,. t.
c. .0 .
v.a)h9IOUhli..
iii... .
1!, . an?
.

92 ‘53::

aw.» Hakka. .r .ih
nirmne. awn!
h '9

.11 l.

91.
"Kai... . .
maisidptv «V In".
. .11
.
l

Cl; An...
) l .L.

.5 ‘
(.1 “urinal .l
fiu.u.«‘.islorlulhnw.l!fi

.!.£nmn.).2..nxll¢

.v 5......

_ {Jinan

nfiwfi

winnhodlx an... ..<. 5.4mm, , . , . .V humm...nwn..rnmwtx.ur
.iutfitawmfismgmri. . .. ‘ . .. . .1. . .. . «fifiMw

 

, chfimmmilJfi ‘ «Mllfiflwuf ~Wu

 

. 1‘ . I .37 IIV . C 14 D... 5. ;
‘ ._.Ml| n.c.u1axd.fl0.d»“u¥; 353.45% -glllfi». 336:5, . .. i1. y , I 2!, 5., w.

 

 

{HESlS

SIIRTYLB RIEAR

ill\lllllllllllllllllllllllllllll zllllllmll

3 1293 015

 

 

 

 

 

This is to certify that the

dissertation entitled

University Waste Characterization
Using Manual Sorting Methods:
A Michigan State University Case Example

presented by l

Marsha M. Crawford

has been accepted towards fulfillment
of the requirements for

Ph.D degree inResmimmmlopment

Major professor

MS U is an Affirmative Argon/Equal Opportunity Institution 0-12771

'

 

 

LIBRARY

Michigan State
University

 

 

 

PLACE IN RETURN 80X to remove thle checkout from your record.

7 TO AVOID F INES retum on or More dete due.

DATE DUE DATE DUE DATE DUE

 

wovzrfizo 3 °

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU IeAnNfltmetIve WM Oppommlylnetltwon

UNIVERSITY WASTE CHARACTERIZATION
USING MANUAL SORTING METHODS:
A MICHIGAN STATE UNIVERSITY CASE EXAMPLE
By

Marsha M. Crawford

A DISSERTATION

Submitted to
Michigan State University
in partial fiIlfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Resource Development

1996

ABSTRACT

UNIVERSITY WASTE CHARACTERIZATION
USING MANUAL SORTING METHODS:
A MICHIGAN STATE UNIVERSITY
CASE EXAMPLE

By

Marsha M. Crawford

This study was conducted to provide a manual sorting model for characterization
of university solid waste streams, using Michigan State University as an example.
Analyzing university waste streams provides a unique challenge, as universities have some
of the characteristics of both municipal and industrial waste streams, but are more difficult
to successfully quantify using either materials flow methodologies or per-capita
estimations. Ifuniversities are to meet state and national waste reduction goals,
characterizing the components of the waste stream is an essential first step in planning
integrated waste management strategies.

The Michigan State University campus buildings were divided into three strata
based on the population characteristics of different segments of the campus: residence
halls, university apartments, and academic/administrative facilities. A minimum of twenty
80-200 pound samples were taken from each of these three waste streams within a two
week period both summer and fall semesters, 1994. The samples were manually sorted

into nine general material categories and the weights of each were recorded. The weights

from these samples were aggregated and analyzed to provide information about the
potential diversion rates for commonly recycled materials. With the addition of
information from existing campus recycling programs, the university waste streams were
compared to the EPA/Franklin materials flow model for municipal waste streams to

highlight differences in composition.

Copyright by
MARSHA MARIE CRAWFORD
1996

ACKNOWLEDGMENTS

I would like to thank the members of my committee, Dr. Eckhart Dersch, Dr.
Cynthia Fridgen, and Dr. Michael Kamrin, for their time and thoughtfiil guidance during
the writing of this document. Special thanks to my committee chair, Dr. Scott Witter, for
his unfailing support and patient editing.

This research would not have been possible without fimding through the Office of
Recycling and Waste Reduction, and the forethought of Peter Pasterz, the director, who
gave me the opportunity to work on this project. His vision is appreciated, as is the work
put in by the recycling office staff and the students in the Department of Resource
Development. Thanks is also extended to Resource Recycling Systems of Ann Arbor,
especially Cheri DeRosario, for technical support and assistance. Although not directly a
part of this research, the staff and student employees of Case Hall Cafeteria gave me my
first opportunity to study trash, and I am gratefiil for their assistance.

Finally, a special thanks to my extended family who have provided a ready ear, and

to my husband Dan and son Karl, who remind me of what is truly important.

TABLE OF CONTENTS

Base_
List of Tables ......................................... ix
List of Figures ........................................ x
Chapter One: Introduction ............................... 1
Municipal Waste Reduction ........................ 5
Industrial Waste Reduction Programs ................ 8
The Role of Universities .......................... 9
Waste Characterization Methodologies ............... 10
Characterizing University Waste Streams ............. 13
Problem Statement .............................. 17
Objectives of Study ............................. 18
Hypotheses ................................... 18
Chapter Two: Review of Literature ....................... 20
Goals of Waste Characterization ................... 20
State and Federal Waste Reduction Guidelines ......... 22
Using Waste Characterization to Meet State and Federal
Guidelines .................................... 24
Recommended Sampling and Evaluation Protocols ....... 25
Choosing Categories of Materials ................... 29
Manual Sorting Considerations ..................... 30

vi

University Waste Characterization Methodologies ....... 32

Chapter Three: Description of Study Method ................ 35
Description of the Research Site ..................... 35
Material Categories .............................. 36
Timing of Study ................................. 37
Sampling Procedure ............................. 37
Manual Sorting Protocol .......................... 40
Procedures for Analysis of Data .................... 48
Limitations and Key Assumptions ................... 50

Collection and Sorting of Samples ............. 50
Data Assumptions ......................... 52
Regulatory and Structural Limitations .......... 53

Chapter Four: Data Analysis ............................ 55

Review of Sampling Protocol ........................ 55
Material Categories ........................ 55
Sample Selection ........................... 55
Sorting Procedure ........................... 57

Evaluating the Effectiveness of a Manual Sorting Model . . . 58

Estimates of Current Diversions From the Waste Stream by

Strata ......................................... 61
University Apartments ....................... 62
Academic and Administrative Buildings ........... 63

vii

Residence Halls ........................... 63
Potential For Additional University Waste Stream Diversion .64

Comparison with the 1994 EPA/Franklin Estimates
for Municipal Solid Waste ......................... 67

Chapter Five: Conclusions .............................. 72

Evaluating the Effectiveness of a Manual Sorting Model for
University Solid Waste Characterization ............... 73

Evaluating the Effectiveness of a Manual Sorting Model in
Predicting Major Components of the University Solid Waste
Stream ........................................ 73

Evaluating the Effectiveness of a Manual Sorting Protocol
Used to Gather Sampling Data ...................... 76

Targeting Further Waste Reduction Efforts ............ 78
Comparison of University and Municipal Waste Streams . . . 80
Recommendations for Further Research ............... 82

Appendix A: Sample Worksheet for Recording Weights of

Material Categories ........................ 83
Appendix B: Routing Schedule for Grounds Department ....... 85
Appendix C: Sample Schedule for Student Employees ......... 86
Appendix D: Waste Sort Training Information ................ 87

Bibliography ........................................... 91

viii

LIST OF TABLES

Table 1. Categories of Materials for Waste Characterization ..... 56

Table 2. Breakdown of Sample Characteristics by
Strata and Semester .................................... 58

Table 3. Percentage Means and Percent Standard Error
for Major Components of the Michigan State University
Waste Stream by Semester and Strata ...................... 60

Table 4. Percentage Means and Percent Standard Error
for Major Components of the Michigan State University
Waste Stream ........................................ 61

Table 5. Waste Stream Diversion from University
Apartments, 1994-95 .................................. 62

Table 6. Waste Stream Diversion from Academic and
Administrative Buildings, 1994-95 ........................ 63

Table 7. Waste Stream Diversions from Residence Halls,
1994-95 ............................................. 64

Table 8. Estimates of Current and Potential Diversion
From the Michigan State University Waste Stream by Strata ...... 65

Table 9. Incorporating the Michigan State University

Waste Characterization Categories Into the Category

Divisions of the EPA/Franklin Model for Municipal

Waste Streams ....................................... 66

Table 10. Comparison of Michigan State University’s
Waste Stream With EPA/Franklin Municipal Figures,
Using the Student T-test ................................ 69

Table 11. Comparison of Michigan State University’s

Waste Stream With EPA/Franklin Municipal Figures
By Strata, Using the Student T-test ........................ 71

ix

LIST OF FIGURES

Figure l. Sectioning Procedure for Sample Selection ........... 39
Figure 2. Diagram of Sorting Site ......................... 41
Figure 3. Material Categories by Container Size and Arrangement

for Sorting ........................................... 42
Figure 4. Transferring a Sample to the Sort Table .............. 46
Figure 5. Student Employees Sorting a Sample ............... 47

, Figure 6. Estimates of Current and Potential Diversion
from the Michigan State University Waste Stream by
Strata .............................................. 65

Figure 7. Comparison of Michigan State University’s Solid Waste
Stream with EPA/Franklin Municipal Solid Waste Stream Estimates
by Material Category Percentage of Total ................... 7O

CHAPTER ONE:

INTRODUCTION ’

In 1971, William Small wrote a book entitled “The Third Pollution,” enumerating
the concern with solid waste disposal and its impact on the environment, as well as the
loss of resources represented by landfilling or incinerating materials. He included a
quotation from then Senator Edmund Muskie saying in part:

Ifwe are to preserve and enhance the quality of our environment, we must find

ways to reduce the quantities of solid waste and lessen the burden on our air and

water resources. (p. 90)

The country as a whole has struggled with this sentiment for over 20 years, agreeing with
Muskie in principle, but finding it difficult to put his words into action. The history of
solid waste legislation illustrates this struggle to find ways to lessen the environmental
burden as well as the risk to human health. The first major legislative response to
increasing amounts of consumer and industrial waste was the Solid Waste Disposal Act
(SWDA) of 1965, which encouraged more environmentally sound methods for waste
disposal. Responsibility for solid waste programs was shified in 1970 from the Department
of Health, Education, and Welfare (HEW) to the newly-created Environmental Protection
Agency (EPA). The Resource Recovery Act, passed the same year, provided some
financial support for solid waste initiatives as well as urging the recovery of energy and
materials as part of efl‘ective waste management programs. Until the passage of the 1976

amendment titled the Resource Conservation and Recovery Act (RCRA), however, solid

waste was still seen primarily as a local issue. Continuing concern over the environment

and human health prompted the federal legislature to include as part of RCRA the
involvement of the EPA in the regulation of waste disposal facilities, both for municipal
solid waste and for hazardous waste. The primary goals of RCRA were to:

1) protect human health and the environment fi'om the potential hazards of waste

disposal,

2) conserve energy and natural resources,

3) reduce the amount of waste generated, including hazardous waste, and

. 4) ensure that wastes are managed in an environmentally sound manner.
(Christensen, 1991, p. 569)

From these goals emerged a “waste management hierarchy,” a list of waste
management alternatives ordered by their ability to conserve resources and protect human
health and the environment. Listed from greatest perceived benefit to least, the hierarchy
includes: source reduction, reuse of materials, recycling (and composting), incineration,
and landfilling. (Young, 1994) Despite the goals of RCRA, and continuing legislative
efforts to reclaim more of the materials used in this country, about 75% of the nation’s
waste stream is still being landfilled. (Wagner, 1995)

The greatest benefit, for both the environment and the economy, is to produce less
waste in the first place. Source reduction, the reduction or elimination of waste prior to
generation, can be accomplished in several ways. (WMI, 1990) One approach is simply to
reduce the amount of waste being generated, through improvements in manufacturing or
packaging technology and more careful inventory practices. Source reduction can also
address waste toxicity by promoting the replacement of a manufacturing material known

to adversely affect human health with one that is less toxic. A third source reduction

strategy is the manufacture and purchase of products with a longer useful life, to decrease

the proliferation of one-use disposable consumer goods. Source reduction strategies can
reduce the amounts of hazardous materials that must be disposed of or otherwise managed
to reduce the risk to human health and the environment. (Williams, 1990)

The second step in the waste hierarchy, reuse of materials, is generally defined as
using a discarded item again for the same purpose it was originally intended. Examples
are the reuse of shipping containers or wooden pallets, as well as finding “new homes” for
outdated machinery instead of discarding it. This has benefits both for the company that
avoids disposal fees and the recipient of the still-useful good, who avoids the capital
outlay necessary to purchase a newer version. Although the materials will eventually need
to be discarded, reusing items instead of making new ones enables the use of fewer
resources without negative environmental impacts.

Recycling is, by comparison, the processing of discarded materials into new
products, such as using crushed glass to make new containers, or melting scrap plastic to
manufacture park benches. Although recycling does not represent as great an economic
benefit as source reduction, and carries some environmental risk, it is seen as a viable
means for reducing the amount of waste that must be landfilled or incinerated at any one
time. In most cases, the recycling of waste materials into new products conserves raw
materials, uses less energy and water than virgin materials, and presents less of an
environmental concern than those waste management options lower on the hierarchy.
(York, 1990) Some materials that currently have recycling markets within the United
States include paper products, glass, metals, construction materials, and some types of

plastics.

Composting is usually included as a subcategory of recycling, and involves the
managed decomposition of discarded organic material, primarily for use as a soil
amendment. (Martin, 1993) Materials that are commonly considered to be compostable
include yard and landscape waste, food waste, animal waste and paper products.
Composting operations can range from backyard “compost piles” composed of leaves and
grass clippings to large indoor facilities with rotating drums to aerate a complex mix of
organic materials.

A less desirable option in the waste management hierarchy is incineration. Also
referred to as combustion, it is the controlled burning of waste materials either as a
method of reducing the volume of waste and/or to generate energy. It can also help
destroy pathogens and some toxic chemicals found in the waste stream. (York, 1990)
Although some materials generate energy through the combustion process, the majority
of waste materials save more energy through the recycling process than is given up during
combustion. (Young, 1994) The ash generated, moreover, must then be disposed of in
dedicated landfills. The fly ash, in particular, is of concern because of the concentrated
accumulation Of heavy metals and dioxins that attach to ash particles, materials that are
considered by many to be toxic to humans.

