17,462,811
247,859,250
11,266
16,899,494
7,210,451
90,130,636
56,332
120,549,726
732,311
1,633,618
1,633,618
554,314,680
2005
9,151,281
1,658,670
34,889,260
5,204,791
17,730,607
251,660,235
1 1 ,439
17,158,652
7,321,025
91,512,813
57,196
122,398,387
743,542
1,658,670
1,658,670
562,815,237
RECYCLED
2003
3,287,423
595,845
12,533,300
1,869,722
6,369,382
90,404,133
4,109
6,163,918
2,629,938
32,874,230
20,546
43,969,283
267,103
595,845
595,845
202,180,626
2004
3,014,882
546,447
1 1 ,494,239
1,714,714
5,841,335
82,909,268
3,769
5,652,905
2,411,906
30,148,825
18,843
40,324,053
244,959
546,447
546,447
185,419,040
2005
3,061,116
554,827
11,670,507
1,741,010
5,930,913
84,180,703
3,826
5,739,593
2,448,893
30,611,165
19,132
40,942,433
248,716
554,827
554,827
188,262,489
INCINERATED
2003
279,575
50,673
1,065,878
159,008
541,676
7,688,300
349
524,202
223,660
2,795,745
1,747
3,739,310
22,715
50,673
50,673
17,194,184
2004
256,397
46,472
977,512
145,826
496,769
7,050,909
320
480,744
205,117
2,563,967
1,602
3,429,306
20,832
46,472
46,472
15,768,717
2005
260,329
47,185
992,503
148,062
504,387
7,159,037
325
488,116
208,263
2,603,286
1,627
3,481,895
21,152
47,185
47,185
16,010,534
                                                                                                     80

-------
EXHIBIT E-5: MATERIAL COMPOSITION OF AGGREGATE E-WASTE VOLUME FOR LCD MONITORS BY MANAGEMENT METHOD
                                              (IN POUNDS)

MATERIAL
Aluminum
Backlight lamp (CCFL)
Color filter pigment
Glass
Indium tin oxide (ITO)
Liquid crystals
Mercury
Plastics
Polyimide alignment layer
Polyvinyl alcohol
Solder (60% tin; 40% lead)
Steel
Transition Metals, other
(Mo, Ti, W)
Total
LANDFILLED
2003
19,956
256
5,543
87,926
85
341
1
264,204
85
1,279
5,629
376,265
256
761 ,826
2004
62,392
800
17,331
274,899
267
1,067
3
826,031
267
4,000
17,598
1,176,387
800
2,381,840
2005
167,868
2,152
46,630
739,624
717
2,870
7
2,222,459
717
10,761
47,347
3,165,103
2,152
6,408,409
RECYCLED
2003
6,675
86
1,854
29,411
29
114
0
88,377
29
428
1,883
125,861
86
254,832
2004
20,870
268
5,797
91,954
89
357
1
276,308
89
1,338
5,886
393,503
268
796,729
2005
56,152
720
15,598
247,405
240
960
2
743,416
240
3,599
15,838
1,058,731
720
2,143,622
INCINERATED
2003
568
7
158
2,501
2
10
0
7,516
2
36
160
10,704
7
21,672
2004
1,775
23
493
7,820
8
30
0
23,498
8
114
501
33,465
23
67,757
2005
4,775
61
1,326
21,040
20
82
0
63,223
20
306
1,347
90,038
61
182,301
                                                                                                      81

-------
EXHIBIT E-6: MATERIAL COMPOSITION OF AGGREGATE E-WASTE VOLUME FOR CRT TELEVISIONS BY MANAGEMENT METHOD
                                               (IN POUNDS)

