ACID HYDROLYSIS OF REFUSE
U S ENVIRONMENTAL PROTECTION AGENCY
-------
This report has been reviewed by the U.S. Environmental Protection
Agency and approved for publication. Approval does not signify that
the contents necessarily reflect the views and policies of the U.S.
Environmental Protection Agency, nor does mention of commercial products
constitute endorsement or recommendation for use by the U.S. Government.
-------
ACID HYDROLYSIS OF REFUSE
This final open-file report (SW-15rg.of) on work
performed under solid waste management research
grant no. EC-00279 to Dartmouth College was written by
ROBERT D. FAGAN and, except for minor changes in the
introductory pages, is reproduced as received from the grantee.
U.S. ENVIRONMENTAL PROTECTION AGENCY
1971
-------
An environmental protection publication
in the solid waste management series (SW-15rg.of).
-------
CONTENTS
ABSTRACT
INTRODUCTION
1. SOLID WASTE AND ITS DISPOSAL 4
1.1 Current Disposal Methods 4
1.2 Refuse Composition 4
1.3 Utilization of Municipal Refuse 8
1.4 Summary of Alternatives 8
2. THE HYDROLYSIS OF CELLULOSE 11
2.1 The Chemistry of Cellulose 11
2.2 Dilute Acid Hydrolysis of Wood Cellulose 13
3. EXPERIMENTAL APPARATUS AND PROCEDURE 20
3.1 Kinetic Model 21
3.2 Isothermal Kinetic Study 22
3.3 Acid Injection Bomb 25
3.4 Nonisothermal Kinetic Study 26
4. EXPERIMENTAL RESULTS AND ANALYSIS 29
4.1 Isothermal Analysis Results 29
4.2 Non-Isothermal Analysis Results 31
4.3 Experimental Analysis Conclusions 41
5. PLANT DESIGN 44
5.1 Separation and Pretreatment System 46
5.2 Hydrolysis Reactor System 51
5.3 Concentration of Sugar Solution and Heat Recovery 57
5.4 Design Calculations 61
6. PLANT ECONOMICS 74
6.1 Capital Costing Procedure 74
6.2 Manufacturing Cost Estimation 78
6.3 Manufacturing Cost Analysis 80
6.4 The Marketability of a Glucose Solution 88
iii
-------
Page
7. GLUCOSE AS A POTENTIAL RAW MATERIAL 95
7.1 Monosodium Glutamate 95
7.2 Citric Acid 96
7.3 Butanol 96
7.4 Lactic Acid 96
7.5 Sorbital and Oxalic Acid 96
7.6 Conclusions 97
8. RECOMMENDATIONS FOR FUTURE WORK 98
8.1 Metal Ion Effect on Hydrolysis 98
8.2 Flow Reactor 99
8.3 General Plant Considerations 99
8.4 Sugar as a Raw Material 100
9. SUMMARY AND CONCLUSIONS 101
REFERENCES 104
Appendix I QUANTITATIVE SACCHARIFICATION 107
Appendix II SUGAR TEST 108
Appendix III NONISOTHERMAL HYDROLYSIS RESULTS 113
Appendix IV COMPUTER PROGRAMS 116
TABLES
1-1 Composition and Analysis of an Average Municipal
Refuse 4
1-2 Municipal Solid Wastes Composition 5
1-3 Packaging Materials Consumption 7
1-4 Alternative Disposal Methods 10
4-1 Quantitative Saccharification 32
4-2 Carbohydrate Composition of Kraft Paper Pulp 32
5-1 Material Balance - Representative 250 Ton Plant 63-64
5-2 Evaporator Design Data 73
6-1 Manufacturing Cost Analysis 79
IV
-------
Page
6-2 Effect of Acid Concentration and Temperature
on Manufacturing Cost 82
6-3 Ethanol Production by Fermentation - Cost Analysis 92
FIGURES
2-1 SUGAR YIELD VERSUS RESIDUAL POTENTIAL SUGAR 18
4-1 SUGAR YIELD VERSUS REACTION TIME 30
4-2 LOG (PRE-EXPONENTIAL FACTORS) VERSUS LOG (ACID
CONG.) 35
4-3 SUGAR YIELD VERSUS NONISOTHERMAL TEMPERATURE (TIME) 37
4-4 SUGAR YIELD (KRAFT FIBERS) VERSUS NONISOTHERMAL
TEMPERATURE 39
4-5 PREDICTED ISOTHERMAL SUGARS VERSUS TIME 42
5-1 SEPARATION SYSTEM 47
5-2 REACTOR SYSTEM 48
5-3 EVAPORATION AND'HEAT RECOVERY 49
5-4 CORROSION RATE OF CARPENTER STEEL 55
5-5 MATERIAL BALANCE SHEET 65
6-1 CAPITAL & MANUFACTURING COST ANALYSIS - REPRESENTATIVE
250 TON PLANT 76-77
6-2 SUGAR COST VERSUS PLANT CAPACITY 84
6-3 CAPITAL COST VERSUS PLANT CAPACITY 85
-------
-------
ABSTRACT
To recover some of the value of municipal solid waste, a new dis-
posal process was studied which uses the cellulose portion of refuse,
mainly paper, as a raw material for the production of sugar by acid
hydrolysis. Experiments to predict the sugar yield were performed,
and the resulting information was used in the design of a proposed
hydrolysis plant to establish the economics of the system.
A sugar yield of 51% is predicted with 1% acid at 230°C. With
this yield it is concluded that sugars from the hydrolysis plant can
be produced at costs competitive with molasses sugars. This conclusion
is valid for municipalities with populations greater than 200,000,
producing refuse containing 50% paper, or populations of 100,000 pro-
ducing refuse containing 60% paper. It is also shown that such sugars
could be used to produce ethanol at a price comparable to the existing
market price. Refuse disposal by acid hydrolysis under these conditions
might save $600,000 a year for a community of 200,000 people which was
previously disposing of its waste at $3/ton.
-------
-------
INTRODUCTION
The lowest predicted disposal cost for modern incineration is
$3/ton. At this rate, solid waste disposal would cost a city of
approximately 200,000 people $600,000 per year. Andrew Porteous, a
1967 graduate of Thayer School of Engineering, Dartmouth College, pro-
posed a disposal process which would utilize refuse as a raw material.
Porteous' process consisted of hydrolyzing the available cellulose con-
tent of refuse, mainly paper, to sugar. The sugar in turn could be
used to produce the saleable product of ethanol by fermentation. His
economic study showed that in some cases such a process would not only
eliminate the disposal cost but make a profit as well.
Because of the great economic potential of the process, a research
program was established to study the process in greater depth. Porteous'
economic analysis was based on the assumptions that (1) paper cellulose
would hydrolyze like wood cellulose, and (2) the kinetic model experi-
mentally determined for wood cellulose at temperatures below 200°C could
be used to predict higher yields at 230°C. In order to test these
assumptions it was necessary to establish new hydrolysis techniques
and perfect quantitative analysis procedures for sugar and cellulose.
The information gained from laboratory work and process analysis was
used to modify Porteous' original process design and reevaluate the
economics of the system.
- 3 -
-------
- 4 -
1. SOLID WASTE AND ITS DISPOSAL
1.1 Current Disposal Methods
There are currently just two acceptable means of solid waste
disposal—incineration and sanitary landfill; the former requires a
large capital investment, and the latter, large areas of available
land. The cost of disposal varies with the area, and ranges from $1 to
$12 per ton with an average value of approximately $4.50 per ton.
1,2 Refuse Composition
A knowledge of the composition of solid waste is necessary, if any
technological advancement is to be made in the field of solid waste
disposal. The classification of solid waste by weight and composition
is hampered by its heterogeneous composition and its offensive nature.
Table 1-1 is a breakdown of solid waste by composition derived from a
(2)
study done by Bell in 1963. Table 1-2 is a compilation of municipal
solid waste composition obtained through the Public Health Service.
The actual composition of municipal refuse varies with the location
sampled and the season of the year. Thus it is difficult to arrive at
a set average composition for the country. Generally, it is possible
to say that paper is the main component of refuse, 40 to 60% by weight,
followed by food waste, and then metal and glass products. It is
believed that there will be an increasing amount of paper and synthetic
materials produced each year. A large part of this waste is derived
(3)
from packaging material. Table 1-3 shows the increasing trend of
solid waste produced from packaging material consumption alone. An
-------
- 5 -
.-1
(2)
Approximate
Percent of Analysis
Total "as received"
Refuse basis, moisture
Ruoo^sn, o 4 'o
Paper, mixed
V:GOC and bark
Grass
j_j r u s ii
Greens
Leaves , rips
Leather
Rubber
Plastics
Oils, Paints
Linoleum
Rags
Sweepings, street
Dirt, household
Unclassified
Focc. wastes, 12%
Garbage
~ a c s
Nor.combustibles , 24%
Xe'callics
Glass and Ceramics
A s n e s
Organic Analysis of Composi
Moisture
Cellulose, Sugar, Starch
Lipids (fats , oils , waxes)
Protein, 6.25N
Other organic (plastics)
Ash, metal /glass , etc.
Analysis of Composite
Moisture 20.73
Carbon 28.00
Total Hydrogen 3.50
Available Hydrogen 0.71
Oxygen 22.35
Nitrogen 0.33
42.0
2.4
4.0
1.5
1.5
5.0
0.3
0.6
0.7
0.8
0.1
0.6
3.0
1.0
0.5
10.0
2.0
8.0
6.0
10.0
te
Percent
20.73
46.63
4.50
2.06
1.15
24.95
100.00
Refuse, As
Sulfur
10.24
20.00
65.00
40.00
62.00
50.00
10.00
1.20
2.00
0.00
2.10
10.00
20.00
3.20
4.00
72.00
0.00
3.00
2.00
10.00
Received Basis
0.16
Non Cora. 24. 93
Ratio
Btu/Lb
C: (H) 39.4
4,917
Btu, dry 6,203
Btu, M
and AF 9,048
Weight
percent
Volatile
Matter
75.94
67.89
26.74
68.46
83.98
64.50
84.34
54.00
20.54
20.26
0.5
0.4
2.68
-------
- 6 -
>-7-
O
' 1
EH
oo
0
0-!
o
u
00
CN M
i EH
rH 00
<.'
C! £3
i — I
X3 Q
(3 HI
EH iJ!
CO
MUNICIPAL
^
>
H
X\
p!
&,,
01
(3 tn
>H td
s
U rj
•H Vl rCi -H 4-1 C
! GO (3 C -P (3 <3
, G • 5 • C (3 O • C
i ,c tn ,c in -H 5 un -o ^
4J 31 4-1 C1 ^Dtn^Gl ,O
I -H > -r( > S 4J > K^-
>HI-i U
* (< *
I I 1 1 1 1 1 i rH ^0 CM
V
rHCOV£)i • •/- ••;
| 'r-i ^ • E-l •'; ^
M >i 00 O/ •»-'
• O -V. r-t • >
CL,rH '-, ;: n •
i — i O *• --"' f •* '•* ~
--rH fi O '- ~'. •
CN r~ in > -r-i £H • £H
in ->s< r-i o'. i ^ X
1 -iH ', O ^ 'J ~-
• • rH 3 ii >i O •
diCIjrHO - -^ EH '•J
>• - O >i -P U -
co c?i rH m 4-) r: X-
H j2 o -i-i n n ~- r;
0 fa O C 0 — ' -i-l
O rH ^ • -rH C: ~^J 'i-
rH ^ > rc o c i;
X5 (3 -n r; C ^ - C
rd£H^>C 3-rJQC-;
EH tn D O U I'D 0 X
c\ U 'j
rH -r-i
co o r*- o r^ o <
m vu o vs 'o
G^ C^ C"i G^i C', 00 C.O
CO 1-1 rH rHrHr-lkr/O
m i cr\ G-.
rH L"l -H
C7~, C d r3 C rj ^ .'i
rH<^4J ^j_^;^_i_)-^
4-1 13 rC 13 rG 13 rw
K (8 CO -H
EH C -H rH 13 U "w rC '"3 r^
Q NCU-C ^^^X r,
X -H -n to t/; _-, L-. 0', c,":
K >H i—i «••!-; -i-i -i-;1 -n -H -i-i
O ir-i i-H rH rH 1-1 rH
K H130OX!X'-3^^
D - 3POD3333
O rJ -4-1^,0^0,^0,^^
U ,-C -P
cue - -
(u -H O 0" in LO u"; (,1 i"i
£ x; 5 c: in tn o", in a;
UUOO OOCGO
tJ- f-i M 'ri !H Vi X
^^ CO-POO COO
< < (3 !-i -H i-, SH « ^ !-i
S^SS si, tn OiSiiCXOiO.
CH &j r,
< -r-i > -H -H -H -rH -1-1
t,Q -rH
- -rd^C^Ai^X.-^
rH rH £: j_, 3 !_| ^| ^ M ;.]
(3 (d 0 OOOOG
0 O • :-,
Wtnri2>H5S S SSi
•H -H o & rd fd ."o rd rj
G Q U !-i S ri '-1 Ui 'r- '^
tr» o cri w~i o c^- 'C"j
OOrHQSOO OOD
tn tn r3 5n SH i-i JH SH ^
33Mpj &,P^&,^PM
4H IH GJ -
(1) • 4J -P -P 4-J 4->
rH rH t," w tn a i: w .0
rd td 4-J .-3 (3 ,-d rj c; .1
S^O,OE: -££:£:^:2;
-H -H I-D SH
ou i3CJrcrcrDi;i-a
•H •(—! O -r". 01 -rH -,-n *<-, -H -r-.
3300^00 ooo
S 2
-------
«"7 »-
Table 1-3
PACKAGING MATERIALS CONSUMPTION
MATERIAL
Paper & Paperboard
Glass
Metals
Wood
Plastics
(Ref. (3) )
MILLIONS
1966
25.2
8.2
7.1
4.1
1.0
OF TONS
1976
36.9
11.9
8.4
4.4
2.5
45.6 64.1
-------
accurate study of waste composition is necessary before a municipality
can make a reliable decision on what method of disposal would be most
economic for its community.
1.3 Utilization of Municipal Refuse
Since paper is the main component of waste, it is desirable to
devise recycling processes using paper as the raw material. Porteous
suggested a process which would convert the cellulose portion of paper,
approximately 75%, to glucose by means of sulfuric acid hydrolysis.
The basis for the hydrolysis is that cellulose is a polymer composed
of individual glucose units. The glucose can be used as a raw material
for fermentation to ethyl alcohol. In this manner, the bulk of the
refuse is converted to a utilizable raw material, which can be sold to
cover the disposal cost of the residual waste. Since the residual
waste is inert, it can be disposed of at a lower cost than can the raw
refuse. Porteous' preliminary study showed the process has an economic
advantage over all other existing means of disposal, thus warranting
further study.
1.4 Summary of Alternatives
Table 1-4 summarizes the available disposal methods and their re-
lated disposal cost. Incineration and sanitary landfill are the only
disposal alternatives open to most cities. Since land is becoming more
scarce, the cost of sanitary landfill will continue to increase for
many municipalities. Although heat recovery systems and other techno-
logical improvements in incineration will help cut cost, there will
always be some associated disposal cost. With this as a basis, it is
-------
- 9 -
Method
Open Burning
Sanitary Landfill
Incineration
Composting
Table 1-4
ALTERNATIVE DISPOSAL METHODS
Comments
No longer legal
Requires large land areas
Must contain antiair
pollution devices
Little available market.
Depends on dumping fee
Cost $/Ton
1-12
3-7
2-7
-------
- 10 -
seen that refuse will have to help pay for its own disposal cost. Using
refuse as a raw material for the production of sugar by acid hydrolysis
is one such attempt at accomplishing this goal.
-------
2. THE HYDROLYSIS OF CELLULOSE
The concept of hydrolyzing cellulose containing materials
to sugar is not a new one. Meiler and Scholler ' in 1923
studied the acid hydrolysis of cellulose dextrine and the first
plant to convert wood to sugar with subsequent fermentation was
built in Germany just prior to World War II. This plant was
built to offset the ethanol shortage which existed during that
period. The principal research conducted in the U.S. was by
Saeman' ' who was contracted by the War Production Board during
World War II to develop a practical process for producing
ethanol from wood waste. Saeman's work established the kinetic
model which describes the hydrolysis of cellulose in dilute
acid solutions.
2.1 The Chemistry of Cellulose
Cellulose is a fibrous tissue found in the cell walls of
plants and trees. It is a polysaccharide composed of long
chains of glucose units, a six carbon carbohydrate, connected
at hydroxyl groups, glycosidic bonds.
OH ^
-11-
-------
-12-
The other polysaccharides which are found in the colls of
plants are called hemicelluloses. They are unrelated to cellu-
lose and consist of pentose units, a five-carbon carbohydrate.
Cellulose is the main component of paper and the amount of
cellulose in the paper varies with the grade of paper produced.
The pulp used for naking paper is prepared from wood by choiaicul
or mechanical treatment. This treatment breaks down the wood
fibers, eliminating the lignin and most of the easily destroyed
hemicellulose. The main methods of pulping, with their result-
ing cellulose contents are: Sulfite 86%, Kraft 80%, and me-
chanical 50%. The exact amount of cellulose in any given type
of paper depends on which pulp or pulps are used and the amount
of filler which is added to the paper. The amount of cellulose
in paper ranges from approximately 95% in rag paper to 50% in
newspaper.
Cellulose and hemicellulose undergo the following hydrol-
ysis reactions with acid acting as a catalysis:
ACID
Cellulose >Hexose > decomposition products acids
ACID
Hemicellulose > Pentose > decomposition products furfural
The hydrolysis reactions can be either homogeneous or hetero-
geneous, depending upon the acid concentration of the solution.
Hemicellulose is an easily hydrolyzable material which will
yield mainly the pentose, xylose, which upon further reaction
produces furfural. The more stable of the two, cellulose,
•upon hydrolysis yields glucose, which subsequently decomposes
to hydroxymethylfurfural and finally levulinic and formic acid.
-------
-13-
In its natural state, cellulose is not only linked by
glycosidic bonds but by hydrogen bonding in a crystalline
lattice. Cellulose is soluble in concentrated solutions of
72?. sulfuric, 35% phosphoric, and 45% hydrochloric acid. In
such solutions the hydrolysis reaction is first order homo-
geneouo and due to the low reaction temperatures proceeds with
very little decomposition of the sugars. A process based upon
high acid concentration would give high yields of sugar, but
the high cost associated with using large amounts of acid
places the process in an undesirable economic state. Dilute
acids, less than 2%, can be used if the reaction temperature is
increased. This reaction is heterogeneous, since the crystal-
line state of the cellulose remains intact. The exact kinetic
mechanism of the reaction is not known but the results have
Kl K2
been modeled with a standard A > B ) C irreversible
reaction rate model. Sulfuric and hydrochloric acid with their
high ionization coefficients are the best catalysis for the
reaction since they give the highest ratio of rate of formation
to rate of decomposition of sugar. Sulfuric acid is usually
preferred for use since it is easier to handle and less costly
than hydrochloric acid.
2.2 Dilute Acid Hydrolysis of Wood Cellulose
Porteous' original process^) for waste disposal by cellu-
lose hydrolysis was based on a kinetic study performed by
SaemanC?) on wood. Saeman attempted to show that the hydrolysis
of wood cellulose in dilute sulfuric acid could be described by
K! K2i KI
an A —> B —* C consecutive reaction with cellulose »
-------
-14-
K2
reducing sugars ) sugar decomposition products. Saeman
performed his experiments in sealed glass bombs heated by
direct steam in a rotating digester.
The rate constant for the decomposition of the sugars, K2,
was determined by reacting sugar solutions in the glass bombs
at temperatures from 170° to 190°C with acid concentrations
from 0.4 to 1.6%. The decomposition of both hexose and pentose
sugars was studied and it was found that glucose, the main
component of cellulose, was the most stable. Plots of the
logarithm of the residual sugar concentrations versus reaction
time gave straight lines. Since the integrated equation of a
first order reaction is, B = B e ^ , the logarithm of the
concentration, log B = log BQ - K0T , will give a straight
line with the slope equal to K2. Saeman's results, therefore,
showed that the reaction was first order. The results of this
experimentation are summarized in the following empirical
equations taken from Saeman's work.
P.ate Constants for Sugar Decomposition
. 32.700
K2 = 1.86 x 1014 C1'02 e RT d>
-32870
K2 = 2.39 x 1014 C1'02 e KT~ (2)
with C = concentration of acid
R = 1.93 cal/g mole deg. K
T = Temp. deg. Kelvin
Equation (1) was derived by measuring the residual reducing
sugar content of the solution and equation (2), by measuring the
residual fermentable sugar content. Since sugars act as a re-
ducing agent, a simple analysis for sugar is a test of the
-------
-15-
rcducing power of the solution. Although valid when degraded
sugars are not present, it gives false values when degraded
sugars are present. This is explained by the additional re-
ducing power of the first oxidation state of glucose. The only
reliable measure is based on equation (2) which is calculated
with a test specific for glucose. This equation was derived
using a pure homogeneous glucose solution and is only a means
of approximating the reaction rate for the actual decomposition
of sugars in the presence of wood cellulose.
The rate constant for the formation of sugar from cellulose,
Kl7 was determined by reacting ground samples of Douglas fir
wood with a dilute acid solution in the heated glass bombs.
