EPA-600/3-77-030
August 1977
Ecological Research Series
FOAM GLASS INSULATION FROM WASTE GLASS
Municipal Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ECOLOGICAL RESEARCH series. This series
describes research on the effects of pollution on humans, plant and animal spe-
cies, and materials. Problems are assessed for their long- and short-term influ-
ences. Investigations include formation, transport, and pathway studies to deter-
mine the fate of pollutants and their effects. This work provides the technical basis
for setting standards to minimize undesirable changes in Jiving organisms in the
aquatic, terrestrial, and atmospheric environments.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/3-77-030
August 1977
FOAM GLASS INSULATION FROM WASTE GLASS
by
Wendell G. Oakseson
June-Gunn Lee
S. K. Goyal
Thayne Robson
Ivan B. Cutler
Department of Materials Science and Engineering
University of Utah
Salt Lake City, Utah 84112
Grant No. R800937-02
Project Officer
Charles J. Rogers
Solid and Hazardous Waste Research Division
Municipal Environmental Research Laboratory
Cincinnati, Ohio 45268
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Municipal Environmental Research
Laboratory, U.S. Environmental Protection Agency, and approved for publica-
tion. Approval does not signify that the contents necessarily reflect the
views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.
-------
FOREWORD
The Environmental Protection Agency was created because of increasing
public and government concern about the dangers of pollution to the health
and welfare of the American people. Noxious air, foul water, and spoiled
land are tragic testimony to the deterioration of our natural environment.
The complexity of that environment and the interplay between its components
require a concentrated and integrated attack on the problem.
Research and development is that necessary first step in problem
solution and it involves defining the problem, measuring its impact, and
searching for solutions. The Municipal Environmental Research Laboratory
develops new and improved technology and systems for the prevention,
treatment, and management of wastewater and solid and hazardous waste pol-
lutant discharges from municipal and community sources, for the preservation
and treatment of public drinking water supplies, and to minimize the adverse
economic, social, health, and aesthetic effects of pollution. This publica-
tion is one of the products of that research; a most vital communications
link between the researcher and the user community.
The objective of this study is to determine if waste glass from municipal
waste streams could be used for production of foam glass insulation both in
bulk or rigid board form and pellet form. Results of the study showed that
water was the best foaming agent for waste glass for micron sized particles
to 0.6 cm pellets, while carbon and calcium carbonate yielded better products
for larger objects. With foamed pellets available to industry the develop-
ment of a market will probably take place by logical processes of replacement
of part of the market for conventional insulation materials.
Francis T. Mayo, Director
Municipal Environmental Research
Laboratory
-------
ABSTRACT
Waste glass has proven to be effective for the production of foam glass
insulation both in the bulk or rigid board form and pellet form. Many
problems, inherent in the use of water, carbon black, and calcium carbonate
as the foaming agents, have been identified and many have been solved by
various techniques.
The foaming agents were not equally effective for the same-size foamed
objects. Water was found to be best suited for micron-sized particles to
0.6 cm pellets, and carbon and CaCO-, yielded better products for larger
objects.
Large amounts of water can be rapidly incorporated into glass by using
a sodium hydroxide (NaOH) solution in a heated autoclave. Smaller amounts
can be incorporated into the glass by placing pellets formed by adding NaOH
to a glass-clay mixture and directly heating in a furnace.
The foaming process with carbon black was examined by analysis of the
density, pore size, and open porosity of the foamed piece. Also, the
addition of clay made the foam glass less soluable to water.
This report was submitted in fulfillment of Grant No. R800937-02 by the
University of Utah under the sponsorship of the Environmental Protection
Agency. Work was completed as of August 1975.
-------
TABLE OF CONTENTS
Page
DISCLAIMER ii
FOREWORD iii
ABSTRACT iv
LIST OF TABLES vi
LIST OF FIGURES viii
ACKNOWLEDGEMENTS x
I. CONCLUSIONS 1
II. RECOMMENDATIONS 3
III. GENERAL INTRODUCTION , 4
IV. WATER AS THE FOAMING AGENT 6
A. Experimental Procedure 8
B. Results and Discussion 31
V. CARBON AS THE FOAMING AGENT 41
A. Experimental Procedure 44
B. Results 46
C. Discussion 56
VI. CALCIUM CARBONATE AS THE FOAMING AGENT 60
A. Experimental Procedure 61
B. Results and Discussion 63
VII. SUMMARY OF FOAMING METHODS 78
VIII. MARKET POTENTIAL OF FOAMED WASTE GLASS 79
IX. SLAB PRODUCTION 102
X. PELLET PRODUCTION 116
REFERENCES ..... 124
-------
LIST OF TABLES
Table Page
1
2
3
4
5
6
7
8
9
10
n
12
13
14
15
16
17
Autoclave Reaction of Glass with Water
Data of Water Absorbed at Different Temperatures,
Time and Water Added
Effect of Hydroxyl Ion Concentration on Absorbed
Water
Data of Water Absorbed for Different Amounts of Water
and One-Normal NAOH Solution Added
Data of Water Absorbed for Different Concentrations
of NAOH Solution
Water Absorbed vs. Time, Temperature, and NAOH Con-
centrations
Carbon Blacks Examined for Foaming Glass
Summary of Foaming Facts
Openings of Pores with Increasing Foaming Time . . .
Properties of Foamed Glass Pellets (Foamed at
850°C for 5 minutes with 2% CaC03)
Clay Additives
Effect of Particle Size of Glass on Foaming ....
Comparative Values of Selected Physical Properties
of Insulation: Rigid
Western Market for Non-Residential Rigid Roof In-
sulation
National Market for Non-Residential Rigid Roof
Insulation
Comparative Values of Selected Physical Properties
of Concrete Aggregate
Light Weight Concrete Aggregate Market for Vermi-
culite and Perlite
11
13
20
20
23
25
43
48
64
71
74
74
83
85
85
86
95
VI
-------
LIST OF TABLES (Continued)
Table Page
18 Value of Light Weight Aggregate Market 95
19 Western Market for Lightweight Concrete Aggregate . . 99
20 National Market for Lightweight Concentrate Aggregate 99
21 Cost of Production Summary - Slab Production 101
22 Cost of Production Summary - Pellet Production .... 101
23 Major Items of Equipment - Slab Production 103
24 Equipment and Plant Cost Summary - Slab Production . . 105
*
25 Estimated Capital Cost - Slab Production 110
26 Estimated Annual Operating Costs - Slab Production . . Ill
27 Direct Labor Requirements - Slab Production 115
28 Major Items of Equipment - Pellet Production 117
29 Equipment and Plant Cost Summary - Pellet Production . 118
30 Estimated Capital Cost - Pellet Production 120
31 Estimated Annual Operating Costs - Pellet Production . 121
32 Direct Labor Requirements - Pellet Production 123
vn
-------
LIST OF FIGURES
Figure Page
1. The effect of water content on the water absorbed
in the corroded glass at 175°C 15
2. The effect of water content on the water absorbed ,fi
in the corroded glass at 200°C ....
3. The effect of water content on the water absorbed
in the corroded glass at 225°C 17
4. The effect of water content on the water absorbed
in the corroded glass at 425°C 18
5. The effect of time and water content on the water
absorption of -35 + 60 mesh glass at 225°C in
saturated steam 19
6. The effect of water or a IN NaOH solution content
on the water absorbed in the corroded glass . . 21
7. The effect of water or a IN NaOH solution content
on the water absorbed in the corroded glass ... 22
8. The effect of NaOH concentration on water ab-
sorbed 24
9. The effect of NaOH concentration on the thickness
of the glass rod reacted 27
10. Water absorption of -100 + °° mesh glass particles
autoclaved in a 80% solution of NaOH for various
times. The insert shows the temperature variation 28
11. Water absorption of -100 + °° mesh glass particles
autoclaved at 150°C for various times 29
12. Reaction rate vs. temperature for 5N - NaOH cata-
lyzed water-glass reaction 30
13. Pore size as a function of time at 850°C ... 49
14. Pore size as a function of time at 850°C ... 49
15. Density as a function of time at 850°C for S-315
and R-1040 carbons 72
viii
-------
LIST OF FIGURES (Continued)
Figure Page
16. Density as a function of time and carbon con-
centration at 850°C 52
17. Density as a function of time and carbon con-
centration at 850°C 53
18a. Water absorption of 0.5% R-1040 carbon black
foamed glass as a function of time at 850°C . 54
18b. Water absorption of 0.5% S-315 carbon black
foamed glass as a function of time at 850°C . 54
18c. Water absorption of F-l carbon black foamed
glass as a function of time at 850°C 55
18d. Water absorption of U-3024-L carbon black
foamed glass as a function of time at 850°C . . 55
19. Cell size versus percent calcium carbonate . . 65
20. Cell size versus foaming time 66
21. Firing time versus % CaCO, vs. temperature
Samples expanded 6 times fts volume 68
22. Volume increase vs. CaCO~ size for 10-20p
cullet at 700°C for one hour and 0.5% CaC03 . . 69
23. Large scale sample foaming schedules (CaCO.,
type) 70
24. Percent weight loss (solubility) versus
percent CaC03 73
25. Solubility change of foam glass in water with
clay additives 75
26. K-factor of CaC03 foam vs. density 76
27. K-factor of CaC03 foam vs. cell size 76
28. Weight spectrum of lightweight concretes ... 90
IX
-------
ACKNOWLEDGEMENTS
The authors wish to thank the following for their assistance:
Mark Olson, Keith Ramsay, Dale Hoggard
-------
I. CONCLUSIONS
1. Waste glass without any special treatment such as cleaning and
color sorting can be used to form a good quality foam glass.
2. The rate determining step of a water-glass reaction is the build-
up of the hydroxide ion (OH) concentration on the glass surface.
3. The rate limitations of the build-up of hydroxide ions on the
surface can be avoided by placing the glass in an alkaline solu-
tion.
4. Acidic solutions reduce the rate of reactions for incorporation
of water.
5. Waste glass with water concentrations in excess of 10% by weight
can consistantly be achieved by autoclaving from 200-275°C in the
presence of hydroxide ion in water. The concentration necessary
was found to be from 1 to about 10 normal NaOH solution.
6. Two percent water can be incorporated into glass by spraying a
5N solution of NaOH onto 325 mesh glass particles and heating at
200°C for 30 minutes.
7. Water impregnated glass can foam glass particles as small as 177
microns.
8. In order to achieve closed pores throughout the entire foamed
body with water as the foaming agent, the piece to be foamed must
be small (65 mesh to .64cm) or the temperature must be reduced
sufficiently so that foaming will not take place only on the
outside. Preheating at 700°C for a few minutes will help make
uniform pores.
9. Foamed bodies from .16cm to 5 cm (2 inches) thick can be obtained
by using carbon black as the gasifying agent.
10. The pH, surface area, and absorbed gases of a carbon black are of
major concern in selecting the optimum carbon black for the
gasifying agent.
11. A larger concentration of carbon and/or a reducing atmosphere will
yield smaller pore size.
12. Higher temperatures or use of activated carbons yield larger pores.
13. Densities of foamed discs of carbon, clay and glass decrease to a
minimum with time at constant temperature and then increase.
-------
Typical minimum densities of .12-.15 g/cc (7.48-9.36 lb/ft3) were
achieved. Pore size of .4-.9 mm were common.
14. Pellets of .32-.64 cm (l/8-l/4inch) diameter can be produced in
a rotary kiln by injecting extruded pellets composed of carbon,
bentonite, clay and glass mixed with dry clay. The clay is added
to prevent the adhesion of the foaming particles to one another
and to the furnace. Densities of-.23g/cc (14.3 lb/ft0) with
overall bulk density of 8.5 lb/ft were obtained. This pellet
has a thin crust.
15. Calcium carbonate can be used to produce a bulk foamed product
with .16-.24 g/cc (10-15 lb/ft ) and 1-3 mm pore size. Large
amounts of CaCO- increase cell size.
16. Milled limestone whose particle size was less than 5y gave the
best results of the carbonate foamed glass.
17. For .32 to 5 cm (1/8 2 inch) size foamed objects carbon, rather
than water or CaCOo. yields the lowest density combined with the
lowest open porosity.
18. Clays decrease the solubility of water in glass but it may be
necessary to dissolve them into the glass structure to make them
effective in decreasing solubility.
19. The market price of foamed waste glass as rigid insulation is
lower than other rigid insulation products.
20. The price and quality of foamed waste glass leads to a strong
market potential for foamed waste glass as a rigid insulation
project.
21. The eleven western states market for foamed waste glass as rigid
insulation would be sufficient to sustain two or three 20/ton/day
plants.
22. The national market for foamed waste glass as rigid insulation
would be sufficient to sustain fourteen 20/ton/day plants.
23. The market potential of foamed waste glass pellets as lightweight
concrete aggregate is weak.
24. The eleven western states market for foamed waste glass as a light-
weight concrete aggregate would be sufficient to sustain only one
20/ton/day foamed waste glass plant.
25. The national market for foamed glass as a lightweight concrete
aggregate would be sufficient to sustain five 20/ton/day glass
P I 911 wS *
-------
II. RECOMMENDATIONS
Small pilot production of foam glass pellets should be initiated.
Either carbon or water could be used to obtain low density pellets.
With foamed pellets available to industry the development of a market
will probably take place by logical processes of replacement of part
of the market for perlite.
-------
III. GENERAL INTRODUCTION
The problem of solid waste disoosal is becoming more acute each
year. Glass, which constitutes approximately six percent of all
solid wastes, is one of the few materials that has not been success-
fully recycled. Annually, as much as twelve million tons of glass
are discarded. Unlike other solid wastes, glass presents no threat
to the world's natural resources since the major components of glass
(silica, limestone and soda ash) are very abundant in nature. Salvage
and recycling of all materials, however, offer the only viable, long-
range solution to the waste disoosal problem.
Urban refuse is composed of metallic, organic and ceramic materials.
To make total solid waste recovery feasible and profitable, all three
of these fractions must be recovered and used. The metallic andorganic
fractions are either already valuable and recoverable or can be in-
cinerated to reconvert thermal energy. Ceramics, on the other hand,
have defied potential reuse for a variety of reasons. They will rot
burn, rot, or otherwise disintegrate over a long period of time.
Consequently, the currently available methods of disposal are by
sanitary land fill or as an aggregate in asphaltic concrete. However,
if a new commodity can be manufactured from the waste glass without
expensive color sorting, sizing and cleaning, the possibility exists
of partially eliminating the problem of disposal and at the same time
establishing a profitable new business. Recent research has investigated
the feasibility of foaming waste glass for use as a premium grade in-
sulation in industrial and commercial applications.
-------
The process followed by Pittsburg-Corning, the only producer of
foam glass, cannot be used to foam waste glass. Pittsburgh-Corning
melts the raw materials, along with an oxidizing agent (a sulfate or
sulfites). They add the carbon during the crushing and grinding of
the melted glass. During foaming the carbon is slowly oxidized and the
sulfate is reduced. The processes involved in this report show that
waste glass can be used to form an excellent foamed material without
beginning with the raw materials.
Foaming of glass can be achieved only when there is a proper
balance between the sintering of glass particles, gas generation from
the foaming agent, and the softening of the glass. To have low density,
it is necessary to avoid both the collapsing and coarsening of indiv-
idual pores. Collapsing of pores arises when the viscosity of glass
is too low or when there is excess gas available for foaming. Coar-
sening of pores is a natural consequence of surface tension that
drives a porous system to reduce its surface area. The free energy
increase of a glass due to pore generation is
AF = y ' A
where: y = surface tension of glass
A = total interfacial area of pores.
To have a stable foam minimal coarsening should take place at the
foaming temperature. The foaming agents used in this study are water,
carbon black and CaCOg.
-------
IV. WATER AS THE FOAMING AGENT
Because gases, such as water, contained in melted glass form
bubbles that interfere with the transparency and strength of glass,
there has been careful study done by scientists regarding glass
manufacturing. These studies have been directed toward processes
for eliminating gases dissolved in glass.
An extensive review of the literature describing water in
to]
glass has been made by Scholze. v ' From this review it can be
concluded that: (1) water is soluble in glassd (2) its solubility
decreases as temperature increases; (3) water forms hydroxyl groups in
glass; (4) water dissolves to form two different tvpes of hydroxyl
groups — one associated with silicon-oxygen-silicon framework and
another type of hydroxyl group associated with the sodium ions and
other modifier ions. Water solubility in glass can greatly reduce
(3)
glass viscosity. ' Combining these concepts one can visualize the
incorporation of water into the glass at temperatures below the glass
point and then during reheating, drive the water off as steam. As the
water escapes from the outside of the glass particle .the glass viscosity
rises immediately as it returns to its dehydrated initial state. The
water trapped inside the more viscous envelope then expands the particle
until the surface tension of the glass is in equilibrium with the
internal steam pressure.
To make the above concept viable for an industrial process, the
kinetics and equilibrium water content of glass as a function of water
vapor pressure and temperature must be investigated. It has long been
6
-------
known that water corrodes the surfaces of glasses. Information about
the nature of the surface attack and changes in the composition of the
corrosive liquid are helpful in evaluating the autoclave process for
water absorption.
As Charles'1 ' reports, when water comes in contact with soda-lime
glass the following reaction takes place:
-Si-0-[Na] + H90 -> -Si-OH + Na+ OH" [H
I J
This reaction is a typical hydrolysis reaction. A free hydroxyl
ion is formed in the process and the second important step in the
glass corrosion takes place as
II. I I .
-Si-O-Si- +OH -> -Si-OH + -SiO [2!
