Air Classification of
Solid Wastes ^
J?
-------
-------
Air Classification of
Solid Wastes
Performance of Experimental Units and Potential Applications for Solid Waste Reclamation
This publication (SW-30c) was written for the Federal solid waste management program
by R.A. BOETTCHER
Stanford Research Institute, Irvine, California
under Contract No. Pti 86-68-157
Environmental protection Agency
LiLroA7. -'•• J""f- v
1 NC.-I.J! \>^3r Drive
Chicago, Illinois 60606
U.S. ENVIRONMENTAL PROTECTION AGENCY
1972
-------
ENVIRONMENTAL PROTECTION AGENCY
An environmental protection publication
in the solid waste management series (SW-30c)
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, B.C. 20402 - Price 75 cents
-------
Foreword
The objective of this laboratory-scale study by the Stanford Research Institute was
to determine the technical feasibility of using air classification to process or treat
selected types of nonhomogeneous, dry solid wastes. Design requirements for pro-
cessing solid wastes with a full-scale commercial unit are estimated and the required
supplemental equipment (shredders, screens, dryers) are identified. Limitations and
advantages of the method are listed.
The air classification method was shown to be applicable to salvage of paper,
recovery of nonferrous metal from shredded automobile body waste, and processing of
compost.
This method was shown to be technically feasible for processing wastepaper.
Furthermore, by analogy from calculations on a model compost-processing operation,
it is thought that the method should also be economically feasible. Limitations for
wastepaper processing are not inherent in the method itself but result rather from the
size of the unit and the character of the shredded feed material.
Nonferrous trash from automobile body processing can be separated without special
feed preparation other than screening of oversize materials. The method permits
separation of materials with only slightly differing densities whose other properties,
such as size, are identical.
Theoretically, in the processing of shredded automobile body materials, air classifi-
cation can improve cleanliness and reduce nonferrous contamination of the steel scrap,
reduce quantities (and cost) of wastes requiring landfill disposal, and increase the
capacity of nonferrous-recovery processing facilities.
It was concluded from bench-scale experiments with the laboratory unit that air
classification is technically feasible for processing semifibrous solid waste materials such
as compost. Compost can be cleaned of glass and other contaminants that reduce its
marketability. Yields of more than 50 percent combined bulk and horticultural-grade
material can be produced. Recovery and reuse of glass could stimulate a greater use of
nonreturnable glass bottles by the beverage industry.
As a unit operation, the advantages of air classification include the following: dry
processing capability; sharp, clean separation capability; high-capacity throughout; low
power requirement: low operating manpower requirement; dust-free operation.
We should like to acknowledge the work of the project officer, Dr. Boyd T. Riley, Jr.,
in coordinating this study.
in
-------
-------
Contents
INTRODUCTION !
Objective and Scope 1
Method of Approach 1
Information Sources 2
Acknowledgments 2
SUMMARY AND CONCLUSIONS 5
LABORATORY-SCALE UNIT EXPERIMENTS
Selection of Solid Wastes for Experimental Program 9
Criteria for Selection 9
Outside Agencies' Consensus 9
Materials Selected and Test Program 11
Sources of Sample Material 12
The Laboratory-Scale Air Classification Unit 13
Background 13
Physical and Operating Characteristics of Laboratory-Scale
Air Classification Unit 14
Modification of Laboratory-Scale Unit to Facilitate
Testing of Selected Wastes 16
Performance of the Laboratory-Scale Unit on Selected Wastes 17
Operating Procedure 17
Performance of Laboratory-Scale Unit on Selected Wastes 18
Compost from Municipal Refuse 19
Automobile Body Trash 19
Municipal Refuse 27
Evaluation of Laboratory-Scale Unit's Performance in Separating
Solid Wastes 31
PILOT-UNIT EXPERIMENTS
Need for Pilot-Scale Experiments for Recovery of Municipal Refuse 35
Pilot-Scale Air Classification Unit 35
Procurement of Shredded Samples and Their Characteristics 35
Procurement of Samples 35
Los Angeles Sample 35
Cincinnati Sample 37
Characteristics of Samples 37
Performance of the Pilot-Scale Unit on High-Paper-Content Feeds 40
Operating Procedure '. 40
Discussion of Results 44
Conclusions and Recommendations Regarding Air Classification of Municipal
and Commercial Waste 48
Effect of Shredding 48
Scale-up Considerations 49
-------
POSSIBLE ROLE OF AIR CLASSIFICATION IN PROCESSING SOLID WASTES 51
Supplemental Equipment Needed for Mechanical Processes of Solid Waste Reclamation 51
Shredders 51
Screens and Driers 52
Scale-up Procedure 53
Estimated Full-Scale Performance of Air Classifier Unit in Processing Solid Waste 56
Compost from Municipal Refuse 56
Automobile Body Trash 56
Municipal Refuse 60
Related Applications 62
REFERENCES 63
APPENDIX A INTERVIEW RECORD FOR SELECTION OF WASTES TO BE
PROCESSED IN PHASE I 67
APPENDIX B REPORT OF SAN DIEGO UTILITIES DEPARTMENT ON
RETORTING OF AUTOMOBILE BODY MATERIAL 71
VI
-------
TABLES
1. Solid Wastes Suggested in Interview Responses for Air Classification 10
2. Typical Fluidizing Velocities and Horsepower Requirements for Air
Classification on Materials Processed in Previous Experimental Work
at SRI 14
3. Fractions Resulting from Screening. Residue after Magnetic Separation
from Automobile Body Shredding 22
4. Sub fractions Resulting from Air Classification. Screen Fraction 3
(Coarse)-Automobile Body Trash 22
5. Subfractions Resulting from Air Classification. Screen Fraction 4
(Fine)-AutomobOe Body Trash 22
6. Physical Characteristics of Subfraction Resulting from Air Classification.
Screen Fraction 3 (Coarse)—Automobile Body Trash 23
7. Physical Characteristics of Subfraction Resulting from Air Classification.
Screen Fraction 4 (Fine)—Automobile Body Trash 23
8. Semiquantitative Spectrographic Analysis of Screened Fractions of Residue
After Magnetic Separation from Automobile Body Shredding 27
9. Semiquantitative Spectrographic Analysis of Air-Classified Subfractions
After Magnetic Separation from Automobile Body Shredding. Screened Fraction 3
(Coarse)-Automobile Borlv Trash 28
10. Semiquantitative Spectrographic Analysis of Air-Classified Subfractions
of Residue After Magnetic Separation from Automobile Body Shredding. Screened
Fraction 4 (Fine)-Automobile Body Trash 29
11. Characteristics of Municipal Refuse Samples as Received for Experimentation
at Stanford Research Institute 30
12. Expected Ranges in Composition of Mixed Municipal Refuse 31
13. Heating Values of Various Types of Municipal Refuse 32
14. Fluidizing Velocities for Selected Pure Components of Refuse
Mixtures in a Straight Pipe and the Zigzag Column 33
15. Shredding Information and Source of Municipal Refuse Samples Processed in
the Pilot Air Classification Unit 37
16. Description of Shredded Municipal Solid Waste, Johnson City, Tennessee,
PHS—TV A Composting Project, Shredded by Gruendler Swing Hammermill,
Model 48-4 38
17. Description of Shredded Municipal Solid Waste, Johnson City, Tennessee,
PHS-TVA Composting Project, Shredded by Dorr-Oliver Rasp , . 39
18. Description of Shredded Municipal Solid Waste, Los Angeles, California,
Scholl Canyon Landfill, Shredded by Williams Rigid-Arm Paper Shredder 40
19. Description of Shredded Municipal Solid Waste, Cincinnati, Ohio,
BSWM Laboratory, Shredded by Williams Hammermill 41
20. Description of Shredded Municipal Solid Waste, Albuquerque, New Mexico,
Eidal International Corp., Shredded by Eidal Model 400, Coarse Grind 42
21. Air Classifications Conducted on Johnson City (Hammermill) Refuse 44
22. Air Classifications Conducted on Johnson City (Rasp) Refuse 45
23. Air Classifications Conducted on Los Angeles Refuse 45
24. Air Classifications Conducted on Cincinnati Refuse 46
25. Air Classifications Conducted on Albuquerque Refuse 47
26. Summary of Column-Loading Data from Tables 21 through 25 50
27. Summary of Data Obtained on Commercial Shredding Equipment for
Municipal Wastes 51
28. List of Factors and Relationships of Importance in Air Classification 53
29. Scale-Up Factors Based on Data Obtained with Laboratory Air Classifier 55
30. Prices (per Ton) of Wastepaper Stock in Major U.S. Markets 62
B-l. Run No. 156 (Materials Retained on 1-in. Screen) 71
B-2. Run No. 157 (Materials Passing 1-in. Screen) 73
vii
-------
FIGURES
Figure 1 Schematic flow diagram of air classification process
for wastepaper recovery from municipal refuse 6
Figure 2 Scientific Separators zigzag air classifier evaluated
by SRI . . . .' 15
Figure 3 Flow pattern of granular material in the Scientific
Separators zigzag air classifier evaluated by SRI 16
Figure 4 Hypothetical cost relationships for shredding and
air classification 18
Figure 5 Compost-processing diagram and photographs of the
original shredded compost and the fractions into
which it was separated by screening and air
classification 20
Figure 6 Automobile body trash—screened fractions 24
Figure 7 Automobile body trash—air classified subfractions 25
Figure 8 Scientific Separators pilot air classification unit 36
Figure 9 Mechanical analysis diagrams representative of various
methods of shredding municipal solid waste 43
Figure 10 Conceptual process flow diagram: air classification
system for production of horticultural-grade
compost from aged stockpile material 57
Figure 11 Process flow block diagram: conventional automobile
body shredding system 58
Figure 12 Process flow block diagram: modified system for auto-
mobile body shredding employing air classification
and including nonferrous metal recovery 59
Figure 13 Process flow block diagram: proposed wet system
of nonferrous metal recovery from automobile
body trash 61
Figure B-l Material retained on 1-in. screen, Run 156 72
via
-------
Air Classification of
Solid Wastes
Since the enactment of the Solid Waste Disposal
Act (PL 89-272) on October 20, 1965, the spotlight
of publicity has been thrown on the Nation's solid
waste management problem. Under the Act, research
and development programs have been initiated to in-
vestigate, study, and develop new and improved
methods of solid waste management.
Those already in the field know that populations
whose living habits are undergoing the changes that
accompany increased affluence are generating in-
creased quantities of waste; that air pollution legisla-
tion that outlaws open burning has brought about
rapid increases in the quantities of solid waste; and
that massive programs of urban renewal, central-city
freeway construction, and replacement of older build-
ings with modern structures have increased demoli-
tion wastes.
Also to be considered, along with the quantitative
and qualitative changes in the material being col-
lected, are the changes in collection and salvage
methods being brought about by continually increas-
ing labor costs. Combined collections have largely
replaced segregated collections and have thereby
reduced collection costs. On-site compaction of com-
mercial wastes containing large quantities of paper
and cardboard is a rapidly growing extension of this
trend. In only a few specialized cases is it now eco-
nomical to salvage rags, newspapers, and corrugated
cardboard boxes by handpicking them from com-
bined collections of municipal solid wastes.
There is a growing realization of the desirability of
reducing the quantity of wastes requiring disposal by
reclaiming a portion of the flow for reuse. There is
also, however, the realization that only processes that
are mechanized and that operate continuously can be
seriously considered. Acceptable processes require a
minimum input of labor per ton of refuse to separate
material into those portions that are reusable and
those that must be discarded. Air classification ap-
pears to be one such process, judged from work done
before this study with Scientific Separators* air clas-
sification unit in handling agricultural materials and
simulated domestic and demolition wastes.
OBJECTIVE AND SCOPE
The objective of the first phase of research re-
ported herein was to make a preliminary determina-
tion of the technical feasibility of using air classifi-
*Mention of commercial products does not imply endorse-
ment by the U.S. Government.
cation to process selected types of nonhomogeneous,
dry solid wastes. A small laboratory-size Scientific
Separators air classifier was available. The feed
mechanism and throat size of the classifier accepted
wastes shredded to a particle size of 1 to 1-1/2 in. A
blower provided air velocities adequate for separation
from refuse of materials with the densities of wood,
glass, or steel. A number of solid wastes were
prepared and processed under different operating
conditions. Methods of waste preparation and pro-
cessing were sought that could be used by refuse
contractors or industrial firms to recover a stream of
salvaged material having commercial value.
First-phase results indicated the desirability of fur-
ther experiments to investigate techniques for paper
stock recovery from combined collections of munici-
pal and commercial wastes. For this purpose a larger
air classification unit was built and operated in the
second phase of the program to do the following: (1)
determine the degree of separation possible in a zig-
zag air classification column capable of processing
paper-containing refuse material 4 to 6 in. in size and
(2) evaluate the influence of various commercially
available shredding methods on the degree of separa-
tion obtainable. The study was limited to technical
aspects of air classification. It did not include com-
parative evaluations of other separation processes,
process economics, market studies, or the economics
of reclaimed products that might be salvaged by pro-
cessing the waste streams.
This final report of the research investigation is
intended to acquaint interested engineers with char-
acteristics of the air classification device and to
suggest potential applications for separating mixed
solid wastes. It also contains speculations on the role
that this equipment might have in the management of
solid waste.
METHOD OF APPROACH
The wastes to be processed in the first-phase work
were selected by the Institute's research team in ac-
cordance with criteria established by the office's pro-
ject officer. Wastes selected for experimentation were
expected to be successfully air classified. Moreover,
selection was made after interviews with representa-
tives of agencies who had extensive solid waste ex-
perience. During these interviews, opinions were
solicited on the kinds of separations that would be
desirable from the standpoint of commercial exploita-
tion. An experimental program was then developed
that established the objectives and outlined tentative
classification procedures for each of the wastes
1
-------
AIR CLASSIFICATION OF SOLID WASTES
selected. The experiments performed illustrated both
the limitations and the advantages of the air classifica-
tion separating process.
For the second phase, five samples of municipal
solid waste were obtained and shredded in different
types of commercial shredding equipment. Two of
these shredded samples were preponderantly of the
largest particle size (4- to 6-in. maximum dimension)
that could be processed in the pilot unit, and three
samples were generally of considerably smaller par-
ticle size. All the shredded material was air classified.
Only limited experimentation was possible for paper
recovery in these initial, larger scale classifier runs;
classifications made were intended primarily to
establish qualitatively the type of separations that
could be effected on a commercial scale and to
indicate characteristics of the shredded material that
would be desirable in specifying the performance of
shredders for reclamation, by air classification, of
paper and other secondary raw materials in municipal
solid waste.
In most cases, the only way to characterize the
wastes and the separated products was to describe
them visually. For the most part this proved
adequate; it was supplemented where possible by
determinations of bulk density for separated fractions
and moisture content* and bulk density of the
material as received. Photographs were taken of a
number of the samples and of the separated streams.
With vehicular scrap samples, i.e., the nonferrous
metal and trash stream after magnetic separation of
shredded automobile body material, visual evaluation
of the effectiveness of the separation was not
possible. Hence, separations were made over a range
of throat velocities, and the separated materials were
analyzed for density, oil and moisture content, per-
centages of combustible organic and magnetic
material, and composition as indicated by spectro-
graphic assay. For certain of the recovered paper
samples, analyses made by the St. Regis Paper Com-
pany's Technical Center are reported.
The performance of Scientific Separators labora-
tory- and pilot-scale air classification units were evalu-
ated by SRI in terms of the technical feasibility of
making the various separations, together with their
advantages and limitations. With appropriate scale-up
factors, the laboratory results are reported in terms of
full-scale process power requirements, limitation on
density, optimum size reduction, and problems that
can be anticipated with various types of granular and
fibrous materials. To illustrate the possible role of air
classification in processing solid wastes, results ob-
tained are presented in the form of preliminary pro-
cess flow sheets showing the shredding, drying,
screening, air classification, and other unit operations
required for reclamation processing of solid waste
materials.
INFORMATION SOURCES
In addition to Boyd T. Riley, Jr., Chief, Waste
Handling and Processing Branch, Solid Waste Manage-
ment Office, U. S. Environmental Protection Agency,
who was designated project officer for this research,
21 government agencies and other organizations were
contacted.t These were the regional representatives:
Region IX, Solid Waste Management Office; Los
Angeles County Sanitation Districts; Los Angeles
Bureau of Sanitation; the municipal refuse retorting
project, being conducted for the Solid Waste Manage-
ment Office by the San Diego Utilities Department
and San Diego State University; the California solid
waste management project, being conducted for the
Solid Waste Management Office by the Bureau of
Vector Control, California State Department of
Public Health, Berkeley, California; the California
integrated systems study in solid waste management,
being conducted for the Solid Waste Management
Office by the California Department of Public Health,
Fresno, California; Sanitary Engineering Research
Center, University of California, Richmond Field
Station, Richmond, California; Utilities Division,
Department of Public Works, Sacramento County,
Sacramento, California; Governmental Refuse
Collection and Disposal Association, Pasadena,
California; Office of the Los Angeles County Engi-
neer, Los Angeles, California; Departments of
engineering, University of California, Los Angeles,
and the University of Southern California; Chemical
Engineering Systems Research, Division of Wood
Fiber Products, Forest Products Laboratory, USDA,
Madison, Wisconsin; Aerojet-General Corporation and
Ralph Stone Engineers (engaged in contract research
and proprietary developments related to solid
wastes); National Metals Company; Clean Steel, Inc.;
Pan American Resources; Universal By-Products, Inc.;
St. Regis Paper Company, Technical Center; United
Paper Stock Co.; U.S. Gypsum Company; Metro-
politan Waste Conversion Corp.; Lone Star Organics,
Inc.
ACKNOWLEDGMENTS
The Institute team wishes to express its apprecia-
tion to the Clean Steel Division of National Metals,
Lone Star Organics, Universal By-Products Company,
Eidel International Corporation, Alpha Beta -Acme
Markets, and the PHS TVA Johnson City,
Tennessee, composting project for cooperation in
providing and for shredding solid waste samples.
These samples were in addition to the samples of
municipal waste provided by the Cincinnati labora-
*Moisture contents in this report are based on the dry
weight of the sample.
t Individuals interviewed are listed in Appendix A.
-------
INTRODUCTION 3
tory of the BSWM*The fact that all samples sub- manufacture was provided by Harry Armstrong,
mitted had already received primary shredding works manager, and A. Chaves of U.S. Gypsum
simplified laboratory operations considerably. Company, by A.C. Veverka of St. Regis Paper
Useful industry information and an understanding Company, and by Charles Pabigan of United Paper
of wastepaper salvage operations were provided by Stock Company. The Institute team is also grateful to
Richard T. Stevens, president of Universal By- Donald A. Hoffman, research coordinator of the San
Products, and by Mr. Stevens' father, Robert W. Diego Utilities Department, and his staff for pyrolysis
Stevens, an independent paper mill consultant. Infor- runs on samples of automobile body material.
mation on the utilization of secondary fibers in paper
*BSWM (Bureau of Solid Waste Management) is the
former designation of the Federal solid waste management
program
-------
-------
Summary and Conclusion
Preliminary research on air classification, as inves-
tigated with laboratory- and pilot-scale zigzag air clas-
sification units, has indicated the feasibility of using
this method of separation to process a number of
types of solid waste mixtures. Satisfactory separations
require the following:
Suitable feed preparation. An air classification unit
alone does not usually constitute a complete solid-
waste-processing system. Operations such as
shredding, drying, and screening must often be com-
bined with the air classifier to achieve optimum
separation.
Particles with maximum dimensions no greater
than three-quarters the least dimension of the column
throat.
Particles that flow in a granular fashion when
fluidized by air. For fibrous materials, this precludes
a shredding method that achieves particle size reduc-
tion but produces detrimental aerodynamic char-
acteristics and an undesirable agglomeration of the
components of the refuse.
The need for feed preparation was revealed by
work with all three general types of waste processed
experimentally. These wastes were municipal refuse;
aged stockpile compost; and nonferrous trash from
automobile body processing.
With municipal refuse, shredding was essential
before air classification. Shredding is not practiced at
the present time, except to a limited extent in re-
covery of usable corrugated and high-grade mixed
papers from commercial and industrial collections.
Shredding is, however, practiced for municipal wastes
being composted and for developmental transfer/
baling, retorting, incineration, and landfill operations.
Thus, it would not represent an additional processing
step in these methods of refuse handling.
Compost that has been aged in outdoor stockpiles
requires additional light shredding to break up lumps
before it can be air classified to remove impurities,
even though it has already been shredded in normal
processing. Needed feed preparation for air classifi-
cation of compost also includes drying and screening,
because of the presence of large amounts of fine
materials that behave aerodynamically the same as
low-bulk-density cellulosic components. Similar re-
processing also may be needed for municipal refuse.
Nonferrous trash from automobile body pro-
cessing could be separated effectively without special
feed preparation in an air classifier designed to handle
quantities available from typical automobile body-
fragmentizing operations. In certain applications, it
probably would be desirable to remove oversize
material by screening to reduce the required size and
power requirements of the air classification device. In
the experiments performed as part of this research,
screening of the automobile body trash was necessary
because of the small throat size of the laboratory
column.
Limitations observed in the operation of the
laboratory zigzag classification unit on municipal
refuse are those related to the size of the unit and the
characteristics of the materials fed—resulting from the
shredding methods used to produce an acceptably
small particle size—rather than from limitations of air
classification itself as a method of separation. First,
these limitations could be one, as was demonstrated
by successful operation of the pilot-scale air classifi-
cation unit. Thus, separating usable wastepaper from
municipal refuse appears to be technically feasible in
an integrated system employing air classification. A
schematic flow diagram of an air classification process
is shown (Figure 1) and an alternate system is
described on p. 48. The high capacities, low equip-
ment cost, and low power requirements of such a
system, as illustrated by calculations for a compost-
processing operation, should make salvage of
municipal refuse by air classification economically
feasible as well. Large markets exist for secondary
fiber of acceptable quality. Consequently, a signifi-
cant contribution can be made to conservation of the
nation's forest resources by increased recycle of
air-classified wastepaper.
Separation of heavy materials from dry, shredded
municipal and commercial refuse can be accom-
plished easily. It is possible to insert an air classfica-
tion column into a vertical run of pneumatic con-
veyor piping handling the output of a refuse shredder
and to remove metal, glass, rocks, rubber, and wood
from the shredded refuse. The cost is estimated to be
less than 10
-------
AIR CLASSIFICATION OF SOLID WASTES
§1
"is
«I
III
6,-g
E-S S
•sir
u .op 3
s
a, b -a
_ J3 c
c -a o
o S u
.2 S
»-i
* ^ ^
rt.!^
-------
SUMMARY AND CONCLUSIONS
tities (and cost) of waste requiring landfill disposal,
and (3) increased capacity of nonferrous recovery
processing facilities.*
Compost can be cleaned of glass and other con-
taminants that reduce its marketability. Yields of
more than 50 percent combined bulk and horti-
cultural-grade material can be produced by an air clas-
sification process developed by laboratory column
operation. In a 30-ton-per-hour plant, the cost is
estimated to be 30* per ton of material processed
(24-hr operation), including amortization.
Other findings of importance include the follow-
ing:
The laboratory- and pilot-scale air classifiers are
useful tools for empirical design.
For a particular type of refuse and a particular
reclamation objective, final system design requires
that separations be optimized by a more extensive
series of experiments than was performed herein.