After a measured period of time the vessels were quenched and
analyzed for the unreacted wood cellulose content. Since wood
is not entirely cellulose and no direct measure of cellulose
content is possible, an indirect method of quantitative sacchari-
fication was employed to measure the cellulose content. The
unreacted wood cellulose, referred to as the residual cellulose,
was washed free of the dilute acid-sugar solution and reacted
under conditions which gave quantitative yields of sugar without
sugar decomposition. Consequently the quantitative saccharifi-
cation procedure described in Appendix I uniquely measured the
residual cellulose of the sample. Obviously when the quanti-
tative saccharification is applied to an unreacted sample of
wood, it determines the original cellulose content of the wood,
which is referred to as the potential sugar yield of the sample.
When the quantitative saccharification is applied to a reacted
sample of wood (residual), it gives the remaining cellulose con-
tent which is referred to as the residual potential sugar yield.
-------
-------
-17-
The high activation energy of rate constant
that although it was derived for a heterogeneous sc
reaction, the reaction is not mass transfer controlled. The
higher the activation energy, the more temperature dependent
the reaction is. Since diffusion is not so temperature de-
pendent, it is concluded that the reaction is not diffusion
controlled.
Kl K2
With the A—> B —> C irreversible reaction, as the
reaction proceeds, the cellulose concentration, A, decreases,
the decomposed products concentration, C, increases, and the
glucose yield, B, goes through a maximum. This yield of in-
terest is that of glucose, the maximum value of which is given
by
max ~ K2
This equation is derived from the integrated equations for
the reaction in section 3.1. Since the empirical equations
(2) and (3) for Kn and K0 show that K-, will increase faster
& •!•
than K2 with an increase in acid concentration or temperature,
the maximum yield of glucose, equation (4), should also increase
-with these conditions. This was shown by Saeman in his origi-
nal work^ ' to be true over the range of 170 to 190°C. On this
basis Porteous extrapolated Saeman's data to 230°C with a 0.4%
acid concentration, and predicted a yield of 55%. Figure 2-1
taken from a more recent publication of Saeman's work'^) shows
that a yield of approximately 47% was obtained under these con-
ditions. There was no explanation of the discrepancy between
the predicted and experimentally obtained results.
K2 \
-------
-18-
Figure 2-1
SUGAR YIELD versus RESIDUAL POTENTIAL SUGAR
'J
-f
\~
LJ
Z
olil
loo
1
BO
1C
U&AK_(%OP INITIAL)
-------
-19-
Sacnan's empirically derived results only hold for the
crystalline portion of cellulose and most cellulosic material
contains both amorphous and crystalline cellulose. £ leman ;. i -
found that cotton cellulose decomposed at a rate different, frcrn
that of wood cellulose. The rate of hydrolysis of paper cellu-
lose, especially at higher temperatures, therefore, cannot be
strictly inferred from Saeman's work since its crystalline
structure may be modified both by the paper making process and
any pretreatment performed in the hydrolysis process. The
kinetics and yields associated with paper cellulose hydrolysis
had to be found experimentally before a hydrolysis process for
refuse paper could be designed.
-------
3. EXPERIMENTAL APPARATUS AND PROCEDURE
An experimental program was established to determine the
ultimate potential of refuse hydrolysis as a solid waste dis-
posal process. Since Porteous based his preliminary economic
study on an extrapolation of Saeman's kinetics for wood cellu-
lose, it was necessary to determine whether such an assumption
was valid. If it were found that paper hydrolyzed at a dif-
ferent rate than wood cellulose, then a model for paper hydrol-
ysis would have to be determined. Although it was not necessary
to determine exactly what reaction was occurring, a model for
predicting the yield under various conditions of hydrolysis
had to be found. Such a model was necessary for the design of
the hydrolysis plant and the determination of the most econom-
ical operating conditions for the plant. Once this information
was obtained, the total economics of the system could be de-
termined.
The laboratory apparatus had to contend with the high
temperature (>220°C), high pressure (>400 psi), and the cor-
rosive atmosphere anticipated for the process. Moreover, the
fact that the yield of glucose reached a maximum in a short
time made the design of a reliable laboratory procedure very
difficult.
A quantitative saccharification procedure which would
completely hydrolyze cellulose to glucose without decomposition
products is found in Reference 9 and summarized in Appendix I.
This experimental analysis is based on initial high acid con-
centration, 72% H2S04, to dissolve the cellulose at room
-20-
-------
-21-
tcr.Vjeraturo, follov/ecl by treatraent with diluted 4% H^SO, at
an elevated, temperature. This procedure determines the poten-
tial glucose in any cellulosic material such as refuse, and is
essential in any kinetic study since the original amount of
cellulose must be known before the kinetic parameters can be
iatvjirpixtccl. The glucose analyses v/crc performed with an
C-toluidine colorirr.etric test described in Reference 9 and
modified by Dr. Sanka, Mary Hitchcock Memorial Hospital, Hanover,
N.H., for work in blood and urine sugar analysis. This test is
outlined in Appendix II. It is a micro test based on the spe-
cific reaction of aldohexoses with 0-toluidine in accordance
with Beers Law.
3 . 1 Kinetic Model
SaemanC?) showed that the hydrolysis of the resistant
wood cellulose is described by an A — i — » B — ^ — > C reaction
governed by the following equations. (A = cellulose, B =
glucose, C = decomposed sugar).
3CS
~ K2CB + K1CA
Integration of (5) and (6) gives
P
JL. = e~Kl6 (7;
r~* T/1 \ ^
C^ = (K2 ~ KX) (8)
CB(MAX) , Kl ; A / K2 I (9)
-------
-22-
inKo - inKi
6* - ± £ (10)
K2 - KI
where CA = concentration of A
Cj. = initial concentration of A
A0
8 = time
a* = time to max yield of B
Kj_ = reaction rate constant for A to B reaction
K = reaction rate constant for B to C reaction
The classical experimental approach for determining rate
constants was used by Saeman. He assumed isothermal conditions
for the reaction and that the heat up and cool down time for
his glass vessels would be negligible when compared with the
raction time for temperatures below 190°C. The amount of
original and subsequent cellulose was determined by a quanti-
tative saccharification. Thus, by reacting the cellulose for
various lengths of time, the log of the cellulose concentra-
tion versus time could be plotted and the first order rate
constant KJL determined. The same procedure was used to de-
termine the first order reaction rate constant K^ with the
cellulose replaced with a known concentration of glucose.
Saeman's experimental equipment limited his reaction tem-
perature to a maximum value of 190°C, but extrapolation of
Saeman's results indicated that a higher yield could be ex-
pected at higher temperatures.
3.2 Isothermal Kinetic Study
The first experimental procedure was designed to determine
if Saeman's data could be extrapolated to temperatures above
190°. It was thought that a simple approach to this problem
would be to assume that Saeman's empirically derived equation (2)
-------
-23-
for the decomposition rate of glucose would hold for higher
tempera.cures. Therefore, knowing the original concentration
of cellulose from a quantitative saccharification ¥-2 from
equation (2) and the yield of glucose versus time from experi-
mentation, it would be possible to determine K^ by a trial
and errGf solution involving" equation1 (8) . If tYiQ de'tSIrnineu
values of K, were the same as those predicted by Saeman's
empirical equation (3), it would be concluded that the hydrol-
ysis of paper cellulose is the same as the hydrolysis of wood
cellulose and Saeman's empirical equations would be valid for
such work.
Initial kinetic studies were performed in 1" x 4" stain-
less steel (316) nipples capped and sealed with Teflon tape.
The reaction temperature was obtained by placing the reaction
vessels into a high temperature oil bath. A temperature-time
history was recorded by sealing a chromel-alumel thermocouple
in one of the vessels. A number of preliminary tests gave in-
/
consistent results with very low yields. Since it was thought
that the acid was reacting with the vessels and causing lower
than expected yields, a 25 x 100 mm glass culture tube was
placed in the stainless steel vessels. This alleviated the
corrosion problem and increased the measured yield by 300%.
A new temperature-time curve was recorded and all remaining
tests were made using the glass liner. 0.2 gr brown paper bag
samples were prepared by cutting up the bag into approximately
1/10" squares. The samples were placed in glass culture tubes
and 5 ml of 0.5 percent sulfuric acid solution was added. The
glass tubes were then sealed in the stainless steel vessels
-------
-24-
and placed in the constant temperature bath for a measured
period of time. At the end of this time the vessels were
quenched in ice water, and analyzed for glucose content. The
heat of reaction for cellulose decomposition is approximately
10 K cal/g-mole. Therefore the adiabatic temperature rise was
calculated to be 2°C under the experimental conditions used
in this study. However, the conditions are far from adiabatic;
hence the error in neglecting heat effects of the reaction was
certainly less than 2°C. An experimental temperature-time
trace indicated isothermal conditions after an initial heat-up
period, the reactor temperature being a constant 4°C below the
indicated bath temperature.
A corrected isothermal reaction time was determined as
follows
where:
ti = reaction time used in kinetic calculations
t^ = time that sample resides in bath
t3 = time to heat sample from room temperature
to 170°C in the bath, as determined from
a heat-up curve
t4 = time to heat from 170°C to final bath
temperature
The main cause of error in kinetic calculations Was the
time to heat the sample from 170°C where the reaction rate
first becomes significant, to the final temperature. This
heat-up time was found to be between 2 and 3 minutes ; thus
for temperatures above 210°C and reaction times approaching
3 minutes, the error due to heat-up was very large.
-------
-25-
3 . 3 Acici Inj_cction Bomb
The isothermal kinetic analysis based on the results of
the experimental procedure outlined in Section 3.2 w ,s in error
due to the time necessary for the bombs to reach isothermal
conditions. It was believed that this error could be elimi-
nated by building an acid injection bomb system. Such a systom
was constructed using a 2" x 4" SS-316 pipe with a glass liner
as the reactor. The reactor, charged with 150 ml of water and
0.5 grams of paper, was heated in a high temperature bath until
it reached the desired temperature. The hydrolysis was ini-
tiated by injecting 1 ml of acid through a three way ball
valve by nitrogen pressure. After a measured period of time
/•
a sample was drawn off and flashed to room temperature. It
was found that the reactor required more than two hours to
reach isothermal conditions and that during this time the
paper had undergone thermal degradation.
The initial paper-water charge volume was kept large
compared to the injected acid volume to prevent the lower
temperature acid from cooling down the reactor. Using less
than 1 ml of acid solution would result in large error in the
final desired acid concentration, due to transfer losses in
the valve and tubing. Although a smaller reactor volume would
have a shorter heat up time, the required high volume ratio
and lower limit on the injected acid volume prevented the use
of such a system. An attempt to circumvent the thermal de-
gradation was made by injecting a fine paper slurry into the
reactor. The reactor was charged with 150 ml of 0.5% acid
solution and brought to the desired temperature. 5 ml of
-------
-26-
watcr with 0.1 gram of powdered paper was then injected into
the reactor. it was found that the paper failed to pass freely
through the valves and tubing into the reactor. The modifi-
cations necessary to allow•such operation precluded its use
for such experimentation.
3<4 Nonisothermal Kinetic Study
Due to the difficulties of obtaining an isothermal system,
a new approach to the problem was attempted. It was believed
that a non-isothermal kinetic analysis could be performed by
means of a modified curve fitting routine. It was assumed that
the proposed A —=-> B —=-) C kinetic model was valid and that
the rate constants followed Arrhenius' equation, K = Ce~A^' .
There are then at a given acid concentration four parameters,
the two pre-exponential constants (C's) and two activation
energies (AH's), which had to be calculated to determine the
overall kinetic model. The following method was used to de-
termine these parameters.
A sample of paper or refuse which had been run through a
Wiley ball mill to produce 2 mm particles was placed in a glass
culture tube with a specified amount of acid solution. This
vessel was placed in a 1" x 4" stainless steel nipple and
sealed. The cap of the nipple was taped with a 1/4" N.P.T.
such that a Swagelock thermocouple well connector could be
screwed into the cap. The thermocouple well consisted of a
1/4" stainless steel tube coated with Teflon tape. A chromel-
alumel thermocouple was placed in the well, which had been
previously filled with oil, to measure the liquid temperature
within the vessel. This vessel was placed in a constant tem-
perature bath which had been set at 260°C. A constant record
-------
-27-
of the temperature-time history was kept by use of a thermo-
couple! chart recorder. Samples were removed from the bath
at various intervals of temperature, quenched in ice water, and
analyzed for sugar content. Six to eleven samples were used
for each acid concentration run, and at the start of each acid
run a sample quenched when the temperature reached 180°C was
analyzed for both sugar and residual potential sugar (cellulose)
content. The data obtained from such a procedure consisted of
1) quantitative potential sugar (cellulose content) at 180°C,
2) sugar yield at intervals of temperature, and 3) temperature-
time history. The differential equations which describe such
a system are:
3 A
33
(11)
- K2B(t) (12)
0 t
where A = concentration of cellulose
B = concentration of sugar
Kl = C1e"AHl/RT = rate constant 1
-AH9/RT
K2 = C2e * = rate constant 2
R = 1.98 cal/g deg. K
T - deg. Kelvin
t = time
With calculated initial conditions on A and B at 180°C, a
temperature-time function derived from a 4th order polynomial
curve fit for the data, and values for C]_, AH]_, C2, AH2, equa-
tions (11) and (12) could be numerically integrated for the
sugar yield as a function of time. It was necessary to search
for the values of C]_, AH]_, C2/ and AH2 which resulted in yields
-------
-28-
of sugar which best agreed with those obtained experimentally.
Once such parameters were determined from the non-isothermal
analysis, the kinetic model could be used to predict the yield
for an isothermal reaction.
A Hooke-Jeeves pattern search^ ' and a Runge-Kutta numer-
ical analysis were used to determine the rato constant para-
meters which minimized the sum of the square errors between
the predicted and the experimental sugar yields. Insurance of
convergence was determined by generating hypothetical experi-
mental yields with four given parameters, and using these
yields to repredict the parameters with the previously des-
cribed computer program. The final sum square error from
this analysis was 1 x 10~6 and convergence was obtained in
120 seconds on a G.E. 630 time-sharing system.
-------
4. 2XPERIMZXTAL RESULTS AND ANALYSIS
The apparatus and procedures described .in Chapter 3 were
used to determine the yields and kinetics of paper hydiolysis.
It was found that an accurate kinetic analysis based on iso-
theraal conditions would not be possible. Reaction half lives
of less than one minute were expected at temperatures >220°C,
and under such fast reaction conditions it was not possible to
build an apparatus with a heat-up time small enough to be
ignored. A non-isothermal kinetic analysis was therefore used
in determining the kinetics of the reaction.
4.1 Isothermal Analysis Results
A preliminary study of acid hydrolysis was performed with
temperature conditions which were to be isothermal. Runs were
made at bath temperatures of 200°C, 210°C, and 220°C. The
yields of sugar on a weight sugar/weight paper basis are shown
in Figure 4-1. A kinetic study was performed on this data as
indicated in Section 3.1, but it was hampered by an inaccurate
quantitative saccharification procedure and the actual non-
isothermal state of the system.
An accurate quantitative saccharification is essential
since a knowledge of the amount of cellulose in the paper is
needed to correct yields to a weight sugar per weight potential
sugar (cellulose) basis. This change of base is necessary since
the kinetic model is for the hydrolysis of cellulose, and paper
is not 100% cellulose. The main difficulties in the quanti-
ta-c_ve saccharification were caused by lack of a suitable low
-29-
-------
-30-
Figure 4-1
SUGAR YIELD versus REACTION TIME
10
\? 2.20°
\
\
10
20
TIME.
-------
-31-
tenpcrature bath for the first step of the analysis and by
using too large a particle size in the procedure. The time
necessary to reach isothermal conditions also caused error
which could not be neglected in a kinetic analysis.
In spite of these difficulties the data obtained showed
that paper could be hydrolyzed to sugars and that the yield, of
sugar passed through a maximum as would be expected if the
proposed reaction sequence held. Further evidence of such a
series reaction is that the time to maximum yield decreases as
the bath temperature increases. The yields, although lower
than expected, did indicate that further experimentation was
justified. Moreover, at the end of this series of experimen-
tation, a better understanding of the hydrolysis reaction and
related chemical analysis was obtained, so that more effort
could be given to the determination of the reaction model
itself.
4.2 Non-Isothermal Analysis Results
The cellulosic materials used for the acid hydrolysis ex-
periments were tested quantitatively for their potential sugar
yields which is an indication of their cellulose content.
Table 4-1 contains the results of this quantitative sacchari-
fication. These yields represent the total experimental poten-
tial aldohexose content of the sample. The fermentable aldo-
hexose sugars are glucose, galactose, and mannose. Table 4-2,
taken from Reference (8), shows the expected percent composi-
tion of these sugars in Kraft paper. The main contribution to
the quantitative sugar yield will come from glucose. Glucose
can be formed from both amorphous and crystalline cellulose.
Mannose and gelactose are mainly derived from non-crystalline
-------
-32-
Table 4-1
QUANTITATIVE SACCHARIFICATIONS
SAMPLE
YIELD
Weight Sugar/Weight Paper %
Ground Kraft Paper
Ground Kraft Paper
Slurried Kraft Paper
Slurried Kraft Paper
Ground Refuse #1
Ground Refuse #1
Ground Refuse #2
Ground Refuse #2
86
84
84.8
80.5
38.0
38.4
52.6
52.6
Table 4-2
CARBOHYDRATE COMPOSITION OF KRAFT PAPER PULP
(8)
Component
Glucose Galactose
Mannose/ Arabinose
Xylose
Uronic Anahydride
Percent
82.4
7.0
9.2
1.4
-------
-33-
polysaccharides . Although the main portion of cellulose in
paper is believed to be in the crystalline form, the exact
amounts of amorphous and crystalline forms are not knc. fin.
Samples of milled Kraft paper, approximate particle size
2mm, were hydrolyzed with 20 ml of 0.2%, 0.5%, and 1% by weight
sulfuric acid solutions. The yields and temperature time his-
tories are shown in Appendix II. Selected samples, quenched
when the temperature reached 180°C, were filtered and reana-
lyzed for their potential sugar yield. In this manner a hy-
drolysis sugar yield and residual potential sugar yield were
known at 180°C. These yields were then used as the initial
conditions for the 4-dimensional search outlined in Section 3.3.
Arrhenius ' equation was used to relate the rate constant to the
reaction temperature. The computer program found in Appendix IV
was used to perform the search which determined the pre-exponential
factors and activation energies that would best fit the data for
the 0.5% acid runs. It was then assumed that these activation
energies could be used for the 0.2% and 1% acid experiments,
thus only requiring a search on the two pre-exponential para-
meters for this data. The Arrhenius equations for the calculated
values are:
0.2% KI = 1.661 x 1019 * e"AHl/RT
K2 = 2.21 x 1014 * e-AH2/RT
IQ -AH-i/RT
0.5% K-L = 7.59 x 10iy * e I
K2 = 2.67 x 1014 * e-AH2/RT
1% KI = 2.95 x 1020 * e'AHl/RT
K2 = 5.57 x 1014
-------
-34-
with
hRI = 45127 cals/g mole
AH. = 32800 cals/g mole
In an acid catalyzed reaction the overall reaction rate is
usually proportional to the acid concentration as well as the
reactant concentration. Although this would normally result in
second order kinetics, the fact the acid concentration is con-
stant allows the rate constant to be expressed as/
K..^ = KQ Cm C = acid concentration Wt. %
where K. is the overall constant K,, K2 shown above.
Since the reaction activation energies are the same for low
acid concentrations, the acid effect will be contained in the
pre-exponential factors of Arrhenius equations for K., and K2/
P^ = PQCm where P. is the pre-exponential of K. in
K = p e~AHl'RT. A plot of the log of the pre-exponential
J» Jx
factors versus the log of the acid concentration should give
straight lines within the slope of the line. Figure 4-2 shows
that this relationship held for K, but not K-. This result can
be explained, since the actual hydrolysis reaction, K,, is known
to be due to cleavage of the cellulose bonds by the hydronium ion
whereas the decomposition of glucose, K_, is a series of sugar
oxidations which are not simple hydrolysis bond cleavage.
These results also showed that at zero acid concentration
and high temperatures, 230°C, thermal hydrolysis would occur, but
low yields would result since the ration K-^/K^ ^; 0.25 is very
small. This explains why thermal degradation or hydrolysis
occurred in the acid injection system described in Section 3.'3.
The accuracy of these fitted values was determined by
performing replicate experiments at approximately 230°C. The
-4
sum square error of these replicates was 5.362 x 10 with
2 degrees of freedom. The largest determined curve fit error
-------
5 '
-35-
Figure 4-2
LOG (PRE-EXPONENTIAL FACTORS) versus LOG (ACID CONC.)
©
Q/
J-1.
O.z.
