xvEPA
United States
Environmental Protection
Agency
Industrial Environmental Research EPA-600/2-78-191
Laboratory July 1978
Cincinnati OH 45268
Research and Development
On-Site Production
of Activated
Carbon From Kraft
Black Liquor
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7 Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-78-191
July 1978
ON-SITE PRODUCTION OF ACTIVATED
CARBON FROM KRAFT BLACK LIQUOR
V. D. Del Bagno
R. L. Miller
J. J. Watkins
St. Regis Paper Company
Cantonment, Florida 32533
Grant No. 120^0 EJU
Project Officer
John S. Ruppersberger
Food and Wood Products Branch
Industrial Environmental Research Laboratory - Cincinnati
Corvallis, Oregon 97330
INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U. S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 1+5268
-------
DISCLAIMER
This report has been reviewed by the Industrial Environmental Research
Laboratory - Cincinnati, 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 trade names or commercial products constitute endorse-
ment or recommendation for use.
11
-------
FOREWORD
When energy and material resources are extracted, processed, converted,
and used, the related pollutional impacts on our environment and even on our
health often require that new and increasingly more efficient pollution
control methods be used. The Industrial Environmental Research Laboratory-
Cincinnati (lERL-Ci) assists in developing and demonstrating new and improved
methodologies that will meet these needs both efficiently and economically.
This report presents the findings of a pilot-scale project on the
production of char and activated carbon from kraft black liquor. The gran-
ular activated carbon compared favorably to commercial activated carbons in
its ability to treat kraft mill wastewater. The results will interest both
industry and regulatory agencies in considering alternatives for effluent
polishing and color removal. Cost estimates are also presented. For fur-
ther information contact the Food arid Wood Products Branch, Industrial
Environmental Research Laboratory - Cincinnati.
David G. Stephan
Director
Industrial Environmental Research Laboratory
Cincinnati
in
-------
ABSTRACT
The overall objective of this grant program was to develop and demon-
strate an economical system for producing a reusable water from kraft pulp
and paper mill effluent. The objective of Part II of the program was to
demonstrate the production and activation of a carbon char derived from
black liquor and develop the economics of on-site production of activated
carbon.
A pilot plant was designed and constructed to produce char via the
St. Regis hydropyrolysis kraft chemical recovery process and to produce
activated carbon from the char. This report includes discussion of labo-
ratory and prepilot work conducted outside the grant program which formed
the basis for pilot plant design. The report describes the plant and
presents operating results.
After a period of optimizing feed pretreatment, temperature, and pres-
sure conditions in the hydropyrolysis section of the plant, about 22,000 kg
of char was produced for activation. The char was converted to a high
quality granular activated carbon having properties which compared favorably
with commercially available carbons. The concept of on-site production of
activated carbon and the use of such carbon for local effluent treatment has
been demonstrated to be technically sound.
The cost of on-site production of hydropyrolysis activated carbon in
quantities required for local effluent treatment is not competitive with
commercially available carbons. To become competitive, the plant would
have to supply carbon for effluent treatment at several locations. To
achieve significantly lower cost would require entry into market
production.
This report was submitted in fulfillment to Grant No. 120^0 EJU under
the partial sponsorship of the U. S. Environmental Protection Agency. This
report covers the period from November 1, 1971 to September 30, 1975 and
work was completed as of September 30, 1975•
IV
-------
CONTENTS
Foreword iii
Abstract vi
Figures vii
Tables ix
Acknowledgement x
1. Introduction 1
2. Conclusions 3
3 • Recommendation k
U. Hydropyrolysis Process Development
Prior to pilot plant design 5
5- Activated Carbon Technology Development
Prior to pilot plant operations 8
Nature of hydropyrolysis char 8
Carbonization ; 10
Activation 12
Regeneration of St. Regis activated carbons Ik
Multiple hearth furnace trial prior
to pilot plant startup 15
6. Pilot Plant Design 19
Design procedure 19
Basis for preliminary engineering and
selection of capacity 19
Description of the process for design purposes 20
Basis for the detailed design and
preparation of the big package 23
7. Pilot Plant Construction and Description of Plant 27
8. Operating Procedure , k-6
Wet end start-up 1*6
Operation on black liquor 1*7
-------
CONTENTS (continued)
Sampling and analytical procedures U8
Dry end operating procedure U8
Operating schedule k&
9. Pilot Plant Operating Results 51
Wet end operational testing period 51
Production of char for dry end
operational testing 5^
Wet end production demonstration 57
Additional production runs 59
Dry end operations 6l
Analysis of pilot plant char 6l
Drying operations 63
Compacting operations 6k
Multiple hearth furnace operations 68
10. Activated Carbon Plant Design and Cost Estimate 73
Overall design criteria 73
Design methods and assumptions' 73
Cost estimates 76
References 83
Appendix &b
vi
-------
FIGURES
Number Page
1 Flow diagram for the hydropyrolysis process 7
2 Surface area vs. activation temperature 18
3 Flow diagram for preliminary pilot plant design 21
k Flow diagram for detailed pilot plant design 25
5 Pilot plant layout drawing 26
6 General view of pilot plant 27
7 Black liquor pretreatment equipment 29
8 Pressurization equipment 30
9 Cross exchanger 31
10 Heating and reaction equipment 32
11 Heating and reaction equipment 33
12 Slurry cooling and pressure let-down equipment 3^
13 Slurry-gas separation equipment 35
1^ Flash-gas combustion equipment 36
15 Filtration and washing equipment 37
l6 Char drying equipment „ 39
17 Compaction equipment kl
18 Compactor system schematic k2
19 Multiple hearth furnace firing controls hk
20 Furnace open for inspection UU
vii
-------
FIGURES ( continued)
Number page
21 Operating schedule 1*9
22 Effect of moisture content on hardness of compacted char ... 67
23 Compactor material balance 69
2k Carbon activation flow diagram 7k
25 Effect of production rate on cost of carbons ".... 79
26 Effect of on-site carbon production on effluent treatment ':cost 8l
viii
-------
TABLES
Number Page
1 Performance of St. Regis Carbon and Atlas Darco on Mill Effluent.. 16
2 Operational Testing Period 52
3 Production of Char for Dry End Testing 55 & 56
4 Ease of Operation - Production Runs for Dry End Testing 58
5 Wet End Production Demonstration 58
6 Additional Wet End Production Runs 60
7 Effect of Temperature on Plugging 61
8 Drying Rates 64
9 Multiple Hearth Runs 70
10 Categorization of St. Regis Activated Carbon, Pittsburgh CAL and
Granular Darco 71
11 Loading Comparison of Activated Carbons 72
12 Cost Estimates for Activation Plants 78
13 Cost Estimates for Effluent Treatment Plants 80
-------
ACKNOWLEDGEMENTS
Completion of this grant program and the prior work ehich led up to the
program involves a list of engineers, scientists, technicians and consultants
too long to be detailed here. The authors gratefully acknowledge their many
contributions.
Among the many St. Regis people, past and present, who contributed
their time and expertise, those who were most directly involved with the
program included: Mr. W. J. Evers, Mr. W. E. Gay, Mrs. L. B. Halfacre,
Mr. R. I. Padgett, Mr. 0. C. Pippin, Dr. D. R. Raymond, Mr. D. N. Robbins,
Mr. J. N. Rockwell, Mr. W. Y. Sirmon, Dr. I. H. Stockel and Mr. L. M. William-
son.
Mr. E. W. Lang, who directed Part I of the grant program, provided
valuable advice and counsel during the continuation into Part II.
The authors acknowledge the patient guidance and the constructive
critiques by the EPA Project Officer, Mr. J. S. Ruppersberger.
Sadly, but gratefully, the authors acknowledge the genius of the late
Mr. W. G. Timpe who invented the hydropyrolysis process, developed the con-
cepts which led to this grant program, and organized the execution of the
program.
-------
SECTION I
INTRODUCTION
St. Regis Paper Company, with partial support by EPA, initiated a
program in July, 1969, to develop and demonstrate an economical system for
producing a reusable water from kraft pulp and paper industry effluent. The
program was organized in two parts. Part I covered the use of activated
carbon in effluent treatment and the pilot plant demonstration was completed
in December, 1972. The objective of Part II as proposed orginally was to
demonstrate the production and activation of a carbon char produced from
black liquor and determine the economics of the mill site preparation of the
activated carbon.
Two reports have been issued. The first report (l) was a survey of the
literature and a review of industry information on effluent treatment and
the reuse of treated water in pulp and paper mills. The second report (2)
described laboratory investigations and the design, construction and opera-
tion of the pilot plant utilizing activated carbon, alone and in sequence
with other treatment steps, to treat the combined effluent streams from an
unbleached kraft pulp and paper operation. Capital and operating cost
estimates were included in the report.
It was concluded from Part I of the program that treatment sequences
involving carbon adsorption can produce reusable water having a remaining
color of 100 APHA-NCASI color units (CU) and a remaining total organic
carbon (TOG) of 100 mg/1. Based on the survey of the literature and industry
information, these had been defined as the tentative criteria for a re-
usable general process water. Treatment for reuse was confirmed as an
attractive alternative to treatment for discharge, particularly in the light
of expected future discharge standards.
It was recommended that a better working definition of reusable water
be established and that grants be initiated to demonstrate actual reuse
of treated effluents on a mill scale. The results of such work could
significantly alter the definition of reusable water and change the relative
economics of treatment for reuse versus treatment for discharge. This work
was not within the scope of the St. Regis program.
Assuming that activated carbon would have application in the production
of reusable water, the original grant proposal had anticipated that the cost
of carbon would have a significant impact on the economics of treatment.
The results of Part I showed this to be true in varying degrees depending on
-------
the treatment sequence. Part II of the grant proposal had suggested that
activated carbon produced on-site from material available on-site would be
lower in cost than activated carbon obtained through normal commercial
channels. Specifically, it was proposed that activated carbon be produced
from kraft black liquor as a by product of hydropyrolysis recovery of kraft
pulping chemicals.
On-site production and use of activated carbon was expected to show
savings in raw material acquisition and shipping costs, savings through
integration of energy inputs and outputs with the hydropyrolysis process,
and savings in packaging and shipping of the activated carbon.
The hydropyrolysis process was conceived and partially developed by
St. Regis prior to the July, 1969, EPA grant. The patent application and
detailed data were submitted to EPA and placed in background. St. Regis
agreed to future licensing of the hydropyrplysis process to those who might
wish to utilize the process as a starting point for on-site production of
activated carbon.
This report describes successful pilot-scale production and activation of
hydropyrolysis char and develops the economics of on-site production of
activated carbon.
-------
SECTION II
CONCLUSIONS
1. A char suitable for production of activated carbon can be obtained via
the hydropyrolysis kraft chemical recovery process. Char production has
been demonstrated in a pilot plant system.
2. Hydropyrolysis reaction conditions must be selected and controlled to pro-
duce acceptable char while avoiding operational problems. A range of
satisfactory operating conditions has been defined.
3. Production of granular activated carbon from hydropyrolysis char has been
demonstrated in a pilot plant system.
4. Hydropyrolysis char is a good starting material for production of
granular activated carbon. The char can be compacted using little or no
added binder material. Activation rates are high.
5. Hydropyrolysis char requires calcining before activation.
6. Hydropyrolysis activated carbon has properties - surface area, hardness,
adsorption performance - which compare favorably with commercially avail-
able carbons.
7. Pilot plant experience and operating data provide a sound basis for com-
mercial plant design and cost estimates.
8. The cost of on-site production of hydropyrolysis activated carbon in
quantities required for local effluent treatment is estimated to be
$1.52/kg (October 1975 dollars). This cost is not competitive with
commercially available carbons at $0.88 - $1.10/kg.
9. The production cost of hydropyrolysis activated carbon decreases as the
production level increases. To become competitive, the plant would have
to supply carbon for effluent treatment, at several locations. To achieve
significantly lower costs would require entry into market production.
10. On site production of activated carbon has an effect on the cost of
effluent treatment, primarily through the carbon regeneration cost
element. Treatment costs become significantly lower if the activated
carbon plant is servicing three or more treatment plants.
-------
SECTION III
RECOMMENDATION
The concept of on-site production of activated carbon, using readily
available on-site materials (black liquor), and the use of such carbon for
local effluent treatment has been demonstrated to be technically sound. No
further work nor application of this technology is recommended at this time
because current economics do not favor this approach to effluent treatment.
There may be specific instances in which the economics would be more
favorable. Also, changing economics conditions could alter this recommenda-
tion in the future.
-------
SECTION IV
HYDROPYROLYSIS PROCESS DEVELOPMENT PRIOR TO PILOT PLANT DESIGN
The desire to eliminate the problems associated with the conventional
kraft recovery system and the potential for economically more attractive
recovery processes prompted St. Regis to undertake a research and develop-
ment program in this area. The economical recovery of new organic chemicals
or other byproducts (for example, activated carbon) would be a long range
objective. Several new processes were developed and pursued by St. Regis
with the hydropyrolysis process resulting as the most attractive for kraft
recovery. Hydropyrolysis is a process for the recovery of pulping chemicals
and energy from kraft black liquor. The process has been described in a US
Patent (3).
In the hydropyrolysis process, kraft black liquor is subjected to high
temperature while being maintained under pressure in the liquid state.
Under these conditions, the organic components of the black liquor are
converted to a carbonaceous solid (char) which is dispersed as a slurry in
the aqueous phase. This reaction of the organics also generates a gaseous
phase which separates from the aqueous slurry when pressure is reduced at
the discharge end of the reaction system. The gaseous product can be
burned to generate heat for input to the hydropyrolysis reaction system and
the sulfur values in the gaseous product can be recycled to the feed black
liquor.
Most of the sulfur and essentially all of the sodium in the black liquor
are retained in the liquid phase of the product slurry. Small fractions of
the sodium and sulfur are found in the char. After filtration and washing,
the char is suitable for combustion as a relatively clean fuel or for thermal
treatment to produce activated carbon. The filtrate essentially has the
composition of green liquor and is reacted with lime (CaO) to regenerate
kraft pulping chemicals.
During the period 1969-1972 St. Regis conducted an intensive laboratory
research program using batch reactors to examine reaction variables:
temperature, time, feed liquor concentration, and effect of recycled gaseous
components. Fundamental experiments were carried out to determine heat
transfer data for black liquor in a temperature and pressure range well
above anything reported in the literature, with attention given to the
influence of a disperded solid phase and a gaseous phase.
.5
-------
Based on the results, a prepilot reactor system was designed and con-
structed. Operation of the prepilot system yielded the data and. operating
experience needed before the pilot plant design phase could be initiated.
Timpe and Bvers (h) have summarized the work completed during the 1969-1972
period, so there is no need to describe the results in any detail.
The results of the laboratory research program and operation of the
prepilot system identified the process flowsheet as that shown in Figure 1.
In this figure, the hydropyrolysis process is divided into unit processes
or operations, which are the following:
1) Pretreatment
2) Pressurization
3) Heat recovery
k) Heating and reaction
5) Pressure letdown
6) Flash gas separation
7) Flash gas combustion and recycle
8) Liquid-solid separation and char washing
Black liquor, after three effects of evaporation and soap skimming,
contains 25% solids, and is fed to the process. Before the black liquor
can be pressurized or reacted, it must be subjected to a pretreatment which
involves the absorption of SO into the liquor. After pretreatment, the
treated black liquor is pressurized by pumping and heated to a reaction
temperature at which char formation proceeds with a reasonable rate. The
heating process occurs in two steps. First, the treated black liquor is
heated by the hot slurry stream leaving the reactor in a heat recovery pro-
cess, then the liquor is heated to reaction temperature by a process which
utilizes an outside source of energy. After heating, the liquor enters the
reactor where the char formation reaction occurs. After leaving the
reactor, the slurry is cooled in the heat recovery unit and flashed by a
pressure reduction process. The flash gas is separated from the slurry,
combused, and recycled to black liquor pretreatment. The char in the slurry
from, flash gas separation is separated from the liquor and the char is
washed to recover entrained pulping chemicals. The liquor and concentrated
wash contain the inorganic pulping chemicals; these are combined and sent to
the caustic plant. The washed char, after drying, can be used as a boiler
fuel or further processed into activated carbon.
-------
BLACK LIQUOR
FROM SOAP SKIMMER
PRETREATMENT
PRESSURIZATION
HEAT
RECOVERY
FLASH GAS
COMBUSTION
&
RECYCLE
FLASH GAS
SEPARATION
PRESSURE
LETDOWN
WATER
HEATING
&
REACTION
LIQUID-SOLID SEPARATION
&
CHAR WASHING
I
LIQUOR
WASH
LIQUOR
WASHED
CHAR
Figure 1. Flow diagram for the hydropyrolysis process
-------
SECTION V
ACTIVATED CARBON TECHNOLOGY DEVELOPMENT
PRIOR TO PILOT PLANT OPERATIONS
INTRODUCTION
This section discusses the nature of hydropyrolysis char and laboratory
production of activated carbon from hydropyrolysis char. The section also
covers vork on regeneration of St. Regis activated carbon and a series of
tests to correlate laboratory fluid bed furnace results with multiple hearth
furnace results. A substantial protion of this work was conducted outside
the EPA grant program.