Empirical determination of design criteria for full-
scale air classification systems and the development
of economic recovery processes, especially for mu-
nicipal refuse, require a pilot-plant research facility
that might best be combined with a demonstration
facility in which municipal solid waste is being
shredded routinely for purposes other than salvage.
The characteristics of material shredded by com-
mercially available shredders are compatible with
sorting by air classification. For predicting the char-
acteristics of individual components of municipal
solid waste after shredding—as related to rotating
speed, feed rate, and other variables of shredder
operation—a cooperative test program with shredder
manufacturers would be desirable.
Commercial interest might be stimulated by the
construction of a demonstration plant for salvage of
paper or other secondary raw materials from refuse.
Nonferrous metal recovery from automobile body
trash should be particularly attractive commercially.
Recovery and reuse of glass could stimulate a
greater use of nonreturnable glass bottles by the
beverage industry.
As a unit operation, the advantages of air classifi-
cation include the following: dry processing capa-
bility; sharp, clean separation capability; high-
capacity throughput; low power requirement; low
operating manpower requirement; dust-free
operation.
Air classification is subject to certain limitations,
the most significant of which are the following:
Feeder and column throat size impose particle-size
limitations for low-capacity systems.
Multiple-column or repetitive operation is required
for more than two-component separations.
Oversize, semifibrous materials require shredding
before classification.
These processing possibilities are illustrated in the body
of this report by process flow block diagrams.
-------
-------
Laboratory-scale Unit Experiments
Selection of Solid Wastes for Experimental Program
Criteria for Selection. In the selection of solid
wastes for the experimental program, the Institute
team was aided by a number of State and local gov-
ernmental agencies, by representatives of firms
engaged in solid waste collection and disposal, and by
solid waste program contractors and grantees engaged
in research and demonstration activities. A series of
personal interviews was employed, supplemented by
telephone contact as needed. During the interviews,
experimental equipment to be used in the separation
process was described and the objectives of the pro-
gram were outlined. Each person interviewed was re-
quested to suggest a number of types of solid wastes
that might be classified experimentally and to
indicate the separations that would be of greatest
potential value. In all cases, the criteria for selection
of materials were the following:
Wastes that by their nature or quantity are
creating a problem at present or may be expected to
become a problem
Wastes that could be separated by air classification
into more easily disposable or reclaimable fractions
Combined wastes whose composition varies at
different times but from which a uniform fraction
might be separated
Wastes possessing reclamation potential—currently
not realizable because of high processing costs of
existing methods and equipment—that might be
feasibly realized by air classification
Outside Agencies' Consensus. Names of individuals
with whom discussions were held are given in the
chronological interview record included in Appendix
A. In conducting these discussions, an attempt was
made to obtain suggestions on various types of wastes
to be processed. Because of the large volume of
wastes from municipal domestic and municipal com-
mercial sources, the majority of interviews centered
on these types of refuse. Industrial wastes were also
considered important because of the opportunities
for by-product recovery or other forms of reclama-
tion. Wastes from the processing of agricultural
materials and municipal demolition wastes were also
discussed, as well as a number of specialized oppor-
tunities, such as cleaning of aged compost from plants
processing municipal refuse, nonferrous metal re-
covery from the trash stream after magnetic separa-
tion of crushed automobile bodies, cleanup of the
char product from refuse - retorting operations, and
reclamation of usable pigment from paint sludge. In
each case, an opinion was requested about the
recovered component that would have greatest value
and the use or uses that could be foreseen if recovery
could be effected. Information was also sought on the
type of contaminants that might be permitted in the
reclaimed product as an indication of the selectivity
of classification that would be required.
Interview responses suggested 10 types of solid
waste from 15 sources for air classification (Table 1).
Items 1, 2, and 3, important as types of wastes from
which usable grades of reclaimed paper might be
obtained, are listed in order of probable ease of clas-
sification. Relatively clean collections of waste paper
from office buildings and large commercial establish-
ments, mainly department stores, are not included,
because such collections are now marketed as mixed
paper.
The preferred grade of mixed paper for secondary
fiber repulping is a material that is shredded before it
is baled. To protect shredding equipment, the salvage
operator removes heavier metal from collected
material before shredding and thereby reduces the
amount of unpulpable trash requiring removal from a
user's pulping equipment and possible damage to this
equipment.
Where collection routes can be so arranged that
only municipal domestic refuse from apartments
predominates, it was the consensus of respondents
that a fairly good grade of wastepaper might be
reclaimed, because the refuse would normally not
contain garbage or garden trimmings. Virtually all the
newer apartments are equipped with garbage grinders,
and landscaping is done by gardeners who are
required to dispose of their own clippings. Municipal
domestic refuse in which collections from single-
family residences predominate might contain a
salvable wastepaper component, but it was expected
that this would be of a lower quality than that from
either shopping centers or apartments because of a
greater number of contaminants. Refuse that was
segregated before collection was, of course, not
considered for air classification.
A substantial amount of corrugated carton mate-
rial is collected from markets. It has been the practice
of paper dealers to collect this material without
charge when it is segregated. Recent developments in
the use of on-site compaction equipment have re-
duced the cost of collecting this material, and this
saving permits the collector to pay a small amount for
it and to supply his customer with the compaction
equipment.
-------
10
AIR CLASSIFICATION OF SOLID WASTES
TABLE 1
SOLID WASTES SUGGESTED IN INTERVIEW RESPONSES FOR AIR CLASSIFICATION
Item
number
1
2
3
4
Type of
solid waste
Municipal,
commercial
Municipal,
domestic
Municipal,
domestic
Municipal,
domestic
Source
Small commercial
establishments not
segregating for
paper salvage
Apartments only
Single-family
residences only
Combined collection,
unsegregated routes
Air classification
product
Wastepaper
Wastepaper
Wastepaper
Beneficiated retort
feed stock, by
exclusion of glass
and metals
Possible reuse
Paper or board manufacture
Paper or board manufacture
Paper or board manufacture
Retort feed
Municipal
demolition
waste
Wood frame houses
Aged compost Composting plants
Retort char
Industrial
wood waste
Retorting of domestic
waste
Retorting of
industrial wastes
(Pan-American
Resources)
Debris from small
cabinet shops
Automobile body Nonferrous metal and
reclamation trash stream after
magnetic separation
Beneficiated feed
stock for
composting process,
by exclusion of
material either not
converted or
converted slowly
Beneficiated feed
stock for
fermentation,
principally cellulosic
material
Broken concrete
Wood waste
Grades of compost;
salvage material-
glass, metal, etc.;
uncomposted waste
Clean char, without
glass or metal
contamination
Clean char, without
glass or metal
contaminants
Wood waste separated
into hardwood and
softwood fractions
Metallics only
Glass or rubber
Composting: alone; with
sewage sludge, or with
animal manures
Paper salvage
Fermentation for synthetic
proteins, alcohol, or other
carbohydrate derivatives
Aggregate for new concrete
Charcoal, paper pulp, or
particle board
Higher value material bagged
for horticultural use;
coarse material for
planting freeway slopes
(mulch)
Water or sewage treatment;
low-ash fuel; carbon black
for rubber compounding:
activated carbon
Charcoal, paper, pulp, or
particle board
Feed to various systems of
nonferrous separation
Secondary raw material
(Continued next page)
-------
LABORATORY-SCALE UNIT EXPERIMENTS
11
TABLE 1 (Concluded)
SOLID WASTES SUGGESTED IN INTERVIEW RESPONSES FOR AIR CLASSIFICATION
Item
number
Type of
solid waste
Source
An classification
product
Possible reuse
Separation of metallics Secondary raw material
(cheaper than
flotation or sweating)
10
11
12
13
14
15
Airplane
fuselage
reclamation
Miscellaneous
industrial
wastes
Agricultural
processing
Agricultural
processing
Agricultural
processing
Agricultural
processing
Shredder output
Ford assembly plant
at Milpitas
Nutshells
Raisin processing;
trash
Cotton ginning trash
Oilseed processing
trash
Metals other than
aluminum
Miscellaneous salvage
Broken nutmeats;
clean shells
Good raisins in trash
stream
Cotton lint
Oilseed meat
Reclaimed metal
Reclaimed metal, paint
pigment, charcoal
Activated charcoal
Raisins
Reclaimed cotton lint
Oil and meal
There are at least five possible uses for products
separated from municipal domestic refuse (Table 1).
Since paper now accounts for more than 50 percent
of domestic collections on a dry-weight basis, paper
salvage for reuse of its cellulose content would be
preferable to any method of utilization that would
require degradation of the cellulose. Reuse in paper
or board manufacture would accomplish this resource
conservation objective. If refuse is converted to
charcoal, with or without recovery of gaseous pro-
ducts and pyroligneous acids, it is desirable to remove
glass and metals from the retort feed. This reduces
process heat requirements and the ash content of the
char. For composting and fermentation, all nonbio-
degradable materials must be removed, the most
difficult of which are plastic wrapping materials and
styrofoam packing.
Composting of domestic refuse collections, as it
has been practiced to date, usually includes hand-
picking to preclean the digester feed prior to shred-
ding, primarily to protect the shredding equipment
but also to recover wastepaper. Besides being eco-
nomically marginal as a salvage operation, hand-
picking leaves a great amount of glass and other
materials in the feed to the composting digester; these
elements reduce the quality and salability of the
compost product. It would be highly desirable to
remove all materials that could be salvaged or could
not be composted from the digester feed as easily as
tin cans are now removed by magnetic separation. It
would also be desirable to clean existing stockpiles of
compost to improve salability of the product. The
removal of glass, metal, and material that either is
uncomposted or cannot be composted appears
possible by air classification.
Among the industrial possibilities, greatest promise
appears to exist in the recovery of nonferrous metals
from automobile body-shredding operations. Pro-
cessing of agricultural materials—while desirable
because of the magnitude of solid waste problems
involved—appears to offer specialized economic
opportunities.
Materials Selected and Test Program. The 15
different solid wastes listed (Table 1) were suggested
by interview respondents as meeting the established
selection criteria. Although it was originally intended
to air classify only five samples of waste, it was
considered feasible—owing to the probable similarity
of a number of the samples of municipal domestic
waste that were suggested and the similarity of
products to be recovered—to process six separate
samples. The six samples selected as having the
greatest probability for successful air classification
were the following:
Sample 1. Single-family domestic waste, Los
Angeles, California. This sample was to be represen-
tative of single-family residences in the higher-than-
average income area of San Fernando Valley, as
collected in the dry weather of late fall.
Sample 2. Single-family domestic waste, Houston,
-------
12
AIR CLASSIFICATION OF SOLID WASTES
Texas. This sample was to represent the material
delivered by municipal collection trucks to the
composing plant of Lone Star Organics.
Sample 3. Domestic waste from multi-family
dwellings, Los Angeles, California. This sample was to
be taken in essentially the same area and at the same
time as sample 1 and was to be from a collection
route where apartments predominate.
Sample 4. Commercial waste, Los Angeles, Cali-
fornia. This sample was to be taken from selected San
Fernando Valley collection routes along with
shopping centers and small commercial establish-
ments predominate.
Sample 5. Industrial waste sample from auto-
mobile body reclamatipn plant in the Los Angeles
harbor area. This sample was to be representative of
the trash stream containing nonferrous metal and
other materials remaining after magnetic separation
of shredded steel scrap from an automobile
body-crushing operation.
Sample 6. Aged Compost. This sample was to be
representative of material from outdoor stockpiles
that has been aged 6 months or more. The source of
this material was the domestic-refuse-composting
plant of Lone Star Organics, Houston, Texas. The
aged compost contained glass, metal, and other
impurities that made it unsuitable for horticultural
use without further cleaning.
For samples 1 and 2, single-family domestic
wastes, the test program consisted of separation runs
on each of the samples to determine the feasibility of
the following operations:
Salvage of wastepaper for use in paper or board
manufacture
Removal of noncompost material
Removal of metal, glass, and plastic for salvage
Removal of materials to improve the wastes for
destructive distillation (retorting)
Removal of materials to improve the wastes for
fermentation of cellulosic constituents
The test program for samples 3 and 4 was limited
to the recovery of a grade of wastepaper stock known
as No. 1 mixed paper. The objective of classification
was to produce a higher quality of mixed paper—from
the standpoint of cleanliness and contamination by
metal, glass, and other noncellulosic products—than
the scrap paper that is normally supplied by salvage
contractors to secondary fiber processors. Both
samples were expected to yield a salvage wastepaper
of higher quality than would be obtainable from
processing of single-family residential refuse.
The components of sample 5, the nonferrous trash
stream from automobile body reclamation, were not
readily identifiable by visual means. The test pro-
gram included, therefore, analysis of the composition
of various fractions of this waste material after initial
screening into four size ranges. The larger size
fractions could not be air classified. Air classifications
were made of a minus 3/8-in. and a 3/8- to 1-in.
screen fractions. Each fraction was separated by air
classification into seven subfractions, with correspon-
ding superficial air column velocities from 500 to
3,000 ft/min. The subfractions could not be de-
scribed adequately by appearance; therefore, a series
of physical and chemical tests was made for more
complete characterization. These tests were to deter-
mine density, oil and moisture content, percentages
of combustible organic and magnetic material, and
chemical composition as indicated by spectrographic
assays.
The test program for the compost material of
sample 6 required shredding and screening followed
by air classification of relatively close-graded
middlings and coarser material. This procedure pro-
vided the following: (1) a lightweight fraction con-
sisting mainly of plastic (nonbiodegradable); (2) a
heavy, oversize fraction requiring regrinding for re-
turn to the compost digester; (3) a clean, granular
material suitable for sale; and (4) two waste streams
for landfill disposal.
More nearly complete test program descriptions,
together with diagrams and photographs, are given in
conjunction with later discussions of laboratory air
classifier performance. In the detailed test program
employed for air classification of each of the selected
types of waste, data were sought on the following:
Characteristics of the material fed
Density (overall and by size fractions)
Moisture content
Particle size gradations
Visual identification of separated fractions
Analytical identification of separated fractions
Processing required for air classification
Drying
Shredding
Screening (removal of oversize material or
separation into fractions for air classification)
Operating parameters for air classification
Throat velocity
Feed rate
Sources of Sample Material. Samples 1,3, and 4
(San Fernando Valley solid waste material) were
provided through the cooperation of Universal
By-Products, Sun Valley, California. Mr. Richard P.
Stevens, president, is a member of the California
State Department of Public Health's 1968 Solid
Waste Advisory Committee. His firm represents an
integrated operation comprising collection, disposal,
by-products salvage and processing, and equipment
sales. This firm is extremely interested in improved
methods for mechanical salvage of wastepaper from
refuse streams and the salvage of nonferrous materials
from automobile body reclamation operations. Since
their mixed paper is now shredded prior to baling, it
-------
LABORATORY-SCALE UNIT EXPERIMENTS
13
was relatively simple for them to provide the Institute
with shredded samples. These were in the 3- to 6-in.
size range that could be reduced further by the
Institute's laboratory shredding equipment. With this
firm it was possible to make a careful selection of
sample material. To obtain samples that were repre-
sentative of a typical collection route at a given
season of the year, an Institute technician accom-
panied the collection truck. He selected a 4-cu-yd
sample (for primary shredding) from the contents of
the truck at the time of discharge at the landfill.
From the shredded material, a 100-gal sample was
selected for subsequent laboratory processing.
Samples 2 and 6 were provided in three 50-gal
steel drums by Lone Star Orgamcs, Houston, Texas.
Sample 2 consisted of two drums, one that had
undergone primary shredding to a 3- to 6- in. size in a
Williams 475-GA mill and the other that had been
shredded to a nominal 1-in. size by a Williams 80-GA
mill. The material had undergone secondary
shredding and required drying prior to laboratory
storage. Compost sample 6 was taken directly from
the outdoor stockpile of aged material and also
required drying.
Approximately 150 Ib of the nonferrous auto-
mobile body trash, sample 5, was selected by Insti-
tute representatives from the outdoor stockpile of
Clean Steel, Inc. The sample was dry, and care was
exercised to obtain a sample that was visually
representative of the material in the stockpile, which
contained several hundred cu yd of waste.
In addition to these six samples, four additional
domestic and commercial refuse samples were sup-
plied from the U.S. Public Health Service's solid
waste program laboratory in Cincinnati, Ohio. For
each type of refuse, one preground sample that had
been passed through a Williams Model 30-S hammer-
mill, without a screen, and one final-grind sample that
had been shredded with a 1-in., round-hole screen
were supplied. Each of these samples was 1 to 2 cu ft
in size and weighed between 15 and 25 Ib. The
domestic waste samples were very wet and required
air drying before they could be stored at the
laboratory.
The Laboratory-Scale Air Classification Unit
Background. Classification is used preponderantly
in the treatment of raw material. By definition, it is
an operation in which a mass of granular particles of
mixed sizes and different specific gravities is allowed
or caused to settle through a fluid that may be either
in motion or substantially at rest. Sizing or screening
is defined as the separation of various sizes of
particles into two or more portions by means of a
screening surface acting as a multiple "go" and "no
go" gauge such that the final portions consist of
particles of more nearly uniform size than those in
the original mixture. Although the definitions of
classification and sizing describe operations that
apply best to free-flowing granular materials, these
operations can obviously be applied to any parti-
culate material. Properly applied, therefore, both
operations (screening and classification) should have
value for separating various types of solid wastes.
Air classification, air sizing, and dust collection all
deal with different facets of the relative motions
between components of mixtures of solids and gases
(in this case, air). The theory and principles that
apply are covered in a number of references1"4 and
will be presented here only in generalities sufficient
for an understanding of the operation of the labora-
tory unit. In the range of particle sizes and densities
that we are concerned with in the air classification of
solid wastes, the considerations of settling velocity,
buoyancy, and interparticle collisions are not impor-
tant. The primary theoretical consideration in air
classification is that of terminal velocity—the con-
stant velocity reached by a particle falling from rest
in a body of gas at rest when the gravitational pull is
equal to the resistance offered by the gas.
Expressions have been developed for terminal
velocity under turbulent, streamline, and transitional
conditions. These expressions generally apply to
spherically shaped particles and involve the particle's
diameter, its specific gravity, and the density and
viscosity of the gas. Constants in these equations
must be determined experimentally and can,
therefore, be determined for irregular fragments as
well as spherical particles. In all cases, the terminal
velocity increases with increasing particle density and
particle size. Particle shape exerts a great deal of
influence on this velocity, particularly for lightweight
fibrous materials. When the flow is confined, electro-
static forces on smaller sizes of these materials can
become as important as gravitational forces. The air
velocity required to float a particle when the current
as a whole is vertically upward is usually different
from the velocity with which the particle settles in
still air, and both are different from the velocity
necessary to transport the particle, as for pneumatic
conveying, when a major component of the current
direction is horizontal. Although related to terminal
velocities and floating velocities, fluidizing velocities
for the zigzag air classifier, as reported herein, are not
directly comparable, because of the acceleration
effects caused by impact of the particle with the
column's walls and because of the special conditions
of turbulence produced by the tortuous airflow.
Zigzag air classification has been pioneered by the
Institute for processing dry mixtures containing
material that can be fluidized and transported in an
airstream. Particles of these mixtures are fractionated
according to density, size, and aerodynamic pro-
perties. Thus, zigzag air classification is somewhat
-------
14
AIR CLASSIFICATION OF SOLID WASTES
analogous to distillation of hydrocarbon liquids in a
petroleum refinery's fractionating column. The zigzag
principle permits separation of materials with only
slightly differing densities whose other properties,
such as size, are identical and has proved in almost all
cases examined to be more efficient than any other
air classification system.
Frequently, the air classifier is used with chopping,
grinding, and screening processes to obtain particles
of nearly uniform characteristics for improved pro-
cess efficiency. Capacity, both in terms of throughput
rate and maximum particle size, can be increased by
increasing the dimensions of the column's throat. The
range of density can be increased by increasing the
maximum throat velocity. Thus, the air classifier is a
flexible separating tool.
As a tool for the empirical design of full-scale
separating systems, the Scientific Separators labora-
tory-size air classifier as evaluated by SRI has been
used for separating products such as the following:
Roasted coffee beans (coffee has been graded to
separate dense, high-quality coffee from improperly
roasted or cull beans).
Seed and grain cleaning (successful work in this
area has led to the design of commercial equipment
that is now being marketed).
Dehydrated alfalfa (dried leaf has been removed
for pelletizing before the stems are sold for cattle
feed).
Oil seed meal (the protein level of rendered meal
can be increased by selective removal of the oil seed
hulls).
Fish protein concentrate (experiments have indi-
cated that a combined process, including air classifica-
tion, may be the most efficient as well as the least
expensive method for controlling fluoride concentra-
tions by the selective removal of bone).
Data on typical performance of the laboratory
unit on materials processed in previous experimental
work at the Institute indicates that increasing throat
velocities and increasing horsepower are required with
increases in the bulk density of material classified
(Table 2).
Physical and Operating Characteristics of
Laboratory-Scale Air Classification Unit.The
Scientific Separators zigzag air classification unit
available at the Southern California Laboratories of
Stanford Research Institute is a 12-stage, zigzag
column (fed at the eighth stage from the bottom),
through which air is drawn by a high-capacity
induction blower. This blower draws air first through
the column and then through a conventional cyclone
separator that removes material passing upward
through the column. Heavy material that cannot be
transported in the airstream at any set velocity moves
countercurrently to the stream and is discharged from
TABLE 2
TYPICAL FLUIDIZING VELOCITIES AND
HORSEPOWER REQUIREMENTS FOR AIR
CLASSIFICATION ON MATERIALS PROCESSED
IN PREVIOUS EXPERIMENTAL WORK AT SRI
Bulk
density
Material (Ib/cu ft)
Estimated range of
Superficial horsepower
velocity requirements
(ft/min) per ton/hour capacity
Dehydrated
alfalfa 5* 800-900
0.45-0.67
Garlicf 20-30 1,450-1,700 0.75-1.1
Almonds* 40 1,750-2,000 0.80-1.3
Raisins 40^t5
Peanuts 45-50
Pinto beans 45-50
2,150*
2,550*
2,600*
1.0-1.5
1.1-1.8
1.1-1.8
*Appioximate.
fRoot crowns being separated from cloves.
tSplits and hulls being separated from whole meats.
the bottom of the column. The separation achieved
by the process can be observed through the trans-
parent side walls of the column. The general arrange-
ment of the unit and a flow diagram are shown
(Figure 2).
The 1 -hp induction blower can provide superficial
velocities of as much as 3,000 ft per min for
fluidizing material in the 2- by 6-in. throat of the
classifier. The flow is controlled by means of a sliding
gate valve, and pressure drop is measured by means of
a manometer. Originally, this manometer operated
across an orifice in the air outlet of the cyclone
separator, but it was found that the direct measure-
ment of the pressure drop across the classifier and the
cyclone could be calibrated in terms of throat
velocity.
The laboratory-scale unit is operated for batch
processing only. Although the rotary airlock feeder
permits continuous feeding, separated overhead
material dropping out of the cyclone separator must
be withdrawn intermittently. A feeder attachment is
available for insertion between the rotary feed valve
and the column that allows fibrous agglomerates to
be broken up and propelled into the column mechani-
cally. A feed hopper with a screw-type conveyor is
also available. A tight cover on this hopper prevents
leakage of air into the column at the feed point, this
leakage being undesirable because it disturbs the
uniformity of the column's flow pattern.