II II
Reaction [2] can occur only after the reaction [1] has already taken
place. In reaction [21 the very strong Si-O-Si bond is broken and
gives rise to another active group, which is capable of reacting with
water as
-SiO" + H20 -» -Si-OH + OH" [3]
It should be noted that once a hydroxide ion is formed or is provided,
reaction [1] is no longer necessary to propogate the water reaction.
The overall reaction between water and glass may be written as:
-Si-O-Si- + H90 -»• 2[-Si-OH] [4]
II2 I
These concepts are consistent with the data of Charles ^ ' who
found that a rather large amount of water could be diffused into glass
-------
at relatively low temperatures and pressures. He followed an interface
between the diffused and undiffused volume of the glass. This boundary
between the diffused and undiffused region appears to proceed linearly
rather than parabolically as would be expected for a diffusion process.
Rana and Douglas ^ ' explain that when typical soda-lime-silica glasses
react with water the initial rate of reaction varies with the square
root of time, suggesting a diffusion-controlled process. Moreover,
the activation energy of the reaction is about the same as that for the
diffusion of sodium ions in the glass. However, it has been emphasized
that this is merely convenient ^ and not to be intended as an assess-
ment of the actual mechanism involved.
A. Experimental Procedure
To determine the effect of pressure and temperature on water
at various times a single autoclave was used. After receiving a charge
of ground glass in steel crucibles, stacked one over the other, the
autoclave was sealed, lowered into a Kanthal wound electric furnace
and heated to the desired temperature. The amount of water absorbed
in the powders was determined. The glass near the top of the crucible
was found to contain more water than the glass near the bottom. The
glass at the top of the crucible was hard and dense, whereas near the
bottom the glass powder was not sintered appreciably. To eliminate
the problem, a crucible was made of wire mesh to allow free passage of
steam in all cases except for extremely dense sintering. The crucible
was hung above the water level so that only steam reacted with the
glass powder. Water did not react with the glass in a reproducible
manner.
8
-------
It was reasoned that perhaps water condensation during heat up
may be the problem. It was hoped that a twin autoclave system would
yield reproducible results because dry steam could be assured and the
reaction could be followed from zero time. One autoclave contained
the glass sample while the other one contained water. After the test
temperature was reached, the steam was allowed to enter the sample
autoclave. Of most interest were the experiments at 225°C for several
hours. The samples contained only 0.3% water. This aroused the sus-
(A\
picion of Charles ^ ' data.
Seven closed end tubes of soda lime glass and borosilicate glass
of o.d. = 8.00 mm to 20.00 mm and i.d. = 5.8 mm to 17.2 mm and 30 cms
in length were prepared. Soda-lime glass rods (o.d. = 4.00 mm) were
placed inside each one of the tubes. These tubes containing rods were
hung with the help of a stand so that the tubes did not touch each
other. Each tube was filled with water up to a height of 15 cms and
the remaining parts of the tubes and rods were to be in contact with
steam. All the tubes were covered with a lid in order to prevent
condensed water from entering into the tubes. This package of tubes
and rods was placed in a 2.5 liter autoclave which had 400 ml of
distilled water to generate steam and to keep the bottom part of the
tubes in water. The autoclave was closed and the thermocouple
(chrome!-alumel) and the pressure gauge were fixed to read the tempera-
ture and pressure.
The autoclave was heated in a Kanthal-wound electric furnace and
*
heated to produce saturated steam at 250°C and was maintained for 6
hours. The furnace was shut off and the autoclave was cooled in the
-------
furnace. After cooling, the package of tubes and rods was taken out
and the o.d. and i.d. of the reacted rods and tubes were measured at
the bottom part which was in contact with water and the top part which
was in contact with steam. The observations made and the results
obtained are tabulated in Table 1. The measurements were easily made
because it was observed that the reacted glass on the surface of
the rods and tubes peeled off like flakes, leaving a transparent
interior. This was used as the boundary between the reacted and
unreacted glass.
The most direct and simple method of determining water content of
water reacted glass was to measure weight loss after heating. The
reacted glass was dried at 115°C for a long time (- 10 hrs). One or
two grams of reacted glass were put into a porcelain crucible and
weighed before and after. The weight was determined within + 0.0001
grams on a single pan balance. The glass was then heated slowly to
900°C, maintained for 30 minutes, and slowly cooled to prevent crucible
breakage. The crucible and glass were then re-weighed and the difference
between the initial and final weighings was assumed to be due to the
water being driven off. The amount of water absorbed in the flakes
was found to be 9.8% on a dry basis.
Initial foaming experiments were done in one inch diameter cru-
cibles which contained some reacted glass flakes and then put directly
into a hot furnace at 900°C and kept for 15 seconds. The flakes ex-
panded or popped to a frothy mass of low density material and the
foamed product looked like commercial perlite.
10
-------
Table 1.
AUTOCLAVE REACTION OF GLASS WITH WATER
Original Size
Sample
Number
1
2
3
4
5
6
7
Tube
OD(mm)
8.0
9.0
9.9
11.0
11.8
15.0
20.0
ID(mm)
5.8
6.6
7.5
8.6
9.4
12.8
17.2
Rod
OD(mm)
4.1
4.0
4.0
4.15
4.15
4.20
4.0
Water
(gm/cm^ of
Glass Surface)
0.0425
0.065
0.0875
o.m
0.131
0.216
0.33
Glass Thickness Reacted
On OD
Top
0.40
0.40
0.50
0.40
0.45
0.00
0.00
of Tube
Bottom
0.00
0.00
0.00
0.00
0.00
0.00
0.00
On OD
Top
0.45
0.45
0.40
0.40
0.45
0.45
0.45
(mm)
of Rod
Bottom
0,35
0.10
0.10
0.10
0.10
0.00
0.00
NOTE: Samples 6 and 7 were pyrex glass. All other glasses were soda-lime-silica.
-------
Common soda-lime glass obtained from crushed clear bottles was
used. Although not analyzed specifically, this glass would be expected
to have composition similar to the composition of any soda-lime glass.
Size reduction of the glass was done by ball milling in a rubber lined
mill using steel balls.
In order to determine if sufficient water could be absorbed by
glass, the absorption of water was studied as a function of time,
temperature, and water in contact with glass powder.
Soda-lime glass powders of -35 + 60 and -100 +200 sieve sizes
with 10, 20, 40, 80 and 160% added water (on a dry basis) were reacted
at different temperatures. For this purpose glass powders and water
were enclosed in small steel bombs (black 3/4" diameter nipple with
caps on both sides) and put together in a large autoclave having 400
.ml water in it, to avoid any pressure difference between the large
autoclave and steel bombs. This arrangement provided good thermal
conductivity. The large autoclave containing individual steel bombs
was introduced into a preheated furnace and the desired constant
temperature was attained in 30 minutes. After heating for the desired
time, the furnace was switched off and the autoclave was allowed to
cool in the furnace, which took 30 minutes to cool down to 100°C.
After cooling to room temperature the steel bombs were opened and
the glass powders were dried for a long time (^ 10 hours) at 115°C.
The reacted water absorbed was found in the same way as in the case of
the previous experiment. The percent of water absorbed was calculated
on a dry basis. The observations obtained are tabulated in Table 2.
12
-------
Table 2.
DATA OF WATER ABSORBED AT DIFFERENT TEMPERATURES, TIME AND WATER ADDED
Size
35 +60
•
100 +200
Added
Water
%
10
20
40
80
160
10
20
40
80
160
200° C
6 hours
1.88
2.18
1.46
0.51
0.54
5.57
6.01
2.15
1.10
1.28
%
225°C
2 hours
4.12
4.50
0.75
1.10
0.94
3.99
4.80
2.28
2.14
1.45
Water Absorbed
175°C
6 hours
0.26
0.21
0.21
0.33
0.39
2.15
1.26
0.66
0.48
0.38
(dry basis) at
225° C 175°C
4 hours 4 hours
5.00
5.83
2.04
0.76
0.65
2.01
1.04
0.49
0.40
0.55
425°C
4 hours
3.79
7.20
8.28
8,07
--
3.42
4.20
7.45
7.76
-------
This is further illustrated in Figures 1 through 5, showing results
with glass powders where water content, temoerature and time of reaction
were variables.
Further to confirm the increased rate of reaction reported by Das
and Douglas ^ ' due to hydroxyl ions concentration, glass powder (-100
+200) was added with 160% water in three different steel bombs. One was
left as is, a few drops of sulphuric acid were added to the second and
a few drops of NaOH solution were added to the third. All these were
autoclaved at 225°C for 6 hours. The water absorbed was found to
increase with hydroxyl ion concentration, as shown in Table 3.
The results of Table 3 indicated the importance of high pH in the
corrosion process and the need for further study of the effect of
alkali. Further experiments were carried out in the same type of
steel bombs with different amounts of a one-normal solution of NaOH.
The observations are shown in Table 4 and drawn in Figures 6 and 7.
In order to find the minimum concentration of NaOH solution
required for the maximum water absorbed, 80% of NaOH solution of dif-
ferent concentrations were added to -35 +60 size glass powder in steel
bombs and autoclaved at 225°C for 4 hours. The results obtained are
shown in Table 5 and drawn in Figure 8.
The percentage of NaOH solution was then varied with respect to
normality and these results are tabulated on Table 6.
With the fact established that NaOH increases the water absorp-
tion of glass, it was then necessary to observe in detail the kinetics
of the reaction. Glass rods were allowed to react for varying amounts
of time and temperature in an autoclave. This data is plotted on
14
-------
- 5
to
o
Q
iLl
CD
§3
CO
00
tr
I75°C
6 hrs
-35 + 60 MESH
-100+200 MESH
0
20
40
60 80 100 120
% WATER ADDED (dry basis)
140
160
180
Figure 1. The effect of water content on the water absorbed in the
corroded glass at 175°C.
-------
CTi
in
'w
o
.a
Q
LJ
CD
CO
CD
200°C
6 hrs
-35 + 60 MESH
-100 + 200 MESH
-D
20
40
60 80 100 120
% WATER ADDED (dry basis)
140
160
180
Figure 2. The effect of water content on the water absorbed in the
corroded glass at 200°C.
-------
.52
'
o
.a
Q
LU
CD
CO
CD
uj 2
225°C
2hrs
-35 + 60 MESH
-100 + 200 MESH
I
0
20
40
60 80 100 120
% WATER ADDED (dry basis)
140
160
180
Figure 3. The effect of water content on the water absorbed in the
corroded glass at 225°C.
-------
lOr
8
00
GO
i — i
c/o
-------
<£>
% WATER ADDED
10 12
TIME (hours)
Figure 5. The effect of time and water content on the water absorption
of -35 + 60 mesh glass at 225°C in saturated steam.
-------
Table 4.
DATA OF WATER ABSORBED FOR DIFFERENT AMOUNTS
OF WATER AND ONE-NORMAL NAOH SOLUTION ADDED
% % % Water Absorbed (dry basis)
Water Added at 225UCaTT75°C
Added IN-NaOH 4 hours 4 hours
Solution -35 +60 size -100 +200 size
10 0 5.00 2.01
20 0 5.83 1.04
40 0 2.04 0.49
80 0 0.76 0.40
160 0 0.65 0.55
0 10 6.05 6.14
0 20 9.73 9.87
0 40 10.11 10.54
0 80 10.66 10.32
0 160 10.08 10.86
Table 3.
EFFECT OF HYDROXYL ION CONCENTRATION ON ABSORBED WATER
Added Reagent H2S04 None NaOH
% Water Absorbed 2.37 7.70 13.55
20
-------
INJ
o
.o
12,
~ 10
8
Q
LU
CD
g 6
(O
CD
a:
UJ 4
I75°C
4 hrs
-100•»• 200 MESH
WATER
NaOH SOLUTION
20 40 60 80 100 120 140
% WATER OR NaOH SOLUTION ADDED (dry basis)
160
180
Figure 6. The effect of water or a IN NaOH solution content on the
water absorbed in the corroded glass.
-------
12,
rv>
ro
~ 10
w
'55
o
-O
•o 8
Q
UJ
DO
§ 6
CO
00
a:
UJ 4
A
A
225°C
4hrs
-35 + 60 MESH
WATER
NaOH SOLUTION
20 40 60 80 100 120 140
% WATER OR NaOH SOLUTION ADDED (dry basis)
Figure 7. The effect of water or a IN NaOH solution content on the
water absorbed in the corroded glass.
160
180
-------
Table 5.
DATA OF WATER ABSORBED FOR DIFFERENT CONCENTRATIONS
OF NAOH SOLUTION
Sample
No.
1
2
3
4
5
6
7
8
9
10
Normality
of
NaOH
Solution
0.00
0.05
0.10
0.50
1.00
2.50
5.00
10.00
15.00
20.00
% Water Absorbed (dry basi
at 225°C, 4 hours, in
-35 + 60 size
when 80% solution added/dry
0.768
0.824
0.966
10.91
10.85
14.75
8.48
7.24
6.97
5.98
s)
basis
23
-------
ro
225°C
4 HOURS
-35 +60 MESH
807 SOLN. ADDED
/o
4 6 8 10 12 14
NORMALITY OF NoOH SOLUTION
Figure 8. The effect of NaOH concentration on water absorbed.
-------
IV
en
Table 6.
WATER ABSORBED VS. TIME, TEMPERATURE, AND NAON CONCENTRATION
Sample
No.
1
2
3
4 :
5
6
7
8
9
Normality
NaOH
1
1
1
2
2
2
4
4
4
Solution
Added
(Dry Basis)
40
80
160
40
80
160
40
80
160
300° C
2 hrs
9.14
9.66
8.98
9.30
9.51
9.88
8.34
8.30
12.27
260°C
2 hrs
10.12
10.01
9.32
9.82
9.52
10.23
8.75
9.33
11.75
225°C
2 hrs
12.80
14.24
15.23
14.01
14.10
14.25
15.67
20.92,
7.66
225°C
1 hr
13.12
14.61
--
16.27
12.54
. __
16.19
17.30
- - _ _
175°C 150°C
2 hrs 1 hr
14.18
23.17 12.3
15.38
15.37
19.74 11.9
13.60
18.61
22.44
14.93
-------
Figure 9. In order to get the water absorption for shorter times a
thermocouple was attached to a steel bomb with -100 +°° mesh powdered
glass and 80% 5N NaOH solution contained inside. It was lowered into
a clay pot in the center of the furnace that was preheated to 200°C.
Other bombs were stacked on top of the temperature monitored bomb and
removed one by one at the indicated times. As can be seen from Figure 10
the temperature fluctuation of the furnace closely paralleled the water
absorption of the glass. The same process was used again but at 150°C
and 80% 5N NaOH. This time, in Figure 11, the water absorption rate was
much slower. Reaction rate vs. temperature for 5N NaOH catalyzed water-
glass reaction is shown in Figure 12. With the activation energy of
20+4 kcal/mole given by R. Charles the time required for the reaction
can be determined for a given temperature (Figure 11).
Following this study the glass powder (-100 + °° mesh) was autoclaved
in 250g quantities contained in tin cans or carbon cloth. These powders,
during autoclaving would settle out and form a hard conoact when 100%
solution was used (a 200% solution would leave a mushy compact). This
compact was broken down into various sized particles -28 to +35 mesh,
-35 +48, -48 +80, -80 + °° and -325 + «. These particles were placed
separately in a furnace at 900°C to 1000°C and fired for times varying
from 15 seconds to 90 sec. The particles were also pressed in com-
pacts and fired at 900-1000°C for various times. Large pieces were
broken from the compacted mass and fired at different heating schedules.
The foaming characteristics were observed.
10 grams type "S" lime composed of 38%/Mg (OH), 57%, Ca (OH)2,
1% MgO was also added to 500 grams of -100 + « glass powder with 500
ml water and autoclaved. The autoclaved glass was then separated into
26
-------
50
ro
40
O
H
O
<
LU
cr
0)
o
X
30
20
10
0
0
FLINT, BOTTOM
225°C, 8hrs.
FLINT, BOTTOM
225°C, 2hrs.
FLINT, BOTTOM
I50°C, 8hrs.
6 8
NORMALITY OF NaOH
10
12
14
Figure 9. The effect of NaOH concentration on the thickness of the
glass rod reacted.
-------
ro
oo
14
12
10
z
UJ
I-
§ 8
o
cc
Ul
I 6
DESIRED TEMP.
ATTAINED
0
30
60 90
TIME (min.)
120
15
30
45
60 75
TIME (min.)
90
105
120
-o
150
135
Figure 10. Water absorption of -1.00 + °° mesh glass particles autoclaved
in a 80% solution of NaOH for various times. The insert
shows the temperature variation.
-------
IN)
IQ
12
10
Z 8
LJ
o
o
LJ
<
DESIRED TEMP.
ATTAINED
15
30
200
O
o
LJ
OC
150
I- 100
LU
Q_
2
LU
50
0
20 40 60
TIME (min.)
45 60
TIME (min.)
75
90
105
80
120
Figure 11. Water absorption of -100 + «> glass particles autoclaved at
150°C for various times.
-------
CO
o
0
O -2
O
-3
-4
1.4
352
TEMPERATURE (°C)
283 227
1.8 2.0
I/T X I03 (°KH)
181
2.2
143
0.17
1.7
17
c
1
LjJ
53
2.4
Figure 12. Reaction rate vs. temperature for 5N - NaOH catalyzed
water-glass reaction.