NATURE OF HYDROPYROLYSIS CHAR
In order to understand better the mechanism by which a carbonaceous
material is activated and to interpret data taken during activation runs, it
is necessary to have a knowledge about the material being treated.
The organic compounds present in black liquor are the modified lignins
and carbohydrates present in wood. Artificial coalification (5) studies
have shown that coal-like bodies are produced by hydrothermal treatment of
lignin and cellulosic substances. These studies also have shown that an
alkaline medium favors the production of substances with weak coking pro-
perties. Therefore} the carbonaceous product from hydropyrolysis of kraft
black liquor should have a structure similiar to low rank coals. It is well
known that good quality activated carbons are produced commercially from
coals (6).
Torikai and Walker (7) have reviewed the literature on activation of
coals. They find that relatively high volatile coals, from lh% to 3,8%
volatile matter, have been activated by various investigators to produce
excellent activated carbons. Conditions of preliminary carbonization and
activation are widely varied. King, MacDougall, and Gilmour (8) have
reported that the best starting materials contain about 38% volatile matter,
82% carbon, and 5$ hydrogen. Torikai and Walker (j) show an analysis of two
bituminous coal lithotypes which were reacted to produce satisfactory
activated carbons.
-------
Proximate Analysis,
Ultimate Analysis,
Volatile Fixed
Moisture Ash matter carbon Sulfur Carbon Hydrogen Nitrogen
Coal A 2.1
Coal B 2.5
3.9 37.6
2.1 38.3
62.4 0.63 85.0 5.6 1.4
61.7 0.88 83.4 5.8 1.6
In order to determine the nature of hydropyrolysis char and compare it
to coals, several analyses were carried out. Among these were analyses for
carbon and hydrogen. Char A was washed and dryed and Char B was the
same material after a carbonizing step.
Carbon
Hydrogen
Char A
Char B
84.53
86.18
3.23
3.18
Comparing the values of Char A with the data for bituminous coals,
it is shown that hydropyrolysis char contains nearly the same weight of car-
bon as the coals reported: 84.53% as opposed to 84.2% average for
two coals. When hydrogen content is compared, however, it is seen that the
hydropyrolysis char has significantly less hydrogen than the coals: 3.23%
for char that has not been thermally treated vs. 5.7% average for coals.
The mole ratio of hydrogen to carbon is also less: 0.458 for the untreated
char as opposed to 0.80 for the coals. Several tests showed volatile
matter to vary between 30 and 40% for experimental char. Comparing this
with the values for coals, we see that the hydropyrolysis chars are quite
similar to coals in volatile content.
Since certain metals present in carbonaceous materials have been re-
ported to act as catalysts in some gasification reactions, Char A
was analyzed for metal ions. The results, calculated as % in the ash, are
shown below. Also included are values for the ash content of Pennsylvania
bituminous coals.
-------
Pennsylvania
Element
Sulfur
Sodium
Iron
Calcium
Magnesium
Chromium
Nickel
Silicon
Aluminum
Titanium
Potassium
Total Ash
Hydropyrolysis char bituminous coal
2.4
10.0 0-2.6
11.0 . 4.5
4.0 7.2
2.0 1.1
6.0
2.0
55.5
17.0
1.7
0-3.0
2.4 3-8
Perry's Chemical Engineers Handbook, Fourth Edition.
These data show that hydropyrolysis char, as a raw material for acti-
vated carbon, ranks on the low side in total ash when compared to bitu-
minous coals, but that this ash is composed of different elements in
different amounts. How this different composition should effect the gasi-
fication or activation of char was not certain at this time.
•CARBONIZATION
It is generally known that, in order for a material to be a candidate
for activated carbon manufacture, it must be carbonaceous, stable to mild
oxidation, have a relatively low ash content, and also contain some internal
pore structure. The reason for such a structure is to provide the activat-
ing gas with a means of entering the carbon particle, to react with it, and
then return to the surroundings as a reaction product. In this manner,
micropores are enlarged, new pores previously unavailable to activating
gases are opened, and the result in development of the surface area neces-
sary for adsorption. The access to the micropores described above is quite
often through macropores that exist naturally in the surface of carbonaceous
materials. The absence of such macropores does not mean that micropores are
not present, but in their absence the development of micropores can be great-
ly inhibited. When macropores are not naturally present, they sometimes can
be produced by a heat treatment step (calcination) prior to activation.
10
-------
It has been shown that hydropyrolysis char has many of the properties
of coal. It is known (9) that when coals are heated from ambient tempera-
ture to temperatures above 400°C , they undergo a transition, similar to the
glass transition in polymers, in the temperature range between 300 and 400°C.
At the same time, a considerable amount of volatile matter escapes. On heat
treatment, a portion of the material melts. During this time, the material
undergoes active decomposition with the elimination of smaller molecules
and formation of carbon-carbon bonds. Thus, escape of volatiles through
the liquid phase creates a porous structure and the carbon-carbon bond forma-
tion which takes place at a higher temperature tends to make the system
rigid, thus retaining the porous structure. Upon further heating, the
carbon-carbon bond formation continues and blocks out some pores already
formed.
Early attempts to activate hydropyrolysis char were unsuccessful, per-
haps owing to the omission of a carbonizing step to develop a macropore
structure. From this it was hypothesized that the carbonization step de-
veloped the necessary macropore structure for successful activation.
Torikai and Walker (7) have suggested a method by which the presence
or absence of an internal pore structure can be detected and evaluated re-
lative to materials known to produce high quality activated carbons. The
BET nitrogen surface area of porous solids is obtained by measuring the ad-
sorption of nitrogen at 77°K (10). Experience with porous carbons has shown
that nitrogen is essentially excluded from pores less than about 5.5A in
diameter under the conditions of measurement (11) . Carbon dioxide surface
area, on the other hand, is measured at 298°K (12). The carbon dioxide
molecule can enter into pores of about 3.0 A. Thus, theocarbon dioxide
surface area is believed to include all pores above 3.0 A in diameter. It
should be realized that this approach gives strictly a measure of micropore
volume which is then converted to an equivalent surface area. Both the BET
nitrogen method and the carbon dioxide Dubinin method have certain theoreti-
cal limitations. These limitations however, are of little concern in evalu-
ating the effectiveness of a micropore structure. It is the magnitude of
the difference between theotwo values that reflects the presence of pores in
the range of 3.0 A to 5.5 A. The larger the difference between the two
areas, the greater is the probability that a satisfactory pore structure
exists.
Initial studies of the carbonization of hydropyrolysis char were
carried out to see if a micropore structure could be developed suitable for
subsequent activation. Hydropyrolysis chars and some coals were analyzed
for their nitrogen and carbon dioxide surface areas.
11
-------
(A) (B)
_ , Sample N2 Area CO. area
SamPle treatment Wg V/g (B)-(A)
Coal A1 None 88.4 169 80.6
Coal A Carbonized 77.4 612 534.6
Coal B None 32.0 148 116.0
Coal B Carbonized 19.1 456 436.9
Hydropyrolysis Char None 0.6 128 127.4
Hydropyrolysis Char Carbonized 36.0 445 409.0
1
Soft coals of the lignite variety.
There are several points of interest in this analysis, Hydropgrolysis
char had nitrogen and carbon dioxide surface areas of 0.6 and 128 m /g,
respectively, before any heat treatment. This showed that the material
already had micropore area in the range of 3.0 to 5-5A in diameter. This
inherent pore structure should be advantageous for subsequent activation.
Upon thermally treating this char, the two areas were enlarged to 36.0
and kh^.Q m /g, respectively, interpreted to mean a substantial increase in
micropore area had been achieved and a number of micropores had been created
by the mechanism previously discussed.
As shown then, hydropyrolysis char after carbonization compares favor-
ably with selected coals in terms of the necessary pore structure for
activation, as measured by comparing nitrogen and carbon dioxide surface
areas.
2
The difference of U09 m /g in these areas is considered very satis-
factory for the material to have potential for activation. In fact,
differences in carbon dioxide and nitrogen areas up to 592 m /g have been
attained which exceeds the values for the coals.
ACTIVATION
As mentioned previously, the properties of activated carbons depend on
the existence within the carbon particle of an extensive micropore
structure. Gasification by steam or carbon dioxide is the principal
method used to modify this microporosity. It has been shown by Marsh and
Rand (13) that the development of porosity is a combination of two
processes: the widening of existing pores and the opening of closed pores.
The process of pore widening has been demonstrated by Kipling and McEnany
(iM.
12
-------
Although initial attempts to activate hydropyrolysis char met with
limited success, the results firmly established the fact that the char could
be activated,and provided an indication of the reactivity of the material.
The original laboratory activation apparatus, a fixed bed design, had
the serious experimental limitation of non-uniform exposure of the carbon
particles to the activating gas. This led to the development of a fluidized
bed reactor system. It was in this apparatus that all future activations
would be accomplished until pilot plant startup.
Activation studies in the fluidized bed reactor included the use of
carbon dioxide with nitrogen, water vapor with nitrogen, and flue gas. Flue
gas, as used in this discussion, refers to the products of the stoichio-
metric combustion of methane with air giving a product (activating) gas
of the composition: 72.7% N
18.2% H^O , and
9.1% C02 .
It must be noted that, although these runs produced a granular carbon,
the particle strength was far below commercial acceptability. The products
should be considered as powdered carbons. Later in the project, the decision
was made to produce a true granular carbon.
To simplify equipment operation and data analysis, the first activations
utilized carbon dioxide gas with nitrogen. Fractional weight loss as a
function of time for various C£L concentrations and temperatures of 750°C
and 850°C5was found to be linear and suggests a rate which is zero order with
respect to the carbon. Differences in results for the different gas concen-
trations indicate that the rate is dependent on gas concentration.
2
Surface areas above 1200 m /g were attained. Surface areas for com-
mercial carbons fall in the range of 550-1000 m /g. To achieve these areas,
hydropyrolysis char would have to be reacted to 50% weight loss with CO-
as the activating gas. At higher CO- concentrations, this weight loss was
attained in 60-80 minutes.
In addition to C09 and N , flue gas contains water vapor. Weight loss
data during the activation of char with water vapor were obtained. The rate
at all water vapor concentration levels is linear and independent of the
carbon present. A rate dependency on the water vapor concentration is
similar in trend to that of C02 at 850°C. For activation with water vapor
at 850°C, the development of total N_ surface area showed a maximum surface
area in the range of 40-60% weight loss. This weight loss Occurred after
kO-80 minutes at water vapor concentrations above 20%. Surface areas from
700 to over 1000 m^/g were obtained in this range of weight loss.
The limited amount of data obtained for flue gas activations at 750°C
and 850°C indicated that the rate is approximately the sum of the rates
13
-------
for the individual gases. Thus, each reaction appears to proceed inde-
pendently. Relating these data to values for coal reactivity showed that
hydropyrolysis char in general is more reactive than coal.
All of the above observations point to the fact that kraft hydropy-
rolysis char is a good material for activated carbon production. Data
taken on carbons produced from this material have shown nitrogen surface
areas and open pore volumes that compare exceptionally well with com-
mercially available carbons.
REGENERATION OF ST. REGIS ACTIVATED CARBONS
Granular carbon is utilized generally in packed beds. The liquid or
gas stream to be treated is forced downward or upward through the beds.
In this manner, each successive layer of carbon in a given bed acts to
remove impurities, with maximum adsorption taking place in the early stages
of contact, and less and less taking place as the gradually purified solution
continues on. Ultimately, the carbon becomes loaded to the point where the
product stream exceeds' specifications. The economics of utilizing granular
activated carbon depend upon its ability to be regenerated. Regeneration is
usually accomplished in a multiple hearth furnace or similar equipment at
elevated temperatures in an atmosphere mildly oxidizing to the carbon.
The response of St. Regis carbons to regeneration was determined by
comparing spent commercial carbons with spent St. Regis carbons under
similar regeneration conditions. Initially, a sample of St. Regis hydropy-
rolysis char was activated and the surface area was determined. The car-
bon's capacity for treating mill effluent was obtained utilizing an
isotherm determination. The fresh activated sample was then exposed to
Pensacola mill effluent to yield a loaded (spent) carbon for a regeneration
s tudy.
There was no previous knowledge of the required regeneration conditions
for spent St. Regis carbon. Also, the amount of carbon available would
permit only two regenerations. For these reasons, granular Atlas Darco
was loaded and then regenerated using a number of different conditions and
the results of these runs were analyzed to set the regeneration conditions
for St. Regis activated carbon.
Atlas Darco was regenerated under a set of conditions varying from
nitrogen gas at 500°C for 3 hours to a flue gas mixture at 850°C for
3 hours. All regenerations were performed in a fluid bed furnace. These
experiments indicated that Darco could be regenerated under rather mild
conditions. This information and a consideration of the high reactivity
of St. Regis carbon indicated that a regeneration condition of 600°C in the
flue gas mixture for 3 hours would be acceptable. The weight loss for St.
Regis carbon under these conditions was higher than anticipated with a re-
generated nitrogen surface area of 810 m /g...... The original activated
-------
carbon had a surface area of 750 m2/g. As the surface area increased with
regeneration, additional activation probably occurred and the conditions
chosen were judged too severe. A second regeneration run performed on St.
Regis spent activated carbon in nitrogen alone at 850°C and 3 hours of
residence time,restored the surface area to 680 m^/g from a spent surface
area of 470 m /g. The isotherm tests on regenerated carbons are tabulated
in Table 1. The flue gas and nitrogen regenerations of St. Regis
carbon were performed at different times and, therefore, the isotherms were
not run simultaneously on the same mill effluent. However, fresh Darco was
employed each time as a standard so the results can be compared.
The data indicated that Darco can be regenerated to approximately the
same level as St. Regis material in nitrogen. However, in an oxidizing
atmosphere, Darco required more vigorous regeneration conditions to regain
its original capacity. For example, at 600°C in flue gas, Darco was not
regenerated to its full capacity. St. Regis activated carbon at these
same conditions underwent additional activation, increasing its capacity.
The original activated St. Regis carbon had a lower capacity than granular
Darco. The further activation performed during the 600°C regeneration step
and the resulting increase in loading capacity indicate that the original
St. Regis material may not have had a fully developed pore structure for
wastewater treatment applications.
The full regeneration of St. Regis carbon at conditions significantly
lower in reactivity than required for a present commercial carbon indicates
a possible energy advantage for the hydropyrolysis char-based carbon. The
lack of sufficient quantities of char and the difference between the fluid
bed used and a commercial multiple hearth make judgements as to yield on
regeneration impossible.
MULTIPLE HEARTH FURNACE TRIAL PRIOR TO PILOT PLANT START-UP
The laboratory work employed a fluid bed furnace. Pilot plant opera-
tion would employ a mulitple hearth furnace not available until later.
Arrangements were made to conduct experiments in a multiple hearth furnace
in the laboratory facilities of the pilot plant furnace supplier. The
experiments were designed to:
(l) Obtain a feel for the relationship between carbon performance in
the fluid bed versus a multiple hearth activation furnace. This
information would enable rapid prescreening of materials and con-
:; ditions using the fluid bed.
f, .
(2) Gain a knowledge of the testing and sampling procedures required
during a furnace run.
(3) Gain operator experience on a multiple hearth furnace to maximize
efficiency during pilot plant startup and operations.
15
-------
TABLE 1. PERFORMANCE OF ST. REGIS CARBON AND ATLAS DARCO OS MILL EFFLUENT
Pensacola No.
Carbon
Granular Darco (Fresh)
Granular Darco (Loaded)
Darco S-51 (Fresh)
Regenerated Granular Darco
Regenerated Granular Darco
Regenerated Granular Darco
Regenerated Granular Darco
St. Regis Carbon (Fresh)
St. Regis Carbon (Loaded)
St. Regis Regenerated
Granular Darco (Fresh)
St. Regis Regenerated
Temp.
-
-
600
850
450
600
_
-
600
_
850
Color
At
400
Atmos .
150
38
230
N2 170
N2 ~ 240
Pseudo Flue Gas 150
Pseudo Flue Gas 120
90
Keg.
Pseudo Flue gas 270
At
200
CU
90
N2 60
Loading
At
100
CU
60
£10
68
<10
30
<10
/v30
~-20
Neg.
56
At
100
CU
72
36
2 Mill Effluent
TOC Loading (mg/g)
130
mg/1
90
50
200
100
50
34
<10
120
25
300
130
73
90
mg/1
50
16
55
27
20
12
<1
15
/•J2
40
55
35
60
mg/1
23
r^ 5
10
5
11
/N/4
-------
It would have been most desirable to operate with hydropyrolysis char.
Unfortunately, the required 1000 kg of char was not available prior to pilot
plant start-up. As a substitute for hydropyrolysis char, a coke used by
some activated carbon producers was obtained.
The furnace operation consisted of five runs including three tempera-
ture levels and two residence times. The gas composition in the multiple
hearth was determined during each run. The conditions were duplicated in
the fluid bed using the same material and gas composition.
A comparison of fluid bed and multiple hearth data is shown in Figure 2,
providing a basis for preliminary selection of pilot plant operating condi-
tions. For a given raw material, carbon tetrachloride numbers (weight %
pickup of CC1, vapor) and bulk density values were found to be quick
indicators of the furnace operation.