When the column is operated to separate a mixture
of two uniform granular materials of different aero-
dynamic characteristics, it is a simple matter to adjust
-------
LABORATORY-SCALE UNIT EXPERIMENTS
15
-------
16
AIR CLASSIFICATION OF SOLID WASTES
the airflow—as the mixture is fed at a uniform rate
into the column—so that conditions of feed rate and
airflow are achieved that permit free flow of solid
material both up and down the column. Under these
conditions, both the heavy and light materials are
separated and reentrained at each step in the column.
A relatively high-speed airstream exists in the
center of the column and along the column's faces
exposed to the upward flow. Active vortices are
created in the pockets of the column and on the faces
along which the heavier material is falling. The falling
material and the upward flowing air that is trans-
porting the lighter material provide a combined
centrifugal and entrainment action (Figure 3). Flow
patterns were further confirmed by observing smoke
patterns with airflow only. These were made at
velocities of approximately 500, 750, and 1,100
ft/min. At the lower velocity, flow in the central core
appeared to be laminar, with vortices at the edges
that project into the stream. At the 750 ft/min throat
velocity, the space occupied by the corner vortices
becomes smaller and the central stream becomes
wider. At velocities higher than 1,000 ft/min, corner
vortices could not be observed; the entire cross
section appeared to be full of turbulent smoke.
CIRCULATION
PATTERN
Figure 3. Flow pattern of granular material in the Scientific
Separators zigzag air classifier evaluated by SRI.
Before the experimental work on solid wastes
reported herein, a series of experiments were made on
solid waste material that included combined domestic
refuse, demolition waste, and a sample of the
automotive shredding trash stream. These preliminary
experiments indicated that air classification could
prove to be a useful unit operation in the continuous
processing of a number of types of solid wastes.
High-capacity separation on materials of different
densities appeared to make the unit particularly
applicable to the removal of concrete, metal, and
similar contaminants from demolition wastes so that
the wood fraction could be used for the production
of paper pulp, fiberboard, or charcoal briquettes. Air
classification appeared to be equally useful as a
pretreatment for combined domestic wastes that were
to be fed to a retort for the production of char oil or
briquetted fuel or were to be processed to recover
coarse cellulose for fiberboard, roofing, or similar
products. The primary difficulty encountered in these
preliminary experiments was caused by the feeder
mechanism of the classifier. The screw-type feeder
used in the preliminary experiments was built pri-
marily for granular materials, and it was believed that
the difficulty could be easily remedied by minor
redesign.
Modification of Laboratory-Scale Unit to Facili-
tate Testing of Selected Wastes. For granular
materials such as beans and seeds, the hopper with a
screw conveyor in the bottom proved to be an
excellent device for introducing the material to be
classified at a uniform rate. Varying the speed of the
screw proved to be an effective way of controlling the
feed rate to the classifier. Initial experiments with
solid wastes demonstrated, however, that the screw
conveyor would be inappropriate for heterogeneous
material because of the ease with which the feeder
could become jammed. In addition, this type of
feeder did not provide a positive seal against entry of
air, and it was necessary to have a tight feed-hopper
cover for proper operation of the column.
These problems were largely remedied for this
project by fabricating a new hopper that incorporated
a four-blade, rotary airlock feeder with rubber blades.
This feeder provided both searing and metering
functions. The maximum particle size of solid
material was limited to 1-in. spheres by the feeder
mechanism. As an accessory for the rotary feeder, a
small electrically driven shaft with prongs extending
radially from it was installed in 1he feed slot between
the rotary feeder and the feed plate of the classifier
column. The shaft projections mated with similar
stationary projections on the bottom of the feed slot
and provided a means of separating agglomerated
fluffy material and projecting it into the airstream.
This so-called "kicker" limited the particle size to no
-------
LABORATORY-SCALE UNIT EXPERIMENTS
17
more than about 3/8-in. diameter. With the stationary
teeth removed, particles up to 1/2-in. or slightly more
in diameter could be handled. Thus, both the column
throat and the feeder mechanism constituted limita-
tions on the size and shape of particles that could be
fed to the column. With most solid wastes, therefore,
a need existed for shredding before air classification.
The question immediately arose whether or not
additional shredding should be employed to prepare
an already shredded material for air classification in
the laboratory column, since this would change the
aerodynamic properties of the material. This applied
particularly to the automobile body trash stream.
Rather than reshred, it was believed that a classifier
column should be designed with a throat large enough
to process the material in the already existing size
range if air classification processing were to be used
commercially for this material. Since the throat size
of the column is related to the capacity of the
column as well as to the size of the particle, the
design of a full-scale column would also take into
account the rate at which it was desired to process
the material. Further, considering commercial scale-
up, there is a relationship between column size and
initial and operating cost and the cost of shredding
or crushing the material for processing by air classifi-
cation. These relationships are suggested by Figure 4,
which indicates that there is an optimum size to
minimize the combined cost of shredding and air
classification.
Another consideration related to particle size must
also be taken into account in deciding what degree of
size reduction is to be employed for laboratory air
classification experiments. The alternative to com-
plete shredding is to remove oversize material by
screening, an operation that is practical if the oversize
constituents are of such a nature that their classifica-
tion characteristics can be determined readily by
judgment.
In addition to the new feed hopper with its rotary
valve and "kicker" attachment, three additional
modifications of facilities were required. New 1 - and
1-1/2 in. screens had to be fabricated for the
laboratory screening equipment. It was necessary to
relocate the laboratory shredder and air classification
column to a roofed, outdoor dock to provide a
working area sufficiently large for both pieces of
equipment and for storage of the material being
processed. It was necessary also to fabricate a new bar
grate for the 5-hp McCormick No. 4E hammermill
used for laboratory shredding. The 1-in. and smaller
round-hole screens previously used with this shredder
produced an overshredded product (almost dry pulp)
when fed a waste with large paper content. Although
not wholly satisfactory, the shredder output with the
1-1/2-in. bar grate represented a significant improve-
ment.
Performance of the Laboratory-Scale
Unit on Selected Wastes
Operating Procedure. When samples of the selected
wastes arrived at the laboratory, the densities and
moisture contents were determined. If the samples
had more than approximately 20 percent moisture
content, they were air dried so that they could be
stored without decomposition. The wastes were also
inspected to determine visually the nature of the
major constituents and the degree of contamination.
Even though all samples received were in the range of
3- to 6-in. particle size, or less, they could not be fed
directly to the small laboratory-scale air classification
unit. Either shredding or removal of the oversize
fractions by screening was necessary before the
material could be processed.
Shredding was usually done dry because wet
material was subject to a greater degree of over-
shredding. On relatively moist samples, however, a
drying effect amounting to as much as a 5 percent
change in moisture content was produced by the
shredding operation. Shredding of materials having a
large paper content at moisture levels approaching
100 percent produced a definite pulping effect.
Shredding at a large moisture content, or even
addition of moisture during the shredding operation,
is claimed to be desirable to reduce the explosion
hazard in • commercial operations. Following shred-
ding, a preliminary air classification run was generally
made on the sample at its air-dried moisture content.
Experiments were made with some materials to
determine whether or not an improvement in the ease
of separating plastic wrapping materials and news-
paper fractions could be brought about by moistening
the sample slightly before air classification. This
appeared to be of some benefit, especially where
electrostatic effects inhibited separation. In commer-
cial practice, however, accurate control of the mois-
ture content of material to be air classified would
complicate the operation to a prohibitive extent.
Screening before air classification often improved
the separating effectiveness of the air classifier by
limiting the range of sizes in the classifier feed. It also
appeared, in this series of experiments, to remove
effectively leaf fragments and other low-density fines
that adhered to paper and cardboard. Three types of
laboratory screens were available as follows: (1) a
shaker screen, which can continuously separate as
many as lour different sized gradations; (2) a
standard Ro-tap unit used with Tyler testing sieves;
and (3) a laundry tumbler-type drier in which the
basket could be operated as a rotating screen.
Screened fractions were air classified to determine the
combination of separations and removals that ap-
peared to produce the best results. In most cases, this
could be determined visually. In some cases, however,
-------
18
AIR CLASSIFICATION OF SOLID WASTES
notably the automobile body trash stream, the
constituents were not readily identifiable, and analy-
tical procedures had to be used to identify the
concentrations being produced by air classification.
In all cases a method was sought that would
involve minimum processing only. The order of
sequential operations was approximately as follows:
Air classification of air-dried sample
Air classification at controlled moisture content
Shredding plus air classification
Screening plus air classification
Shredding plus screening followed by air
classification
From the initial series of experiments, a detailed
test program was developed for each waste sample.
The samples were then run according to the detailed
procedure and results evaluated. If improvement in
separating efficiency or concentration of recoverable
fractions appeared desirable or achievable, the de-
tailed test procedure was modified and the series of
experiments rerun. Results reported for the individual
wastes may thus be the outcome of a number of
minor improvements of the initial test program, the
fundamental objective remaining the same and only
the procedures being modified.
No standard methods of determining the physical
characteristics of solid waste could be found. For
moisture content, the determinations made for this
study followed procedures described in ASTM D143,
Tests of Small Clear Timber Specimens, Sections 122
through 125. Bulk density tests for building mate-
rials, such as those described in ASTM C519-63T for
fibrous, loose-fill building insulation, appeared to be
most adaptable to density determinations for solid
wastes. If the waste were to be reduced first to a
uniform moisture content, a great deal of information
might be obtained about its physical nature by
conducting a three-step density measurement. Ideally,
these density determinations would be made (1) after
the waste had been loosely packed into a container,
according to some standard method of procedure; (2)
after compression to 10 psi (conventional packer
forces amount to 7 to 15 psi); and (3) after rebound.
Unfortunately, time did not permit development of
this density procedure to the point where it could be
reliably employed. Uncontrolled, loose-fill density
determinations made for the material as received were
used instead and are the bulk densities reported
herein.
The determination of moisture and volatile oil
content for the automobile body trash samples
followed procedures described in ASTM Dl800-63,
Moisture and Creosote Preservative in Wood. These
determinations were made on the screened and
air-classified fractions of this malerial. Moisture con-
tents were also spot checked by drying samples in an
oven.
Performance of the Laboratory-Scale Unit on
Selected Wastes. Detailed procedures for operating
the air classification unit, together with other neces-
s
COST OF
AIR CLASSIFICATION
LARGE
DECREASING
PARTICLE SIZE
SMALL
Figure 4. Hypothetical cost relationships for shredding and air classification.
-------
LABORATORY-SCALE UNIT EXPERIMENTS
19
sary unit operations of a pilot-plant system for the
laboratory processing of each solid waste sample,
were determined by the approach outlined in the
previous paragraph. Operating procedures and the
performance of the laboratory unit are reported in
this section for the three general classifications of
solid wastes that were processed in this experimental
program. The characteristics of the wastes, the
procedures for their processing, and results are now
reported.
Compost from Municipal Refuse. The compost
material, aged approximately 6 months in the out-
door stockpile of the Lone Star Organics waste
disposal plant at Houston, Texas, was received in a
single 55-gal steel drum. The net weight of the sample
was 233 Ib and its calculated bulk density was 32.7
Ib/cu ft (833 Ib/cu yd). On the assumption that the
barrel had been filled before shipment, the sample
had compacted to 87 percent of its original volume
during the 10-day railway express shipment from
Houston to Los Angeles.
At the Houston plant*all waste material processed
is collected from residential routes. After manual
selection and sorting, followed by magnetic separa-
tion to remove tin cans, the waste is shredded, mixed
with sewage sludge to a moisture content of 60 to 70
percent (dry basis) and continuously digested while
being aerated and mechanically agited. After 6 days'
residence time, during which aeration is regulated so
that the temperature increases from 135 F to a final
temperature of approximately 170 F, the material is
discharged to an outdoor stockpile for completion of
the composting action.
When received, the compost sample was too wet
for storage (moisture content estimated at 80 to 100
percent) and had to be air dried. Before being air
dried, a small portion of the sample was air classified
at velocities of 1,100 and 1,600 ft/min. Although the
sample ran well, the separations were unsatisfactory
as determined by visual observation. After being air
dried to approximately 20 percent moisture content
(dry basis), the sample was shredded through a 1-in.,
round-hole screen to break up lumps in the dried
material. Dedusting at a superficial velocity of 600
ft/min effectively removed fines, fibrous material,
and plastic wrap. There materials were present,
however, in only very small amounts. Some feeder
trouble and a small amount of column clogging were
encountered. Separation into two fractions at 800
ft/min, followed by screening to remove fines, did
not produce a substantially improved product. A
shredded sample was then screened before air classifi-
cation to remove glass, dirt, and the like from the
compost. Classfication of material remaining on No. 8
screen at 400, 500, and 1,100 ft/min produced
separations that appeared to have little commercial
value.
On the basis of these preliminary experiments, a
test program was developed. This program is dia-
grammed (Figure 5), and photographs of the original
shredded compost material and the fractions into
which it was separated by a combination of screening
and air classification are presented.
Figure 5 indicates that approximately 20 percent
by weight of the stockpile material can be recovered
as horticultural-grade compost. The air-classified
bottoms of the coarser screened material might be
ground, and this product would conceivably have a
market as bulk compost for use in landscaping of free-
way slopes and other such areas. The diagram also
indicates the possibility of the materials being
returned to the compost digester, if it contains a
considerable amount of uncomposted material that
would benefit by additional digestion. When this
material is combined with the horticultural material,
55 to 60 percent recovery of the stockpile is
obtained. Approximately 35 to 40 percent would
require ultimate disposal as landfill, and approxi-
mately 5 percent could be burned to obtain heat for
drying, if required.
A small amount of work was done with the
screened fines, and it appears that air classification of
this fraction followed by tabling of the overhead
might recover as much as half of the stabilized
organic fines. Air classification alone does not pro-
duce a satisfactory product, because of its contamina-
tion by fine glass. The presence of glass would also
create slag problems if this fraction were to be
burned.
Automobile Body Trash. It was considered desir-
able from the standpoint of scale-up of laboratory
results to commercial operation to remove oversize
material from the automobile body trash sample by
screening and to experiment with air classification of
those portions of the sample that could be processed
iri the laboratory without further shredding. (The
hammermill shredding operation that produces the
trash stream being reported on in this section has
been described by Ralph Stone and Company.5)
Essentially, this stream consists of all material that
cannot be removed magnetically from the output of a
5,000-hp, 600-rpm hammermill that is fed junked
automobile bodies from which easily salvageable
copper items (such as radiators) tires, and in some
cases engines and transmissions, have been removed.
This material was divided into four fractions as
*For descriptions of this plant, see Piescott, J. H.
Composting plant converts refuse into organic soil condi-
tioner. Chemical Engineering, p. 232-234, Nov. 6, 1967 and
American City, Compost works in Houston. Oct. 1967.
-------
20
AIR CLASSIFICATION OF SOLID WASTES
o
s-
•s
•a
•s
II
II
It
•o u
gi'rt
ll
It
+1>
-------
LABORATORY-SCALE UNIT EXPERIMENTS
21
follows: fraction 1 consisted of material retained on a
1-1/2-in. screen; fraction 2, of material passing a
1-1/2-in. screen but retained on a 1-in. screen;
fraction 3, of material passing a 1-in. screen but
retained on a 3/8-in. screen; and fraction 4, of
material passing the 3/8-in. screen. Fractions 1 and 2
could not be fed to the laboratory classifier. A
description must suffice to characterize these
materials. Fractions 3 and 4 were characterized by air
classification.
Fraction 1-more than 1-1/2 in. in size. This
sample appeared to be equally divided between
organic material (rubber, cardboard, cloth, and paper)
and inorganic material (metal and dirt). There was a
surprising amount of fine dirt, rust, and glass frag-
ments that were carried through with the coarse
fractions by being entrapped in shreds of fiber,
cardboard, and cloth. Rubber moldings, hose, and
extruded seals predominated; many chunks were 6 to
8 in. long or longer and 3/4- to 1-1/2-in. maximum
transverse dimension. Rubber- and asphalt-
impregnated cardboard sheet material were also
present in significant quantity, along with other
combustible sheet material and fabrics of various
sorts such as seat cover materials. The metal material,
which appeared to be about one-fourth to one-third
of the sample by volume and probably more than 50
percent by weight, consisted of chrome-plated pot
metal (zinc) and aluminum trim, a large crumpled
piece of l/16-in.-thick sheet zinc about 18-in. long by
6-in. wide, and some copper and iron parts. Identifi-
able objects were door handles, an ignition coil, the
armature from a starter or generator, and a small
electric motor. There were also several small electrical
components and a portion of the heavy glass from a
head lamp. There was little wood or wire.
Fraction 2—maximum size between 1 and 1-1/2 in.
This sample was predominantly organic, metals and
glass accounting for only about 10 to 15 percent of
the total volume. There were many 2- to 3-in. lengths
of rubber molding and a number of plastic parts. The
greatest amount of material consisted, however, of
roughly equidimensional pieces of shredded sheet
cardboard, sheet rubber, upholstery padding, floor
matting, and fabric. The metal fragments were pri-
marily diecast zinc, many from chrome-plated trim
parts. Hydraulic brake wheel cylinder plungers were
recognized. There was little wire or wood and very
little iron. Fines were present but in somewhat lesser
quantity than in fraction 1. There were several
chunks of broken safety glass held together by the
safety laminate.
Fraction 3-3/8 to 1 in. in size. This sample was air
classified to determine its composition. Most of the
material was oil soaked and dirt impregnated and was
fibrous; the fraction was homogeneous. Metal frag-
ments were not easily recognizable although insulated
wire could be identified. Small splinters of wood were
also present.
Fraction 4—smaller than 3/8 in. in size. This
fraction was air classified. It was oily and fibrous,
appearing to be about an equal mixture of fiber and
dirt. Except for its oil-soaked appearance, it re-
sembled the material found in the dustbag of a
household vacuum cleaner.
The relative sizes of the samples, by weight and by
volume, and their densities are listed (Table 3).
Fractions 3 and 4 were each fed to the air
classification column and separated into seven sub-
fractions at column differentials (fluidizing velocities)
that were arbitrarily selected to give reasonably sized
sub fractions for analysis. The original fractions were
first classified at a column differential of 0.2 in. of
water column differential. The overhead fraction was
collected and the bottoms fraction passed through
the column again at a column differential of 0.5 in. of
water. This overhead fraction was collected and the
bottoms fraction passed through the column again at
1 in. of water column differential. This procedure was
repeated until separations had been made and over-
head fractions obtained from column differentials of
0.2, 0.5, 1, 2, 4, and 8 in. of water column
differential, representing superficial velocities in the
column throat of 400, 600, 800, 1,100, 1,600 and
2,500 ft/min, respectively. The 8-in. bottoms sample
was also collected. Compositions and analyses of
these subfractions are given (Tables 4 through 7).
Visual examination revealed little about the origi-
nal fractions or about the subfractions separated by
air classification; photographs were taken of auto-
mobile body trash fractions selected as typical and
are reproduced in Figures 6 and 7.
Because of its semigranular nature and the rela-
tively high proportion of dense material, the auto-
mobile body trash samples fed well and could be
classified at high rates with no difficulty.
Results obtained by pyrolizing fractions 3 and 4 at
approximately 1,500 F in an electrically heated batch
retort are given in Appendix B, which contains the
report of the San Diego Utilities Department. The
conclusions drawn from this report are as follows:
1. Both fractions have large contents of inert
material such as metal, dirt, and glass, separated inerts
plus ash in the char exceeding 50 percent of the
original sample.
2. In the coarse fraction, the calorific values of
the chars are almost four times those of the fines,
because of the large ash content of the char in the
fine material. The total heating value is about 2,100
Btu/lb for the coarse fraction and 1,400 Btu/lb for
the fine fraction. Approximately 55 percent of the
heat produced comes from the gas in the fines and
-------
22
AIR CLASSIFICATION OF SOLID WASTES
TABLE 3
FRACTIONS RESULTING FROM SCREENING RESIDUE
AFTER MAGNETIC SEPARATION FROM AUTOMOBILE BODY SHREDDING
Fraction Screen size particle
number retained or
1 1-1/2-in. square hole
2 1-in. mesh, 16-gauge,
square hole
3 10-mm round hole
4
Total sample
Weight
db)
37.5
17.0
30.0
64.0
148.5
Volume
(cu ft)
1.07
0.72
1.15
1.42
4.36
Percentage of
total sample
By weight
25.2
11.4
20.2
43.2
100.0
By volume
24.5
16.5
26.4
32.6
100.0
Bulk
density
(Ib/cu ft)
35.0
23.6
26.0
45.0
34.0 (avg)
TABLE 4
SUBFRACTIONS RESULTING FROM AIR CLASSIFICATION.
SCREEN FRACTION 3 (COARSE)-AUTOMOBILE BODY TRASH
Designation
3,8-in. BTMf
3,8-in. OH*
1,4-in. OH
3, 2-in. OH
3, 1-in. OH
3, 0.5-in. OH
3, 0.2-in. OH
Total sample
Superficial
velocity
(ft/min)
2,500
2,500
1,600
1,100
800
600
400
Percentage of
Weight
(g)
506.1
275.2
199.5
144.2
122.6
81.4
5.7
1,334.7**
total
By weight
37.9
20.6
15.0
10.8
9.2
6.1
0.4
100.0
sample
By volume
9.2
13.3
14.7
19.1
34.6
7.7
1.4
100.0
Dry density*
(G/cc)
1.09
0.41
0.27
0.15
0.07
0.21
0.08
(Ib/cu ft)
68.0
25.6
16.8
9.4
4.4
13.1
5.0
*Volume of all samples weighed for density determination was 50 cc.
•f Screen fraction 3, 8-in. manometer reading, bottoms (BTM) subtraction.
*Screen fraction, 3, 8-in. manometer reading, overhead (OH) subfraction.
**46.2 g of wire removed before weighing.
TABLE 5
SUBFRACTIONS RESULTING FROM AIR CLASSIFICATION.
SCREEN FRACTION 4 (FINE)-AUTOMOBILE BODY TRASH
Designation
4, 8-in. BTMf
4, 8-in. OH*
4, 4-in. OH
4, 2-in. OH
4, 1-in. OH
4, 0.5-in. OH
4, 0.2-in. OH
Total sample
Superficial
velocity
(ft/min)
2,500
2,500
1,600
1,100
800
600
400
Weight
(g)
283.7
376.6
532.5
154.7
80.1
120.9
64.8
1,613.3
Percentage of
total sample
By weight
17.5
23.3
32.9
9.6
5.1
7.5
4.1
100.0
By volume
6.1
12.7
40.8
15.9
9.1
9.5
5.9
100.0
Dry density*
(G/cc)
1.54
0.98
0.43
0.32
0.29
0.42
0.36
(Ib/cu ft)
96.0
61.2
26.8
20.0
18.1
26.2
22.5
* Volume of all samples weighed for density determination was 50 cc.
fScreen fraction 4, 8-in. manometer reading, bottoms (BTM) subfraction.
tScreen fraction 4, 8-in. manometer reading, overhead (OH) subfraction.
-------
LABORATORY-SCALE UNIT EXPERIMENTS
23
only 30 percent from the gas in the coarse material.
Spectrographic assays were performed on fractions
3 and 4, and the results are listed (Table 8). Results
of similar assays on the air-classified subfractions are
also presented (Tables 9 and 10).
Spectrographic analyses are usually made on ores,
the material being vaporized in a carbon arc. It is first
necessary, therefore, to reduce a sample to ash, and
this was done. The ash from the entire fine material
sample that was analyzed represented 68 percent of
the original screen fraction, while that from the
coarse material represented 59.5 percent of the
original screen fraction. For the subfractions, the
percentage ash ranged from 40 to 85.