-------
small particle sizes. Also, some pellets were formed by squirting a
fine mist of 5N solution of NaOH into a rotating mixer which contained
250gm of 325 mesh glass powder and 4% bentonite. The pellets rolled
themselves into various sizes. These pellets were put into an aluminum
foil container inside the autoclave with 350 ml of water and heated
to 200°C. The autoclave was allowed to cool overnight. The samples
were then fired at 900°C which gave uniform pores. At 1000°C uneven
pore size distribution was evident.
Because such a small amount of water was placed in the bottom of
the autoclave, serious thought was taken that perhaps not all the
pressure or water vapor was needed. The pellets were made as before
but this time they were placed directly into a 200°C furnace for 1/2
hr to allow the water-glass reaction to take place. Then they were
fired at varying temperatures and times. Pellets were also produced
by extrusion but had poorer quality.
B. Results and Discussion
A range of temperatures and pressures would have been investi-
gated but discouraging results in foaming the product brought the
experiments to a narrow range of temperatures restricted to low tem-
peratures by sintering and diffusion constraints. The glass particles
begin to sinter at about 600°C under normal conditions. Sintering
during reaction with water is undesirable because of the associated
reduction in surface area. Therefore, maximum allowable temperatures
were of the order of 250°C in a water vapor atmosphere because of the
(3)
sintering enhancement due to water vapor.v ' Below 175°C the dif-
fusion time became increasingly longer.
31
-------
The results in Table '1 confirm the data of Charles that the glass
reacts faster in steam than mwater at the same temperature and pres-
sure. The penetration rate was also equal* to that given by Charles. '";
Initial penetration or surface layer formation took,approximately 2 ••• -;
hours to form a 25 \im thick layer and then .in the second stage 425-urn
is reacted in 4 hours, which corresponds well with Charles'M20 ym/hr
at 250°C ,in saturated steam.
The higher rate of reaction in the glass/vapor phase is* explained •
by Charles-as due to a-pH increase,at glass/vapor interface inrcontrast
to a dilution of pH at the liquid/glass, interface.-;i Higher, concentra- ';
tion of hydroxyl ions at the vapor/glass;interface break silica network-
bonding according to reaction [21. As the corrosion,on-tube numbers 6 i
and 7'of pyrex glass (Table .1) leads to the conclusion that dissolution
of fused silica and borosilicate glasses in water or steam alone is-
difficult even at relatively high temperatures;- .This may be .due to <
lo\
the special phase separated structure of these glasses. V •' These •*-
o
glasses have a silica matrix with a very'fine (20 to,30 A diameter)
second phase of sodium borosilicate. The matr.ix or, low .soda .content:
appears to be resistant to water corrosion by suppression of reaction
[1].
- The sharp boundary obtained between the reacted and unreacted
glass suggests the corrosion reaction can occur at:an interface., As the
hydroxyl ion concentration builds up, the reaction proceeds more and1
more rapidly.;>- Kinetically the reaction can be separated into :two
stages. H In the first, the 'amount -Of water absorbed or thickness
32
-------
reacted is proportional to the square root of time, whereas in the
second the amount of water absorbed is proportional to time. The
rate of the first stage is controlled by interdiffusion of hydrogen
and alkali ions, and the second involves a surface reaction.
Table 2 summarizes the data on water absorbed in glass powders
at different temperatures with different amounts of water in contact.
Figure 1 illustrates the weight gain of glass at 175°C in 6 hours.
The weight gain is small and the rate gradually tapers off. Even
smaller size powder has absorbed little water, not sufficient to foam
the particles. These data demonstrate the importance of higher tempera-
tures, which leads to Figures 2 and 3. At 200°C the water absorbed
is about 6% when 20% water is in contact in 6 hours. Still at higher
temperatures (225°C) the water absorbed was 5% in only 2 hours, whereas
at very high temperatures (425°C) the water absorbed increased to 8%
(Figure 4). At high temperatures the solubility limit is probably
reached quickly. Increased pressure simply raises that solubility
limit. When the temperature is too low, the solubility limit is high
but the rate is too slow for equilibrium to be reached in these experi-
ments. It should be noted that rapid corrosion required small water
contents, high surface area and, of course, proceeded more rapidly at
higher temperatures and pressures. Here the amount of water in contact
with glass powder determines the hydroxyl ion concentrations and,
consequently, the rate of breaking the silica chains and water absorbed.
The results of Table 3 confirmed the necessity for higher pH for
rapid reaction and need for the further study of the effect of alkali
concentration in contact with glass powder.
33
-------
The results of Figures 6 and 7 make it evident that glass reacts
faster with water when the hydroxyl ion concentration is high. Glass
is more rapidly attacked in alkaline solutions than in neutral or acid
solutions because the alkali supplies hydroxyl ions for reaction [2]
with the silica network and gives rise to more silanol groups. From
Figures 6 and 7 it seems that there is a critical amount of NaOH solu-
tion necessary for the highest rate of reaction, above which the rate
of reaction becomes constant. Addition of more NaOH solution does not
increase water absorption. This may be due to the saturation point
reached in the glass structure to accommodate the maximum number of
hydroxyl groups.
Glasses react at a very slow rate with acids. Attack of glasses
by acids differs from that by water in that any alkali dissolved is
neutralized by the acid. Also, the alkali and basic oxide components
may be preferentially dissolved, leaving a silica surface layer which
reduces the rate of attack with time. Presence of a protective surface
(9)
film was found by Brueche and Poppa v ' having a thickness of - 0.1 u.
Reaction [1] leads to a weakened network through which large
ions can rapidly diffuse. The results of Eisenman ^ ' are consistent
with this requirement, since he found that in the hydrated layer
Na diffused only about 5 to 10 times faster than K , whereas in the
dry glass the Na is about 1000 times more mobile than K+.
This can be explained on the ion-exchange mechanism. As the ion
exchange proceeds, the surface layer becomes more open and the water
molecule can diffuse in it more rapidly, thus increasing the rate of
reaction of water with the silica lattice. Also, the silicon-oxygen
34
-------
bonds in the surface layer may be more reactive after ion exchange
because of the strain resulting from the replacement of a larger ion
(Na ) with a smaller ion (H ). Therefore, the water-glass reaction is
an accelerating process which, at least initially, can be related to
Na ion diffusion in the bulk glass.
Figure 8 makes it clear that there is one critical strength of
NaOH solution which will allow maximum water to be absorbed in the
glass at the highest rate of reaction. Perhaps higher concentration
of NaOH solution forms a protective film or stops the Na ion exchange
and stops or slows the reaction from going further. In either case,
it is quite evident that there is an unreacted core at the center of
the glass powders which was confirmed by breaking the particles and
looking at the broken pieces under the microscope.
Even higher amount of weight loss of samples 9 and 10 may be due
to the NaOH dried up on the surface of the particles, which lost
water on heating at 900°C and gained hydroscopic water when kept in
atmosphere for a few hours.
Regardless of pressure, temperature, amount of water, or NaOH
solution, or size of glass powder, the results were, with minor varia-
tion, essentially reproducible.
There are complications to the simple model of water incorporation.
No doubt some sodium is leached out, and some of the water, sodium
and silica is tied up in crystalline compounds. This is possibly why
the activation energies measured by Charles do not correspond to sodium
diffusion or glass corrosion energies. From Table 6 it can be seen
that varying the NaOH concentration from 1 to 4 Normal along with
35
-------
varying the amount of the solution, didn't effect the results to any
great extent. At 300°C for 2 hours the water absorption was approxi-
mately 10%. At 225°C the water absorption was about 14%* It ,can
be seen from this that the temperature of autoclaving has a larger
effect than changing normality from 1 to 4 or changing the pressure
(1246-psia at 300°C and 343-psia at 225°C).
From Figure 9 it was found that the thickness of the redacted :
glass increased as the temperature increased as would be expected by
rate of reaction control. It is interesting to note that concentrations
above 8N NaOH have no additional effect on the rate of reaction. This
could be explained by the NaOH forming a protective coat on the glass,
or that a saturation point of hydroxide groups that can be accomodated,
in the glass structure has been reached.
Figure 10 and its insert show that within 15 minutes after the
samples reached 200°C the reaction had ended. As the temperature
rises and falls so does the water content. This shows how sensitive
the reaction is to temperature fluctuations. Since there isn't a .wide
separation of temperatures involved in the reaction, the rate dependence
on temperature should be related as •-..?,-.-
k=Ae-Ea/RT
By taking Charles' data that Ea = 20 + kcal/mole and log k = -Ea/(2.303RT)
+ A the rate of the reaction can be determined. From Figure 11 it is
seen that the rate is essentially linear, giving k proportional to
^ ; where t is the time for the reaction to take place. The reaction
time at 150°C in 5N NaOH is 50.25 minutes or 3015 seconds, k then is
36
-------
-4 -1
proportional to 3.317 x 10 sec . Using this point and the slope of
-20,000/(2.303R) a line can be constructed that shows the rate of
reaction at various temperatures using 80% of a 5N NaOH solution. This
graph appears on Figure 12.
When the autoclaved glass powder with 5-14% water was spaced so
that the individual particles didn't touch on a stainless steel tray it
was found that glass particles smaller than 80 mesh did not foam.
This was probably due to the water being immediately driven from the
small partjcles before a viscous covering could form on the outside of
the particle to trap the water closer to the center.
Particles larger than 80 mesh would foam enough to float. The 35
mesh particles produce from 5-7 pores within their structure. After
foaming was completed for the differing water content of 5 and 14%, it
was observed that there wasn't much difference in the final foamed
particles. Autoclaved glass of about 325 mesh size was sprinkled .32
to .64 cm (1/8 - 1/4 inch) deep on a tray and fired at TOOO°C. It was
observed that it would foam uniformly. When the sample was thicker
foaming would begin on the outside and leave a dense inner core. At
850°C powders pressed at 7000psi and .64 cm (1/4 inches) thick didn't
foam at all.
The hard sedimented autoclaved glass was also used intact to pro-
duce a foam. The autoclaved glass was broken into large pieces and
fired. At 850°C a 1.28 cm (1/2 inch) thick piece was fired for TO
minutes and relatively uniform pores of 1 mm were formed. However at
900°C the foaming action was observed to be .much more rapid. The density
37
-------
of .28 g/cc was obtained with an open porosity of 60%. It was noted
that even though there were large pores inside, a crust was formed on
the outside that made the piece relatively impervious to water.
On further tests it was noted that there was probably too much
water absorbed into the glass to give good results and that good foaming
would occur on the outside while the inside would remain dense. To
verify this, two identical samples were obtained. One was placed into
a furnace for 10 minutes at 700°C. Then both samples were placed
directly into an 850°C furnace for 7 min. The foamed pieces were
observed and noted that the pre-fired sample had 1/2 the pore size of
the sample that was not prefired. This could be explained because of
the lower water content during the final firing. The density obtained
for the prefired sample was .46g/cc.
An autoclaved lime-glass mixture at 200°C yielded a water absorbancy
of 12%. This further verifies the fact that the hydroxide ion is a
necessary step in the water glass reaction. Small autoclaved glass
particles (-35 +48 mesh) would yield foamed particles that would float
when fired at 900°C for 30-45 seconds. However, even at higher tempera-
tures a pressed disc wouldn't foam.
When the 5N NaOH mist was used to pelletize .62 cm pellets from
glass-clay mixture in the rotary mixer, the water absorption was found
to be 12% in the pellets that were autoclaved-, while the 5N pellets
not autoclaved, but heated to 200°C, had 2% water absorption. It was
found that both products would foam. At 850°C the pores of the 12%
glass pellets were about twice as large as the 2% with a density of
38
-------
.29g/cc. The 2% glass had a density of .38g/cc for the same time of 2 min
in the furnace. The crust of the 2% glass was denser and thicker than
the 12%. At 900°C the pores of 12% were smaller than the 2% for the
same firing time. The 12% glass had a bulk density of .25g/cc while
the 2% had a denisty of .44g/cc. The reason the pores of the 12%
absorbed water were smaller than the 2% is that glass at higher water
content is less viscous at a given temperature which would make smaller
pore stabilization possible. A possible explanation for the 2% glass
increasing in density at higher temperatures while the 12% does not is
that a greater proportion of the 2% water is removed from the glass
before the sintering mechanism can trap very large amounts of the
water at higher temperatures. A large amount of the 12% water in the
glass could escape but a large amount of water would remain to foam the
glass. With lowered viscosity and more foaming agent, a less dense
particle is produced from the 12% water reacted glass than the 2% at
higher temperatures.
A large pellet from the autoclaved sample with 12% absorbed water
was prepared by squirting a 5N NaOH solution, whose final foamed
dimension was 5 cm when fired at 900°C. A density of .14g/cc with
uniform pores was obtained. However, a high water absorption into the
foamed object was 64%.
From this data it is evident that more than enough water can be
put into the glass to foam it and that foaming of very small particles
is possible when using water as the foaming agent. For large objects
it appears that pelletizing in a rotary mixer before autoclaving will
yield the best results for pore uniformity and density, although high
39
-------
in open porosity as compared to sedimented compacts which require
closer temperature controls.
Pellets were also extruded but the foam produced after firing was
coarser and less uniform. This confirms the tendency obtained throughout,
that the best results come from the most loosely packed mixture.
These loosely packed structures probably allow more of the gases to
escape before sintering begins and the foaming is more controllable
due to the smaller amounts of the vapor.
40
-------
V. CARBON AS THE FOAMING AGENT
The carbon foaming process is largely deoendent on the type of
carbon that is used. Carbon blacks available on the market today
are very diverse in their individual characteristics. The carbon
blacks examined as the foaming agent were the furnace blacks, channel
blacks and activated carbons. These carbons differ by the manner in
which they were produced. The activated carbons are carbons that
were deposited at low temperature and are free from absorbed and
stabilized hydrocarbons on their surfaces. Furnace blacks have
oxygen contents below 1.5% while channel blacks have between 3-4% oxygen.
Carbon black particles have extensive surface areas. Surface
2 2
areas ranging from 25 m /g to 1000 m /g were studied. This large
surface area is due to the interconnecting pores which oermeate the
particle. However, it has been observed that the most oorous carbon
doesn't necessarily absorb the largest amount of gas. It has been
proposed from this that the pores should be about the same size as
the absorbed gas molecules. There is some evidence that the gases
absorbed are condensed into a liquid. When gases are absorbed onto
the carbon heat is evolved. The heat of absorption of CO^ is 6800-
7800 g cal/mole while the heat of liquifaction is 6250 g cal/mole. ^ '
'(12)
In addition to this Melscherlick, according to Mantel ' observed
that COo, at 12°C and atmospheric pressure, was absorbed into the
carbon and occupied 1/56 its initial volume. This being the case,
there could be large volumes of gas on the surface of the carbon.
41
-------
In addition to the absorbed gases there are other elements that
are chemically bound to the carbon. It has been shown that the
carbon black contains .2-1% hydrogen, .1-4% oxygen and up to 1% sul-
(11}
phur. ' The reason hydrogen is present is because the carbon
black is generally made from hydrocarbons. Oxygen appears from the
oxidation of the hydrocarbons.
Table 7 shows that the various carbons have difficult pH values.
This is due primarily to the carbon dioxide which is chemically or
quasi-chemically bonded to the surface due to oxidation. With increas-
ing C0?-complex concentration more surface ionization results,
giving a larger H ion concentration. Other functional groups such
as quinone and phenolic groups have been identified. With these
acid radicals present, the carbons become more hydrophilic.
As the carbons are heated to the 850° range several changes
can take place. One of these is the evolution of the absorbed gas.
With increasing temperature, carbon dioxide comes off first, followed
by carbon monoxide, hydrogen, and if present, sulphur dioxide.
Sulphur dioxide can react in the following manner
2S02 + 3C = C02 + 2CO + S
or under a carbon monoxide atmosphere
S02 + 2CO = 2C02 + S
The second reaction is important since 1 less mole of gaseous products
are formed than is reacted. This would mean that when the S0? and
COo were desorbed they would react to give 1 less mole of C0?, thus
decreasing the volume of gas effectively evolved.
42
-------
Table 7.
CARBON BLACKS EXAMINED FOR FOAMING GLASS
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19-
F -
C -
LB-
Act.
Carbon
Black
Continex F-l
S-315
N-330
P-108
Neo-Spectra 2
Neo-Spectra 4
Continex F-3
Raveen 2000
Raveen 1040
Raveen 14
Raveen 3500
HAF #4
Raveen 1255
U-3024-L
XZ (Nutshell)
NW (Hardwood)
YD (Nutshell)
410 (Nutshell)
JF (Nutshell)
Furnace Black
Channel Black
Lamp Black
- Activated Carbon
Particle
Carbon Size
Type urn
F
F
F
Act.
C
C
F
F
F
LB
F
F
F
F
Act.
Act.
Act.
Act.
Act.
80
.0280*
.0798*
.05-. 15*
13
15
47
17
29
--
16
—
22
50
98%thru*
95%thru*
90%thru**
70%thru**
.20
-------
The carbon with its absorbed gases can be mixed with glass
particles. At 850°C the particles sinter together and form a viscous
liquid surrounding the carbon. The gases evolve from the carbon
particle and cause a gaseous pore to develope. As the pores expand
they are hindered by contact with another expanding pore; so that the
larger the distance between nucliation sites, (carbon particles) the
larger will be the pores. The glass-carbon mixture ceases to expand
when the gas inside the pores come into equilibrium with the surface
tension of the glass.