The operator exposure to the multiple hearth furnace during these runs
proved extremely beneficial during later pilot plant work.
17
-------
ao
"B
w
u
1200
1000
800
600
400
2 HOUR RETENTION
MULTIPLE
HEARTH
I I I 1 i I
800
900
1000
o
2100
isno,
1500
1200
900
600
4 HOUR RETENTION
MULTIPLE
HEARTH
I I
J I
800 900 1000
TEMPERATURE (°C)
Figure 2. Surface area vs. activation temperature.
18
-------
SECTION VI
PILOT PLANT DESIGN
DESIGN PROCEDURE
Following a review of the results of laboratory and prepilot operations
on October 26, 1971, EPA authorized St. Regis to proceed with Part II of the
grant. Pilot plant design efforts began in January, 1972.
A etecision was made to use an outside engineering company to design the
pilot plant. After preliminary screening, six companies were invited to
submit proposals and a contract was awarded in May, 1972. At this time, the
engineering contractor was given detailed process flowsheets, material and
energy balances, equipment criteria, physical properties of process streams,
alternate processing routes, and alternate equipment considerations.
Preliminary cost estimate was considerably higher than anticipated.
In order to stay within budget limitation, it was necessary to adopt a
number of compromises and simplifications. Major among these was the
decision to settle on a single heat exchanger for regenerative heat recovery
and a single stage of pressure let-down. The changes made during the design
phase will become apparent by comparison of "Description of Process for
Design Purposes.", which appears later in this section of the report, and
"Description of Plant" in Section VII.
BASIS FOR PRELIMINARY ENGINEERING AND SELECTION OF CAPACITY
The absence of commercial or pilot plant experience with hydropyrolysis
char, an untried material outside the laboratory, made it impossible to ac-
curately size elements of the char preparation and activation system. This
lack of experience was due to the unavailability of sufficient char to per-^
form test work on production equipment. This fact led to two decisions:
(1) Nft) attempt should be made to integrate the dry end processes of drying,
blending, compaction, and activation into a continuous operation, and (2) N;O
requirement should be imposed which required the dry end of the process to
integrate with the wet end in the pilot plant operations. The dry end would
then be a batching operation through the several unit operations and opera-
ting manpower requirements would be reduced because the wet end and the dry
end would not operate at the same time.
Relieved of the requirement of operating as an integrated drying, com-
paction, and activation train, each component of the dry end was sized at the
minimum level for accurate scale-up:
19
-------
(1) After discussions with several companies which deal in the design
of multiple hearth furnaces it was decided that a 0.762 m(30 in)
I.D. unit was the minimum that should be installed.
(2) The compactor, a 136 kg (300 Ib) per hour minimum machine, would
have the greatest capacity of the three components. This size was
the smallest available using a commercial configuration. Other
units of a laboratory nature with lower capacities were available
but were not considered suitable for scale up needs. For lack of
better information the compactor was to be designed assuming a
carbon black as feed.
(3) The dryer was specified to maintain adequate feed for the expected
run time on the dry end.
For wet end (hydropyrolysis char production) design, 2.5 cm'(1 in )
piping was considered to be the practical minimum in the reactor and heat
exchange areas. This fact, combined with the flow velocities required to
maintain char in suspension, dictated a nominal feed rate of 0.63 J./sec
(10 gpm).
DESCRIPTION OF THE PROCESS FOR DESIGN PURPOSES
The flowsheet given to the engineering contractor as a basis for the
preliminary pilot plant design is shown in Figure 3. The process can be
broken up into basic unit operations, similar to what was done in con-
nection with Figure 1 in Section IV. These unit processes will be dis-
cussed first for the wet end, then for the dry end of the pilot plant.
Wet End
Feed—
The feed to the hydropyrolysis process should be extracted from the mill
evaporator system after separation of tall oil soap. At this point in the
evaporator system, black liquor has reached approximately 25% solid concen-
tration. Storage at the mill source of black liquor should be sufficient
to provide at least two days supply to the pilot plant.
From the storage tank in the mill, the liquor should go directly to
holding tanks at the pilot plant. These holding tanks can be used for the
gas recycle operation. For the pilot plant design, the gas recycle opera-
tion will be simulated by absorbing gases from cylinders into the liquor,
using some type of in-line mixing device. The gas will be injected into the
liquor while recirculating the contents of the holding tanks.
Pressurization--
This operation, being a comtinous pressurization process, can be best
accomplished through the use of some type of pump. Possible types include a
staged-centrifugal, reciprocating piston, or reciprocating plunger pump.
20
-------
HEAT RECOVERY
BLACK LIQUOR^
FROM SOAP SKIMMER
STORAGE
g
4.
L
5ATMENT
1
PRETREATMENT
*•— GAS
HIGH
PRES-
SURE
PUMP
COMPACTION
ACTIVATED
CARBON
TO STORAGE
VAPOR
WWV
DRYER
PRESSURE
LETDOWN
WASHED
CHAR
WASHER
FILTER
HEATING
WASH
STORAGE
I
FILTRATE
STORAGE
HEATING
&
REACTION
TO CAUSTICIZING
Figure 3. Flow diagram for preliminary pilot plant design.
-------
Heat Recovery—
In order to discuss this operation, since it will involve the recovery
of energy from the hot slurry stream after reaction, some mention of the
reaction and pressure letdown operations will have to be made prior to
discussing them in detail. The heat recovery section should consist of two
heat exchangers. The first exchanger (the order of numbering is the same as
the order of temperature, i.e., as the number of the exchanger increases,
the average temperature in the exchanger increases also) shall heat the
treated black liquor by a countercurrent thermal contact with the hot slurry
from the first pressure letdown and flash separation (the order of numbering
here is the reverse of the above). The second exchanger shall heat the
treated liquor by a countercurrent thermal contact with the hot vapor from
the first pressure letdown and flash separation. The two heat exchangers
can be of any common industrial type, i.e., shell and tube, double pipe,
etc. The design of these exchangers is complicated by the fact that streams
exchanging energy are under high pressures, at high temperatures, consist of
multiphase mixtures, and have a complex chemical composition.
Heating and Reaction—
This operation should provide for heating the liquor leaving the second
heat exchanger of the heat recovery section to the desired reaction tempera-
ture and provide for the heat of reaction. The operation should provide
also for the requirements of the hydropyrolysis reaction. The heater will
provide the temperature requirement, but the reaction time requirement will
have to be provided by another piece of equipment: the reactor. The reactor
can be any of the common industrial types, e.g., tubular, continuous stirred
tank, etc. The heater can also be any of the common industrial types, or the
reactor and heater could be combined into the same unit. As it was in the
case with the exchangers in the heat recovery section, the design of the
heater and reactor will be complicated by the high process pressure, high
process temperature, multiphase mixture, and complex chemical composition.
Pressure letdown—
The slurry product from the reactor should be cooled.below 95°C in order
to prevent excessive flashing of water when reducing the process pressure.
The slurry product is cooled by the heat exchangers in the heat recovery
section. The pressure letdown operation occurs in two steps. The first
pressure reduction occurs on the slurry stream immediately as it leaves the
reactor. The second pressure reduction occurs on the slurry leaving the
first heat recovery exchanger. The slurry at this point should be at a
temperature less than 95°C. ;
Flash Gas Separation—
The flash gas separation operation occurs in two steps. The first step
occurs immediately after the first pressure letdown. The gas which is sepa-
rated is used as the heating medium in the second heat recovery exchanger.
As this vapor is cooled, the water which it contains is condensed. The
second step occurs after the vapor stream has been cooled, has left the
second heat recovery exchanger, and has undergone another pressure re-
duction. The actual flash gas (non-condensible portion) is separated in
22
-------
this step. The condensate, or liquid which leaves the second step of flash
gas separation, is used for washing the filter cake.
Flash Gas Combustion and Recycle—
The flash gas leaving the second step in the flash gas separation is
combusted in a waste gas incinerator. In the pilot plant no attempt will
be made to recycle the combusted flash gas. Recycle will be simulated by
S02 cylinder gas.
Liquid-Solid Separation and Char Washing—
The liquid-solid separation operation is best accomplished by filtra-
tion. The filtration can be carried out by using any of the common indus-
trial filtration equipment, e.g., rotary vacuum, belt, extractor, or pressure
leaf filters. The liquor which is separated from the slurry (filtrate) is
sent to a storage tank and from there to the caustic plant in the kraft
mill. The char from filtration goes to washing where any entrained filtrate
is washed away. Again, the best way to accomplish this operation would be
with industrial filtration equipment. The wash liquor goes to storage and
from there to the caustic plant. The washed char will be collected and
stored in some type of portable equipment.
Dry End
Drying—
The washed char from the wet end must be dried before the operations
of compaction and activation can be performed. Any common industrial dryer
can be used for this purpose, e.g., batch vacuum, continuous direct fired
or continuous indirect fired dryer,
Compaction—
After drying, the char is compacted to a certain particle size and is
then ready for activation. The compactor, for lack of better information, will
be designed assuming a carbon black feed.
Activation—
The activation will be carried out on granular rather than powdered
char. The type furnace desired is the Herrshoff, or multiple hearth. After
activation, the carbon will be stored in some type of portable equipment.
Alternate Processing Routes and Equipment Consideration
Several alternate processing routes, which consisted of modifications
made to the flowsheet as shown in Figure 3, were suggested to engineering
contractor for consideration. Also, several equipment designs or configura-
tions were suggested for his consideration. Due to the proprietary nature
of these matters, they will not be discussed in any detail.
£as,is for the Detailed Design and—P-reparatioft of the %iet•-Package
As mentioned earlier, the pilot plant cost indicated by the preliminary
estimate was much higher than had been anticipated. Therefore, steps were
taken to reduce the cost of the pilot plant. This was achieved by modifying
23
-------
the basic flowsheet as shown in Figure 3 and eliminating the expensive
alternate processing routes. The equipment selected for each operation was
chosen as the least expensive item that still allowed for operation of the
pilot plant as an experimental data gathering unit. Simplified instru-
mentation, controls, and piping were used in as many places as possible in
order to reduce costs.
A major process revision consisted of removing the first pressure let-
down and flash gas separation step. The revised process flowsheet is shown
in Figure 4. This flowsheet shows the process upon which the detailed de-
sign and bid package were based. It should be noted that with the removal
of the first pressure letdown and flash gas separation units, some heat ex-
change capacity was sacrificed in the heat recovery area. Because of this,
a heat exchanger to cool the slurry below 95°C was added upstream from the
pressure letdown unit.
24
-------
BLACK LIQUOR.
FROM SOAP SKIMMER
NJ
ui
STORAGE
ACTIVATION
PRETREATMEN1
*-
PRE
TREATMEI
GAS
HIGH
PRES-
SURE
PUMP
HEAT RECOVERY
COMPACTION
ACTIVATED
' CARBON
TO STORAGE
VA
FOR
wwv
DRYER
HEATING
HEATING
&
REACTION
TO CAUSTICIZING
Figure 4. Flow diagram for detailed pilot plant engineering.
-------
IXJ
ON
\m
1
1 —
1
1 1
1 1
1 1 j 1 1 j J I — '.
JL " il_ 11- „ .-.11-
11 11 if . JL
i
i
1 _... J.
r._7_ zi
V\ i
K\ /
A&T
•&F¥\
R/ x
41.+J/
Figure 5. Pilot plant layout drawing.
-------
SECTION VII
PILOT PLANT CONSTRUCTION AND DESCRIPTION OF PLANT
SUMMARY OF THE CONSTRUCTION PHASE
Six construction companies, which previously had expressed an interest
in bidding on the job, were provided with the bid package prepared by the
design engineers and were invited to bid. Bids were received from five of
these on March 1, 1973; the sixth company did not bid. Bids were analyzed
and it was recommended that a contract be awarded to the low bidder. A
contract was signed in April. Construction began in May, 1973 and was
completed in May, 1974.
LAYOUT AND PILOT PLANT EQUIPMENT DETAILS
Figure 4 (Section VI) showed the pilot plant basic flow sheet. Figure
5 is a layout drawing of the plant showing the arrangement of major equip-
ment installed. These two figures should be used for orientation in con-
nection with the plant description below. Also, Figure 6, a general view
of the pilot plant from the north, will help the reader to visualize the
arrangement in the layout drawing.
Figure 6. General view of pilot plant.
27
-------
In the foreground, the cross exchanger, reactor preheater, reactor, and
slurry cooler can be seen very easily. Feed pretreatment and the high
pressure feed pumps are toward the right rear of the picture. The filter-
washer and dry end equipment are hidden to the rear in this view.
Black Liquor Storage
The equipment in this area consists of the black liquor storage tank
(T-l), storage tank heater (H-l), and black liquor transfer pump (P-l).
(Refer to Figure 5 for equipment designation). This equipment is not
located in the pilot plant, but is near the evaporators in the No. 2 pulp
mill approximately 100 m west of the pilot plant. This area is connected
with the pilot plant by a 10 cm (4 in ) carbon steel transfer pipe.
The black liquor storage tank is a 2.74 m (9 ft) diameter by 3.20 m
(10.5 ft) high vertical, cylindrical vessel with a 18.9 m (5000 gal)
capacity. The tank is fitted with a 0.46 m (1.5 ft) high self supporting
conical head. It was designed to operate at atmospheric pressure and to
hold black liquor at temperatures in the range 65 tp 95°C. The tank is con-
structed of 0.64 cm (1/4 in.) carbon steel plate and insulated. The tank
is connected to the fourth effect evaporator inlet and outlet so 25% or 30%
black liquor sol.ids can be drawn off. The purpose of this storage tank is
to allow the pilot plant to operate independent of evaporator operation.
The storage tank heater is mounted on the black liquor storage tank
for the purpose of keeping the black liquor hot. The heater is a 7.5 kw
unit with an immersion element.
The black liquor transfer pump is a ductile iron centrifugal with a
5.59 kw (7.5 hp) motor. It was sized to deliver 4.73 I/sec (75 gpm) when
pumping black liquor at 82°C and 31.5 kkgf/tn (45psig) discharge pressure.
Black Liquor Pretreatment
The equipment contained in the pretreatment area consists of the black
liquor feed tanks (T-2A and T-2B), tank heaters (H-2 and H-3), black liquor
feed pumps (P-2A and P-2B), and the inline mixer (M-5).
This area of the pilot plant is pictured in Figure 7. The inline mixer
can be seen at the extreme lower left. The two large cylindrical vessels
are the black liquor feed tanks. One of the black liquor feed pumps can be
seen in the lower middle section of the figure. Some of the gas cylinders and
the manifolds to which they are attached can be seen in the lower right hand
portion of the picture.
The black liquor feed tanks are each 1.68 m (5.5 ft) diameter by 3.66 m
(12 ft) high vertical* cylindrical vessels with dished heads. The capacity
of each tank is 7.6 m (2000 gal ). The design pressure and temperature
were 3.5 kkgf/m (5 psig) and 65 to 95°C. The material of construction is
0.95 cm (3/8 in ) carbon steel plate. Each tank is insulated. The 10 cm
(4 in ) pipe from the black liquor storage area is connected to each of the
tanks. Each tank is connected to the inline mixer.
28
-------
Figure 7. Black liquor pretreatment equipment-
Each of the black liquor feed tanks is equipped with a tank heater.
Heater H-2 is a 66 kw immersion type-unit; heater H-3 is a 4.5 kw
immersion-type unit.
The black liquor feed pumps are ductile iron centrifugal pumps with
4.73 kw (7.5 hp) motors. The design capacity was 3.15 I/sec (50 gpm) when
pumping black liquor at 82°C and 39.2 kkgf/m (56 psig) discharge pressure.
The pumps circulate liquor from the tanks T-2A and T-2B through the inline
mixer and back to T-2A and T-2B. These pumps also feed treated liquor to
the high-pressure pumps P-4A and P-4B.
The inline mixer is a device used to blend components in the process
pipe line. The mixer fits directly into the pipe line. It has a mixing
chamber with about 7.6 1 (2 gal ) capacity. The liquid from the process
pipe line flows in one end of this chamber and out the other. In the
center of this chamber is an agitator paddle. The gas is injected into the
bottom of the chamber. The agitator is driven by a 0.75 kw (1 hp) motor.
The material of construction is stainless steel.
This operation simulates the recovery of combusted product gas from the
process. Normally, only the sulfur dioxide component of the product gas is
simulated but the system also provides for nitrogen, oxygen, and carbon
dioxide gas feeds.
29
-------
Pressurization
The equipment contained in the pressurization area consists of the high
pressure feed pumps (P-4A and P-4B), low pressure pulsation dampener (V-l),
and high pressure pulsation dampener (V-2).
M
Figure 8. Pressurization equipment.
The actual area in the pilot plant is show in Figure 8. The two pumps
are clearly visible. Two pipe manifolds are seen in the lower portion of
the picture. The upper manifold is the pump inlet and the lower manifold
is the pump discharge. The low pressure pulsation dampener is the cylin-
drical vessel in the central portion of the picture. The high pressure
pulsation dampener is not visible in this picture, but it can be seen in
a picture showing the reactor preheater (Figure 11).