0.3
0-4 .>
/WT. %)
0-5 0-fe 0.7
-------
-36-
M -3
was 3.3 x 10 with 4 degrees of freedom. The statistical "F"
ratio is therefore 6.15 which is well within the 19.25 "F"
ratio at the 95% confidence interval. This test shows that
there is not a significant difference between the fit and the
experimental error. The actual experimental data points and
the predicted curves are shown in Figure 4=3, Another qu&§-
tion concerning the accuracy of the experimental data arises
from the ice water quench time. It was found that approxi-
mately 20 seconds are required to lower the reactor tempera-
ture- from 230°C to 170°C, This time would appear to con-
tribute a great deal of error at high temperatures and acid
concentration. Although this assumption would be true if the
reaction continued at the 230°C rate, it is not true when it
is considered that the temperature is continually dropping, as
is the reaction rate. Using the reaction rates calculated for
a 1% acid solution, a numerical integration was performed for
a linear drop in temperature from 230 to 170°C with an initial
sugar concentration of 11%, This calculation revealed that
approximately a 11 less ,.:i yield could be expected due co the
quench. Since the condition used for this test is the most
severe encountered, the error due to quench time was ignored:
The reliability of the predicted rate constant parameters
can further be demonstrated by their ability to predict yields
with temperature time curves other than those used for deter-
mining them. Initial runs with the high temperature bath pro-
duced widely different temperature time curves; this was
found to be due to convection currents established by the
-------
-37-
Figure 4-3
SUGAR YIELD versus NONISOTHERMAL TEMPERATURE (TIME)
.N 50-
130
\QO
ioo
ZlO
TEMP&
A
51
Tir-iE
-------
-38-
bath circulation pump. This problem was remedied by placing
a metal beaker filled with oil in the bath, thus acting as a
buffer between the actual bath convection currents and the test
reactor. Three of these variable temperature time curves were
used in conjunction with the calculated rate constants to pre-
dict the associated sugar yield curve. The 'following yields
were obtained in this manner.
0.5% Acid
Curve
I
II
III
Quench
Time
8.6 min.
5.6 min.
6 . 1 min.
Experimental
Yield
0.30
0.24
0.325
Predicted
Yield
0.30
0.21
0.315
It is seen from these results that the agreement between pre-
dicted and experimental yields is very good.
Since the proposed hydrolysis plant will operate with a
paper slurry, it was necessary to determine whether a signifi-
cant change in rate constants would occur if fibrous material
were used in place of ground particles. A sample of Kraft paper
was pulped with water in a standard Waring blender. This sample
was then dried and 0.5 gram samples were prepared from the fi-
brous sheet. These samples were mixed with 20 ml of 0.5% acid
producing a fibrous slurry which was then hydrolyzed. The
agreement between predicted and experimental results, Figure 4-4,
indicates that the determined relationship holds for fibrous
samples as well as ground samples. It is therefore possible
to predict isothermal yields for plant operations using' a pulped
paper slurry.
-------
-39-
Ficjure 4-4
UGAR YIELD (KRAFT FIBERS) versus NONISOTHERMAL TEMPERATURE
-------
-40-
Referring again to Figure 4-3, it is noticed that approxi-
mately 6 to 11% sugar yields were obtained at lower tempera-
tures between 170°C and 180°C. Saeman's rate constants for
crystalline cellulose indicates that yields of this magnitude,
at low temperatures and short heat up, should not be obtained.
It is believed that the discrepancy aan be explained by the
different regions of cellulose hydrolyzed. Saeman's original
kinetic data applied to what was called the resistant portion
of cellulose. He discovered that when working with untreated
ground wood, all plots of residual potential sugar versus time,
extrapolated back to zero time, gave a value for the potential
sugar yield below that obtained experimentally by quantitative
saccharification. It was believed that the difference between
the original 66.6% potential sugar and the extrapolated 44% at
zero time was due to easily hydrolyz,able amorphous cellulose.
His calculated rate constants, therefore, held only for the
resistant portion of cellulose. It is believed, from the
hydrolysis of paper results, that it also contains a portion
of easily hydrolyzable cellulose.
This kinetic study shows that the easily hydrolyzable
portion of cellulose accounts for the sugar yields at lew tem-
perature values and that the kinetic parameters estimated here
are for the residual cellulose. It is not valid to assume that
these kinetics only apply to residual crystalline cellulose.
Browning explains that during hydrolysis, cleavage in the
amorphous regions can cause rearrangement and crystallization
of the amorphous cellulose. The calculated rate constants foi
paper hydrolysis may therefore be a combination of rate contra:.•
-------
-41-
for the amorphous-crystalline cellulose. The important point
is that the model was shown to predict the correct yield under
various hydrolysis conditions. It is of no great importance
for the engineering plant design to determine exactly what
reaction is occurring; it is important to be able to predict
the yield under various conditions of hydrolysis.
The predicted isothermal yields of sugar from paper cellu-
lose are shown in Figure 4-5. For these calculations it was
assumed that 10% of the original potential sugar yield would
hydrolyze instantaneously. This value is based on the experi-
mentally determined sugar yields of 6 to 11% from the easily
hydrolyzed portion at low temperatures. This curve gives the
residence times necessary to obtain maximum yield at various
temperatures and acid concentration. If the original potential
sugar content of the paper refuse is known, this curve will
give the predicted isothermal yields of sugar.
The original prediction by Saeman for wood that an in-
crease in acid concentration and temperature causes an increase
in overall glucose yield is shown to hold true. 10% yield in-
crease is shown to occur by doubling the acid concentration at
230°C and a 10 degree rise in temperature results in approxi-
mately a 5% yield increase.
4. 3 Experimental Analysis Conclusions
It was shown by the experimental analysis that paper
cellulose does hydrolyze to sugar and that the hydrolysis
reaction could be modeled by the A — L>B — 2_^.c irreversibl
reaction model used by Saeman. The reaction rate constants
-------
-42-
Figure 4-5
PREDICTED ISOTHERMAL SUGARS versus TIME
-<— ("A")
-------
-43-
were found to be different from those calculated by Saeman,
but as predicted, they both increased as temperature and acid
concentration increased, with Kj_ increasing faster -_han K2.
The increase of the ratio K-j_/K2 was not as pronounced as
expected, and the yield of sugar from paper cellulose at
higher temperatures was therefore lower than that predicted
by Saeman. Porteous originally predicted a yield of 55% with
Saeman1s data at 230°C whereas the calculated kinetic model
predicted a yield of 41%. This difference is reaction rates
was thought to be due to the amorphous cellulose content of
paper cellulose. It was also shown that the predicted
reaction rates held when pulped paper fibers were hydrolyzed.
It was therefore concluded that the calculated kinetics could
be used in the design of a refuse hydrolysis plant under iso-
thermal reaction conditions.
-------
5. PLANT DESIGN
The information gained from the hydrolysis experimen-
tation was used in the design of a refuse hydrolysis plant.
As much of the non-cellulosic material as possible must be
separated out of the raw refuse, since additional acid solu-
tion would be required for this useless material. In addition
to separation, the refuse must be pretr"eated to produce fine
particle sizes which can be hydrolyzed without mass transfer
slowing down the reaction. Porteous' preliminary design pre-
treatment system consisted of dry pulverization followed by
flotation separation and secondary shredding. Although the
heavier metals would be separated out by such a system, the
lighter plastics, bottles, plastic wrappings, rubber, and
any pieces of cut metal entrained in the paper would be car-
ried over to the secondary shredder. Moreover, secondary
shredding produces particle sizes of approximately one inch
which would result in lower sugar yields due to the lack of
intimate contact of the acid with the paper fibers. A proven
separation system, used in pretreating waste paper for reuse
as paper pulp, was adapted in place of Porteous1 system. Al-
most complete separation of the above materials would be ob-
tained and the resulting paper slurry would consist of fine
fibers, the size controlled by the pulper operation, which
were proven to be easily hydrolyzable.
A continuous flow reactor, as recommended by Porteous,
was incorporated in the design with an additional feature of
a cellulose recycle stream. The experimental results indicated
-44-
-------
-45-
chat approximately 25% of the cellulose would not be hydro-
lyzed and the recycle stream was used to hydrolyze this resid-
ual potential sugar source. Since the experimental ;
-------
-46-
sugar by addition of sulfuric acid and heated water. Unreacted
cellulose is recycled after flash cooling and liquid separation.
Following acid neutralization, the final operation of concen-
trating the liquors is performed in a feed forward multiple-
effect evaporator system. The inputs to the hydrolysis plant
are refuse, process water, sulfuric acid, and lime. The out-
puts from the system are scrap iron, other metals, plastic
inerts, hydrolysis waste, and a concentrated glucose sugar
solution. A general plant flow sheet is shown in Figure 5-1
through 5-3.
5.1 Separation and Pretreatment System
The purpose of the separation and pretreatment system is
to eliminate as much of the unhydrolyzable portion of refuse
as possible and convert the refuse to a pulp slurry for the
hydrolysis reactor. A system which will accomplish this task
is already in existence. The paper industry employs such a
system to pretreat waste paper for re-use as paper pulp. This
system is described in reference (12) and is the basis for the
hydrolysis plant pretreatment process.
A diagram of the separation system is shown in Figure 5-1.
The incoming raw refuse is stored in a silo storage hopper. A
ventilation line connects the storage hopper with the direct
fired heater. A continuous flow of air is drawn from the
storage hopper to the heater, thus eliminating the odor given
off by the raw refuse while in storage. The storage hopper
feeds the refuse to a magnetic pulley conveyor which rejects
the scrap iron portion of the refuse and feeds the refuse to
-------
-47-
z
o
5
I
ai
in
LL
-------
-48-
•0
b
a
UJ
-------
-49-
-------
-50-
th e hydropulper. The hydropulper breaks up the refuse and its
"junker" carries out the larger pieces of unpulped refuse, i.e.
tin cans, plastic bottles, non ferrous metals, and unbroken
glass. A screen trap follows the hydropulper and eliminates
the plastic films and other lighter pieces of trash. The
dirt and finer pieces of glass are removed in a cyclona sepa-
rator. Enough water is added to the hydropulper to produce a
2 to 3% consistency slurry which is the recommended consistency
for such an operation. The pulp is dewatered in a screw press
to approximately 50% by weight water. At this point the refuse
is ready for the hydrolysis process.
It was assumed for calculations that the synthetics, glass,
metal, and inert (rocks, etc.) portion of the refuse will be
separated from the pulp. This means the pulp will consist of
essentially rubbish (paper, leaves, and wood) and garbage. The
pulp quality demands in the paper industry are much more strin-
gent than for the hydrolysis process. It is therefore believed
that this separation system should be adequate for the hydrol-
ysis process.
Equipment Design Information
Storage Hopper:
Standard reinforced concrete. Design capacity based
on 3 days storage at approximately 500 Ib/yd .
Hydropulper System:
Designed and manufactured by Black Clowson, Middleton,
Ohio. Capacities up to 850 tons/day. Horsepower
requirements approximately 20 H.P. day/ton.
-------
-51-
5.2 Hydrolysis Reactor System
Figure 5-2 is a flow diagram for the hydrolysis reactor
system. Freshly pulped refuse is passed from the screw press
into a series of Moyno pumps which bring the pulp to the de-
sired reactor pressure. A stream junction at the inlet to the
continuous flow reactor mixes the fresh pulp with recycled
cellulose and enough acid solution to reach the required reac-
tion temperature, acid concentration, and liquid to solid
ratio. The acid is injected into the heated water stream
prior to the mixing junction. After passing through the flow
reactor, the solid-liquid slurry is flash cooled to quench
the hydrolysis reaction. Vapor from the flash chamber is used
as a heat source in the evaporation system. The remaining
slurry is separated in a pressurized continuous flow centri-
fuge. The liquid continues down stream while the solid stream
is partially bled of inerts and cellulose before entering a
series of Moyno pumps which feed the unreacted cellulose into
the mixing junction as a recycle stream.
The hydrolysis reactor was chosen from the three standard
chemical reactors: batch, continuous stirred tank, and tubular
flow reactor. The operations of charging, discharging, and
cleaning of the batch reactor require an exorbitant amount of
time and material handling which result in high cost for large
scale operation. Although well suited for small amounts of
costly reactants, the batch reactor would be undesirable for
the large amounts of material processed in a hydrolysis plant.
The continuous stirred tank reactor is not advised for use
Kl K2
in an A *B }C series reaction, since the theoretical
-------
-52-
maximum batch yield of B (glucose) is unobtainable. This can
be seen when one realizes that a CSTR is assumed to be well
mixed with outlet reactant concentration equal to the bulk
reactant concentration. An unreacted stream of A is continu-
ously added to the outlet concentration. Therefore the reac-
tion rate is not proportional to the inlet concentration/ but
to the outlet concentration. Reference (13) presents a graph-
ical display of the maximum obtainable yield and required
Kl K2
volume of a CSTR for an A—=•) B —=^ C consecutive reaction
with variable values of K-j_ and K2» It is clearly shown that
a CSTR gives a lower yield of B, and requires a greater volume
to reach its maximum yield than does a plug flow tubular reactor.
Variations and combinations of batch, semi-^batch, series
CST, reactor configurations are conceivable, but the most simple
and reliable reactor for the hydrolysis reaction would be the
tubular flow reactor with a recycle stream. Such a reactor with
plug flow will give the high yield of a batch reactor and the
production capacity of a CSTR.
The design dimensions of the flow reactor will be a func-
tion of required residence time, through-put (fresh and recycle
reactants), degree of axial diffusion (divergence from plug
flow), and the desired flow velocity. The residence time is
the time to maximum yield of converted sugar. The residence
time will be determined by the reaction temperature and acid
concentration. The through-put capacity of the reactor is
governed by the total refuse handling capacity of the plant and
the amount of recycled cellulose. The size of the recycled
stream is dictated by the amount of incoming unhydrolyzable
-------
-53-
material and unreacted cellulose. The desired recycle ratio
is ultimately determined from cost considerations. That is, as
;he recycle stream is increased, the manufacturing anc, initial
capital cost of the plant will increase as will the product
rate of sugar. The final design recycle flow will be chosen
as that rate which minimizes the total production cost.
Divergence from plug flow of the slurry will produce a
lower expected yield and increase the required capacity of
the reactor. A study of the two phase flow'of wood pulp fibers
is found in References (14, 15, 16). Such fibers exhibit three
regions of flow with increase in flow velocity. The laminar
region is characterized by a plug flow of fibers surrounded by
a thin water annulus. As the velocity increases, the fibers
begin to break up in what is termed the "transition region".
This trend continues until the fibers are completely agitated
in a region of turbulent flow. The velocity profiles associ-
ated with the flow of pulp fibers cannot be described by a
Newtonian fluid model. In Newtonian fluids an increase in
velocity results in blunter velocity profiles, whereas in the
case of pulp, an increase in fiber flow results in a sharper
profile. This tendency toward plug flow at low velocities in-
creases with an increase in fiber concentration. It was assumed
that such a fiber flow will exist in the hydrolysis reactor.
Therefore, the assumption of plug flow in the reactor will be
valid if the flow velocity is in the range of laminar fiber flow.
The transition velocity will be the limiting velocity which will
produce plug flow. Although there is no available expression
for calculating this velocity, it can be obtained empirically
-------
-54-
from the curves in Reference (16) . The transition velocity
occurs at approximately the same value for a given concen-
tration at various pipe sizes. For a five percent consistency
slurry the transition velocity is approximately 6 ft/sec. This ,
velocity is high enough to insure a wide range for the reactor
flow velocity. Some consideration should be given to sedimen-
tation or settling out of the solids at very low velocities,
but if desired, this effect can be reduced by a vertical flow
reactor. The optimum velocity will ultimately have to be de-
termined from a pilot plant flow reactor system.
The material of construction for the reactor must be able
to withstand the corrosive effect of the <1% sulfuric acid so-
lution at temperatures above 400°F. The selection of a material
for these conditions is hampered by the fact that most corrosion
test data are limited to the boiling point of the acid solution
at atmospheric conditions. Reference (17) indicated that the
use of nickel based, nickel containing materials, and high
chromium content stainless steel alloys might be applicable for
such conditions. Hydrolysis experiments were conducted in
cylinders of Monel , Hastalloy-B, chrome-moly, stainless 316,
and Carpenter 20 alloys. The hydrolysis in vessels of Carpenter
20 stainless steel gave the highest yield of glucose. Although
these results did not show a complete study of the corrosion
problem, they did indicate that Carpenter 20 had the best poten-
tial as a material for constructing the flow reactor. An im-
proved variety of Carpenter 20 designated as No. 20Cb-3 has
been developed. A corrosion chart for this material, taken
from a Carpenter Steel Co. technical data sheet, is found in
Figure 5-4. From this chart it can be seen that Carpenter
-------
-55-
Figure 5-4
CORROSION RATE OF CARPENTER STEEL
No. 20Cb and No. 20Cb-3
u.
ul
Ui
o
o
C°C
-------
-56-
20cB-3 exhibits very good corrosion properties in dilute I^SO^
acid at high temperatures. For design and cost considerations
it is assumed that the reactor will be made of carbon steel
clad with Carpenter 20Cb-3.' At a rate of 30 mpy/ one inch
cladding would last for twenty years.
Equipment Design Information
Reactor:
Tubular flow reactor. Design pressure 500 psi.
Carpenter 20Cb-3 cladding
Plug flow at 1 ft/sec
Flash Chamber:
Assumed residence time for adequate vapor separation of
2 minutes. Constructed of Carpenter 20 steel. Design
pressure 150 psi.
Acid Storage Tank:
14 day design capacity.
Constructed of carbon steel.
Centrifuge:
Pressurized (150 psi) continuous variety.
Recommended - Sharpies Super-D-Canter model with continuous
screw conveyor for solid separation. Maximum design
capacity limited to liquid stream of 350 gal/min.
Horsepower » liquid flow rate. H.P. ?a gal/min.
Constructed of Carpenter 20 steel.
Assumed solid moisture content at discharge of 50%.
Slurry Pumps:
Recommended Moyno positive displacement.
Maximum pressure 1000 psi.
Maximum capacity 500 gal/min.
-------
-57-
Feed pump constructed of steel rotor and housing.
Recycle pump constructed of Carpenter 20 steel.
Water Pumps:
Standard steel motorized centrifugal pumps.
Acid Pump:
Positive displacement controlled volume feed pump.
Constructed of Carpenter 20 steel.
5.3 Concentration of Sugar Solution and Heat Recovery
Most processes which use glucose as a raw material require
at least a 6% solution (Torula yeast production) and preferably
a 12% solution (ethanol fermentation) of glucose. With a liquid
to solid ratio of 20 to 1 the glucose solution leaving the cen-
trifuge will be approximately 2% by weight glucose. The glucose
solution must therefore be concentrated after leaving the hydrol-
ysis section of the plant.
An ideal process would perform the required concentrating
while allowing the high temperature stream of water to act as
a preheat or feed for the hydrolysis reactor. Two relatively
new processes, as cited in a recent Doctorate thesis^ ', which
are being tested for use in concentrating spent sulfite liquor,
perform the task to some degree. They are reverse osmosis and
electrodialysis. Reverse osmosis has proven to be effective in
concentrating liquors to 10% by weight solids. If the sugar
solution contained only glucose, this would be a feasible method
of concentration. The difficulty is that the hydrolyzed solu-
tion will contain, in addition to glucose, decomposed sugars
and other large organic molecules from the garbage portion of
the refuse. Therefore, the glucose concentration will be
-------
-58-
initially 2%, but the total solute concentration may be greater
than 4%. The upper limit of glucose concentration by reverse
osmosis will be approximately 5% which is below the 12% re-
quired concentration. Although feasible for the lower limits
of concentration, reverse osmosis, in its present state of de-
velopment, would not ba acceptable for most sugar fermentation
processes. Electrodialysis has been tested experimentally but
as of yet not proven economically feasible on such separation
processes. Thus at its state of the art, it cannot be recom-
mended as a means for concentrating the sugar solution.
One proven means of concentration which will allow partial
use of the high temperature stream as preheat is feed forward
multi-effect evaporation. The proposed evaporator scheme is
shown in Figure 5-3. The recommended effects are short tube
vertical evaporators. They were chosen because of their ability
to be cleaned with a minimum amount of effort.
The liquid stream from the recycle centrifuge is neutral-
ized by means of lime addition. The lime is injected in a
slurry form, using an in-line mixer to insure that complete
neutralization occurs. The lime, or CaCo3, reacts with the
sulfuric acid to produce water, CaSQ^, and CC^. After neu-
tralization the C02 is removed by a gas separation vessel and
the CaSC>4 which precipitates out of solution is separated by
a pressurized centrifuge. The operation of neutralization
followed by separation is performed to produce a non-corrosive
solution and to decrease scaling in the evaporators. Lime is
used as the base because of its comparably low cost. If the
glucose solution will be used for ethanol fermentation, it may
-------
-59-
be advisable to use ammonium hydroxide as the neutralizing
agent. The ammonium sulfate produced by such neutralization
could then be used as a nutrient for fermentation.
The evaporator system was designed to make maximum use of
the high temperature of the feed stream. The upper limit of
preheat for the reactor water feed stream is set by the tem-
perature of the flash chamber quench. This quench temperature
was 177°C (350°F) since at this temperature the hydrolysis re-
action will be essentially zero. The latent heat of the vapor
stream from the flash chamber is used to evaporate the product
stream in the first effect and to raise the temperature of the
direct fired heater feed water stream. It was assumed that the
vapor from the flash chamber would be free enough of impurities
to be condensed directly in the heater water stream, thus elim-
inating a condenser and fresh feed water. This assumption was
again made for the vapor formed and condensed in each effect.
The condensate from each effect (other than the last) is
pumped to the condensing chamber where it is combined with the
flash vapor and make up feed from the last effect's condenser.
This heated stream is passed through the direct fired heater
where it is brought to the required temperature for the hydrol-
ysis reactor. A bleed off from the preheat stream could be
used to prevent impurities build up if the assumption of clean
vapor is not valid.
The number of effects used in the evaporation system was
determined by an economic analysis of the evaporator-heater
system. That is, for a given initial and final concentration
of sugar, reactor feed water temperature, and flash condition,
-------
-60-
the optimum use of the potential heat source was made to mini-
mize the required capital investment of the evaporator heater
system and the utility cost of heating oil. It was found that
for initial and final concentrations of 2 and 12%, the optimum
number of stages would be 6. This optimum number does not vary
with feed capacity, but it will vary with the degree of concen-
tration necessary. At 4% to 12% the number of stages was de-
termined to be 5. This increase in initial concentration would
correspond to a decrease in the hydrolysis liquid to solid ratio.