The relative abundance of the various basic ele-
ments is disclosed from a Spectrographic analysis.
However, the elemental combinations and the mineral
forms in which the metallic elements occur probably
TABLE 6
PHYSICAL CHARACTERISTICS OF SUBFRACTION
RESULTING FROM AIR CLASSIFICATION.
SCREEN FRACTION 3 (COARSE)-AUTOMOBILE BODY TRASH
Percentage
Designation
3, 8-in. BTMt
3, 8-m. OH**
3, 4-m. OH
3, 2-in. OH
3, 1-in. OH
3, 0.5-in. OH
Total sample
Superficial
velocity
(ft/min)
2,500
2,500
1,600
1,100
800
600
Percentage
moisture*
0.9
2.4
4.1
-
6.3
4.7
3.3
Percentage
extractable
oil*
12.7
28.4
35.8
-
14.6
13.4
24.2
Percentage
magnetic
material
33.2
9.5
3.0
2.5
2.0
1.0
--
combustible
materialf
Run 1 Run 2
21.0
46.1
61.5
67.0
61.0
44.0
41.4
19.7
53.3
61.0
60.5
55.0
54.4
--
*Following procedure described in ASTMD-1860.
fFollowmg procedure described in ASTM D271-58, 600 C oven. Run 1, 20-g samples from initial
experimental classification; Run 2, approximately 500-g samples (57 to 585 g) from production
separation.
tScreen fraction 3, 8-in. manometer reading, bottoms (BTM) subfraction.
**Screen fraction 3, 8-in. manometer reading, overhead (OH) subfraction.
TABLE 7
PHYSICAL CHARACTERISTICS OF SUBFRACTION
RESULTING FROM AIR CLASSIFICATION.
SCREEN FRACTION 4 (FINE)-AUTOMOBILE BODY TRASH
Percentage
Designation
4, 8-in. BTM*
4, 8-m. OH**
4,4-in. OH
4, 2-in. OH
4, 1-in. OH
4, 0.5-in. OH
4, 0.2-in. OH
Total sample
Superficial
velocity
(ft/min)
2,500
2,500
1,600
1,100
800
600
400
Percentage
moisture*
0.5
0.6
1.4
1.7
1.5
2.5
2.9
1.5
Percentage
extractable
oil*
15.1
13.7
9.7
10.0
10.2
1.1
10.3
Percentage
magnetic
material
36.5
24.5
25.5
14.0
16.0
15.5
7.5
17.5
combustible
material!
Run 1 Run 2
3.0
18.9
24.5
27.6
24.5
24.0
25.5
25.5
_ _
14.8
19.1
27.1
30.6
30.0
30.9
*Following procedures described in ASTM D-1860.
fFollowing procedures described in ASTMD 271-58, 600 oven. Run 1, 20 g samples from initial
experimental classification; Run 2, approximately 500 g samples (123 to 405 g) from production
separation.
tScreen fraction 4, 8-in. manometer reading, bottoms (BTM) subfraction.
**Screen Fraction 4, 8-in. manometer reading, overhead (OH) subfraction.
-------
24
AIR CLASSIFICATION OF SOLID WASTES
o
H
<
o:
1
o
O
I-
-------
LABORATORY-SCALE UNIT EXPERIMENTS
25
ei
!" a.
8
I
e
•o
.a
x
•O
2
O Q.
o
o
a
.1
-------
26
AIR CLASSIFICATION OF SOLID WASTES
f
I- u-
O
u
c
.2
1
I
•g
JD
—
-------
LABORATORY-SCALE UNIT EXPERIMENTS
27
TABLE 8
SEMIQUANTITATIVE SPECTROGRAPHIC
ANALYSIS OF SCREENED FRACTIONS OF
RESIDUE AFTER MAGNETIC SEPARATION
FROM AUTOMOBILE BODY SHREDDING*
Ash from Ash from
fraction 3, fraction 4,
coarse materialf fine material*
(% by weight) (% by weight)
Silicon
Iron
Copper
Calcium
Aluminum
Zinc
Magnesium
Chromium
Barium
Boron
Titanium
Lead
Tin
Manganese
Nickel
Molybdenum
Vanadium
Sodium
Silver
Zirconium •
Cobalt
Strontium
Antimony
Potassium
Other elements
10.9
32 (magnetic)
1.1
2.1
5.5
6.8
1.3
0.046
0.14
0.02
0.88
1.8
0.17
0.051
0.20
0.0032
0.007
Trace
0.0009
0.011
0.0058
0.011
0.14
Nil
Nil
16
21
1.6
3.8
5.1
2.1
1.7
0.095
0.16
0.016
0.91
1.1
0.090
0.077
0.31
0.0097
0.0069
Trace
0.0042
0.019
0.042
0.015
Nil
Trace
Nil
*Refuse material ashed at 600 C, per ASTM D271-58, to
produce samples for spectrographic assay. Total will not
equal 100%, because gases do not show. Quartz, for instance,
is 28 parts silicon and 32 parts oxygen, which would show
46.7% silicon. For most ores, the total is less than 50%.
Source: Materials Engineering Co. (MECO) analysis dated
October 11,1968.
fAsh was 59.5%, by weight, of original fraction. Course
material designated MECO Sample 38699.
iAsh was 68%, by weight, of original fraction. Fine
material designated MECO Sample 38698.
bear little resemblance to the combinations and forms
within a naturally occurring ore. The total sample
appears to be fairly rich in aluminum, zinc, copper,
and lead, with zinc concentrated in the coarse
fraction. Such concentration could make it economi-
cal to process this material—especially so, when it is
realized that quantities of from 50 to 200 tons per
day can be obtained free at automobile body-
processing plants in metropolitan areas.
From data on the air-classified subfractions
(Tables 9 and 10), zinc was further concentrated in
the heavy bottoms of the coarse screened material.
More than 12 percent of the total zinc in material
that was less than 1 in. in particle size was concen-
trated in this subfraction, and this concentration
permitted recovery of as much as 13 Ib of zinc per
ton of the total nonmagnetic trash stream. Concentra-
tions of other materials were indicated that may offer
promise when the material is considered for pro-
cessing as an ore. These materials include iron,
chromium, copper, and possibly lead or titanium.
Most of these metals or oxides appear to be concen-
trated in the heavy subfraction, either the overhead
or bottoms, and this circumstance would make for a
high-volume, low-cost separating operation.
Municipal refuse. Eight solid waste shredded
samples of domestic and commercial solid wastes that
were primarily wastepaper were obtained from the
following locations:
Los Angeles, California (San Fernando Valley)
Domestic waste from single-family residences
Commercial wastes
Houston, Texas
Domestic waste from single-family residences,
coarse and fine grind
Cincinnati, Ohio
Domestic waste, coarse and fine grind
Commercial waste, coarse and fine grind
All samples were from combined collections. Sample
sizes ranged from 15 to 100 Ib; all samples provided
an adequate quantity of waste for air classification
experimentation. Information on these samples as
they were received is summarized (Table 11).
We do not intend in this report to add to the
confusion on solid waste terminology that is already
extensive in statutes, literature, and in the vocabu-
laries of professionals and laymen alike. Although we
may use somewhat interchangeably terms such as
garbage, rubbish, refuse, trash, and solid wastes, we
accept the definition of solid wastes as used by the
California State Department of Public Health in their
recent publication on solid wastes management,^
which defines solid waste as "all those materials that
are solid or semisolid and that the possessor no
longer considers of sufficient value to retain." Thus
the term "solid wastes" is all inclusive and embraces
all types of classifications, sources, and properties.
Expected ranges in percentage composition of
mixed municipal refuse from U.S-. cities are as follows
(Table 12): paper, 37 to 60; metallics, 7 to 10; food,
12 to 18; other materials (leaves, wood, glass, plastic),
1 to 12. It is not, however, to be expected that any
single sample of refuse would follow this composition
exactly. The routing of collection vehicles as well as
seasonal and economic factors will certainly influence
the composition of any given refuse sample.
From visual inspection, it can be said that the
Cincinnati refuse samples contained most of the
components listed, and the composition fell within
the ranges given (Table 12). The Houston refuse
contained less newsprint and cardboard in the pri-
-------
28
AIR CLASSIFICATION OF SOLID WASTES
TABLE 9
SEMIQUANTITATIVE SPECTROGRAPHIC ANALYSIS OF
AIR-CLASSIFIED SUBFRACTIONS AFTER MAGNETIC
SEPARATION FROM AUTOMOBILE BODY SHREDDING.
SCREENED FRACTION 3 (COARSE)-AUTOMOBILE BODY TRASH*
Percentage, by weight, from ash sample of subtraction
3, 8.0 BTMf 3, 8.0 OH* 3, 4.0 OH 3,2.0 OH 3, 1.0 OH 3, 0.5 OH
Silicon
Iron
Copper
Calcium
Aluminum
Zinc
Magnesium
Chromium
Barium
Boron
Titanium
Lead
Tin
Manganese
Nickel
Molybdenum
Vanadium
Sodium
Silver
Zirconium
Cobalt
Strontium
Antimony
Other elements
Total percentage
ash by weight,
at original
sample
17
24
2.0
4.2
4.9
7.6
2.0
0.026
0.34
0.0098
0.17
4.0
0.096
0.14
0.25
0.0073
Trace
Trace
0.00039
Nil
0.0084
0.070
0.21
Nil
80.3
0.77
21
0.42
0.26
1.5
0.20
0.043
3.2
Nil
Nil
0.057
0.028
0.041
0.19
0.075
0.012
0.0072
Nil
Nil
Nil
Trace
Trace
Nil
Nil
46.7
20
11
3.2
5.0
4.1
1.5
2.1
0.052
0.58
0.027
0.87
0.81
0.13
0.12
0.11
0.0044
0.0025
Trace
0.00029
0.020
0.013
0.10
Nil
Nil
39.0
12
6.9
2.8
2.8
2.4
0.81
1.2
0.027
0.48
0.026
0.42
1.1
0.20
0.13
0.072
0.0055
0.0021
Trace
0.00019
0.013
0.0046
0.058
Nil
Nil
39.5
12
18
26
22
3.1
1.9
0.92
0.037
0.23
0.038
0.51
0.38
0.19
0.076
0.076
0.0067
0.0027
Trace
0.00026
0.019
0.013
0.056
Nil
Nil
45.0
17
12
0.55
2.8
4.5
2.4
1.4
0.059
0.31
0.037
0.91
1.2
0.072
0.091
0.094
0.0078
0.0038
Trace
0.00032
0.026
0.0084
0.062
Nil
Nil
45.6
*MECO Samples 39384-39389. Refuse material ashed at 600 C, per ASTM D271-58, to produce
samples for spectrographic assay. Total will not equal 100%, because gases do not show. Quartz, for
instance, is 28 parts silicon and 32 parts oxygen, which would show 46.7% silicon. For most ores, the
total is less than 50%.
•(•Screen fraction 3, 8-in. manometer reading, bottoms (BTM) subfraction.
tScreen fraction 3, 8-in. manometer reading, overhead, (OH) subfraction.
mary shredded sample, because easily salvageable
material of this type had already been handpicked
from the sample. The Houston material (after sec-
ondary shredding) not only contained less newsprint
and cardboard but also had no ferrous metal, since
this had been removed by magnetic separation before
the fine-shredding operation.
Because the Los Angeles samples were selected
purposely to demonstrate differences of paper re-
covery potential, they showed a variation in both
quantity and quality of paper content. In addition,
the domestic waste sample was believed to contain
less food wastes than the national average because of
the more prevalent use of garbage grinders in the Los
Angeles area. Apparently, because this sample was
collected in the late fall from a relatively high-income
suburban area, it was also characterized by a much
higher than normal percentage of garden clippings
and dried leaves. However, for the purpose of
investigating the separating effectiveness of the air
classification unit on mixed municipal refuse, the
actual composition of the waste processed was not of
paramount importance. It was important that
material that would normally be expected to be
present in the collected waste and that would
constitute a contaminant of the intended air classifi-
cation product be present in the samples classified to
determine whether satisfactory removal could be
effected. In this respect, the presence of a larger-than-
normal quantity of yard wastes in the San Fernando
Valley single-family domestic sample was of benefit
because this type of contamination is especially
detrimental to reuse of a salvaged, mixed-paper
product. The same comment pertains to the unusu-
ally large amounts of sawdust and floor sweepings in
the San Fernando Valley commercial sample.
-------
LABORATORY-SCALE UNIT EXPERIMENTS
29
TABLE 10
SEMIQUANTITATIVE SPECTROGRAPHIC ANALYSIS OF
AIR-CLASSIFIED SUBFRACTIONS OF RESIDUE AFTER MAGNETIC
SEPARATION FROM AUTOMOBILE BODY SHREDDING.
SCREENED FRACTION 4 (FINE) AUTOMOBILE BODY TRASH*
Percentage, by weight, from ash sample of subfraction
Silicon
Iron
Copper
Caicium
Alumium
Zinc
Magnesium
Chromium
Barium
Boron
Titanium
Lead
Tin
Manganese
Nickel
Molybdenum
Vanadium
Sodium
Silver
Zirconium
Cobalt
Strontium
Antimony
Other elements
Total percentage ash,
by weight, of
original sample
4, 8.0 BTMf
21
11
0.60
3.9
4.3
0.61
1.7
0.10
0.16
0.010
0.42
1.7
0.037
0.0081
0.096
0.0067
Trace
4.4
Trace
0.016
0.018
0.066
Trace
Nil
--
4, 8.0 OH*
12
46
0.63
2.9
1.9
1.6
1.8
0.048
0.17
0.0084
0.22
0.83
0.037
0.17
0.19
0.0080
Trace
2.3
0.0027
Nil
0.0078
0.058
Nil
Nil
85.2
4, 4.0 OH
20
9.1
0.029
4.6
4.5
0.42
2.1
0.14
0.30
0.013
0.58
0.64
0030
0.086
0.034
0.0060
0.0013
3.0
Trace
0.016
0.0045
0.11
Nil
Nil
80.9
4, 2.0 OH
24
16
0.037
4.5
5.2
0.61
2.2
0.065
0.43
0.022
1.1
0.96
0.035
0.086
0.31
0.0065
0.0019
3.9
Trace
0.020
0.011
0.14
Nil
Nil
72.9
4, 1.0 OH
19
11
0.032
2.5
4.5
0.30
1.5
0.047
0.44
0.015
1.2
0.65
0.053
0.083
0.061
0.0067
0.0017
2.2
Nil
0.025
0.0069
0.080
Nil
Nil
69.4
4, 0.5 OH
17
13
0.047
2.5
4.7
1.3
1.3
0.033
0.37
0.014
0.81
0.83
0.035
0.12
0.070
0.0067
0.0038
0.84
0.0013
0.021
0.0072
0.12
Nil
Nil
70.0
4, 0.2 OH
20
14
0.071
3.4
5.7
0.93
1.7
0.039
0.43
0.027
0.59
0.91
0.044
0.17
0.065
0.0091
0.0036
0.80
0.0011
0.026
0.0078
0.090
Nil
Nil
69.1
*MECO Samples 38384-39389. Refuse material ashed at 600 C, per ASTM D271-58, to produce
samples for spectrographic assay. Total will not equal 100%, because gases do not show. Quartz, for
instance, is 28 parts silicon and 32 parts oxygen, which would show 46.7% silicon. For most ores, the
total is less than 50%.
jScreen fraction 4, 8-in. manometer reading, bottoms (BTM) subfraction.
*Screen fraction 4, 8-in. manometer reading, overhead (OH) subfraction.
One factor related to refuse composition and the
effectiveness with which components can be sepa-
rated for salvage or reuse is the amount of time that
elapses between refuse collection and the separation
process. In a commercial application, refuse would
probably be separated immediately after collection.
In the Institute's laboratory experiments, this was not
physically possible. In many cases, the samples had to
be stored for a month or more while preliminary
investigations were being carried out to determine the
most advantageous detailed test program. The effect
of storage, it is believed, is to make more difficult the
separation of fine, gritty contaminants that may
embed themselves in, or become cemented to, paper
or cardboard surfaces.
Another characteristic of solid waste that is likely
to be of interest in processing studies is its calorific
value. Drying may become necessary for air classifica-
tion of certain wastes. If so, a portion of the waste or
a separated fraction most likely will be burned to
provide the heat needed. For that reason, the heating
values (2,500 to 8,500 Btu/lb) of various types of
municipal refuse are listed (Table 13).
A general problem encountered by the Institute's
research team with all paper-containing wastes was
that of overshredding. Overshredding produces an
effect similar to that for which a "Jordan" or a beater
is used in the original paper-making process—that is,
to cut and scuff the edges of each fiber so that it will
cling strongly to another fiber. The small size of the
overshredded particles reduces the air velocity re-
quired for fluidization (and thus the available separat-
ing force) and the fuzzy nature of the particles causes
the paper and cardboard to agglomerate and thus
form a floe that picks up and carries with it a great
deal of the other light and fine material. It was found
-------
30
AIR CLASSIFICATION OF SOLID WASTES
TABLE 11
CHARACTERISTICS OF MUNICIPAL REFUSE SAMPLES AS RECEIVED
FOR EXPERIMENTATION AT STANFORD RESEARCH INSTITUTE
Dry net
weight of
sample
Source and type of sample (IDj
San Fernando Valley,
California
Domestic waste, 94
single family
Domestic waste,
apartments
Commercial waste 100
Maximum
particle
dimension
(in.)
12-14
Segregated
12-14
Air -dried
bulk density
(Lb/cu ft) (Lb/cu yd)
7.0 189
sample not obtainable
7.2 195
Percentage
moisture
content
as received Remarks
3 5 More than 25% yard
trimmings and dry
leaves
20 Considerable sawdust
and other wood waste;
high percentage corru-
gated material
Houston, Texas
Domestic waste,
single family
Primary shredding
Secondary
shredding
47
70
Mostly 1 in.
or less; 8-in.
plastic strips
(length)
6.6
10.9
178
294
25
59
Slightly over-shredded
Cincinnati, Ohio
Domestic waste,
single family
Preground (Sample
79-H) 14
Final grind (Sample
80-H) 16
Commercial waste
Preground (Sample
78-H) 21
Final grind
(Sample 81-H)
25
12
1-1 1/2
12-16
1 1/2-2
149
163
14
Very heterogeneous,
but constituents
recognizable.
Semipulped
High percentage of
corrugated material.
Overshredded
*Air dried, approximately 20%.
that a major improvement resulted as the particle
sizes became larger; this was particularly the case
when the edges were not ragged and torn but were
sharply cut. Runs on mixed newspaper and corru-
gated cardboard that had been cut with a punch into
1-in. rounds revealed that sufficient differences
existed in the fluidizing velocities to permit effective
separation by air classification. In a column designed
for volume processing, the column throat and allow-
able particle size would be much larger than in the
laboratory equipment, and this circumstance would
make very effective separation possible.
When fibrous agglomerates of paper and cardboard
formed in the column (this also happened with wet
compost material but to a lesser extent), the column
immediately lost capacity, separation efficiency
dropped owing to the overload condition, and
column clogging usually resulted. An expedient for
remedying this condition was to "pulse" the column
as follows: At the point of clogging, if the air inflow
to the column was alternately stopped and started by
covering the air inlet at the bottom of the column,
sufficient additional agitation could often be pro-
duced to break up the floe and to maintain effective
column action. Done rapidly, this did not affect
separation. Occasional purging at higher velocity was
sometimes also necessary. The obvious solution to the
problem of column clogging in larger scale operations
-------
LABORATORY-SCALE UNIT EXPERIMENTS
31
TABLE 12
EXPECTED RANGES IN COMPOSITION
OF MIXED MUNICIPAL REFUSE*
Percentage composition
as received
(dry weight basis)
Component
Paper
Newsprint
Cardboard
Other
Metallics
Ferrous
Nonferrous
Food
Yard (leaves, etc.)
Wood
Glass
Plastic
Miscellaneous
Total
Anticipated range
37-60
7-15
4-18
26-37
7-10
6-8
1-2
12-18
4-10
1-4
6-12
1-3
<5
Nominal
55
12
11
32
9
7.5
1.5
14
5
4
9
1
3
100
*Moisture content: range, 20%-40%; nominal, 30%. Based
on data contained in references 7 through 13. Source:
Personal communication, Battelle Memorial Institute.
would be to shred the material in a way that reduces
the tendency for agglomeration, that is, to cut rather
than tear. The problem, of course, would be much
less important in a larger sized column, if not totally
eliminated.
Fluidizing velocities observed when the laboratory
column was operated on selected pure components of
refuse mixtures are summarized (Table 14). In this
tabulation of experimental data, an attempt was
made to show (for fibrous materials) the effects of
the size of a particle and the sharpness of the edges of
the shredded material. When fluidizing velocities for
paper and cardboard in a 6-in. straight pipe are
compared with those in the zigzag column, the
straight-pipe velocities are approximately one-third
lower than those in the column for the same material
(dry, shredded newspaper and cardboard and dry, cut
newspaper and corrugated cardboard rounds). Shred-
ded and cut newspaper appear to differ little in
fluidizing velocities in either the column or the
straight pipe, and there is no observable difference in
straight-pipe velocities for large- or small-cut paper
specimens. On the other hand, cut cardboard requires
a one-fourth to one-third greater velocity than the
shredded material, and large squares require about 50
percent greater velocity for suspension than small
rounds do. This is of considerable significance for
separation; whereas there exists a 250 ft/min in
velocity difference between cardboard and newspaper
(about 50 percent of the lower fluidizing velocity)
when finely shredded, a difference of 650 ft/min in
velocity exists between large pieces (almost 200
percent of the lower fluidizing velocity) when they
are cut to minimize the aerodynamic effect of torn
edges. The greater air velocities would also permit
greater particle agitation for removal of entrapped
fine particles.
The performance of the laboratory unit on the
eight samples of paper-containing wastes essentially
involved only two test programs. In a number of
respects, these two programs were not even mutually
exclusive. The first program dealt with single-family
domestic wastes from various sources. The second
program dealt with domestic wastes from multiple-
family dwellings and with commercial wastes. Re-
covery of a commercially usable grade of waste paper
stock was the objective of both programs. Addition-
ally, the first program was also directed toward
removal of nonbiodegradable organic and inorganic
material that interfered with composting, fermenta-
tion, and retorting from the usable cellulosic fraction.
It was considered desirable to effect removal, if
possible, so that the metal, glass, and plastic con-
taminants might also have a certain salvage value.
Evaluation of Laboratory-Scale Unit's Perfor-
mance in Separating Solid Wastes. From the stand-
point of paper recovery, results of classification
experiments with the laboratory-scale unit were
inconclusive. Accomplishment of the subobjectives of
processing paper-containing waste—metal and glass
removal, removal of nonbiodegradable material, and
pretreatment for retorting—was easily demonstrated
and would present no problem in commercial opera-
tion. In developing guidelines for the production of a
usable recovered paper product, it was realized that
recovery of secondary fiber for reuse does not require
separation into the paper stock grades that are now
offered in the salvage paper market. These grades are
dictated more by paper salvage practices, such as
segregation by source, and the opportunities for
economical hand selection than by users' specifi-
cations. To be of maximum value, however, a recover-
ed product should take full advantage of the potential
that air classification offers for removing all
noncellulosic fines from the collected material. Gritty
contaminants are much more objectionable in paper
making than water-soluble stains that the waste
paper may have acquired as an ingredient of mixed
refuse.