Both of the high pressure feed pumps are reciprocating triplex plunger
pumps with 22.4 kw (30 hp) motors and variable speed drives. The pump
heads, being in contact with the process fluid, are 316 stainless steel. In
the pump P-4A, the valves are hardened stainless steel but in P-4B, the
valves are normal stainless steel. The pumps discharge into a 316 stainless
steel 2.5 cm (1 in ) Schedule 80 pipe manifold.
-------
The low pressure pulsation dampener is located on the line that feeds
the high pressure feed pump inlets. Its purpose is to reduce the liquid
pulsation pressure in the suction line of the high pressure plunger pumps.
This pulsation dampener is a cartridge type which uses nitrogen gas for
stabilization. It has a volume capacity for 4.9 1 (300 in ) of nitrogen
gas.
The high pressure pulsation dampener is located on the line leading
from the discharge of the high pressure feed pumps. Its purpose is to re-
duce the liquid pulsation pressure in the discharge line.
Heat Recovery
The equipment contained in the heat recovery area consists of the cross
exchanger (H-4). A picture of the cross exchanger .is shown in Figure 9.
This picture shows the cross exchanger as viewed from the front of the pilot
plant at about waist level. An additional view of this large piece of
equipment can be seen in Figure 6.
Figure 9. Cross exchanger.
The cross exchanger is a double pipe heat exchanger with 5 hairpins.
The inner pipe is 2.5 cm (1 in ) Schedule 80; the outer pipe is 5 cm (2 in )
Schedule 160. Both are 316 stainless steel. The cross exchanger was de-
signed so that the inner pipe can be removed for cleaning.
y-
-------
The cross exchanger heats the feed stream to a temperature approaching
the desired reaction temperature by regenerative heat exchange with the pro-
duct stream, thus cooling the product stream and achieving heat.economy.
Heating and Reaction
The equipment contained in the heating and reaction area consists of
the reactor preheater (H-6) and reactor (V-3). Views of the reactor pre-
heater and reactor from the front of the pilot plant are shown in Figures
10 and 11. The cyclindrical vessel in the upper middle portion of Figure
11 is the high pressure pulsation dampener.
Figure 10. Heating and reaction equipment.
The reactor preheater is a convection type, non-radiant heater which
utilizes natural gas as fuel. The process fluid makes thermal contact with
the heating medium by flowing thru 316 S.S. 2.5 cm (1 in ) Schedule 80 pipe.
The reactor is an adiabatic tubular reactor. The actual tubular portion
of the reactor is constructed of 316 S.S. 3.8 cm (1.5 in ) Schedule 80 pipe.
The reactor is made adiabatic by an insulated rectangular box around the
pipe bundle.
32
-------
Figure 11. Heating and reaction equipment.
Pressure Let-down
The equipment contained in the pressure letdown area consists of the
slurry cooler (H-7) and the pressure let-down valve (PCV-64). A picture of
this area is shown in Figure 12. The pressure let-down valve can be seen in
the middle left portion of the picture. The slurry cooler is seen in the
right hand portion of the picture.
The slurry cooler is a heat exchanger used to cool the slurry from the
cross exchanger below 95 C. This is necessary before the pressure let-down
operation can be carried out; otherwise the water in the slurry would flash.
The slurry cooler is a double-pipe, six-hairpin, heat exchanger with slurry
flowing through the inner pipe and water flowing through the annular space
between the inner and outer pipes. The inner pipe is 316 S.S. 2.5 cm (1 in)
Schedule 80 pipe, while the outer pipe is a carbon steel 5 cm (2 in)
Schedule 40 pipe.
33
-------
The pressure letdown valve is a 1.27 cm (1/2 in ) angle-body control
valve. The body is constructed of 316 S.S., and the trim is linear with C
of 1.5. The trim is constructed of stellite.
Figure 12. Slurry cooling and pressure let-down equipment.
Flash Gas Separation
The equipment contained in the flash gas separation area consists of the
slurry-gas separator (V-8). A picture of the slurry-gas separator is shown
in Figure 13. This is a view of the separator as one would see it when
standing next to the pressure let-down valve. The separator is the cylin-
drical vessel in the right hand portion of the figure.
The slurry-gas separator is a cylindrical vessel 0.32 m (12.75 in) in
outside diameter and 1.83 m (6 ft) in height. The vessel is constructed of
316 S.S. The design pressure and temperature was 3.5 kkgf/m^ (5 psig) and
95°C. The vessel is insulated. The separator is connected to the pressure
let-down valve. (316 S.S. 2.5 cm 1 in). Schedule 80 pipe was used for inter-
connecting all the equipment from the high pressure feed pump discharge to
34
-------
the separator inlet, The slurry which leaves the separator and flows to
the filtration area exits from the bottom of the separator by a 3.8 cm
(1.5 in ) carbon steel pipe. The gas outlet is at the top of the separator,
A 316 S.S. 5 cm (2 in ) Schedule 10 pipeline connects the separator gas out-
let with flash-gas combustion.
Figure 13. Slurry-gas separation equipment.
Flash Gas Combustion
The equipment contained in the flash-gas combustion area consists of
the afterburner (M-22) shown in Figure 14.
The afterburner is a self-supporting, natural-gas-fired, incinerator
for gases that are produced in the hydropyrolysis reaction. Its purpose is
to oxidize the gas products from the slurry-gas separator and the filter
system vacuum pump discharge. The unit is failure safe, operating with a
positive ignition pilot system to maintain a continuous flame.
35
-------
Figure 14. Flash gas combustion equipment.
Liquid-Solid Separation and Char Washing
The equipment contained in the liquid-solid separation and char washing
area consists of the filter feed tanks (T-7A and T-7B), filter feed tank
agitators (M-18A and M-18B), filter feed pump (P-6), filter system (M-8),
filter beltwash pump (P-7), filtrate storage tanks (T-12A and T-12B), con-
centrated wash storage tank (T-13), and filtrate-concentrated wash pump
(P-ll). A picture of the filter system as viewed from the south end of the
pilot plant is shown in Figure 15. In this figure, the belt drum drive,
cloth discharge roll, and vacuum manifold can be seen clearly. In the upper
left hand corner of the figure, the slurry-gas separator can be seen. In
the middle left hand portion of the figure, the filter feed tank (T-7B) can be
seen.
-------
Figure 15. Filtration and washing equipment.
The filter feed tanks T-7A and T-7B are identical. Each tank is a 1.23m
(4 ft) diameter by 1.68 m (5.5 ft) high cylindrical-vessel with 0.30 m (1 ft)
deep conical bottom. The capacity of each is 1.9 m (500 gal ). The
material of construction is carbon steel and each tank is insulated.
The filter feed pump (P-6)Jis a ductile iron centrifugal pump with a
0.56 kw (3/4 hp) motor. It is connected to the filter feed tank recycle
lines and the filter system feed line by a 2.5 cm (1 in ) carbon steel mani-
fold. The filter belt wash pump (P-7) is a cast iron centrifugal self priming
pump with a 0.37 kw (1/2 hp) motor.
The filter feed tank agitators M-18A and M-18B are 26.7 cm (10.5 in)
propeller type agitators with a single impeller. The shafts are carbon
steel and the propellers are stainless steel. The agitators have 0.56 kw
(3/4 hp) motors with variable speed drives. They operate from 0 to 350 rpm.
37
-------
The filter system M-8 is a top feed extractor belt filter which con-
tinuously dewaters and washes the hydropyrolysis char. The filter is
approximately 6.1 m (20 ft) long with a 0.30 m (1 ft) wide transmission
belt. This provides about 1.67 m2 (18 ft2) of active area. The unit is skid
mounted with all accessories including controls, control panel, motors,
safety devices (overload protection, branch circuit fuses,etc.), and control
voltage transformer prewired to single end point connections. Piping be-
tween individual components and accessories is provided on the skid. All
structural elements are carbon steel.
The filter transmission belt is nylon cord with neoprene external
cover and has a tensile rating of 37.5 kgf/cm (210 Ib/in ) of width. Stain-
less steel fasteners are provided for joints. The drainage grids are poly-
propylene with steel grid fasteners. The edge seals are neoprene.
The vacuum box is made of mild steel with four compartments and four
outlets. The cloth belt filter medium is nylon and has a stainless steel
clipper lace. After cake discharge, two spray pipes are provided for
washing the cloth. The spray pipe nozzles are V-jet type on 6.35 cm
(2.5 in ) centers. A flbergla^rs reinforced polyester wash collecting pan is
provided for the full length of the filter.
The extractor top feed assembly is composed of four steel overflow weir
boxes: one for the hydropyrolysis slurry feed and three for the wash water
feed to the wash stages. The hydropyrolysis slurry feed box has a neoprene
wiper material and special side dams and seals for maintaining a slurry pool
at the feed end. The filter has a 1.49 kw (2 tip) motor with variable speed
drive and a heavy duty enclosed worm gear reducer. The belt speed ranges
from a minimum of 0.91 m/min (3 ft/min) to a maximum of 4.57 m/min (15 ft^iin) .
The following accessories complete the filter assembly:
(1) Four 0.36 m (14 in ) diameter by 1.22 m (4 ft) high cylindrical
vacuum pump receivers made of carbon steel reinforced for vacuum service.
o
(2) One cast iron water seal vacuum pump with 10.2 m /min (360 cfm)
capacity at 559 mm (22 in ) of Hg vacuum and 1170 rpm. The vacuum pump
motor is 18.64 kw (25 hp) and 1800 rpm.
(3) Four self-priming, centerline discharge, ductile iron, centrifugal
filtrate pumps. The pump motors are 1.12 kw (1.5 hp) at 1750 rpm.
The filtrate storage tanks T-12A and T-12B are each 1.23 m (4 ft)
diameter by 1.68 m (5.5 ft) high cylindrical vessels with 1.9 m3 (500 gal )
capacity. The tanks are carbon steel, insulated, and have flat tops. Both
tanks are connected to the filtrate pump under the first suction box of the
filter.
The concentrated wash storage tank T-13 is a 1.52 m (5 ft) diameter by
1.68 m (5.5 ft) high cylindrical vessel with 2.84 m3 (750 gal ) capacity.
The tank is carbon steel, insulated, and has a flat top. It is connected
to the filtrate pump under the second suction box of the filter.
38
-------
The filtrate-concentrated wash pump(P-ll)is a ductile iron centrifugal
pump with a 5.6 kw (7.5 hp) motor. The pump transfers filtrate from tanks
T-12A, T-12B, and T-13 to the No. 2 mill green liquor clarifier.
Drying
Equipment in the drying area consists of the vacuum cake dryer (M-10).
Figure 16 is a picture of the dryer.
Figure 16. Char drying equipment.
The pilot equipment used to dry hydropyrolysis char is a Stokes Model
//59 vacuum dryer. This system is a rotary batch dryer consisting of a
stationary cylinder with a rotating central shaft. The capacity, running
half full, is 0.283 m (10 ft ). Welded to the shaft are spiral agitator
and scraper blades. The cylinder is jacketed to receive a heating medium,-
which is steam for the pilot plant. Shell working pressure is 87.9 kkgf/m
(125 psig) with full vacuum. Maximum temperature rating is 343°C.
, .
-------
The central agitator shaft, but not the blades, is heated by steam.
Total heating surface in the shell and shaft is 3.76 m2 (40.5 ft2). The
agitator, driven at 6 rpm by the 1.12 kw (1.5 hp) drive, continuously mixes
and distributes the material over the heated surface of the cylinder.
The dryer is charged through the loading port at the top. When the
drying is complete, the bottom discharge port is opened and the large outer
spirals rapidly sweep the material out of the dryer. Cooling water may be
passed through the jacket to reduce char temperature before discharge.
Water vapor removed from the char filter cake in the dryer cylinder passes
through a vacuum line and enters a shell and tube condenser.
The pilot plant batch dryer is only a tool to produce dry char for
subsequent operations. Although the unit can provide some drying rate data,
it does not provide full information for selection of production units.
When needed, this can be accomplished by testing char samples in dryer
vendor facilities.
Blending
The design objective of the dry end of the pilot plant is to produce a
granular activated carbon. Hydropyrolysis char is produced as a fine powder
and requires compaction to attain the desired particle size. In order for
some carbonaceous materials to attain sufficient strength during the compac-
tion to withstand further processing, a binder is sometimes required. There-
fore, the plant was equipped with a drum rotator to blend char with a
powdered binder. Additional blending equipment was not considered required
initially as units were available on a rental basis if necessary.
Early in the pilot operation it was determined that the dryer with its
spiral agitators could be efficiently utilized as a blender. Thus the
dryer acquired two duties: a batch dryer and batch blender.
Compaction System
The compaction system (M-ll) is a unitized assembly of a mixer, feed
screw, press rolls, granulator, and screen pack, with associated conveying
equipment, manufactured by K. G. Industries (#5274-25CS). Figure 17 is a
picture of the assembly and Figure 18 is a schematic drawing of the
arrangement.
Feed char is contained in a hopper and flow to the compactor is con-
trolled with a slide valve. The raw feed passes through a mixer where it is
blended with recycled fines. The mixer is of the paddle type. The paddles
are angled to push, the material toward the roll feed screw. Some paddles
have reverse angle to back mix the char for better blending. The mixer is
driven at a constant speed of 29.2 rpm by a 7.46 kw (10 hp) motor operating
through a chain drive and speed reducing gearbox.
40
-------
Figure 17. Compaction equipment.
At the end of the mixer, the char drops into the feed screw where the
entrained air is pressed out and the char is predensified for optimum press
roll operation. The feed screw is highly loaded and is a one piece forging.
Speed is variable from 30 to 295 rpm by means of a varispeed drive operating
off a 2.24 kw (3 hp) motor.
To accomplish compaction, the predensified material passes between two
rolls. These rolls are machined from 440 stainless steel and have cor-^
rugated faces. The roll loading force may be varied from 0 to 13,600 kg
(30,000 Ib). This force is developed with a hydraulic pump and a hydraulic
cylinder and lever arm on each roll. An accumulator is included in the
hydraulic system to maintain pressure while compensating for variable char
-------
COMPACTOR
ROLLS
Figure 18. Compactor system schematic.
42
-------
loading on the rolls. Both rolls are positively driven at 10.5 rpm by a
3.73 kw (5 hp) motor through a gearbox. Roll speed is not variable but can
be changed by modifying the gearing in the drive.
The corrugated rolls produce a wavey ribbon of compacted char. This
material falls through a chute and is pushed into the granulator knives by
means of a feed screw. The knives and screw operate on the same shaft and
are Driven by a 2.24 kw (3 hp) motor. The char is broken up by the knives
until it can pass through the granulator screen spinning around the knives
and in opposite direction to the blades. The granulator screen is driven
through a variable speed drive of 0.75 kw (1 hp) .
Granulated char is transported by means of a bucket elevator powered by
a .37 kw (1/2 hp) motor to a Sweco screen. In this screen, the oversize is
removed and recycled back to the granulator, and the -7 + 20 product is re-
moved and stored in drums. The fines are recycled back to the mixer to be
blended with the raw feed char.
Multiple Hearth Furnace System
The furnace system (M-12) includes a multiple hearth activation (or
calcining) furnace, associated firing controls, an afterburner, flue gas
scrubber and stack assembled by EIMCO-BSP; Envirotech Systems Inc.
The multiple hearth furnace is pictured in Figures 19 and 20. The
furnace consists of six circular hearths of cast refractory construction,
stacked vertically. The hearths are 0.762 m (30 in.) in diameter. Figure
19 is an outside view of the furnace from the firing side. Gas pressure
safety devices are at the bottom of the picture. Five sets of gas burner
controls and steam injection controls are shown, each serving one of the
hearths. Figure 20 is a view of the hearths with the furnace open and
the rabble arm assembly removed. Flue gas exhaust ports are visible.
The furnace is fed from a 304 stainless steel hopper above the furnace.
Feed rate is controlled by a calibrated 10 cm (4 in ) diameter variable
speed screw conveyor. Material passes from the screw directly into the
top of the furnace.
Compacted char is dropped onto the periphery of the top hearth, called
an "in" hearth, and is rabbled toward the center where it drops to the
hearth below, called an "out" hearth. The material is then rabbled toward
the outside of this hearth and drops through holes to the hearth below- A
vertical rotating shaft through the center of the hearths carries 4 rabble
arms per hearth with 5 rabble blades per arm. These blades stir the charge
and move it in a spiral path across each hearth. The arms and shaft are 310
stainless steel to resist high temperatures. Commercial units, which contain
much larger rabble assemblies, have air cooled arms and in severe duty the
arms are insulated. The pilot furnace, owing to its small size, includes
neither of these options.
43
-------
Figure 19. Multiple hearth furnace
firing controls-
Figure 20. Furnace open for
inspection-
The feed passes in and out across the six hearths until it is discharged
from the sixth and lowest hearth. The carbon is removed from a chute be-
low the bottom hearth by a 10 cm (4 in ) diameter screw conveyor which is
water jacketed and contains a water spray. The jacket insures minimum
metal temperatures and cools the hot carbon by conduction. Rapid and com-
plete carbon cooling is insured by the spray so that it may be discharged
safely. Secondly, the water vaporized by the hot carbon creates a steam
blanket and positive pressure in the screw area, lessening the possibility
of atmospheric oxygen entering the furnace. This would cause uncontrolled
burning of the carbon. All parts in contact with the hot carbon are 304
stainless steel.