MATERIAL
Aluminum
Antimony
Cadmium
Chromium
Copper
Glass
Lead
Plastic
Steel/Iron
Tin
Zinc
Total
LANDFILLED
2003
2,774,706
603,197
12,064
1,206
27,747,061
661,103,884
55,494,122
366,743,761
88,066,758
1,327,033
361,918
1,204,235,711
2004
2,849,519
619,461
12,389
1,239
28,495,188
678,928,821
56,990,375
376,632,047
90,441,248
1,362,813
371,676
1,236,704,776
2005
2,941,666
639,493
12,790
1,279
29,416,658
700,883,844
58,833,315
388,811,475
93,365,913
1,406,884
383,696
1,276,697,012
RECYCLED
2003
440,221
95,700
1,914
191
4,402,209
104,887,422
8,804,419
58,185,723
13,972,230
210,540
57,420
191,057,990
2004
452,090
98,281
1,966
197
4,520,903
107,715,437
9,041,807
59,754,549
14,348,954
216,217
58,968
196,209,369
2005
466,710
101,459
2,029
203
4,667,099
111,198,711
9,334,198
61,686,876
14,812,967
223,209
60,875
202,554,337
INCINERATED
2003
78,933
17,159
343
34
789,327
18,806,574
1,578,654
10,432,844
2,505,255
37,750
10,296
34,257,170
2004
81,061
17,622
352
35
810,609
19,313,644
1,621,218
10,714,138
2,572,803
38,768
10,573
35,180,825
2005
83,682
18,192
364
36
836,822
19,938,204
1,673,645
1 1 ,060,609
2,656,002
40,022
10,915
36,318,493
                                                                                                       82

-------
EXHIBIT E-7: MATERIAL COMPOSITION OF AGGREGATE E-WASTE VOLUME FOR CELL PHONES BY MANAGEMENT METHOD
                                             (IN POUNDS)

MATERIAL
Aluminum
Copper
Iron
Lead
Mercury
Nickel
Plastics
Silica
Silver
Tin
Zinc
Total
LANDFILLED
2003
1,218,182
2,571,718
1,082,829
121,818
135,354
135,354
6,226,266
541,414
121,818
135,354
406,061
12,696,168
2004
1,382,025
2,917,609
1,228,467
138,203
153,558
153,558
7,063,684
614,233
138,203
153,558
460,675
14,403,773
2005
1,653,485
3,490,691
1,469,765
165,349
183,721
183,721
8,451,147
734,882
165,349
183,721
551,162
17,232,991
RECYCLED
2003
298,278
629,698
265,136
29,828
33,142
33,142
1,524,532
132,568
29,828
33,142
99,426
3,108,720
2004
338,396
714,391
300,796
33,840
37,600
37,600
1,729,578
150,398
33,840
37,600
112,799
3,526,836
2005
404,864
854,713
359,879
40,486
44,985
44,985
2,069,305
179,940
40,486
44,985
134,955
4,219,584
INCINERATED
2003
34,654
73,158
30,803
3,465
3,850
3,850
177,120
15,402
3,465
3,850
11,551
361,171
2004
39,315
82,998
34,946
3,931
4,368
4,368
200,942
17,473
3,931
4,368
13,105
409,747
2005
47,037
99,300
41,811
4,704
5,226
5,226
240,412
20,905
4,704
5,226
15,679
490,231
                                                                                                    83

-------
EXHIBIT E-8: MATERIAL COMPOSITION OF AGGREGATE E-WASTE VOLUME FOR KEYBOARDS BY MANAGEMENT METHOD
                                            (IN POUNDS)

MATERIAL
Copper/plastic (Wire)
IC/resin/copper/iron (1C
Board)
Plastic (Button)
Plastic (Shell)
Total
LANDFILLED
2003
5,316,831
29,148,499
8,800,993
26,416,916
69,683,239
2004
6,079,790
33,331,273
10,063,925
30,207,711
79,682,699
2005
6,300,846
34,543,170
10,429,841
31,306,038
82,579,894
RECYCLED
2003
1,932,093
10,592,328
3,198,210
9,599,693
25,322,325
2004
2,209,346
12,112,315
3,657,149
10,977,238
28,956,047
2005
2,289,676
12,552,708
3,790,120
11,376,361
30,008,864
INCINERATED
2003
151,249
829,194
250,364
751 ,488
1,982,295
2004
172,953
948,182
286,291
859,326
2,266,752
2005
179,242
982,657
296,700
890,570
2,349,169
                                                                                                    84

-------