In tha laboratory column, the degree of cleaning
and separation considered suitable to permit paper
salvage from mixed waste for reuse in the paper-
making process was not achieved. Experience with
-------
32
AIR CLASSIFICATION OF SOLID WASTES
TABLE 13
HEATING VALUES OF VARIOUS TYPES OF MUNICIPAL REFUSE*
Source
Commercial and
light industrial
establishments
Principal
components
Highly combustible
waste such as
paper, wood,
cardboard
cartons,
plastic, or
rubber scraps
Percentage
Approximate Percentage incom-
comtJosition moisture bustible
(% by weight) content solids
100% trash 10 5
Heating
value Auxiliary Recommended
of refuse, fuel minimum burner
as fired (Btu/lb input (Btu/lb
(Btu/lb) of waste) of waste)
8,500 0 0
Combined collec- Combustible waste
tion from such as paper,
domestic, cartons, rags,
commercial, wood scraps,
and industrial combustible
„ . 80% trash
sources floor sweepings,
with some
putrescible cook-
ing residues and
food wastes
Combined collec- Trash, garbage, and 50% trash
tion domestic garden clippings 50% vegetable
waste from matter
residences only
Markets; restaur- Food wastes,
25
10
6,500
50
4,300
1,500
ant, hotel,
club, and
institutional
kitchens
including ani-
mal, fruit, and
vegetable
residues from
preparation and 35% trash
cooking of
foods
70
2,500
1,500
3,000
*A classification of wastes based on satisfactory incinerator operation is given in paragraph 2.1, section 3, of "Code of
Recommended Practices for Non-Domestic Incinerators," a supplement to the City of Chicago's Air Pollution, Control
Ordinance as revised October 7, 1968. Adapted from Essenhigh, R. H. Incineration-a practical and scientific approach,
Environmental Science and Technology, 2 (7): 530, July 1968.
the column provides, however, some guidance for
future efforts in this area. Difficulties were due to the
small particle size required by the laboratory classifi-
cation column and the high aerodynamic drag of the
finely shredded waste material. Thus, only very low
fluidizing velocities and correspondingly weak sepa-
rating forces (gravity and air velocity) resulted. Under
these conditions, electrostatic forces also have a
powerful influence. (To permit viewing, the sidewalls
of the laboratory column are made of plexiglass,
which is a good dielectric material and contributes to
electrostatic interference.)
For removal of noncellulosic fines from dry waste
samples, a combination of screening and air classifi-
cation appeared to be effective. Screening was neces-
sary because of the similarity in aerodynamic char-
acteristics of the fine dust and grit and the paper
constituents of the samples. Air classification of the
coarser, paper-containing fractions after screening was
difficult because of the small differences in density of
the components (remaining fines, paper, cardboard,
and plastic) and the intimate mixture produced by
the shredding and screening operations. The over-
shredded material from a conventional hammermill
tends to agglomerate, forming a floe of paper and
cardboard that picks up and carries with it a great
deal of other light and fine material.
Equipment for shredding and screening in
commercial applications must be selected that mini-
mizes the dry pulping and felting effects of these
operations on a fibrous material. In a column
designed for volume processing, the column throat
and allowable particle size would be much larger than
in the laboratory equipment, arid these conditions
would make more effective separation possible.
Performance of the air classification column on
-------
LABORATORY-SCALE UNIT EXPERIMENTS
33
TABLE 14
FLUIDIZING VELOCITIES FOR SELECTED PURE
COMPONENTS OF REFUSE MIXTURES IN A STRAIGHT
PIPE AND THE ZIGZAG COLUMN
Velocity (ft/min)
Component
Zigzag classifier
with 2-in. throat
Straight
6-m. diameter pipe
Plastic wrapping Less than 400
(shirt bags) (electrostatic)
Dry, shredded newspaper 400-500 350
(25% moisture)
Dry, cut newspaper
1-in. rounds 500 350
3-in. squares -- 350
Agglomerates of dry, shredded
newspaper and cardboard 600
Moist, shredded newspaper 750
(35% moisture)
Dry, shredded corrugated 700-750 450-500
cardboard
Dry, cut corrugated cardboard
1-m. rounds 980 700
3-m. squares -- 1,000
Styrofoam packing material 750-1,000
(electrostatic)
Foam rubber
(1/2-in. squares) 2,200
Ground glass, metal, and stone 2 500-3 000
fragments (from automobile
body trash stream)
Solid rubber 3,500
(1/2-in. squares)
compost and automobile body trash, which is more
granular than paper-type trash, was very satisfactory.
Empirically obtained operating data appear sufficient
to permit scale-up for the design and construction of
a full-size unit, if desired. Process flow diagrams are
presented and performance of typical commercial
plants are discussed in the last part of this report.
Minor variations of the preliminary designs developed
from the current set of experimental data will be
desirable, inasmuch as the objective of this research
was simply to demonstrate the technical feasibility of
air classification. Additional experimental data should
be obtained before actual plants are designed or built
to optimize the desired nonferrous metal concen-
trations for automobile body trash and to produce a
quantity and quality of compost that satisfies both
the compost marketing and refuse disposal require-
ments of a compost plant operator.
It can be concluded from these bench-scale experi-
ments that air classification is technically feasible for
processing semifibrous solid waste materials. For
light, more fibrous materials, the results were favor-
able but inconclusive. Additional research is necessary
to demonstrate a workable process for paper stock
recovery from combined collections of municipal and
commercial wastes. This work must include experi-
ments with a larger sized air classification column and
investigations into the performance of commercially
available shredding and screening equipment on feeds
with large paper contents. Improvement of commer-
cially available feed preparation equipment (shredders
and screens) may be necessary for use in an air
classification system.
-------
-------
Pilot-unit Experiments
Need for Pilot-Scale Experiments for
Recovery of Municipal Refuse
The degree of cleaning and separation considered
suitable to permit paper salvage from mixed waste for
reuse as secondary fiber was not achieved in the
laboratory column. Experience with the small column
provided, however, some guidance for future efforts.
Difficulties were due to the small particle size
required by the laboratory classification unit and the
high aerodynamic drag of finely shredded waste
material. Thus, only very low fluidizing velocities and
correspondingly weak separating forces (gravity and
air velocity) resulted. The overshredded material from
a conventional hammermill tended to agglomerate,
forming a floe of paper and cardboard that picked up
and carried with it other light and fine material.
Additional experimentation was, therefore,
recommended to indicate how air classification might
be employed in a workable process for paper stock
recovery from combined collections of municipal and
commercial wastes. This work included experiments
with a larger sized air classification column and
investigations into the performance of commercially
available shredding equipment on high-paper-content
feeds.
Pilot-Scale Air Classification Unit
The pilot air classification unit designed and built
especially for the recommended paper separation
experiments is shown (Figure 8). It is a 10-stage
column with viewports and a two-stage column
section that can be used either at the top or the
bottom of the main column. (This effectively permits
feeding at alternate positions on the column.) The
column throat is 6 by 12 in. in cross section; thus,
shredded refuse can be handled in which maximum
particle size of the paper fractions is somewhat more
than 4 in. In developing design criteria for the
column, this throat size was selected as the minimum
practical size for separating refuse shredded in the
same way it would be shredded for a commercial
operation. Column appurtenances were provided that
perform the same functions as those on the labora-
tory-scale unit described previously—induction
blower; feed hopper with rotary, airlock feeder; a
cyclone for separating overhead material from the
airstream; and a manometer indicating pressure drop
across the column. In addition, the 23-in. diameter
Carter-Day high-velocity cyclone was equipped with
an 8-in. rotary discharge valve that made possible the
continuous operation of the unit, and a 12-in.
portable belt elevator was provided for uniform
delivery of material to the feed hopper. The 5-hp
induction blower with slide-gate control permitted air
velocities in the column throat to be varied between
300 and 2,500 fpm.
In all work reported herein, the column was
operated with six stages above and four below the
feed point.
Procurement of Shredded Samples and
Their Characteristics
From experience in obtaining samples of material
for laboratory-scale separations, it was believed that
shredded municipal refuse for the pilot unit could be
obtained locally (to eliminate shipping charge) and
without cost for shredding. This was not true. It was
found more difficult to obtain samples several cu yd
in size and sufficient for continuous operation of the
new column than it was to obtain the smaller samples
(generally less than 50 Ib or 0.1 to 0.3 cu yd) that
were donated for batch operation of the small
laboratory unit. It was originally estimated that a
150- to 200-lb sample would be required for each run
of the large column and that reuse of samples was
possible but that physical degradation of the samples
by continuous handling and rehandling would be
limiting. This also was not true. In practice 50- to
100-lb, or smaller, samples proved to be adequate,
and the separated component could be remixed and
rerun five or more times. Losses of material during
feeding became a more significant limitation than
physical degradation of the samples.
To fulfill the objectives of the proposed research it
was necessary to obtain shredded samples of munici-
pal solid waste that were representative of actual
refuse and that exhibited the characteristic size
reduction patterns produced by several types of
commercial shredding equipment.
Procurement of Samples. Initially, only two sam-
ples were available. These were the following.
Los Angeles Sample. A rigid-arm, 75 hp Williams
shredder for shredding corrugated boxboard in a
baling installation at a major supermarket warehouse
was used to shred a synthetic mixed-paper sample to
a nominal 4- to 5-in. particle size. Corrugated board
(15 to 20 percent by weight) was supplied by the
warehouse operator; newspaper (30 to 35 percent by
35
-------
36
AIR CLASSIFICATION OF SOLID WASTES
Figure 8. Scientific Separators pilot air classification unit.
-------
PILOT-UNIT EXPERIMENTS
37
weight) was supplied by SRI; miscellaneous paper (50
to 55 percent by weight) was handpicked at a local
landfill. The combined mixed-paper sample, together
with a small amount of film plastic, was shredded in
the Williams mill. Refuse constituents other than
paper were obtained also by picking a representative
sample at a landfill. The County Sanitation Districts
of Los Angeles, having a small experimental shredding
plant at a landfill, shredded the heavy fraction. Heavy
and light material were to be blended after shredding
for runs in the air classification pilot-unit column.
Cincinnati Sample. A 450-lb sample of shredded
municipal refuse was shipped by motor freight from
the Cincinnati laboratory of BSWM. Ground in one
pass (no screen or bar grate) by the laboratory's
100-hp Williams hammermill, the particle size of the
paper and cardboard proved too large for separation
in the air classifier. It became necessary, therefore, to
hand separate paper from other refuse material for
further size reduction.
Quantities of both samples were adequate, but the
large maximum particle size of paper and cardboard
made them unsuitable for continuous column opera-
tion. An attempt was made to reshred these samples
in a forage chopper that employed a rotating cutter-
head similar to a lawn mower, but this was unsuccess-
ful. Other shearing devices that would produce the
desired cleanly cut edges on paper particles, such as
commercial brush chippers, were investigated; none
of these was suitable.
Characteristics of Samples. Five different samples
of shredded refuse representative of five different
types of shredding equipment were eventually investi-
gated. Two of these samples were preponderantly of
the largest particle size (4- to 6-in. maximum dimen-
sion) that could be processed in the pilot unit, and
three samples were generally of considerably smaller
particle size. The fine-particle-size samples were pro-
vided by the PHS-TVA composting project and by
the Eidal International Corporation, the refuse being
obtained by them from Johnson City, Tennessee, and
Albuquerque, New Mexico, respectively.
Data on the shredded refuse samples used for
experimental purposes are summarized (Table 15).
Size classification by components is shown for each
sample (Tables 16-20). The size gradations are shown
graphically (Figure 9). The Cincinnati and Los
Angeles samples that were indicated as "coarse"
material were really dissimilar. The former was well
graded, having only 20 percent of the total sample
weight in the 1/2- to 1-1/2-in. size range. The latter
had more than 80 percent in this range and contained
essentially no fines. Like the Cincinnati sample, the
Albuquerque sample was well graded; only 45 percent
was in the 1/2- to 1-1/2-in. size range, with 45
percent finer than and 20 percent coarser than the
range. The Johnson City hammermill and rasp
samples were very similar. Both contained approxi-
mately one-third (by weight) of material finer than
the screened range; the balance was in this range with
little or no material in excess of the 1-1/2-in. size.
The screen sizes selected for sieve analysis corre-
sponded with those used by the Forest Products
Laboratory of the U.S. Department of Agriculture in
preliminary work on dry separation of paper com-
ponents from shredded municipal waste from
Madison, Wisconsin. The material categories were
those used by Black-Clawson Company Research
Center and the Forest Product Laboratory in the
work they are doing to reclaim usable grades of
secondary paper fiber from municipal refuse.
TABLE 15
SHREDDING INFORMATION AND SOURCE OF MUNICIPAL REFUSE SAMPLES
PROCESSED IN THE PILOT AIR CLASSIFICATION UNIT
Source
Supplier
Manufacturer and
tvne of shredder
Particle
size
Remarks
Johnson City, Tennessee
Hammermill
Rasp
Los Angeles, California
(Scholl Canyon Landfill)
Cincinnati, Ohio
PHS-TVA composting project
PHS-TVA composting project
SRI and L.A. County Sanitation
Bureau of Solid Waste Management
Albuquerque, New Mexico Eidal International Corp.
Gruendler, Model 48-4 Fine
(250 hp)
Don-Oliver (80 hp) Fine
Williams, rigid-arm Coarse
paper shredder (75hp)
Williams, swinging-arm Coarse
refuse shredder
(100 hp)
Eidal Model 400, coarse Fine
grind (400 hp)
Shredded wet
No cans or bottles
Shredded wet
-------
38
AIR CLASSIFICATION OF SOLID WASTES
TABLE 16
DESCRIPTION OF SHREDDED MUNICIPAL SOLID WASTE, JOHNSON CITY, TENNESSEE,
PHS-TVA COMPOSTING PROJECT, SHREDDED BY GRUENDLER SWING HAMMERMILL, MODEL 484
Material
Total,** general appearance
Group I (includes metals,
glass, and dirt)
Group II (film plastic)
Group II, other (includes
food waste, heavy plastic,
yard waste, cloth, wood,
and lint)
Group III, general appearance
Newspaper
Magazines
Brown paper bags,
corrugated containers, etc
Miscellaneous paper, food
wrappers, and cartons
Description of material in size range indicated
More than 1-1/2 in.*
1/2 to 1-1/2 in.f
Mostly paper, generally similar to rasper material
but less stringy and twisted
Negligiblef f
Significant amount
but very small
compared with
paper
Negligible
Negligiblef f
Significant amount,
but very small
compared with
paper
Negligible
Less than 1/2 in.*
Approximately same as
rasper material
Significant amount of
dirt and very fine glass*
Trace only
Probably most of this
fraction. Much of it
fine, unidentified
fibrous material
Approximately same as rasper material but paper fragments less twisted and rumpled
Significant but less
than in middle cut
Significant amount
Significant amount
Possibly half of
total cut
Possibly half of
total cut
Significant amount
Significant amount
Significant amount
Lots of finely shredded
paper, not identifiable
as to kind
Lots of finely shredded
paper, not identifiable
as to kind
Lots of finely shredded
paper, not identifiable
as to kind
Lots of finely shredded
paper, not identifiable
as to kind
*By weight 4.6 percent by volume negligible.
fBy weight 55.5%; 74% by volume.
tBy weight 39.9%; 26% by volume.
**Moisture content, 20%-25% (dry basis); weight screened 21.7 Ib; volume, 2.75 cu ft; bulk density, 8.0 Ib/cu ft.
ffFerrous metal apparently removed magnetically before shredding.
-------
PILOT-UNIT EXPERIMENTS
39
TABLE 17
DESCRIPTION OF SHREDDED MUNICIPAL SOLID WASTE, JOHNSON CITY, TENNESSEE,
PHS-TVA COMPOSTING PROJECT SHREDDED BY DORR-OLIVER RASP
Description of material in size range indicated
Material
More than 1-1/2 in.*
1/2 to 1-1/2 m.t
Less than 1/2 in. t
Total,** general appearance Very small amount;
rumpled, dirty
Group I (includes metals,
glass, and dirt)
Group II (film plastic)
Group II, other (includes
food waste, heavy
plastic, yard waste,
cloth, and wood)
None visiblef f
Significant amount, but
only small fraction of
total
Trace only (several
large pieces of
heavy plastic)
Finely and uniformly
shredded, with stringy,
twisted appearance
Trace only of metals and
glassff (glass pieces
to 3/4 in.)
Significant amount but
only small fraction of
total
Trace only (some leaves
evident)
Group III, general appearance Probably more than 90% of the fractions consists of assorted
paper 1/2 to 1-1/2 in. size
Mostly newsprint, but
much unidentifiable
Newspaper
Probably more than any other single type of paper
Dense, dark-colored,
fibrous matter, much of
it unidentifiable
Trace only of metals and
glassft (significant amount
of fine dirt and glass
to 1/8 in.)
Trace only
More than 50% of
fraction probably in
this category
Much of fibrous mate-
rial finely shredded
paper
At least 1/4 of this
cut finely shredded
paper but not
classifiable as to kind
Magazines
Brown paper bags,
corrugated containers, etc
Miscellaneous paper, food
wrappers, cartons (waxed),
kleenex, and other very
light paper
Negligible
Negligible
At least half of
total paper content
Negligible
Negligible
At least half of
total paper content
At least 1/2 of this cut
finely shredded paper
but not classified as to
kind
At least 1/2 of this cut
finely shredded paper
but not classifiable as
to kind
At least 1/4 of this cut
finely shredded paper
but not classified as
to kind
*By weight 2.0%; percent by volume negligible.
fBy weight 65.0%; 73% by volume.
tBy weight 33.0%; 27% by volume.
**Moisture content, 45% (dry basis); weight screened 30.0 Ib; volume, 2.75 cu ft; bulk density, 10.9 Ib/cu ft.
t|Ferrous metal apparently removed magnetically before shredding.
-------
40
AIR CLASSIFICATION OF SOLID WASTES
TABLE 18
DESCRIPTION OF SHREDDED MUNICIPAL SOLID WASTE, LOS ANGELES, CALIFORNIA,
SCHOLL CANYON LANDFILL
SHREDDED BY WILLIAMS RIGID-ARM PAPER SHREDDER
Material
More than 1-1/2 in.*
Description of material in size range indicated
1/2 to 1-1/2 in.f
Less than 1/2 in.*
Total,** general appearance
Group I (includes metals,
glass, and dirt)
Group II (film plastic)
Group II, other (includes
food waste, heavy plastic,
yard waste, cloth, and
wood)
Group III, general appearance
Newspaper
Magazines
Brown paper bags,
corrugated containers, etc
Miscellaneous paper
Miscellaneous paper, fairly uniform in size; httle evidence
of crumpling; all very clean
None
None
None
None
None
None
Entire sample was made up of paper and cardboard,
chopped to size with fairly minimal fraying;
very few pieces more than 5 in. in largest
dimension.
20% in this size range
None
20% in this size range
Present
80% in this size range
None
80% in this size range
Present
Trace of dust and some
glass, resembling
floor sweepings
Trace
None
None
None
None
None
None
None
*By weight 17.0%.
fBy weight 83.0%.
tBy weight 0%.
**Moisture content, approximately 10%; synthetic dump stock composed of: 1 part (15%-20% by weight) corrugated
cardboard; 2 parts (30%-35% by weight) newspaper; 3 parts (50%-55% by weight) miscellaneous paper hand picked from
refuse; weight screened, 3.90 Ib; original shredded sample contained significant quantity of paper requiring hand tearing before
material could be fed to air classifier.
Performance of the Pilot-Scale Unit on
High-Paper-Content Feeds
As mentioned previously, only limited experi-
mentation was possible with the 6- by 12-in.-throat-
size pilot unit because of feeding limitations and
difficulty in obtaining shredded municipal refuse
samples. A further constraint was the lack of pro-
cedures for analyzing the recovered paper fractions.
The scope of the project did not include development
of these procedures; consequently, results of the
degree of separation obtainable are reported primarily
as descriptions of the materials contained in the
various fractions. Some work toward quantitative
analysis was done by a commercial client, the St.
Regis Paper Company Technical Center, on air-
classified fractions of municipal refuse produced
under conditions identical to those reported herein.
Their quantitative results are considered proprietary,
but qualitative generalizations are reported where this
information has been released to us.
Operating Procedure. No operation other than air
classification was employed in making separations of
the five commercially shredded municipal refuse
samples described previously. The operating proce-
dure consisted of conducting a number of runs with
30-to 60-Ib raw feed samples, and fractions
separated therefrom, at constant superficial velocities
in, and uniform feed rate to, the air classification
column. The purpose was to determine the following
data for each shredding method:
Velocity required to effect potentially useful
separations
Weight percentage yields of the separated frac-
-------
PILOT-UNIT EXPERIMENTS
41
TABLE 19
DESCRIPTION OF SHREDDED MUNICIPAL SOLID WASTE, CINCINNATI, OHIO,
BSWM LABORATORY
SHREDDED BY WILLIAMS HAMMERMILL*
Description of material in size range indicated
Material
More than 1-1/2 in.f
1/2 to 1-1/2 in. i
Less than 1/2 in.**
Total.tt general appearance Overwhelmingly paper and fairly clean
Group I (includes metals,
glass, and dirt)
Group II (film plastic)
Group II, other (includes
food waste, heavy plastic,
yard waste, cloth, and
wood)
Group III, general appearance
Newspaper
Magazines
Brown paper bags, corrugated
containers, etc.
Miscellaneous paper
Trace only (several
battered tin cans)
Very little
Very little
Trace only
Very little
Very little
Very little evidence of twisting, rumpling, or fraying
of pieces; many very large pieces, more than 5 in.
in largest dimension
About 1/2 of total
fraction
Present
Present
Present
About 1/2 of total
fraction
Present
Present
Present
Coarser than correspond-
ing fraction from
other samples screened
Mostly dirt; some
glass very finely
ground
Negligible
Wood especially present;
other fine, unidenti-
fied fibrous material
Possibly more paper
than in corresponding
fraction from other
samples
Some paper present but
undifferentiated
None
None
None
*Close breaker bar setting; no grate.
|By weight 47.5%.
*By weight 24.2%.
**By weight 28.3%.
ffMoisture content, 10%-15%; weight screened, 13.3 Ib: original shredded sample contained significant quantity of paper
requiring hand tearing before material could be fed to air classifier.
tions, for process design material balance calcula-
tions
Maximum allowable feed rate for column throat
size calculations
Combined influence of feed rate and air velocity
on degree of separation obtainable
Results from the five different methods of shredding
were compared and overall results evaluated from the
standpoint of future process design.
The general procedure employed was to make two
runs on a sample to recover a paper-rich middle
fraction: a deducting run to separate light fines and
film plastic and a run to separate heavy constituents
such as metal, rocks, glass, rubber, heavy plastic, and
wood. These separations were followed by experi-
ments to split the paper fraction into chemically and
mechanically pulped material, corresponding roughly
to (1) containers, which would include cardboard
boxes, grocery bags, milk cartons, and similar items,
and (2) newsprint.
At low velocities the column loading, that is,
pounds of solids per pound of air, had little influence
on the separations produced at low feed rates.