44
-------
Five burner systems are included with the furnace, firing the lower
five hearths. This allows maximum flexibility for developing scale-up data.
It may not be necessary to fire all of the lower hearths in a commercial
installation.
The lower five hearths of the furnace may be fired independently or
receive steam injection. Heated gases flow countercurrently to heat the char
to reaction temperature and carry on the activation or calcination reactions.
Burners are of the nozzle mixing type, wherein the fuel and air are mixed
intimately as they enter the burner tunnel. This design prevents flash
backs and intraburner explosions. The burners are stable from very high
excess air ratios up to 50% excess fuel.
Auxiliary Equipment
Water Supply—
The equipment in this area consists of the water pump suction tank
(T-3) and the water supply booster pump (P-14) . The water pump suction tank
is a 0.61 m (2 ft) diameter by 0.91 m (3 ft) high cylindrical vessel. The
water supply booster pump is a ductile iron centrifugal pump with a 14.9 kw
(20 hp) motor. It can deliver 11 I/sec (175 gpm).
Steam Boiler—
The steam boiler (M-9) is used primarily to provide steam for the multi-
ple hearth furnace system and dryer; however, small amounts are used to make
hot wash water for the filter system and to vaporize the liquid S0» from the
gas cylinders. The steam boiler is an automatic electric type steam genera-
tor capable of operating in the range of 3.5 to 56.3 kkgf/m2 (5 to 80 psig)
with a capacity of 98 kw (10 boiler hp).
Air Compressor—
The air compressor unit (M-19) supplies the air for the pneumatic in-
struments and for any pneumatically powered equipment in the pilot plant.
The compressor has a capacity of 14.16 £/sec (30 scfm) and operates at a
receiver outlet pressure of 56.3 kkgf/m (80 psig). The compressor unit
has a horizontal air receiver and is a two cylinder, single stage, air
cooled unit. A water aftercooler to cool the air to a temperature 5.5°C
above the cooling water inlet temperature is provided on the compressor.
Instrument Air Dryer—
The instrument air dryer (M-20) is used to remove moisture from the
compressed air prior to its use in the pneumatic instruments. The air
dryer is designed to operate at a flowrate of 14.16 I/sec (30 scfm) and at
inlet conditions of 70.3 kkgf/m2 (100 psig) air pressure. The dryer is an
electric reactivated dual tower using desiccant for drying. The desiccant
is silica gel.
-------
SECTION VIII
OPERATING PROCEDURE
INTRODUCTION
This section describes pilot plant start-up procedure, normal operating
procedure, sampling and testing schedule, and the operating schedule. A flow
diagram of the pilot plant was shown in Figure 4. This flow diagram and the
pilot plant layout shown in Figure 5 should be used for reference in the
discussion which follows.
WET END START-UP
The wet end is started up on water. The water pump suction tank (T-3)
is filled and the water supply booster pump (P-14) started. This supplies
water to the suction of the high-pressure feed pump (P-4). After checking
that all upstream and downstream process valves are in their proper position
and that the pressure letdown valve (PCV-64) is activated (i.e. fully open),
P-4 is started. The variable speed drive selector on P-4 is then set at
the position which will deliver the flowrate desired for the run conditions.
The control for PCV-64, which is always in the manual mode when the wet end
is being started up, is adjusted until the system presure rises to the
level desired for the run conditions. When the pressure has become stable,
the control for PCV-64 is switched to the automatic mode. At this point,
the water is flowing from P-4 through the crossexchanger (H-4) tubeside, to
the reactor preheater (H-6), to the reactor (V-3), to the crossexchanger
shellside, to the slurry cooler (H-7), to PCV-64, and then to the slurry-
gas separator (V-8). During start-up, the water from V-8 is channeled to
the trench. After checking that all valves in the line extending from V-8
to the afterburner (M-22) are in their proper position, the after burner is
fired off. Next, cooling water is introduced to the shell of the slurry
cooler (H-7). At this point, the reactor preheater (H-6) can be started.
It has a temperature control unit which controls the process temperature of
the fluid exiting H-6. During start-up, the temperature control unit is
placed in the manual mode and left there while the system heats up. When
the temperature of the process fluid leaving H-6 is within about 15£ of the
desired run temperature, the temperature controller is switched to the auto-
matic mode. When the temperature reaches the desired level and becomes
steady, the reaction system (this includes the crossexchanger, reactor pre-
heater, and reactor) is ready to process black liquor.
46
-------
The filter system (M-8) is started up on water. This is accomplished
by starting the vacuum pump and the drum roll which drives the transmission
belt, then feeding water thru the feed pump (P-6) to the weir box over the
No. 1 suction box. Water flowing_into the No. 1 vacuum receiver is discarded.
Next, water is fed to the filter cloth wash spray nozzles and the filter
belt-wash pump (P-7) is turned on. As the water sprays on the filter cloth,
it drops to the collection pan located under the filter. The pump(P-7)draws
all the water out of this pan and sends it to the feed box over the No. 4
suction box. The water flowing into this feed box is the wash for the char.
The wash rate for the run is controlled by the flowrate of water sent to the
cloth spray nozzles and is set for the desired run conditions.
The filtrate pumps on all four vacuum receiverstare now started. Since
the char washing operation is a three stage countercurrent process, the
water flowing into the No. 4 vacuum receiver is pumped to the feed box over
the No. 3 suction box. The water flowing into the No. 3 vacuum receiver is
pumped to the feed box over the No. 2 suction box. The water flowing into
the No. 2 vacuum receiver is discarded.
OPERATION ON BLACK LIQUOR
The black liquor storage tank (T-l) always contains enough black liquor
to supply the pilot plant for about two days. The storage tank heater (H-l)
maintains the temperature in T-l. The amount of black liquor in T-l is held
constant by a level sensing device which is connected to a flow control
valve. Whenever it is desired to operate the wet end, the black liquor
transfer pump (P-l) is started and the amount of black liquor needed for
the run conditions is pumped into the black liquor feed tanks (T-2A and
T-2B) . The tank heaters (H-2 and H-3) are used to keep the black liquor '
hot while it is in these tanks.
Pretreatment is a batch operation. There is adequate time during wet
section start-up to pretreat the contents of the feed tanks and, during
sustained operation, time to pretreat one tank while the other is feeding
the system. Black liquor is circulated through the inline mixer (M-5) using
either or both black liquor feed pumps (P2A and P2B) . Valves to the high
pressure feed pump (P-4) suction are closed during pretreatment. Sulfur
dioxide gas is added to the black liquor as it circulates through the inline
mixer (M-5). After pretreatment, black liquor circulation is continued until
the reaction system is ready to receive the liquor.
When the reaction system is ready, treated black liquor is switched to
P-4. The water feed to the P-4 suction is shut off and the valves on the
recycle line to T-2A and T-2B are closed. At this point, the black liquor
is feeding P-4 and being pumped through the reaction system. The water exiting.
the slurry-gas separator is discarded until the slurry appears. When the
slurry appears, the flow is switched to one of the filter feed tanks (T-7A
or T-7B). The flash gas will be flowing to the afterburner at this time.
When sufficient slurry has accumulated in T-7A or T-7B water to the
filter feed pump (P-6) is turned off and slurry is fed to P-6 which delivers
47
-------
slurry to the weir box over the No. 1 suction box on the filter system (M-8).
The slurry is separated at the No. 1 suction box.into filtrate and unwashed
char streams. Filtrate from No. 1 vacuum receiver is switched to the
filtrate storage tanks (T-12A and T-12B). At this point the vapor discharge
from the vacuum pump is checked to make sure that it is flowing to the
afterburner.
When the washed char appears at the discharge roll and begins to fall
from the cloth freely, a 208 1 (55 gal ) drum is placed there to collect the
cake. The filter will now run by itself, except for adjustments that have
to be made periodically to keep the discharge cake thickness uniform and
keep the wash over the No. 2, 3, and 4 suction boxes from flooding.
SAMPLING AND ANALYTICAL PROCEDURES
Samples are taken once during every run when the wet end is running
steadily. The following is a list of the process streams which are sampled
during a run;
black liquor before pretreatment;
black liquor after pretreatment ;
vapor from V-8;
slurry from V-8;
filtrate (liquor from No. 1 suction box);
u'hwashed char (char over No. 1 suction box);
concentrated wash (liquor from No. 2 suction box);
Washed char (char over No. 4 suction box); and
belt wash (liquid feed over No. 4 suction box).
The samples taken during each run are analyzed in the laboratory using
the procedures and analytical equipment described in Appendix A. The re-
sults of these analyses, coupled with the flowrate data taken during the
runs, allow the construction of material balances. The analytical data also
provide characterization of the product streams.
DRY END OPERATING PROCEDURE
The dry end operations of drying, blending, compacting, calcining, and
activating were all conducted independently with product storage between
operations. (Attempts to calcine on the upper hearths and then activate in
the same pass through the furnace encountered control problems. Thereafter,
the furnace was set up for either calcination or activation and these
operations were conducted separately.) Operating procedures should be
apparent from the equipment functional descriptions in Section VII and
should require no amplification in this Section.
OPERATING SCHEDULE
The pilot plant operating schedule, as shown in Figure 21, spans a
period of 16 months. The pilot plant activities, as covered in this report,
began in June 1974 and were completed in September 1975. The schedule can
48
-------
WET END SHAKEDOWN
WET END OPERATIONAL TESTING
WET END CHAR PRODUCTION
DRY END SHAKEDOWN
DRY END OPERATIONAL TESTING
WETEND DEMONSTRATION OF
*CONTINUOUS OPERATION
WET END CHAR PRODUCTION
DRY END ACTIVATED
CARBON PRODUCTION
J J A
S 0
1974
N D J F M A M
J J
1975
A S 0
Figure 21. Operating schedule.
49
-------
be broken down into three periods for the wet end and three similar periods
for the dry end.
The shakedown period was the time during which the equipment was de-
bugged and equipment operating procedures were established. The shakedown
period for both ends covered a period of 2 months.
The operational testing period was the time when various process condi-
tions were tested to see which conditions would yield an acceptable product.
For the wet end, this time period lasted about one month; for the dry end,
this time period was spread over 7 months. The time required here was much
longer because dry end operations depended on availability of char from the
wet end.
50
-------
SECTION IX
PILOT PLANT OPERATING RESULTS
WET END OPERATIONAL TESTING PERIOD
The operational testing period lasted about one month and was devoted
to adjusting process conditions to find the range which would yield a char
product suitable for activation. A total of fifteen pilot plant runs was
made during this period.
Initial pilot plant conditions were selected based on previous batch
and prepilot experience. These conditions were pretreatment to pH 11.5,
pressure 1758 kkgf/m2 C2500 psig), and temperature 345°C (655°F). As ex-
pected, pilot plant performance did not precisely duplicate batch and pre-
pilot system performance under the apparently same conditions. It was
necessary to adapt and adjust the initial conditions to avoid operating
problems and to produce an acceptable char.
There are three primary process variables available for adjustment.
These are the amount of S02 used for black liquor pretreatment (represented
by pK of the pretreated liquor), the operating pressure, and the reaction
temperature. The fifteen operational testing runs are summarized in Table 2.
The pressure drop (i&P) across the system is listed as an indication of
fouling or plugging. It is the A P between the tubeside inlet to the
cross exchanger and the outlet from the reactor measured during the heat-
up time on water prior to switching to black liquor. The A P shown for Run 1
represents a clean system.
The visual appearance of the char product, its filtering ability, and
the ability to wash the filter cake are listed to describe the acceptability
of the char. "Normal" appearance indicates a fine material, forming a dry-
looking cake, which does not flow on standing (not thixotropic) . Run 1
produced a char which could not be filtered. The ,&P increased rapidly while
running on black liquor and resulted in rupture of several pressure relief
discs, terminating the run. It was felt that the rapid increase in A P was
caused by a salting-out effect brought on by a vapor phase increasing pro-
portionately to the liquid phase in the system. The supposed vapor phase
would be incapable of maintaining the solids in suspension and would leave
solids deposits on the walls. This explanation later proved to be incorrect
but was the basis for selecting pressure and temperature conditions for
Run 2.
51
-------
TABLE 2. OPERATIONAL TESTING PERIOD
NJ
Pre treatment
Run pH
1
2
3
A
5
6
7
8
9
10
11
12
13
14
15
11.5
11.0
11.0
9.5
10.0
10.5
10.5
10.0
10.0
10.0
9.5
9.5
10.0
10.0
10.0
Pressure
kkgf/m*
1758
1934
1934
1934
1934
1934
1934
1934
1934
1934
1934
1934
1934
1934
1934
Temperature
•c
345
335
335
335
335
335
340
340 '
340
350
340
340
340
335
340
Black liquor
solids
Z
N.A
25.0
32.5
27.5
30.0
27.5
20.0
30.0
27.5
25.0
27.5
30.0
25.0
25.0
30.0
System8
A P Char
kkgf/m appearance
52.7
70.3
70.3
77.3
87.9
87.9
87.9
87.9
87.9
94.9
98.4
98.4
87.9
87.9
87.9
Tarry
Grainy
Grainy
Slightly
thixo tropic
Normal
Normal
Normal
Normal
Normal
No char
collected
Thixotropic
Thixotropic
Thixotropic
Normal
Normal
.
Filtering
ability
Not
filtered
Moderate
Moderate
Good
Good
Good
Good
Good
Good
Not
filtered
Good
Good
Good
Good
Good
Cake washing
ability
Not washed
Bad
Bad
Good
Good
Good
Good
Good
Good
Hot washed
Good
Good
Good
Good
Good
Runnine ability
Terminated by R.D. failure.
Smooth. A? held constant at
70.3 kkgf/m2 on black liquor.
Terminated by R.D. failure.
Smooth. ^ P held constant at
87.9 kkgf/m2 on black liquor.
Smooth. A P held constant at
105.5 kkgf/m2 on black liquor.
Smooth. A P held constant at
94.9 kkgf/m2 on black liquor.
Smooth. £ P held constant at
87.9 kkgf/m2 on black liquor.
Smooth. A ? held constant at
87.9 kkgf/m2 on black liquor.
Smooth. A P held constant at
116.0 kkgf/m2 on black liquor.
Terminated by R.D. failure.
Smooth. A F held constant at
105.5 kkgf/m2 on black liquor.
Smooth. A P held constant at
105.5 kkgf/m2 on black liquor.
Smooth. <£\ P held constant at
87.9 kkgf/m2 on black liquor.
Smooth. A P held constant at
105.5 kkgf/m2 on black liquor.
R.D. failure (and "run termination)
was avoided, but & P increased
from 87.9 at start to 351.6 kkgf/m2
at end of the run.
a) This is Che pressure drop through the system on water
before the run on black liquor was started.
-------
For Run 2. operating pressure was increased to 1934 kkgf/m2 C2750 psie)
and reaction temperature was decreased to 335°C (6356F). Based on batch and
prepilot experience, SO pretreatment was increased to give a pH of 11 0 to
improve the properties of the char.
The P on water at the beginning of Run 2 was somewhat higher than the
P on the clean system, indicating some build-up of deposits during Run 1.
Run 2 proceeded very smoothly on black liquor and the A P held steady. The
slurry filtered easily and produced a grainy char which looked like black
sand. However, the char could not be washed on the three-stage counter-
current extractor belt. All attempts to wash the char resulted in a flooded
cloth. In a separate experiment, it was shown that the resistance of this
filter cake to the passage of water was very high.
Run 3 was designed to be a duplicate of Run 2. Again, the slurry filter-
ed easily, and a grainy char was produced which could not be washed.
The filter cake had a high resistance to the passage of water. Contrary to
the experience during Run 2, the A P rose rapidly during Run 3 and the run
was terminated by pressure relief disc failure. The black liquor received
from the mill evaporators for Run 3 had a solids concentration of 32.5%
compared to 25.0% solids received for Run 2. It appeared that high black
liquor solids concentration contributed to the plugging problem.
In Run 4, feed liquor SO™ pretreatment was increased further to give a
pH of 9.53ystem pressure and reaction temperature were unchanged. The A P
on water before this run reflected deposits which occurred during the pre-
vious run. Early in Run 4, there was a further increase in A P. perhaps a
residual effect of Run 3, then the A P held steady, and the run was com-
pleted without incident. The slurry filtered easily, the char had "normal"
fine-textured appearance and washed very easily on the countercurrent ex-
tractor. The filter cake was slightly thixotropic, which presented some
difficulty in subsequent solids handling and drying of the char.
In Run 5, the S09 pretreatment level was decreased to a pH of 10.0 in
search of a level where the filtration and washing would be satisfactory
without introducing thixotropic behavior of the filter cake. This objective
was realized in Run 5. The run showed no evidence of adding to the deposit
problem.
The pretreatment was decreased again in Run 6 to explore the gap be-
tween pH-s of 11.0 and 10.0. Results were comparable to Run 5.