Six effects were used in the plant design and cost estimates;
this would allow one unit to be used in standby if maintenance
problems develop.
Equipment Design Information
Evaporators:
Standard vertical-tube evaporators.
Heat transfer area determined by flow capacity.
Constructed with cast .iron shell and copper tubes.
Condenser:
Standard shell and tube construction.
Cooling flow through tubes.
Constructed of carbon steel.
Atmospheric pressure.
Condensing Chamber:
Volume based on 2 minutes residence time.
Standard steel construction.
Design pressure 150 psi.
-------
-61-
Direct Fired Heater:
Cylindrical construction.
Carbon steel tubes.
Design pressure 500 psi.
In Line Mixer:
Carpenter 20 steel.
Nettco Flomix is recommended (Nettco Corp., Everett, Mass.)
Centrifuge:
Sharpies Super 0 - Hydrator continuous operation.
Ordinary steel construction.
Pressurized at 150 psi.
Pumps - Water:
Motorized centrifugal pumps
Carbon steel construction.
Pumps - Slurry:
Moyno slurry pump.
5.4 Design Calculations
The basic design calculations were generalized so that a
computer program with variable plant inputs could be written.
The inputs to the program consist of:
1. Total plant refuse processed.
2. Fractional input of paper
3. Percent cellulose in paper
4. Fraction garbage in refuse
5. Fraction of refuse separated before hydrolysis
6. Fraction of inerts in hydrolysis feed
7. Liquid to solid ratio for hydrolysis
-------
-62-
8. Recycle ratio
9. Acid concentration
10. Reaction temperature
11. Time to maximum yield
12. Chemical yield of glucose.
The separable portion of refuse was assumed to be that which
can be separated in the hydropulper system (metals, plastics,
stones, dirt, etc.). The garbage portion of refuse flows into
the hydrolysis reactor with the paper. Although the majority
of the garbage, organic plant waste, will be decomposed, it was
assumed that fifty percent would not be hydrolyzed. This por-
tion would add to the inert portion of the hydrolysis recycle
stream and therefore give a conservative estimate for the re-
actor size and final glucose concentration. The inert fraction
would include the clay paper additives in addition to the unhy-
drolyzable portion of the garbage.
A. Material Balance (Appendix IV)
The following assumptions were made in making the material
balance:
1) Specific volume of paper slurry 62.4 lb/ft3
2) The liquid content leaving the pre-hydrolysis screw
press will be 50%.
3) 66 Be' sulfuric acid (93% H2S04)
4) Solid stream from centrifuges will contain 50% moisture.
5) Degree of refuse separation described previously in 5.4
6) Lime slurry contains 50% liquid.
A computer program print out of a material balance and a gener-
alized flow sheet follows in Table 5-1 and Figure 5-5.
-------
-63-
Table 5-1
MATERIAL BALANCE
REPRESENTATIVE 250 TON PLANT
? 250
INPUT FRACTION PAPER,GARBAGE,SEPARABLES IN WASTE
? .6,.15,.25
INPU1 FRATION CELLULOSE IN PAPER
? .8
INPUT TIME TO MAX YIELD IN MIN
? 1
INPU'l CELLULOSE FRACTION CONVERTED AND SUGAR PRODUCED
? .75,.55
INPUT FRACT ACID,SOLID 10 LIQUID.REACTION TEMP CENT
? .004,.1,230
INITIAL FRACTION INERTS 0.28125
INPUT II .INERT RATIO FINAL
? .4
DO YOU WANT MATERIAL FLOW BALANCE YES OR NO
? YES
SEPARATION SYSTEM
SEPARATOR
WASTE
WATER
IN
250
12500
OUT T/D
187.5
375
IN
10.4167
520.833
**************************** ********************
*********************************************
FLASK
WATER
IN T/D
2173.22
OUT T/D
1917.73
IN T/KR
90.551
OUT T/HR
7.8125
15.625
REACTOR
CELLULOSE
WATER
ACID
SUGAR
D SUGAR
S INERTS
L INERTS
IN
131 .875
2183.75
9.39247
0
0
52.75
33.75
OUT T/D
32.9687
2173.22
9.39247
79.7844
29.0125
52.75
33.75
IN
5.49479
90.9896
0.391353
0
0
2.19792
1.40625
OUT T/HR
1.3737
90.551
0.391353
3.32435
1.20885
2.19792
1.40625
OUT T/HR
79.9053
*•****#•(<•#***** ******* ***************-*#****#*******
CENTRIFUGE #1
CELLULOSE
WATER
ACID
SUGAR
D SUGAR
S INERTS
L INERTS
*******************BLEED AND
MATERIAL BLD-T/D
CELLULOSE 21.0937
S INSRTS 33.75
WATER 54.8437
ACID 9 .62221 E-2
PLUS SMALL AMOUNTS OF SUGAR
IN
1.3737
' 79.9053
0.391353
3.32435
1.20885
2.19792
1.40625
LIQUID OUT
0
76.3336
0.385087
3.17657
1.15512
0
1.34374
SOLID STREAM
1.3737
3.57161
6.26631 E-3
0.147781
5.37386 E-2
2.19792
6.25136 E-2
RECYCLE*******************
TON S/HR
REC-T/D
11.875
19.
30.875
5.41695 E-2
AND LIQUID INERTS
BLD-T/HR
0.878906
1.40625
2.28516
4.00925 E-3
REC-T/HR
0.494792
0.791667
1.28646
2.25706 E-
*******************************************
-------
-64-
32*
Table 5-1 (Continued)
NEUTRALIZER
WATER
ACID
CASO4
C02
LIME
SUG^R
D SUGAR
L INERTS
IN T/D
1841.44
9.24208
0
0
9.4307
76.2376
27.7228
32.2497
OUT T/D
1843.14
0
12.8257
4.14951
0
76.2376
27.7228
32.2497
IN T/HR
76.7266
0.385087
0
0
0.392946
3.17657
1.15512
1.34374
****#****#*********«*********«*******************#
OUT T/HR
76.7973
0
0.534406
0.172896
3.17657
1.15512
1.34374
CENTRIFUGE 2
WATER
CAS04
CO 2
SUGAR
D SUGAR
L INERTS
IN
76.7973
0.534406
4.14951
3.17657
1.15512
1.34374
LIQ OUT
76.2629
0
0
3.17565
1.15478
1.34335
SOL OUT
0.534406
0.534406
0
9.21027 E-4
3.34919 E-4
3.89608 E-4
TONS/HR
*#*#****#*##***#***#********************************
REACT WATER T/HR AT TEMP
81 .8906 T/HR 244.561 DEC CENT 472.209 DEC F
TEMP TO HEATER DEC F AND C
329.809
165.45
EVAPORATORS
STEAM FROM FLASH USED 2.9652 TONS/HR
EFFECT VAPOR TONS/HR
1 4.86472
2 6 . 58 32 7
3 7.98176
4 9.1766
5 10.0442
6 10.9322
-------
s-s
.ILJ" ."f T1
-------
-66-
B. Reactor Design and Recycle Calculation
The design of the reactor vessel is determined by the flow
capacity, flow velocity, and required residence time. The total
flow through the reactor is found by a recycle material balance
on the reactor system.
The recycle calculation is performed on the solid portion
of the refuse. It is only necessary to consider the cellulose
and inert portion of the stream. The soluble portion of non-
cellulosic materials and liquid will be separated by the con-
tinuous centrifuge and therefore will not be included in the
recycle loop.
Let ( C^ = fresh feed of cellulose Tons/hr
Ij_ = fresh feed of inerts Tons/hr
C2 = recycled cellulose Tons/hr
R, = entering ratio of inerts to cellulose
R2 = exit ratio of inerts to cellulose
B = amount of cellulose bled off Tons/hr
Y = fractional conversion of cellulose
YI = fraction yield of sugar
The final ratio of inerts to cellulose at the exit of the
reactor will be:
Rj_ (C^ + C2) R^
^ Y (CI + C2) Y
At steady state the amount of inerts entering the reactor will
equal the amount of inerts bled off.
1^ = R2-B
A material balance around the bleed point requires that the
-------
-67-
(1 - Y) (C1 + C2) = (C2 + B)
B - (l-Y)Ci
C = -
* Y
Thus once values for R^ and Y are set the total amount
of cellulose and inerts in the reactor can be determined.
The total amount of solid material in the reactor will in-
clude, in addition to the cellulose and inerts, the non-hydro-
lyzable portion of garbage (G). Total solids will be:
S__ f+ i^ f\ _L f^ ,i TJ / f^ i_ ^ \ •
*~ o i WT T v* o * *»T VNO "• >* o /
Ju *• J. a. t»
From the solid to liquid ratio, L, the necessary liquid flow rate
is found to be W = LxS and the total weight of material will
be M = S + W Tons/hr. With the assumed value of 62.4 lb/ft3
the total volumetric flow rate can be determined (Q). This
value in conjunction with the residence time T determine the
reactor volume, V = QxT. The necessary cross sectional area
and length can be found from this volumetric flow, residence
time, and flow velocity of 1 ft/sec.
C. Energy Balance
The heat load to the system is a function of plant capacity,
liquid to solid ratio, recycle ratio, reaction temperature, and
preheated feed water temperature. The first three of these are
directly associated with the required liquid flow stream in the
reactor (W) . The temperature of this stream (T-^) is set by the
reaction temperature. The preheat temperature (T2) is calcu-
lated by an energy balance on the flash and evaporator section.
V. 6 Tj:^ + FT£
T2 = -J- + E
\ c + F
01 e
-------
-68-
Where V = vapor from flash chamber, Ib/hr
X = latent heat of vapor, BTu/lb.
W = water flow rate to reactor, Ib/hr
T3 rn. •,
C = heat capacity, 1
T = temperature from effect e °F
C = condensate from effect e —
e nr
T = make up feed temperature °F
F = make up feed flow Ib/hr
The required heat load to the system will then be
Q = (Tx - T2)WCp
This is supplied by the direct fired heater with oil at
1.5E5 —=-y with 80% conversion efficiency.
D. Flash Chamber Design
The amount of material vaporized in the flash chamber can
be determined by an adiabatic enthalpy balance where: The
enthalpy of the feed equals the combined enthalpy of the vapor
and liquid stream leaving the chamber.
F HL = V HV + L HL enthalpy balance
i o o
F = V + L mass balance
F HL = V H^ + (F - V)H^
j_ o o
H^ - HL
=• = — r fraction flashed
£ -rrV TjJ-l
il Jtl_
o o
For a flash from 230°C to 177°C
» = 0.117
The size of the flash chamber is determined by the cross
-------
-69-
(19)
of the chamber is usually between 7 and 12 ft. r to prevent
splashing and therefore liquid entrainment. The allowable vapor
2
flow rate (G) Ib/hr.ft is determined by an empirical correla-
tion based on a decontamination factor, DF = weight vapor/
weight of entrained liquid. From Reference (20) the recom-
mended flow rate for DF » 10/000 at 300°F is 200 Ib/hr.ft2.
The recommended flash temperature is 350°F, therefore it is
(19)
necessary to use the empirical correlation found in Perry
,1
G = (T 'P (P£ - Pg)
With G = Ib/hr.ft2 of vapor
p = vapor density lb/ft3 at flash temp.
p^ = liquid density lb/ft3 at flash temp.
C1 = empirically correlated factor
C1 = 80 at DF of 10,000
For T = 350°F
G = 320 Ib/hr.ft2
For a total flow rate of 2 x 10 Ib/hr the cross sectional area
2 3
required is 69 ft and the total volume is 8300 ft . This volume
will result in a vapor residence time of 2:0.7 minutes. For the
computer program a residence time of 2 minutes was used to cal-
culate a nominal flash chamber volume. This will give a con-
servative estimate of the volume required for the preliminary
design.
E. Evaporator Design
The design procedure for feed forward multi-effect evapo-
(21)
ration systems was taken from Kern . In such a design the
temperatures and pressures in the first and last effects are
fixed. It is usually assumed that equal areas will be used and
-------
-70-
that under such conditions the pressure difference between
effects will be approximately equal. The.equal area restric-
tion is imposed, partially, because it is less expensive to
build a system with equal area effects. From these pressures
the saturated liquid temperature can be found. The actual tem-
perature and pressure drop in the system will adjust itself
during operation according to the actual heat transfer coef-
ficient in each stage. The steam supply for the first effect
is taken from the flash chamber and the vapor formed in each
effect is used to evaporate the liquor in the subsequent effect.
This liquor will boil at a lower temperature than the condens-
ing vapor because its saturation pressure is lower than the
vapor's, due to the staged pressure drop.
The steam and surface requirements for the multi-effect
evaporation system are computed by performing a heat balance
across each effect and an overall material balance on the system.
WSX + WfCf(Tf-T1) = W-^ Heat balance 1st effect
i
W- -,\- -, + (w:=- Z Wn) C._i (T- -,-T.) = W.X. Heat balance on
i-l x-1 f n=1 n 3.-1 1-1 J x x subsequent effects
TT
E = z, W_ Material balance
n=l n
A = wsxs
1 u-(T -T ) Required surface area
where Cf = specific heat of feed BTu/lb°P
TF = feed temperature °F
WF = feed Ib/hr
TS = saturation temperature of steam °F
-------
-71-
W = steam to first effect Ib/hr
O
E = total required evaporation Ib/hr
C. = specific heat of liquor in effect i BT,/lb°F
T. = boiling point of liquor in effect i °F
W^ = vapor removed in effect i Ib/hr
A£ a latent heat o£ vapor i BTu/lb
A- = area of effect i ft^
U. = heat transfer coefficient for effect i BTu/hr.ft2 °F
Therefore for a system composed of K effects there will
be K + 1 unknowns, the required steam input and K vapor rates,
and K + 1 equations. These equations were solved by matrix
inversion in a variable input computer program. If the areas
are not equal the pressure drops, and therefore temperature dif-
ferences, are adjusted until they are equal.
It was assumed that the boiling point rise due to solute
content will be negligible. This assumption was based on infor-
(21}
mation obtained from Kern v ' which indicated that over the
range of 0 to 20°Brix (£? % weight sugar) the.BPR was approxi-
mately 1°. It was further assumed that the specific heat of
the solution would be 1 BTu/lb°F. Over the same range of con-
centration C_ varies from 1 to 0.9.
£T
The heat transfer coefficient in each stage will be con-
trolled by the temperature, temperature difference, the viscosity
of the liquor, and scale formation in the effect. The standard
method of calculating an overall heat transfer coefficient from
the individual resistances is not used in practice. Most in-
formation concerning the heat transfer coefficients is obtained
from operating experience with the given type of effects.
-------
-72-
Reference (19) gives a range of experimental values for ver-
tical tube evaporation with natural water circulation of 200 -
500 BTu/hr.ft2°F. Kerr^19) gives data for operational transfer
coefficients obtained with a feed forward system used in concen-
trating cane sugar to 50° Brix. The transfer coefficients are
from 100 to 450 over a temperature range of 220 to 120°P.
(21)
Kern ' also presents data for concentrating cane sugars from
approximately 13 to 50°Brix with AT=23°F and a temperature range
of 274 to 180°F.
The proposed system for concentrating the hydrolysis sugars
operates with a AT2?23°F, temperatures from 350 to 212°F, and
sugar concentrations from 2 to 12%. Since other solids will be
mixed with the sugars, the actual concentrations will be from
approximately 4 to 24%. Given the previously mentioned experi-
mental data and the above specifications, a range of coefficients
from 500 to 250 was used in the design calculations. The data
in Table 5-2 were used in the design of the multi-effect system.
-------
-73-
Table 5-2
EVAPORATOR DESIGN DATA
Steam Press,
Chest psia
1
2
3
4
5
6
To CONDENSER
135
100
74
57
35
25
14.7
Temp F6
350
327
304
282
260
240
212
L,
BTu/lb
870
889
905
924
939
952
970
U' o
BTu/hr ft2
500
480
450
410
370
250
500
-------
6. PLANT ECONOMICS
Many of the design parameters for plant operation cannot
be determined without an economic evaluation of the hydrolysis
plant. That is, variables such as recycle rate, acid concen-
tration, reaction temperature, and residence time must be set
such that the manufacturing cost of glucose is minimized. A
knowledge of the effect of total refuse input and its compo-
sition on the manufacturing cost must also be determined, be-
fore a decision can be made concerning the economic potential
of refuse disposal by paper hydrolysis. It would require a
great amount of labor to perform an overall economic analysis
such as this, without the use of a plant-simulating computer
program.
The previously described generalized material flow program
(Section 5.4) generates the information necessary to size the
individual plant components. The size of each component can
then be used to determine its purchased and installed cost. From
these costs, the total fixed capital investment is calculated.
The material balance is also used to determine the total manufac-
turing cost per pound of glucose.
6.1 Capital Costing Procedure (Appendix IV)
Estimates of equipment costs, not including the hydropulper
system, were taken from a recent article by C.E. Guthrie^2^.
The equipment cost information in this article included, in ad-
dition to purchase costs, the size exponential factor for each
piece of equipment and a direct material and labor factor used
to determine the total installed cost. The M & L factor includes:
-74-
-------
-75-
piping, concrete, steel, instruments, electric material, insu-
lation, paint, and labor necessary to install the component.
The hydropulper system cost was estimated by the Black Clawson
Co. for an 80 ton/day capacity system with an 0.6 factor for
scale-up. The cost information in Reference (22) was for mid-
1968 and a 6% escalation was used to predict the capital invest-
ment for mid-1969.
To arrive at a value for the final fixed capital invest-
ment, a percentage of the installed equipment cost was used to
determine the additional plant cost.
Building Cost = 20% of IEC (Installed Equipment Cost)
Outdoor-indoor type construction
Freight & Taxes = 8% of IEC
Construction = 17.8% of IEC
Engineering = 10% of IEC
Contingency & Contractor Fee = 18% of Direct Plant Cost
Working Capital = 15% of Direct Plant Cost
The above percentages were taken from Reference (23) and are
accepted values used in preliminary cost estimations.
A printout of the equipment sizes and cost for a nominal
250 ton plant are shown in Figure 6-1. The manufacturing costs
in this example represent 24 hours of continuous operation.
Given any set of initial conditions for plant operation, the
computer program will calculate the presented information. In
this manner, it is possible to determine the total manufacturing
cost per pound of sugar for any municipality with its own local
plant operating cost and refuse composition, 'in order to use
this program to evaluate a plant for any location, see Appendix IV.
-------
-76-
Figure 6-1
CAPITAL & MANUFACTURING COST ANALYSIS
REPRESENTATIVE 250 TON PLANT
r
DO YOU WANT EQUIPMENT SIZE AND COST YES OR
? YES
*>*******CQST ANALYSIS*******************
EQUIPMENT
HOPPER
CONVEYOR
KYPROPULP
RFACTOR
;>CID STORAGE
FLASH CHAM
CENT 1
NrUT
LIME ST
CENT 1
HEATER
OIL ST
EVAP
CCND
COND POT
PUMP
NO
SIZE
27000
100
250
47.6612
121078.
97.2111
283.956
304.302
2218.99
304.302
2.33224 E+7
43900.9
1521.98
1291.5
87.49
FT" 3/KR
500.801
12019.2
751.202
6.85696
412.326
937.577
, 18.4916
FT" 3
FT
TONS/DAY
FT" 3
GAL
FT" 3
GAL/MIN
GAL/MIN
FT" 3
GAL/MIN
BTU/HR
GAL
FT" 2
FT" 2
FT" 3
H.P.
2.73062
65.5349
27. 3062
0.311563
14.9881
34.0809
0.168042
P COST
9732.57
9978.68
133363.
9089.98
5668.84
10461.5
79387.6
3711.69
1026.93
17643.9
74047.6
8939.09
349953.
10628.1 •
3101.27
P.C.
1190.92
6217.17
3943.51
660.805
2013.16
4425.2
279.412
IN COST
10705.8
16165. 5
213381.
13332.
10487.4
13782.6
127020.
5938.7
1129.62
28230.2
120698.
16537.3
664911.
24869.7
6202.53
i.e.
2870.12
14983.4
9503.86
1592.54
4851.72
10664.7
673.382
TOTAL PURCHASED EQUUIPMENT COST 790192. DOLLARS
10TAL INSTALLED EQUIPMENT COST 1.39764 E+6 DOLLARS
BUILDING COST
FREIGHT AND TAXES
CONSTRUCTION COST
ENGINEERING COST
279529. DOLLARS
111811. DOLLARS
248781. DOLLARS
139764. DOLLARS
DIRECT PLANT COST 2.17753 E+6 DOLLARS
CONTINGENCY AND CONTRACTOR FEE 391955. DOLLARS
FIXED CAPITAL INVESTMENT
WORKING CAPITAL
2.56948 E+6 DOLLARS
385423. DOLLARS
(continued)
-------
-77-
Figure 6-1 (continued)
•****************MANUFACTURING COST********************
INPUT DUMPING FEE PER TON
? *
input cost for disposal of waste per ton
? 3
INPU1 INTEREST RATE AND YEARS OF OPERATION
? .04,20
RAW MATRIALS
ACID 9.24208 TONS/DAY
LIME 9.4307 TONS/DAY
UTILITIES
WATER 150902. GALS
ELECT 12984.3 KILOWATTS
4390.09 gal
86.0937 TONS/DAY
72 ERS
24 HRS
oil
WASTE DISPOSAL
LA BOR
SUPER
FRIN3E BEN
MAINTAIN 0
SUPPLIES 0
FIXED CHARGES
TAXES 0
INSURANCE 0
cap return
GENERAL COST
PAY OVER 0
LAB 0
PLANT OVER 0
TONS /DAY OF GLUCOSE 75.7071
TOTAL COST/DAY NO DUMP FEE
COST/TON OF GLUCOSE NO DUMP
COST/LB OF GLUCOSE NO DUMP
TOTAL COST/DAY WITH DUMP FEE
COST/TON WITH DUMP FEE
COST/LB WITH DUMP FEE
TIME : 6.365 SEC.