Consequently, runs to establish separating velocities
for splitting the paper fractions were made initially at
low feed rates. As the feed rate was increased, the
sharpness of separation decreased until the column
operation became unstable and the column became
choked because of overloading. The maximum allow-
able feed rate for full-size column design, therefore,
will be selected in practice to produce column
loadings between these two limiting values. The
number of experimental runs possible for determining
feed rates was, however, insufficient in this series of
experiments, to establish limiting column loadings
with any degree of reliability. Assessment of potenti-
-------
42
AIR CLASSIFICATION OF SOLID WASTES
ally useful separations of the paper fraction probably
can be quantified best by producing and strength
testing a hand sample of paper from the recovered
material. Paper industry standards cover the necessary
laboratory procedures, since this type of testing is
used routinely to check mill "furnishes," that is,
blends of paper-making ingredients.
It is realized that the strength properties of paper
produced from recycled material are subject to
degradation both by contamination with foreign
material in refuse and by biological action. The
degree of biological degradation depends on the
activity of compost-type organisms, moisture
content, and time. In laboratory experiments, the
time between collection of a refuse sample and pro-
duction of a test sheet of paper from suitable recover-
ed fractions is considerably longer than the corre-
sponding production cycle would be if the process
were commercialized. To minimize degradation in
samples provided to St. Regis Paper, a sterilant was
added to the refuse samples when received and again
when separated fractions were sent to St. Regis for
TABLE 20
DESCRIPTION OF SHREDDED MUNICIPAL SOLID WASTE, ALBUQUERQUE, NEW MEXICO,
EIDAL INTERNATIONAL CORP.
SHREDDED BY EIDAL MODEL 400, COARSE GRIND*
Description of material in size range indicated
Material
More than 1-1/2 rn.f
1/2 to 1-1/2 in.*
Less than 1/2 in.**
Total.ft general appearance
Group 1 (includes metals, glass,
and dirt)
Group II (film plastic)
Group II, other (includes food
waste, heavy plastic, yard
waste, cloth, and wood)
Group III, general appearance
Newspaper
Magazines
Brown paper bags, corrugated
containers, etc
Miscellaneous paper
Primarily paper; moderately twisted and
rumpled; 1/2 to 1-1/2 in. same appeared
finer than middle cut from other samples
Negligible
Negligible
Small, but significant amounts
Negligible Present
Very conspicuous
difference between
coarse and medium
cuts
Fragments look a lot
more twisted and
rumpled than those
in coarse cut
All kinds of paper present in about equal amounts
All kinds of paper present in about equal amounts
All kinds of paper present in about equal amounts
All kinds of paper present in about equal amounts
Very fine and uniform
Present (mostly dirt)
Negligible
Much unidentified
fibrous material;
some wood definitely
present
Undifferentiated, finely
shredded paper
probably about 1/2 of
this fraction
Undifferentiated, finely
shredded paper
probably about 1/2 of
this fraction
Undifferentiated, finely
shredded paper,
probably about 1/2
of this fraction
Undifferentiated, finely
shredded paper
probably about 1/2
of this fraction
*Shredded with water spray.
fBy weight 18.8%.
*By weight 35.9%.
**By weight 45.3%.
ftMoisture content, 8%-13%;
weight screened, 2.7 Ib (sample too small for accurate volumetric measurement).
-------
PILOT-UNIT EXPERIMENTS
43
UJ
CC
o
z
in
2
CD 40
z
UJ
o
o:
JOHNSON CITY-
HAMMERMILL
RECTANGULAR WIRE SCREEN MESH SIZE
Figure 9. Mechanical analysis diagrams representative of various methods of shredding municipal
solid waste. See Table 15 for manufacturer and type of shredding used. See Tables 16-20 for
additional detail on refuse samples and gradation by components.
-------
44
AIR CLASSIFICATION OF SOLID WASTES
TABLE 21
AIR CLASSIFICATIONS CONDUCTED ON JOHNSON CITY (HAMMERMILL) REFUSE
Bulk
Weight density Column's
of feed of feed ^ Feed
sample sample velocity rate
Run Number Feed (Ib) (Ib/cu ft) (fpm) Qb/mir
3/11-1*
3/11-2
3/11-3*
3/11-4
3/12-1
3/12-2
3/12-3
4/8-19(1)
4/8-19(2)
4/8-19(3)
Raw refuse 46.0 8.0 2,000
(air dried)
Lights from 40.0 -- 500
Run 3/1 1-1
Raw refuse 64.0 -- 2,000
(as received)
Lights from 61.0 -- 400
Run 3/1 1-3
Heavies from 22.0 -- 1,100-
Run3/ll-4 1,200
Lights from 18.5 -- 800
Run 3/12-1
Lights from 13.0 -- 650
Run 3/12-2
Wetted raw 25.8 -- 1,800
refuse
Lights from 24.4 -- 1,000
Run 4/8-19(1)
Lights from 18.0 -- 600
Run 4/8-19(2)
Light overhead fraction Heavy bottom fraction
Sample Sample
taken for taken for
St. Regis St. Regis
Percent of feed analysis Percent of feed analysis
By By of paper By By of paper
i) weight volume contentf weight volume contentf
95.0 -- -- 5.0
7.7 -- -- 92.3
95.0 -- -- 5.0
3.0 -- -- 97.0
86.0 96 -- 14.0 4 HM-A
84.0 82.5 -- 16.0 17.5 HM-B
8.0 27 HM-D 92.0 73 HM-C
93.0 -- - 7.0 -- 19-1H
74.0 -- -• 26.0 -- 19-2H
17.0 -- 19-3L 83.0 -- 19-3H
*Moisture contents: Run 3/11-1, 20%-25%.
Run 3/11-3, 35%-40%.
t Designations in column assigned by St. Regis.
analysis. The sterilant selected was a commercial
grade of chloripicrin (nitro chloroform) a lachryma-
tor sold under the trade name of Larvacide 100 by
the Morton Chemical Company. It is a heavy liquid
that vaporizes readily when sprinkled into a barrel or
plastic bag containing a sample of refuse. Its toxicity
is intermediate between chlorine and phosphine. In
agricultural practice, it is used as a soil sterilant and a
fumigant for grain.
Discussion of Results. The 45 air classification
runs made on the five samples of shredded municipal
refuse are summarized (Tables 21 through 25). The
most nearly comprehensive series of runs is that made
on Cincinnati refuse (Runs 4/10-1 through 4/10-5).
Two other series that could be considered typical in
that they might be adapted to commercial processing
are those designated Runs 4/9-1 through 4/9-3 made
on air-dried Albuquerque refuse and Runs 4/94
through 4/9-6 made on wetted raw refuse from
Albuquerque. These runs will be described to illu-
strate the sorting procedure and to indicate the
separations achieved. The 4/10 series of runs started
with a dedusting run at a velocity of 300 to 400 fpm
(Table 24). This classification removed overhead a
small quantity of very light fines, dust balls, and a
major portion of the film plastic contained in the raw
feed sample. The next separation, in which the feed
consisted of the column bottom cut from the 300-to
400-fpm run, was made at 700 to 800 fpm. This
overhead fraction was very clean, containing only
small amounts of fine material, and was predomi-
nantly news and magazine stock, that is, groundwood
papers. There was a somewhat lesser amount of
-------
PILOT-UNIT EXPERIMENTS
TABLE 22
AIR CLASSIFICATIONS CONDUCTED ON JOHNSON CITY (RASP) REFUSE
45
Run Number
3/11-1*
3/11-2
3/11-3
3/11-4
3/11-5*
3/11-6
3/12-1
3/12-2
Feed
Raw refuse
(air dried)
Raw refuse
(air dried)
Raw refuse
(air dried)
Lights from
Run 3/11-2
Raw refuse
(as received)
Lights from
Run 3/1 1-5
Heavies from
Run 3/1 1-4
Lights from
Run 3/12-1
Bulk
Weight density
of feed of feed
sample sample
(Ib) (lb/cu ft)
10.9
56.4 10.9
56.0 10.9
45.0
91.0
76.0
15.0
10.0
Column's
air
velocity
(fpm)
2,500
2,000
1,500
500
2,000
500
1,000-
1,200
800
Light overhead
Feed Percent of feed
rate By By
(Ib/min) weight volume
fraction
Sample
taken for
St. Regis
analysis
of paper
contentf
Heavy bottom
Percent of feed
By By
weight volume
fraction
Sample
taken for
St. Regis
analysis
of paper
contentf
Not satisfactory
87.5
15.0 80.4
14.0
87.4
3.0
65-70 80
56.0 62
--
--
--
--
--
--
Rasp-C
12.5
19.6
86.0
12.6
97.0
30-35 20
44.0 38
--
--
--
--
--
Rasp-A
Rasp-B
*Moisture contents: Run 3/11-1, 45%.
Run3/ll-5,50%-55%.
fDesignations in column assigned by St. Regis.
TABLE 23
AIR CLASSIFICATIONS CONDUCTED ON LOS ANGELES REFUSE
Run Number
Feed
Bulk
Weight density Column's
of feed of feed air Feed
sample sample velocity rate
Light overhead fraction Heavy bottom fraction
Sample Sample
taken for taken for
Percent St. Regis Percent St. Regis
of feed Bulk analysis of feed, Bulk analysis
by density of paper by density of paper
(Ib) (lb/cu ft) (fpm) (Ib/min) weight (lb/cu ft) content* weight (lb/cu ft) content*
4/6-2
4/6-3
4/6-4
4/6-5
Synthetic
dump stock
Synthetic
dump stock
Synthetic
dump stock
Synthetic
dump stock
18.0 2.0
18.0 2.0
8.0 2.0
22.0
600
900
1,100
800-
900
0.5 55.0
0.7 67.0
1.4 65.0
1.0 56.0
2-L 45.0
3-L 33.0
35.0
5-L 46.0
2-H
3-H
--
5-H
*Designations assigned by St. Regis.
-------
46
AIR CLASSIFICATION OF SOLID WASTES
TABLE 24
AIR CLASSIFICATIONS CONDUCTED ON CINCINNATI REFUSE
Run Number
4/8-9
4/8-11
4/8-12
4/8-13
4/8-16
4/8-17
4/8-18
4/10-1
4/10-2
4/10-3
4/ltM
4/10-5
Feed
Raw refuse
Raw refuse
Heavies from
Run 4/8-11
Lights from
Run 4/8-12
Wetted raw
refuse
Heavies from
Run 4/8-16
Heavies from
Run 4/8-17
Raw refuse
Heavies from
Run 4/10-1
Heavies from
Run 4/10-2
Heavies from
Run 4/10-3
Heavies from
Run 4/10-4
*Designations assigned by
Bulk
Weight density
of feed of feed
sample sample
(lb) (Ib/cu ft)
12.0
12.0
7.0 5.3
4.5
11.0
7.9
3.5
28.5
28.0
20.25 --
17.5
14.5
St. Regis.
Light overhead fraction Heavy bottom fraction
Sample Sample
taken for taken for
Column's Percent St. Regis Percent St. Regis
air Feed of feed, Bulk analysis of feed. Bulk analysis
velocity
(fpm)
700
500-
600
1,800
800
500-
600
1,000
1,800
300-
400
700-
800
900-
1,000
1,100-
1,200
1,200-
1,300
rate by
(Ib/min) weight
2.0 43.5
38.0
75.0
39.0
25.0
56.0
29.0
2.0
28.0
14.0
17.0
3.0
density of paper by
(Ib/cu ft) content* weight
9-L 56.5
1.6 -- 62.0
25:0
13-L 61.0
16-L 75.0
17-L 44.0
18-L 71.0
A 98.0
B 72.0
C 86.0
D 83.0
E 97.0
density of paper
(Ib/cu ft) content*
9-H
5.3
12-H
13-H
_ _
18-H
_ _
- _
..
--
F
chemically pulped paper consisting of approximately
equal quantities of bleached material (bleached car-
ton stock, bleached kraft paper toweling, tissue, and
writing paper) and unbleached material (board, bag,
and general unbleached paper). The overhead fraction
also contained some large shreds of film plastic and
light styrofoam. At 900 to 1,000 fpm, the overhead
fraction still contained small amounts of film plastic
and was contaminated by fine dust, glass, and small
wood splinters. The proportions of groundwood and
chemically pulped papers were essentially reversed
from those observed in the 700- to 800-fpm overhead
fractions. There was very little corrugated material in
this sample.
At a separating velocity of 1,100 to 1,200 fpm, the
overhead, paper-containing fraction obtained from
the heavy fraction from the 900-to 1,000-fpm run
was very similar to the previously obtained overhead
fraction, with a decreasing amount of news and an
increasing proportion of currugated board. This trend
was continued in the light fraction at the 1,200- to
1,300-fpm velocity, the heavy fraction from this run
containing almost entirely corrugated stock, contami-
nated with more than 50 peicent (by weight) of
shredded metal, wood chips, heavy plastic, glass, and
other, similar high-density materials. These contami-
nants were predominantly of small particle size and
probably could have been removed quite easily from
the paper by a 2-in. screen. They were also very
effectively removed by air classification at 1,800 fpm,
as observed in Run 4/8-12 on a sample of the same
refuse.
The effect of wetting the refuse feed before
separation, tried in the hope that wetting might
-------
PILOT-UNIT EXPERIMENTS
47
improve the removal of film plastic, is best demon-
strated by Runs 4/9-1 through 4/9-3 and Runs 4/94
through 4/9-6 on Albuquerque refuse (Table 25).
Essentially the same procedure was followed as that
used for the Cincinnati refuse, in which the heavy
fractions were successively rerun at higher separating
velocities. As expected (compare the light fraction
from Run 4/9-4 with the light fraction from Run
4/9-1), a greater proportion of material (18 percent
compared with 11 percent) was removed in the
dedusting cut, because this cut could be made at a
higher velocity without removing excessive amounts
of newspaper. Thus, the film plastic contamination of
subsequent paper cuts was reduced. The wetted feed
material required a somewhat higher velocity (1,000
to 1,100 fpm compared with 900 fpm) to remove the
overhead fraction containing newspaper, but the
percentages were essentially the same (33 percent and
34 percent, respectively, for dry and wetted feed
separations). Both newpaper fractions were slightly
contaminated with fine dirt; there was no particular
evidence, however, that wetting caused dirt to
adhere to the recovered paper. Both final cuts for
recovery of kraft and corrugated paper fractions
could be made at 1,400 fpm, and recoveries were
similar (57 percent dry and 60 percent wetted). The
heavy, non paper fraction from the 1,400-fpm
separation in both cases contained aluminum, shred-
ded tire fragments, shreds of inner tube, glass, tin
cans, and the like.
The air classification procedure described for
Cincinnati and Albuquerque refuse appeared to pro-
duce better separations than a procedure wherein
heavy material was separated initially at high velocity
and the overhead from each separation was run at
successively lower velocities. (Refer to Runs
TABLE 25
AIR CLASSIFICATIONS CONDUCTED ON ALBUQUERQUE REFUSE
Light overhead fraction Heavy bottom fraction
Sample Sample
taken for taken for
Percent St. Regis Percent St. Regis
of feed, Bulk analysis °ffeed Bulk analysis
by density of paper by density of paper
weight (Ib/cu ft) content* weight (Ib/cu ft) content*
Bulk
Weight density
of feed of feed
sample sample
Run Number Feed (H>) (Ib/cu ft)
Column's
air Feed
velocity rate
(fpm) (Ib/min)
4/7-5
Raw refuse 21.5 7.0
1,100-
1,200
1.0
56.0
44.0
4/7-6
4/7-7
4/7-8
4/7-10
4/9-1
4/9-2
4/9-3
4/9-4
4/9-5
4/9-6
Raw refuse
Heavies from
Run 4/7-6
Heavies from
Run 4/7-6
Lights from
Run 4/7-7
Raw refuse
Heavies from
Run 4/9-1
Heavies from
Run 4/9-2
Wetted raw
refuse
Heavies from
Run 4/9-4
Heavies from
Run 4/9-5
*Designations assigned by
47.0 7.0
34.5
34.0
7.0
54.0
48.0
32.0
60.0
45.0
27.0
St. Regis.
900 7.0
1,100- 7.0
1,200
1,200- 15.0
1,300
850- 2.3
900
500-
750
900
1,400
700-
800
1,000-
1,100
1,400
27.0 - - - 6-L
22.5
14.0 --- 8-L
30.0 - - - 10-L
11.0
33.0
57.0
18.0
34.0
60.0
73.0
87.5
86.0
70.0
89.0
67.0
43.0
82.0
66.0
40.0
...
...
8-H
10-H
...
...
...
...
...
...
-------
48
AIR CLASSIFICATION OF SOLID WASTES
4/8-19(1) through 4/8-19(3) on Johnson City
hammermilled refuse.) It appears useful, especially on
low-velocity runs, to have heavy fractions in the feed
to prevent column clogging by a sort of reflux action.
Although not tried, experimental results indicate that
the addition of heavy material such as gravel could be
beneficial for improving separating efficiency and
increasing throughput. The recommendation can be
made that, of the two separations needed for isolating
a paper-rich middle fraction from domestic and
commercial solid waste, separating the light material
and then removing the heavy fraction normally will
be found most effective.
Regarding removal of the heavy fraction, indica-
tions are that the sharp separation provided by air
classification may not be required simply to separate
heavy contaminants from the corrugated and other
strength grade papers left with the bottom product of
prior air classifier separations. In the process design of
an integrated recovery system, other dry separating
methods—such as screening, ballistic techniques, or
even a simple air cascade—that have lower separation
efficiencies might be acceptable if their costs are
sufficiently lower. It will also be desirable to investi-
gate removal of all heavy material initially by air
classification (alone or in combination with another
method, or methods, that will meet the desired
recovery objective) and in subsequent cleaning and
separation of the paper fractions by air classification
to use synthetic heavy material, such as gravel, for
reflux.
A very simple process for paper recovery based on
these principles can be visualized in which the
shredder serves primarily to break apart the com-
pacted refuse as delivered by packer-type collection
trucks, producing a paper component particle size in
the 6- to 12-in. range. A screen would be employed to
remove the same 6-in. material, which would be
largely cans, broken bottles, and dirt. The screen
would employ a dust removal system designed so as
to remove film plastic also, and the recovered paper
would then be air classified for separation into usable
grades. If necessary to remove large plastic bottles
screened out with the recovered paper mixture, a
simple ballistic device could be incorporated as an
appurtenance in the air classifier's feeding mechan-
ism.
Conclusions and Recommendations Regarding Air
Classification of Municipal and Commercial Waste
Most of the conclusions reached concerned the
effect of shredding on separation of municipal and
commercial waste by air classification. In addition,
certain recommendations can be made regarding
future work needed to provide data for the rational
design of full-scale air classification equipment. A
complete pilot plant that would employ air classifi-
cation and would operate as a research facility to
develop an economical recover;/ process is visualized
as the proper vehicle for accomplishing these further
design developments. This research facility might best
be incorporated into a demonstration facility in
which municipal solid waste was being shredded
routinely for landfill, compaction, incineration, or
other purposes.
Effect of Shredding. Characteristics of shredded
refuse have a predominant effect on the separability
of components to be extracted from it for recovery,
the process by which recovery can be effected, and
often, on the ultimate use to which recovered
material can be put. Size characteristics of the
shredded municipal refuse samples from Johnson
City, Los Angeles, Cincinnati, and Albuquerque were
prepared to provide data on the output product (not
only for the paper fi action but also for all other
components of the refuse) that is representative of
the particular type of shredder used. Differences in
the overall range of particle size have been discussed
previously. Regarding specific materials, it will be
useful to note first the similarities in the output
product from all four types of shredders and then to
note their differences.
Regarding glass, it can be said that any impact mill
produces fine particles from a brittle material. All
glass was in the minus 1/2-in. size range, most of it
reduced to a coarse sand. The rasper produced the
largest glass size, with noticeable quantities approxi-
mately 1/8 in. in size and some pieces to 3/8 in.
Ferrous metal had been removed from the
Johnson City and Los Angeles samples, and so little
can be said regarding the performance of the rasper,
the rigid-arm Williams, and the Gruendler mills on tin
cans. Generally, hammermills shred and crumple light
metal into medium-bulk density balls that are in the
1/2- to 1-1/2-in. size range. The large Williams
swinging-arm mill, operating without a grate, battered
but did not shred most cans in its single-pass
reduction of Cincinnati refuse.
Film plastic is a difficult contaminant to remove
from recovered paper. Shredder 'output of this
material is, therefore, important. It would be desir-
able to reduce this material to small fragments and at
the same time to produce large-size particles of paper.
There was a great deal of film plastic material larger
than 1-1/2 in. in the output of all shredders except
the Eidal. Although this machine also produced
relatively small-size particles of paper, these paper
particles were twisted and crumpled, and this con-
dition increased their bulk density and permitted
better separations from the small particles of film
plastic.
Most of the miscellaneous Group II material (food
-------
PILOT-UNIT EXPERIMENTS
49
waste, heavy plastic, yard waste, cloth, and wood) is
heavier than paper and appears in the rejected
bottoms fraction from air classification. Thus, it is
desirable that it be as large a particle size as possible.
All shredders appeared to deliver a large quantity of
this material in the less than 1/2-in. size range. With
the exception of the rasper products the amounts
delivered represented essentially the total quantity of
such material.
It is difficult to identify the origin of individual
particles of paper in a shredded refuse sample. This
was done, however, by St. Regis Paper in their
proprietary studies on air-classified fractions of paper
separated from municipal refuse. In the SRI study, an
attempt was made to estimate from the general
appearance of a sample the relative amounts of four
different categories of paper products present in three
different size ranges, after shredding. There was little
apparent difference in the behavior of different types
of paper in a given mill. The size distribution and the
nature of the shredded material were generally the
same for all types of paper products. An exception
was noted, however, in the case of newspaper
shredded by the Williams mills: because they were
being operated without bar grates or discharge
screens, both the rigid-arm and swinging-arm mills
passed large pieces of folded newspaper that had been
torn to only one-half or one-third of their original
size.
The following conclusions were reached regarding
the effect on air classification of the five methods of
shredding employed.
1. The method of shredding has little effect on
the separation velocities for which a full-size air
classification column would be designed, because a
considerable degree of operating flexibility must be
provided.
2. The method of shredding has little effect on
the separation of heavy materials from the paper
fraction.
3. The method of shredding influences the separa-
tion of film plastic from paper. Separation is more
effective when the material is wetted before shred-
ding and when the shredding action is of the rasp
type or by roller-bit hammers, as in the Eidal mill.
4. The method of shredding may have an influ-
ence on cleanliness of the paper fraction and ability
of air classifications alone to produce a product that
can be inexpensively cleaned by wet screening after
repulping. It appears that the Williams hammermills
have some advantage in this respect.
5. The methods of shredding investigated appar-
ently do not have significant influence on paper
separations that can be effected by air classification,
which in any case are not complete but should rather
be regarded as "beneficiations." Sharply sheared edges
are not essential when paper fragments are large.
Excessive recirculation in a shredder is,however,to be
avoided since it produces a "dry pulped" paper
product. There is indication that the rigid-arm hogger
used by paper balers may produce a paper fraction
that can be most easily separated, but the feed sample
shredded in this type of unit for experimentation
with the air classifier contained only the paper
fractions of refuse, and so these observations are not
conclusive.
Public Health Service Report No. 1908s presents
cost and performance characteristics of equipment
for various unit operations involved in refuse pro-
cessing. It does not, however, provide a basis for
estimating the characteristics of various components
of mixed municipal solid waste when the waste is
shredded by mills employing different grinding prin-
ciples. It would be desirable to determine for a given
mill the effects on product characteristics of rotating
speed, grate or screen size, and feed rate to the mill in
order to be able to specify the grinding principles to
be employed or the characteristics desired in the
shredded product. Because of the empirical nature of
the problem, apparently only full-scale tests would
yield significant information. A program of this sort
might be set up with the cooperation of interested
shredder manufacturers as a follow-on study.