Runs 5 and 6 defined an area of satisfactory operation and product
quality between pH's 10.0 and 10.5. So far,all acceptable runs had been made
at one reaction temperature (335°C) , In Runs 7, 8, and 9, the temperature
was increased 5°C to 340°C and this temperature condition was used xn com-
bination with pretreatment in the range of pH's Trom 10.0 to ™>*-™*Astern
operated smoothly and product properties were satisfactory. The A P held
steady through Runs 7 and 8, making a total of four consecutive runs with
no evidence of plugging or deposits. During Run 9, supposedly identical to
Run 8, the A? rose initially from 87.9 kkgf/m to 116.0 kkgf/m , then held
eonstant. The cause of this phenomenon was not understood until later.
53
-------
It will be remembered that a temperature of 345°C, used in Run 1,
resulted in a plugged system. However, the pretreatment and system pressure
conditions also were different from those employed in later successful runs.
Therefore, it could not be concluded definitely that the 345°C temperature
contributed to plugging. Run 10 was designed to use known acceptable pre-
treatment and pressure conditions at higher temperatures. The reaction
temperature was set at 350°C. This is 10 to 15°C higher than the previously
demonstrated satisfactory range. Run 10 was terminated almost immediately
by excessive pressure which caused relief discs to rupture. Not enough
slurry was produced to operate the filter.
In Run 4, (pH 9.5, pressure 1934 kkgf/m2, temperature 335°C), the
system operated well and the only problem with the product char was a
slightly thixotropic behavior. The thixotropy was overcome by increasing
S0? pretreatment. Runs 11 and 12 were made to determine whether a tempera-
ture change could have overcome the problem. Pretreatment to pH 9.5 was
used with a temperature of 340°C. Both runs under these conditions pro-
duced char with greater thixotropy, further demonstrating the unsatisfac-
tory results of pretreatment to pH 9.5. Run 13 used the conditions (pH
10.0, 1934 kkgf/m2, 340°C) which proved satisfactory in earlier runs but
the reactor was by-passed to shorten the reaction time. This run utilized
only the cross exchanger and the reactor preheater. There were no operating
problems but the char was thixotropic.
At this point, conditions for satisfactory operation and product
quality could be defjned as pretreatment conditions between pH 10.0 and 10.5,
pressure 1934 kkgf/m , and temperature between 335°C and 340°C. Run 9,
however, which employed conditions within this range, resulted in a A P
•while Tunning on black liquor and left some deposits in the systems.
This suggested that the lower temperature might be superior. Runs 14 and
15 were made to test this assumption. Run 14 at 335°C proceeded smoothly.
Run 15 at 340°C did not proceed smoothly. Relief disc failure was avoided
but the system A P increased from 87.9 kkgf/m to 351.6 kkgf/m2 when the
run was stopped. It appeared that the lower temperature was better.
At this point, it was not understood why Runs 8, 9, and 15, all cond-
ducted at apparently the same conditions, gave different results. Run 8
proceeded smoothly, Run 9 showed a small increase in A P, and Run 15
threatened relief disc failure. Later, it became apparent that the dif-
ferences reflected the extent of control over the temperature.
PRODUCTION OF CHAR FOR DRY END OPERATIONAL TESTING
At the completion of the wet end operational testing period, conditions
had been identified which would yield a char which could be handled and
processed in the dry end. A program was initiated to produce the char
needed for the dry end operational testing period. Twenty char production
runs were made and results are summarized in Table 3. Most of these runs
54
-------
TABLE 3. PRODUCTION OF CHAR FOR DRY END TESTING
Run
16
17
18
Pre treatment
PH
10.0
10.0
10.0
Pressure
kkgf/m2
1776
1776
1776
Temperature
•c
335
330
340
Black liquor
solids
Z
30.0
30.0
35.0
System"
1 D
»^ Y n
kkgf/«r
112.5
123.0
123.0
Char
appearance
Normal
Normal
Normal
Filtering
ability
Good
Good
Good
Cake washing
ability
Good
Good
Good
Runnlne ability
Smooth. 4 P held constant at
140.6 kkgf/m2 on black liquor.
Smooth. A P held constant at
140.6 kkgf/m2 on black liquor.
R.D. failure in the initial part
19 10.0
20 10.0
21 10.0
22 10.0
23 10.0
24 10.5
1934
1934
1934
1934
1776
1934
335
335
340
340
340
335
25.0
25.0
25.0
27.5
140*6 Grainy Slurry not
filtered
158.2 Run terminated
before any slurry
could be collected
105.5
130.1
Normal
Normal
Good
Good
25.0 175.8 Normal Good
25.0
140.6 Normal
Good
of the run; but PSV reseated and
the run was continued. Run was
not terminated, but A P rose
from 123.0 to 527.3 kkgf/m2 on
black liquor when the run was
stopped.
Slurry not Run was not terminated by R.D.
filtered failure, but A P rose from
140.6 to 386.7 kkgf/m2 when run
was stopped.
The run was terminated by K.D.
failure. Run time was very short.
Good The run proceeded smoothly with
A t rising Initially to 140.6
kkgf/m2 on black liquor and hold-
ing steady.
Good R.D. failure in the Initial part
of the run; but PSV reseated and
run was continued. This time the
run proceeded smoothly with 4 P
holding constant at 193.4 kkgf/m2.
Good Run was not terminated by R.D.
failure but 4 PTose from 175.8
to 632.8 kkgf/o2 when run was
stopped.
Good Smooth with A ? rising initially
to 158.2 kkgf/m2 and holding constant.
(continued)
-------
TABLE 3 (continued)
Slack liquor
Pretreatment Pressure Temperature solids
Run pH kkgf/m2 "C . Z
System"
A P , Char Filtering Cake washing
appearance ability ability _ ..
Runnine ability
Ul
25
26
27
28
29
30
31
32
33
34
35
10.0
10.0
10.0
10.0
9.5
10.0
10.0
10.0
10.0
10.0
10.0
1934
1776
1776
1934
1934
1934
1934
1934
1934
2109
1776
340
335
340
345
340
345
340
340
340
340
330
27.5
25.0
30.0
27.5
27.5
25.0
27.5
30.0
25.0
30.0
27.5
158.2
203.9
239.1
140.6
123.0
112.5
123.0
123.0
140.6
140.6
203.9
Normal
Normal
Normal
Slightly
thixotropic
Slightly
grainy
Normal
Normal
Normal
Good
Good
Good
Grainy Moderate
Good
Good
Good
Good
Normal Good
Normal Good
Good
Good
Good
Good
Bad
Good
Fair
Good
Good
Good
Good
Good
Run was not terminated by R.D.
failure, but 4P rose from 158.2
to 316.4 kkgf/m when run was
stopped.
Run was not terminated by R.D.
failure, but 4 P had risen to
281.2 kkgf/m2 when run t*as stopped.
Run was not terminated by R.D.
failure, but A f had risen to
773.4 kkgf/m2 when run was stopped.
Run was not terminated by R.D.
failure, but & P had risen to
667.9 kkgf/m2 wiien run was stopped.
Smooth with /3 P rising initially
to 175.8 kk|>f/m2 ,ind hoidinf. constant.
The run proceeded in moderate fash-
ion with A P rising slowly to
281.2 kkgf/m2 during the run.
Smooth with /IP holding constant
at 140.6 kkgf/m2 after an initial
rise ro that value.
Run was not terminated by R.D.
but A P had risen to 632.8 kkgf/m2
when run was stopped.
Smooth with 4 P rising initially
to 158.2 kkgf/m and holding constant.
Run was not terminated by R.D. fail-
ure, but & P had risen to 808.6
kkgf/in2 when the run was stopped.
The run proceeded in moderate fash-
ion with & P rising slowly to
_298.8 kkgf/m2 during the run.
a) This is the pressure drop through the system on water
before the run on black liquor was started,
-------
were made within the proven limits of pretreatment and temperature
conditions. Seventeen out of the twenty runs produced 10,000 kg of
acceptable char from a total of 77,400 liters of black liquor. Yield of
char based on black liquor solids was about 23%.
Although this campaign was successful in producing the char required
for dry end operation, three out of twenty runs were terminated or hindered
by relief disc failure and another eight runs showed large increases in A P
during operation on black liquor. After Run 20, the system was shut down
for mechanical cleaning of the reactor preheater and after Run 27, the
reactor was cleaned.
Table 4 shows the distribution of runs according to ease of operation
as a function of the operating conditions. It is apparent that reaction
temperature is the controlling variable and that temperature preferably
should be limited to 330°C for trouble-free operation. As will be seen
later, utilization of this information was complicated by an overdesign of
the pilot plant equipment.
Dry end operational testing began as soon as char started to come from
the wet end. Early results indicated that the sodium content of the char,
averaging 1.55% on a dry weight basis, was too high (see discussion of dry
end operation). An acceptable level was about 0.5% sodium. Laboratory
experiments showed that this level could be reached easily by acid washing
the char. All char on hand and all subsequent production was acid washed.
WET END PRODUCTION DEMONSTRATION
Based on the analysis of Runs 1-35, which pointed to reaction tempera-
ture as the variable controlling the deposit problem, five Runs (36-40)
were made to explore lower temperatures and to retrain the operating person-
nel before atempting a sustained run. Results of these runs are shown in
Table 5. All of these runs proceeded smoothly and produced char which
filtered and washed easily. The lowest temperature (320°C) produced a
slightly thixotropic char. The five runs processed 19,000 liters of black
liquor and produced 2,500 kg of char for dry end operations.
During two earlier runs at 330°C (Runs 17 and 35), it was observed that
the temperature control was difficult. The temperature tended to rise above
the control point and it was necessary to turn off the reactor preheater oc-
casionally to avoid this increase. This problem became more pronounced
during operation at still.lower temperatures and the reactor preheater had
to be operated in on-off cycles. It became apparent that the reactor pre-
heater had been oversized for the actual heat duty required by the process.
The low temperature runs (36-40) contributed very little to increased
AP. However, the system contained substantial deposits built up by Runs
28-34, so the system was mechanically cleaned in preparation for sustained
operation. The continuous run, Run 41, is listed in Table 5. Average
57
-------
Ui
00
11.5
Number of runs 1
Percent terminated or 100
hindered by relief disc
failure
Percent showing large 0
A P increase without
relief disc failure
Percent eood operation 0
^^^^•^^•^•••••^••M^fllMl
PH
11. 0 10.5
2 3
50 0
0 0
50 100
^•^••••^•'^••^^••••••^•^^•^^•^•^^^••••••M^^M^^H
Pressure
10.0 9.5 1776
25 4 8
16 0 25
36 0 38
48 100 38
•I^IWHt^M^^HHM^BIHM
kkgf/m2
1934 2109
26 1
IS 0
19 100
66 0
1 TESTING
m^^^mmmmmammi^m*fm^mmmmmmm*m*^*mm~~~*~~~~^**—~~*^**i~a^^*~^~"^^^
Temperature *C
330 335 340 345 350
2 11 17 3 1
0 18 12 33 100
0 18 30 33 0
100 64 58 33 0
TABLF 5 UFT VKn PunniirTTnN nFMnw^TRATTpM
Black liquor
Pre treatment Pressure Temperature
Run pH kkgf/m2 "C
36 10.0 1776 330
37 10.0 1776 325
38 10.0 1776 320
39 10.0 1776 325
40 10.0 1776 330
41 10.0 1776 330
solids
X
27.5
30.0
27.5
25.0
25.0
27.5
System*
A P Char
kki'f/n appearance
203.9 Normal
203.9 Normal
246.1 Slightly :
thixotropic
228.5 Normal
228.5 Normal
52.7 Normal
Filtering
ability
Good
Good
Good
Good
Good
Good
Cake washing
ability Running ability
Good Smooth with A P rising Initially
to 210.9 kkgf/m2 and holding.
Good Smooth with A P having risen to
246.1 kkgf/m2 when the run was stopped.
Good Smooth with & P holding at
228.5 kkgf/m2.
Good Smooth with A P holding at
228.5 kkgf/m2.
Good Smooth with A P holding constant
at 228.5 kkgf/m2.
Good This wan the continuous run demon-
stration. The run proceeded smoothly
for most part with the A P in-
creasing slowly during run. The
run was terminated after 27 hours
by R.D. failure. The A P had
increased to 984.3 kkgf/m2 when R.D.
a) This IB the pressure drop through the system, on water
before the run on black liquor was started.
-------
conditions were 10.0 pH, 1776 kkgf /m2, and 330°C. The char product was nor-
mal in appearance and filtered and washed easily. Operation generally was
smooth but there was a gradual increase in A? culminating in relief disc
failure after 27 hours of operation. It was most difficult to control the
reaction temperature at 330°C and the temperature climbed above this level
whenever the reactor preheater was on. Each time this happened, the system
4 P increased. If there had been a way to avoid the temperature surges,
it is believed that the run could have continued indefinitely.
The continuous run processed 67,500 liters of black liquor and pro-
duced 8,700 kg of char, a yield of about 23.3%6asedon black liquor solids.
ADDITIONAL PRODUCTION RUNS
The char produced through and including Run 41 was insufficient to
supply the planned dry end operating schedule. A series of fourteen runs,
Runs 42-55, was required to produce the additional char. Results are sum-
marized in Table 6.
After the 27 hour run (Run 41) the system had a high AP, i.e. bad
deposits. The reactor was cleaned to reduce the deposits, but when opera-
tion was resumed the A P still was quite high and it became apparent
that the reactor preheater also was dirty. Operation was smooth but some-
what cumbersome. After Run 47, the system was given a complete cleaning to
simplify operation. This series of runs provided an opportunity to check
the conclusion that excessive reaction temperature was the cause for de-
posits. While the reactor was being cleaned prior to Run 42, the cross
exchanger was modified to throw more heat duty into the reactor preheater
and bring its load within a range which could be controlled. During Runs
42-55, temperature never was allowed to rise higher than 330°C and several
lower temperatures were investigated. As can be seen in Table 6, increases
in A P were minimal during this series of fourteen runs and there were no
relief disc failures nor threats of failures. The amount of black liquor
processed was 96,400 liters,and 14,100 kg of char was produced, giving a
yield of about 26.1% based on black liquor solids. This production was about
one and one-half times the production during the 27 hour continuous run.
The information and observations developed during this series of
fifty-five runs convincingly relates the deposit (plugging) problem to the
control of reaction temperature. With an adiabatic tubular reactor
configuration, such as the pilot plant reactor, relatively low temperature
is required to avoid rapid formation of char in a localized area. With
other reactor configurations designed to distribute the reaction over a
larger area, higher operating temperatures should be feasible.
The true effect of temperature was difficult to detect in the pilot
plant because temperature control deteriorated as the temperature was re-
duced. The control problem resulted from greater than expected efficiency
of heat transfer in the cross exchanger and from the exothermal heat of
reaction which was not recognized in smaller scale work. Thus, the heat
input required from the reactor preheater was reduced by the higher tempera-
ture entering the preheater and was further reduced by the exotherm. As
59
-------
Run
42
43
44
45
46
47
48
49
50
51
52
53
54
55
Pre treatment
PH
10.0
10.5
10.0
10.5
10.5
10.5
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
Pressure
kkflf/m^
1758
1758
1758
1758
1758
1758
1758
1758
1758
1758
1758
1758
1758
1758
Temperature
"C
330
320
320
315
325
330
330
320
330
320
320
310
310
310
Black liquor
solids
Z
30.0
30.0
30.0
30.0
30.0
30.0
25.0
32.5
27.5
20.0
>
27.5
25.0
25.0
20.0
,_
System
A P ,
kkgf/mz
246.1
298.8
281.2
351.6
351.6
246.1
35.2
52.7
52.7
70.3
87.9
105.5
105.5
105.5
Char
appearance
Normal
Normal
Normal
Normal
Normal
Normal
formal
Normal
Normal
Normal
Normal
Normal
Normal
Normal
Filtering
ability
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
Cake washing
ability
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
i( Running ability
Sroooth with A- P rising slowly to
351.6 kkgf/m2.
Smooth with 4 P rising initially
to 351.6 kkgf/m2 and holding constant.
Smooth with A 7 rising initially
to 369.1 kkgf/m2 and holding constant.
Smooth with A t rising initially
to 386,7 kkgf/m2 and holding cons tant.
Smooth with /i P rising slowly to
509.7 kkgf/m?.
Smooth with & P rising slowly to
421.9 kkgf/m2 and holding constant.
Smooth with & V rising initially
to 52.7 kkgf/m2 and holding constant.
Smooth with A ? risins initially
to 87.9 kkgf/m2 and holding cons tant.
Smooth with A P rising initially
to 87.9 kkgf/m2 and holding constant.
Smooth with A P rising initially
to 87.9 kkgf/in2 and holding constant.
Smooth with A P rising initially
to 105,5 kkgf/m2 and holding constant.
Smooth with A P holding steady at
123.0 kkgf/m2.
Smooth with A P rising initially
to 123.0 kkgf/m2 and holding steady.
Smooth with A r rising initially
-------
the temperature control point was decreased, the demand on the preheater
became so low that the input could not be controlled. Although reducing
the temperature alleviated deposits in the reactor, the surges brought on
by control problems caused too much reaction to occur in the preheater and
created deposits there.
The relative freedom from deposits in the later stages of this campaign
demonstrate that the process can be operated free of deposit problems by
proper thermal sizing of heat recovery, heating, and reaction equipment and
by close control of temperature at the stage where the reaction is pro-
ceeding rapidly.