Landfills, once no more than open pits, have improved dramatically in the last
twenty years. Referred to as sanitary landfills, modern land-based disposal systems are
designed to reduce the interaction of discarded materials with the surrounding land, water,
and air. Some of the devices that are utilized to prevent this interaction include leachate

collection systems, plastic and composite liners and covers, and in some cases, methane

collection systems. Although landfill facilities are much more carefiilly constructed and

' maintained than in the past, they represent a net loss in the resources available for future
manufacturing. In addition, the potential for environmental contamination remains.
Sanitary landfills have become, in the United States, the preferred option for final disposal
of waste products as the past alternatives of land and ocean dumping have been curtailed.
(Alexander, 1993)

The Environmental Protection Agency (EPA) uses the term “integrated waste
management” to represent the application of the waste hierarchy to local and regional
situations, with the understanding that the different waste management Options will be
represented in varying proportions depending on the characteristics and conditions of the
region. The goals for waste reduction set by the EPA do not specify which options are to
be used; the term “waste reduction” encompasses source reduction, as well as reuse of
materials, recycling, composting, and by some definitions, combustion. Many states have
set volume goals as well, with waste reduction ofien measured as a reduction in the
amount of waste entering landfills, or landfills and incinerators, regardless of the processes

involved. (Fishbein, 1992)

Municipal Waste Reduction

Within municipalities, waste reduction and recycling programs have found favor as
an alternative to increasing solid waste disposal costs. Coupled with a rising population in
many areas, the amount of trash being discarded continues to go up: compared to 1960

estimates of approximately 2.5 pounds per person per day (Wagner, 1995), the

Environmental Protection Agency estimated an average of 4.4 pounds per person per day
in 1993. (EPA, 1994) This figure includes discards from households as well as
commercial and institutional sectors, but does not include manufacturing, construction,
agricultural and mining wastes. (Alexander, 1993) An annual survey of state programs by
the environmental journal Biocycle, which includes industrial waste disposed of in
municipal facilities, places this figure even higher, at an average of 6.5 pounds per person
per day for the same year. (Steuteville, 1995)

As the volume Of solid waste increases, the number of operating landfills
continues to decrease, as does the number of Operating combustion and waste-tO-energy
facilities. (Decision-Maker’s Guide, 1989) Although the national capacity for land
disposal Of municipal waste remains steady, the trend is toward larger, but fewer, sanitary
landfills. (Rathje, 1992) As regional landfills close because of exceeded capacity or
environmental concerns, increasingly restrictive legislation and opposition by community
groups have prevented the Opening of new sites in many areas. (F essenden, 1989) This
has resulted in increasing collection and disposal fees for many communities, making
recycling and waste reduction a more attractive alternative financially, particularly in the
urban areas of the Northeast and Northwest.

Siting of new landfill facilities, if it happens at all, can take years and run into the
millions of dollars. (Bachman, 1995) Part of the reason for this is more restrictive
legislation regarding the siting and Operation of new facilities. The other part is the
NIMBY, or “Not in My Backyard,” syndrome that manifests within communities

regarding the potential siting of a new landfill or combustion facility. There are several

reasons for this opposition, one of them being concern for the environment and human
health. There is also a “lack of public trust” toward both the Operators and the regulators
of landfill and combustion facilities. Community residents that live near a proposed
landfill facility are concerned about property values and do not want to accept risk they
see as involuntary and beyond their control, especially if the waste is generated outside the
community. (Williams, 1990)

The NIMBY syndrome that has blocked the construction of new landfill and
combustion facilities does not seem to be present to the same extent with regard to the
siting of recycling and composting facilities. (Blumberg, 1989) From 1988 to 1990,
Biocycle estimated growth in the number of materials recovery facilities (“MRF’s”) in the
United States from 16 to 92, and composting facilities from 651 to 1407. (Steuteville,
1992) By 1994, the same source estimated the number of MRF’s in the United States at
1243, and composting facilities at 3202. A MRF in this instance is defined as a facility
that sorts, bales and markets recyclables, although this varies somewhat from state to
state. (Steuteville, 1995)

The growth in recycling and composting facilities serves as one informal measure
of increasing municipal support for waste reduction programs. Another way to measure
support for these programs is to look at the diversion rates reported by municipalities.
Waste diversion refers to materials diverted for recycling and composting compared to
previous discard totals. (Martin, 1993) In 1990, some municipalities were achieving
diversion rates of greater than forty percent of their solid waste streams, primarily on the

East Coast where landfill costs are high. (Platt, 1991) Nationwide, at least fifteen states

reported municipal solid waste diversion rates Of twenty-five percent or more of their

waste streams for 1994. (Steuteville, 1995)

Industrial “gste Reduction Programs

The industrial sector has also embraced waste reduction and recycling measures,
primarily for the benefits of cost avoidance, although human health concerns play a part as
well. Because industrial processes often include hazardous materials, reduction in the
quantity or toxicity of materials used can mean lower liability and cleanup costs for
manufacturing companies in addition to lower disposal fees. (OTA, 1989) Generally,
industrial waste reduction efforts fall into four categories: 1) modifying a technology,
process or piece of machinery, 2) product redesign, 3) use of alternate input (raw)
materials, and 4) improving the handling, storage or inventory of materials within an
operational framework. (Rappaport, 1993)

Using waste reduction strategies has paid off for many companies, both monetarily
and in terms of public image. (Makower, 1994) In 1985, 3M’s highly publicized
Pollution Prevention Pays program was credited with $3 00 million in cost savings and a
50 % reduction in wastes over the first ten years of the program. (Oldenburg, 1987)
Current estimates of cost savings from the program are close to $600 million. (Makower,
1994) Dow Coming and Ciba—Geigy reported similar levels Of cost savings follOwing the
introduction of waste reduction efforts, combining process improvements, materials
management and by-product exchange programs in an effort to reduce the amount and

toxicity of their waste streams. (Huisingh, 1982) Apple Computers’ decision to use a

recycled content corrugated material for shipping instead of bleached, virgin material has

saved the company over $3 million since 1991. (Saunders, 1993)

The Role Of Universities

Universities have a significant role to play in recycling and waste reduction efforts,
both in their capacity as research institutions and because of the types of waste streams
found on large campuses that could enable universities to implement effective waste
reduction programs. (Hegberg, 1992) Instead of leading the way, however, many
universities are lagging behind the municipal and industrial sectors in recycling and source
reduction efforts. (Abbott, 1992) In 1991, a survey of campus recycling programs found
that of those colleges and universities that had programs, about half were run by students
on a volunteer basis, without the necessary support of campus administrators. Nearly
seventy-eight percent of the roughly 3,500 colleges and universities in the country
reported having recycling programs by 1993, although the type of program and estimated
diversion of materials varies widely. (DeBell, 1994) Ithaca College in New York, for
example, has an extensive paper products, metals, glass, plastics, oils, landscape waste and
food waste recycling program. (Schoonover, 1994) Typically, however, colleges and
universities have limited programs, collecting Office paper or corrugated cardboard from
academic buildings or residence halls.

Like their municipal counterparts, those universities that have committed to

comprehensive recycling programs have experienced significant reductions in disposal

10

costs. During 1994, the University of Illinois recycled 455 tons of materials, with a net
savings of $67, 000.1 (Hoss, 1996) Officials at Miami University, a school of 16,000 in
Oxford, Ohio, determined that the total cost for a materials recovery facility to process
materials from a diverse recycling program was one-third less than the cost of landfill
disposal of those items, and provided an on-site learning Opportunity for students. (Gaski,
1993) Cost estimates for recycling at The University of Colorado were about half of the

cost of waste disposal. (RRS, 1991)

Waste Chmcterization Methodologies

 

For any sector of the economy to implement effective waste reduction measures, it
is important to first identify the quantity and composition of the waste stream in question.
(Tchobanoglous, 1993) Waste characterization can provide figures for planning collection
programs, estimating total potential diversion from the waste stream, and identifying
materials that should be targeted for waste reduction or recycling. This process can be
accomplished in a variety of ways, depending on the sector in question.

For municipal waste streams, the Environmental Protection Agency publishes a
periodic report, derived from production information for materials by weight, as well as
information derived from end products found in the waste stream. This materials flow
methodology is used to provide national average figures for municipal waste streams, and
is thought to provide a more complete picture than could be Obtained from other methods

of waste characterization. Physical samples can be subject to regional differences, such as

 

‘ Net savings is defined as the disposal cost avoidance - the total recycling cost.

ll

climate, sociO-economic status, industrial or agricultural specializations and state
governmental regulations, all of which may affect the composition of the waste stream .
(Environmental Protection Agency, 1994)

Franklin, Associates, in conjunction with the Environmental Protection Agency,
has used figures from the United States Department of Commerce as well as trade
association data to provide information about materials and products. Allowances are
made for manufacturing losses, exports and imports, and materials that are destroyed or
diverted during use. Adjustments are also made for combustion activity, which are not
included in totals. Organic products, such as food and landscape wastes, are estimated
from widespread sampling procedures. In total, the figures published include residential,
commercial, and institutional wastes. Some of the materials that are not included using
this methodology include demolition and construction wastes, incinerator ash, and
remnants of products lefi in containers. (EPA,1994) Although the majority of the figures
using this method are calculated without physically sampling the waste stream, other
studies of municipal solid waste streams, as well as Franklin’s own confirming samples,
are fairly close to these numbers. A study done in Brevard County, Florida concluded:

“...the sum of the subcategories from Franklin and those obtained locally, do not

vary widely. Since no major differences were observed it was concluded that the

Franklin subcategories of the percentages of materials in the United States

municipal solid waste streams could be utilized to estimate those same

components in the state of Florida with reasonable accuracy.”
(EPA, 1990, p. 100)

Another form of this method is materials mass balance analysis. This method

works best for determining the total amount of solid waste generated, by identifying all

12

activities that generate solid waste within a given boundary and then using that
information to mathematically derive the rate of accumulation. It is not as USCfiJI for
determining the composition of the waste stream. (Tchobanoglous, 1993)

There is also a small, but significant, body of literature on the composition of
materials found in municipal landfills and how these materials translate into information
regarding consumer decision-making. Excavation of landfills to examine the types and
amounts of waste can be helpful for targeting materials (like excess packaging or
disposable diapers) for which alternate purchasing decisions could be made. (Rathje,
1989) The figures derived from these excavations are compatible with the Franklin
materials flow percentages. (Alexander, 1993)

There is a developing body of literature on waste audits and life cycle analyses,
primarily for use within the industrial sector. Both processes are designed to identify
materials used during the process of manufacture in an effort to reduce or replace
materials that are considered hazardous to human health and the environment. (Sullivan,
1995) RCRA refers to this endeavor as waste minimization, a term which embraces both
source reduction and recycling of solid and liquid wastes in an effort to reduce the amount
of hazardous material that must be disposed of in landfills or through incineration.
(Oldenburg, 1987)

Life Cycle Analysis (LCA) is a process used to track the materials used in the
manufacturing process, usually for a single product, in an effort to identify the potential
tradeoffs involved in design, material or production changes. Opportunities for

environmentally positive changes in the manufacturing process can be identified.

13

Conversely, the process can be used to analyze proposed changes that would, in the long
run, only trade one type of pollution for another. In order to consider all possible effects,
an attempt is made to examine all Of the stages of a product’s life cycle, from the
collection of raw materials to its eventual disposal. The steps of a life cycle analysis
usually include goal definition ( purpose of the study), inventory analysis (quantifying the
inputs and outputs of the manufacturing process), impact assessment (looking at the
qualitative or quantitative impacts Of those inputs/outputs), and improvement analysis
(identifying areas that could be changed, with an end goal Of environmental improvement).
Life cycle analyses are recognized as a positive step for characterizing industrial waste
streams, although the methodology for this process is still being developed. (Sullivan,

1995)

_C_ha_racterizing University Waste Streams

Unfortunately, none of these methods for identifying waste stream characteristics
is best suited for university campuses. The estimates of waste stream composition
calculated by the Environmental Protection Agency and other organizations for municipal
solid waste using materials flow methods are good starting points, but are not specific to
university waste streams, which may differ significantly in content from those of
municipalities. (Creighton, 1993) Municipal landfill excavations can be very helpfirl in
making decisions about consumer habits, but have the limitation of not providing
purchasing information for specific segments of the population, like college students.

(Delirr, 1992)

14

Except in very limited capacities, universities also do not resemble industry, as
there is seldom an identifiable operational “process” for making physical products that can
be targeted and identified. Nor are waste materials generally considered hazardous, with
the exception of materials from laboratories and research centers located on university
campuses. Industrial waste characterization approaches could be USCfiJI for this segment
of the university waste stream, although hazardous materials usually represent a relatively
small percentage of the total solid waste generated and may be disposed of differently
from the majority of the solid waste stream.

Some of the unique characteristics of university waste not commonly encountered
in municipal or industrial settings include: predictable seasonal fluctuations of activities
and on-campus populations, concentrated densities of same-age residents, intertwined
research, academic and athletic/event facilities, as well as (in some cases) large amounts
of landscape or agricultural waste. As most campuses do not have available facilities for
handling solid waste on site, waste management decisions must take into account
municipal relationships or constraints regarding contracts with available waste
management contractors. (Giannone, 1995)

In addition, there are mechanisms within the typical university structure for fairly
effective distribution of information related to recycling and reduction education
programs. A large body of students is available, many of whom possess levels of
environmental concern Often greater than those of the general population. (Watson, 1990)

Many university campuses have school-year populations larger than municipalities with

15

successful recycling programs, (F012, 1990) indicating the potential for sustainable levels
of materials recovery.

Some of the differences that exist between university campuses can affect the
generalizability of waste stream characterization. Climate, size of the campus (particularly
the extent of on-campus housing facilities), its mission toward “commuting” or part-time
students versus fiIll-time students, the types of programs and degrees that are offered, and
state or local legislation are some of the variables that can affect the characteristics of the
waste stream at any given university. (Friedman, 1993) From a study comparing the
waste generation rates for seventeen colleges and universities, smaller colleges had larger
per person rates of waste generation than larger ones, and universities with on-campus
student housing had significantly greater rates of waste generation than colleges serving
primarily a commuting population. (Hegberg, 1992)

These differences notwithstanding, the types of activities and populations found on
university campuses make comparisons possible and provide direction for recycling and
waste reduction programs. The use of a manual sorting approach combined with
information on current levels of waste disposal and diversion can provide more accurate
information about a specific university, as well as increasing the waste characterization
information available for other universities making program decisions. The manual
sorting approach to waste characterization entails physically dividing representative
samples of the solid waste stream into categories and using the resulting measurements to
estimate the percentage weight and volume for each category within the solid waste

stream. Choice of categories to be examined vary, but are usually based on components

16

of the waste stream that are potentially recoverable through recycling, composting or
reuse. (Alexander, 1992) This provides a “snapshot” of the waste stream, and can
pinpoint the areas upon which waste reduction and recycling programs should focus.

Although the standard deviations of previous studies utilizing this method of waste
characterization have varied by as much as twenty-five percent, (Gay, 1993) careful
planning and division of the campus into like strata can reduce this variability considerably.
If the goal of the waste characterization is to project future diversion capacity, for those
universities considering the implementation of a transfer station or materials recovery
facility, repeated characterizations during the time it takes to develop and find such a
facility can provide specific numbers to supplement other projection methods. (Alter,
1991)

Waste reduction efforts can provide benefits both economically and for the
environment. To develop effective waste reduction programs, however, it is necessary to
have a clear picture of the types and amounts of materials found in the solid waste stream.
Although there are a variety of methods for estimating this information, a well-planned
manual sorting protocol analyzing representative samples of the waste stream in question
can provide information of the necessary precision to predict potential capture rates or
to evaluate current diversion programs.