READY
0
0
0
0
0
0
0
295.747
113.168
37.7256
162.303
439.009
258.281
216
84
45.
356.873
26.7655
142.749
71.3745
603.964
71.7655
71.7655
239.218
3235.71 DOLLARS
42.7398 DOLLARS
2.13699 E-2 DOLLARS
2235.71 DOLLARS
29.531 DOLLARS
1.47655 E-2 DOLLARS
-------
-78- . y;
6.2 Manufacturing Cost Estimation
The amount of raw material required for a given plant size
is calculated from the material balance. The raw materials
purchase prices were taken from a June 16, 1969 edition of the
Oil, Paint and Drug Report in which the current prices f.o.b.
New York are given for all major chemicals.
Utility costs were calculated from rates given in Ref. (24)
for 1967. These rates will vary with the given plant location.
Mid-range values were used for calculation purposes.
Labor costs were based on three men operating the equip-
ment with one supervisor, and three shifts. This amounts to
one man per plant section with one overall supervisor. Labor
rates are reported in U.S. Department of Labor publications and
a yearly index for such rates is given in Reference (24).
Other direct costs such as fringe benefits, maintenance,
and repairs are calculated on a percentage basis, as are the
indirect costs of payroll overhead, laboratory and plant over-
head. These percentages were taken from References (23) and (24),
The fixed charges such as capital return, taxes, and insur-
ance are all based on the fixed capital investment. The charges
for the original capital investment depend upon the source of
the capital investment. It was assumed that these charges would
be 4% for 20 years if the plant is built by the municipality,
and 10% for 20 years if by private industry. From these interest
rates and years of operation an overall capital recovery factor
can be determined. The taxes and insurance are a percentage of
the fixed capital investment. Table 6-1 gives the values used
for calculating the direct, indirect, and fixed cost of manu-
facturing.
-------
-79-
Table 6-1
MANUFACTURING COST ANALYSIS
Direct Cost
Raw Materials
H2S04 66%
CAC03
Utilities
Oil
Water
Electricity
Labor
3 Laborers
Supervisor
Maintenance (Materials & Labor)
Supplies
Fringe Benefits
$32.30/ton
$13.50/ton
$ 0.10/gal.
$0.25/1000 gal.
$0.010/KWHR
$3.00/hr.
$3.50/hr.
5% of F.C.I.
13% of Maintenance Material
15% of Labor Cost
Indirect Cost
Payroll Overhead
Laboratory
Plant Overhead
Fixed Charges
Capital Investment
Municipal
Private
Taxes
Insurance
15% of Total Labor Cost
15% of Total Labor Cost
50% of Total Labor
4%, 20 years
10%, 20 years
2% of FCI
1% of FCI
-------
-80-
In addition to these operating costs there will be the
cost of disposing of unhydrolyzed waste from the plant and pos-
sible credit for disposing of the original waste. The charge
for disposing of plant waste will be lower than that associ-
ated with raw refuse disposal since it is compact and already
separated. For calculation purposes/ it will be assumed to cost
$3.00; approximately the lower limit of incineration cost. A
dumping fee of $4.00 was used to demonstrate the effect of waste
disposal credit on the total manufacturing cost. As stated
earlier, the dumping fee is the charge for waste disposal assigned
to the municipality by the plant. It can vary from zero to a
value which is just under the cost of the next best means of
disposal.
6.3 Manufacturing Cost Analysis
The operational plant variables such as temperature, acid
concentration, liquid to solid ratio, and recycle ratio, must
be set such that the sugar manufacturing cost is minimized. The
reaction temperature and acid concentration effect on sugar cost
are independent of the recycle ratio but not the liquid to solid
ratio. If a lower liquid to solid ratio is used, then an in-
crease in acid concentration will not increase the total acid
cost as appreciably as it would at a higher ratio. It is not
possible to determine what the lowest feasible liquid to solid
ratio is without data from a flow reactor experimental apparatus.
(26)
It has been shown by Saeman et alv ;that values as low as 3 to 1
produce consistent hydrolysis yields. The hydrolysis plant
relies on thorough mixing of the pre-pulped cellulosic material
and acid solution prior to passage through the flow reactor.
Due to such consideration, a middle range value of 10 to 1 was
-------
-81-
chosen for the liquid to solid ratio. A study of the acid
concentration and temperature effects on the operating cost was
conducted for a hypothetical 250 ton plant with 40 and * 0%
paper contents. The cellulose content of paper (amorphous and
crystalline) was set at 80%. The expected sugar yields were
taken from the experimental analysis section of this report.
A study of the sugar manufacturing price was performed for
ranges of temperature from 220 - 240°C and acid concentrations
of 0.2 to 1%. It is believed that extrapolating yields for
acid concentrations outside of this range would not be possible.
Not enough information concerning the exact effect of the acid
catalysis on yield is known to allow this. Above 240°C the
reaction residence time is below 10 seconds. Such short resi-
dence times give little margin for process control and it would
be unrealistic to consider such conditions for plant operation.
The manufacturing costs for the conditions considered are pre-
sented in Table 6-2. These costs range from 8.5 to 2.7C/lb.
The costs with a dumping fee credit of $4.00 range from 7.3 to
2 cents/lb. The lowest cost occurs at a 1% acid concentration
and a 230°C reaction temperature.
The optimum recycle rate for these, 230°C, 1% and 60% paper,
was found to be a 0.4 ratio of inerts to cellulose in the reactor,
A range of 0.3 to 1 was studied and minimum cost occurred at 0.4.
The optimum recycle setting varies with reaction conditions,
yield, L/S ratio, and plant capacity. It would therefore be
impossible to set a value which would be optimal for any given
hydrolysis plant.
-------
-82-
Table 6-2
EFFECT OF ACID CONCENTRATION AND
TEMPERATURE ON MANUFACTURING COST
COST 0 lb,
Acid
% Paper Concentration
40 .2
60 .2
40 .5
60 .5
40 1.
60 1.
Temperature °C
220° 230° 240°
8.5«
6.0
5.0
3.7
• 4.8
3.3
7.4« 6.2$
5.4 4.4
4.7 4.3
3.2 2.8
4.2 *
2.7 *
*Not feasible (reaction time too short)
-------
-83- ' ,-
The effect of overall plant capacity on the manufacturing
cost of sugar was studied for various refuse compositions.
Table 1-2 demonstrated that a wide range of municipal refuse
compositions can be expected. Three different refuse jompo-
sitions: 40% paper, 30% garbage, 30% separables; 50% paper,
17% garbage, 33% separables; 60% paper, 15% garbage, 25% sep-
arables, all with paper containing 75% cellulose, were used to
determine the manufacturing cost/lb. sugar. In this manner the
expected cost can easily be determined for the various types of
communities. A basis of 5 Ib. of refuse per capita per day
was used in calculating population from total plant capacity.
Figure 6-2 presents this calculation and Figure 6-3 shows the
approximate fixed capital investment for such plants. Opera-
tional variables used in these calculations were 230°C, 1%
acid, and a 10 to 1 liquid to solid ratio. Although the re-
cycle ratio was not optimized for each plant size, it was held
in a suboptimum range. In this range the recycle ratio had
little effect on total cost and was close to the actual optimum
value.
Figure 6-2 also shows the range of sugar cost which would
result from using blackstrap molasses as a raw material for a
fermentation process. Blackstrap molasses is the waste syrup
from a sugar cane crystallization process. It is the most
widely used fermentation raw material and therefore would be the
most competitive sugar containing raw material. Blackstrap
molasses contains approximately 55% sugar by weight including
sucrose, glucose, and fructose. Since this molasses is an
agricultural commodity, its selling price fluctuates a great
-------
-84-
Figure £-2
SUGAR COST .versus PLANT CAPACITY,
/•>
h
3
M
V
ov.
Z'dL
5 1
3 .§
G-3
a
ii
40V. MsPEfc CONTBHT
CO*Tfi VWITMCJUT PUMI*
OP
COtt Of
PLANT CAPACITY CT&NS WASTE/DAY]
o
H
\co
aoo
-4 - 1
300
1 - 1 - 1 —
iooo
400
-4 1
CORR.ESPONPIN& POPULATION SOURCE (xlO )
-------
-85-
Figure 6-3
CAPITAL COST versus PLANT CAPACITY
-------
-86-
(21\
deal. The Commodity Year 1968 Book gives the selling price
for molasses up to February 1968/ and the ranges for 1966/ 1967,
and 1968 are 11 to 12, 17 to 18.75, and 17 cents/gallon. This
fluctuation in selling price greatly affects the ability of a
process to produce a fermentation product at a constant market
price, which ia on© of the main reasons for the §witch from fer-
mentation to synthetic production of many products. Hydrolysis
of refuse would produce sugar at a more constant cost and over
the years the actual production price should' decrease, since
the trend is toward a higher level of refuse paper composition.
The main advantage of using molasses is its high sugar concen-
tration which allows it to be used for any fermentation process,
with economical shipment of the raw material over long distances.
Plants which operate at cost below that of molasses can be
considered as producing a saleable raw material. It is noticed
that in some cases it would be necessary to charge a credit for
the refuse processed to operate in a competitive range. In the
case of a municipality this would require absorbing some of the
cost of production. If the necessary charges are below the
existing cost of waste disposal, then it would be of an economic
advantage to build a hydrolysis plant. If the plant is owned
by a private industry then it would require that the plant
charge the community a set cost for disposing of its refuse.
The quoted prices given in the graph are for a capital
return factor based on 4% interest for 20 years. If owned
privately, a factor of 10% for 20 years would have to be used.
With this value the cost per pound increases approximately
0.2 cents/lb. Another factor which is not shown in this graph
-------
-87-
is the liquid to solid ratio effect. A 10 to 1 ratio was used
for the calculations but this is not necessarily the lowest
possible ratio. It was found for a 500 ton plant operating
with 50% paper that, at a 7 to 1 ratio, the manufacturing price
drops from 2.8 cents/lb. to 2 cents per Ib., and correspondingly
an increase to 13 to 1 increases the cost to 3.4 eente/lb, This
points out the dramatic effect the liquid to solid ratio has on
overall production cost.
For the conditions used, it is therefore shown that the
following plants could produce a 12% aldohexose solution which
would be competitive with molasses on a cost/lb. sugar basis:
a 40% paper composition would reach the competitive range at
400 tons with a dumping charge; a 50% paper composition would
be competitive at the 400 ton level without a dumping charge
and at the 200 ton capacity with a dumping charge; a 60%
paper composition would be competitive at the 200 ton level
with no dumping fee and the 100 ton level with a dumping fee.
The minimum expected costs at 1000 tons capacity, 60% paper,
with and without a dumping credit, are 2 and 1.3 cents/lb.
respectively. These values are well below the average cost
associated with molasses of ~2.5 cents/lb. of sugar.
The lowest predicted disposal price for the most economical
incineration process is approximately $3.00/ton, presently
~$4.00/ton. Although sanitary landfill can undercut this cost
in some areas, there are many cities which have no available
close land sites which enable them to even approach $3..00/ton.
In such municipalities, with refuse paper contents greater than
50%, refuse hydrolysis would be a means to cut disposal cost and
produce sugar at a competitive market price with molasses.
-------
-88-
6.4 The Marketability of a Glucose Solution
It is assumed that the sugars produced by the hydrolysis
are fermentable. This seems to be a valid assumption since the
aldohexose sugars of glucose and mannose, which are produced,
are the same sugars used in most fermentation processes. It
(28}
has also been proven by Saeman et alv ' that these same sugars
obtained by wood hydrolysis are fermentable. In some cases
contamination of the fermentation product occurs, such as by
the S02 in spent sulfite liquors. Such difficulties are not
foreseen with the hydrolysis plant product which, in addition
to being free of such inorganics as SO^, should be free of micro
organism contamination due to the high temperatures of the
reaction.
The total U.S. consumption of blackstrap molasses in 1967
was 700 million gallons. Approximately 300 million gallons were
used in the industrial production of drugs, citric acid, vinegar,
and ethanol. This indicates that approximately one million tons
of such sugars were consumed in processes which could equally
as well use the product of a hydrolysis plant. This does not
necessarily mean that a greater market is not available for the
raw sugar solution produced by hydrolysis, since at a lower
price than molasses, this product may reopen the markets that
have since been closed by synthetic processes. The main draw-
back to producing sugar solutions by hydrolysis is the low
concentrations produced. That is, at concentrations of 10 to
15% (molasses is approximately 55%) the cost of shipment to
distant areas may preclude its ability to compete with molasses
prices. Consideration must therefore be given to building a
-------
-89-
plant, which uses the hydrolysis plant product as a raw material,
in conjunction with the hydrolysis plant. The construction of
a fully integrated source and consumption plant has t le advan-
tage of cutting the overall cost of production for both the rav;
sugar and the final product of ethanol, citric acid, etc. The
construction of such a plant with the final product of ethanol
was suggested by Porteous.
Ethanol or ethyl alcohol was originally totally produced
by fermentation of molasses, but as of 1954, synthetic processes
using ethylene hydrolysis and hydration began to replace fermen-
tation. Presently the only large fermentation plant in exis-
tence is operated by Publicker Industries, Philadelphia, Pa.
This change of processes was due to the variable price of black-
strap molasses and the low price of ethylene. Of the possible
products which can be produced by fermentation, ethanol has the
largest available market; approximately 520 million gallons
were produced in 1963. As of June 16, 1969, ethanol was selling
at $0.52/gallon. There are a myriad of uses for ethanol as
both a raw material and as a solvent. If the selling price
were reduced, there would most likely be a much larger demand
for ethanol.
Browning'°' states that sugar must be produced at approxi-
mately one-third the 'cost of ethylene before fermentation would
be competitive with synthetic ethanol production. This applies
to the production of ethanol by private industry where a margin
of profit is expected. Using the current selling prices of
ethylene, 3.25£/lb., this requires that sugar be produced at
1.08C. Even with a dumping fee of $4.00, the lowest predicted
-------
-90-
sugar cost is 1.3$. This stringent requirement indicates that
it would be very difficult to persuade a private concern to
build and operate an ethanol producing refuse disposal plant
under the present cost considerations. It should be kept in
mind that these costs are based on the current knowledge of
the system, and that a decrease in cost could result if it Wefe
found that a flow reactor could operate at a lower liquid to
solid ratio. It is believed that if a municipality were will-
ing to build and operate an ethanol plant, it could dispose of
its refuse without charge. Since such a plant would not have
to operate at a profit, its sole manufacturing criterion would
be to produce ethanol at the break-even point. This would re-
quire that the municipality enter the chemical sales field but
this factor should not dissuade the municipality from such an
advantageous means of refuse disposal.
An estimate of the ethanol manufacturing cost can be ob-
tained by a cost analysis performed on the raw material required
(29)
for 1000 gallons of ethanol , ana an updated investment cost
per unit of capacity. The cost associated with the sulfuric
acid used to adjust the pH is excluded, since the pH can be
fixed by the degree of neutralization in the hydrolysis plant.
Since ethanol production by fermentation is a dying art, the
latest capital cost per year capacity was $78 in 1950. This
value is presumably high, since in most cases^30' the actual
cost increase over such a long period of time is below that
indicated by cost indexing methods. This smaller increase is
primarily due to the technological advancements in the manufac-
ture of process equipment and materials. A value of $100/ton
-------
year would be considered more realistic. Using this value,
the indirect and direct cost of manufacturing are calculated
on a percentage basis as indicated in Section 6.1. Table 6-3
contains the ethanol by fermentation cost estimation.
The manufacturing price of ethanol is calculated as
$0.52/gallon. This result implies that ethanol could be pro-
duced at a value equal to the existing selling price. There-
fore for a 500 ton plant with 50% paper, ethanol production
would be an economical use for the refuse produced sugars. It
is also seen from these results why private industry would not
be willing to enter the hydrolysis-ethanol industry with the
predicted sugar cost. The overall capital cost for a 500 ton
hydrolysis plant and a 5.4 million gallons ethanol plant would
be 5.2 million dollars. The expected profit before taxes with
a $4.00 disposal charge would be 8.4 hundred thousand dollars,
which is only a 16% return on the original investment. The
fixed manufacturing cost included a capital recovery factor
which calculated both the yearly cost necessary to recover the
original capital investment and an interest charge of 4% per
year; therefore all calculated percentage returns on investment
(23}
are in addition to this 4% charge. Vilbrandt and Dryden
indicate that a 45% return is expected before taxes on a high
risk venture such as this. An alternative would be the construc-
tion of only the ethanol plant by the chemical company, which
results in a return of 49%, since the capital investment does
not include the municipal built hydrolysis plant. This return
would require the municipality to absorb the necessary dumping
fee charge. It has been pointed out that a break-even point
with no dumping fee could be obtained by total plant operation
by the city. Therefore, construction of the total plant by the
community would be the most logical decision.
-------
-92-
Table 6-3
ETHANOL PRODUCTION BY FERMENTATION
COST ANALYSIS
FERMENTATION PLANT (5.4 million gal/year)
$1.7 million
RAW MATERIALS (excluding sugar)
(NH4)2S04
Steam
Process Water
Cooling Water
Electricity
225 Ibs.
750,000 Ibs.
150,000 gals.
630,000 gals.
1600 KWHR
LABOR
2 Laborers
1 Supervisor
MAINTENANCE (Material & Labor)
SUPPLIES
FRINGE BENEFITS
PAYROLL OVERHEAD
LABORATORY
PLANT OVERHEAD
FIXED CHARGES
Capital Return 4%, 2Qyears
Taxes
Insurance
$ 2.85
37.50
15.00
31.50
16.00
$102.85
$144.00
84.00
$228.00
$235.00
26.00
34.00
34.00
34.00
114.00
$350.00
98.00
47.00
$495.00
MANUFACTURING PRICE, NOT INCLUDING SUGAR
COST GAL., NOT INCLUDING SUGAR
SUGAR COST (Basis 500 ton Refuse, 50% Paper)
15.3 Ibs. sugar/gal.
$0.028/lb, No Dumping
$0.018/lb. at $4.00 dumping
TOTAL MANUFACTURING PRICE
NO DUMPING FEE
WITH $4.00 DUMPING FEE
CURRENT SELLING PRICE
$1302.8s/DAY
$0.087/GAL.
$0.43/gal.
$0.275/gal.
$0.518/gal,
$0.362/gal,
$0.52/gal.
-------
-93-
The capital investment for the total process would be 5.2
million dollars. Although this investment is very high, it
can be justified by the savings incurred by the zero disposal
cost. If the municipality were presently paying $4.30/ton for
disposal, a hypothetical profit of $720,000 would result. This
is approximately a 141 return on the investment which is well
above any other expected returns for a city run operation.
These calculations have been made on an assumption of a 50%
paper composition; if the actual plant paper were higher there
would be an additional profit due to the ethanol production.
The conclusion drawn from this preliminary cost analysis of
ethanol production is that such a process must be considered
as a possible user of hydrolysis sugars.
Chapter 7 describes other chemicals which may be possible
sources of demand for the hydrolysis plant product. With these
chemicals, as well as ethanol, a market study must be made to
determine whether expanding markets exist for the product. The
necessity of such a study is illustrated by the magnitude of
ethanol production capable of a 500 ton refuse plant, which
could produce approximately 5.5 million gallons per year. Al-
though the existing ethanol demand could absorb the production
from a small number of such plants, it could not withstand the
impact which would occur if every city of over 200,000 people
began producing ethanol.
Although refuse hydrolysis is not a panacea for the refuse
disposal problem, it does possess the potential to alleviate
the problem in many municipalities. Cities which produce
-------
•94-
refuse high in paper content and in large quantities have
various alternatives open to them: 1) own and operate a com-
bination hydrolysis-chemical plant, 2) produce sugar by hy-
drolysis with sales to a local chemical producer, and 3) entice
private industry into establishing a combination plant in the
community. The final decision will be determined by the
ability of the municipality to raise the necessary original
capital investment and the competitive nature of the product.
-------
-95-
7. GLUCOSE AS A POTENTIAL RAW MATERIAL
The production of ethanol by glucose fermentation has been
outlined in Section 6.4. As stated in that section, g]acose
can be used in most processes which use molasses as a raw mate-
rial. Thus if the glucose solution from the hydrolysis process
is sold at a lower price than molasses on a cost/lb sugar basis,
then it may have potential as a raw material for production of
chemicals other than ethanol. Such processes which use molasses
are described in the following sections. The pertinent infor-
mation which must be considered when evaluating such processes
are: competitive processes which do not require molasses
(synthetic), the existing market, yield of product per Ib. of
sugar, and the selling price of the product.
It is not known a priori what the effect of producing a
given product at a lower price will be on the market demand.
It is possible that additional markets or uses will be estab-
lished, thus increasing the demand for the product. Such in-
formation can only be obtained through a thorough market analysis,
It is essential that such information be obtained since it would
be foolish to attempt to produce a product for an already satu-
rated market.