Table 7 is an amplification of Table 4 to include all fifty-five runs
and clearly shows the critical role of temperature.
Table 7. EFFECT OF TEMPERATURE ON PLUGGING
Temperature Percent of runs with
°C little or no plugging
310
315
320
325
330
335
340
345
350
100
100
100
100
100
64
59
33
0
DRY END OPERATIONS
Analysis of Pilot Plant Char
In Section V, the nature of hydropyrolysis char produced on a laboratory
scale was discussed. To relate the properties of pilot plant char to the
laboratory char, representative pilot plant chars were analyzed by an inde-
pendent testing laboratory. Proximate analysis, ultimate analysis, and fusion
temperature of the ash were determined on char which had been acid washed
and dried.
61
-------
PROXIMATE ANALYSIS OF PILOT PLANT CHAR
(percent)
As received Dry basis
Moisture 7.71
Ash 2.82 3.06
Volatile 32.71 35.44
Fixed carbon 56.76 61.50
ULTIMATE ANALYSIS OF PILOT PLANT CHAR
(percent)
As received Dry basis
Moisture 7.71 -
Carbon 72.31 78.35
Hydrogen 4.49 4.87
Nitrogen 0.22 0.24
Oxygen 10.81 11.70
Sulfur 1.62 1.76
Chlorine 0.02 0.02
Ash 2.82 3.06
FUSION TEMPERATURE OF ASH
Reducing Oxidizing
atmosphere atmosphere
Initial deformation 882 888
Softening (H=W)a 904 954
Softening (H=1/2W) 921 982
Fluid 960 1038
tl is cone height, W is cone width
The data show that pilot plant char contained less carbon and more hydrogen
than the laboratory char and the coals described in Section V. Considering
the adjustments in hydropyrolysis conditions which were required between
the laboratory and pilot operations, some change in char properties would
be expected. Volatile matter, the most important factor, was within the
acceptable range for production of good activated carbons. Fusion temper-
atures of the ash indicated that slagging should not be a problem at
calcination (600° to 750°C) and activation (750° to 850°C) temperatures
appropriate for hydropyrolysis char.
62
-------
Drying Operations
As noted in the pilot plant description, the pilot plant dryer is a
bacch unit selected to serve as a tool for drying the char. It was not
expected to provide design data adequate for commercial equipment selection.
It did provide data to simplify later trials on continuous equipment at
vendor facilities.
Proper dryer charge was found to be 91 to 136 kg wet filter cake at
40% to 60% solids. As mentioned in the equipment description, the dryer
could tolerate 87.9 kkgf/m2 (125 psig) pressure at 343°C. Unfortunately,
this steam pressure and temperature develops a drying rate (vapor load)
far in excess of the capability of the condenser/receiver system, flooding
the vacuum pump. Good operation was obtained at a steam pressure of
4.2 kkgf/mz (6 psig).
Table 8 gives drying rate data for a variety of loads, temperatures.
and vacuum levels. Comparing Run 3 to Run 4 and Run 8 to Run 9, the
effect of changes in vacuum at constant temperature is shown. Runs 1, 5,
and 6 utilized the same temperature and vacuum levels but the char load in
the dryer was varied from 54 kg to 159 kg. It was mentioned earlier that
operation at high temperatures would cause a vapor removal rate high
enough to flood the vacuum pump. In Run 7, initial drying was done with
the system open to atmosphere, then vacuum was used after the maximum
vapor load was passed. Although this did alleviate the flooding problem,
the drying rate without vacuum was too slow> and use of higher temperatures
to increase the drying rate, with the system exposed to air, would risk
combustion of the char.
As noted earlier, hydropyrolysis char has a chemical reactivity
greater than coal. This reactivity, although an advantage during the
activation step, gives the char pyroforic properties, i.e. the heat of
adsorption of oxygen from air raises the temperature to the ignition level.
Laboratory data indicate that the maximum temperature at which hydropyrolysis
char can be exposed to air is 50°Cf Thus, although the char is processed
at temperatures above 50°C, it must be cooled to or below this temperature
before discharge and exposure to air.
Continuous Dryer Trial—
To obtain data for commercial design, acid washed char filter cake was
dried in vendor facilities using a LUWA D-210 continuous dryer operating on
the thin film principle. At a steam heating pressure of 105.5 kkgf/m
(150 psig), char was dried to 5% to 7% moisture at product rates of 34 to
41 kg/hr/m* of heated surface. Performance under vacuum up to 300 mm Hg
was not enough better than atmospheric operation to justify the added
capital and the greater difficulty of feeding wet cake into a vacuum. As
noted earlier, steps to exclude air would be required.
The product was a good powder. The only problem experienced was
bridging of the wet filter cake in the dryer feed hopper. A positive
feeding system is required rather than simple gravity flow.
63
-------
Runb
1
3
4
5
6
7
8
9
Load,
kg
Standard
(91-113 kg)
Standard
Standard
Standard
159 kg
Approx. 54 kg
Standard
Standard-
till 1 r,
Shell
temperature,
°C
106
102
102
106
106
111
111
111
Absolute
vacuum,
mm Hg
150
50
200
150
150
0-2003
200
100
Initial
moisture,
%
47
34
33
36.5
35.5
41.5
48
36.5
Time to
10% moisture,
rain
167
212
95
126
150
235
103
83
Time to
5% moisture,
min
188
240
122
139
170
255
128
98
alnitial drying with no vacuum.
Run 2 is not tabulated due to equipment malfunction.
Compacting Operations
In the original concept, this grant program visualized treatment of
effluent with a powdered carbon and use of the loaded carbon as fuel after
one adsorption. Part I of the program soon showed that the economics favored
a granular carbon which would be regenerated thermally between loadings.
Hydropyrolysis char typically is a uniformly fine material. Under some wet
end operating conditions, a somewhat coarser product was obtained.
SCREEN ANALYSIS OF DRY CHAR
(weight percent)
Sieve fraction Typical product Coarse product
+35 mesh
-35 +60
-60 + 120
-120 + 200
-200 + 325
-325 mesh
2.7
1.7
1.0
2.8
29.1
62.8
0.3
0.0
3.8
31.4
32.0
32.5
64
-------
There are a number of methods used commercially to convert a powder
into a granular material. Among these are compaction, agglomeration, and
extrusion. Usually, a binder material is added to strengthen the granules.
Based on inquiries and evaluation of existing technology, it was judged that
the compaction method using a roll press would be the best selection for
hydropyrolysis char.
Strength (hardness) is an important property of granular carbons,
which must have sufficient strength to withstand the production steps of
calcination and activation and repeated regenerations with minimum attrition
and breakage. Hardness can be evaluated by a destructive testing procedure
in which a dry sample is shaken for 3 hours on a reinforced 60-mesh screen
with steel balls. The percentage of the granules surviving in the original
size range is a measure of hardness. Relatively speaking, the greater
percent survival, the stronger the carbon.
To study compacting variables, the pilot equipment was operated
initially to produce small (100 to 200 kg) quantities of compacted char.
This is about one hour of compactor operation. Later as additional
expertise was developed, sufficient material was processed for multiple
hearth furnace operation.
Effect of Binder Type—
In general, coal tar pitch is considered to be a good binder for
carbons which must be subjected to high temperature conditions and retain
their strength. One way to classify this binder is according to melting
point. I-t: was believed initially that the use of pitches of different melting
points might vary product quality. By lowering the melting point, the binder
might become more tacky or even flow when subjected to the frictional heat
of the compactor. This effect could enhance strength by causing better
adhesion. A second, and possibly a counteracting effect, is the increase in
binder carbon content with increasing melting point. It would be expected
that the higher the carbon content the stronger would be the binder
structure after calcination.
To evaluate this variable, binders in the melting point range of
98° to 143°C were used and the compacted products were activated and
hardness tested. Within the experimental error of the testing, no change
in product strength could be attributed to change in binder melting points.
The 143°C melting point binder was chosen for future pilot operation as it
was available commercially in a pulverized form. This binder is designated
as No. 1' Ebony core binder without cereal binder or oil.
Effect of Binder Amount—
Hydropyrolysis char was compacted using no pitch binder, 5% binder
and 10% binder. The three compacted and granulated products were calcined
for 1 hour in the multiple hearth furnace at 650° to 700°C. The calcined
products were tested for hardness and compared to a commercial granular
activated carbon.
65
-------
HARDNESS AFTER CALCINING
(Percent retained in original 16x20 mesh range)
Product Hardness
Char, no binder 77
Char, 5% binder 85
Char, 10% binder 84
Commercial activated 69
carbon
These results indicated that compacted and calcined hydropyrolysis char
should resist attrition during activation. The data do not show that
hydropyrolysis activated carbon will be stronger than commercial activated
carbon because strength normally decreases during the activation step. The
data show that binder addition above 5% probably would not significantly
increase strength. Further testing to determine optimum binder addition
was not conducted. In fact, it was felt that no added binder would be
needed but the decision was made to use binder as insurance in subsequent
operations.
Effect of Char Moisture Content—
A series of compaction runs was made varying the char moisture content.
The char contained 5% binder. Equipment operation was observed and the
products were calcined and tested for hardness. The results are shown in
Figure 22. There are four zones of operation:
Zone 1: moisture 2% and below - The char is too dry for the feed
screw to grab the material and predensify it for the roll press.
The char simply flows through the feed screw and rolls with no
compaction.
Zone 2: moisture 2% to 8% - This is the operable range for the
equipment. As moisture increases above 2%, frictional forces in
the feed screw increase and heat is generated, causing softening
of the binder. Product quality increases to a maximum at about
8% moisture.
Zone 3: transition at 8% moisture - Operation is unstable.
Product quality varies from very good when the system can operate,
to poor when the feed screw lacks the power to predensify the char.
The transition at about 8% moisture is abrupt; to allow a margin
for errors, safe operating range was set at 6% to 7.3% moisture.
Zone 4: moisture above 8% - The feed screw locks up and the
compactor can only be operated intermittently. The rolls do
not have time to heat up and the system does not reach steady state.
Quality is consistently poor during the brief intervals when
anything is produced.
66
-------
VO
O
1
00
O
1
o
z
M
M
cn
fr*
•xl
O
I
cn
cn
w
la
O*
O
I
Ul
O
I
O
—o-
FEED SCREW LOCKS UP
O
SAFE
// A OPERATING
; RANGE
1
12
1
10
I I I I I
8 6
MOISTURE IN FEED CHAR (%)
POOR IF ANY
PRODUCT
1
2
Figure 22. Effect of moisture content on hardness of compacted char.
67
-------
Material Balance—
After establishing acceptable conditions of 5% binder, 6% to 7.3%
moisture, and a feed rate of 135 kg/hr, compactor operation was standardized
at these conditions and, thereafter, the compactor simply became a tool
to prepare material for multiple hearth furnace operations. During the
later operating periods, a material balance was made around the compaction
system for commercial design purposes. The balance is shown in Figure 23.
Compactor Vendor Tests—
Prior to pilot plant construction, hydropyrolysis char was not available
in quantities sufficient for testing. The pilot plant compactor was designed
assuming a feed similar to carbon black and included a feed screw with
diminishing flights. The design of a compactor feed screw is particularly
sensitive to the nature of the material and the pilot plant feed screw
jammed frequently during operation on hydropyrolysis char. The screw was
compacting the material beyond the point of squeezing out air and became
packed full of compressed char.
When char became available, tests were run in vendor (K-G Industries)
facilities. A feed screw similar to the one in the pilot plant unit was
tried but no product could be made owing to bridging in the K-G hopper and
failure of the material to fill the screw. A high compression screw was
then tried but it developed high loads and the material was compacted in
the screw; the screw jammed owing to over-compression. The best product
was obtained when a uniform-flight feed screw (no compression) was used.
The major problem during this testing was the slow free flow of the
material to fill the screw. The material should be pushed mechanically into
the screw, or the screw should have more flights above the feeder insert
to allow time for filling.
The vendor tests and pilot plant operating experience indicated that
some moisture in the char will improve feeding behavior and reduce dusting.
Too much moisture will aggravate bridging in the feed hopper.
During the compactor runs in vendor facilities, granulator knife
speed was investigated on a KG 8-6 granulator. The granulator was operated
at 300, 600,and 900 rpm knife speeds. The best particle size distribution
was obtained at the intermediate speed.
Multiple Hearth Furnace Operations
Initial operation of the multiple hearth furnace utilized a compacted
coal in order to conserve the supply of hydropyrolysis char. During this
shakedown period, operators were trained and equipment was debugged. The
first run on compacted hydropyrolysis char indicated that physical equipment
changes were required to handle this material and the modifications were
made.
68
-------
205
-1-4 mesh
-4 + 8
-8 + 20
-20 + 40
-40
0%
6.2%
44.8%
21.5%
27.5%
135 (FEED)
SCREEN
COMPACTOR
ROLLS
39 (FINES)
31 (OVERSIZE)
174
GRANULATOR
135 (PRODUCT)
44 mesh
-4 + 8
-8 + 20
-20 + 40
-40
0.6%
80.2%
17.2%
2.0%
Figure 23. Compactor material balance.
69
-------
The first useful product data came from Run 2.
through 6 are shown In Table 9.
Data from Runs 2
Table 9.
MULTIPLE HEARTH RUNS
Run Condition
2 1
2
3
k
5
6
7t hr 11 min
4 hr 5 min
2 hr 58 min
5 hr 50 min
5 hr 50 min
-------
CALCINING
(Percent volatile In product)
1
2
Sample
Feed
Hearth
Hearth
Hearth 3
Hearth 4
Hearth 5
Hearth 6
Product
Volatile
33.9
23.3
5.3
5.6
5.2
5.9
5.6
5.9
The apparent differences beyond hearth 2 are within sampling error.
Devolatilization takes place on hearths 1 and 2, but it is known that time
at temperature beyond devolatilization aids in setting the carbon structure
and is beneficial. Therefore, the 1 hour calcination time was adopted for
all future calcinations.
During Run 3 on calcined char, the furnace burners operated poorly and
control was difficult but good activated carbon was produced. This carbon
was categorized and compared with two commercial carbons, with the results
shown in Tables 10 and 11.
Table 10. CATEGORIZATION OF ST. REGIS ACTIVATED CARBON, PITTSBURGH CAL
AND GRANULAR BARCO
Carbon
St. Regis
Pittsburgh CAt
Granular Barco
N2
Surface
area
m2/gm
720
1040
S60 .
Apparent^-
density
B/cc
.44
.43
49
% Ash
D.W.B.
6.1
6.5
1A.S
GC14
No.
48
79
•50 .
Iodine
No.
1168
1264
6S7,
Hardness
Wgt. % surviving test
-16+20 mat!. -2CM-35 matl.
80 84
64
49
Untapped density.
2Granular Darco was not available in -16+20 mesh size.
71
-------
Table 11. LOADING COMPARISON OF ACTIVATED CARBONS1
St. Regis Pittsburgh
Carbon CAL
Color concentration, cu
Color loading, cu/g
Ratio STR/CAL
STR/DARCO
TOC concentration, mg/1
TOC loading, mg/g
Ratio STR/CAL
STR/DARCO
1000
250
1.47
1.14
160
150
1.25
3.3
500
190
5.6
1.27
100
40
1.5
1.6
100 1000 500 100
100 170 34 <10
1.18
22. 160 100 2P.
16 120 27 16
1.0
1.0
DARCO
Granular
1000 500 100
220 150 85
160 100 2P.
45 25 17
BOD concentration, mg/1
BOD loading, mg/g
Ratio STR/CAL
STR/DARCO
120
2.2 <0.17
1.2 ^0.17
120
9
90
6
120
16
90
6
1.
Isotherms were run on 0.05% Black Liquor simulated effluent as
mill upsets prevented obtaining typical mill effluents. APHA Standard Method
used.
The data showed that activated hydropyrolysis char is competitive with
the commercial carbons. Hardness of the hydropyrolysis product appeared to
be better than the commercial products. Isotherm tests showed performance
levels which established that hydropyrolysis char can be activated to
acceptable quality for effluent treatment.
After Run 3, furnace modifications and repairs were made to provide
better operation. Run 4 was a production run to calcine sufficient char for
a sustained demonstration of activation. Run 5 was a short activation run
to check the effects of the modifications made to the furnace system.
Hearth 3 showed higher activation than the final product, indicating that
retention time should be shortened.
Run 6 constituted the final demonstration of activated carbon production.
Production was sustained for 96 hours and all available char was processed.
Maximum activation again was reached on hearth 3, indicating that shorter
(1 hour) retention time would have produced a better final product. The
increase in activation rate experienced during Runs 5 and 6 is attributed to
improved burner operation and control.
72
-------
SECTION X
ACTIVATED CARBON PLANT DESIGN AND COST ESTIMATE
OVERALL DESIGN CRITERIA
Engineering design studies were made for the sizing of four different
plants that would be capable of producing activated carbon. The microlime-
carbon method for treatment of unbleached kraft effluent, as outlined in a
previous EPA report prepared by St. Regis (2), was used as the basis on
which the activated carbon plants were designed. The four plants were
sized to produce the makeup requirements for one, two, three, and four
such microlime-carbon treatment facilities.