It is hoped this study will provide information about the effectiveness of using a
manual sorting design to evaluate the solid waste stream found within a university setting.
The study is also designed to highlight the differences between estimates for municipal

waste stream composition and the composition of university waste streams, using

17

Michigan State University as an example. Adding to the available information on solid
waste characterization, this study should serve as a basis for comparison for those
institutions conducting similar studies. It should also prove helpful as a rough estimate of

composition for those institutions unable to conduct a waste characterization themselves.

Problem Statement

The purpose of this research is to find answers to two questions: 1) “How reliable
is waste characterization, using a manual sorting design, in the evaluation of current and
potential levels of recycling and waste reduction activities for university populations?”
and 2) “In what ways do university solid waste streams differ fi'om those of
municipalities?”

Universities should be leading the way in the effort to meet state and federal waste
reduction guidelines. The first step toward developing effective recycling and waste
reduction programs is to establish a clear picture of the types and quantities of materials
present in the waste stream. Although there are good estimates for the components of the
solid waste stream available for municipalities, these estimates are not representative of
university waste streams, and do not provide a usable model for integrated waste
management planning for this segment of the population. Similarly, waste characterization
models for the industrial sector, although very effective in reducing the amount of waste
generated on an industry-by-industry basis, do not provide a usable model for integrated

waste management for universities.

18

Objectives of the Study

The objectives of this study are to:

1. evaluate the effectiveness of a manual sorting model for characterizing the solid
waste streams within a university setting.

2. quantify weights of specific materials being diverted from the waste stream via
current Michigan State University recycling programs and estimate potential
diversion rates for commonly recycled materials, for the purpose of complying
with state and federal goals for waste reduction.

3. identify significant differences in composition between Michigan State
University’s solid waste stream and current United States municipal solid waste

streams.

Hymtheses

There are two hypotheses that will be examined within this study. They are:

1. Ho: Using a manual sorting model with fewer than thirty samples, the percentages
by weight of major components (greater than five percent) of a university solid
waste stream cannot be estimated with an eighty percent confidence level and a
: two percent error from the mean.
H1: Using a manual sorting model with fewer than thirty samples, the percentages
by weight of major components (greater than five percent) of a university solid
waste stream can be estimated with an eighty percent confidence level and a

: two percent error fi'om the mean.

l9

2. Ho: The percentages of major components of a university solid waste stream do
not differ significantly from the estimated percentages of major components of
municipal solid waste streams, at a ninety percent confidence level.

H1: The percentages of major components of a university solid waste stream do
differ significantly from the estimated percentages of major components of

municipal solid waste streams, at a ninety percent confidence level.

In the second hypothesis, “major components” refers to the categories used by the
Environmental Protection Agency in Characterization of Municipal Solid fiste in the
United States: 1994 Update: paper, glass, metals, plastics, wood products, food waste,

and landscape waste/yard trimmings.

CHAPTER TWO:

REVIEW OF LITERATURE

G_oals of Waste Characterization

Waste characterization can be defined as “any program which involves a logical
and systematic approach to obtaining and analyzing data on one or more solid waste
streams or substreams.” (p. 2, MDNR, 1986) Waste characterization studies can serve a
variety of functions, depending on the purpose and structure of the study. In general, the
goal of waste characterization is to “identify the sources, characteristics and quantities” of
the waste stream. (Powell, 1993) For the purpose of meeting state and federal guidelines,
as well as planning effective waste management programs, a more specific set of goals
would be to: 1) identify materials that should be targeted for reduction and recycling
programs, and 2) evaluate the success of any current programs in removing materials
from the waste stream. Individual waste characterizations can also contribute to the
general body of knowledge, both to assist others in the field (EPA, 1989) and to
highlight changes in composition due to legislation, technology, or changing attitudes.
(Clarke, 1992)

Identifying the kinds of materials to be targeted for waste reduction and recycling
programs should be based on several criteria. First, those materials contributing a
significantly large percentage (greater than 5%) of the waste stream, either by weight or
volume, should be examined for possible inclusion in recycling or waste reduction
programs. (Gay, 1993) Although weight is the most commonly used material

measurement, the capacities of collection vehicles make volume a consideration as well.

21

measurement, the capacities of collection vehicles make volume a consideration as well.
(Tchobanoglous, 1993) In the case of cardboard, for example, the weight of the material
might not be considered significant, but the volume could make a large difference in truck
capacity and the necessary number of “runs” a collection vehicle would need to make.

The second criteria is related to markets; are there industries within a reasonable
distance willing to take or buy the material? If there is no local market for it, there is the
potential of needing to stockpile the material until a large enough amount is amassed to
interest a more distant buyer, or of combining materials with a nearby municipality.

(Solid Waste Management Task Force, 1989)

The third consideration concerns the mechanics of collection. Is the material
relatively easy to collect? Is there capacity within the collection system to accommodate
additional pickup without capital outlay for new equipment? Any potential purchase of
additional capital equipment needs to be measured against estimated cost savings of
diverting the material from the landfill and the attendant tipping fees. (Sound Resource
Management, 1993)

In geographic areas with high transportation costs or landfill tipping fees, recycling
programs can divert a significant amount of materials, and provide a reduction in waste
disposal costs. Recycling programs in areas with low or moderate landfill tipping fees are
usually not money-making propositions, due to market fluctuations and the difficulty of
competition with subsidized virgin materials. The costs of collecting and processing the
recyclables may outweigh the cost of landfill disposal. Although sale of the collected

materials may offset any additional expense, the market for recyclables remains somewhat

22

unpredictable, making it difficult to rely on this as a source of revenue. (Apotheker, 1995)
For a community or institution, using a waste characterization to determine the criteria as
listed above can reduce the costs of operating a recovery program and ensure that the
most economical materials are chosen to meet waste reduction goals. (Jablonowski,
1995)

Performing a waste characterization is also a positive way to measure the success
of current programs that may already be in place. Comparing the types and amounts of
materials found in the waste stream characterization with known amounts from recovery
programs can give a good estimate of the proportion of material diverted from the waste
stream during the time period of the study. The necessity of compliance with state
recycling and reduction goals can make this an important consideration, especially for
public institutions, although there are private colleges and universities that are complying

with these goals as well. (Creighton, 1993)

State and Federal Waste Reduction Guidelines

The Resource Conservation and Recovery Act of 1976 asked that individual states
“establish and justify priorities and timing” (p. 44) for the development of integrated
solid waste management plans. This includes source reduction, which is defined as any
practice that reduces the toxicity, volume, or weight of materials in the waste stream prior
to disposal or diversion. Source reduction does not technically include recycling, which is
the recovery of post-consumer and production discards for use as raw materials in

manufacturing, or direct reuse of materials through exchanges or salvage programs.

23

(Curbing Waste, 1990) For the purposes of this study, the term “waste reduction” is
Understood to encompass any practice that reduces the amount of materials being
landfilled or incinerated, including source reduction, reuse or recycling.

The US. Environmental Protection Agency has set goals of diverting twenty-five
percent of the nation’s solid waste through recycling and source reduction programs by
1992. (U .8. Environmental Protection Agency, 1990) It hopes to raise that to thirty
percent by the year 2000. (US. Environmental Protection Agency, 1994) Municipal
waste streams seem able, so far, to met these guidelines: one estimate put the national
1994 MSW, or municipal solid waste, recovery rate at twenty-three percent for 1994, up
from nineteen percent in 1993. (Steuteville, 1995)

State goals vary widely, with states setting specific percentages that range from
twenty-one percent by the year 2000 for the state of Delaware to seventy percent at an
unnamed future date for the state of Rhode Island. (Steuteville, 1995) Most state
programs do not refer specifically to universities and colleges, but a few require these
institutions to develop specific integrated waste management plans that meet the state
goals for waste reduction.

In Illinois, for example, the 1990 adoption of PA86-1 363 included a section known

’

as the “College Recycling Law,’ specifically targeted toward two and four year public
colleges and universities. It requires a 10 year comprehensive waste reduction plan and a
40% reduction in solid waste disposal by the year 2000. (Hegberg, 1992) The state of

Tennessee enacted The Solid Waste Management Act of 1991, requiring that all public

24

colleges and universities develop waste management plans focusing on recycling and
source reduction. (Powell, 1993)

Michigan currently does not have specific legislation focused on universities, but
has set statewide recycling goals to reduce solid waste disposal fifty percent by the year
2005. (Steuteville, 1995) Some public universities within the state show indications of
moving toward that goal: The University of Michigan reports a waste diversion rate of
twenty-three percent, Central Michigan University twenty-five percent, and Western
Michigan University twenty-seven percent. Michigan State University, by comparison,
reports diversion rates of thirteen percent; an indication of a significant gap between
current waste diversion program totals and state goals, as well as the efforts of other

Michigan universities. (Michigan State University Recycling Coalition, 1995)

Using Waste Chmcterizafion to Meet Stzgaand Fedegral Guidelines

As previously stated, waste characterization studies can be instrumental in
targeting materials for waste diversion programs as well as estimating the effectiveness of
recovery programs already in place. In municipal recycling programs, this can be referred
to as “participation rates,” or the number of households participating in recycling
programs compared to the total number served by the program. Another way to measure
this is to use the “diversion rate,” a more inclusive term that includes all material diverted
from the waste stream, compared to all material diverted plus all material discarded.

(Riley, 1993) This second method is more easily calculated at the university level and can

25

be derived from waste characterization methodologies like manual sorting along with data
collected by diversion programs operating within the university.

After deciding the purpose of the waste characterization, planning the study itself
entails making decisions about the structure and analysis. Some of the important factors
to consider prior to initiating a manual sorting procedure include sampling design and the
procedures for choosing the samples, the organization of the sorting process itself to
reduce possible bias, and the waste streams and categories of materials that should be

examined.

Recommended Sampling and Evaluation Techniques

The EPA’s recommendations for solid waste characterization sampling , although
primarily concerned with hazardous materials, has applicability for any type of solid waste.
The recommendation for designing a sampling plan is that representative samples be used,
or samples “exhibiting average properties of the whole waste.” This can mean using
simple random techniques, stratified random techniques (for use when differences are
observed in the waste stream either spatially or over time), and systematic or authoritative
random sampling, which is planned to focus on or avoid known anomalies in the waste
stream being studied. Related to this, there is the specification that “enough samples be
collected over a period of time sufiicient to represent the variability of the waste.” (p. 9-
6) The Environmental Protection Agency also recommends an eighty percent confidence
level when evaluating the representativeness of the sample. (Environmental Protection

Agency, 1986)

26

accuracy and precision requirements of the intended data uses. At the same time, it needs
to maximize the data output for a given budget.” (p. 10) A 200-300 pound sample is
recommended, with the number of samples varying depending on the type of waste stream
and the component being measured. Results should be stated in terms of means and
standard deviation. (MDNR, 1986)

Another source states that a sample size of approximately 200 pounds varies
insignificantly from larger sample sizes for municipal waste, if taken from a truckload of
representative waste material. As most truckloads are much larger than this, the load
should be quartered repeatedly until a sample of the appropriate size is reached. This
source suggests that results be considered in terms of mean, standard deviation from the
norm and the variation coefficient to determine how representative the chosen samples are
of the waste stream being examined. (Tchobanoglous, 1977)

The University of Illinois Center for Solid Waste Management and Research
procedure for conducting a manual sort (adapted from a method developed for municipal
solid waste by the American Society for Testing and Materials) recommends a 200-300
pound section taken per vehicle load of MSW. The samples should be collected during a
period of at least one week, with the suggestion that at least twenty and preferably thirty
samples be taken to ensure a normal distribution. The exact number of samples necessary
should be estimated from the expected mean and the statistical confidence level and
precision desired. (Hegberg, 1992)

Martin, in an analysis of manual sorting protocols for municipal waste streams,

found that forty samples, each weighing two hundred pounds, would yield a two percent

27

error at an eighty percent level of confidence. For the same level of error, estimated
sample numbers of sixty-three were suggested to reach a ninety percent level of
confidence, with ninety samples necessary to reach a ninety-five percent confidence level.
It was also noted that material categories found in smaller percentage categories tended to
have a relatively larger amount of error on a proportionate basis than those materials
composing a larger percentage of the waste stream. Basing the number of samples on the
larger components of the waste stream provided the opportunity to further reduce the
number of samples needed to twenty to provide a two percent error at an eighty percent
confidence level. (Martin, 1993)

Other examples of sampling protocols can be drawn from the municipal sector. A
field study undertaken by the Ventura County Solid Waste Management Department used
a total of 307 samples to characterize the municipal solid waste stream for five
jurisdictions. These samples were characterized by residential, commercial, and industrial
populations, and sorted into six general categories of materials. Residential samples were
taken from individual home garbage cans set out for collection, using random address
selection. Larger samples from the commercial and industrial stratas were divided into
grids from which random samples were selected. The resulting data was examined for a
sample mean in pounds per person per day as well as the percent of total and standard
deviation for each material group. The confidence level used was ninety percent. The
total tons disposed and diversion rates by material type were estimated as well. Results
were also examined by jurisdiction, with significant differences in waste characteristics and

amounts found to exist between jurisdictions. (Delin, 1992)

28

Elias Musa, conducting a study to determine optimum sample size for municipal
waste characterization, used a systematic sampling procedure, a variation of random
sampling that is useful in field work. His method for this was to sample one household
out of each twenty-five, following the routing of standard collection vehicles. His
recommendation for the number of samples needed was based on the projected volume of
materials in the waste stream: for items thought to compose large percentages, sample
numbers of thirty to forty were recommended for a ninety-five percent confidence level.
For items thought to be less common, Musa recommended at least two hundred samples
to obtain the same degree of accuracy. (Musa, 1981)

Sampling protocols can also be evaluated qualitatively. John H. Martin listed the
following criteria used to evaluate waste assessment protocols: 1) The protocol should

require the minimum time necessary to attain the desired confidence level, 2) the resulting
data should be accurate, reliable and valid, 3) consideration should be made of the pattern
in which trucks bring discards to the landfill, 4) the peOple involved in the sorting process
should be able to understand the protocol with less than four hours training, and 5) there
Should be minimal disruption of landfill operations. (1993)

Each of these studies provided insight on the sampling techniques commonly used
in the field of waste characterization. Although there is not a firm consensus in this area,
an eighty percent level of confidence can be determined to provide an acceptable basis for
evaluation of sample integrity. To reach this level of precision, at least for major material

categories, at least twenty and preferably thirty samples are necessary as a minimum.