A. Experimental Procedure
The glass used in these experiments was the ordinary soda-lime
glass that was obtained in broken angular pieces from window panes and
bottles from a local hospital. These were then crushed in roller
mills. The glass was then placed in onegallon milling drums half
filled with large metal balls and allowed to mill for approximately
thirty-six hours. This procedure yielded glass particles that would
pass through a 325-mesh sieve.
The glass powder was then mixed with standard one-percent bentonite
and varying percentages of each carbon in a vibratory mill. The
clay was incorporated in this study since it is anticipated that the
industrial application would utilize extrusion methods for forming
the green piece; it is desired that this study be applicable to the
extrusion method.
The glass mixture was pressed initially at 2493 psi, 3739 psi,
and 4986 psi. Finally, 2493 psi was chosen to be the standard pressure
44
-------
because it led to less capping of the compact. The dimensions of
the pressed compacts were 3.18 cm (1.25 inches) in diameter and
about .94 cm (3/8 inches) high.
A standardizing run was made on all carbons at 850°C with a 1%
carbon concentration in a Thermolyne furance. Two samples were then
selected that had thin crusts and uniform pores, and two others
which had large pores. These carbons were U-3024-L, F-l, R-1040, and
S-315 respectively. Varying compositions were also tried and in all
cases the time in the furnace was thirty minutes.
The method employed for firing consisted of pre-heating the
furnace to the indicated temperature and placing the pressed compacts
into the furnace. After thirty minutes the samples were removed with
tongs and allowed to cool under insulation outside the furnace.
The process of foaming was also noted as a function of time,
temperature and content. Ten samples were all placed in the preheated
furnace (Lindberg) at the same time and removed one at a time at the
specified intervals. Observations were recorded for various carbons
and density and water absorption measurements were made. The pressed
compacts during firing expanded to about 6.36 cm (2 1/2 inches) in
diameter. However, because of the crust and a few localized giant
pores, caused probably by pressing, it was thought that the density
measurements would be best calculated by measuring a selected sample
cut out from the middle of the specimen and then weighing this sample.
Dividing the weight by the volume of the sample yielded the density.
Unfortunately, this method gave results of + .05 g/cc. Water absorbancy
was determined by dry weighing the samples and then placing them in
45
-------
an apparatus to draw a vacuum. After a vacuum was achieved, water
with a little soap to reduce surface tension was allowed to enter
the evacuated flask and penetrate all the open pores. The sample
was weighed wet to determine the water absorption. This became a
measure of the interconnection of the pore space.
Finally, to check the reducing atmosphere effect the compacts
were placed in stainless steel beakers and covered by carbon cloth.
Care was taken not to restrict the expanding motion of the glass by
merely placing the carbon cloth on top so that only the atmosphere
effect would enter in.
Following this set of experiments larger samples were foamed
and found to be of good quality up to about 2 inches thick.
Pellets were fabricated by extruding a 2% Bentonite, 2% carbon
and glass mixture through a .24 cm (3/32 inch) die. These extruded
rods were allowed to dry and then chopped up into .32 cm (1/8 inch)
lengths. These pellets were mixed with dried ball clay to prevent
sticking of the pellets to each other at a 1:2 weight ratio and
inserted into a 2 inch diameter stainless steel tube 91 cm (36 inches)
long rotating at about 6 RPM through a Kanthal tube furnace on a
slope of 2.54 cm (1 inch) per 18 cm (7 linear inches). It took
about 15 minutes for the mixture to go through the length of the
tube at 850-900°C. The clay was then screened off and used again.
B. Results
After a screening examination of the available carbons, the
results of this study were limited to the acid R-1040 and S-315 and
the basic F-l and U-3024-L carbons. The results will be generalized
where possible, unless otherwise noted.
46
-------
It is interesting to note the sequential steps that occur during
the foaming operation. Immediately after placing the pressed compact
into the preheated furnace the carbon on the sharp corners of the
disc immediately oxidizes as evidenced by the whitening of this
area. The sintered shell begins to form and reaches its maximum
thickness at about eight minutes. The pores then begin forming from
the outside in, as evidenced by the larger pores near the shell.
As the process continues the gas bubbles make their way to the crust
and coalesce, making the pores in the crust larger. Much of the same
process is taking place inside. With increasing time the pores
coalesce because of the impinging foam surfaces and continued gas
generation by the carbon. As this process continues the open porosity
increases as shown by the water absorption. During continued heating
the carbon disappears, the remaining gases diffuse out, and the
foamed compact begins to shrink.
In examining the various influences that exist in the actual
foaming; several classifications emerge (1) the factors influencing
the formation of small pores (2) the factors governing large pore
sizes and (3) the crust formation. These influences are summarized
on Table 8.
For small pore formation, an increase of carbon content decreased
the pore size, as seen by Figures 13 and 14. It was also noted that
many of the samples had significant quantities of absorbed sulphur,
since after firing, when a spatula was used to break the foamed
piece, an H^S odor was noted. Since sulfur was available in the
47
-------
Table 8.
SUMMARY OF FOAMING FACTS
To achieve:
Smaller Pores
1. increase carbon content
2. add desulphurizers
3. use reducing atmosphere
Larger Pores
1. use activated carbons
2. use coal
3. lower carbon content
4. increase temperature
Decrease Crust Thickness
1. cover compact while firing
2. increase carbon content
Increase Crust Thickness
1. add iron oxide
48
-------
1.2
•-a i% F-I
0.8
UJ
N
CO
UJ
o:
0.4
O 0.5% F-I
0.5% R-1040
0
30 60 90
TIME (min.)
120
Figure 13. Pore size as a function of time at
850°C.
1.3
0.5% U-3024-L
I 0.9
N
CO
UJ
O 0.5
Q.
1% U-3024-L
0.5% S-315
O.I
0
30 60 90
TIME (min.)
120
Figure 14. Pore size as a function of time at
850°C.
49
-------
carbon blacks and In the glass, the exact source is difficult to
trace. However, when 1% powdered manganese was added the odor was
not present and smaller pores were evidenced. Another factor causing
smaller .pores was the use of a reducing atmosphere; and in certain
instances sulphur was deposited on the container surface. This had
the large effect of decreasing the pore size by about a factor of two
for the acid and basic carbons considered.
The large pores were generated, in general, by the use of any
of the activated carbons. These caused pores to be foamed on the
order of 6mm in diameter; as opposed to an average of .6-lmm. Coal
also generated large pores. For the carbon blacks tried, it was
found that decreasing the carbon content and/or increasing the tempera-
ture led to larger pore formation.
The crust or shell of the foamed glass pieces is minimized by
covering the sample during firing, increasing the carbon content, or
covering the compact with a carbon cloth. It was found that the
crust was thickened by addition of powdered iron oxide, and that the
R-1040 carbon developed a coherent crust faster than the basic F-l
type. When considering the two basic and the two acidic carbons, the
pH seemed to be the largest factor in producing large pore size -
the bases being more active than the acids. For a given pH the
largest surface area then took effect in producing the larger pore
size. When 0.5% acid carbon was used, it was found that it foamed
in a general manner comparable to the basic carbons. However, at 1%
the acid carbon caused large localized pores with a fine matrix
50
-------
while the 1% bases foamed much like the 0.5% carbon but with smaller
pores at the same temperature.
From Figures 15, 16 and 17 it can be seen that a higher carbon
content achieved a lower density than the lower content for a given
carbon. The minimum densities range from 0.12-.15 gm/cc. The acidic
carbons also have a sharper minimum range and approach minimum density
faster than the bases.
As can be seen by the water absorption data from Figure 18, a
linear rate of increase for the 0.5% carbons is maintained for the
first twenty-five minutes; and except for F-l, the first forty minutes.
The rate of water absorption for both acid and basic carbons is 0.1
gm HpO/ gm sample/min. For the acidic types the water absorbancy
reaches its peak at 40-45 minutes and begins to decrease because of
the collapsing pore walls.
Figures 13 and 14 show an approximate linear increase in the
rate of pore growth for the different carbons in the early stages of
foaming. For the 0.5% and 1% variations in carbon, it was seen that
the 1% carbon content had a slightly slower rate of pore growth than
the 0.5% mixture. The most pronounced effect was in the U-3024-L
carbon where the 0.5% rate was about .022 mm/min. and for the 1% .it
was .013 mm/min. It is possible that the carbon addition has increased
the viscosity of the glass.
When the acidic carbons were placed in an aqueous medium the
carbon particles were wetted and went into suspension while the
basic types remained on the surface. The U-3024-L carbon wouldn't wet
51
-------
0.4
0.3
e
s
>
H
CO
z
LU
Q
0.2
O.I
0.5% S-315
0.5% R-1040
0
40 80
TIME (min.)
120
160
Figure 15. Density as a function of time at 850°C
for S-315 and R-1040 carbons.
0.4
o
o
e
o>
CO
z
u
o
0.3
0.2
O.I
1% F-l
0.5% F-
0
40 80
TIME (min.)
120
160
Figure 16. Density as a function of time and carbon
concentration at 850°C.
52
-------
0.4
u
^
E
o>
0.3
CO
I °2
O.I
1% U-3024-L
0.5% U-3024-L
40 80
TIME (min.)
120
160
Figure 17. Density as a function of time and
carbon concentration at 850°C.
53
-------
O.
e
o
u»
I-
o.
o:
o
CO
CD
0
10
30
50
TIME (min.)
70
Figure 18a. Water absorption of 0.5% R-1040 carbon
black foamed glass as a function of time
at 850°C.
—
Q.
e
o
in
O
CJ
X
Z
O
H-
Q.
o:
o
CO
CD
0
10
30 50
TIME (min.)
70
90
Figure 18b. Water absorption of 0.5% S-315 carbon
black foamed glass as a function of time
at 850°C.
54
-------
9)
OL
E
O
CO
O
CM
X
o>
Q.
DC
O
CO
CQ
0
10
0.5% F-l
1% F-l
30 50
TIME (min.)
70
90
Figure 18c. Water absorption of F-l carbon black
foamed glass as a function of time
at 850°C.
0)
a.
E
o
(A
O
04
I
•S
•z.
o
t 2
o:
o
CO
GO
0
10
0.5% U-3024-L
% U-3024-L
30 50 70
TIME (min.)
90
Figure 18d. Water absorption of U-3024-L carbon
black foamed glass as a function of
time at 850°C.
55
-------
at all, while about one-half of the Fl carbon was suspended. In
comparing the non-wettability of water on the surface of the carbon
it was noted that: U-3024-L>F-1>R-1040 and S-315. Similarly, the
pore size of the samples follows: U-3024-L>F-1>R-1040 and S-315.
The pellets with low (.5 - 1%) carbon content didn't yield good
results at 850-900°C because much of the carbon oxidized before foam-
ing could take place. However, at 2% carbon content the foaming was
effective yielding a bulk pellet density of .23 g/cc or of 8.5 Ib/ft .
The pellets had relatively non-porous crusts whose water absorption
was 10%.
The larger sized foam boards produced had less uniform cell
structure than the smaller bulk objects. This is due to the uneven
heating rate of the large compacts caused by the insulating qualities
of the outer foamed glass. This problem was less acute for decreas-
ing thicknesses of the final foamed product. Products up to 3.81 cm
(1 1/2 inches) were easily made with relatively good uniformity.
C. Discussion
From the results of this study it is clearly seen that there
are several processes that contribute to the foaming phenomenon.
Immediately upon firing, the carbon begins oxidizing and continues
through this process until, at the conclusion of the firing, the
density rises and porosity decreases along with a change of color of
the inside from black to cream. A reducing atmosphere retards or
eliminates completely the crust on the surface. It is apparent,
however, that oxidation is not the major source of gas that foams the
glass.
56
-------
Due to the fact that the furnace is preheated, a pellet placed
in the furnace will become hotter on the outside faster than the
inside; and so foaming will begin on the outside, since it will seal
on the outside first and the gases will be trapped. In going to
higher and higher temperatures the pore size increases because of two
effects. One is the fact that the pores will be sealed by the glass
faster since it is less viscous at higher temperatures and the other
is that once the envelope has formed, less resistance to the foaming
pressures will be exerted because of the less viscous nature of
the glass.
A possible reason for the correlation of the non-wetting carbons
tested, with the magnitude of pore size and the pH of the material,
has to do with the degree of oxidation of the carbon. Since the
non-wetting carbons are basic, this indicates that the oxidation
hasn't progressed very far and has left many hydrocarbons still
present on the surface. If many oxygen groups were present, these
carbons would act in a hydrophilic manner. This would leave more
hydrocarbons to desorb and give larger pores in the mixture. One of
the bases (F-l) is slightly more wettable and its pH is .5 less than
the U-3024-L carbon yet the pores in the foam are slightly smaller.
This means that it has been more completely oxidized and less hydro-
carbons are present for foaming. This is again reinforced because
of the extremely large pore size of the coal-glass foam. The coal was
composed of hydrocarbon chains with no activation applied. Unfortun-
ately it is hard to obtain coal in the size range needed to obtain
small sized foam.
57
-------
The activated carbons followed a different mechanism since all
of the hydrocarbons are carefully removed, leaving an extensive area
for the gases to be absorbed. Since there aren't many hydrocarbons
present only the absorbed gases foam the glass.
The carbon blacks studied indicate that for a given pH the
largest surface area gives the largest pores simply because more
gases are allowed to be absorbed on a larger area.
A secondary effect on foaming would be the number of sites
available for C0? and SO^ bonding, and the gasification of the carbon
itself. These would be the only way that the acidic carbons would be
able to foam. This is verified because in a CO atmosphere the pores
are reduced in size. According to the reaction SCL + 2CO = 200^
+ 1/2S, three moles of gas are reduced to two moles of gas. This
would reduce the pore pressure and decrease pore size. The carbon
monoxide would also inhibit gasification of the carbon according to
the mechanism listed in the introduction. To further verify that the
SOg is present, the manganese removed the H?S odor as well as decreased
the pore size. The manganese was able to reduce the pore size by
possibly taking a gaseous component and turning it into a solid phase
MnS; and thereby removing the partial pressure of the S0? gas.
The acids have a narrower minimum range in the density than the
basic carbons because less gases are produced which can diffuse out
and densification can begin much sooner.
The absorption or open porosity increases linearly since less
particles are foaming at the same rate, and as time passes, larger
pores form due to the coalescence which, when connected to a channel,
increases the length of the open pore.
58
-------
The more foaming nuclei (carbon particles) there are, the
smaller the pores. This is because all the carbon particles give
off gases at about the same rate and are thus growing at the same
rate. However, when a pore begins impinging on another, its growth
stops. The smaller the concentration of foaming nuclei, the larger
the pores will become before impingment occurs. For the same reason
the higher concentrations of carbon in the carbon-glass mixture
reach minimum density faster because all the available area is
foamed faster because of the more numerous nuclei.
Finally, because the rate of pore size increase is smaller for
the higher percent carbon; it is thought that the glass through
which the pore is expanding is slightly more viscous due to the
higher carbon content.
59
-------
VI. CALCIUM CARBONATE AS THE.FOAMING AGENT
CaCO,, can also be used as a foaming agent. In this case the
O
glass is ground to an average particle size of one to twenty microns.
The glass does not require color sorting, sizing, or cleaning. Calcium
carbonate (CaC03) and bentonite are next mixed with the ground glass.
The bentonite is added to improve the plasticity of the mixture so
that it can be easily extruded or pressed into blocks. As the dried
mixture is placed into a furnace at approximately 800°C, the CaC03
reacts with the glass (largely Si02) as follows:
CaC03 + Si02 •*• CaSi03 + C02 (g)
The CaO being thus incorporated acts as a glass modifier reducing the
viscosity of the glass. At this same temperature, the glass particles
begin to sinter. The sintering prevents the carbon dioxide gas from
escaping by sealing off the oassage ways. The pressure of the gas can
then expand the molten glass into a low-density cellular structure.
This process is used to produce both large blocks or slabs and pellets
of foamed glass.
Foam glass in the form of pellets possibly can provide wider
applications for foam glass as a structural insulating material.
Combining with proper binders, the pellets can be manufactured in any
size or shape without waste. Another advantage is that glass can be
foamed more uniformly due to faster heat transfer throughout the
pellets. One drawback is the formation of a thick skin on the surface
of the CaC03 foamed pellets which increases the overall density and
reduces the insulating characteristics of the pellets.
60
-------
Foam glass in the form of pellets possibly can provide wider
applications for foam glass as a structural insulating material.
Combining with proper binders, the pellets can be manufactured in any
size or shape without waste. Another advantage is that glass can be
foamed more uniformly due to faster heat transfer throughout the
pellets. One drawback is the formation of a thick skin on the surface
of the CaCO-j foamed pellets which increases the overall density and
reduces the insulating characteristics of the pellets.
Exposure to moisture over a significant length of time may cause
the glass insulation to deteriorate, resulting in a loss of insulating
quality and structural strength. In order to lower the solubility,
bentonite clay is mixed with the glass prior to foaming. The rela-
tively high content of Al,,03 in the clay decreases the solubility of
the glass. The function of the clay is then two fold - to lower the
solubility and to act as a plasticizing agent for extrusion.
A. Experimental Procedure
Clear soda-lime cullet is used throughout the experiments Grind-
ing was done in a two foot diameter ball mill with steel balls. Each
batch containing thirty to forty pounds of glass was ground for two
hours. The final size ranged from one to twenty microns. The CaC03
and bentonite was mixed with the glass in smaller rubber lined ball
mills 20 cm (8 inch diameter) using alumina balls.