Scale-Up Considerations. Once satisfactory separa-
tion has been achieved at either laboratory or pilot
scale, the single most important factor in scale-up to
commercial-size operations is the permissible column
loading. Satisfactory separation of wastepaper from
municipal refuse was not obtained in the laboratory
unit, but it was achieved in the pilot unit. In addition,
some information was developed on column loadings
in pilot-unit experiments; however, data from which
to draw general conclusions were insufficient.
Column-loading data abstracted from Tables 21
through 25 are summarized (Table 26). The calcu-
lated column loadings all represent feed rates at which
satisfactory column operation was being achieved.
With the exception of Run 4/7-5 on Albuquerque
refuse, all loadings are believed to be near the
maximum that could be used for design purposes.
From these data there appears to be some correlation
between allowable column loading and feed bulk
density for material similar to municipal refuse.
Intuitively, it is believed that an expression might
be developed for feed rate, w (Ib/min per sq ft of
throat area), in terms of column air velocity, V (fpm),
and bulk density, BD (Ib/cu ft), of the form:
w = K. V. (BD)
where K is a dimensionless constant related to the
experimentally determined column loading. Such an
expression would be dimensionally correct. Bulk
-------
50
AIR CLASSIFICATION OF SOLID WASTES
TABLE 26
SUMMARY OF COLUMN-LOADING DATA
FROM TABLES 21 THROUGH 25
Source of refuse,
Run Number
Johnson City (rasp)
3/11-3
Los Angeles
4/6-2
4/6-5
4/6-3
4/6-4
Cincinnati
4/8-9
Albuquerque
4/7-6
4/7-5
4/7-7
4/7-8
4/7-10
Feed material
Raw refuse, air dried
Synthetic dump stock
Synthetic dump stock
Synthetic dump stock
Synthetic dump stock
Raw refuse
Raw refuse
Raw refuse
73% bottom fraction
from Run 4/7-6
7 3% bottom fraction
from Run 4/7-6
22.5% overhead
fraction from
Run 4/7-7
Column's
Air Velocity— V
(ft/min)
1,500
600
800-900
900
1,100
700
900
1,100-1,200
1,100-1,200
1,200-1,300
850-900
Feed rate-w
(Ib/min/ft
of throat area)
30.0
1.0
2.0
1.4
2.8
4.0
14.0
2.0*
14.0
30.0
4.6
Estimated feed
bulk density-BD
(Ib/cu ft)
10.9
2.0
2.0
2.0
2.0
4.0-5.0
7.0
7.0
8.0-10.0
8.0-10.0
2.0-3.0
Calculated
loading
(Ib solids/lb air)
0.27
0.02
0.03
0.02
0.03
0.08
0.21
0.02
0.16
0.32
0.07
*Low feed rate to check efficiency of separation.
density is, however, an extremely unreliable par-
ameter as presently determined for refuse. A more
usable empirical expression for feed rate should be
achievable with additional experimentation employ-
ing other parameters of greater precision that are
functions of shape, size, and physical properties of
the refuse particles.
Work with rational design formulas should be
included in any future program of research on air
classification. Such research would be desirable to
investigate basic separation relationships for pre-
dicting not only allowable feed rates but also separa-
tion efficiency as related to feed rate and the number
of stages above and below the feed point; column
appurtenances more suitable to high-capacity
columns than the rotary, airlock devices used as
airseals for the feed and overhead discharge streams;
and similar design variables.
-------
Possible Role of Air Classification
in Processing Solid Wastes
Supplemental Equipment Needed for Mechanical
Processes of Solid Waste Reclamation
Air classification alone is a complete separation
process only for certain specialized operations, such
as the cleaning or density grading of seeds. For many
product separations it is necessary to prepare a
material to some extent before it is classified. For
solid wastes, these preparatory steps are likely to be
shredding, screening, and drying.
Shredders. Research results confirm the impor-
tance of shredding to successful air classification. The
problem of interpreting laboratory results is related
to the scale of the operation, since wastes must be
shredded to a considerably smaller size for laboratory
processing than for commercial processing. Some
manufacturer's literature was obtained on shredders
recommended for, or presently being used on, solid
wastes, and a report on one shredding project14 being
conducted under Federal solid waste program spon-
sorship was reviewed. This information was supple-
mented by readings15 in the current literature and
communication with compost plant operators who
were shredding domestic solid waste. The purpose
was to obtain a better understanding of the scale-up
problem related to shredding and to develop costs
that might be useful in preliminary assessments of
economic feasibility. The information obtained on
shredders is summarized (Table 27).
TABLE 27
SUMMARY OF DATA OBTAINED ON COMMERCIAL SHREDDING
EQUIPMENT FOR MUNICIPAL WASTES*
Shredder
Manufacturer designation
Gondard Hammermill
Operating or
rated
Capacity Weight
(tons/hr) Horsepower (Ib)
5-10 (operating; 150 (estimated) --
less than
capacity)
Equipment
Cost
$127,000
(includes
conveyors)
Total estimated
operating cost
(dollars/ton)
$2.0046.00
(test oper-
ation)
$1.00-$1.50
(estimated for
production
operation)
Eidal SW-200 shredder
International SW-100 shredder
SW-20 mini-null
Centriblast Crusher-disinte-
Corporation grator (Joy,
Model COM
4830HD)
80 (rated)
40 (rated)
3 (rated)
20 to 100
(rated)
1,400 or 2,000 200,000
700 or 1,000 100,000
80 7,000
300-1,500
$100,000
$40,000 and up
without motor
$70,000 for
20 ton/hr unit
at Gainesville,
Florida
Williams Patent HammermiU,
Crusher & Model 475
Pulverizer
Company!
GA (primary)
Hammermill,
Model 80 GA
(secondary)
60 (operating) 500
50 300
Francis &
John S Trace
Company
Lanway pulver-
iser (hammer-
mill)
40-50
450
37,000 $50,000
21,000 $30,000
44,800 $100,000
$1.00-$ 1.50, exclusive
of maintenance,
at Gainesville,
Florida (high
maintenance cost
reported)
$0.65 (operation)
$0.10 (maintenance)
$0.50 (operation
and maintenance)
(Continued next page)
51
-------
52
AIR CLASSIFICATION OF SOLID WASTES
TABLE 27 (Concluded)
SUMMARY OF DATA OBTAINED ON COMMERCIAL SHREDDING
EQUIPMENT FOR MUNICIPAL WASTES*
Manufacturer
Shredder
designation
Operating or
rated
capacity
(tons/hr)
Weight
Horsepower (Ib)
Equipment
cost
Total estimated
operating cost
(dollars/ton)
American
Pulverizer
Hammermill
Recommended for municipal wastes only if sorting could be accomplished prior to
grinding
Von Roll, Ltd.
The Heil Co.
Jeffrey Mfg.
Company
Jeffrey Mfg.
Company
Koehring Co.
The Pettibone
Companies
Gruendler
Crusher and
Pulverizer Co.
Bulky waste 15 50 89,600 $90,000
crusher (Jaw
shear)
Tollemache 15 200 -- $75,000
(vertical
hammermill
with ballistic
rejection)
G-28-B garbage 50-60 200 28,000
grinder (wet
hammermill)
Impact crusher 80-1,000 75-1,250 14,900
(hammermill) 110,500
Fox heavy-duty 80 15-150
forage harvester
(reel-type shear)
Bulldog refuse 35 (rated)
shredder
Stationary and 15 to 300
portable (rated)
refuse-
pulverizing
plants
*See also Waste volume reduction by pulverisation, crushing, and shearing. Paper presented by P.K. Patrick, Department of
Public Health Engineering (Refuse Disposal Branch). Greater London Council, at the 69th Annual Conference of the Institute
of Public Cleansing (British), June 1967.
f Data from Metro-Waste compost plant, Houston, Texas, and from San Diego, California, study o.f refuse baling.
With the exception of Eidal and Von Roll
machines and the reel-type agricultural product
chopper, all the shredders identified in Table 27
appear to be versions of conventional hammermills.
The Eidal mill is reported to be a vertical-shaft,
roller-bit shredder. The Gondard and Tollemache
units are claimed to incorporate an effective method
of heavy metal separation by ballistic action.
Estimated shredding costs claimed are found to be
comparable with those reported in the APWA rail
haul study (80* per ton minimum), the USPHS
Fresno, California, solid wastes management project
($0.90 to $1.40 per ton), and shredding costs
estimated in connection with the San Diego,
California, refuse baling transfer station. The latter
costs are based on actual cost experience at the Lone
Star Organics (Metro-Waste, Inc.) composting plant in
Houston, Texas. At this location, shredding costs that
initially were as high as $2.00 per ton have been
reduced to approximately 65* per ton by improving
maintenance practices, increasing throughput, and
decreasing the time between shutdowns. The current
target at this plant for secondary shredding costs is
35* to 40* per ton.
Screens and Driers. Information on available
screens and driers can be obtained from References 1,
2, and 3. The type of lightweight, low-cost equipment
developed for cleaning and drying of cotton may have
application to processing of municipal wastes. In
particular, the reel-type cleaner-drier for seed cotton
appears to be well suited to the removal of fine
material prior to air classification and to be economi-
cal, in that it can also be used for drying when
needed.
-------
PROCESSING SOLID WASTES
53
Scale-Up Procedure
The theory of air classification has already been
discussed briefly. In the discussion of the test
program, mention was made of factors that influence
the performance of an air classification unit. These
factors are elaborated (Table 28), both those related
to the material being processed and those related to
the variables in column operation. Moreover,the units
of measurement used in this report are given and
certain qualitative effects associated with various
factors are indicated (Table 28).
Because of the empirical nature of the relation-
ships among the factors that influence air classifier
operation, it is necessary to establish experimentally
those relationships of importance for any desired
separation. This involves (1) determination that the
mixture of materials can be processed satisfactorily
and (2) establishment of the degree of separation that
is desirable or possible. When a satisfactory separation
has been achieved at small scale, it is then possible to
expand the operation to the scale at which com-
mercial processing is contemplated.
For compost, both requirements could be satis-
fied; the material could be handled in the laboratory
unit and, by comparison with other acceptable
compost materials (such as dried sewage sludge and
dairy manure), performance criteria for the quality
requirements of commercial separation could be
approximated. Thus, it was possible, on the basis of
results obtained in this study, to estimate the
performance of a full-scale plant for compost pro-
cessing. A process flow diagram for a 30-ton/hr plant
is presented on page 57. Calculations for this flow
diagram were based on the scale-up factors given
(Table 29).
As an example of scale-up procedure, the follow-
TABLE 28
LIST OF FACTORS AND RELATIONSHIPS OF
IMPORTANCE IN AIR CLASSIFICATION
Factor
Units
Remarks on qualitative effects
Material being processed
Bulk density
Feed
Overhead
Bottoms
Particle size
Feed
Overhead
Bottoms
Particle gradation
Feed
Overhead
Bottoms
Particle shape
Feed
Overhead
Bottoms
Lb/cu ft
Max and min
sieve sizes
Percentage passing
and retained on
sieves of various
sizes
Length/diameter
ratio
Often not determined for
pilot column operation.
Column loading is more
important operating
parameter
Influences column clogging
due to "bridging"
Moisture content
Feed
Overhead
Bottoms
Tendency to agglomerate
Feed
Overhead
Bottoms
Percentage, by Drying can be effected
weight by using heated air;
air recirculation is
possible
Squeezing, in feed mechanism,
and impacting or electro-
static charges, in column,
may produce sticky surfaces
on material that cause
particle agglomeration or
surface buildup. Over-
shredding of fibrous materials
also causes agglomeration
(Continued next page)
-------
54
AIR CLASSIFICATION OF SOLID WASTES
TABLE 28 (concluded)
Factor
Units
Remarks on qualitative effects
Column operation
Fluidizmg velocity
Column loading
Effectiveness of
separation
Column capacity
Superficial velocity,
cfm/thioat area
Solids-to-air ratio,
(Ib solids/lb air)
Overlap in particle
gradation between
overhead and
bottoms
Lb/hr
Most important parameter.
Must be maintained for
scale-up.
Important factor in column
capacity. Varies between
0.2 and 0.8 for light
materials. May be as
low as 0.02 for shredded
paper.
Close separations require
light column loadings
Related directly to maximum
possible loading at
acceptable separating
effectiveness
Column pressure
drop
No load
At given loading
In. of water
In. of water
Characteristic of column
Depends on solids-to-air
ratio
ing calculations are presented for a 30-
ton/hr classifier for use on compost:
Quantity to be processed
Stockpile material
Dried material
Compost bulk density
Stockpile at 80% -100%
moisture content
Stockpile air dried to
approximately 20%
moisture content
Quantities to be air
classified (see Figure 4)
Coarse screenings
Middlings
Fluidizing velocities (from
laboratory column results)
Dedusting of coarse
screenings
Separation of middlings
30 ton/hr
20 ton/hr
32.7 Ib/cu ft
19.0 Ib/cu ft
42.5% x 20 ton/
hr = 8.5 ton/hr
27.5% x 20 ton/
hr = 5.5 ton/hr
550-600 ft/mm
850-900 ft/min
Required air quantities
Coarse screenings
284 Ib/min + 1.0
284 Ib air/min
Middlings
8.5 ton/hr =
8.5 x 2,000 =
60
284 Ib/min
Ib solids/lb air =
284 Ib air/min
+ 0.075 = 3,790
cfm air
5.5 ton/hr =
5.5 x 2,000 =
60
183 Ib/min
183 Ib/min ± 0.8 Ib solids/lb air = 229 Ib air/min
229 Ib/min + 0.075 = 3060 cfm air
Required throat cross sections
Coarse screenings
Middlings
3,790 cfm =
550
6.9 sq ft = 994
sq in
3,060 cfm _
850
3.6 sq ft =
518 sqin
-------
PROCESSING SOLID WASTES
55
TABLE 29
SCALE-UP FACTORS
BASED ON DATA OBTAINED WITH LABORATORY AIR CLASSIFIER
Ratio or Unit
Remarks
Throat ratios
Total area
Column width
Throat width
Column height ratio
Flow ratio
Pressure drop
Power requirement
Throat area of prototype/throat
area of pilot unit
Width of prototype/width of
pilot unit
Throat of prototype/throat
of pilot unit
Height of prototype/height
of pilot unit
Cfm of prototype/cfm of pilot
unit
In. of water
Horsepower
Column capacity scale-up factor
Variable, to give desired column
capacity at necessary fluidizing
velocity
Variable, widening throat increases
column height. Actual prototype
throat width determines maximum
particle size (oversize) that can
be fed
Geometrical scale-up; wider throat
requires taller column for same
number of flow reversals or
"plates." Reduced number of
effective "plates" reduces
separation efficiency
Numerically the same as throat area ratio,
since same superficial velocity must be
maintained
For same throat velocity and loading,
column pressure drop of prototype is
same as for pilot unit. Add approximately
2-in. water for prototype cyclone
Calculated from prototype blower cfm and
pressure drop. Not determined for pilot
unit
(Use 12 in. x 82 in. for coarse screenings and 8 in.
x 65 in. for middlings.)
Column height for 8-in. throat width
8 H- 2 x 42 in. (height of laboratory column)
= 168 in. = 14 ft for 12-stage unit (8-stage
unit would be two-thirds this height)
Pressure drop and blower
horsepower requirements
Total pressure drop
2.0 in. water
Column pressure drop
Cyclone separator
pressure drop
0.6 in. of water
(from lab column
results at 1.5 x
Ap for air only)
1.4 in. water (estimated)
Blower horsepower for 3,790 cfm at 3-in. static
pressure = 4.0 hp
Middlings
column pressure drop
Cyclone separator
pressure drop
2.0 in. water (from
lab column re-
sults at 1.5 x Ap
for air only)
2.0 in. water (estimated)
Total pressure drop 4.0 in. water
Blower horsepower for 3,060 cfm at 4-in. static
pressure = 3.4 hp
-------
56
AIR CLASSIFICATION OF SOLID WASTES
For automobile body waste, only one of the
conditions required for scale-up was satisfied. The
waste could be separated satisfactorily in the small
laboratory column, but no information was provided
on the beneficiated feedstock requirements for sub-
sequent processing for nonferrous-metal recovery.
Determination of these requirements was outside the
scope of this research. Consequently, the process flow
diagram (Figure 10), though typical of a recovery
process employing air classification, would require
additional laboratory work to optimize the separa-
tions in a particular crushing method and with a
specified objective for nonferrous-metal recovery.
Experimentation with the recovery of wastepaper
stock from combined refuse did not produce satis-
factory separation in the laboratory air classification
unit, because of the combined problems of shredding
and the small throat size of the laboratory unit.
Special scale-up difficulties are created when it is not
possible to classify a material that is of the same
particle size and aerodynamic characteristics as those
that will exist in the full-scale prototype equipment.
If the reduced-size particles can be satisfactorily
classified, the fluidizing velocity for the prototype
particle size must be estimated from laboratory
results and used in calculations of the column's cross
section, capacity, air requirements, and horsepower.
Whereas theoretical relationships between particle
size and density and the terminal velocity are of some
assistance in making such estimates,* they are of
limited practical usefulness because of the differences
(acceleration effects and turbulence) discussed pre-
viously. Since satisfactory classification could not be
effected, the process flow diagram developed for
paper recovery processing is indicative only of the
conditions anticipated for a full-size column and
larger size particles with more sharply cut edges
(relative to the size of particle) than were obtained in
the laboratory shredder.
Estimated Full-Scale Performance of Air Classifier
Unit in Processing Solid Waste
Compost from Municipal Refuse. Based on the
calculations presented previously, two 14-ft air clas-
sifiers have been selected for a 30-ton/hr compost
plant. The classifier for removing plastic and other
light material from the coarse screenings before
recycling or bulk sale is an 8-stage unit with a throat
size of 12 by 82 in. The column for separating glass
and other heavy contaminants from horticultural-
grade compost is 12 stage and has a throat size of 8
by 65 in. On each column, 5-hp blowers would be
*In actual practice on a larger experimental column,
fludizing velocities would be obtained for separation of
several sizes of particles and extrapolated to the prototype
particle size.
required, but the operating power requirement would
normally be less than 7.5 hp for both units. The two
columns, complete with blowers and cyclones, would
cost approximately $42,500. A complete processing
plant would also require equipment for shredding and
screening as well as facilities for in-plant materials
handling. The makeup of such a plant is shown by the
conceptual process flow diagram (Figure 10). This
plant would handle 30 tons o.f stockpile compost per
hour at 80 to 100 percent moisture content, after it
had been air dried to a maximum moisture content of
20 percent. The actual plant throughput of air-dried
compost is 20 tons/hr. When the unit's capacity on
air-dried material is used as a base, an estimate of cost
suitable for preliminary comparison with other
methods of compost processing indicates that the
capital investment in such a plant, excluding land,
would approximate $5,000/ton/hr capacity in this
size range. The plant's operating cost, including
depreciation, is estimated to be 50«/ton (dry basis)
for a single-shift operation and 30*/ton for three-shift
operation. Of the latter figure, approximately 20*/ton
is attributable to the cost of shredding.
Automobile Body Trash. The process for non-
ferrous metal recovery from automobile body trash
could not be developed to the extent possible for
compost. Experimentation for design of a complete
reclamation process was outside the scope of the
research program. Air classification runs on auto-
mobile body material demonstrated only that close
separations could be made; there was no attempt to
optimize these separations or to concentrate any
particular metallic constituent, The results achieved
were, however, sufficient to permit comparisons to be
made of conceptual air classification processing with
the process currently being employed for ferrous
metal separation and a proposed process for non-
ferrous recovery. Process diagrams are given (Figures
11,12, and 13).
A typical automobile body-shredding operation is
diagrammed (Figure 11). Because of air pollution
regulations, upholsteiy and other combustibles can-
not be removed by burning before shredding. Conse-
quently, magnetic separation leaves this material in
the nonferrous stream. Also as a consequence of not
being able to burn is an explosion hazard from the
dust resulting from the shredding and all subsequent
material-handling operations; this dust must be col-
lected to prevent air pollution as well as to control
the explosion hazard.
Air classification could improve the conventional
separating process (Figure 12). Magnetic separation is
still employed; however, its effectiveness is increased
by the prior removal of all fibrous material by air
classification, which also eliminates handpicking to
remove contaminants such as rags, rubber, and copper
-------
PROCESSING SOLID WASTES
57
E 3
S3
-------
58
AIR CLASSIFICATION OF SOLID WASTES
TO DUST
COLLECTION
SYSTEM
TO DUST
COLLECTION
SYSTEM
SHREDDED STEEL (PRIMARILY BALLED SHEET a HEAVY FRAGMENTS TO 6" SIZE I
30-185 T/HR 100-150 */CUFT BULK DENSITY
HAND CLEANING
BY PICKING
OF COPPER WIRE
SPRING LEAVES
RAGS & RUBBER
TO DUST
COLLECTION
SYSTEM
TO DUST
COLLECTION
SYSTEM
NON FERROUS METAL & TRASH
15-20 T/HR 35 /CU FT BULK DENSITY
TRUCK HAUL TO DISPOSAL SITE OR
BELT CONVEYOR TO WET PROCESS
OF NON FERROUS METAL RECOVERY
Figure 11. Process flow block diagram: conventional automobile body shredding system.
-------
PROCESSING SOLID WASTES
59
CRANE
HANDLING
FROM
STOCKPILE
LARGE PI
1
Ifc. w
^^ SCRE
TO DUST
COLLECTION
SYSTEM
t
-* 8^"^ -^ ""V"'A^ -
CONVEYOR HAMMER MILL
CONVEYOR QR SHREDDER
150-200 T/HR
TO DUST
COLLECTION
SYSTEM
t
BUCKET AIR
ELEVATOR CLASSIFIER
135-190
CLEAN SHREDDED STEEL 130-180 T/HR
fc BELT CONVEYOR
TO GONDOLA CARS
ECES NON FERROUS METAL, RUBBER, GLASS AND OTHER INORGANICS 5-1
1 | HAND PICKING OF
~-T - LARGE PIECES OF ' ^.
ENING |NON FERROUS METAL ,
•I 1 | ( IF DESIRED )
RUBBER. FirvrflTinij CLLAN
DIRT FLOTATION GLASS
BELT NON FERROUS MET
CONVEYOR SEPARATION PROC
OTOTE BIN ( 3-5 T/HR )
MAGNETIC
SEPARATOR
3 T/HR
HL TO
ESSING
)
AND GLASS SPPABAT.™ RUBBER^ „,_,..„.
Fl
OF
ft [MRT
9ROUS AND OTHER ORGANIC MATERIAL PL
FERROUS AND NON FERROUS METAL, RU
US SMALL PIECES
ST.DIRT.ETC (10-IST/HR)
TO DUST
COLLECTION
SYSTEM
PNEUMATIC
TRUCK HAUL TO
DISPOSAL SITE
( 5-10 T/HR )
TO NON FERROUS
METAL SEPARATION
PROCESSING
( 0-3 T/HR )
GRANULAR IRON
TO SALVAGE
(5-7 T/HR)
Figure 12. Process flow block diagram: modified system for automobile body shredding employing air classification
and including nonferrous metal recovery.