 

29

Choosing Categories of Margials

There are wide variations among waste characterization studies regarding choice
of materials categories. Most ofien, categories are chosen based on the intended waste
management strategies. For instance, a study to determine the potential amount of
materials available for a waste-to-energy facility would choose different categories than a
study evaluating the success of current recycling programs. For combustion, categories
would be grouped by their Btu, or British thermal unit, potential. Papers and wood
products would be of greater interest than metals and glass. (NREL, 1992) For the
purpose of developing or evaluating a recycling program, the solid waste stream would be
evaluated for those materials currently or potentially recyclable, like paper, metals, glass
and plastic. Different paper products would need to be evaluated separately, as the
market price for each varies. (OTA, 1990) Composting operations would be interested in
what is ofien called the “putrescibles” component of the solid waste stream: items that
decay fairly quickly like food, yard waste, and other organic materials. (Martin, 1993)
Planning for waste reduction would necessitate examining manufactured materials that
could be purchased in a reusable format instead of disposable as well as materials for
which a lower-use alternative could be substituted. Some examples are computer printer
cartridges, disposable cups and plates, batteries and light bulbs. (Curbing Waste, 1990)
Program planning that combines these waste management options would need to plan
materials categories with a compromise between ease of sorting and desired information
about composition. Larger numbers of categories increase sorting time and may reduce

precision for categories constituting very small percentages of the waste stream. At the

 

30

same time, too few categories may not provide the desired information about the waste

stream.

m1 SortingConsiderations

The methods used for physically sorting solid waste materials vary considerably,
based on study objectives, personnel available, physical facilities, and budgetary
constraints. Some of the primary considerations include staffing, safety equipment, sorting
equipment and sorting site. Not surprisingly, finding individuals willing to manually sort
solid waste is a difficulty common to this form of waste characterization. The issue of
safety for the sorting personnel is important to prevent injuries from sharp items or
contaminated waste products. The choice of a sorting site and sorting equipment can be
essential to maintaining sample integrity.

The Michigan Department of Natural Resources recommends that manual sorting
studies be conducted at a landfill facility with three to five workers and a knowledgeable
supervisor. Personal safety equipment should consist of at least heavy boots, gloves, eye
protection and a mask (respirator). Minimum sorting equipment includes a sorting table
with a screen to separate fine material, sort containers, and a scale for weighing material,
with a plastic sheet covering the sort site. (MDNR, 1986)

The Stagdard Handbook of Environmental Engineering (Corbitt, 1990)
recommends the sorting be done indoors, with room for both equipment and collection
vehicles depositing samples. Each often crew members should be assigned one material

to collect (with a rolling barrel) from the sample located on the floor of the sort area.

31

When containers are full, they are to be wheeled to the weighing area and the crew
member continues with an empty barrel. Oversized materials are segregated and
photographed. After the pile of waste is reduced in size, it can be transferred to a sort
table, with one-inch square screening. Crew members should be provided with long-
sleeved coveralls, puncture resistant gloves, and heavy workboots. Unfortunately, an
indoor facility with the necessary characteristics was not available for this study, so this
protocol was not considered as a model.

The waste characterization study conducted at East Tennessee State University
used student volunteers working in shifts, with 10-15 peOple working at any given time for
the two days of the sort. The students were provided with Tyvek suits, eye protection and
butyl gloves to prevent accidents, and signs were posted with safety rules. The samples
were sorted on tables into plastic containers and weighed. The study was conducted
outdoors on the campus near the physical plant. (Powell, 1993)

For the Tufts University (Medford Campus) waste sort, students were also the
majority of the work force. They were provided or asked to bring gloves and some form
of eye protection. The sort was also conducted on the campus, near the dumpster being
sorted. The waste material was emptied from the dumpster onto the ground and taken by
students to specified nearby sites labeled with the chosen categories of materials. The
sorted materials were then placed in clear plastic bags and weighed. (Creighton, 1993)

All of these campus protocols were instructive when planning this waste

characterization study. After considering the study objectives and available facilities, it

32

was detemrined that the protocol outlined by the Department of Natural Resources was

the best starting point for planning the protocol for the manual sorting of the samples.

University Waste Char‘acterization Methodologi_ea

A waste stream assessment plan for East Tennessee State University divided the
waste stream into nine strata and manually sorted one sample from each strata into
thirteen categories of materials. Samples consisted of an individual dumpster from each
strata and ranged in weight from 345 to 1,252 pounds. The samples were chosen using a
variation on random sampling referred to by the Environmental Protection Agency as
“authoritative sampling”: the samples were identified by knowledgeable parties as most
representative of the stratum. (Environmental Protection Agency, 1986) Total weights
for the campus were determined by weighing selected dumpsters periodically and
extrapolating total weight for the academic year. Using one-way and two-way ANOVA’S,
the sampling results were tested for any difference in means between the strata for the
material categories selected. Significant differences between strata were not found in this
case, and the author concluded that the number of samples was sufliciently representative
of the campus waste stream. (Powell, 1993)

A study undertaken by Tufts University considered in part the different waste
strata divisions present on college campuses. Tufts divided the waste streams into
administrative buildings, dormitories and dining halls. For the purpose of a materials flow
assessment, the campus was firrther divided into housing, ofiices, labs, dining services,

athletic facilities, grounds, and other buildings. An assessment of the University of North

33

Carolina, cited in the Tufts study, used the division of administration buildings, academic
departments, and residence halls. (Creighton, 1993) This study did conduct waste
characterization sampling on a limited basis, but was primarily a comparison of other
partial waste stream studies conducted on college campuses.

A compilation study of other university programs used the divisions of educational,
administrative, residential and dedicated purpose areas like libraries and computer centers
to differentiate the waste streams. The study also found significant differences in
generation rates between colleges with significant on or near campus housing and those
colleges that were serving primarily commuting students. Larger universities also were
shown to have less per-person waste generation than smaller universities. This study
found variability in the waste stream on a monthly basis, with larger amounts of materials
present during the months considered to be part of the traditional school year. Existing
recycling and composting programs were thought to influence waste stream composition
as well. (Hegberg, 1992)

The studies of waste streams on college campuses vary in their choices of strata,
making comparison difficult. Adopting the strata divisions used by the University of
North Carolina should make possible some limited comparison and makes a useful
structure for analyzing the differences between waste streams. These studies highlight
some characteristics of college waste streams to be considered when designing a protocol.

Overall, as this review of the literature base on waste characterization
demonstrates, there is not yet a high degree of consensus regarding the type of protocol to

be used when conducting a manual waste characterization study. After a review of

34

available information, as well as examining the characteristics and informational needs of
the Michigan State University campus, it was decided that no one of the protocols
described in this review was able to meet all of the study objectives. Therefore, the
research approach is a compilation of strategies used in other waste characterization
studies, modified for the chosen research site. It is hoped the resulting protocol will serve

as a useful model for other universities planning waste characterization studies.

CHAPTER THREE:

DESCRIPTION OF STUDY METHOD

Description of the Research Site

The main campus of Michigan State University is situated on 1500 acres, and
enrolled approximately 41,000 students at both the undergraduate and graduate levels for
the 1994-1995 academic year. The residence hall system consists of twenty-five buildings
with fourteen cafeterias which feed and house approximately 14, 500 students during the
academic year (September to May). In addition, there are university apartments that
house approximately 2200 students and their families (about 5400 adults and children)
year round. Academic facilities on campus include a veterinary school, two medical
schools, eight sports facilities, and forty-five academic and office buildings, twelve of
which have laboratory or medical facilities. There are also several dedicated or special
event facilities, including a performing arts center, a union with food service, a
planetarium and a laundry. The large number of on-campus students as well as the large
variety of buildings and academic programs provided a large, comprehensive waste stream
from which to sample, and gave a greater probability that information derived fiom this
study will be applicable to other programs.

The campus was divided into three strata for analysis: residence halls, university
apartments and academic/administrative buildings, including athletic and dedicated
facilities. The decision for this division was based on: 1) previous studies of university

waste stream analysis, 2) a preliminary characterization conducted in the spring of 1994

35

36

with limited samples to establish the potential differences between strata, 3) the routing
availability of the grounds department to provide dedicated samples for analysis and 4) the
intention of using the results of the study to plan waste reduction programs specific to
these areas of the campus. Agricultural buildings not on the main campus were excluded
for two reasons: lack of access to collection routes for these buildings and lack of proper
equipment to sort the waste stream from these facilities, which consists primarily of animal

waste and bedding.

Material Categories

Examining the available literature regarding material categories for solid waste
characterizations, it became apparent that there wasn’t a general consensus on the types of
categories that should be used. Therefore, the categories chosen were a combination of
other suggested sorting classifications. They were chosen to provide information about
several facets of the waste hierarchy, for the specific planning needs of the Oflice of Waste
Reduction at Michigan State University. Some of the categories were chosen to provide
information about recycling (paper, plastic, metal, and glass). Another category were
chosen to provide information about compostable materials (organics, sometimes referred
to as putrescibles), and the remaining categories were chosen to provide information about
materials that could be reduced in amount or reused instead of being discarded
(construction debris, textiles and manufactured goods). Within each of these eight
categories were several subcategories, for a total of forty-eight material divisions. A

complete list of categories and subcategories is given in Figure 3, found in the discussion

37

of sorting procedure. The use of these categories was designed to provide information
either by major category for comparison with other programs, or by subcategories, for a

more precise focus on specific materials found within the waste stream.

Timing of‘ Study

The first part of the waste characterization study was conducted during the last
two weeks of July, 1994. The timing of the study was designed to provide a
representative waste stream to sample, as this time period fell roughly in the middle of the
summer semester. As other studies have shown, the beginning and end of the semester
are not representative because students are moving into and out of the system. (Powell,
1993)

The second part of the study was scheduled for the last two weeks of October,
again to avoid the unrepresentative nature of the beginning of the term as well as wet or
snowy weather common to November and December. Unfortunately, the study was
unable to avoid the wet weather. This, in addition to scheduling problems caused by
students’ class schedules, caused the fall segment of the waste characterization to be

extended a third week in order to obtain enough samples.

Sampling Procedure
The Grounds Department at Michigan State University, which is responsible for
the collection and hauling of campus solid waste, agreed to bring loads of solid waste

from the university to a privately owned landfill in nearby Lansing during the time period

38

collection and delivery of materials from only one strata at a time. For the summer study,
samples were taken from each load picked up by the grounds department from the
campus. Exceptions to this were loads in which more than one strata had been combined,
special events waste was included, or adequate samples had been already been taken from
that strata, in which case samples were not taken from that vehicle load. Special events
waste in this case refers to an event called the Michigan Festival, which is a one-time
influx of a population not primarily related to campus activities and not representative of
the usual campus waste stream.

Almost all loads came to the landfill in thirty-eight cubic yard compactor trucks,
most of which were considered to be full, and therefore compacted to the same degree.
Because the residence halls do not house undergraduate students on a regular basis in the
summer, resulting in a smaller and less consistent waste stream, this was the most difficult
strata from which to collect firll, compacted loads. Although one residence hall was in
firll-time operation (Owen Graduate Hall) and some residence balls were housing high
school students for athletic camps and orientations, most of the loads fi'om this strata were
partial loads and therefore somewhat less compacted than those fiom the other two strata.

The front-loader trucks that brought the majority of the loads to the landfill ejected
their loads in the form of a “loaf”, a roughly rectangular mound of refuse that measured
approximately 30’ x 10’ x 8’. Each of these loads was then divided into twelve sections
(6 for partially full loads) and 1-2 sections were randomly selected with the aid of a
random number generator. These sections were then divided into 100-200 pound samples

with the aid of landfill personnel and a small bulldozer that transferred the samples onto

39

with the aid of landfill personnel and a small bulldozer that transferred the samples onto
tarps at the sorting site. An effort was made to ensure the samples included material fi'om
both the top and bottom of the section to reduce the potential for bias from weight-related
shifting of materials. The samples that had been placed on tarps near the sorting area were
then covered to reduce moisture gain from precipitation and loss from evaporation until

they could be sorted. A diagram of this sectioning procedure is given in Figure 1.

 

 

 

 

--------‘ ---------

 

 

 

 

 

 

 

:\
\

I

I

I

*-

'\

' \

1

I

I

I

 

 

 

 

 

 

 

10 11 ,/’ 12

Figure 1. Sectioning Procedure for Sample Selection

40

Manual Sorting Protocol

The sorting area was located in the landfill near the working face, with a 1-2 foot
pad of sand covering the landfill floor. A large tent was erected, both for storage of
equipment and as shelter for potential bad weather. It also served as a place to store
student protective gear and other belongings. The actual sorting was conducted primarily
outside the tent, unless there was rainy weather. The weighing area was located just
inside the door of the tent. A battery operated digital floor scale with increments to one-
tenth of a pound was used for weighing of materials. A diagram of the sorting site is
given in Figure 2.

Plastic containers of three different sizes were used for sorting and weighing
materials. The size container chosen for each subcategory was based on the expected
volume of each material. A color-coded label in a plastic sleeve was attached to each
container, with all materials within the same category given the same color label. This
helped both in the daily set up procedure and to avoid confirsion during sorting and
weighing. Material containers from the same category were grouped together near the
sort table with one person nominally responsible for each major category. The containers
were weighed prior to the sorting process and labeled with tare weights. They were also
weighed periodically during the sorting process to avoid the possibility of inaccurate
measurement due to occasional buildup of debris or sand. The approximate organization

of material containers and the size containers used is indicated in Figure 3.

41

a: /‘|/

! % Weighing
Station

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Roll-off
for
Waste
Recycling X
Barrels
Sort
X Table X PM“
m
2 g
0
E
a;
>
'5
“3; Tarps for
9 Samples
o- .
2
O

 

 

Working Face

Figure 2. Diagram of Sorting Site

Corrugated Junk Mail? W

Pizza Boxes While ()fiice

Newsprint Magazines Books
Mixed Oflice Paperboard

Other Paper
Groundwood

Food

Film
PETE Col. HDPE
Nat. HDPE Polypro.
PS Foam

Other Plastics

Large Containers

42

Cloth Other Manufactured
Other Textiles HH W
Clothing Food

Small Appliances

Sort
Table

 

 

 

 

g2

HHW Food

Medium Containers

Clear Glass Brown Glass
Green Glass Aerosols

Steel Cans Ferrous
Deposit Cont. Aluminum
Alum. Cont. Non-Ferrous

Food

Landscape Other Organics

Other Wood Lumber
Lather/Rubber Drywall
Asphalt/Concrete Other Constr.

Liquids

Small Containers

Figure 3. Material Categories by Container Size and Arrangement for Sorting

43

The samples were sorted manually by a team of four or five student employees.
Very large items that did not fit into the containers were removed from the samples first
and weighed separately. The remainder of the sample was placed on the sort table and
sorted into the categories listed in Figure 3. A sort table with one inch wire grids was
used, which enabled materials smaller than one inch to fall through the grid. The
preliminary study conducted in the spring of 1994 showed a significantly greater amount
of sorting time necessary for each sample when materials smaller than one inch were
included in the sorting procedure, without a significant increase in the sample weight of
individual materials. This material was weighed separately and added to load totals as a
sub-category of organic materials called residuals. Some of the materials typical for this
category included soil, sawdust, fragments of glass and plastic, small paper and foil
wrappers, and food crumbs.