The standard size disk-shaped samples for foaming tests each
weighed 65 grams and are 6.5 cm in diameter. The samples were placed
directly into a preheated electric furnace for a set length of time,
61
-------
and then they were withdrawn directly to room temperature. In addition
to these standard samples, larger samples 23 cm (9 inches) by 13 cm (5
inches) by 3.8 cm (1 1/2 inches) were prepared to assure that the
relationships between the various parameters are consistent for dif-
ferent sizes.
When working to form foam glass pellets the waste plate glass
was crushed and ground in a ceramic ball mill for one day and the
particle size distribution was determined by Coulter Counter as:
< 3y ; 25%
3 ^ lOp ; 50%
15 > 10y ; 25%
The surface area of this glass powder also obtained by BET test is
2
3.34m /g. The powder was mixed with 2% CaC03 and 4% high-swelling
bentonite and extruded into cylinder shape with 3 different diameters.
The extruded samples were dried and cut into small pellets of differing
lengths. The pellets were foamed at various temperatures and time
and foaming characteristics were observed. The procedure for the
solubility test was as follows:
1. Crush the foamed glass and sieve to a -40 to +50 mesh size.
2. Wash with alcohol to remove fines from the surface of the particles.
3. Dry in drying oven at 120°C for 12 hours.
4. Weigh a sample approximately 7-1/2 grams to 10 grams.
5. Add 250 ml distilled water to sample (in 250 ml plastic beaker with
lid).
62
-------
6. Place beaker in 90°C water bath for six hours.
7. Filter and dry in oven at 120°C for 12 hours.
8. Weigh residue and determine loss.
In order to see the effect of glass particle size on the density of
the foamed glass, four samples composed of different particle sizes
were prepared. ^ ' Each sample was mixed with 1% CaCQ, and pressed
into the form of a disc at 500 psi. After foaming for 5 minutes at
800°C, the change in pore size and bulk density of the products was
observed.
Thermal conductivity measurements were done using the rapid K-factor
method and standardized to Pittsburg Corning foam with .4 BTU/hr. °F
2
ft. /in. The K-factor of CaC03 foam vs density and cell size was deter-
mined.
B. Results and Discussion
During earlier investigation, glass containing 1.0 to 2.0 percent
CaCOo was foamed at temperatures ranging from 700°C to 800°C, often
resulting in an open-cell foam structure. A closed-cell structure,
which improves the insulative qualities of the foam, was obtained by
foaming at temperatures near 800°C using approximately 0.5 to 1.0
percent CaCO,. However as noted on Table 9, as time in foaming increases
the open porosity also increases. The cell size as a function of the
percent CaCO, and the foaming time are shown in Figures 19 and 20
respectively.
As shown in Figure 19 the type of CaC03 affects the foaming
characteristics. Milled limestone and reagent grade CaC03 have
similar size distribution and structure. However, the foaming quali-
ties are significantly different between the two. A high-purity
63
-------
Table 9.
OPENING OF PORES WITH INCREASING FOAMING TIME
Sample Foaming Time Hater Absorbed (wt.%)
Foamed at 850°C 3 minutes 25.0%
!%CaCO
5 minutes 84.5
8 minutes 138.5
64
-------
3.0 r-
2.0
Q
i
01 LJ
N
CO
1.0
A MILLED LIMESTONEC CaCOJ
D REAGENT GRADE CaCO,
0
LO
2.0
3.0
% CALCIUM CARBONATE
Figure 19. Cell size versus percent calcium carbonate.
-------
3.0
2.0
cr>
E
E
Q
UJ
N
CO
_J
_J
LJ
O
o
O
1.0
0
30
60
90
120
TIME (min)
Figure 20. Cell size versus foaming time.
-------
precipitated CaCCL was tested and found to be inferior to the milled
limestone. Since the reactivity of the CaCO- is inversely related to
the surface area, the precipitated CaC03 probably reacts before the
surrounding glass particles can sinter to seal off the passageways.
Figure 21 compares the firing temperature time and percent CaCO- added.
Uniform cell size was only possible when the entire furnace was
within 25°C of the operating temperature and fired according to the
other variables shown. The milled limestone and reagent grade CaCCL
produce higher quality foam as the particle size of the CaCO^ decreases
as shown in Figure 22. The best foaming occurred when the CaCO,
particles were from 10 to 50 y.
Because of the ease in producing large amounts of milled glass,
large 7.6 cm (3 inches) by 30 cm (12 inches) by 60 cm (24 inches)
pieces of CaCOo foam were produced. When working with large pieces
care had to be taken in heating during foaming. Figure 23 gives
examples of the heating curves used. Only the one curve would produce
uniform small (^ 2 mm) pores. Total heating time was tyoically four
hours for the largest pieces.
Table 10 shows the properties of the various pellets which
were foamed at the same conditions. The density of bulk foam block
was also listed for comparison. Much higher densities of foamed
glasses were obtained for foamed pellets. It should also be noted
that the densities of foamed pellets increase with decreasing sizes
of the original extruded pellets. These results can be explained by
the so called "skin effect". When a pellet is heated, the surface
67
-------
IOO
90
8O
LU
oo 2
I-
CD
cr
70
6O
5O
40
650
REGION OF UNIFORM
CELLS SIZED l.0-2.0mm
1.0% CaC03
CaC03
2.O% CaCO
700
750 800
FIRING TEMPERATURE (°C)
850
Figure 21. Firing time versus % CaCO- vs. temperature. Samples expanded
6 times its volume.
-------
3OO
£ 200
CO
<
UJ
or
o
2
100
•"<•*
m o
O
O.OI
325 mesh
,
200 mesh
lOOmesh
40 mesh
0.02
0.05
O.I
SIZE (mm)
0.2
0.5
1.0
Figure 22. Volume increase vs. CaCO, size for 10-20y cullet at 700°C for one
hour and 0*5%
-------
750'
T
NON-UNIFORM
FOAM
750*
500£
NON-UNIFORM
FOAM
750C
5OO<
NON-UNIFORM
FOAM
800C
UNIFORM FOAM CELLS
Figure 23. Large scale sample foaming schedules (CaCOg type).
-------
Table 10.
PROPERTIES OF FOAMED GLASS PELLETS
(FOAMED AT 850°C FOR 5 MINUTES WITH 2% CAC03)
Sample
Pellet #1
#2
#3
Block foam
Glass
Size of Sample
before foaming
Diameter
1.27cm
1/2"
.48cm
3/16"
.19cm
3/40"
Height
1.27cm
1/2"
.48cm
0.370
.19cm
3/40"
Bulk Density 3
after foaming (g/cm )
0.312
0.410
0.230
71
-------
looses most of the C02 before the glass seals off its passage.
Eventually this portion sinters without any foaming and results in a
high density product.
The solubility of foamed glass in water decreases as the percent
of clay increases. The solubility, however, is not affected by the
CaC03 content (Figure 24). The typical chemical analysis of the clays
used are shown in Table 11.
The higher the alumina content of the glass due to the addition
of the clays, the lower the solubility of the glass. Figure 25 shows
the solubility as a function of clay additives. The minimum amount
of bentonite needed to extrude the glass mixture is two percent. At
this percentage, the solubility has decreased by 24%
Clay content greater than approximately 6% lowers the quality of
the foam by increasing density and decreasing cell uniformity. The
sample with low-swelling bentonite, foamed at a lower temperature
(780°C) does not show any effect of clay addition while the same sample,
foamed at a higher temperature, does.
This result may indicate that there is a minimum temperature for
each clay to be dissolved in the glass structure and affect the
solubility of glass. As can be seen from Table 12 the best foam for
uniformity was for glass with a particle size below 10 y. The density
attained here was ^.4 g/cm . As seen from Figures 26 and 27 the K
factor is a function of density and cell size whose minimum value is
about .4 BTU/hr °F ft2/in.
72
-------
.30
.20
oo
CO
CO
O
_J
O
.10
O
O
O
O
O
0
1.0
2.0
3.0
% CALCIUM CARBONATE
Figure 24. Percent weight loss (solubility) versus percent CaCO,
-------
Table 11.
CLAY ADDITIVES
Clay
Source
Ball Clay
(Old mine #4)
Low-Swel1 ing
bentonite
Plantsite
(Harrisvi1le
Brick Plant)
High Swelling
bentonite
Kentucky-Tennessee
Clay Company
American Colloid
Company
local
Main Composition (wt.%)
Si02 A1203 Na20+K20 CaO
52.0 31.2
57.0 20.0
55.0 12.0
64.0 21.0
1.3 0.4
1.0 2.0
3.0 10.0
3.0 0.5
Sample No.
1
2
3
4
Table 12.
EFFECT OF PARTICLE SIZE OF GLASS ON FOAMING
Properties of Foam
Particle Size (y) Pore Size (mm) Bulk Density (g/cm3) Uniformity
10^43
43^61
6K74
^0.40
^0.45
^0.55
MJ.80
Best
Good
Bad
Bad
74
-------
en
1.6
1.4
0.4
0.2
0
LOW-SWELLING BENTONITE
(FOAMED AT 780°C)
FOAMED
AT 800°CT
FOAM GLASS BY CORNING
_L
• HIGH-SWELLING
BENTONITE
LOW-SWELLING BENTONITE
I
I
6 8 10
ADDITIVE CONTENT (%)
15
17
Figure 25. Solubility change of foam glass in water with clay additives.
-------
0.32
0.54
0.50
0.46
m o.42
cr
O
o
0.38
0.34
10
12
16
18
DENSITY p(lb/ft3)
Figure 26. K-factor of CaCO^ foam vs. density.
06
01
05
0.2 0.4 0.6 0.6 1.0 1.2 1.4
CELL SIZE-DIA (mm)
20
1.6
1.8
Figure 27. K-factor CaC03 foam vs. cell size.
2.0
76
-------
The properties of the CaC03 glass compared well with commercially
produced foam glass. Density varied around 10 pounds per cubic
foot. Compressive strength averaged over 8 kg/cm2 (125 pounds per
square inch) and is sensitive to annealing. The high insulating
quality, as indicated by the K-factor, varied with density and cell
size. The foamed glass will not burn but does have a tendency to
abrade under a flame.
77
-------
VII. SUMMARY OF FOAMING METHODS
Each foaming agent has some characteristics that distinguishes
it from another that would make it useful in a particular application.
Water impregnated glass can foam particles as small as 177 microns.
This will be an asset as a filler material as well as an insulation
since good foaming can easily be done in pellets up to .62 cm in
diameter. In proceeding to larger dimensions considerable control of
temperature and water content is necessary to obtain uniform pores.
Carbon black foaming is limited from a small pellet size of .32
cm to large slabs. The foaming produces uniform pores and is reproducible
if the furnace temperature is uniform. The crust or shell is not
very thick and can essentially be eliminated if foamed in a reducing
atmosphere.
Calcium carbonate produces pores that are more than twice the
size of the carbon foamed pieces and has a thicker crust associated
with it. The bulk density was also a little higher than the carbon
foamed glass.
From these observations it appears that water is best suited for
foaming glass from a 177 micron size to pellets .32 cm in diameter.
Carbon black gives the best results for objects between .64 cm and 5 cm.
78
-------
VIII. MARKET POTENTIAL OF FOAMED WASTE GLASS
Insulation and Concrete Aggregate Markets
The development of foamed waste glass stems from a current trend
in the United States to conserve natural resources and to decrease
pollution. Solid waste refuse is one form of that pollution. Recycling
and re-use of the refuse conserves that portion of the original raw
material which is replaced in a new product by the recycled material.
If components of refuse can some day be economically separated, recycled,
and marketed by industry, the pollution from solid waste will be reduced
significantly. Research concerning the recycling of solid waste has pro-
vided economical ways to recover various components of the solid waste
refuse. The implementation of solid waste separation and use of the
output on a widespread commercial basis remains to be done.
Solid waste is composed of metallic, organic and ceramic material.
The metallic and organic fractions are either already valuable and recov-
erable or can be incinerated to recover the fraction as thermal energy.
The ceramics fraction, however, is by comparison of low value. This is
because the raw materials are generally non-critical and widely available.
i
Glass is part of the ceramic fraction and constitutes about 6h percent of
all solid waste. The glass fraction of solid waste can be separated.
There are more potential uses for salvaged waste glass than there
is glass available now or in the foreseeable future. If the salvaged
waste glass from Los Angeles, St. Louis, New York, Washington, D.C.,
Detroit, San Francisco, Denver and Chicago was crushed and used for
79
-------
asphalt aggregate in those cities, only 50 percent of the aggregate
needed could be provided. Color separation is required for some
uses such as recycling into bottles. Another way to recycle waste
glass is to foam it and use the foamed waste glass rather than the
broken glass. More research has been done on the use of crushed waste
glass than on foamed waste glass—probably because crushed glass was
more readily available. Foamed waste glass requires another process
(foaming) to produce a usable raw material. However, in many instances
such as concrete aggregate, the properties of foamed glass seem to pro-
vide a broader base for use than the crushed glass.
Uses for foamed glass have been suggested in previous literature,
but little has been published to evaluate the market potential of
foamed waste glass.
The objective of this study is to determine the market demand for
foamed waste glass. Market demand is analyzed with respect to the two
principal uses of foamed waste glass: (1) rigid insulation and (2)
lightweight concrete aggregate. In each case the market demand is
determined on a national and regional basis. The discussion of market
demand is followed by an analysis of the production cost of foamed
waste glass. Production cost information is relevant because these
costs determine the market price of the product and are therefore very
important indicators of market potential. The market price of foamed
waste glass is particularly significant because of the price differential
between foamed waste glass and existing rigid insulation products.
80
-------
Other desirable features of foamed glass insulation include its
strength and rigidity. Flexible, blanket, and batt insulations cannot
support any significant weight. Because of the strength and rigidity
of foamed glass, it will not slump or compress. This eliminates voids
(26)
and sagging. The high compressive strength and low density also make
it an ideal construction material. Foamed glass will support an average,
ultimate load of seven tons per square foot with a compressive strength
greater than 100 psi. Therefore, less material needs to be used to
create 'supporting structures. Other than the minimal steel support
structure, four-inch thick foamed glass has been used as the exterior
wall of a building. Tapered blocks may be used to create a roof deck,
and thin slabs of foamed glass have been used as insulation in parking
(25)
decks and over concrete sub-floors. The rigidity of the cellular foam
in combination with the vapor barrier created by cellular glass is
excellent for many tasks in which no other insulation can meet the
specifications.
Because of the incombustible material from which foamed glass is
made, no fireproofing is necessary. This is especially important in
public buildings where fire hazards are of prime concern.
Other types of rigid insulation are not competitive with foamed
glass in uses where the object to be insulated requires a vapor barrier.
A. The Market Potential for Rigid Insulation
Rigid insulation is an important construction material used in non-
residential construction. A small segment of this rigid insulation
market has been captured by FOAMGLAS, a Pittsburgh Corning Product,
81
-------
which is relatively expensive and is used only in high quality non-
residential construction. Foamed waste glass, as produced by the (Table 13)
University of Utah, possesses many of the highly desirable properties
of FOAMGLAS, i.e., waterproof, incombustible, rigid, and dimensionally
stable. Although there are many physical similarities between FOAMGLAS
and foamed waste glass there is a significant price difference between
the two products. Pittsburgh Corning FOAMGLAS is priced at 24-26
-------
Table 13
COMPARATIVE VALUES OF SELECTED PHYSICAL PROPERTIES
OF INSULATION: RIGID
Property
University of Utah
Foamed Glass
Pittsburgh-Corning (P.C.)
Foamed Glass
Other
Size
(dimension In inches)
Assumed similar to
Pittsburgh-Corning (P.C.)
Assumed similar to P.C.
24 x 18 x
Other sizes are
available or cut or molded
lesser sizes or shapes desired4
oo
OJ
Density
(lb./ft3)
10 - 15'
(19)
Perlite 3 - 4 <16>
Compressive Strength
(psi) > 1000-9 '
100
(24)
Thermal
Conductivity
.4
(19)
.4<25>
Perlite .3 - .5
Combustability
Will not burn
(25)
Will not burn
(18)
Perlite - Will not
'Note: Numbered footnotes in parentheses refer to references at end of text material.
aSuch as tapered thickness for roofing and curved for pipes.
-------
foot and none of them have the wide range of qualities possessed by
foamed waste glass. The high quality and low price of foamed waste glass
would increase the size of the market from $25 million to $100 million
(value of entire non-residential roof insulation market). Nevertheless,
limited application for the residential construction market would continue.
Given the price differential and inherent quality advantages of foamed
waste glass, it is assumed that foamed waste glass would capture 20 percent
of the roof insulation market. Tables 15 and 16 supply important informa-
tion about the size and value of the rigid insulation market and the
number of 20/ton/day plants required to meet projected production levels.
The data shows that the foamed waste glass market for rigid insula-
tion in the Western United States has been sufficient to support two
20/ton/day plants and the national market has been sufficient to support
fourteen 20/ton/day plants.
B. Characteristics of Foamed Waste Glass as Lightweight Concrete Aggregate
The term "lightweight aggregate" describes a range of concrete aggre-
gates that have a weight considerably below that occurring in sand and
gravel. Lightweight aggregates range from the extremely light materials
used for insulative and non-structural concrete all the way to expanded
clays and shales used for structural concrete. In general, the low
density lightweight aggregate concretes at the left end of the concrete
spectrum (Figure 28)are used primarily for insulating purposes, as they
have relatively low compressive strength; while those in the middle range
are used for insulation and fill. The lightweight concretes at the
upper end of the spectrum develop excellent compressive strength and are
found in a number of structural applications.