-------
60
AIR CLASSIFICATION OF SOLID WASTES
wire. Air classification also more effectively dedusts
the ferrous stream than the air cascades frequently
employed for this purpose do.
Further advantages of air classification are evident
when the processing system is expanded to include
the recovery of nonferrous metals. A wet process
under construction for separating rubber, aluminum,
zinc, copper, and stainless steel from the trash stream
currently disposed of to landfill is shown (Figure 13).
Even though approximately 15 tons of material is
recovered per hour, the weight of material requiring
disposal is not reduced (and disposal costs remain
unchanged), because the fibrous material has become
water soaked in the process. When air classification is
employed (Figure 12), the wet screening and flota-
tion is done on a smaller stream of material, and this
circumstance reduces the required size of new equip-
ment 01 increases the capacity of an existing installa-
tion. The material for landfill disposal is dry, having
been removed by air classification rather than wet
screening, and its weight is less by approximately 50
percent.
In summary, advantages of the modified pro-
cessing system employing air classification are as
follows:
It improves efficiency of magnetic separations by
removal of fibrous material from magnetic separator
feed.
It eliminates handpicking for final cleanup of steel.
It reduces dust content of steel.
It reduces cost of dust collection installation.
It reduces amount of material requiring landfill
disposal.
It permits simple, hand salvage operation on larger
fragments of nonferrous metal, if desired, by separa-
tion from fibrous trash.
It increases capacity of wet screening and flotation
equipment.
Municipal Refuse. Because of the size limitations
of the laboratory air classification column and the
accompanying problems of shredding the refuse to be
processed, as discussed previously, it was not possible
to demonstrate conclusively that the salvage of a
usable grade of paper from municipal refuse was
technically feasible. It is believed, however, that a
new grade of clean and uniformly shredded paper
consisting of mixed newspaper, kraft paper stock, and
corrugated cardboard—with little or no styrofoam
packing and sheet plastic and only small percentages
of paper or paperboard that is highly filled (such as
magazine stock) or that contains waterproofing (such
as waxed paper, butter cartons, and the like)—could
be produced economically on a commercial scale
from combined domestic wastes. Such a reclaimed
product would have value, and might open up a new
market for secondary fiber in the paper industry.
The recycling of wastes can have important effects
on waste management costs. In the United States,
paper currently accounts foi about 50 percent, by
weight, of municipal solid wastes. Secondary fiber
(wastepaper) represents the second largest supply of
fibrous raw material, following wood pulp, for the
U.S. paper industry. Between 10 and 20 percent of
this paper is now salvaged and reused, according to
one source.16
On the assumption that there is a market for
recycled paper, this same source estimates waste
management costs for each level of recycling. In New
York City, for instance, where paper not recycled is
disposed of with other solid wastes by incineration,
subsequent disposal of the incinerator residue being
by sanitary landfill, it is estimated that the annual
disposal cost for 20 percent tecycle is about one and
one-half times the cost for 80 percent recycle, a
difference of approximately $100 million annually
for the New York region.
Another source17 estimates that, in 1966, 10
million tons of paper stock (wastepaper) were re-
cycled and became raw material for new products.
The U.S. Forest Service predicts the volume will
reach 17 million tons by the year 2000. Employment
in 1958 to collect this material and get it to market
involved 10,000 employees with a payroll of $45
million. Wastepaper provides about 25 percent of the
raw material for the paper and paperboard industries.
The total value of paper stock (wastepaper to
consuming mills) is greater than $300 million per
year. A beneficial effect on conservation was the fact
that 12,800,000 cords (13 million acres) of trees did
not have to be cut in 1966, because of the 10 million
tons of wastepaper that were used in place of wood
(that is, raw material).
Comparative paper reuse in the United States and
foreign countries for which data are available is
estimated to be:
United States
United Kingdom
West Germany
Japan
10% (minimum estimate)
25% (maximum estimate)
27% '
33%
46%
U.S. prices of wastepaper stock are given (Table 30).
New markets are probably essential for a new
recycled paper product because of the quantities
involved. The city of Los Angeles, excluding the
smaller cities of the metropolitan area and com-
mercial collectors in the unincorporated areas of Los
Angeles County, picks up approximately 5,000 tons
of domestic waste daily. During the 1970s, a number
of municipal transfer stations to serve this area will
become justified on the basis of transportation
economics alone; the minimum size of the transfer
station will be 300 ton/d. Los Angeles basin cities,
-------
PROCESSING SOLID WASTES
61
° •
UJ L
J n-
H »S
< f
> 1
r 0-
c a.
UJ 0 o
d k1
i
<
es
z
^ L
_l _l IE
3 < I
< W \
X O UJ 1-
te-v a- »- ^
0 « « S
K ° J,
-° !
h
a;
o
t— r t *-
tf — ' UJ
> w >
S °
_l
UJ
< o
I
= a:
WET
SCREENING
i
O -
a a
ffi
iZ
J .
DEWATER ING
i
» UJ >
UJ m Z
-1 0
UJ <->
.
1
0 UJ
UJ 0
1- <
< IT
> 0
UJ h-
-1 (/)
UJ
D z
o
Ct
I
1 K.
~
CONVEYOR
1
01
o: -j
5S 2
^» * Ul
y° §
* 0
ct
or
lu
U.
a
o
s
«
^
UJ
o
U3
Or
Uj
CD
00
3
Or
&
S
[•TI
J3
eQ
J3
>,
T3
5
X:
•S
2
O
E
o
3
« » i
ID *
H ^ >•
or — B
CM UJ.E £
' ^* 8
Z ^ 0
~ ° —
UJ — z O "«
1- IE 0 T 0
cn x 1- J, E
\ 0)
-J 1- I - "S.
< -J §
UJ C
UJ (0 u ^
° s yg I
_/ > Z 0. g
^ -1 ;; a
< < 2 z -s
I <" en 0 >-^
S 2 a 5 ll
3 i a- t S tS
H ' S< -B &
i 1 &s °s
S ? 8" ||
a, Z J< a|
535^ El
CC < N S S >,
1 1 1 *S
•^2
- §
1 -° I
* =
z < z Q n
050 G-j
r~ uj i_ s ^
^ t o: Z S K -2
^ 1- 0 „. IE pc
o S < £ g
_l z "• G
U- UJ UJ m 0
S w 2 °
,H>*
-------
62
AIR CLASSIFICATON OF SOLID WASTES
such as Santa Monica and Beverly Hills, have been
operating refuse transfer stations for a number of
years. In Orange County, two transfer stations have
operated since 1964 and each now processes more
than 500 tons/day. A third transfer station recently
has been completed. Future expansion of these
stations to a capacity of 800 to 1,000 tons/day is
possible. Thus, if only facilities already in operation or
contemplated in the greater Los Angeles area are
considered, a minimum of 3,000 tons of refuse per
day will probably be handled through municipally
built transfer facilities by 1975. If salvage contractors
were to operate these stations under no-cost contracts
and process the refuse to recover only 25 percent of
the incoming material as mixed paper, an assured
supply of 750 tons/day would result. At a price of
only $5.00 per ton, the dollar value of the recovered
paper would be more than $1 million per year. In
addition, Los Angeles and Orange counties and their
municipalities would save the present cost of transfer
station operation, haul costs would be reduced 25
percent, and landfill life would be extended.
Related Applications
The ease with which noncombustible material can
be removed from municipal refuse has been demon-
strated in the laboratory. Air classification is being
incorporated in the process flow diagrams of the
400-ton/day combustion power unit now under
The primary air classifier m the pneumatic conveying
system between the shredder and the drum dryer of Figure
13 would perform this separating function fn any process
requiring shredded refuse.
development. This unit will produce electrical power
in a gas turbogenerator from solid waste that has been
gasified by combustion or pyrolysis in a high-
pressure, fluidized bed retort.
When refuse is already shredded, as for composting
or fluidized bed combustion, glass, dirt, rubber,
metallics, and wood and heavy plastics (which typi-
cally represent 20 to 25 percent by weight of
municipal refuse) can be removed for less than 10*
per ton by air classification.* This opens up many
possibilities for recovering usable materials such as
ferrous metal and glass. It is known that secondary-
materials processors are keenly aware of, and are
interested in, all developments related to new salvage
opportunities. The Sanitary Engineering Research
Laboratory of the University of California at
Berkeley is interested in salvage as a facet of solid
wastes management.l *
One further application that appears to merit
investigation by researchers working with refuse
incineration is the removal of heavy materials from
the incinerator feed by shredding and air classifi-
cation. Besides the salvage possibilities that would be
created, incinerator operation and maintenance costs
might be reduced by such removal. New methods of
combustion, such as fluidized-bed and grateless
systems employing injection of shredded material by
blowers or ram-type packers, become worthy of
consideration. In conventional incineration, the
recovery of materials from the ashed residue by air
classification, though not yet investigated by the
Institute or others, appears to offer some promise for
economical recovery and leuse of secondary
materials.
TABLE 30
PRICES (PER TON) OF WASTEPAPER STOCK IN MAJOR U.S MARKETS*
Other grades of paper stockf
Number of
other grades
Market
New York
Chicago
Boston
Pittsburgh
Philadelphia
Folded news
$19.00-$28.00
(3 grades)
$16.00418.00
$12.00-$13.00
$18.00-$20.00
$24.00-$31.00
(3 grades)
No. 1 mixed
$11.00
$3.00-$4.00
$5.0046.00
$6.0048.00
$7.00
Old corrugated
$13.00-$14.00
$11.00-$13.00
$15.00-516.00
$16.00-$18.00
$16.00-$17.00
available
36
8
None
15
11
Price range*
$2.00-$65.00
$2.00-$55.00
-
$5.00450.00
$2.00457.50
*Secondary Raw Materials, 11 (6): 59, Nov. 1968.
fSee Paper stock standards and practices. Circular PS-66, Paper Stock Institute of America. January 1966.
*Low grade is typically mixed books and magazines, and high grade is either bleached, unpnnted sulfite cuttings or
No. 1 hard, white envelope cuttings.
-------
References
1. TAGGART, A. F. Handbook of mineral dressing; ores and industrial minerals. Section
9. New York, John Wiley & Sons, Inc., 1945.
2. PERRY, J. H., ed. Chemical engineers' handbook. 4th ed. New York, McGraw-Hill
Book Company, 1963. 1,898 p.
3. AGRICULTURAL RESEARCH SERVICE U.S. DEPARTMENT OF AGRICULTURE.
Handbook for cotton ginners. Agriculture Handbook No. 260. Washington, U.S.
Government Printing Office, Feb. 1964. 121 p.
4. MONASEBIAN, D. Principles and applications of air classifiers. American Laboratory,
p. 10-18, Dec 1968.
5. RALPH STONE AND COMPANY, INC. Copper control in vehicular scrap with special
emphasis on component design. Los Angeles, Mar. 1968. 109 p.
6. CALIFORNIA STATE DEPARTMENT OF PUBLIC HEALTH. California solid waste
planning study, v.l. Interim report. Status of solid waste management.
[Sacramento], 1968.
7. ROGUS, C. A. Refuse quantities and characteristics. In Proceedings; National
Conference on Solid Waste Research, University of Chicago Center for Continuing
Education, Dec. 1963. American Public Works Association Research Foundation,
Feb. 1964. p. 17-27.
8. BELL, J. M. Characteristics of municipal refuse. In Proceedings; National Conference
on Solid Waste Research, Chicago, Dec. 1963. American Public Works Association,
1964. p. 28-38.
9. Personal communication. E. R. KAISER, and C. D. ZEIT, New York University, to N.
L. DROBNY, Battelle Memorial Institute, Aug. 1966.
10. Personal communication. J. HOUSER, Fairfield Engineering Company, to N. L.
DROBNY, Batteile Memorial Institute, Aug. 16, 1967.
11. Personal communication. J. E. HEER, JR., University of Louisville, to N. L. DROBNY,
Battelle Memorial Institute, Sept. 5, 1967.
12. Personal communication. W. GALLER, University of North Carolina, to N. L.
DROBNY, Battelle Memorial Institute, July 25,1967.
13. LESLIE, T., J. KENNEDY, and G. GARLAND. A study of residential solid waste
generation variables. Unpublished data.
14. Solid waste reduction/salvage plant; an interim report; City of Madison pilot plant
demonstration project, June 14 to December 31, 1967. Washington, U.S.
Government Printing Office, 1968. 25 p.
15. DROBNY, N. L., H. E. HULL, and R. F. TESTIN. Recovery and utilization of
municipal solid waste; a summary of available cost and performance characteristics
of unit processes and systems. Public Health Service Publication No. 1908.
Washington, U.S. Government Printing Office, 1971. 118 p.
16. BOWER, B. T., G. P. LARSON, A. MICHAELS, and W. M. PHILLIPS. Waste
management; generation and disposal of solid, liquid, and gaseous wastes in the
New York region. Bulletin 107. New York, Regional Plan Association, Inc., Mar.
1968.107 p.
17. GOLUEKE, C. G. Comprehensive studies of solid wastes management; abstracts and
excerpts from the literature. Berkeley, Sanitary Engineering Research Laboratory
Report No. 68-3. Berkeley, School of Public Health, University of California, 1968.
p. 236.
18. GOLUEKE, C. G., and P. H. McGAUHEY. Comprehensive studies of solid wastes
management. Sanitary Engineering Research Laboratory Report No. 67-7.
Berkeley, University of California, May 1967. p. 105-117.
63
-------
-------
Appendix A
INTERVIEW RECORD FOR SELECTION OF WASTES
TO BE PROCESSED IN PHASE I
-------
-------
APPENDIX A
67
APPENDIX A
INTERVIEW RECORD FOR SELECTION OF WASTES TO BE PROCESSED IN PHASE I
Date
Individual and organization
Suggested waste
Remarks
September 4, 1968
September 6,1968
Septembers, 1968
September 6,1968
September 10,1968
September 12, 1968
September 13,1968
September 17, 1968
September 17, 1968
September 18,1968
Harry Faversham, vice president,
and James W. Moberg, chief
engineer, Clean Steel Inc.,
division of National Metal
& Steel Corp.
Joseph Edberg, engineer,
Pan-American Resources,
Inc. (by phone)
Nathan Herman, industrial
waste enforcement
engineer, Los Angeles
County engineer's staff
Frank Dair, division engineer,
solid waste disposal, Los
Angeles County Sanitation
Districts
Richard P. Stevens, president,
Universal By-Products, Inc.
John Gault, project engineer
and Henry Giles, refuse
superintendent, City of
Pasadena
Don Hoffman, technical studies
coordinator, Fred Meyer and
Mike Noe, San Diego Utilities
Department Laboratory,
San Diego, California
Robert B. Laursen, senior
engineer, Utilities Division
Department of Public Works,
Sacramento County, California
Steven Klein and Clark Weddle,
University of California at
Berkeley, Sanitary Engineering
Research Laboratory,
Richmond, California
Professor P. H. McGauhey,
director, University of
California at Berkeley,
Sanitary Engineering Research
Laboratory, Richmond,
California (by phone)
Trash stream from automobile body
shredding
Municipal waste for steel recovery
Trash stream from automobile body
shredding
Domestic waste
Demolition waste
Cabinet shop wood waste
Three municipal waste streams
of decreasing quality for paper
salvage:
Retail store collections
Apartments collections
Single-family collections
Supplied sample of
automobile body trash
Interested in separating glass
and metal from retort
char and in recovering
pigments from paint
sludge
Suggests grant application
to develop automobile
body nonferrous recovery
process
Recovery of cellulose, steel,
and wood for paper pulp
and rubble for core
Domestic trash to remove glass,
metal, and other incombustible
material for better retorting
Universal By-Products can
supply samples of all three
suggested wastes
Made reference to Los
Angeles By-Products
Made reference to Kelso
Kelp Products
Sacramento County took over
collection in northern area of
county from contractor as of
July 1, 1968
Working with Dr. C, Golueke on
anaerobic digestion of solid
wastes
Will put SRI on mailing list for
field station research reports
-------
68
AIR CLASSIFICATION OF SOLID WASTES
Date
Individual and organization
Suggested waste
Remarks
September 18, 1968
September 18,1968
September 20,1968
September 23,1968,
Peter A. Rogers, State of
California, Department
of Public Health, Bureau
of Vector Control,
Berkeley, California
D. M. Keagy, regional
representative, USPHS,
solid waste program,
San Francisco, California
Professor A. Bush, sanitary
engineering, University
of California at Los
Angeles (by phone)
Jack Betz, assistant director,
and C. Imel, research
engineer, City of Los
Angeles, Department of
Public Works, Bureau of
Sanitation
Shredded municipal refuse for
separation of compost and
noncompost materials
Airplane salvage
Ash from aluminum smelting
Paper from municipal waste for
production of alcohol by
fermentation of carbohydrates
Grading of finished compost;
initial separation into
fractions for composting
and retorting, plus
uncompostable fraction.
Suggests contact with SIRA
Corp., Los Gatos
Suggests contacting Davis-
Monthan Air Force Base,
Tucson, Arizona
Questions economics of
shredding and classifying
September 23,1968
September 24, 1968
September 30,1968
October 14, 1968
•October 22,1968
October 29,1968
November 14,1968
Tom Conrad and Bob Stearns,
Ralph Stone Engineers,
Los Angeles, California
Frank Bowerman, Aerojet-
General Corp., El Monte,
California
Paul Maier, Bureau of Vector
Control, California
Department of Public Health,
Fresno, California (by phone)
Victor Brown, president, Metro-
Waste, Inc., Wheaton,
Illinois (by phone)
Harry Armstrong, works manager,
U. S. Gypsum Paper Mill,
Southgate, California
Jerry Vaughan, plant manager,
Lone Star Organics, Houston,
Texas (by phone)
John Siracusa, president,
Sira Corp., Los Gatos,
California
Industrial waste probably has
greatest potential for
economic feasibility
Industrial waste, paper
salvage and compost
Agricultural wastes such
as nut hulls (i.e., walnuts
and almonds)
Mechanical extraction of
all salvage now handpicked
from domestic refuse
Recovery of clean mixed paper
from grades presently
purchased
Compost separation into salable
grades. Salvage and removal
of noncompost material from
feed to plant
Domestic refuse
Can supply costs on
shredding
Dense compaction is new
direction for disposal
Will supply shredded
samples from Houston
plant of Lone Star
Organics
Desires assured supply
without metal or
plastic contamination;
additional cost for
clean paper is difficult
to justify
Sira system does not
contemplate salvage;
package unit available
for burning of pulver-
ized refuse (can be
used as drier)
-------
Appendix B
REPORT OF SAN DIEGO UTILITIES DEPARTMENT
ON RETORTING OF AUTOMOBILE BODY MATERIAL
-------
-------
APPENDIX B
TABLE B-l
RUN NO. 156 (MATERIAL RETAINED ON 1-in. SCREEN)
71
Gas data, by constituent
H2 CH4 CO CO2 C2H4 C2H6
Composition (%) 33.70 20.24 18.60 17.58 5.63 1.71
Metered gas (cu ft) 2.800 1.682 1.546 1.461 0.468 0.142
Cu ft gas at standard 2.539 1.525 1.402 1.325 0.424 0.129
temperature and
pressure
Gram factor 2.547 20.268 35.385 55.598 35.438 37.985
Mass (g) 6.467 30.908 49.609 73.667 15.026 4.900
Handbook Btu/cu ft-low 290 963 341 - - 1,631 1,703
Evolved gas (Btu) 736 1,469 478 - - 692 220
Material pyrolyzed: 2,308 g + 453.59 g/lb = 5.088 Ib
Cu ft gas/lb material pyrolyzed = 1.443
Btu/lb material: 3,595 Btu -r 5.088 Ib = 707
Gas (Btu/cu ft). 3,595 Btu -r- 7.344 cu ft = 490
Lb gas evolved: 180.577 g + 453.59 g/lb = 0.398
Lb gas evolved/lb material pyrolyzed: 0.398 Ib + 5.088 Ib = 0.078
Barometric pressure (corrected for elevation): 29.42 in. Hg
Room temperature: 23 C
Volume correction factor to standard temperature and pressure: 0.9069
Mass balance
Pyrolysis products Amount (g) Percentage
Inerts* 820.000 35.53
Chart 864.000 37.44
Condensables 400.000 17.33
Gas 180.577 7.82
Total 2,264.577 98.12
Material pyrolyzed. 2,308.000 g
Mass accounted for . 98.12%
*Before calorimetry, pyrolysis residue passed through a No. 25 sieve to remove gross particles of metal and
Material retained on No. 25 sieve designated as "inerts" and so shown in Mass Balance.
tMaterial passing No. 25 sieve.
Pyrolysis residue passing No. 25 sieve
Proximate analysis (moisture-free basis) - ASTM D271-58
Volatile material (%) 9.12 Calorimetry (Parr bomb)
Fixed carbon (%) 19.02
Ash (%) 71.86 Heating value (Btu/lb) 3,840
100.00
Total
97.46
8.099
7.344
180.577
3,595
glass.
-------
72
AIR CLASSIFICATION OF SOLID WASTES
/
\
/
s'
X
1
I
o
o
10
\
a
o
Uj
10
$
3
o
O
5
\ J
\ff
\
o
o
o
o
o
Q3H313W - StfO
o
o
UJ
~J
0.
3
O
O
O
It
UJ
1
\
o
o
o
o
o
o
r-~§-~
r
a:
o uj
- SVO.O.J
-------
APPENDIX B
TABLE B-2
RUN NO. 157 (MATERIAL PASSING 1-in. SCREEN)
73
Gas Data, By Constituent
CH4
CO
CO,
temperature
and pressure
2.547
10.453
290
1,190
20.268
36.482
963
1,733
35.385
56.758
341
547
Gram factor
Mass (g)
Handbook Btu/cu ft—low
Evolved gas (Btu)
Material pyrolyzed: 2,770 g -r 453.59 g/lb
Cu ft gas/lb material pyrolyzed
Btu/lb material: 4,720 Btu -r 6.107 Ib
Gas (Btu/cu ft):4,720 Btu ^- 10.697 cu ft
Lb gas evolved: 265.969 g -r 453.59 g/lb
Lb gas evolved/lb material pyrolyzed: 0.586 Ib -=• 6.107 Ib
Barometric pressure (conected for elevation): 29.55 in. Hg
Room temperature: 22 C
Volume correction factor to standard temperature and pressure:
55.598
134.992
= 6.107 Ib
= 1.752
= 773
= 441
= 0.586
= 0.096
0.9140
35.438
22.574
1,631
1,039
37.985
4.710
1,703
211
Total
Composition (%)
Metered gas (cu ft)
Cu ft gas at standard
37.73
4.490
4.104
16.55
1.969
1.800
14.75
1.755
1.604
22.33
2.657
2.428
5.86
0.697
0.637
1.14
0.136
0.124
98.36
11.704
10.697
265.969
4,720
Mass Balance
Pyrolysis Products
Amount I
Char*
Condensables
Gas
1,650.600
455.600
265.969
2,372.169
Material pyrolyzed: 2,770 g
Mass accounted for: 86.64%
Percentage
59.59
16.45
9.60
85.64
* Pyrolysis residue
Proximate analysis (moisture-free basis)-ASTM D271-58
Volatile material
Fixed carbon
Ash
Percentage
5.26
2.08
92.66
100.00
Calorimetry (Parr bomb)
Heating value (Btu/lb) 1,073
S GO\ IiRNVj KN I PRINTING OFFIC ? 1972—484-484/151
-------
------- |