After the samples were sorted into plastic containers, they were taken to the
weighing area where the weight, container tare and approximate volume were recorded on
the worksheet shown in Appendix A, and then either emptied into a rollofi’ container for
disposal or a large plastic container for recycling. The information recorded on the
worksheets was then transferred to the database program EXCEL for analysis.2 Initially,
a portable computer was on site for this purpose, but the lack of a power supply and the
threat posed by blowing sand and dust made this impractical.

With roughly the same group of students for the duration of the two week
summer study, the time necessary to sort a sample and prepare for the next one averaged

approximately thirty to forty nrinutes, depending on the size of the sample. The number of

 

2

copyright Microsoft Corporation

44

persons involved in weighing and estimating volume was kept to a minimum to reduce
recording bias, with the researcher weighing and recording the majority of sample weight
information.

Safety equipment for personnel included Tyvek coveralls, heavy shoes or boots,
hats, eye protection, and paper respirator masks. Two pairs of gloves were used, one of
woven stainless steel or Kevlar to resist penetration by sharp objects and an outer glove of
polypropylene to repel moisture. A two-hour training was scheduled to communicate
safety issues, sort procedures and material categories.

For the fall waste characterization, a similar method was used, with three
variations. The residence halls were in the process of installing compaction systems,
which meant some of the loads from this strata were delivered by the grounds department
in ten or fifteen cubic yard compacting dumpsters. The same random selection process
was used, but modified for the smaller volume of the dumpsters, and fewer samples were
taken fi'om these leads. The increased amounts of waste material for the fall semester,
(compared to summer) necessitated samples be taken for each strata from approximately
every fourth truck that picked up dedicated (individual strata) loads on the campus, using
a routing schedule worked out with the grounds department. This schedule, given in
Appendix B, was based on established pickup routes, vehicle availability, and the sampling
needs of the study.

Fall semester classes and less flexible schedules for the students sorting materials
meant that it was not possible to use the same five students for the three week sampling

period. Approximately twenty-five students were scheduled in two four hour shifts each

45

weekday, with most students working two or more shifts per week. A sample schedule is
given in Appendix C. An attempt was made to compensate for the larger number of
sorters by using the students who had participated in the summer study as “shift
supervisors,” with at least one of these students present for each four hour shift. In
addition, the director of the Office of Recycling and Waste Reduction and the researcher
were on site most of the time. Students were involved in a three hour training session on
safety procedures and categories of materials. An outline of the training session and
informational materials given to student employees involved in the fall sort is provided in
Appendix D. In addition to the training on correct sorting procedures, material containers
were examined briefly before weighing by the shift supervisor to remove any incorrectly
sorted material. Sample sorting times averaged fifty to sixty minutes, and the average
number of samples sorted per day was six. This was partially due to scheduling difficulties
and the students being less familiar with the sorting process, and partially due to poor
weather conditions on several of the sort days. Photographs of the sorting procedure are

detailed in Figures 4 and 5.

(I

K

 

Figure 4. Transferring a Sample to the Sort Table

47

 

Figure 5. Student Employees Sorting a Sample

48

The first priority for analyzing the data from the waste characterization study was
to establish the representativeness of the samples to the waste streams from which they
were taken. In other words, the samples within each strata were compared to their
respective means to find the level of potential error. The samples from
academic/administrative buildings, residence halls and university apartments were
aggregated separately to determine the sample mean percentage for each major category
(those categories comprising more than five percent of the waste stream by weight) as
well as the standard error from the mean within an eighty percent level of confidence for
both the summer and fall sorting periods. The eighty percent confidence level was chosen
because of the wide variation common to this form of waste characterization, (Gay, 1993)
and the understanding that the purpose of the data in this case was to provide an estimate
of the major components of the waste stream.

A second measure of the effectiveness of this sampling protocol was to examine
the ease of use and the reproducibility of the method used. For this, Martin’s criteria for
waste assessment protocols was used as a guideline. These criteria included
considerations like the training of employees, routing of collection vehicles, and the
reliability of the data collected.

To evaluate the second objective, the sample mean percentages for each of the
campus waste streams were then aggregated with information from current campus
landfill diversion programs and compared to the percentages of materials estimated by the

Environmental Protection Agency for municipal waste streams using multiple student

49

t-tests at a ninety percent level of confidence. Student t-tests highlight significant
differences between sample means and the population mean, in this case represented by
the EPA/Franan model. This provides an indication of significant differences in the
relative amounts of materials found in the solid waste streams of municipalities compared
to those found within a university setting, and determines whether the second hypothesis
should be accepted.

For the first objective of the study, the student t-tests were used to identify which
categories of materials differ significantly between municipal and university waste streams.
This, then, becomes the basis for examination of the individual strata to identify which
materials should be targeted for potential diversion from the waste stream to meet state
waste reduction goals.

To meet the second and third objectives, estimates of the total weights of major
materials (as defined in the first hypothesis) for the year were calculated using the
percentages from the waste characterization and the reported total weight of all general
refuse materials collected by the grounds department for each strata. For the purposes of
the study, fall characterization percentages were used to calculate total material weights
for both fall and spring semesters, making the assumption that the spring waste stream
(and campus population) was similar to that analyzed in the fall. Information for the
months of May, June, July and August was used to calculate summer diversion, and the
remaining months were considered to be part of the fall and spring semesters.

Weights of materials diverted from the waste stream were derived fi'om records

kept by those programs currently in operation at the university. These included an oflice

50

paper and newspaper recycling program run by the Office of Recycling and Waste
Reduction, a cafeteria recycling program in one of the residence halls (Brody Complex), a
corrugated cardboard recycling program undertaken by the Grounds Department, and
student run materials recovery programs in several other residence halls. These figures
were aggregated and used to estimate: 1) total yearly percentage of materials (1994-95)
diverted from the campus waste stream, excluding agricultural and special events waste
2) percentage of each major category (paper, metals, organics) currently being diverted
from the waste stream and 3) estimated weight of commonly recyclable materials that

could potentially be diverted from the waste stream on a yearly basis.

Limitations and Key Assumptions

Collection agd Sorting of Samples

There are some limitations to this type of study. As other studies of this type have
found, there is some uncertainty with the scheduling of samples and data collection, which
means the sampling method cannot be completely random. (Tchobanoglous, 1977)
Pickup schedules planned with the grounds department are still subject to the vagaries of
unforeseen situations that may necessitate last-minute schedule changes and missed
samples. The Grounds Department was not able to designate a separate truck for this
purpose, which would have allowed for a more rigorous random sampling method. It is
hoped that the adherence to random sampling technique within loads and the large number
of samples from each strata would somewhat compensate for this limitation. Because the

sorting was conducted outside in an open landfill, precipitation on several days limited the

51

number of samples that could be sorted on a given day, interfering with the scheduled
numbers of samples to be sorted as well as potentially increasing measured material
weights, especially for paper products. An effort was made to limit this bias by covering
samples, sorting under shelter if the weather was wet or by calling off sorting for any day
for which there was too much precipitation.

A firrther limitation is volume calculation of materials. Because of the variation in
structure and compressibility for different materials, the estimations were assumed to be
somewhat inaccurate. For this reason, attention has been given primarily to data based on
weight, as is customary for this type of study. The main purpose of the volume data was
to provide an estimate for calculating equipment needs, and therefore is not included in
this analysis.

A third limitation was the safety concern involved with sorting some samples.
Because of concerns for the safety of the students sorting materials, samples found to
contain excessive amounts of hazardous materials (such as unlabeled chemicals, animal
waste, or torn biohazard bags) or large amounts of broken glass or other sharps from
laboratories were discarded without being sampled. These types of samples constituted
approximately ten percent (5 samples) of the total samples taken fi'om the
academic/administrative stratum (no samples of this type were found within the residence
hall or university apartments strata), and therefore the figures for this category may be

closer to six percent than the five percent figure given.

52

Data Assum tions

When analyzing the results of the data collections, some assumptions were made
due to limitations of time and budgets. One was the lack of samples specific to the spring
semester. Although it was not possible to collect extensive samples for the spring
semester, it is assumed for this study that they would be similar to the data collected
during the fall semester for the following reasons: 1) a much smaller preliminary study
was conducted during the spring of 1994 showing results similar to those found in the fall
of 1994, 2) for municipal solid waste, one of the biggest variations between Fall and
Spring waste streams is related to yard waste, which at MSU is composted or left on
lawns and is therefore not a significant part of the waste stream, 3) other activities on
campus, as well as student population, remain fairly consistent with the exception of
moving out procedures at the end of the spring term. It was not feasible to conduct
separate characterizations of waste generated at the beginning and end of the semesters,
and it should be noted that reporting by residence hall managers indicated that this strata
may differ somewhat at these times, with increased amounts of corrugated cardboard at
the beginning of the fall semester and increased amounts of textiles, paper, manufactured
goods and wood waste from lofts at the end of the spring semester. However, much of
this material is currently being diverted by residence hall managers for reuse. (Drake,

1992)

53

Ragglatory and Structural Limitations

Each state varies somewhat in its legislation related to waste reduction. This can
have an effect on the composition of the waste streams within the state in question.
Likewise, each university can be assumed to have different types of structures in place that
might affect the results of a waste characterization study. The regulatory and structural
limitations mentioned here were in place prior to 1990 and are not included in an analysis
of the total waste stream. They may, however, impact a comparison between Michigan
State University and other universities and should be taken into account in that event.

The State of Michigan passed a mandatory beverage container deposit law in
1976. This provided for a ten cent deposit for all plastic (PET), glass and aluminum
containers used for carbonated beverages. It is difficult to predict the extent to which this
may have influenced the amounts of those materials found in the waste stream, but it is
likely that much less of these materials would be present than in states without beverage
container deposit legislation. Although there are no numbers available specific to the
Michigan State University campus, the average return rate for these containers within the
State of Michigan is estimated to exceed ninety percent. (Pasterz, 1996)

Michigan State University, like many land grant universities, has extensive open
land in and around the main campus. The majority of landscape waste from the campus
has been composted or otherwise reused by the grounds department for more than five
years, and is therefore not included in the figures presented here. It can be assumed that
percentages for landscape waste for universities without this type of program would be

higher.

54

The university has also operated a salvage facility since early in the twentieth
century. Used manufactured goods belonging to the university are sent to the facility for
resale to the general public or donation to service organizations. This includes items
ranging from furniture and appliances to computers and other electronic equipment.
Although this policy is not always followed, the salvage facility diverts a sizable
percentage of manufactured items that may have influenced the amount of materials from

that category found during the waste assessment.

CHAPTER FOUR:

DATA ANALYSIS

Review of Sampling Protocol

Because there is no one accepted protocol for conducting manual waste
characterizations, the design of the sampling protocol contains components of accepted
research methodologies, as well as components that are unique to this study. The design
of the study will be reviewed briefly prior to examining the objectives of the study.

Material Categories

Eight major groupings of materials were used, with subcategories for each, for a
total of 48 sort categories. The categories chosen were a composite of other suggested
category groupings, and were chosen for the specific informational needs of the campus
Office of Recycling and Waste Reduction. The large number of categories served two
purposes: to give a more complete picture of the waste stream as an information base for
future recycling and waste reduction efforts as well as to provide the opportunity for
combining categories in a flexible manner for comparison with other waste assessment
data. A list of the categories and subcategories is given in Table l.

S_ample Selection

Samples were taken fi'om truck loads that had picked up waste materials from one
of three building types on campus: residence halls, university apartments, and academic or
administrative buildings. Both of the sorting periods were scheduled for the approximate

middle of the semester, to avoid the anomalous waste streams found during the

55

56

Table 1. Categories of Materials for Waste Characterization

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1) Paper 2) Plastics 3) Glass
White Office Films Clear Container
Mixed Office Clear PET, #1 Green Container
Magazines/Glossy Natural HDPE, #2 Brown Container
Corrugated Colored HDPE, #2 Other Glass
Pizza Boxes Polystyrene Foam
Paperboard Other Plastics
Bound Groundwooe
Books
Newsprint
Junk Mail
Other Paper
4) Metals 5) Organics 6) Construction
Alum. Cans/Foil Food Asphalt/Concrete
Other Aluminum Landscape Waste Drywall
Steel Cans Wood Waste Lumber
Ferrous Leather/Rubber Other CD
Non-Ferrous Residuals
Aerosol Cans Liquids
Other Orgnics
7) Textiles 8) Manufactured/Other
Cloth Small Appliances
Carpetipg Clothing
Other Furniture
Packaged Goods
HHW/Medicines

 

 

 

Deposit Containers

 

 

 

57

the beginning and the end of an academic term. During the two week long summer
semester waste characterization, samples were taken from every truck load that fit this
description. During the three week long fall waste characterization, samples were taken
from every fourth truckload, because of the increased volume of solid waste generated
during the school year. One or two of twelve sections was taken from each of these leads
using a random number system, and divided into approximately 200 pound sorting
samples.

Sortirg Procedure

The samples were sorted one at a time on a sort table with one inch mesh into the
48 subcategories detailed above. Four to five students were involved sorting the samples
into three differently sized plastic containers that were marked with the material name and
tare weights. The samples were then weighed by material and the weights recorded on
a worksheet for later input into an EXCEL spreadsheet. Using EXCEL, each sample was
converted to percentage figures, dividing each material category by the total weight of the
sample taken. This allowed for comparison between samples without regard to the
variation in weight of individual samples.

A total of 160 samples were taken from the campus waste stream, with a combined
weight of 28,000 pounds (14 tons). Although an effort was made to select samples with
a weight of approximately 200 pounds, the average weight of the samples was somewhat
less than this. A breakdown of the number of samples taken from each strata and their

average weight is given in Table 2.

Table 2. Breakdown of Sample Characteristics by Strata and Semester

58

 

 

 

 

 

 

 

 

Academic! University Residence
Administrative Apartments Halls
Sum F all Sum Fall Sum Fall
Number of samples taken 22 23 25 25 40 27
Average weight of samples 164 134 181 171 198 178
(in pounds)
Weight variation among samples 80- 79- 117- 97- 91- 80-
(in pounds) 319 199 290 323 375 285
Estimated weight of waste stream 775 1550 451 901 1090 2181
during time of study (in tons)
Estimated percentage of the waste 1.8% 1.4% 4.3% 3.3% 3.1% 1.5%
stream being sampled (for 2 week
time period)

 

 

 

 

 

 

 

Evaluating the Effectiveness of A Manual Sorting Model

The first objective of the study was to examine the effectiveness of using a manual

sorting model for characterizing the solid waste stream within a university setting. As

Table 2 illustrates, the number of samples taken for each segment of the waste stream

was less than 30, with the exception of the residence hall strata during the summer

semester. This is a smaller number of samples than is generally recommended for this

type of waste characterization. The contention of this study is that this is an adequate

number of samples to characterize the Michigan State University waste stream, when it is

divided into strata with similar characteristics. In this case, that means that the summer

semester waste stream should be considered separately from the academic school year

waste stream, and that dividing the campus into residence hall (RH), university apartments

 

59

(UA), and academic/administrative buildings (AA) will improve the precision of a manual
waste characterization.