The following information for the various chemicals is
taken from Reference (29) . Most market and cost information
are for the year 1963.
7.1 Monosodium Glutamate
Monosodium glutamate is a food flavoring produced by the
fermentation of glucose with "micrococcus glutamicus". It re-
quires a 15% glucose solution and produces approximately 0.27
-------
-96-
pounds of monosodium glutamate per Ib. of glucose. The produc-
tion is on the increase with approximately 31 million pounds
being produced in 1963. As of June 16, 1969 its selling price
was $0.47/lb. Although synthetic processes are being developed,
fermentation was the only economical means as of 1963.
7.2 Citric Acid
Citric acid is produced strictly by fermentation and is
primarily used in the beverage industry. A 15 to 20 percent
sugar solution is fermented by "aspergillus niger". Approxi-
mately 118 million pounds were produced in 1963 and sold at
$0.34/lb. One pound of citric acid is produced for approxi-
mately two pounds of sugar.
7.3 Butanol
Butanol was originally produced by fermentation of a 5%
sugar solution but has since been synthetically produced from
acetaldehyde. The yield is 20% by weight of original sugar.
Three hundred million pounds were produced in 19.63. The present
selling price is $0.13/lb. It is used as a solvent and a raw
material for butyl acetate, resins, and plasticizers.
7.4 Lactic Acid
Lactic acid is primarily used in the food industry and its
market is small — 5 million pounds/year in 1963. It is pro-
duced by fermentation of a 15% glucose solution with an 85%
yield, or by the hydrolysis of lactonittrite. The synthetic
process was just under development in 1963. Its current selling
price if $0.17/lb for a 50% technical, grade solution.
7.5 Sorbital and Oxalic Acid
Sorbital and oxalic acid can also be produced from a glucose
solution but they require a 50% and 60% solution. Such high
-------
-97-
concentrations would require extensive evaporation of the hy-
drolysis plant sugars. It is therefore unlikely that such
products could economically be produced from these sugars.
7.6 Conclusions
Of the above chemicals, monosodium glutamate and citric
acid have the greatest potential for economic production with
refuse produced sugars. Both can use low sugar concentrations,
approximately 15%, which could be economically produced by a
refuse hydrolysis plant. In addition, the only available means
of producing these chemicals is by fermentation, thus eliminating
competition from synthetic processes. The present market, as in
the case of all fermentation products, may not be large enough
to absorb the added production of many new plants, but this
cannot be unequivocally stated without further market research.
All such possible uses for sugar must be considered, since the
final solution will lie in establishing many such outlets for
the refuse sugars.
-------
-98-
8. RECOMMENDATIONS FOR FUTURE WORK
Although sufficient experimentation was conducted to estab-
lish the economic potential of the process, there is still addi-
tional experimentation which must be done on a small scale before
a pilot plant can be built. This work includes the study of the
effect of metal ions on the hydrolysis reaction, the ability to
react a slurry at low liquid to solid ratios, and the ferment-
ability of the sugars produced. This experimentation would lead
to a more accurate pilot plant design, thus increasing the proba-
bility of successfully producing a product at a marketable cost.
8.1 Metal Ion Effect on Hydrolysis
When metal ions were present, low sugar hydrolysis yields
occurred. Such ions were found to interfere with the 0-toluidine
sugar test resulting in false, low yields. Although it is be-
lieved that such ions do not interfere with the actual hydrol-
ysis, this hypothesis has not been proven. An analysis procedure,
using ethylene diamine tetra acetic acid to nullify the reac-
tivity of metal ions, is being developed. With such a procedure
a sample containing metal ions can be hydrolyzed and analyzed
for sugar content with and without addition of the EDTA solution.
If the metal ions only affect the sugar test, the hydrolyzed
solution containing EDTA should show higher yields than the
solution without EDTA. Moreover, if the iron containing hydro-
lyzed solution treated with EDTA gives yields comparable with
hydrolyzed solutions from paper without iron, then this would
be positive evidence that metal ions do not interfere with the
actual hydrolysis.
-------
-99-
Since almost any alloy metals used in constructing the
hydrolysis reactor will corrode to some degree, it is essential
to determine whether metal ions actually affect the hydrolysis
reaction. In addition, if metal ions do not affect the hydrol-
ysis, waste acids could be used as the acid catalysis. Waste
pickling aeids from the steal industry, which contain metal
ions, may then be used in the hydrolysis process, thereby
reducing the manufacturing cost of sugar.
8.2 Flow Reactor
It is essential that a model flow reactor system be built.
The enormous effect of the liquid to solid ratio on the manu-
facturing cost has been shown. A flow reactor system must be
tested in order to determine what the minimum feasible liquid
to solid ratio is for both the hydrolysis reaction and the
equipment used. A flow reactor is also needed to produce sugar
in large quantities from actual refuse. This would enable a reason-
able scale of sugar fermentation, and thus determine whether
refuse contains any materials which would contaminate fermen-
tation.
8.3 General Plant Considerations
The main manufacturing cost of the hydrolysis plant is the
oil necessary to heat the slurry to the reaction temperature.
If a system other than evaporation could be used to concentrate '
the sugars, a more efficient heat recovery system could be
designed. Reverse osmosis can be used to concentrate sugar
solutions, and future experimentation may prove that it is more
economic than evaporation.
Presently there is no internally proposed method for hy-
drolysis waste disposal. It may prove economic to burn the
residual waste and use the generated heat as an additional
-------
-100-
preheat source. Before this can be proposed, it is necessary
to determine the BTu potential of the waste. In addition to
hydrolysis waste, outlets for scrap metals should be determined
and possible uses for CAS04 explored.
8.4 Sugar as a Raw Material
Chapter 7 listed chemicals other than ethanol which can
be produced by sugar fermentation. Although the production of
ethanol seems to be economically feasible, a cost and market
analysis should also be performed on these other chemicals,
It may be found that there would be a larger economic advantage
in producing some other chemical. If such a feasibility study
indicates an existing economic potential for another chemical,
then an experimental program based on this chemical should be
established.
-------
-101-
9. SUMMARY AND CONCLUSIONS
A survey of existing municipal refuse disposal processes
showed that high temperature incineration is one of the important
conventional disposal methods, and the projected disposal cost of
such a plant can be around $3/ton. The main component of refuse
was found to be paper, 40 to 60%, and, as shown by Porteous, the
paper content of refuse is increasing.
Porteous' process for refuse disposal, which utilizes the
refuse paper content as a raw material for sugar production by
acid hydrolysis, was studied. Experimentation was conducted to
determine a kinetic model for paper hydrolysis which was then
compared with Saeman's model for wood cellulose hydrolysis.
Kl K2
Saeman's predicted A »• B >• C irreversible reaction model
was found to accurately describe the reaction, but the determined
rate constants were found to differ from those predicted by Saeman.
Yields, approximately 75% of those predicted by Porteous with
Saeman's kinetics, were obtained.
A hydrolysis plant design, using the experimentally
determined kinetics, was proposed. Modifications of Porteous'
original design included, a hydropulper separation and pretreat-
ment system adopted for use from those used in the waste paper
pulping industry, a recycle stream for the continuous flow
reactor system, and a multi-effect evaporation system for
concentrating the sugar solution.
A generalized computer program with variable refuse
tonnage, refuse composition, and operating conditions, was
-------
-102-
used to size the individual plant components and calculate
the total plant and manufacturing cost. A 1% acid solution
and a 230°C reaction temperature, yielding 51% cellulose-
sugar conversion, were found to be the most economical hydrol-
ysis conditions. The manufacturing cost of a 12% sugar
solution for various refuse compositions and plant capacities
was calculated and compared to the market price of molasses
sugars on a cost/lb. sugar basis. This comparison showed
that under the proposed operating conditions cities with
populations greater than 200,000 with refuse containing 50%
paper, and 100,000 people with refuse containing 60% paper,
could produce sugar by acid hydrolysis at a cost comparable
to the existing market price of molasses and at a zero refuse
disposal cost to the city. Moreover, a preliminary economic
study showed that ethanol could be economically produced by
fermentation of the sugars. Other potential uses for the
sugar solutions include the production of monosodium glutamate
and citric acid.
Porteous' economic study indicated that a 250 ton refuse
hydrolysis and ethanol fermentation plant could make $4.00/ton
profit with refuse containing 60% paper, and $0.77/ton profit
with refuse containing 40% paper. Although this study showed
that the hydrolysis process was not as economically attractive
as originally conceived, it still indicated that this method
of refuse disposal could eliminate the refuse disposal cost
for many U.S. cities.
-------
-103-
Before the final step of establishing a pilot plant is
taken, it was recommended that a small scale flow reactor be
built. Such a reactor would be able to confirm the pre-
dicted isothermal yields and produce sugars in large enough
quantities for fermentation experimentation.
-------
-104-
REFERENCES
1. Decisions of the Policy Planning Council, Refuse Disposal
Report for New York City, September 27, 1967
2. Bell, John M., "Characteristics of Municipal Refuse";
Proceedings National Conference On Solid Waste Research,
American Public Works Assoc., Special Report No. 29,
February 1914
3. "Garbage: Rosy New Future as a Raw Material", Chemical
Engineering, April 22, 1968.
4. "Conversion of Organic Solid Wastes Into Yeast", Report
prepared for the Solid Waste Program of Public Health
Service by Ionics, Inc., Mass., 1968
5. Porteous, A., "Towards a Profitable Means of Municipal
Refuse Disposal", ASME Pub. 67-WA/PID-2
6. Meiler & Scholler, dissertations, Techn 'Hochshule,
Munich, 1923
7. Saeman, J., "Kinetics of Wood Saccharification", I & EC,
January 1945
8. Browning, B. , The Chemistry of Wood , Interscience
Publishers, New York, 1963
9. Saeman, Buhl & Harris, "Quantitative Saccharification of
Wood and Cellulose", I & EC, Vol. 17 #1, January 1945
10. Wilde & Beightler, Foundations of Optimization,
Prentice Hall, 1967
11. Browning, Methods of Wood Chemistry, Interscience Pub-
lishers, New York, 1967
12. Casey, J., Pulp and Paper, Vol. 1, Interscience Pub-
lishers, New York, 1966
13. Levenspiel, Chemical Reactor Engineering, Wiley & Sons,
New York, 1962
14. Daily & Bugliarello, "Basic Data for Dilute Fiber Suspen-
sions in Uniform Flow with Shear", Tappi, July 1961,
Vol. 44 #7
15. Daily & Bugliarello, "Rheological Models and Laminar Shear
Flow of Fiber Suspensions", Tappi, Dec. 1961, Vol. 44, #12
16. Stepanoff, Pumps and Blowers, Two-Phase Flow, Wiley & Sons,
1965
-------
-105-
17. ror.tana and Greene, Corrosion Engineering, Xc-Graw-Hill,
New York, 1967
13. G::ulich, G., "Improved Methods of Spent Sulfite Liquor
Disposal", D.E. Thesis, Thayer School of Engineering,
Hanover, N . K . , June 1969
19. 1-orry , Chemical Engineering Handbook, Fourth Edition,
McGraw-Hill, New York, N.Y., 1953
:C. "Evaporation", Chor-i^njl Enrrincsring, December S, 1963
21. Kern, Process Heat Transfer, McGraw-Hill, 1950
22. Guthrie, "Capital Cost Estimating", Chemical Engineering,
March 24, IS69
23. Vilbrandt and Dryden, Chemical Engineering Plant Design,
McGraw Hill, New York, N.Y., 1959
24. Peters and Timmerhaus, Plant Design and Economics for
Chemical Engineering, McGraw Hill, New York, 1968
25. Aris, Chemical Engineering Cost Estimation , McGraw Hill,
New York, 1955
26. Saeman, U.S. Forest Research Labs, Madison, Wisconsin,
Personal Communication, June 1969
27. "Commodity Year 1968 Book", Commodity Research Bureau Inc.,
140 Broadway, New York, N.Y.
28. Harris, "Fermentation of Douglas Fir Hydrolysis by S.
Cerevisiae", I & EC, September 1946
29. Faith, Keyes, Clark, Industrial Chemicals, Wiley & Sons,
New York, 1965
30. Haselbaith, Chemical Engineering , 74(25):214, 1967
-------
-------
-107-
Apper.dix I
QUANTITATIVE SACCHARIFICATION
Quantitative saccharif ica-cion is an analytical technique
which hydrolyzes cellulose to glucose with a ininiir.um amount of
glucose decomposition. Reference (9) from which the procedure
was taken, indicates that chemical yields of greater than S5%
can be achieved.
Procedure:
A ground cellulosic sample of 0.35 grams is weighed and
mixed with 5 ml. of 72% sulfuric acid that has been cooled to
15°C. This mixture is placed in a water bath at 30°C. This
temperature is maintained for 45 minutes and during this time
the sample is stirred at 5 to 10 minute intervals. After the
required time, the mixture is diluted in an Erlenmeyer flash
with 140 ml. of water. The diluted solution is autoclaved
for one hour at 15 psi. This yield will be the maximum poten-
tial aldohexose yield of the sample weight aldohexose x 100>
weight sample
For a pure cellulose sample the potential glucose yield should
be
-------
-108- •
Appendix II
SUGAR TEST
(EXPERIMENTAL WORK BY WILLIAM ELLSWORTH, SENIOR CHEMIST)
Two methods of measuring the sugar content of the hydrolysis
product were compared: spectre-photometry and ferricyanide oxi- ,
dation. The latter test measures the total reducing power or
the solution including partially decomposed sugars. The first .
technique measures only the aldohexose sugars which are the
sugars of interest since they are fermentable. All glucose
yields found in this report have been obtained by use of the
0-toluidine colorimetric test.
It was found that metal ions cause a false reading to occur
with the colorimetric test, but that decomposed sugars have no
effect on the test results. A summary of the test procedures
and the experimental study done on the 0-toluidine test follows.
A. Spectrophotometry:
Principle: Aldohexoses react with 0-toluidine reagent to form
a green colored complex. The solution is then put in a spectro-
photoineter. The color follows Beers Law, i.e. color is propor-
tional to concentration of sugar up ro 1000 mg%, where mg% equals
number of milligrams per hundred milliliters of solution.
Procedure:
(1) 0.1 ml of sugar solution is mixed with 6 ml of reagent.
(2) Mixture placed for 8 minutes in a boiling water bath,
then cooled to room temperature.
(3) Read in spectrophotometer at 630 mu.
(4) % sugar found from calibration curve. Calibration is
done using standard glucose solutions.
Reagent: 60 ml 0-toluidine + 1.5 ml thiourea.. Dilute
this to 1000 ml with glacial acetic acid.
-------
-109-
Rcf erer.ccs :
This method is a modification of several in the literature:
(1) Nature, Vol. 183, p. 108, 1959
(2) Clin. Chera. Acta, Vol. 7, p. 140, 1962
(3) Clin. Chem. 8, p. 215, 1962
^* Ox^^tj-on with Ferricyanide:
I & 'EcT:~AnaTytica 1 Edition/ Vol. 9 (1937) , p. 228
In this test, sugar is oxidized using ferricyanide. Excess
ferricyanide is titrated with eerie sulfate.
Procedure:
(1) 5 ml of sugar solution (not containing more than 3.5 mg
sugar) + 5 cc of alkaline ferricyanide. Solution is
placed in boiling water bath for 5 minutes/ then cooled.
(2) Add 5 ml of 5% ^304 to above solution. Titrate against
0.01% eerie sulfate using setopaline-C indicator.
Comparison of Results
Method A estimates only the glucose, whereas Method B esti-
mates all kinds of sugars — pentose, etc.
Comparison of results:
% Sugar (gms sugar/100 grams paper)
Test A Test B
Method A 28.81 29.83
Method B 31.55 31.87
C. Analysis of Spectrophotoinetric Test Reliability
1. Determination of Effect of Ferrous Ion
This test was run by taking a standard glucose solution
and adding ferrous sulfate to it to give various concentrations
of iron.
-------
-110-
Run
1) Standard Solution
2) Solution + 0.01% FeS04
3) Solution + 0.1% FeSO.
4) Solution +0.5% FeS04
5) Solution +1.0% FeS04
Experimental Percent
Result Error
120 mg %
88
46
44
39
26.6
61.6
63.4
67.5
Ferrous sulfate was chosen because iron makes up about 50%
of Carpenter 20 CB3 and is the most reactive nmetal in it. •
The results show that even a very small amount of ferrous
ion causes a low reading.
2. Determination of Effect of Carpenter 20 CB3 on Tests.
A 1.6g (1/2" of 1/4" tubing) piece of CB3-plus 20 ml of
0.5% H2S04 was put into the constant temperature bath at 268°C.
It was left for 30 minutes and then the 15 ml of it that hadn't
boiled out of the glass liner was diluted to 41 ml.
A 102 mg% standard solution was used in this and the rest
of the experiments.
Run Experimental Expected Error
1) Standard Solution (SS)
2) 2 ml of SS + 2 ml H20
3) 2 ml of SS + 2 ml test
4) 5 ml of SS + 2 ml H20
5) 5 ml of SS + 2 ml test
6) 10 ml of SS + 2 ml H20
7) 10 ml of SS + 2 ml test
A similar test gave almost the same results.
106 mg %
49
42
72
69
85
79
102 mg %
51
51
73
73
85
85
3.9%
3.9%
17.6%
1.4%
5.5%
0.0%
7.1%
-------
-111-
This experiment definitely indicates that, at high tem-
peratures over an extended period of time, the 0.5% H2S04 dis-
solves enough iron to affect the glucose test. The fact that
when the test solution was added to the standard glucose solu-
tion the glucose concentration indicated was always lower than
when just water was used, is consistent with experiment #1.
The fact that the error goes down with increasing dilution of
the test solution is also in agreement.
3. Paper Hydrolysis with No Metal Contact to determine the
effect of decomposed sugars in the glucose test.
The paper hydrolysis was carried out using 20 ml of 0.5%
H2S04 + 0.2g of paper in a glass lined reactor. It was put
into the bath at 265°C for 15 minutes, then diluted to 50 ml
with H20.
Run Experimental Expected Error
1) Hydrolysis product (HP) 51 mg %
2) 5 ml of HP + 5 ml SS 80 77 mg% 3.9%
3} 5 ml of HP + 10 ml SS 85 84 1.2%
4) 5 ml of HP + 25 ml SS 95 94 1.1%
5) 5 ml of HP + 50 ml SS 93 97 4.1%
6) 5 ml of HP + 75 ml SS 101 99 2.0%
7) 2 ml of HP + 2 ml H90 24 25.5 0.6%
^
8) 2 ml of HP + 5 ml H20 13 14.6 1.1%
9) 2 ml of HP + 2 ml SS 79 77 2.6%
These results indicate some error, but it is probably
within experimental error. See the next experiment.
-------
-U2- . ':
4. Test to determine the overall accuracy of the dilution
method of c necking the glucose test.
Run, Experimental Expected Error
1) 2 ml of SS + 2 ml H20 45 mg % 51 mg % 11.82
2) 5 ml of SS + 5 ml H20 50 51 2.0%
3) 5 ml of SS + 10 ml H20 34 34 0.0%
4) SS alone 101 102 i.Ofc
5) SS alone 106 102 3.9%
See also runs 2, 4, and 6 of experiment #2.
The average percent error of these eight runs is 3.0%.
There is some dilution error in these tests, and also some error
in the glucose determinations.
5. Conclusion
The test is accurate to 3 - 4 percent, and is more consistent
if all the tests are performed simultaneously than if they are
done separately. Care must be taken that no ferrous ion is in
the solution.
From other references it was learned that the amount of time
that the samples are heated is crucial, and that any metal ion
affects the tests.
It has also been determined that allowing the sample to sit
in the test tube with the 0-toluidine solution for a short time
before heating it does not affect the test appreciably.
-------
Appendix III
NONISOTHERMAL HYDROLYSIS RESULTS
TEMPERATURE TIME HISTORY AND YIELD
0.2% Acid Runs
0.5 gram Samples
20 ml Liquid
Observations: Base tenp. = 100°C. The body of
table reads temp, in °C.
IQuenched 5.4
7.2
11
Run No . -
Time
/M- r\ \
j V •»-»• • /
i
: o
1
! 2
1 3
: A
\ 5
! 6
i 7
i 8
! 9
: 10
: 11
i 12
i 13
; 14
.Yield
V7/W %
paper
1 2
'
100 ' 100
120 119
138 139
153 '. 154
167 • 167
179 179
183 ' 189
i 200
' 202
1
,
;
j
|
i
i
i
1.4 ; 5.9
'
! 3
i
I
j
[
! 100
' 120
1 139
i 155
I 169
. 180
! 191
1 201
: 209
! 211
>
I
1
!
1
1
1
' 7.8
1
4
100
122
143
157
171
183
195
204
214
222
8.7
5 6
t
I
I
100 100
121 121
140 141
156 156
171 170
183 182
194 192
204 202
212 210
220 217
228 : 224
233 230
235
237
i i
,
12.8 ! 11.8
j
12.4!
Temp.