84
-------
Table 14
Western Market for Non-Residential Rigid Roof Insulation
(millions of dollars and board feet)
Year
1971
1972
1973
1974
1975
Total Value
Non-Resid.
Construction
$4,051.5
4,770.7
5,980.9
6,271.9
5,822.9a
Total Value
of Non-Resid.
Roof Systems
$125.6
147.9
185.4
194.4
180.5"
Mkt. Value of
Rigid Roof
Insulation
$12.6
14.8
18.5
19.4
18,1C
Value of
20% of the
Rigid Roof
Insul. Mkt.
$2.5
3.0
3.7
3.9
3.6
No. of Bd.
Ft. of 20%
of the Roof
Insul. Mkt.
31.8
35.9
41.7
41.4
36.0°
No. of 20
Ton/ Day
Plants Rqd.
2.27
2.56
2.98
2.96
2.57'
Table 15
National Market for Non-Residential Rigid Roof Insulation
(millions of dollars and board feet)
Year
1971
1972
1973
1974
1975
Total Value
Non-Resid.
Construction
$25,590.2
27,118.9
31,761.4
33,859.5
30,336.3"
Total Value
of Non-Resid.
Roof Systems
$793.3
840.7
984.6
1,049.6
940.4b
Mkt. Value of
Rigid Roof
Insulation
$74.3
84.1
98.5
105.0
94. Oc
Value of
20% of the
Rigid Roof
Insul. Mkt.
S15.9
16.8
19.7
21.0
18.8
No. of Bd.
Ft. of 20%
of the Roof
Insul. Mkt.
202.3
201.2
222.1
222.9
188.0d
No. of 20
Ton/Day
Plants Rqd.
14.45
14.37
15.86
15.92
13.43"
(Source and Footnotes apply to both Table 14 and Table 15).
Source: Compiled by Bureau of Economic and Business Research, University of Utah, Salt Lake City, Utah, 1978.
aMcGraw-Hill, Building Cost Calculator and Valuation Guide (New York, New York: McGraw-Hill Information Sys-
tems Company) 1976.
McGraw-Hill, 1976 Dodge Construction System Costs (New York, New York: McGraw-Hill Information Systems
Company) 1975. From this source it was determined that 3,1% of the cost of non-residential construction was the cost of
the roof system.
GSame footnote as b. 10% of the cost of a roof system is the cost of insulation.
Determined by dividing the value of market by average board foot cost of roof insulation,
Determined by dividing the total annual production of a 20 ton/day foamed waste glass plant into millions of
board feet of 20% of the roof insulation market.
85
-------
Table 16
COMPARATIVE VALUES OF SELECTED PHYSICAL PROPERTIES OF
CONCRETE AGGREGATE
CO
01
Property
Size
(diameter in inches)
University of Utah
Foamed Glass
Pittsburgh-Corning (P.C.)
Foamed Glass
Lightweight Aggregate
Assumed similar to
Pittsburgh-Coming (P.C.)
Assumed similar to P.C.
Assumed similar to P.C.
.12- .25(35)*
.25 - .625<27>
.94 - 1.89
(15)
Other
Density
(bulk, Ib./ft9)
14 - 20 Assumed similar to P
.C. 6 - 8<15>
Perlite 7 1/2 - 20
(17, 31, 21)
Compressive Strength
(psi)
>100
(19)
100
(27)*
Perlite .27 - .40 (16)
*Note: Numbered footnotes in parentheses refer to references at end of text material.
aWill not settle.
-------
Table 16 (Continued)
Property
University of Utah
Foamed Glass
Pittsburgh-Corning (P.C.)
Foamed Glass
Other
Temperature
Range
Lightweight Aggregate^
Up to 800°F(18)
-3000 to 1,000°F(15)
Up to 1,500°F
(18)
00
Weight
(lb./ft.9)
Lightweight Concrete Properties
Assumed similar to P.C.
30 - 40
(15)
42 - 60(15>
Perlite 20 - 80(16)
(17)
Vermiculite 35 - 75
Clay and Shale 90 - 110(17>
Cure
Time
Assumed similar to P.C.
Short
(15)
Compressive Strength
(psi)
Assumed similar to P.C.
Achieved quickly
(15)
Vermiculite SO
Perlite 95 - 400 (**>
Gilsonite 1,047 (17) -
Perlite 1,000- 2,000
CA special Perlite for stucco and precast masonry.
-------
Table 16 (Continued)
oo
CO
University of Utah Pittsburgh-Corning (P.C.)
Property Foamed Glass Foamed Glass Other
Lightweight Concrete Properties
Dry Shrink
(percent) Gilsonite .034l '
Thermal ,
Conductivity Perlite .5 - .8
Clay and ( b
Shale 1.7 - 5.2V '
Other Water requirement
drastically reduced
Any lightweight has an insulation value approximately four times greater than ordinary concrete.
-------
Table 16 (Continued)
Property
University of Utah
Foamed Glass
Pittsburgh-Corning (P.C.)
Foamed Glass
Other
Thermal
Conductivity
BTU/wq. ft., hr.
in., degrees F.
Lightweight Aggregate
.4C"9)
.42 - .45
(15)
Diatomite .6 - .
Perlite .3 - .5(18')b
Vermiculite .4 - .
CO
Alkali/Acid
Resistance
Assumed similar to P.Co
Resistance includes
hydrofloric and
phosphoric (15)
Soluability in
Water
24% reduced because of
6% bentonite in foamed
glass
-------
Figure 28. WEIGHT SPECTRUM OF LIGHWEIGHT CONCRETES*
*Source: Lightweight Concrete—Hi story, Applications, Economics.
Washington, D.C.: Expanded Shale Clay and Slate Institute, 1971.
90
-------
C. Characteristics of Foamed Waste Glass as Rigid Insulation
Rigid insulation is insulation which is self-supporting and holds its
shape. Rigid foamed glass insulation is characterized by good strength
characteristics, water resistance, almost perfect resistance to the
migration of moisture and poor abrasion resistance. It may be produced
in blocks, boards, molded, or cut to a shape.
The wood fiber insulation board commonly used in residential construc-
(19)
tion has a thermal conductivity lower than foamed glass. In situations
where vapor barriers are unnecessary, this gives an advantageous position
to the wood fiber insulation. It is normal to install a vapor barrier
between the ambient conditions and the insulated object because of the
intrusion of moisture into insulation which would increase the conductivity.
The best vapor barriers available are blocks of cellular material.
With the exception of cellular (foamed) glass, all rigid cellular insula-
(18)
tions will either transmit liquid water or water vapor. Foamed glass
only requires sealing at the joints because foam glass does not absorb
moisture and is completely impervious to vapors, having a permeability
(24)
rating of 0.00. This vapor barrier also acts as a barrier to rodents
and vermin which is especially important in cold storage units.
One of the physical properties of foamed glass which limits its use
(19)
is the melting point. The foam is initially created at 700° - 800° C.
Therefore, the use of foamed glass is limited to lower temperatures.
When it is desirable to use foamed glass with objects of higher tempera-
tures to create a vapor barrier, an intermediate high temperature insula-
tion is used between the high temperature object and the foamed glass.
91
-------
The concrete spectrum shows the relative weights and differing
applications of the various lightweight aggregates now in use. At the
extreme left are concretes using vermiculite and perlite, which are
sometimes referred to as the "super lightweights". Concrete can be
made with these aggregates weighing as little as 15 or 20 pounds per
cubic foot.
Next are the natural aggregates, pumice and scoria. These can be
made into concrete weighing about 25 or 30 pounds, and it also may run
as high as 65 pounds per cubic foot. Overlapping these are concretes
using coal cinders, with a range from 75 to 120 pounds, and expanded
shale, clay and slate aggregates. Beyond the natural aggregates are
aggregates such as sand and gravel and crushed stone, which produce con-
ventional concretes weighing 135 to 150 pounds per cubic foot.
Some of the physical characteristics of foamed glass and other
products are listed in Table 16. A comparison of these characteristics
indicates that perlite, vermiculite and foamed glass are very similar.
The prime consideration is the density of the aggregate which determines
the density and strength of the concrete. The density of foamed glass is
about the same range as perlite density. The density could be extended
to higher values by less foaming if this were desired.
J. Craig Phillips of the Riverside Cement Company, Riverside, Cali-
fornia, indicates in his research that glass in concrete and specifically
mansonry block seems feasible because there is a lack of significant
(22,23)
glass water reaction. He also indicates that although there is a
slight decrease in strength at 18 months the blocks continued to show a
good strength trend. After the 18-month time period, the test blocks,
92
-------
which had been cured in high humidity to encourage a glass cement
reaction, exceeded the strength of normal, yard-cured blocks. It is
assumed that if crushed glass has such good results when used in con-
crete then foamed glass will also maintain similar properties. The
foamed glass at the University of Utah also contains bentonite which
retards the reaction of glass with water and this should decrease any
potential reactions.
Tests using glass in concrete have shown that test values are within
the American Society for Testing Materials (ASTM) code specification.
The tests were run using crushed glass in concrete rather than foamed
•(22,30)
glass. The tests met the ASTM compressive strength and linear expan-
tion tests.
The use of lightweight aggregate in masonry blocks permits increased
labor productivity because the decreased weight makes for greater speed
and ease of handling. Similar considerations apply in the case of pre-
cast elements using lightweight structural concrete. However, there are
decreases in compressive strength which accompany lighter weights.
The foamed glass could be used in either a foamed bead or crushed
block form. The crushed block would provide a rougher bonding surface
(15)
but the bead provides a greater ease of handling.
D. The Market Potential for Lightweight Concrete Aggregate
Density is the critical variable in determining the use of concrete
and makes it possible to divide the concrete aggregate market into three
basic categories according to the density characteristics of the product.
The density requirements of various concrete products have important impli-
cations for the potential of foamed waste glass as a lightweight concrete
aggregate due to the difficulty of producing high density foamed glass.
93
-------
This important production restriction effectively eliminates foamed waste
glass from the structural concrete market (except possibly for use in
concrete blocks) because concrete used for such purposes must have high
density characteristics to provide adequate structural strength. The
market for foamed waste glass is further reduced by the competitive price
disadvantage of foamed waste glass with respect to pumice and scoria.
The price per short ton of pumice or scoria in 1974 was $2.23 compared to
$66-$93 a short ton for foamed waste glass. It is obvious that only in
those cases or regions where pumice and scoria are unavailable will there
be any opportunity to capture a share of this market.
This means the market for foamed glass as a lightweight aggregate is
limited to insulation concrete which is dominated by two super lightweight
aggregates—vermiculite and perlite. Both of these lightweight aggregates
require high transportation costs which limits the extent of their markets.
These transportation costs make entry into the market attractive, parti-
cularly in those areas currently located long distances from a lightweight
aggregate source. Such conditions obviously offer the most favorable
circumstances for market entry but it should be emphasized that foamed
waste glass is also competitive in those market areas located near a
vermiculite or perlite source. For example, Utah Lumber Company, a Salt
Lake firm.is an expander of vermiculite and supplies most of the state
with the product at approximately $17.50 a cubic foot. Foamed waste
glass can be produced at a price which would be very competitive provided
the other barriers to market entry are surmountable.
Table 17 shows the number of short tons of vermiculite and perlite
used as lightweight aggregates in the United States from 1969-1973. The
data in these tables shows that the lightweight aggregate market has been
94
-------
Table 17
Light Weight Concrete Aggregate Market for Vermiculite and Perlite
(short tons)
Year Vermiculite Perlite Total
1969
1970
1971
1972
1973
110,000
88,000
84,000
104,000
137,000
108,500
79,000
77,000
84,200
75,2(50
218,500
167,000
161,000
188,200
212,200
Source: U.S. Department of the Interior, Minerals Yearbook, Volume I, 1969-1973 (Washington, D.C.: U.S. Govern-
ment Printing Office).
Table 18
Value of Light Weight Aggregate Market
Year
1969
1970
1971
1972
1973
Vermiculite
(dollars/ton)
$ 79.66
85.11
99.93
100.31
106.44
Perlite
(dollars/ton)
$54.95
60.00
60.07
60.17
67.02
Total Value
(millions of dollars)
$14.8
12.3
12.6
15.0
19.6
Source: U.S. Department o« the interior. Minerals Yearbook, Volume I, 1969-1973 (Washington, D.C.: U.S. Govern-
ment Printing Office).
95
-------
very cyclical, being heavily influenced by non-residential construction.
Nevertheless, the lightweight aggregate industry is optimistic about its
future. The energy crisis and the high cost of fuels have forced energy
conservation measures on the building industry leading to a rising demand
for improved techniques and methods of insulation. It is against the
backdrop of the energy crisis that the excellent insulation properties of
lightweight concrete are expected to provide expanding markets for the
industry. The data in these tables do not show any recent dramatic
increases in the use of lightweight aggregates to verify the industry's
optimism but industry spokesmen do agree that 1975 data will show an
increase in the use of lightweight aggregates for insulation concrete.
Of course, any foamed glass producers would benefit from increased demand
for lightweight aggregates for insulation concrete.
The barriers to entry of lightweight aggregate industry for foamed
waste glass are not severe. There are no significant advantages in
economies of scale that would give established firms a competitive advan-
tage, nor is there significant product differentiation or cost advantage
for established firms. Perhaps the largest barrier to entry would be
"product loyalty" for the vermiculite and perlite products. Nevertheless,
entry of foamed waste glass into the lightweight aggregate market seems
possible, but there are some important considerations about the nature of
this market which should be discussed to properly assess the potential of
foamed waste glass as a lightweight aggregate.
The data in Tables 17 and 18 show the total national market in 1973 of
212,240 short tons for lightweight concrete aggregates, valued at
$19,600,000. Given the price and product similarities of foamed waste
glass to lightweight concrete aggregates, foamed waste glass should be
96
-------
able to capture some part of the lightweight aggregate market, but it
should be stressed that the lightweight aggregate market is a very small
market. Five foamed waste glass plants each producing 20 tons of foamed
glass pellets per day would need to capture 10 percent of the current
lightweight aggregate market to be profitable. The total value of this
penetration of foamed waste glass into the lightweight aggregate market
would be only $2 million and represents production of 21,000 tons of
foamed waste glass pellets.
The demand for foamed glass in the eleven state western region was
examined to get a better understanding of the structure of the lightweight
aggregate market.
In 1974 the non-residential construction industry in this region was
valued at $6.3 billion. Of this $6.3 billion valuation, only $6.2 million
was spent on lightweight concrete aggregates. This is a dramatic illus-
tration of the limited market for lightweight concrete aggregates. It is
this limited market which will require a foamed waste glass plant to
capture a significant share of the lightweight concrete aggregate market
if the operation is to be profitable.
For instance, in the eleven state western region, a 20/ton/day foamed
waste glass plant would produce 7,300 tons of foamed waste glass pellets
a year. Priced at $80 a ton this production level means a total output
value of $584,000. Assuming a market capture rate of 10 percent the
lightweight concrete aggregate market in the eleven western states would
be sufficient to sustain only one 20/ton/day foamed waste glass plant.
97
-------
The small market for lightweight concrete aggregates would require
a foamed waste glass plant to broaden its market and produce slab-
foamed waste glass for rigid insulation purposes.
Table 19 and 20 give regional and national market demand data.
98
-------
Table 19
Western Market for Lightweight Concrete Aggregate
(millions of dollars)
Year
1971
1972
1973
1974
1975
Total Value
Non-Resid.
Construction
84,051.5
4,770.7
5,980.9
6,271.9
5,822.9"
Total Value
of Concrete
$121.6
143.1
179.4
188.2
174.7
Total Value
of Concrete
Aggregate
$24.3
28.6
35.9
37.6
34.9
Total Value
of Lightweight
Concrete
Aggregate
S3.9
4.6
5.7
6.0
5.6
Value of
10% of the
Lightweight
Concrete
Aggregate Mkt.
$.390
.460
.570
.600
.560
No. of 20
Ton/Day
Plants Rqd.
.7
.8
1.0
1.0
1.0
Table 20
National Market for Lightweight Concentrate Aggregate
(millions of dollars)
Year
1971
1972
1973
1974
1975
Total Value
Non-Resid.
Construction
$25,590.2
27,118.9
31,761.4
33,859.5
30,336.3a
Total Value
of Concrete
$ 767.7
813.6
952.8
1,015.8
910.1b
Total Value
of Concrete
Aggregate
$153. 5
162.7
190.1
203.2
182.0C
Total Value
of Lightweight
Concrete
Aggregate
S24.5
26.0
. 30.4
32.5
29.1d
Value of
10% of the
Lightweight
Concrete
Aggregate Mkt.
$2.45
2.60
3.04
3.25
2.91
No. of 20
Ton/Day
Plants Rqd.
4.2
4.5
5.2
5.6
5.0"
(Source and Footnotes apply to both Table 7 and Table 8.)
Source: Compiled by Bureau of Economic and Business Research, University of Utah, Salt Lake City, Utah, 1976.
BMcGraw-Hill, Building Cost Calculator and Valuation Guide (New York, New York: McGraw.Hill Information
Systems Company) 1976.
bMcGraw-Hill, 1976 Dodge Construction System Costs (New York. New York: McGraw-Hill Information Systems
Companyl 1975. From this source it was determined that 3% of the cost of non-residential construction was the cost of
concrete.