The method used to establish the representativeness of the samples entailed first
identifying those categories that made up the majority of the waste stream, by strata and
semester. Major components are defined as those materials that compose five percent or
more of the waste stream by weight. Within each strata, the samples were then compared
to their respective mean percentages to find the level of potential error. Using the
software program EXCEL, the standard error from the mean at an eighty percent
confidence level was calculated for each component by strata and semester. These figures
are given in Table 3.

To evaluate the effectiveness of this model from the data presented in Table 3, it is
necessary to examine the first hypothesis of the study, which states:

Ho: Using a manual sorting model with fewer than thirty samples, the percentages by
weight of major components of a university solid waste stream cannot be estimated within
an eighty percent confidence level and a 3: two percent error from the mean.

H1: Using a manual sorting model with fewer than thirty samples, the percentages by
weight can be estimated with an eighty percent confidence level and a j; two percent error
from the mean.

The results of these calculations show it is possible in most cases to estimate the mean
percentages of major components of each waste stream within an eighty percent
confidence level and a two percent error, using this methodology. Given these results,
the null hypothesis should be rejected. For those components that had a greater than two

percent error, further comments are made in Chapter Five. For the sake of comparison,

the same procedure was used to calculate the standard error from the mean for the major

60

Table 3. Percentage Means and Percent Standard Error of Major Components of
the Michigan State University Waste Stream by Semester and Strata

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Strata Material Percentage of Standard %
Category the Waste Error from
Stream Mean
AA-Summer (n = 20) l. White/Mixed OP 8.6 1.0
2. Corrugated 10.9 1.5
3. Newsprint 5.6 0.87
4. Other Paper 14.8 1.5
5. other organics 7.4 2.3
6. HHW 6.1 1.7
AA-Fall (n = 23) l. white/mixed OP 6.7 0.63
2. corrugated 9.6 1.4
3. newsprint 15.8 2.8
4. other paper 12.4 1.4
UA-Summer (n = 25) 1. other paper 9.2 0.75
2. food 29.5 1.9
3. HHW 7.7 0.9
UA—Fall (n = 25) 1. corrugated 5.3 1.1
2. newsprint 6.8 1.2
3. other paper 11.2 1.5
4. plastic film 5.1 0.49
5. food 25 2.3
6. HHW 7.6 1.1
RH—Summer (n = 27) 1. corrugated 7.7 0.76
2. other paper 10.2 1.1
3. plastic film 6.2 0.56
4. food 23.5 2.8
5. construction 5.5 1.2
RH-Fall (n = 40) 1. corrugated 8.2 1.1
2. newsprint 9.2 1.4
3. other paper 16 1.3
4. plastic film 7.2 0.8
5. food 16.8 1.7

 

 

 

 

 

 

 

61

components of the waste stream as a whole, using the entire population of 160 samples.

The results are given in Table 4.

Table 4. Percentage Means and Standard Error from the Mean for Major
Components of the Michigan State University Waste Stream

 

 

 

 

 

 

 

 

 

Material Category Percentage of the Standard % Error
Waste Stream From Mean
1. Corrugated 18.4 0.46
2. Newppfint (ONP) 7.4 0.64
3. Other Paper 12.1 0.55
4. Food 18.4 1.16
n = 160

 

 

Estimates of Current Diversions from the Waste Stream byStrata

The second objective of this study was to quantify the weights of specific materials
being diverted from the waste stream through current Michigan State University recycling
programs, and then use these figures and the information from the waste characterization
to estimate the potential diversion rates for commonly recycled materials, for the purpose
of complying with state and federal goals for waste reduction.

This was accomplished by first identifying those programs on the campus that
divert materials from each of the three stratas used in the study. The types and amounts of
materials diverted was estimated from records kept by those programs involved in
recycling efforts on the Michigan State University Campus. This includes the Office of
Recycling and Waste Reduction, the Grounds Department, and a residence hall recycling

program run by undergraduate students. Because some estimates were made by months,

62

the months of September through April were considered to be part of the Fall/Spring
semesters, and the months May through August were considered to be part of the summer
semester. Although the final calculation for potential diversion is given as an academic
year total, the current diversion amount by program is separated into F all/Spring semesters
and Summer semester because some programs were not in operation during the summer.
Diversions being made from the solid waste stream at Michigan State University at the
time of this study are described below according to strata.

A drop-off program run by the Office of Recycling and Waste Reduction collected
a total of 38.1 tons of recyclable material from the university apartments during the
1994-95 academic year. This constituted 2.7% of the university apartments waste stream.

A breakdown of amounts by weights and materials is given in Table 5.

Table 5. Waste Stream Diversion from University Apartments, 1994-95

 

 

Material Summer Amounts Fall/Spring
(tons) Amounts (tons)
1. Papers 7.7 16.2

(cardboard, office
paper, newspaper, magazines)

 

 

 

 

2. Container glass 1.6 4.8
3. Plastics 1.4 2.7
(clear/colored HDPE,

clear PET)

4. Metals 0.96 2.6

 

 

 

(tin/steel, aluminum)

 

63

A paper pick-up program run by the Office of Recycling and Waste Reduction
collected 805.7 tons of materials from academic and administrative buildings for the
1994-95 academic year. This constituted 25.7% of the academic and administrative solid
waste stream for this time period. A breakdown of materials and amounts is given in

Table 6.

Table 6. Waste Stream Diversion from Academic and Administrative Buildings,
1994-95

 

 

 

Material Summer Amounts Fall/Spring
(tong Amounts Qons)
Paper 270.7 535
(white/mixed office
paper, newspaper,
magazines and otheQ

 

 

 

 

Four programs were in operation during the 1994-95 academic year to collect
residence hall materials. One was a volunteer program run by undergraduate students in
some of the residence halls for door-to-door collection from residents. The second was a
cafeteria-based collection program run by cafeteria and residence hall personnel. The third
was a corrugated cardboard collection program run by the Michigan State University
Grounds Department for part of the year. There was also a limited program through the
Office of Recycling and Waste Reduction, estimated at five percent of program totals. The
combined weight of these programs is 254.6 tons, which constitutes 7.2% of the residence

hall waste stream. A breakdown of these figures is given in Table 7.

64

Table 7. Waste Stream Diversion from Residence Halls, 1994-95

 

 

 

 

 

 

 

 

 

Material Summer Amounts Fall/Spring
(tons) Amounts (tons)

1. Paper 50 197
(white/mixed office paper,
newspaper,corrugated
cardboard,magazines and other)
2. Metals 0.5 3
(aluminum, tin/steel)
3. Container glass N/A 2.7
4. Plastic v 0.1 1.3
(clear/colored PET, clear/
colored HDPE, Styrofoam)

 

Overall, 1098.4 tons of materials were diverted from the university waste stream
by all of the recycling programs combined. With a total estimated waste generation rate of
8046.4 tons for the year, (including waste disposed and waste diverted) the overall

recycling/diversion rate for the university was estimated to be 13.6 %.

Potential for Additional University Waste Stream Diversflg

The goal of the second research objective was to estimate potential diversion rates
for the University, using the figures estimated above and the composition information
derived from the waste characterization. Potential diversion rates were calculated using
the percentages derived from the waste sort for commonly recycled materials, including:
white and mixed office paper, newspaper, magazines, corrugated cardboard, metal

containers and foil, HDPE (#2) and PET (#1) plastic containers, and glass containers. The

percentage amounts by weight for these material categories derived from the waste sort

were multiplied against the total estimated weight of the waste stream for each strata.

Using the resulting figures from this, estimates for the current and potential diversion

amounts by weight for the 1994-95 academic year are given in Table 8. A graphic

representation is given in Figure 6.

Table 8. Estimates of Current and Potential Diversion from the Michigan State
University Waste Stream by Strata

 

Current Diversion

Additional

Current and

 

 

 

 

Amounts (tons) Potential for Potential
Diversion (tons) Diversion
Percentage
Amounts By
Strata
Academic/ 805 .7 797 51%
Administrative
University 38 344.8 27.5%
Apartments
Residence Halls 254.6 1096 38.3%

 

 

 

 

Overall, assuming the distribution of materials discarded on the campus remains

the same, there is the potential for diverting approximately 2,238 additional tons of

material per year, for an overall recycling rate of 41.4 % of the solid waste stream. This

is much greater than is currently being realized on the campus. Implementation of more

comprehensive recycling programs, even without additional source reduction or

composting programs, would put the University much closer to the diversion rates being

realized at other state universities and well on the way to reaching future state recycling

goals of 50% waste reduction by the year 2005.

 

66

    

    

Academic/Administrative Buildings. 1994-95 Residence Halls. 1 994-95
Current Waste
Current Waste SM
Other Stream Other 0W Paternal Waste
19% 0' I 3' 5'0" 44% 14% Stream
26% Diversion
(Recvcmsl
25%
Potential Waste Potential Waste
Stream Stream Potential Waste
Dwersion Dwersrun Stream
(Composting) (R ”yang, Diversion
30% 25% (CW!!!)
1 7%
University Apartments, 1994-95 Michigan sum University. 1994-95
31'9"“ Potentral Waste Stream 0
m We“ Stream
‘ (Recycling)
25%
Potential Waste
Stream
Diversion
(Cumming)
0‘5

Current Waste Stream Diversion: Current programs in operation, including office
paper, corrugated, newsprint, magazines, #1 and #2 plastics, and metal/glass containers
Potential Waste Stream Diversion (Recycling): Same materials as listed above, but
with a more comprehensive collection system

Potential Waste Stream Diversion (Composting): Including the diversion of food and
other organic materials

Figure 6. Estimate of Current and Potential Diversion From the Michigan State
University Waste Stream, By Strata

67

mafison with 1994 EPA/Franklin Estimates for Municipal Waste Stfla_m_s

The third objective of the study was to identify significant differences between
Michigan State University’s solid waste stream and current United States municipal solid
waste stream estimates. The EPA/Franklin model was chosen as the best indicator of the
composition of municipal solid waste streams, with the estimates used here taken from the
1994 report of 1993 waste stream data. There were three major calculations used to
complete a comparison between the percentage composition of the MSU solid waste
stream and the EPA/Franklin estimates for municipal waste streams.

The first step was to group the subcategories used in the university waste stream
characterization using the category divisions from the EPA/Franklin model. In most cases,
divisions were similar. An illustration of how those subcategories were grouped is given
in Table 9. The mean percentages for subcategories were added to obtain percentages for
divisions in those cases where more than one subcategory was incorporated.

Table 9. Incorporating the Michigan State University Waste Characterization

Categories Into the Category Divisions of the EPA/Franklin Model for Municipal
Waste Streams

EPA . metal
MSU . all metal

MSU ' . food

EPA . wood
MSU ' lumber wood

EPA . other

MSU :

and wood waste construction debris
textiles and manufactured

food

 

68

The second step was to aggregate the mean percentages for all samples, regardless
of strata and semester, to determine the overall mean percentage for each of the EPA
categories. These figures were then used to estimate the weight in tons for each category
for the academic year 1994-95. This was done so that the weights of materials diverted
from the waste stream could be added back in for the purpose of this comparison, as the
figures given in the EPA/Franklin model are prior to waste diversion (recycling) programs.
After these figures were added to the total, the weight in tons was recalculated as
percentage figures using the computer program EXCEL.

The hypothesis for determining significance is given as:

Ho: The percentage of major categories of a university solid waste stream do not differ
significantly fiom the estimated percentages of major categories of municipal waste
streams, at a ninety percent confidence level.

H1: The percentages of major categories of a university solid waste stream do differ
significantly form the estimated percentages of major categories of municipal solid waste
streams, at a ninety percent confidence level.

Using these aggregated mean percentages as representative of the University’s
total waste stream prior to diversion for recycling, a student t-test was used to test for

significant differences between the two waste streams at a ninety percent confidence

interval for each of the major categories. The formula used is:

t= (X-g) ltl : tor/2

s/n

where X = mean sample value for a given strata and material
u = population mean as calculated by EPA n = number of samples
3 = standard deviation of the sample mean
t (112 = 1.711 with 24 degrees of freedom and a 90% confidence interval

69

Results of the t-tests for each category are given in Table 10. Using the definition for

significance given in the hypothesis, the mean percentages for municipal waste streams are

significantly different than those of the University, for every category except plastics. This

indicates that the null hypothesis should be rejected in this case. A graphic comparison of

the mean percentages for each material category is shown in Figure 7.

Table 10. Comparison of Michigan State University’s Solid Waste Stream
with EPA/Franklin Municipal Waste Stream Figures Using T-test

 

 

 

 

 

 

 

 

 

Material University Population SD T-value Conclusion
Mean % Mean %
paper 48.4 37.6 15.4 9 reject null
(more than MSW)
plastic 8.7 9.3 5 1.5 do not reject null
(similar to MSW)
glass 2.9 6.6 2.8 16.8 reject null
(less than MSW)
metal 3.2 8.3 3.8 17 reject null
(less than MSW)
food 15.9 6.7 14.6 7.9 reject null
(more than MSW)
wood 1.1 6.6 2.4 28 reject null
(less than MSW)
landscape 1.2 15.9 2.5 74 reject null
(less than MSW)
other 18.2 9.0 2. 5 46 reject null
(more than MSW)
Using a t-value of 1.658, with n=l62 and a two-tailed alpha of .052

 

 

2 t value is taken from Statstical Concepts grid Methods, Gouri K. Bhattacharyya and Richard A. Johnson,
John Wiley and Sons: New York, 1977.

 

 

70

 

     
    
 

 

5%

45‘ f IMSUbyweight j
l

40. 5 DMSWbyweight l

 

f.
7

   

 

 

’5
35 “'4 e4
4 in: g
m. in. u
m as
Percent h“ E5
by 0'9 3;; :3
e“ ’23:
:11 2::
fit :2
104 a.» .4
if». M
5-1 it: T33
H ‘--:
o_ 4.3;

g i3

Figure 7. Comparison of Michigan State University’s Solid Waste Stream with
EPA/Franklin Municipal Solid Waste stream Estimates by Material Category
Percentage of Total

71

A comparison of the EPA/Franklin municipal solid waste estimates was also done
separately with each of the three university strata, using the same mathematical process.
Results are given in Table 11. Out of the eight material divisions, significant differences in
mean percentage amounts were found for almost all categories within each strata. This is
an indication that Michigan Sate University’s waste streams should be considered
significantly different than municipal waste streams inmost respects, with appropriate
modifications in the types of materials that should be targeted for diversion and reduction

programs.