Observation:
6 =
Temp, to time curve for run was:
A + BT + CT2 + DT3 + ET4'
A = 99.9254
3 = 23.0599
C = -1.78918
D = 0.105227
E = -3.11787 E-3
-113-
-------
w
H
K
O
£-)
O'i
H
H
H
c1
•H
HH
O
O
CN
g
(0
01
CN
•
O
•H
O
in
o
iH
O
H
a\
CO
r^
V£>
m
M1
ro
CN
iH
• •
*
O
K
C
D
«
ocNroinc^cNforor-ioo^critN
OfM-^Ln^COCTiOiHrHCNCNfO
+
oro'sDfo^
O CN -5T VD r- 00 ^.O H rH (N 00 CTi
O H CTl in O O
o CN ro in r- oo
H H H H H H
+
OCNOt^HCMCSlHO^VDCN^H
ocN^inr^ooo^ooiHCNCNcn
oo
00
•
•*
in
•
H
CN
•
0
H
r~
•
00
CO
l£
Time | 4
Quenched
w •
cd
.p
•H
4-1
-H
rO •
•H £
g
0
c
>i £
iH
H
QJ
T3
i-4
A C
•P 0
H
3
O •
4H
"3 .
0) E
>
cd
CP
(C
-P
«
t3
o
XJ*
(rt P-<
W
Q)
X! +
-P
CO
5-1 EH
0 Q
4-4
+
4-)
•HCN
IH EH
u
0)
> +
5-1
3 E-i
U ffl
Q) +
M
IT! <
3
a1 u
w
a>
-P
yj
to
Q)
i-5
C
•H
&
a
•H
0)
•rl
EH
C
•H
04
§
EH
II
CD
CO
in
*sj*
CD
•
CT\
CTi
CTl
CO ^
CO <7l
•<* CN
vo co
• •
CO rH
CN 1
Hit
H
« U
CN
1
W
rH
cn
f*^
CO
o>
«
00
II
11
Q
ro
i
W
CO
CN
r~-
^D
O
•
CN
I
W
i
WJ
CO
rH
U
•
D1
cn
>w
O
g
^
O
5-1
5-1
W
-114-
-------
-115-
TEMPERATURE TIME HISTORY & YIELDS
1% Acid Runs
0.5 gram Samples
20 ml of Liquid
Run No,
Time , ;
(Min.)
0 100 100
1 118 119
2 127 137
3 152 152
4 165 165
5 175 177
6 183
7
o
9
10
11 :
13
Yield
w/w paper | 9'97 u-85
Time j
Quenched 4'8 5'6
100
119
127
153
165
177
186
191
19.70
6.6
100
119
137
152
165
177
187
197
202
i
!
; 30.60
7.6
100
119
137
152
165
177
188
197
203
205
! 31.30
8.2
i
i
100
119
137
152
165
177
187
197
203
210
31.56
8.6
i
100 i
119
137
152
165
177
188
197
203
210
219
220
(
19.08
10.4
100 i
119 i
137 ]
152 !
165 j
177 j
(
187 !
197 !
204 1
211
219
224 j
227 i
i
i
i
3.0
I
|
11.6 |
!
CJ
-P
(U
A
£
o
EH
Observation: Temp, to time curve for run was:
0 = A + BT + CT2 + DT3 + ET4
A = 99.6517
B = 21.3828
C = -1.52206
D = 7.43993E-2
E = -0.001729
-------
Appendix IV
COMPUTER PROGRAMS
A. HOOKE-JEEVES SEARCH AND RUNGE-KUTTA NUMERICAL
INTEGRATION
Used for calculating parameters of Arrhenius' equations
which best fit the data of the non-isothermal analysis.
The Hooke-Jeeves search was originally written for two
dimensions by Steve Smith, Tuck-Thayer, 1967. It was modi-
fied for four dimensions and incorporated with a 4th order
Runge-Kutta integration routine written by Professor Converse,
Thayer School of Engineering. A program flow sheet and
internal program documentation are given to help explain its
operation.
-116-
-------
-117-
HOOKE JEEVES SEARCH
INPUT
Yields & Times
To Runge Kutta ';
Subroutine
INITIALIZATION
Set X' 's i=l to 4 (Arrhenius Parameters)
Set D-L'S i=l to 4 (Step Sizes) ••
Set L = Search Convergence Criterion '
Let Bi = Xi
r" "
! Calculate Sum Sq Error With
i X,- = B, + D.
= B - D±
i Save Suboptiumum X.!s
MINIMUM OF
FUNCTION
-------
-118-
From Hooke-Jeeves Search
It'*
RUNGE KUTTA SUBROUTINE
T~
INITIALIZATION
Set X(l) = Cellulose content
X(2) = Sugar content
t = Time of initial conditions
tn = Time of final conditions
Fourth Order
RUNGE KUTTA
EQUATIONS
Temp T = f(t)
= g1 (XX, X2,T)
= g"(X3,X4,T)
X(2) = h"(K1,K2,t)
iYES
Calculate:
Sum of Squares
! Error of fit
To Hooke-Jeeves Search :
-------
/?
-11,9-,
1 ?."" This program calculates the 4 parameters of 2 Arrhenius
2 RFM equations which best fit the curve sf a non-isothermal
? RE.M kinetic analvsis.
50 PRINT' "INPUT NUMBER OF TIME POINTS*'
52 i "1 T9=7
60 PRINT "INPUT TIME (WIN), YIELD (FRACT)
«-2 FOR N = 1 TO T9
6* RF/'D T(N),Y(N)
65 Nr.XT N
C6 DAI A 5.6, .11,6.6, .1995,7.6,. 306, 8. 2, .313, 8. 6,. 3156
66 DATA 10.2, . 190F.11.6, .03
100 REM THIS PROGRAM DOES A KOOKE-JSEVES PATTERN SEARCH AS
110 ' DESCRIBED IN SECTION 7-08 OF WILDE AND BEIGHTLER ,
120 ' "FOUNDATIONS OF OPTIMIZATION," PP. 307-310. PRESENTLY THE
13G ' PROGRAM IS LIMITED TO FOUR INDEPENDENT VARIELES, XI. X2 ,X3 , X4
150 ' STATEMENT 250 LET Xl= INITIAL VALUE OF 1ST PRE-EXPOTENTIAL
160 ' STATEMENT 260 LET X2= INITIAL VALUE OF 1ST ACT ENERGY
170 ' STATEMENT 270 LET X3=INITIAL VALUE OF 2ND PRE-EXPOTENTIAL
160 ' STATEMENT 280 LET X4 = INITIAL VALUE OF ' 2ND ACT ENERGY
190 ' STATEMENT 290 LET Dl = INITIAL VALUE OF XI STEP SIZE
200 ' STATEMENT 300 LET D2 = SAME X2
210 ' STATEMENT 310 LET D3 = SAME X3
120 ' STATEMENT 320 L?;T D4 = SAME X4
230 ' SUE ROUTINE 1490 RUNGE-KUTTA INTEGRATION
240 ' *-*-*r##*******#*###£i^j) OF DESCRIPTION***""'***'***'**********
250 LET Xl= 600000 'ACTUALLY 6E19 SEE LINE 2334
260 LET X2=45000
270 LET X3=24000 'ACTUALLY 2.4E14 SEE LIME 2335
280 LET X4=32800
290 LET D1=1E5' GIVE LARGE SEARCH MARGIN
300 LET D2=1000' KEEP IN FEASIBLE RANGE
310 LET D3=1E4' GIVE LARGE SEARCH MARGIN
320 LET D4=1000' KEEP IN FEASIBLE RANGE
330 LET Pl=0
340 LET P2=0
350 L?T P3=0
360 LET P4=0
370 DIM P(300,4)
SCO LET P(0,1)=X1
390 LET P( 0,2)^X2
400 LET P(0, 3)=X3
410 LFT P(0,4)=X4
420 GO SUE 1490
430 LET FO=F
450 FOR 1= 1 TO 290
460 GO SUB 1490
^70 IF F>FO THEN 500
480 LET FO=F
-------
-120-
SE£R (continued)
490 GO TO 540
SCO LET P1=X1
510 LEI P2=X2
520 LET P3=X3
530 LET P4=X4
5-0 LET XO = XI
550 1,?1 n*Xl+Dl
560 GO SUE 1490
570 IF F>FO THEN 600
5 SO LET FO = F
590 GO TO 670
600 LET Xl=XO
610 LET X1=X1-D1
620 GO SUB 1490
630 IF F> FO THEN 660
b40 LET F0= F
650 GO TO 670
660 LEI XI = XO
670 LET XO=X2
6SO LET X2= X2+D2
690 GO SUB 1490
700 IF F> FO THEN 730
710 LET FO = F
720 GO TO 800
730 LET X2= XO
740 LET X2= X2-D2
750 GO SUE 1490
760 IF F>FO THEN 790
770 LET FO=F
760 GO TO 800
790 LET X2= XO
£00 LET X0= X3
£10 LET X3= X3+D3
&20 GO SUB 1490
£30 IF F>FO THEN 860
640 LEI F=FO
650 GO TO 930
£60 LET X3=XO
£70 LET X3= X3-D3
f£0 GC SUB 1490
690 IF F>FO THEN 920
900 IET F0= F
910 GO TO 930
920 LET X3= XO
930 LEI X0= X4
940 LET X4= X4+D4
950 GO SUE 1490
960 IF F> FO THEN 990
970 LET FO=F
960 GO TO 1190
-------
-121-
SFr R (continued)
990 LET X4 = XO
1CCO LET X4 = X4-D4
1010 CO SUB 1^90
1020 IF F>FC THEN 1050
1030 LET FO=F
1C'-0 GO TO 1190
1050 LET X4=XO
1C6C ' ***********#EKD OF ABOVE DESCRIBED ROUTINE***************
1070 ' THE FOLLOWING TWO TESTS DETERMINE IF THE NEW BASE POINT OR
10F-0 ' ANY OF ITS SURROUNDING PERTURBATIONS ARE BETTER THAN THE LAST
1090 ' BASE POINT. IF NOT THEN A NEW BASE POINT MUST BE CALCU-
1100 ' LATSD USING A SMALLER STEP SIZE.
1110 IF P1<>X1 THEN 1190
1120 IF P2<>X2 THEN 1190
1130 IF P3<> X3 THEN 1190
1140 IF P4<> X4 THEN 1190
1150 LET Xl= F(l-l.l)
1160 LET X2 = P(I-1,2)
1170 LEI X3= P(1-1,3)
1180 LET X4 = P(I-1,4)
1190 LET P(I,1)=X1
1200 LET P(I,2)=X2
1210 LET P(I,3)=X3
1220 LET P(I,4)=X4
1 2?0 ' THE FOLLOWING FOUR STATEMENTS CALCULATE AND ASSIGN
1240 ' VALUES TO THE NEW BASE POINT.
1 250 FOR J= 1 TO 4
1260 LET P(I+1,J)=2*P(I,J)-P(I-1,J)
1270 NEXT J
1280 LET Xl= P(I+1,1)
1290 LET X2 = P( 1+1,2)
1300 LET X3=P(1+1,3)
1310 LET X4 = P( 1+1,4 )
1320 LET Z9=D1+D2+D3+D4
1321 IF Z9<100 THEN 1430
1330 ' THE NEXT FOUR STATMENTS REDUCE TEE STEP SIZE IF THE PROCEEDING
1340 ' SEARCH DID NOT FIND A NEW MINIMUM.
1350 FOR J= 1 TO 4
1360 IF P(I+1,J )<:>P(I,J ) THEN 1420
1370 TEXT J
1380 LET Dl=Dl/2
1390 LET D2=D2/2
1400 LET D4= D4/2
1410 LET D3= D3/2
1420 NEXT I
1430 PRINT "MINIMUM OF FUNCTION IS"FO
14-^0 PRINT "COORDINATES ARE"xi;x2^x3;X4
1450 PRINT' " FINAL STEP SIZES ARE D1;D2;D3;D4
1452 FOR J= 1 TO T9
1453 PRINT SQR(E(J)),Y(J)
-------
/.
-122-
£E£R (continued)
145-i NEXT J
1460 PRINT
1^70 PRINT"IT";I
14i?C GOTO 2500
1490 ' THIS IS START OF RUNGE-K.UTTA
1491 IF XI<0 THEN 1530' PUT RESTRAINT ON SEARCH VARIABLES
1193 IF X2<0 THEN 1530
14°4 I? X3 Tl THEN 2360 ' END OF INTEGRATION
2311 FOR J=l TO T9' PICK OUT ERROR OF EACH POINT
2212 IF A:-S(T-T(J))>.OI THEN 2317
2313 LKT E(J) = (X(2)-Y(J))~2
2314 IF E(J)>500 THEN 2341
2317 NEXT J
2320 C-CTO 1970
-------
-123-
fr,/>R (continued)
7330' C WILL BE ABSOLUTE TEMP -TIME FUNCTION
2331 LET C=273+99 .6517 + 21.3828*T-1.52206*T's2+7.4399E-2*r~3
2332 LET C=C-1 ,729E-3*T~4
2333' ^ RRHE NIUS FUNCTIONS FOLLOW
2331 LET Kl= XI *!El4*EXP(-X2/(1 .98*C ) )
2336 LET K2= X3*lElO*EXP(-X4/(1.98*C))
2337' CONSTRAINS ON K1.K2
2338 IF Kl>20 THEN 2341
23^0 IF K2<20 TEEN 2345
2341 LET F=F+10
23^2 GOTO 2400
23-M ' DIFFERENTIAL EQUATIONS FOLLOW
2345 LET G (1 )=-Kl*X (1)
23^6 LET G(2)=-K2*X(2)+K1*X(1)
2350 RETURN
2355' CALCULATE SUM SQUARE ERROR
2360 LET F=0
2370 FOR J= 1 TO T9
2380 LET F=F+E(J)
2381 ''EXT J
2382 LFT J5=J5+1
2383 IF J5=100 THEN 2395
2384 IF J5=160 THEN 2395
2385 IF J5=200 THFN 2395
2386 IF J5=250 THEN 2395
2387 IF J5=350 THEN 2395
2388 IF TIM>320 TEEN 2395
2389 GOTO 2400
2395 PRINT F.XI,X2,X3,X4
2396 FOR J=l TO T9
2397 PRINT' SCR (E (J ) ) , Y (J )
2398 NEXT J
2399 PRINT "l",I
2400 IF J=100 THEN 2402
2401 GOTO 2405
2402 PRINT J
2405 RETURN
2500 El£>
-------
-124-
B. HYDROLYSIS PLANT SIMULATION PROGRAM
by R. Fagan
The hydrolysis plant simulation program can be used to
calculate the manufacturing cost of sugar for any city with
local refuse composition and local operational cost. Internal
documentation is given to aid the reader in following the pro-
gram. The inputs to the system are either external or internal,
External inputs are requested by the program at run time, and
internal inputs must be changed before run time.
Line Number
External Inputs
230 - 270 Refuse tonnage and composition
270 - 500 Plant operational variables
A. yield
B. reactor residence time
C. acid concentration & temperature
D. solid to liquid ratio
E. recycle ratio
3790 Dumping credit
3810 Waste disposal charge
3830 Interest rate and years of operation
1820-3010
3332
3370-1442
3570-3780
850-1590
3070-3250
3450-3544
3900-4165
Internal Inputs
Equipment cost parameters
Equipment cost update %
Factors for calculating total fixed capital
cost from installed equipment cost
Manufacturing raw material cost and indirect
manufacturing cost factors
Outputs
Material balance
Equipment size and cost
Fixed capital investment
Manufacturing cost analysis
-------
-125-
By inserting these variables, an economic feasibility
study can easily be performed on any city which might wish
to use acid hydrolysis to dispose of its refuse.
-------
-126-
COST3
100 DIM S(30),P(30),I(30)
110 DTK K(30)
120 DIMT(30)
130 LET A$(I)="CELLU^OSE"
140 LET A$(2 ) = "WATER '
150 LET A$(3)="ACID"
160 LET A$(4)= 'SUGAR
170 LET A$(5)='*D SUGAR*'
180 LET A$(6)=*S INERTS"
190 LET A$(7)="L INERTS"
200 LET A$(8)=' LIME
210 LET A$(9)="SEPARABLES"
220 REM *******MATERIAL BALANCE**************
230 PR INT" INPUT TOTAL TONS WASTE*'
240 INPUT S' TOTAL TONS/DAY WASTE
250 let jl=s
260 PRINT "INPUT FRACTION PAPER.GARBAGE,SEPARABLES IN WASTE
265 INPUT1 F1,F4,F3
270 PRINT"INPUT FRATION CELLULOSE IN PAPER
272 INPUT F2 ' ;
273 LET F5=1-F2
280 PR I NT" INPUT' TIME TO MAX YIELD IN MIN"
290 INPUT' T5
3co FRINT"INPUT CELLULOSE FRACTION CONVERTED AND SUGAR PRODUCED'
310 INPUT Y1,Y2
311 LET Y3=1-Y1
320 PRINT"INPUT FRACT ACID,SOLID TO LIQUID,REACTION TEMP CENT
330 INPUT C2,C1,T
340 LET C2=C2*100/93'CORRECT FOR 93$ ACID
360 LET P= F1*S
370 LET C= F2*P' CELL
360 LEI M=F3*S' SEP
390 LET G= F4*-S' GAR
400 LEI1 K= F5*P' CLAYS
410 LET Sl= S-M' SOLIDS TO REACTOR
420 LET 1= ,5*( G+K)' TOTAL INERTS TO SYSTEM
430 LET L= S1/C1' TONS/DAY
440 LET X= L-S1' WATER ADDED AFTER HYDRO
450 LET Tl= ((L+S1)*T-2*S1*20)/X' REQUIRED TEMP
460 LET Al= C2*L' TONS/DAY ACID
470 REM****************RECYCLE PROBLEM*********************
480 PRINT"INITIAL FRACTION INERTS" ;i^c
490 PR INT"INPUT II,INERT RATIO FINAL*
500 INPUT II
510 LET R4= I1/Y3' RATIO INERTS TO CELL OUT
520 LET B= I/R4' TOTAL CELL BLEED
530 LET R= (B-Y3*C)/(Y3-1)' RECYCLED CELL
540 LET Rl= C+R' TOTAL CELL IN REACTOR
542 REM
543 REM
-------
-127*
55C Li: RZ = Y3---R1' TOTAL CELL LEFT
5f,0 LEI f<3= l.r-tY2*Rl' TOTAL WEIGHT SUGAR FORMED
570 LET S3= I+C-s-R+Il*(C+R)' TOTAL SOLIDS IN REACTOR
530 LIT v;= S3/C1' TOTAL WATER RSQ
soo LET ;-.3 = c2•--;•:' TOTAL ACID REQ
£00 LET '/1=',;-L-R-R4---R 'ADDITIONAL WATER TO RECYCLE
61 G LEI F=VS*-;•:+. l^(D-rGl)
620 LET A<1=A3-A1-A3/ (W-F )*(R+R4*R ) 'ADDED ACID
630 REM Gl= GLUCOSE OUT' WITH WATER
640 LET GI= R3-R3/(W-F)*(R2-*-Il*Rl)
650 LET D= (I-Y3-Y2)*Rr*l.l' TOTAL DECOMPOSED SUGAR IN WASTE
rfoO LET Dl= D-D/( VJ-F ) * (R2+I1*R1)
670 LET D2= !-!/(W-?)*(R2+I1*R1)
CSO LET V?.'.=W-.1*(D+G1)
690 LEI W5-W4-V9*V-7
7CO LET T2 = ( (V/l-f2---(R"R4+R))«-T-2^(K*-R4+R)*177)/V7l
705 LET 12 = (Ti-X+T2-V7l)/W'TEMP OF IX WATER CENT
710 LET A5=A3-A4-Al' ACID OUT WITH SOLIDS
720 LET £7= Y2-C-1.1' SUGAR PRODUCED IF NO RECYCLE
730 LET T.>;9=W5-R2-R1*II
7<0 LEI" N6= W9-:-(A3-A5)/98-*-13' WATER OUT OF NEUT
750 LET N7=N6-(P.3-A5)/9G*13S
760 LS'I L9=(A3-A5)/e8* 100 'LIME NEEDED
770 LET G2=(:.3-A5)/98*136*(G1/(N6+L9))
7£0 LET G3= G1-G2 ' SUGAR TONS/DAY OUT OF PLANT
790 LET S9=3-Kl-'--'B-rM/2'TONS PER DAY OF SOLID WASTE
600 PRI_\T"DO YOU WANT MATERIAL FLOW BALANCE YES OR NO"
sic INPUT y.$
820 IF MS="NO" THEN 1580
830 PRI:;T"SEPARATION SYSTEM"
640 PRINT
650 PRINT"SEPARATOR","IN","OUT T/D","IN","OUT T/HR"
660 PRINT "WASTE",S,S-M,S/24,(S-M)/24
S70 Fr,I\\T A$(2),S/.02, ( £-K )/. 5. ( S )/. 02/24 ,( S-M )/. 5/24
e:c PRINT
c 90 PRINT "•"•^-•*«^i:-*****•»•*******«-************#*#**«•*#•«•*#****#"
900 PAINT
910 PRINT"REACTOR","IN","our T/D","IN","our T/HR"
920 PRINT ;.$(!),Rl,Y3*RlpRl/24,Y3*Rl/24
930 PRIKT Ac(2)„W,W4,W/24,W4/24
940 PRINT At(3)sA3,A3,A3/24,A3/24
950 PRINT Ac(4),Z9,R3,Z9,R3/24
9cO PRI>?i A$(5},Z9,D,Z9,D/24
970 PRINT .\S(6),IlttRl,Il*Rl,Il*Rl/24.Il*Rl/24
960 PRib:T A$( 7) ,1,1,1/24,1/24
990 PRINT
IO-;C ?RI I\7I
"; 050 PRi\T"';i'**********************************************
1060 PRINT
1070 PRINT"FLASK","IN T/D","OUT T/D","IN T/HR","our T/HR"
-------OCR error (C:\Conversion\JobRoot\Region 5\00000BC1\tiff\2000ZA4Q.tif): Unspecified error
-------
-129-
COST 3 (continued)
1590 LET El= G3/74
j. c 00 R^,''i '*"•""*'*'*• w..-•*>--A-w.\--X---»
1G10 LET Fl= U*200G/24/62.4'FT~3/!-IR OF FLOW TO FLASH
1020 LSI E2= F1/.134' GAL/HR TO FLASH
1C30 LET S5= S4/.134' GAL/HR THROUGH REACTOR
IvVO LET C5= (33-l)/24' IONS SOLID/RR TO 1ST CENT
1650 LET C&= V.'5/24*iOOO*2/70'FT'"'3/liR TO CENT OF WATER
1&60 LET Kl=W9*2COO/24/&2.4'FT'VKR TO CENT1
1C70 LET N2 = N1/.134' GAL/HR TO CENT 2
ItSO LET N3= (A3-A5)/9S*100/24' TCNS/KR OF LIME TO CENT
1690 LET N8 = N7*2000/63/24' FT"3 /HR TO EVAP
1700 LET N9=NS/.134'GAL/MIN
1710 LET S3 = SI
1720 LET Pl= S3/. 02*1000*2/24/62.4' FT~3 TO PUMP 1 TO PULPER
1730 LET S2= El/(N7*2000/24)' FRACTION SUGAR IN LIQUID
174-0 LET ?2= Wl*2000/62.4/24'FT~3/KR TO 1ST MIX
1750 LET P3= L*2000/62.4/24' FT~3/KR TO FRESH REFUSE
1"60 LET P4=(A1+A4)*2000/24/(62.4*1.8)'FT~3/HR OF ACID
1770 LET P5= (s3/.C2-S3/.5)*1000*2/24/62.4'FT"3/HR PULP RECYCLE
17£0 LET P6= P1-P5' FRESH WATER TO PULPER
1730 LET Hl= P2+P3' THROUGH HEATER AND COND
1SOCPRINT
1S10 PRINT
1620 R'"".V-U-WW-X--X--X-V.-'---^-EQUIPMENT COST*****************
1330 1<2M STORAGE HOPPER 3 DAY CAPACITY
1S40 REM ASSUME REFUSE IS 500 LB/YD~'3
1£50 LET G(l}= 3*S*2000/500*9'FT~3
1660 LET ?(1)=S(1}".9
1370 LET I(1)=1.1*P(1)
IcoO REM CONVEYOR 18 INCH PULLEY MAGNETIC TYPE
1S90 LSI S(2)=100'FT
1900 LEI P(2)= S(2)".65*450*3/250+1000
1910 LEI 1(2)= P(2)*1.62
1920 LET H(2)=l
1920 Ii5M SEPARATION KYDROPULPER SYSTEM
1940 RIM
1950 LET S(3}= Sl'TONS/DAY
I960 LET ?(3;=(S(3)/80)".6*8E4
1970 LET 1(3)= P(3)*1.6
I960 LET E(3)=:20*S/24
1990 I
-------
-130-
COSTS (continued)