C20% of the cost of concrete is the cost of aggregate.
dU.S. Department of the Interior, Minerals Yearbook. Volume 1,1969-1973 (Washington, D.C.: U.S. Government
Printing Office).
"Determined by dividing value of annual production of a 20 ton/day plant into value of 10% of lightweight
concrete aggregate market.
99
-------
E. Plant and Production Costs
Plant and production costs for slab production are determined
for four different daily outputs of 10, 20, 30 and 40 tons per day.
A summary of the cost for these plants is shown in Table 21.
A 20 percent rate of return on investment (before taxes) requires
that the foamed waste glass slabs be sold at $.10 per board foot
for the 10/ton/day plant and $.065 per board foot for the 40/ton/day
plant.
Since the selling price of the commercially available cellular
foam glass is approximately $.24 per board foot, it appears that the
method of foaming waste glass as presented in this report is highly
competitive with other similar products.
Plant and production cost for pellet production is determined
for plants producing 10 and 40/tons/day of foamed pellets, Table 22.
100
-------
Table 21
Cost of Production Summary - Slab Production
Plant A
Daily Output
(tons/day) 10
Annual Output
(bdft/year) 7,000,000
Total Capital
Costs $ 1,217,590
Annual Opera-
ting Costs $ 452,420
Cost per
board foot $.06463
Plant B
20
Plant C
30
Plant D
40
14,000,000 21,000,000 28,000,000
$ 1,662,540 $ 2,058,370 $ 2,625,240
$747,250 $1,030,540 $1,331,900
$.05337 $.04908
$.04757
Table 22
Cost of Production Summary - Pellet Production
Daily Output (tons/day)
Annual Output (Ibs/year)
Total Capital Costs
Annual Operating Costs
Costs per Pound
Plant A
10
7,000,000
$ 637,500
$ 326,310
$.04661
Plant B
40
28,000,000
$ 1,142,330
$ 923,980
$.03297
The market prices derived from the above tables were adjusted for infla-
tion when used in the regional and national demand calculations.
101
-------
IX. SLAB PRODUCTION
Plant and production costs are estimated for four different production
levels of 10, 20, 30 and 40/tons/day. The final size of the slabs are
18 inches by 24 inches by 3 inches.
A. Capital Cost
The capital cost estimate is of the general type called a "study
(32)
estimate" by Weaver, et al. This type of estimate, prepared from a flow
chart and a minimum of equipment data, can be expected to be within 30
percent of the actual cost. Major items of equipment for the various
production levels are listed in Table 23. Table 24 shows the equipment
and plant costs. Factors for foundations, piping, instruments, etc.,
(14,20)
covers additional costs for installation. Estimates for some equip-
ment such as the kiln include installation costs.
The field indirect, which covers field supervision, inspection, tempo-
rary construction, equipment rental, and payroll overhead is estimated at
10 percent of the direct cost. Engineering, and administrative and over-
head are each estimated at 5 percent of total construction cost. A con-
tingency allowance of 10 percent and a contractor's fee of 5 percent are
also included.
Plant facilities and plant utilities are estimated at 1 percent of
the plant cost. Start-up cost equals total operating cost for one month.
Working capital is defined as the funds in addition to fixed capital,
land investment and start-up costs that must be provided to operate the
plant. Working capital is estimated from the following items: (1) raw
materials inventory (cost of raw materials for one month), (2) product
102
-------
Table 23
Major Items of Equipment - Slab Production
o
oo
Item
Hopper
Jaw Crusher
Storage bins (CaCC^J
(bentonite)
(ground glass)
Conveyor belts
Impact grinder and classifier
with compressor, dust
collector, fan, etc.
Extruder
Pre-kiln cutter
Dryer
Tunnel kiln
Kiln cars
Post-kiln cutter
Packager
Front-end loader
' •
Quantity
1
1
1
1
1
2
1
1
1
1
1
82
1
f
1
1
Description
10 tons/day
6' x 4'
15" x 18" opening
3' diameter x 7' high
6' diameter x 9' high
7' diameter x 9' high
14" x 12'
6000 Ibs/hr
*
17" diameter ave, 325 hp
20" guillotine cutter
85' long, 16 car capacity
85' long, 8 car capacity
5' x 10'
5" x 18" maximum cross-
section
strapping machine
18 ft capacity
•^••••^'•^••••••^•^•••••••MHMM'HHIM-HIMMM'VVMHMI^^
Quantity
1
1
1
1
1
2
1
1
1
1
1
75
1
1
1"
•HMMIMlaMWIMIBMMIIIIIIII^V^
Description
20 tons/day
6' x 4'
15" x 18" opening
5' diameter x 6' high
8' diameter x 8' high
8' diameter x 10' high
14" x 12'
6000 Ibs/hr
17" diameter ave, 325 hp
20" guillotine cutter
128' long, 24 car capacity
128' long, 12 car capacity
7' x 10'
5" x 18" maximum cross-section
strapping machine
18 ft3 capacity
— — — — • r-- - ,
-------
Table 23 (Continued)
Major Items of Equipment - Slab Production
Item Quantity
Hopper 1
Jaw Crusher 1
Storage Bins (CaCOj) 1
(bentonite) 1
(ground glass) 1
Conveyor belts 2
Impact grinder and classifier 1
with compressor, dust
collector, fan, etc.
Extruder 1
Pre-kiln cutter 1
Dryer 1
Tunnel kiln 1
Kiln cars 75
Post-kiln cutter 1
Packager 1
Front-end loader 1
Description Quantity
30 tons/day
6' x 4'
15" x 18" opening
5' diameter x 8' high
8' diameter x 10' high
8' diameter x 14' high
14" x 12'
6000 Ibs/hr
17" diameter ave, 325 hp
20" guillotine cutter
156' long, 30 car capacity
156' long, 15 car capacity
40
1
1
1
1
1
2
1
1
1
1
1
81 x 10' 105
5" x 18" maximum cross-
section
strapping machine
18 ft3 capacity
1
1
1
Description
tons/day
6' x 41
15" x 18" opening
6' diameter x 8' high
8' diameter x 12' h'igh
8' diameter x 18' high
14" x 12'
6000 Ibs/hr
17" diameter ave, 325 hp
20" guillotine cutter
128' long, 24 car capacity
128' long, 12 car capacity
7' x 10'
5" x 18" maximum cross-section
strapping machine
18 ft3 capacity
-------
Table 2k
Equipment and Plant Cost Summary - Slab Production
o
en
Hopper
Jaw crusher
Storage bins (3)
Conveyor belts (2)
Impact grinder and classifier
Compressor, dust collector, fan, etc.
Extruder
Pre-kiln cutter
Dryer
Tunnel kiln
Kiln cars and trackage
Post-kiln cutter
Packager
Fork!ift - truck and manual
Trucks - open bed and dump truck
Front-end loader
TOTAL
Building
Land
10 tons/day
Cost
$ 1 ,400
3,000
2,000
4,600
46,000
39 ,000
53,000
6,000
40,000
$ 195,000
Labor
$ 60 $
400
100
800
500
500
500
300
500
$ 3,660 $
$
,'
Total
1,460
3,400
2,100
5,400
46,500
39,500
53,500
6,300
40,000
127,500
63,460
40,500
200
11,260
22,000
7,000
470,000
220,700
6,000
TOTAL
$ 696,730
20 tons/day
Cost
$ 1 ,400
3,000
3,000
4,600
46,000
39,000
53,000
6,000
40,000
$ 196,000
Labor
$ 60 $
400
150
800
500
500
500
300
500
$ 3,710 $
$
$
Total
1,460
3,400
3,150
5,400
46,500
39,500
53,500
6,300
70,000
192,000
63,460
40,500
200
IT, 260
25,000
7,000
576,830
347,400
8,000
932,230
-------
Table 2k (Continued)
10 tons/day
20 tons/day
Equipment cost x factor Indicated
Foundations .05
Structures .04
Electrical .03
Painting .01
Instrumentation .01
Piping .03
— . Subtotal
o
CTt
TOTAL DIRECT COSTS
Field Indirect - 10% of Total Direct Costs
TOTAL CONSTRUCTION COSTS
Engineering 5% of Total Construction Costs
Administration Overhead - 5% Construction Costs
Subtotal
Contingency - 10% of above subtotal
Subtotal
Contractor Fee - 5% of above subtotal
TOTAL COST
$
$
$
$
$
$
$
$
$
$
Total
9,750
7,800
5,850
1,950
1,950
5,850
33,150
729,930
72,990
802,920
40,150
40,150
883,220
88,320
971 ,540
48,580
1,020,120
$
$
$
$
$
$
$
$
$
$
Total
9,800
7,840
5,880
1,960
1,960
5,880
33,320
965,550
96,550
1,062,100
53,100
53,100
1,168,300
116,830
1,285,130
64,260
1,349,390
-------
Table
30
Equipment cost x factor indicated
Foundations .05
Structures .04
Electrical .03
Painting .01
Instrumentation .01
Piping .03
Subtotal
TOTAL DIRECT COSTS
Field Indirect - 10% of Total Direct Costs
TOTAL CONSTRUCTION COSTS
Engineering - 5% of Total Construction Costs
Administration Overhead - 5% Construction Costs
Subtotal
Contingency - 10% of above subtotal
Subtotal
Contractor Fee - 5% of above subtotal
24 (Cont
tons/day
$
$
$
$
$
$
$
$
$
$
inued)
Total
9,860
7,890
5,920
1,970
1,970
5,920
33,530
1,165,540
116,550
1,282,090
64,100
64,100
1,410,290
141,030
1 ,551 ,320
77,570
40 tons/day
Total
$ 9,900
7,920
5,940
1,980
1,980
5,940
$ 33,660
$ 1,481,160
$ 148,120
$ 1,629,280
$ 81 ,460
81,640
$ 1,792,200
$ 179,220
$ 1 ,971 ,420
$ 98,570
TOTAL COST
$ 1,628,890
$ 2,069,990
-------
Table 2k (Continued)
Equipment and Plant Cost Summary - Slab Production
o
GO
Hopper
Jaw crusher
Storage bins (3)
Conveyor belts (2)
Impact grinder and classifier
Compressor, dust collector, fan, etc.
Extruder
Pre-kiln cutter
Dryer
Tunnel kiln
Kiln cars and trackage
Post-kiln cutter
Packager
Forklift - truck and manual
Trucks - open bed and dump truck
Front-end loader
TOTAL
Building
Land
TOTAL
30 tons/day
Cost
$ 1 ,400
3,000
4,200
4,600
46,000
39,000
53,000
6,000
40 ,000
$ 197,200
Labor
$ 60
400
200
800
500
500
500
300
500
$ 3,760
Total
$ 1 ,460
3,400
4,400
5,400
46,500
39,500
53,500
6,300
92,000
229,500
85,290
40,500
200
11,260
38,000
7,000
$ 664,210
$ 457,800
10,000
$1,132,010
40 tons/day
Cost
$ 1 ,400
3,000
5,000
4,600
46,500
39,000
53,000
6,000
40 ,000
$ 198,000
Labor
$ 60
400
250
800
500
500
500
300
500
$ 3,810
Total
$ 1 ,460
3,400
5,250
5,400
46,500
39,500
53,000
6,300
122,500
384,000
100,330
40,500
200
11,260
41 ,000
7,000
$ 868,100
$ 567,400
12,000
$ 1,447,500
-------
and in-process inventory (total operating cost for one month), (3) accounts
receivable (total operating cost for one month), and (4) available cash
(direct expenses for one month table #25).
B. Operating Costs
The estimated operating costs (Table 26) are based on 350 days of
operation per year, which allows 15 days for inspection, maintenance, and
unscheduled interruptions. These costs are divided into direct, indirect,
and fixed costs.
Direct costs include raw materials, packaging costs, utilities,
i
labor, maintenance, payroll overhead, and operating supplies. Direct
labor requirements are shown in Table 27. Supervisory costs are esti-
mated at 12 percent of the labor costs. Payroll overhead includes vaca-
tions, social security, and fringe benefits.
Indirect costs are estimated at 30 percent of direct labor and main-
tenance. It includes the expenses of accounting, plant protection and
safety, plant administration, and company overhead. Fixed cost includes
property taxes, insurance, and depreciation. Depreciation is based on a
straight-line, 15-year period.
109
-------
Table 25
Estimated Capital Cost - Slab Production
Fixed Capital:
Equipment and Plant Cost
Start-Up Cost
Plant Facilities
Plant Utilities
Fixed Capital Cost
Working Capital:
Raw Materials
Product In-Process Inventory
Accounts Receivable
Available Cash
Working Capital Costs
Total Capital Costs
10 tons/day
$ 1,020,120
33,420
20,400
20,400
$ 1,094,340
$ 3,670
63,240
31 ,620
24,720
$ 123,250
$ 1,217,590
20 tons/day
$ 1,349,390
45,730
26,990
26,990
$ 1,449,100
$ 7,350
108,440
54,220
43,430
$ 213,440
$ 1,662,540
30 tons/day
$ 1,628,890
63,300
32,580
32,580
$ 1,757,350
$ 11,030
152,230
76,120
61 ,640
$ 301 ,020
$ 2,058,370
40 tons/day
$ 2,069,990
81 ,600
41 ,400
41 ,400
$ 2,234,390
$ • 14,710
197,160
98,580
80,400
$ 390,850
$ 2,625,240
-------
Table 26
Estimated Annual Operating Costs - Slab Production
10 tons/day
Direct Cost:
Raw Materials:
Glass at $12/ton
Bentonite at $6/ton
CaC03 at $30/ton
TOTAL
Packaging Materials
TOTAL
Utilities
Natural Gas at $.35/MCF
Electric Power at $.015/kwhr
Grinding
Other equipment
TOTAL
Direct Labor
Labor at $3.35/hour
Supervision at 12% of labor
TOTAL
Plant Maintenance
Labor at $3.35/hour
Materials
TOTAL
20 tons/day
Annual
$
$
$
$
42
1
1
44
50
50
Costs
,000
,050
,070
J20
,920
,120
Cost/board
.600
.015
.015
.630
.727
.727
foot
t
t
t
t
Annual
$ 84
2
2
i. 88
$ 101
$_ 101
Costs
,000
,100
,130
^230
,840
.840
Cost/board foot
.600 t
.015
.'015
.630 t
.727 *
.727 *
$ 5,000
.071 t
$ 10,000
.071 *
L
$
L
$
6
3
14
•• "
78
9
87
E-S— — — -s
14
25
39
,150
,000
J.50
,400
,410
48!°
,000
,320
,320
,088
.043
^20g t
1.120
-------
r-o
Table 26 (Continued)
Estimated Annual Operating Costs - Slab Production
30 tons/day
Direct Cost:
Raw Materials:
Glass at $12/ton
Bentonite at $6/ton
CaCO, at $30/ton
TOTAL
Packaging Materials
TOTAL
Utilities
Natural Gas at $.35/MCF
Electric Power at $.015/kwhr
Grinding
Other Equipment
TOTAL
Direct Labor
Labor at $3.35/hour
Supervision at 12% of labor
TOTAL
Plant Maintenance
Labor at $3.25/hour
Materials
TOTAL
Annual Costs Cost/board foot
$
$
$
$
$
$
$
$
$
126,000
3,150
3,200
132,350
153,760
153,760
15,000
18,450
9,000
42,450
188,300
22,600
210,900
21 ,000
41 ,450
.600 t
.015
.015
.630 i
.727 t
.727 i
.071 t
.088
.043
.202 4
.896 *
.108
1.004 i
.100 i
.197
$ 62,450 .297.*
40 tons/day
Annual Costs Cost/board foot
$
$
$
$
$
$
$
$
$
168,000
4,200
4,270
176,470
203,680
203.680
20,000
24,600
12,000
56,600
239,400
28,730
268,130
24,500
• 54,770
.600 it
.015
.015
.630 it
.727 it
.727 i
.071 t
.088
.043
.202 t
.855 it
.102
.957 t
.087 «t
.196
ji Zii_229. -i-28,3 i
-------
Table 26 (Continued)
10 tons/day
20 tons/day
Direct Cost: (cont)
Payroll overhead - 255S of payroll
Operating supplies - 20% of plant
maintenance
TOTAL
Subtotal
Miscellaneous costs - 10%. of above
subtotal
TOTAL DIRECT COSTS
Indirect Costs - 30% of Direct Labor
and Maintenance
Marketing expense - 8% of Direct Cost
TOTAL
Fixed Cost
Taxes - 1,0% of Total Plant Cost
Insurance, 1.0% of Total Plant Cost
VARIABLE.COSTS
Depreciation - 15 year life
TOTAL OPERATING COSTS
Annual Costs
$ 25,450
7,860
i_J3t31£
$ 269,630
$ 26,960
$J_9_6_._590
$ 38,740
23,730
$ 359.060
$ 10,200
10,200
|_379_.46Q
$ 72,960
$ 452,420
Cost/board foot
.364 t
.112
JZ6= *
3.852 «
.385 t
4^1237 t
.553 it
.339
£ J29 t
.146 t
.146
5.42J) t
1.042 t
6.463 t
Annual Costs
$ 42,200
10,310
LJ?jLJJ£
$ 437,740
$ 47,370
i_5?-LJ-lP-
$ 60,860
41 ,690
1623,660
j— „,,,, _. — _j . ._
$ 13,490
13,490
i 650^640
db^=r===:=.-l==ir.=r=
$ 96,610
$ 747,250
Cost/board foot
.301
-------
Table 26 (Continued)
30 tons/day
40 tons/day
Direct Costs: (cont)
Payroll Overhead - 25% of payroll
Operating Supplies - 20% of plant
Maintenance
TOTAL
Subtotal
Miscellaneous costs - 10% of ^bove
subtotal
TOTAL DIRECT COSTS
Indirect Costs - 30% of Direct Labor
and Maintenance
Marketing expense - 8% of Direct Cost
TOTAL
Fixed Cost
Taxes - 1.0% of Total Plant Cost
Insurance, 1.0% of Total Plant Cost
VARIABLE COSTS
Depreciation - 15 year life
TOTAL OPERATING COSTS
Annual Costs Cost/board foot
$
$
$
$
JL.