Table 11. Comparison of the Michigan State University Waste Stream with
EPA/Franklin Municipal Waste Stream Figures by Strata Using T-test

 

 

 

 

 

 

 

 

 

 

 

 

Strata and semester n= Results
Academic/Administrative, summer 22 all materials significant
Academic/Administrative, fall 23 all materials significant except
plastic

University Apartments, summer 25 all materials significant

University Apartments, fall 25 all materials significant except
paper

Residence Halls, summer 40 all materials significant except
paper

Residence Halls, fall 27 all materials significant

Using a t-value of 1.658, with a two-tailed alpha of .05.

 

CHAPTER FIVE:

SUMMARY AND CONCLUSIONS

The costs of solid waste land disposal are rising. This is partially a result of
concern about perceived environmental and public health risks, as well as legislative
measures that limit where and how landfills can operate. To reduce concern about these
potential risks and increase the effective use of available resources, alternatives to
landfilling should be an integral part of any waste management plan.

Characterizing the waste stream is a necessary first step in the development of
effective waste reduction programs, which include source reduction and reuse strategies as
well as recycling and composting programs. It is the contention of this study that a
manual sorting model is an effective method for characterizing university waste streams.

University waste streams present both unique challenges and unique opportunities
for waste reduction. Some of the distinguishing characteristics of university waste streams
that make them different than either municipal or industrial waste streams are: 1) the
densely populated residence hall system, with large numbers of transient same-age
students, 2) the academic scheduling found on a university campus, including semester
anivals and departures, and 3) the types of activities and functions that are common to
universities. Because there is no one established protocol for conducting a manual waste
characterization, these considerations had to be taken into account when designing this

manual sorting model.

72

73

Because of the academic calendar, the sampling procedure was scheduled to take
place during the middle of the semester to avoid the waste streams associated with
activities at the beginning and the end of the semester. Sampling was scheduled for both
the summer and the fall semesters of 1994, to determine if there were major differences
between semesters in the composition of the waste stream. The campus waste stream was
divided into three strata: residence halls, academic and administrative buildings, and
university apartments. This division was based on the assumption that each segment of the
campus waste stream would have distinct characteristics. The Grounds Department at
MSU delivered truckloads from each of the three strata. Samples weighing approximately
100-200 pounds were taken using random sampling methods. The samples were sorted by
student employees into 48 material categories chosen to provide a precise picture of the

waste stream composition.

Evaluating the Effectiveness of the Manual Sorting Model in Predicting Major
Commnents of the Univers_ity Waste Strem

The first objective of this study was to evaluate the effectiveness of a manual
sorting model for characterizing the waste stream within a university setting. The
effectiveness of this manual sorting protocol was evaluated primarily by examining the
information gathered for major components within each of the three stratas. Major
components in this case refer to those materials found to comprise 2 5% of the waste
stream by weight. The samples taken for each semester and strata were compared to their

respective mean percentage values at an 80% level of confidence to evaluate if the mean

74

percentage for major components could be predicted within a :t 2% error. Stated as a
hypothesis,

Ho: Using a manual sorting model with fewer than thirty samples, the percentages by
weight of major components (greater than 5 percent) of a university solid waste stream
cannot be estimated with an eighty percent confidence level and a :l: two percent error
from the mean.

H1: Using a manual sorting model with fewer than thirty samples, the percentages by
weight of major components (greater than 5 percent) of a university solid waste stream
can be estimated with an eighty percent confidence level and a i two percent error from
the mean.

For all but three materials, it was possible to predict the mean percentage by
weight for each material within this 2% standard error. The null hypothesis was rejected
in this case. This indicates that the manual sorting model used provides a highly
acceptable means for identifying the major components of the solid waste stream within a
university setting. Dividing the waste stream into three strata with similar characteristics
reduced the number of samples needed to attain a usable precision for waste
characterizations.

Those materials that were not accounted for within the two percent error include
food (UA fall and RH summer) , newsprint (AA Fall) and other organics (AA summer).
With the exception of other organics, these materials comprise a large (greater than
fifteen percent) segment of their respective waste streams. The error for each of these
categories was greater than two percent, but less than three percent. Possible
explanations for these variations, based on researcher observation, include: 1) food waste

in university apartments during the fall semester- the greater part of the population within

this strata rely on university assistantships or state aid. Both of these income sources are

75

received on a monthly basis, meaning that the timing of food purchase and consumption
are probably influenced, with a greater consumption and discard rate immediately
following disbursement and a lesser consumption for the remainder of the month; 2) food
waste in residence halls during summer semester- there is no long-term population of
students in the residence halls during the summer, with the exception of Owen graduate
hall. The residence halls are not operating on a continuous basis, and some are not
occupied at all. The intemrittent operation of the residence hall cafeterias helps account
for the great variability in food waste generation between samples during this time period;
3) newspaper waste in academic and administrative buildings during fall semester-there are
three factors that may account for the higher variability of newspapers in the waste stream:
weekly newspaper subscriptions by staff and faculty, the delivery schedule of the campus
newspaper, and variability between buildings with staff participation in the campus
recycling programs. Discarded newspapers are picked up by the Office of Recycling, but
the participation among academic and administrative units is voluntary and variable.
Observation during the waste sort indicated newspapers usually came in groups, probably
daily discards of newspapers that were not read and were not placed in recycling
containers; and 4) other organic waste from academic and administrative buildings during
summer semester- this appears to be due to animal bedding and other materials that were
only generated by certain buildings involved with animal research and care, causing a wide
variation in sample amounts. Very little plant matter was noted, although soils were

periodically found in the waste stream from this strata.

76

Muating the Effectiveness of the Manual Sorting Protocol Used to Gather Sampling
Data

Although the quantitative measurement of this manual sorting model is the best
indication of its effectiveness in predicting the components of the waste stream,
effectiveness can also be evaluated in a qualitative manner. Martin’s criteria for evaluating
waste stream analysis protocols made a usefirl framework for evaluating the protocol used
in this waste characterization.

The first criterion was that the protocol should require the minimum time
necessary to attain the desired confidence level. To achieve the desired confidence level of
80% for this study, at least twenty samples needed to be taken from each strata during
both the summer and fall semester sampling periods. To reach this goal of sixty samples, it
was estimated that four sorters could sort eight samples per eight hour day. This estimate
meant that eight days would be needed for the sorting procedure, with one day for setting
up the site and one day for breaking it down. For the summer sort, this estimate was fairly
accurate: as stated in Chapter Three, each sample took approximately 30-40 minutes to
sort, with additional time needed for setting up the site in the morning and storing the
materials at night. Because the site did not have cleanup facilities, additional time was
needed for lunch-time cleanup and transportation. Fifteen minute breaks were observed
every four hours. For the fall sort, it was estimated that additional time would be needed
to sort the samples due to the relative inexperience of the student employees, as well as
the additional time needed for transportation to and flour the campus. This was indeed the
case, with sample sort times increasing to an average of 50-60 minutes, with six samples

sorted per day. Although more time was needed to sort the samples in the second

77

instance, it is felt the accuracy of the results was not significantly affected, and that this
type of schedule is more feasible for those waste sorts utilizing student assistance during
the academic school year.

Martin’s second criterion was that the resulting data be accurate, reliable, and
valid. For the purposes of this study, a confidence level of eighty percent was adequate.
In most cases, the data gathered using this protocol were accurate within a small amount
of error at this confidence level. There is always a tradeoff between the number of
samples gathered and the total cost of the study. In this case, the required precision was
obtained while staying within desired cost parameters.

Two of Martin’s criteria relate specifically to the assumption that manual sorting
protocols are conducted at sanitary landfill facilities. The criterion that consideration
should be given to the pattern in which trucks bring waste to the landfill was important to
this study because of the need to examine each of the three strata separately. A schedule
had been set up with the Grounds department prior to the beginning of the study, and
frequent radio contact helped insure that the correct types of loads were being delivered
for sampling. The criterion that landfill operations should be disrupted as little as possible
was also taken into account in this study. Some of the ways the protocol tried to avoid
disrupting the normal schedule include: sorting near the working face, agreeing to specific
beginning and ending dates, keeping lines of communication open with landfill staff, and
following all traffic and safety guidelines.

An important criterion for this or any protocol is the training of employees on

waste sort procedures and safety considerations. This manual sorting model enabled

78

employees to understand the protocol with less than four hours of training during both the
summer and fall semesters. Training of sorters is essential to the collection of accurate
data as well as the safety of individuals. Comparing the summer and fall sorting
procedures, it is certainly desirable to have the same employees sorting all samples, but
using students to sort samples means that this is not always possible. Four hour shifts
were a feasible alternative, and there did not seem to be any difference in the accuracy of
the resulting data. A very usefirl supplement to the training session was the supervision

during the fall sort by those individuals already experienced in the sorting protocol.

T_argeting Further Waste Reduction Efforts

The second objective of this study was “to provide the opportunity to identify
those materials and stratas that would provide the best opportunity for firrther waste
reduction efforts.” Currently, the university is not meeting state and federal goals for
waste reduction, and lags behind other public universities within the state regarding the
amount of waste being diverted to recycling programs.

The first step in increasing the amount of waste being diverted to recycling
programs is to identify those materials that should be considered for inclusion. As
previously stated, the choice of targeted materials should be based primarily on three
considerations. One consideration is the amount of material present in the waste stream;
the more material that is available for diversion, the more successful will be the effect on
the waste steam as a whole, and the more likely there will be a market for the materials

that are diverted. Another consideration is ease of collection and transportation of

79

materials. If materials are very bulky (large volume to weight ratio), requiring larger
collection vehicles or more frequent collection, or present a hazard during storage or
collection, there will be greater costs associated with them, making their diversion less
attractive. Finally, there needs to be a market for the materials, making it easier to find
the collection programs and ensuring that diverted materials are used to complete the
recycling process.

The material categories previously identified as having an established market value
are: white and mixed office paper, newspaper, magazines, corrugated cardboard, metal
containers and foil, HDPE and PET plastic containers, and glass containers. Of these,
corrugated and newsprint compose 25% of the waste stream overall, and office paper
composes 25% of the academic/administrative strata. These materials represent the
potential to divert an additional 2800 tons of waste materials from the University’s waste
stream, with an overall potential recycling rate of over 40%. Simultaneous efforts toward
source reduction or reuse, which should be an important component of any effective waste
management program, could firrther reduce the amount of waste being landfilled.

The introduction of a more comprehensive waste diversion plan that included
compostable materials in addition to the recycling materials already listed would mean the
potential for greater reduction in both the amount of waste material that would need to be
landfilled, and disposal costs. If the figures from the categories of food, wood, landscape,
and other organic materials were added to the materials listed above, the total potential

diversion rate would be over 60% of the university waste stream.

80

Cormiarison of Universityand Municipal Waste Stream

The third objective of the study was “to compare the waste stream analysis for the
university to the EPA/Franklin model for municipal waste streams, to highlight the types
of components differences found within university waste streams.” The hypothesis was:
Ho: The percentages of major components of a university solid waste stream do not differ
significantly from the estimated percentages of major components of municipal solid waste
streams, at a ninety percent confidence level.

H1: The percentages of major components of a university solid waste stream do differ
significantly from the estimated percentages of major components of municipal solid waste
streams, at a ninety percent confidence level.

Results from student t-tests indicated the percentages of material categories in
most cases were significantly different (0.90) from municipal waste streams, thereby
indicating the rejection of the null hypothesis. Based on this analysis, it is clear that
university waste streams contain larger amounts of paper and food than municipal waste
streams and smaller amounts of glass, metal, landscape waste and wood. The amount of
plastic found in the university waste stream was similar to that estimated to be present in
municipal waste streams.

Breaking the university waste stream down into the three strata of
academic/administrative, university apartments, and residence halls, most material
categories continued to be significantly different in amount than municipal waste streams.
This breakdown can serve as a basis for targeting waste diversion programs to the
individual characteristics of each strata. Paper products, found in large amounts within

two of the three strata, are one example of ways to use this analysis to target appropriate

materials when planning waste diversion programs.

81

Paper was not found to be significant during the fall in the University Apartments
and during the summer in the residence halls. The research makes clear that the
academic/administrative units are the most consistent and largest discard source for paper
products, making this strata the best focus for paper recovery programs. In addition, the
academic/administrative strata contributes the largest percentage of the waste stream
(51%), increasing the potential amount of paper recovery for the campus. Although paper
was not found to be significant in residence halls in the summer, the greater discard rate
for paper products during the school year indicates recycling programs would be more
cost effective if paper pickup programs for residence halls are only active during that time
period.

The amount of food waste was significantly greater than municipal systems for
both the residence hall strata and the university apartments strata, and significantly less for
the academic/administrative units. This shows that any program intended to divert food
(organic) waste would be of most benefit for those strata that produce significantly more
food waste than municipal systems, and should not include academic and administrative
buildings.

For universities unable to conduct a waste characterization, this study provides
insight into the composition of their waste streams and direction for planning diversion
programs. Regional and methodological differences will undoubtedly result in some
differences; however, this study provides baseline information and a starting point for the

establishment of a recycling and integrated waste management plan.

82

Recommendations for Further Research:
Because the data base for this type of research is very limited, there is a great need
for additional research in this area. The following studies are recommended:

1) Conducting waste assessment on other university campuses using the same
protocol to see if the results are transferable to another setting.

2) Designing a waste assessment protocol with fewer material categories, to
examine the theory that those categories for which very small amounts of materials
are present are not accurately represented in a manual sorting protocol.

3) Conducting a similar study on the same university campus following the
introduction of more comprehensive waste diversion programs, which would
allow for quantification of the actual diversion rate compared to the potential
diversion rate as estimated in this study.

4) Using the data from this study to estimate and compare the average cost per ton
for the development and operation of waste diversion systems that a) collect easily
marketed recyclables, b) collect all recyclable material and c) collect recyclables
and compostibles, and compare these costs to current waste disposal costs.

APPENDICES

83

Appendix A

Sample Worksheet for Recording Weights of Material Categories

1 White Ofi'rce
2 Mixed Office
3

4 C

5 Piaa Boxes
6

7 Bound Groundwood
8 Bound Books
9
10 Junk Mail
11 Other

Plastics
12 Film
13 Clear PETE
14 Natural HDPE
15 Colored HDPE
16 PP
l PSFoam
18 Plastics

Glass
19 Clear Containers
20 Green Containers
21 Brown Containers
22 Glass

Metals
23 Alum Containers/Foil
24 Aluminum
25 Steel Cans
26 Ferrous Metal
27 Non-Ferrous
28 Aerosol Cans

Containers

30 Food

31 Waste
32 Wood Waste
33 Leather. Rubber .
34 Residual

35 '

36

 

84

Appendix A (cont’d)

Material

38
39

C onstructron Debris
40 oncrete

4| Drwvall
42 Lumber
43 Other

Manufactured Goods
44 Small
45 '
Furniture
4 Other Manufact.
HHW

 

85

Appendix B

Routing Schedule for Grounds Department

 

GROUNDS SCHEDULE-

FALL WASTE SORT

AM. PM.