2090 LET S(5)=(A1+A4)*10*2000/(1.83*62.4)*74.8'GAL
2100 LET P(5)=(S5/1E5)~.63*15000.
2110 LE-: I(5)=P(5)*1.85
2120 REM 5LASK CHAMBER 2 KIN RESIDENCE TIME
:.130 LET S(6)= F1/60*2'FT~2
21-'0 LET P(6)= (S(6)/62.8r.65* 2500
2150 LET C(6) = ?(6)* 1.05*3-P(6)
2: SO LET 1(6)= P(6)* 2+C(6)
2170 LET P(S)= C(6)+P(6)
2i?0 REM CENTRIFUGE ONE
2190 LET S(7)= C6/(.134*60)' GAL MIN OF WATER 'TO CENT
2195 I? S(7)>350 THEN 2220
2200 LET P(7)= (S(7)/350 )~.6*9E4
2210 LET 1(7)= P(7)» 1.6
2211 LET X7=l
2212 LET h(7)=S(7)
2213 GOTO 2230
2220 REX
2222 LET P(7)=9E4 + ((S(7)-350 )/350)~.6*9E4
2223 LET I(7)=?(7)*1.6
2224 LET X7=2
2225 LET H(7)=S(7)
^230 RSM CONTINUOS FLOW NEUTRALIZER
2250 REM IN LINE KIXER
2260 REM DESIGN 150 PSI
2270 LET S(8)=N1/.134/60'GAL/MIN
2260 LET P(8) = (s(8)/500r.6*5000
23CO LET I(S)=P(8)*1.6
2305 LET H(6)=S(8)/5CO*5
2310 REM LIME STORAGE
^1° r^ ;>(9) = (A3-A5)*lGO/9e*2000*10/85'FT~3 FOR 10 DAYS
•^JJ Li^'i ?^9;=S(9) .9
2340 LET I(9)=l.l*p(9)
235C REM CEl^RIFUGE 2
3c,0 LET S(10)= N2/60' GAL MIN
•57H ~ .rrrn rj / I I-. \ - I r- I 1 r\ \ / -, ~i r \ ~ m.
7"57n , .
j. ~ t ^ lj^
23SO LET 1(10)= P(10)» 1.6
2390 LET li(10) = (S(10)/375)*100'K.P.
2400 REM ;,UX DIRECT FIRED HEATER
2-110 GO SUB 4170
2<11 LET X5=(K4-212)*5/9+100
2420 REM T25.T1
2<22 ?RIXT"TEKP TO HEATER DEG F AND c ,K4,i<5
24?3 px""vT
2-',25 FRIXT"EVAPORATCRS"
2425 ?RINT"STZAM FROM FLASH,USED";w£i,i)/2000; TONS/ER
2427 PRiXT"EFFECT","VAPOR","TONS/KR
242c i-'OR J=2 TO 7
2.429 PRINT J-l, W(J , 1 )/2000
2430 NEXT J
-------
COST 3 (cor.t inued)
LFT S (II } = (X+W 1) * (T2-K 5 )* 9/5* 2000/24 'BTU/HR
1440 LET P(ll)= (S(11)/5E6)~.85*20000
2450 LEI I(11)= P(ll)*1.63
2450 REM OIL rXD OIL STORAGE BASED ON 1.5E5 BTU/GAL
247C LET C=(S(11)/1.5E5/.85)'GAL/HR AT 85 % EFF
2-4?C> LET 01= 0*24' G.-L/DAY
2490 LSI S(12)= 01*10' GAL FOR 10 DAYS
2500 IF S(12)>4E4 THEN 2540
2510 LET P(12)= (S(12 )/lE4)~.30*1600
2520LST I(12)= P(12)* 1.4
2530 GOTO 2560
25^0 LET P(12)= (S(12)/3E5)~.63*3E4
255G LET 1(12)= P(12)* 1.85
2560 RSF *##w-,i-*pTjMPS****#***
2570 REX ** FRESH WATER TO PULPER
2580 REM OUT PSI= 60
2590 LET S(13)= P6' FT~3/KR TO PULPER
2600 LET K(13)= 60* S(13)*7.27E-5
2610 LET P(13)= (3(13)7.134/60*60/1000)".52*600
H(13)= E(13)/.8 :
I(13)= P(13)*2.41
2640 REK PULPER RECYCLE PUMP
2650 LET S(14 )= P5
2fa60 LET p(14)= (S(14 )/.134/60*60/1000)".52*600
2670 LET K(14)= 60*S(14)*7.27E-5
26SO LET I(14)= P(14)*2.41
2690 LET E(14)= H(14)/.8
2700 REX XOYNO SCREW FUMP FROM PULPER TO REACTOR
2710 LET S(15)= (S1+2*S1)*2000/24/62.4'FT~3/HR
2720 LET E(15)= 400*S(15)*7.27E-5
27?C LET P(15)= (S(15)/.134/60*400/1000)".52*600
2740 LET H(15)= K(15)/.8
2750 LET I(15)= P(15)*2.41
27r.O REM I12S04 FEED PUMP
2770 REM CONTROLLED VOLUME PUMP
?7fO LET S(16)= P4
27SO LET 1-1(16)= 500* S (16 )*7 . 27E-5
2800 LET P(16)=(S(16)/.134/60*500/1000)*.7*1200
2810 LET K(16)= H(16)/.8
2S20 LET I (16)= P(16)*2.41
2830 REI'i. REAC10R KOYNO RECYCLE PUMP
2640 LFT S(17)= (Wl+R+R4*R )/62. 4*2000/2-4
2650 LET K(17)= S(17 )*500*7.27E-5
26SO LET P(17)= (S(17)/.134/60*200/1000)".52*600
2870 LET I (17)= P(17)*2.41
2F.60 REM FEED WATER TO HEATER PUMP
2t90 LET S(18)=(P+W(1,1))/62.4'FT"3/ER
2900 LET H(18)=S(18)*400*7.27E-5/.8
2910 LET P(18)=(S(18)/.134/60*400/1000)".52*600
2920 LET I(18)= P(18)*2.41
-------
-132-
COSTS (continued)
:930 REX MOYNO SCREW FOR LIMEE
2940 LEI S(19)= 2*N3/85*2000
2950 LET h(19)= S(19)*ICO/.8*7.27E-5
7950 LSI ?(19)=(S(19)/.134/60*100/1000)".52*600
2970 LEI 1(19)= P(19)*2.41
2980 RE>': 6 STAGE EVAPORATION SYSTEM
29S2 LET P(20)=S(20)".53*1200*6
2983 LfT I(20)=P(20)*1.9
2990 REM CONDENSER
2';93 LET P(21)=(S(21)/1000)". 65*9000
2995 LET I(21)=P(21)*2.34
3000 LET P(22)=(S(22)/62.8)".65*2500
3030 LET I(22)=P(22)*2
3015 PRINT
3016 PRINT
3017 PRINT"DO YOU WANT EQUIPMENT SIZE AND COST YES OR NO
30IS INPUT Y$
3019 IF Y$="NO" THEN 3290
3070 pRi>s;'i>"*****-****CCST ANALYSIS*******************,,
3cso PRTMT"EOUIPMENT"/'SIZS"," ","P COST","IN COST-"
3090 PRINT"HOPPER",s(i),"FT"3",p(i),i(i)
3100 PRINT"CONVEYOR",S(2),"FT",P(2 ),I(2 )
sue PRINT"HYDROPULP',s,'TONS/DAY' ,p(3),i(3)
3120 PRINT"REACTOR" ,s(4), FT"3",p(4),i(4)
2130 PRINT"ACID STORAGE',s(5),"GAL",p(5),i(5)
3140 ?RINT"FLASE cKAMIf,s(6), FT"3",p(6),i(6)
3150 PRINT"CENT";X7,S(7),"GAL/MIN" , P ( 7 ), I ( 7 )
3160 PRINT"NSUT",S(S),"GAL/MIN"tp(8).i(8)
31VU PRIXT"LIKE ST",S(9),"FT"3",?(9),I(9)
3160 PRINT"CENT i",s(10),"GAL/MIN",p(10 ).i(10 )
3190 PRi:rr"HEATER' ,s( 11), "BTU/HR",P(II). 1(11)
3200 PRINT"OIL ST",S(12),"GAL",P(12).I(12)
3210 PRINT"EVAP",s(20),' FT~2",P(20),i(20)
3220 PRINT"CCND",s(21),"FT"2",P(21).i(21)
3230 ?3!NT"COND PO1" , S( 22 }, "FT'^3" , P( 22 ), I( 22 )
32^-0 PRI",T
3250 PRINT"PUXP","FT"3/hR","K.P.","P.C.","l.C."
3260 FOR 1= 13 TO 19
3270 PRINT" ",S(l).H(l).P(l)fl(l)
3230 NEXT I
p=o
33CC LET 1-0
3310 FOR 1= l TO 22
3320 LEI P= P+P(I )
3330 LEri T= T+I(I)
2350 PRINT
3352 LE'i ?=?*••!. 06'6% ESCALATION FOR KID 1969
3353 LET T= T*1.06'65S ESCALATION
3360 PRINT
-------
-133-
" ;B; ^DOLLARS"
]'_ jF-^" DOLLARS'^
" ;C ; "DOLLARS"
" ;E; "DOLLARS"
>t;D; DOLLARS
" "
;Y8; DOLLARS
COSTS (continued)
3370 ?RINT"TOTAL PURCHASED EOUUIPMENT COST";P;"DOLLARS"
3360 PRINT'"TOTAL INSTALLED EQUIPMENT COST" ;T ; "DOLLARS"
3390 LET B= . 2*T
34CO LET F=.08*1?'FREIGHT AND TAXES
3410 LET C= . i'/i.V--T' CON STRUCT ION
3420 LET E=.10*T'ENGINEERING
3430 LET D=T+B+F+C+E
3440 LET Cl=.lb*D'CONTINGENCY AND CONTRACTORS FEE
3442 LET Y£= . 15*(D+Cl)
3450 PR INT"BUILDING COST
3460 PRINT "FREIGHT AND TAXES
3470 PRINT"CONSTRUCTION COST
3480 PRINT"ENGINEERING COST
34 SO PRINT
3500 PRINT "DIRECT PLANT COST
3510 PRINT"CONTINGENCY AND CONTRACTOR FEE ";C1; DOLLARS
3520 PRINT
3530 PRINT"FIXED CAPITAL INVESTMENT ";D+ci;"DOLLARS
3540 PRINT
3544 PRINT"WORKING CAPITAL
3550 PRINT
3560 PRINT"****************MANUF AC-TURING COST*********************
3570 LET A=(A1+A4)*32'COST OF ACID
3580 LETL= (A3-A5)/9S*100*12'COST OF LIME
3590 LET W= G3/.12'TONS OF WATER UESED
3600 LET W2=W*2000/62.4/.134'GALS
3610 let w3=w2/1000*.25'cost of water
3620 LET 03=01*0.1'OIL COST
363CLET K=0
3640 FOR 1= 1 TO 22
3650 LET H= H+K(I)
3660 NEXT I
3670 LET E=E*24*.7487'KILOWATT HOURS
36SOLE1 E1=S*1.25*.010'COST PER DAY WITH 25% SURPLUS
3690 LSI Li=3*3*24'LABOR 3 MEN 24ER AT 3.00
3700 LET L2=3.5*1*24'SUPERVISION
3710 let f=.15*(ll+12)'fringe benefits
3720 LET M=(D+C1)*. 05/360 'MAIN AND REPAIRS AT 5% OF FCI
3730 LET S=.15*M/2'SUPPLIES 15% OF MAIN
3740 LET Fl=.02':i-(D+Cl)/360'TAXES 2 % OF FCI
3750 LET F2=.01*(D+C1)/360'INSURANCE
3760 LET P = .15*(Ll+L2+M/2)'PAYROLL OVERHEAD
3770 LET'P2= P'LAB WORK
3780 LET P3=.5*(Ll+L2+M/2)'PLANI OVERHEAD AT 50% OF LABOR
379C FRINT"INPUT DUMPING FEE PER TON"
3800 inputMdl
3810 print"input cost for disposal of waste per ton"
3820 INPUT D2
3830 PRINT"INPUT INTEREST RATE AND YEARS OF OPERATION"
3640 INPUT R1.R2
-------
-134-
(continued)
CCS13
3£42 LET C2=(Rl*( l
3«?50 LET C=C2*(D+C1+Y8)/3&0
3S6G REM CAPITAL RETURN
3870 let t=a-;-l+w3+o3+el+ll+12+f+m+s+f
3&SO LEr: T=T+D2-''E9
3900 PRINT 'RAW MATRIALS(<
3910 PRINI "ACID" .A.1+A4, "TONS/DAY" , A
3920 PRINT"LIME", (A3-A5)/96*ioo, "TONS/DAY ,L
3930 PR INT "UTILITIES"
39*0 PR I NT " WAT ER " , W 2 f " G AL £ " , W 3
3950 PRINT "ELEGI".?,, "KILOWATTS", El
3960 print "oil" ,01 , "gal" A0l*0 . 1
3970 PR INT "WASTE 'DISPOSAL , E9, "TONS/DAY" ,D2*E9
39&o PRIKT"LAEOR" . 72 , "HRS" ,LI
3990 PRINT;; SUPER", 24, ;;KRs;;,L2t,
4000 print FRINGE BEN , , , f
4010 PRIMT"y.AIMTAIN" ,Z9,Z9,M
4020 PKINT"SUPPLIES" ,z9,z9,s
4025 PRINT"FIXED CHARGES"
4030 PRINT"TAXES",Z9i,Z9,Fl
4040 PRIN'l"lNSURAKCE"J(iZ9,Z9|,F2
4050 print" cap return ," ", ",c
4055 PRIKT/'GENSRAL COST"
4060PRINT"PAY OVER" , Z9,Z9 ,P
4070 ?RINI"LAB" ,Z9,Z9,,P2
4oso PRIKT"?LANT OVER ,Z9,Z9,P3
4090 PRINT
4095 ?RINT"TONS /DAY OF GLUCOSE" ;G3
4100 PRINT "TOTAL COST/DAY NO DUMP FEE
4110 PRINT
^120 PRIXT"COST/TOX OF GLUCOSE NO DUMP
4130 ?RI:^"CO£T/L3 0? GLUCOSE NO DUMP
4140 PRINT
4145 PRINT "TOTAL COST /DAY WITH DUMP FEE
4150 PR INT "COST /TON WITH DUKP FEE
41 60 PRINT "COST/L3 WITH DUMP FEE
4165 GOTO 51 &0
41 66 KEK*-*-"-~**-*w*****EVAPORATOR CALCULATION******************
4170 LET N-6
41£0 DIM 3(7,7),W(7.1).F(7.1).R(7.7)
4190 /.AT E=ZER
4270 FOR 1= 0 TO N
42SO READ T(l),L(l)
4290 L7.1 X(I)=T(I)
4300 NEXT I
4310 LET C1=G3/N7
4320 LET C2=.12
4330 LET ?= N7---2000/24
4340 FOR 1= 1 TO N-t-1
;T; "DOLLARS
DOLLARS
;T/G3/2000; DOLLARS
" ;Ti; "DOLLARS"
" ;T1/G 3 ; "DOLLARS"
" ;T1/G 3/2000 ; "DOLLARS*'
-------
-135-
COST3
4350
4360
4370
4390
4381
4400
4410
4420
4430
4440
4450
4460
4470
4480
4490
4500
4510
4520
4530
4540
4550
4560
4570
4580
4590
4600
4610
4620
4630
4640
4650
4660
4670
4680
4690
4700
4710
4720
4730
4740
4750
4760
4770
4780
4790
4600
4810
4820
4830
4840
(continued)
READ U(I)
NEXT I
LET G=(X+v:l)*2000/24
DATA 350,870,327,889,304,905,282,924,260,939,240.952,212,970
DATA 500,480,450,410,370,250,500
LET WO=F
LET S8=F*C1'SOLIDS IN FEED
LET P=SS/C2'PRODUCT RATE
LET E= WO-P' LBS/KR EVAPORATED IN TOTAL SYSTEM
FOR 1= 1 TO N
LET T(I)= X(I-1)-X(I)
NEXT I
LET B(l,l)= L(0)
LET B(l,2)= -L(l)
B(l,3)=0
LEI
LET B(l,4)=0
LET B(2,l)=0
LET B(2,2)= L(1)-T(2)
LET B(2,3)=-L(2)
LET B(2,4)=0
LET B(3,l)=0.
LET B(3,2)= -T(3)
LET B(3,3)=L(2)-T(3)
LET E(3,4)=-L(3)
IF N>3 THEN 4640
LET B(4,l)=0
LET B(4,2)=l
LET B(4,3)=1
LET B(4,4)=l
LETF(N+1,1)=E
FOR 1= 1 TO N
LET F(1,1)= -WO*T(I)
NEXT I
IF N=3 THEN 4810
FOR J=4 TO N
LET B(J,1)=0
FOR 1= 2 TO J-l
LET B(J,I)=-T(J)
NEXT I
LET B(J,J)=L(J-1)-T(J)
LET B(J,J+1)=-L(J)
NEXT J
LET B(N+1,1)=0
FOR 1= 2 TO N+l
LET B(N+1,I)=1
NEXT I
KAT R= INV(B)
MAT W= R*F
REM CALCULAT AREAS
FOR 1= 1 TO N
-------
-136- i •*
COSTS (continued)
<650 LET A(I)= W(I,1)*L(I-1)/(U(I)*T(I))
4860 NEXT I
4870 LET G2= W(N+1, 1 )*L (N )/140
48SO LET T4= 212-70
4890 LET T5 = 2
4900 LET T6= (T4-T5)/(LOG(T4/T5) )
4910 LET A(N + 1)=W(N+1,1)*L(N)/(U(N + 1)*T6)
4920 PRINT
4930 PRINT
49^0 FOR 1= 1 TO N
4950 LET Z8=Z8+W(I,1)*X(I-1)
4960 LET Z7=Z7+W(I,1)
4970 NEXT I
49fQ LET Kl=(Z84.W(7,l)*212+P*212)/(Z7+W(7.1) + P)
5050 LET S(22)=(X+Wl)*2000/62.4*2/24/60
5055 LET K2=(W*V9*2000/24-W(ltl))*870/((X+Wl)*2000/24)
5056 LET K4=K1+K2'F DEG INTO HEATER
5C57 LET Y-0
5080 FOR 1= 1 TO N
5090 LET Y= Y+A(l)
5100 NEXT I
5110 LET Y2= Y/N
5152 LET S(20)=Y2
5153 LET S(21)=A(N+1)
5157 RETURN
5160 END
yo 651
------- |