$
$
$
$
$
$ i
57,980
12,490
70,470
672,380
67,240
739.620
82,010
59,170
880,800
16,290
16,290
913,380
117,160
,030,540
.276 i
.059
.335 t
3.202 t
.320 i
3.522 t
.391 t
.282
4.194 t
.078 *
.078
4.349 i
.558 i
4.908 t
Annual Costs
$ 73,160
15,850
$ 89,010
$ 873,160
$ 87,320
L^Lm
$ 104,220
76,840
$ 1,141,540
$ 20,700
20,700
$ 1J82L.940
$ 148,960
$ 1,331 ,900
Cost/board fool
.261 t
.057
.318 t
3.118 i
.312 .4
JL430 t
.372 *
.274
4.077 i
.074 i
.074
4.225 4
,532 i
4.757 t
-------
Table 27
Direct Labor Requirements - Slab Production
Raw Materials Handling
Grinding Equipment
Dryer and Kiln
Car Loading and Unloading
Post - Kiln Cutter
Packager
Warehouse
Total
10 tons/day
Shift
per
Day
1
1
3
1
1
1
1
Days
per
Week
5
5
7
5
5
5
5
Total
Men
1.0
1.0
4.2
2.0
1.0
1.0
1.0
11.2
20 tons/day
Shift
per
Day
2
2
3
2
2
2
2
Days
per
Week
5
5
7
5
5
5
5
Total
Men
1.5
1.5
6.3
4.0
2.0
2.0
2.0
19.3
30 tons/day
Shift
per
Day
3
3
3
3
3
3
3
Days
per
Week
5
5
7
5
5
5
5
Total
Men
2.0
2.0
7.9
6.0
3.0
3.0
3.0
26.9
40 tons/day
Shift
per
Day
3
3
3
3
3
3
3
Days
per
Week
7
7
7
7
7
7
7
Total
Men
2.5
2.5
9.2
8.0
4.0
4.0
4.0
34.2
-------
X. PELLET PRODUCTION
Plant and production costs are estimated for two different production
levels of 10 and 40/tons/day. The foamed pellet size ranges from % inch
to 5/8 inch in diameter. The capital and operating costs are estimated
using the same method as described in Appendix A. Major items of equip-
ment are listed in Table 28. The plant and total capital costs are
tabulated in Tables 29 and 30. Annual operating costs are shown for the
two plant sizes in Table 31. Direct labor requirements are shown in
Table 32 Supervision is estimated at 18 percent of labor cost.
116
-------
Table 28
Major Items of Equipment - Pellet Production
Item
Hopper
Jaw crusher
Storage bins
Conveyor belts
Impact grinder and classifier
with compressor, dust
collector, fan, etc.
Extruder
Rotary dryer
Rotary kiln and pollution
control equipment
Packager
Front-end loader
Quantity
10
1
1
1
1
3
1
1
1
1
1
1
Description
tons/day
6' x 4'
15" x 18" opening
6' diameter x 9' high
7' diameter x 9' high
14" x 12'
6000 Ibs/hr
6" ave.
2' I D x 25' long
2' I D x 25' l.ong
3 ft3 bags
18 ft3 capacity
Quantity Description
40 tons/day
1 6' x 4'
1 15" x 18" opening
1 8' diameter x 12"1 high
1 8' diameter x 18' high
3 14" x 12'
1 6000 Ibs/hr
1 6" ave.
1 4' I D x 30' long
1 4' I D x 30' long
1 3 ft3 bags
1 18 ft3 capacity
-------
CO
Table 29
Equipment and Plant Cost Summary - Pellet Production
Hopper
Jaw crusher
Storage bins (2)
Conveyor belts (3)
Impact grinder and classifier
Compressor, dust collector, fan, etc.
Extruder
Rotary dryer
Rotary kiln
Pollution control equipment
Packager
Forklifts - truck and manual
Trucks - open bed and dump truck
Front-end loader
TOTAL
Building
Land
10 tons/day
Cost
$ 1 ,400
3,000
3,000
6,900
46,000
39 ,000
7,500
15,000
30 ,000
10,000
2,500
$ 164,300
Labor
$ 60 $
400
200
1,200
500
500
500
1,500
3,500
500
100
$ 8,960 $
$
Total
1,460
3,400
3,200
8,100
46,500
39,500
8,000
16,500
33,500
10,500
2,600
11,260
22,000
7,000
213,520
127,700
6,000
TOTAL
$ 133,700
40 tons/day
Cost
$ 1 ,400
3,000
3,000
6,900
46,000
39,000
7,500
22,500
45,000
15,000
2,500
$ 191 ,800
Labor
$ 60 $
400
200
1,200
500
500
500
2,000
4.500
800
100
$10,760 $
$
$
Total
1,460
3,400
3,200
8,100
46,500
39,500
8,000
24,500
49,500
15,800
2,600
11,260
40,000
7,000
260,820
290,600
10,000
561,420
-------
Tabl
10
s
Equipment cost x factor indicated
Foundations .05
Structures .04
Electrical .03
Painting .01
Instrumentation .01
Piping .03
Subtotal
TOTAL DIRECT COSTS
Field Indirect - 10% of Total Direct Costs
TOTAL CONSTRUCTION COSTS
Engineering - 5% of Total Construction Costs
Administration Overhead - 5% Construction Costs
Subtotal
Contingency - 10% of above subtotal
Subtotal
Contractor Fee - 5% of above subtotal
e 29 (Cont
tons/day
$
$
$
$
$
$
$
$
$
$
inued)
Total
8,220
6,570
4,930
1,640
1,640
4,930
27,930
375,150
37,520
412,670
20,630
20,630
453,930
45,390
499,320
24,970
40 tons/day
Total
$ 9,590
7,670
5,750
1,920
1,920
5,750
$ 32,600
$ 594,020
$ 59,400
$ 653,420
$ 32,670
32,670
$ 718,760
$ 71 ,880
$ 790,640
$ 39,530
TOTAL COST
$ 524,290
$ 830,170
-------
Table 30
Estimated Capital Cost - Pellet Production
ro
o
Fixed Capital:
Equipment and Plant Cost
Start-up Cost
Plant Facilities
Plant Utilities
Fixed Capital Cost
Working Capital
Raw Materials
Product In-Process Inventory
Accounts Receivable
Available Cash
Working Capital Cost
Total Capital Cost
10 tons/day 40 tons/day
$ 524,290
21,270
10,490
10,490
$ 566,540
$ 3,670
24,050
24,050
19,260
$ 71,030
$ 637,570
$ 830,170
60,750
16,600
16,600
$ 924,120
$ 14,710
71,860
71,860
59,780
$ 218,210
$1,142,330
-------
Table 31
Estimated Annual Operating Costs - Pellet Production
10 tons/day
Direct Cost:
Raw Materials:
Glass at $12/ton
Bentonite at $6/ton
CaC03 at $30/ton
TOTAL
Packaging Materials
TOTAL
Utilities
Natural gas at $.35/MCF
Electric power at $.015/kwhr
Grinding
Other equipment
TOTAL
Direct Labor
Labor at $3.35/hour
Supervision at 18% of Labor
TOTAL
Plant Maintenance
Labor at $3.35/hour
Materials
TOTAL
40 tons/day
Annual Costs Cost/board foot
$
$
$
$
$
$
$
$
$
42,000
1,050
1,070
44.120
33,340
33^340
2,500
6,150
3,000
11,650
57,400
10,330
67,730
14,000
13,320
.600 t
.015
.015
.630 t
.476 t
.476 t
.036 t
.088
.043
.167 t
.820 t
.148
.968 t
.200 t
.190
Annual Costs Cost/board foot
$ 168,000
4,200
4,270
$ 176,470
$ 133,360
$ 133,360
$ 10,000
24,600
12,000
$ 46,600
$ 155.400
27,970
$ 183,370
$ 24,500
25,770
.600
-------
IN3
no
Direct Costs: (cont)
Payroll Overhead - 25% of payroll
Operating Supplies - 20% of Plant
Maintenance
TOTAL
Subtotal
Miscellaneous costs - 10% of above
subtotal
TOTAL
Indirect Costs - 30% of Direct Labor
and Maintenance
Marketing Expense
TOTAL
Fixed Cost
Taxes - 1.0% of Total Plant Cost
Insurance - 1.0% of Total Plant Cost
VARIABLE COSTS
Depreciation - 15 year life
TOTAL OPERATING COSTS
Table 31 (Continued)
10 tons/day
40 tons/day
Annual Costs Cost/board foot
$
JL
$
$
$
$
$
$
$
$
20,430
5,460
25^890
210,050
21,010
231 ,060
28,520
18,480
278,060
5,240
5,240
288,540
37,770
.292 4
.078
.370
-------
Table 32
Direct Labor Requirements - Pellet Production
CO
taw Materials Handling
Grinding Equipment
Dryer and Kiln
Packager
Warehouse
Total
10 tons/day 40 tons/day
Shifts
1
1
3
1
1
Days
per
Week
5
5
7
5
5
Total
Men
1.0
1.0
4.2
1.0
1.0
8.2
Shifts
3
3
3
3
3
Days
per
Week
7
7
7
7
7
Total
Men
2.5
2.5
9.2
4.0
4.0
22.2
-------
REFERENCES
1. B. D. Marchant, "Foamed Glass Insulation from Waste Glass",
report to N.S.F. Utec MSE 72-201, December 1972.
2. H. Scholze, "Rases and Water in Glass I"
Glass Ind 47^ (10) 546-551 (1966)
"II" Glass Ind 47^ (11) 622-628 (1966)
"III" Glass Ind 47 (16) 670-675 (1966)
3. I. B. Cutler, "Effect of Water Vapor on the Sintering of Glass
Powder Compacts", J. Amer. Ceram. Soc. 52_ (1) 11-13 (1969).
4. R. J. Charles, "Static Fatigue of Glass I", J. Appl. Phys.,
29_ (11) 1549-1 553 (1958).
5. M. A. Rana and R. W. Douglas, "Reactions between Glass and Water
I", Phys. and Chem. of Glasses 2^ (6) 79-145 (1961).
6. Ibid, II Phys. and Chem. of Glasses 2_ (6) 196-205 (1961).
7. C. R. Das and R. W. Douglas, "Studies on the Reaction Between
Water and Glass III", Phys. and Chem. of Glass 8^ (5) 178-184 (1967).
8. R. H. Doremos, "Glass Science", John Wiley & Sons, New York, 129
(1973).
9. E. Brueche and H. Poppa, "Glastechn", Ber. 3p_ (1957) 163.
10. G. Eiserman, "Glass Electrodes for Hydrogen and Other Cations",
M. Dekker, New York (1967) 133.
11. Mantel!, C. L. Industrial Carbons. N.Y., D. Van Nostranel Co,
1928. 151.
12. Ibid. p. 150.
13. Kraus, G. "Interactions of Elastomers and Reinforcing Fillers,"
Rubber Chemistry and RCT Technology, 38: 1071, 1965.
14. H. C. Bauman, Fundamentals of Cost Engineering in the Chemical
Industry, Reinhold Publishing Co., New York, 1964, p. 364.
15. Geoff Edwards, "Lighter Concrete Has Glass Base," Engineering
209 (April 3, 1970): 327.
16. U. S. Department of Interior, Bureau of Mines, "Perlite," By
Timothy C. May, Mineral Facts and Problems, 1970 ed. (Washington,
D.C.: Government Printing Office, 1970), p. 655.
124
-------
17. U. S. Department of the Interior, Bureau of Mines, Minerals
Yearbook Area Reports: Domestic, 1972 (Washington, D.C.I
Government Printing Office, 1974.
18. John F. Malloy, Thermal Insulation (New York: Van Nostrad
Reinhold Company, 1969).
19. Bruce D. Marchant, "Foamed Glass Insulation from Waste Glass,"
(Salt Lake City, Utah: Materials Science and Engineering
Department, University of Utah, December 1972. Final Report to
National Science Foundation on Research Grant GY-9641).
i
20. H. E. Mills, "Costs of Process Equipment," Chem. Eng., 71, 133-156
(1964). —
21. Perlite Institute, Basic Facts About Perl He—The White Aggregate
(New York: The Perlite Institute, n.d.).
22. J. Craig Phillips, "Refuse Glass Aggregate in Portland Cement
Concrete," Proceedings of Symposium on Utilization of Waste Glass
in Secondary Products (Albuquerque, New Mexico: University of
New Mexico, Technology/Application Center, January 1973).
23. Personal Interview with J. Craig Phillips, Research Engineer with
Riverside Cement Company, August 1974.
24. Pittsburgh-Corning Corporation "FOAMGLAS-Board Insulation,"
March 1970 (Product Data Sheet issued irregularly by Pittsburgh
Corning Corporation, Pittsburgh, Pennsylvania).
25. Pittsburgh-Corning Corporation, "FOAMGLAS-Board Insulation for
Plazas, Parking Decks, Concrete Floors," June 1970 (Product Data
Sheet issued Irregularly by Pittsburgh-Corning Corporation,
Pittsburgh, Pennsylvania).
26. Pittsburgh-Corning Corporation, "FOAMGLAS-Cellular Glass Insulation,"
December 1967 (Product Data Sheet issued irregularly by Pittsburgh-
Corning Corporation, Pittsburgh. Pennsylvania).
27. Pittsburgh-Corning Corporation, "New CELRAMIC—Glass Nodules for
the Thermosetting Resins," (Product Data Sheet issued-irregularly
by Pittsburgh-Corning Corporation, Pittsburgh, Pennsylvania).
28. Telephone conversation with representative of Pittsburgh-Corning
Corporation, Sadalia, Missouri, November 1974.
29. Robert J. Ryder and John H. Abrahams, Jr., "Separation of Glass
From Municipal Refuse-- A Review," Solid Waste Resources Conference,
sponsored by Battelle Memorial Institute, Columbus, Ohio, May 31, 1971.
125
-------
30. S. Russell Stearns, "Glass Aggregate in Concrete," Proceedings
of Symposium on Utilization of Waste Glass in Secondary Products
(Albuquerque, New Mexico: University of New Mexico, Technology
Application Center, January 1973).
31. J. Derle Thorpe, Assistant Professor of Civil Engineering, Utah
State University, personal conversation, August 1974.
32. J. B. Weaver, H. C. Bauman, and W. F. Heneghan, "Cost and Profita-
bility Estimation,"Chapter in Perry Chemical Engineers' Handbook
(R. H. Perry, C. H. Chilton, and S. D. Kirkpatrick, eds.) McGraw-
Hill Book Company, New York, 1963, p. 26.
126
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
gEPORT NO.
EPA-600/3-77-Q30
3. RECIPIENT'S ACCESSION-NO.
AND SUBTITLE
5. REPORT DATE
August 1977 (Issuing Date)
Foam Glass Insulation from Waste Glass
6. PERFORMING ORGANIZATION CODE
7Wendell G. Oakseson', June-Gunn Lee, S. K. Goyal,
Thayne Robson, and Ivan B. Cutler
8. PERFORMING ORGANIZATION REPORT NO.
tfpERFORMING ORGANIZATION NAME AND ADDRESS
'Department of Material Science & Engineering
University of Utah
Salt Lake City, Utah 84112
10. PROGRAM ELEMENT NO.
1DB314; ROAP 21BFS
11. CONTRACT/GRANT NO.
R800937-02
12. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research Laboratory--Cin. ,OH
Office of Research & Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA/600/14
15. SUPPLEMENTARY NOTES
Project Officer: Charles J. Rogers (513) 684-7881
16. ABSTRACT • —
Waste glass has proven to be effective for the production of foam glass insulation
both in the bulk or rigid board form and pellet form. Problems inherent with the
use of water, carbon black and calcium carbonate as the foaming agents, have been
identified and many have been solved by various techniques.
Water was found to be best suited for micron sized particles to 0.6 cm pellets,
while carbon and CaC03 yielded better products for larger objects.
Large amounts of water can be rapidly incorporated into glass by using a sodium
hydroxide (NaOH) solution in a heated autoclave. Smaller amounts can be incorporated
into the glass by placing pellets formed by adding NaOH to a glass-clay mixture and
directly heating in a furnace.
The foaming process with carbon black was examined by analysis of the density, pore
size, and open porosity of the foamed piece. Also, the addition of clay made
foam glass less soluble to water.
DESCRIPTORS
KEY WORDS AND DOCUMENT ANALYSIS
b.lDENTIFIERS/OPEN ENDED TERMS
water
foam
glass
utilization
insulation
pellets
calcium carbonates
autoclaves
carbon black
sodium hydroxide
foaming agents
"•DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
foam production
municipal solid waste
COSATI Field/Group
11B
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
21. NO. OF PAGES
137
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
PA Form 2220-1 (9-73)
^-GOVERNMENT PRINTING OFrlCE: 1977-757-056/6491
127
------- |