Monday. October 24th Mon 0 ct 4th
Academic/Administrative ”hi/95W Apartments
Tu d r m TU t in

Residence Halls Residence Halls

W nesdo t in W n 1 th

University Apartments Academic/Administrative

W W

Academic [Adminmflves Residence Halls

Fri 1 2 r Fri ct th
Pending-

Pending

86

Appendix C

Sample Schedule for Student Employees

TUESDAY

Pete
Janey

 

Marsha

 

1. Sara

2. Tisha
3. Candice
4. Cheryl
5. Don

TUESDAY

 

 

1. Dana

2. Danielle
3. Jeff

4. Andrea
5. Tisha

6. Justin

 

 

 

 

 

87

Waste Sort Training Information

Waste Sort Training

I. WELCOME!

A. Introductions
B. Why we're doing this...

11. SAFETY

A. Make sure you have all of your safety equipment, and that it is labeled
B. Sorting technique to avoid risk

C. Immunizations (tetanus, hepatitis B)

D. What to do if there is medical or hazardous waste

E. Reporting accidents

III. STRATA AND CATEGORIES

A. Three main strata we're working with
B. Waste sort categoires
1. list of materials
2. examples of categoires
C. What to do with "white goods" or other odd items
D. Setting up samples, Grounds schedule

IV. EQUIPMENT AND SORT PROCEDURE

A. Setting up sort containers
B. Presorting (p.m.)
C. Loading samples on sort table
I. pull out heavy/bulky materials
2. try to avoid adding sand to load
D. Sorting
1. pick a group of materials and try to stay with same group
2. one food bucket for each side of table
E. Weighing containers
1. each person is responsible for their own containers, but please help
someone else if you finish
2. take materials to supervisor to weigh, calling off the tare and estimating
volume to the nearest tenth (.1, .5, full)

88
Appendix D (cont'd)

3. After materials have been weighed, either empty into rolloff or

appropriate recycling barrel

4. Set up containers for next sample

5. Make sure resiudal material form sort table is emptied and weighed
F. Cleanup

1. For am. shift, leave tale and containers out for p.m. crew unless

weather inidcates no p.m. sort

2. Hang up equipment and make sure all containers are empty

3. For pm. shift, make sure table and containers, as well as equipment, are

in tent, and tent is fastened

4. Don't forget to sign the payroll sheet in the van!

V. HOUSEKEEPING

A. Payroll status
B. Schedule
1. any last minute changes?
2. who to call if you can't meet the van at Salvage
C. Transportation
1. van from Salvage
2. campus pick up
3. driving seperately
D. Ifthe weather is bad...

89

Appendix D (cont'd)

WASTE SORT UPDATE

October 7, 1994

Schedulinaandhymu:

Schedules as we know them are attached to this sheet. Ifyour schedule will be
different, please let Marsha or Pete know. Payroll runs from Wednesdays to
Tuesdays; please make sure you sign the payroll sheet that will be kept in the van. If
you do not sign the sheet, we will not know to pay you!

I . . _
There are two training sessions- one on potential pathogens and the other on waste
sort procedure. It is important that you attend both of these! Ifyou cannot attend
the Friday, October 7 meeting, there are alternate dates. If you are unable to attend
the Friday, October 14 meeting (from 1:00-2:30) please let us know so that we can
schedule an alternate time.

mm

For the morning shift, we plan to be at Granger at 8:00. The van will leave Salvage
at 7:40, and swing by Wells Hall at 7:45 before heading to Granger. Ifyou have
made alternate arrangements for transportation, please let the student supervisor
for your shift know. The morning shift will be leaving about 11:45 to go back to
Wells Hall at 12:35, and head out to Granger. Again, if you need to drive
separately, let us know. The afternoon shift will be leaving Granger about 4:40 to
swing by Wells Hall and back to Salvage. There is parking available at Salvage.

Bathing:

Long pants and long-sleeved shirts are recommended for safety reasons. Please also
wear closed-teed shoes with firm soles. On rainy days, boots are recommended.
Coveralls, filter masks, gloves, goggles and hats wo;; be provided. You may want to
bring an extra pair of shoes, so you can leave your "sorting shoes" on site.

E I]! . I [H .
Drinking water will be available, as well as a cooler if you wish to bring other
beverages or snacks. There is a bathroom and sink available, although not directly
on site. We will try to find a storage system so that gear may be left in the tent
between shifts.

90

Appendix D (cont'd)

WHERE SHOULD IT GO?

Here is a list of some items that are commonly found, and the category they should

be placed in:

Batteries

Beverage containers, carbonated
Beverages containers, non-carb.
Metal can lids

Candy wrappers, chip bags
Ceramic dishes or vases
Cleaning supplies, partially full
Cleaning supplies, empty
Curling irons/hair dryers
Chairs, desks, stools

Cleaning sponges

Computer paper

Computer paper, groundwood
Diapers '

Envelopes with windows or unopened
Lab glassware (Pyrex, Kimax)
Latex paint cans
Linoleum/molding

Masking Tape, in wads

Cereal boxes (unopened)

Cereal boxes (half full)

Cereal boxes (empty)

Paper towels

Pens/Pencils (Usable)

Phone books, schedule books
Prescription Medications

Shoes (Usable)

Shoes (Unusable)

Window or plate glass

HHW

Deposit containers
By container type
Steel cans

Film

Other glass

HHW

By container type
Small appliances
Furniture

Other plastics

White office

Mixed paper

HHW

Junkmail

Other glass

Other CD

Other CD

Other CD

Other Manufactured
Food

Paperboard

Other paper

Other Manufactured
Bound groundwood
HHW

Other Manufactured
Leather or Plastics
Other Glass

BIBLIOGRAPHY

91

BIBLIOGRAPHY

Abbott, Stan. “Academic Awakening.” Geoggaphical Magazine Feb. 1992: 5.
Alexander, Judd H. In Defense of Garbage. Westport, CT: Praeger Publishers, 1993.

Alexander, Michael. Market-Based Planning and Implementation Guide for Local
Recycling Proggams. Lexington, KY: The Council of State Governments, 1992.

Alter, Harvey. Materials Recovery from Municipal Waste. New York: Marcel Dekker,
Inc., 1991.

Apotheker, Steve and Jerry Powell. “How Are We Doing?: The 1994 Report and More.”
Resource Recycling July 1995: 34-47.

Bachman, Jody. Personal interview. October 1995.

Blumberg, Louis and Robert Gottlieb. War on Waste: Can America Win the Battle with
Garbage? Washington, DC: Island Press, 1989.

Christensen, John W. Global Science. Dubuque, Iowa: Kendall/Hunt Publishing, 1991.

Clarke, Marjorie J. “Waste Characterization Studies and the Solid Waste Hierarchy.”
Resource Recycling February 1992: 75-84.

Corbitt, Robert A. Standard Handbook of Environmental Engineering New York:
McGraw-Hill, 1990. .

Creighton, Sarah et a1. Solid mste Reduction in Colleges and Universities. Center
for Environmental Management, Tufts University 1993.

Curbing Waste in a Throwa_way World: Report of the Talk Force on Solid Waste
Management. Washington, DC: The National Governors’ Association, 1990.

Data Summary of Municipal Solid Waste Management Alternaaiiaea, vol. I. National
Renewable Energy Laboratory for the U. S. Department of Energy, October 1992.

DeBell, Jack. “Coming of Age: Recycling on Campus.” Resource Recycling September
1994: 44-52.

92

Delin, Mark. “Source Sampling for Waste Characterization.” Biocycle July 1992: 76-
79.

Drake, William. Personal interview. May 1992.

Economics of Recycling and Recycled Materials. Seattle, WA: Sound Resource
Management Group, Inc. for the Clean Washington Center, 1993.

Fessenden, Ford. “Landfills Head for the Scrap Heap,” in Rush to Burn: Solving
America’s Garage Crisis? Washington, DC: Island Press, 1989.

Fields, Scott. “Recycling: Reading the Bottom Line.” Touchstone April 1992: 27-
31.

Fishbein, Bette K. and Caroline Gelb. Making Less Garbage: A Planning Guide for
Communties. New York: Inform, Inc., 1992.

F012, David H. and Joseph M. Hazlett. “A National Survey of Local Government
Recycling Programs.” Resource Recycling December 1990: 82-86.

Franklin, Marjorie. “Solid Waste Stream Characteristics,” in Integgated Solid Waste

Management: Options for Legislative Actior_t, edited by Dr. Frank Kreith,
Schnectedy, NY: Genium Publishing Co., 1990.

Friedman, Fred. “Do We Practice What We Teach?” Waste Age June 1993: 131-
142.

Gaski, Stephen K. “Operating a MRF on Campus.” Biocycle August 1993: 57-59.

Gay, Alan E., et al. “Cost-Effective Solid Waste Characterization Methodology.” Journal
of Environmental Engineering July/August 1993: 631-644.

Geiser, Ken and Frances H. Irwin, eds. Rethinking the Materials We Use: A New Focus
for Pollution Policy. Washington, DC: World Wildlife Fund, 1993.

Giannone, Ann. “University Waste Management Programs,” (unpublished) 1995.

Hegberg, Bruce, et al. University and College Solid Waste Reduction and Rmcling.
ChicagozUniversity of Illinois Center for Solid Waste Management and Research,

1992.

 

Hoss, Timothy. E-mail to the author. February 1996.

93

Huisingh, Donald and Vicki Bailey, eds. Making Pollution Prevention Pay: Ecology with
Econpmy as a Poligy. New York: Pergamon Press, 1982.

Jablonski, Ed. “Cutting the Cost of Recycling.” American Cig and Coung April
1995:46.

Makower, Joel. Beyond the Bottom LinezPutting Social Responsibility to Work for Your
Business and the World. New York: Simon And Schuster, 1994.

Martin, John H. “The Development and Testing of a Standardized Protocol for Analyzing
the Waste Stream.” Diss. West Virginia University, 1993.

Michigan Department of Natural Resources. Solid Waste Stream Assessment Guidebook.
June 1986.

Michigan State University Recycling Coalition. “Call for Recycling,” (unpublished) 1995.

Musa, Elias and GE. Ho. “Optimum Sample Size in Refuse Analysis.” Journal of
Environmental Engineering 107(6): 1247-1260.

Oldenburg, Kirsten U. and Joel Hirschhom. “A New Strategy to Avoid Pollution.”
Environment March 1987: 17-45.

Pasterz, Peter. Personal Interview. February, 1996.

Platt, Brenda, et al. Beyond Forty Percent: Record-Setting Recycling and Cpmposting
Progzams. Washington, DC: Island Press, 1991.

Powell, David L. “A Solid Waste Reduction Plan for East Tennesee State University.”
Thesis. East Tennesee State University, 1993.

Rappaport, Ann. Development aid Transfer of Pollution Prevention Technology;
Westport, CT: Quorum Books, 1993.

Rathje, William L. and Cullen Murphy. Rubbish! New York: HarperCollins, 1992.
Rathje, William L. “Rubbish!” The Atlantic Monthfi December 1989: 99-110.

“Reducing Waste Saves Money.” Resource RecyclingSystems News_li_n_e,
Spring/Summer 1991: 1-3.

Riley, Barbara. “Using Waste Characterization in Program Design”, in The Biocycle
Guide to Maximum Recycling, Emmaus, PA: JG Press, Inc., 1993.

94

Saunders, Tedd and Loretta McGovern. The Bottom Line of Green is Black: Strategies

for Creating Profitable and Environementally Sound Businesses. New York:
HarperCollins, 1993.

Schoonover, Jennifer, “Composting Joins Recycling on College Campus.” Biogycle
February 1994: 56-57.

Small, William E. Third Pollution: The National Problem of Solid Waste Disposal._
New York: Praeger Publishers, 1971.

 

Stead, W. Edward and Jean G. Stead. Management for a Small Planet. London: Sage
Publications, 1992.

Steuteville, Robert, “The State of Garbage, Part One.” Biocycle April 1995: 54-63.
--- . “What’s New in the Waste Stream.” Biocycle October 1992: 40-41.

Sullivan, John L. and Steven B. Young. “Life Cycle Analysis/Assessment.” Advanced
Materials and Processes February 1995: 37-40.

Tchobanoglous, George, et al. Integrated Solid Waste Management: Engineering
Principles and Management Issues. New York: McGraw-Hill, 1993.

--- . Solid WasteszEngineering and ManagementEIssues. New York: McGraw-I-Iill, 1977.

US. Department of Commerce, Office of Technology Assessment. Facing America’s
I_rash: What Next for Municipal Solid Waste? Washington, DC: GPO, October
1989.

US. Environmental Protection Agency. Characterization of Municipal Solid Waste in the
United States: 1994 Update. Washington, DC: GPO, November 1994.

--- . --- . Charactengation of Municipal Solid Wage in the United States: 1990 Updat_e._
Washington, DC: GPO, June 1990.

--- . ~-- . Decision-Mg“ Guide to Solid Waste Management. Washington, DC:
GPO, November 1989.

--- . --- . “Guides to Pollution Prevention: Research and Educational Institutions.”
Washington, DC: GPO, June 1990.

--- . --- . RCRA Orientaaion Manual. 1990 edition. Washington, DC: GPO, 1990.

--- . --- . Solid Waste and Emergency Response. The Solid Waite Dilemma: An Agenda
for Action. Washington, DC: GPO, February 1989.

95

--- . --- . Office of Solid Waste. Test Methods for Evaluating Solid Waste vol. II,
Washington, DC: GPO, 1986.

 

Wagner, Joan. “The Three “R’s” of Solid Waste and the Population Factor for a
Sustainable Planet.” The American Biology Teacher February 1995: 88-91.

Waste Management, Inc. and Piper & Marbury. Mste Reduction: Policy and Practice,
New York: Executive Enterprises Publications Co., Inc., 1990.

“Waste Reduction and Recycling at University of Michigan Housing Facilities.” (not
published) final report by Resource Recycling Systems for the Solid Waste Task
Force, Ann Arbor, MI, 1988.

Watson, Tom. “Recycling Goes to College.” Resource Recycling April 1990: 76- 81.

 

Williams, Marcia E. “Integrated Solid Waste Management,” in Integrated Solid Waata
Management: Options for Legislative Action, ed. by Dr. Frank Kreith.
Schnectedy, NY: Genium Publishing Corporation, 1989.

 

World Wildlife Fund. Gettingat the Source: Strategies for Reduciag Municipal Solid
Waste. Baltimore, MD: WWF Publications, 1991.

York, Kimberly. “Solid Waste Management: A Study of Options and An Initial
Recyclable Materials Waste Stream Analysis for Waco, Texas.” Thesis. Baylor
University, 1990.

Young, John E. and Aaron Sachs. The Next Efficiency Revolution: Creating a Sustainable
Materials Economy. Washington, DC: Worldwatch Institute, 1994.

“71111111111111111111113