An engineering design study was conducted independently by St. Regis
for an approximately 330 tons/day pulp equivalent hydropyrolysis plant.
Results indicated a cost of $17.9 million in May,1975 dollars. It is
obvious that a hydropyrolysis system could not be justified economically
on the basis of activated carbon production. The justification for a
hydropyrolysis plant must be based on recovery of pulping chemicals
and energy. In this study, it was assumed that a hydropyrolysis facility
would be present and that a fraction of the total char sufficient to feed
the carbon activation plant would be available. The energy content of the
char is 7500 cal/kg. Char is charged to the activated carbon plant at
its fuel value.
DESIGN METHODS AND ASSUMPTIONS
A flow diagram for the proposed activated carbon plant is given in
Figure 24. This flow diagram shows the major pieces of equipment that are
used in the process. They are as follows:
dry char storage
compactor
compacted char storage
multiple hearth furnace
Calcined char storage
activated Carbon storage
73
-------
FROM
HYDROPYROLYSIS
TO TREATMENT PLANT
OR SHIPMENT
DRY CHAR
STORAGE
_E
ACTIVATED
CARBON
STORAGE
COMPACTOR SYSTEM
oo
FLUE GAS TO
FINES SEPARATION
AND DISPOSAL
STEAM
-f-4-"-h
COMPACTED
CHAR
STORAGE
CALCINED
CHAR
STORAGE
•AIR & FUEL
FURNACE IN
ACTIVATION MODE
FLUE GAS TO
FINES SEPARATION
AND DISPOSAL
AIR & FUEL
' FURNACE IN
CALCINING MODE
Figure 24. Carbon activation flow diagram.
-------
The storage bins are sized to provide a certain hold up time, which
varies from bin to bin according to the solid material to be stored.
However, since the cost of a bin is directly related to the volume of
material it can hold, the size of the bins is referred to their capacity
in cubic meters. The bulk density of the material to be stored in the bin
is used to convert the hold-up time into a volumetric capacity. The bulk
densities for the materials involved in the plant are:
Material Bulk density, kg/m3
CCompacted carbon 708
Regenerated carbon 400
(Calcined carbon 634
Activated carbon 500
The bins are constructed of carbon steel and have 30° tangent to tangent
conical bottoms.
The compactor is sized to generate a certain capacity in kg/hr of
-7+20 mesh product. The equipment combines the operations of mixing,
compaction, granulation, and screening into one unit. The operation of
binder blending is not included. In pilot compactor operations (see
Section IX), it was noted that binder probably would not be needed, but
5% coal tar pitch was added for insurance. After activation, the carbons
were harder than commercial carbons, which strengthened the opinion that
no added binder was needed.
The calcination and activation operations are performed in the multiple
hearth furnace. The furnace is equipped with a belt feeder and rotary-lock
inlet valve for the solids. The rabble arm shaft is air cooled, as are
the rabble arms. Capability for steam injection on each hearth is provided.
The assembly for the treatment of combustion off-gases consists of an
afterburner, scrubber, and discharge stack. A cooler is provided at the
furnace discharge for the purpose of reducing the carbon product temperature.
The design-rates for activation and calcination are 2.44 kg/nr/hr and
34.18 kg/m /hr, respectively. The calcination rate is based directly on
pilot plant results; activation rate is extrapolated from pilot plant
operating data.
Material handling in the plant is accomplished primarily by belt
conveyers and bucket elevators. The use of screw conveyers is avoided,
since they cause attrition in the carbon product.
Since the operations of calcination, activation, and regeneration all
are carried out in a multiple hearth furnace, there exists the possibility
of process economy by combining some of these operations. Clearly, three
different approaches can be taken: 1) Each operation can have its own
furnace. 2) The calcination and activation operations can be scheduled
into one furnace and regeneration performed in another. 3) All three
75
-------
operations can be scheduled into the same furnace. In approaches one and
two, the regeneration furnace would be included in the treatment facility.
However, in approach three the furnace would be located in the activation
facility. This means that the regeneration operation would have to be
separated from the treatment system. Certain complications immediately
arise because of this. Firstly, the cost of the furnace must be divided
between the activated carbon plant and the effluent treatment plant.
Secondly, the furnace operating time must be divided between calcining,
activating, and regenerating. Approach number one is the simplest in
terms of design. But approaches two and three could be more economical and,
of course, the most economical design should be selected.
2
For sizing the regeneration furnace, a rate of 7.32 kg/m /hr is used
based on data developed in a previous report (2). Also from that report,
the carbon dosage rate for the microlime-carbon effluent treatment plant
is 453.6 kg/hr. The expected loss per regeneration cycle is 10%, so the
fresh carbon make-up rate is 45.4 kg/hr.
The cost of activated carbon should be reduced by increasing the
production rate. To do so would require that the activation plant supply
carbon to more than one effluent treatment facility or that the surplus
be sold. Design calculations were made for carbon activation plants
supplying make-up carbon to two, three>and four microlime-carbon treatment
systems, i.e. rates of 90.8, 136.25and 181.6 kg/hr.
COST ESTIMATES
Cost estimates were prepared for the activated carbon facilities
discussed above. The purpose in making these various estimates was to
select the plant design which yields the lowest cost carbon. These
estimates are believed to be within ± 30% of the actual costs. The bases
for the estimates were the same as those used in the plant designs. All
cost figures are reported in October 1975 dollars.
The capital costs for the activated carbon plants were based on a
confidential study on an activated carbon system prepared for St. Regis.
The capital cost for each plant was derived by estimating the costs of the
major pieces of equipment - storage bins, Compactor, multiple hearth
furnace - and then applying a factor to scale up to the fotal plant
cost. The costs of the storage bins, compactor, and activation furnace
were estimated from quotes received during the last six months. The
costs of the regeneration system were estimated from the data of Hutchins (15)
and a confidential study prepared for St. Regis.
76
-------
The bases for determining the operating expenses were as follows:
Fixed Expenses
Salaried Personnel - 0.4 man at $16,000/man yr.
Salary expense - 10% of salary
Amortization - 6.25% of total cost of investment (TCI) per year.
Insurance and taxes - 2% TCI/yr.
Semi-Variable Expenses
Repairs and maintenance - 5% TCI/yr
Telephone and telegraph - 8% of salary
Travel - 15% of salary
Variable Expenses
Operating labor - 1.5 people per shift at $7.05/hr
Payroll expense - 20% of operating labor
Operating supplies - calculated directly
Fuel - $7.94/million kg-cal
Purchased power - $0.03/kwh
Steam - $4.52/1000 kg
Basic raw material - hydropyrolysis char at its fuel value
For the production of activated carbon the following utility require-
ments must be met:
Fuel - 9500 kg cal/kg of activated carbon
Steam - 4 kg steam/kg of activated carbon
Purchased power - 0.1014 kwh/kg of activated carbon
77
-------
Results of Cost Estimates
Capital and operating costs are shown in Table 12. In each estimate,
the capital costs are listed for both the activation facilities and the
regeneration facilities. The cost of the regeneration facility is
absorbed into the effluent treatment plant and does not enter into the
operating cost for the production of activated carbon. It is given for
the purpose of showing how the various plant designs influence the
regeneration cost.
TABLE 12. COST ESTIMATES POP ArTTVATTr.« m AXTJC -
Estimate number 1 J. .3. j^ 5_ 5
New carbon production (kg/hr) 45.4 45.4 45.4 90.8 136.2 181.6
Regenerated carbon Production (kg/hr) 408.2 408.2 408.2 408.2 408.2 408.2
Capital cost ($ million)
Total 8.110 5.625 4.315 4.755 5.165 5.600
Allocated to regeneration 2.180 2.180 1.695 1.415 1.140 1.000
Allocated to calcination 5.930 3.445 2.620 3.340 4.025 4.600
and activation
Operating expenses ($/kg of activated carbon)
Fixed
Semi-variable
JZariable
Total
0.614
0.364
.. 0.544
0.389 0.312 0.266
0.232 0.186 0.159
0.638 0.600 0.544
Estimates 1, 2 and 3 deal with plants which are capable of producing
the make-up activated carbon and meeting the carbon regeneration require-
ments for one microlime-carbon treatment facility operating at a carbon
dosage rate of 453.6 kg/hr. Estimate 1 considers a plant with the operations
of calcination, activatioii and regeneration being carried out in three
separate furnaces. Estimate 2 considers a plant with calcination and
activation scheduled into one furnace and regeneration carried out in a
separate furnace. Estimate 3 considers a plant with calcination, activation,
and regeneration scheduled into a single furnace. It is apparent that
the design with all three functions scheduled into one furnace yields the
lowest costs. The other two alternatives were eliminated from any further
consideration.
78
-------
Estimates 4, 5>and 6 deal with larger plants capable of producing the
make-up activated carbon for two, three,and four microlime-carban effluent
treatment facilities. Each estimate used the design where calcination,
activation, and regeneration were scheduled into one furnace. The regenerated
carbon production was that needed by one (local) treatment plant. In each
case, scheduling of the three functions was optimized using a computor-
program and the lowest overall operating cost was selected. Costs were
allocated to carbon production and regeneration on the basis of the time
fractions of furnace utilization.
As one goes from Estimate 3 to 4 to 5 to 6, it is seen that the capital
cost for the activation facility increases. This is expected since the
activated carbon production is increasing. The capital cost for regeneration
decreases (see p. 80). As the plant size increases, reductions in fixed
and semi-variable expenses per unit of production significantly reduce the
cost of activated carbon. This is partially offset by the packaging and
shipping costs (variable expense) for moving carbon from the larger plants
to other using locations.
Conclusions from the Cost Estimates
Figure 25 is a plot of activated carbon cost versus plant capacities
considered in this study.
1.6
1.4
1.2
1.0
0.8
0.6
0.4
COST RANGE FOR
COMMERCIAL WATER
GRADE CARBONS
/ S / /~
/ / / / / .
/ / / / S / /'
-/- £, J!—'—
50
100 150
ACTIVATED CARBON PRODUCTION (kg/hr)
200
250
Figure 25. Effect of production rate on cost of carbon.
79
-------
The shaded area in Figure 25 shows the cost range for commercial water
grade activated carbons, i.e. $1.10/kg to $0.88/kg. It becomes obvious that
one could buy make-up carbon from commercial sources at a cost no higher
than it could be made on site from hydropyrolysis char in the plants
considered here.. However, if one has more than the equivalent of four
treatment plants of the size that has been dealt with here and wishes to
produce the make-up carbon for all these plants at one site, then the carbon
could be produced at a lower cost than commercial carbons. The chances
of this happening are remote. It is apparent that in order to produce an
activated carbon from hydropyrolysis char at a significantly lower cost than
commercial carbons, one would have to build a plant comparable in size to
most commercial carbon production plants. In other words, one would have
to be planning on getting into the activated carbon business.
Effect on Effluent Treatment Costs
The results in Table 12 showed that the capital cost for carbon regen-
eration could be decreased through association with the process of producing
on-site activated carbon. This influences the cost of effluent treatment.
As the basis for comparison, consider the same microlime-carbon treatment
plant that has been used throughout this study. The costs for this effluent
treatment plant in January 1973 dollars have been estimated in some detail
in a previous EPA report made by St. Regis (2). The same procedure used
for developing costs in that report is used here. Table 13 shows the devel-
opment of microlime-carbon treatment costs when five different sources of
activated carbon are used. These costs are in October 1975 dollars and they
do not include any credit for water reuse.
TABLE 13. COST ESTIMATES FOR EFFLUENT TREATMENT PLANTS
Estimate number
Carbon source
Carbon cost ($/kg)
1 2
Commercial On site
0.992 1.522
145
On site On site On site
1.259 1.098 0.969
Capital cost ($ million)
Regeneration
Other X
Operating expenses
Fixed
2.180 1.695
8.471 8.471
($l/m3)
0.099 0.094
Make-up carbon 0.030 0.046
Other'1"
Total
to
0.057 0.057
0.186 0.197
Lime treatment 2.694
pH adjust 0.271
Carbon adsorption 3.954
Backwash 0.112
Initial carbon 1.440
Inventory
1.415 1.140 1.000
8.471 8.471 8.471
0.092 0.089 0.088
0.038 0.033 0.029
0.057 0.057 0.057
,0.187 0.179 0.174
.tb's Labor, utilities, supplies, etc. not influenced.
80
-------
Estimate 1 develops the treatment costs when commercial carbon at a
price of $0.99/kg is used as the carbon make-up source. Estimates 2 thru
5 develop the treatment costs when the carbon is made on site at the four
different plants considered in Estimates 3, 4, 5, and 6 of Table 12. By
comparing Estimates 2 thru 5 with Estimate 1, it can be seen that the use of
activated carbon produced on site from hydropyrolysis char can result in a
lower treatment cost when compared with the use of a commercial carbon.
Figure 26 shows the operating cost of the treatment plant using on-site
produced activated carbon plotted versus the production capacity of the
activated carbon plant used to generate the make-up carbon. The shaded
portion shows the operating cost range of the treatment facility if commercial
carbon is used as the make-up source. This again clearly shows that an
effluent treatment facility can be operated at a lower cost if on-site
produced activated carbon is used instead of commercial sources. Of course,
for the cost to be lower one must be producing the make-up activated carbon
at one location for use in the equivalent of three or more microlime-carbon
treatment plants each with a capacity of 3785 m of effluent per day.
0.20
0.19
0.18
•rr7-j-/—f~j^-f—'~-r-/—7~
w%m
,y_y_<^ i_/_ *. j: ./><_<- e-i
COST RANGE
USING
. CARSOHS
_i
200
0.17
50
100
ACTIVATED CAXBOV PRODUCTION (kg/hr)
rigute 26. Effect of on-nte carbon production on effluent treatment cost.
81
-------
181.6 kg/hr of carbon for use at one location. Carbon production rates
greater than 181.6 kg/hr can result in carbons that are lower in cost
than commercial grades. But the true reason for this study was to invest-
igate the production of on-site activated carbons for the purpose of
lowering effluent treatment costs. This purpose can be accomplished if
an on-site production facility with a capacity of 136.2 kg/hr or more of
activated carbon is available. This means that the carbon make-up require-
ments for the equivalent of at least three effluent treatment facilities
would have to be produced at the location of one plant.
82
-------
REFERENCES
1. Timpe, W.G., Lang, E.W., and Miller, R.L. "Kraft Pulping Effluent
Treatment and Reuse - State of the Art." EPA-R2-73-164. February 1973.
2. Lang, E.W., Timpe, W.G. and Miller, R.L. "Activated Carbon Treatment
of Unbleached Kraft Effluent for Reuse." EPA-660/2-75-004. April 1975.
3. Timpe, W.G., "Pyrolysis of Spent Pulping Liquors." U.S. Patent
No. 3,762,989. October 1973.
4. Timpe, W.G., Evers, W.J. "The Hydropyrolysis Recovery Process."
Tappi 56 (8): 100. August 1973.
5. Van Krevelen. Coal Science. 1964.
6. Mattson, J.S., and Mark, H.B. Jr. Activated Carbon - Surface Chemistry
and Adsorption from Solution. 1971.
7. Torikai, N. and Walker, P.L. Jr. "Activation of Bituminous Coal Lithotypes
in Carbon Dioxide," Office of Coal Research November 1968.
8. King, J.G., MacDougall, H., and Gilmour, H. Paper No. 47, Dept. of Scien-
tific and Coal Research, His Majesty's Stationary Office, London. 1938.
9. Lowry, H.H. (Ed). Supplementary Volume - Chemistry of Coal Utilization.
1936
10. Brunauer, Emmett, and Teller. "Adsorption of Gases in Multi-Molecular
Layer.'' Proceedings of American Chemical Society. February 1958.
11. Nandi, S.P., Walker, P.L., and Austin, L.G. Vol. II, Physics and
Chemistry of Carbon. 1966.
12. Dubinin, M.M. "The Potential Theory of Adsorption of Gases and Vapors for
Adsorbents with Energetically Non-Uniform Surfaces." Chemical Reviews 60:
235-41. 1960.
13. Marsh, H. and Rand, B. Carbon 9. p. 47. 1971.
14. Kipling, J.J. and McEnany, E. 2nd Conference on Industrial Carbon and
Graphite, London, p. 380. 1966.
15. Hutchins, R.A. "Thermal Regeneration Costs." Chemical Engineering
Progress 71(5) : 80. 1975.
83
-------
APPENDIX A
ANALYTICAL PROCEDURES
The methods used for analyses of samples in the work covered by this report
are listed below. The standard method number given refers to the method
in the 13th edition of Standard Methods for Examination of Water and Waste-
water, APHA (1971). Methods modified by St. Regis Paper Company are avail-
able on request.
TOG and TIC - Std Method 138 using a Beckman 915 total carbon analyzer;
sample filtered through Whatman No. 2 paper filter.
_p_H - Std Method 144A.
Na0S - Mercuric chloride titration method according to E. Bilberg,Norsk
2.
Skogindustri, 11/58 470 (1958), modified by St. Regis Paper Company.
Na?S and Na?S?0 - by unpublished methods developed by St. Regis Paper ,
Company.
Sulfate - Std Method 156B, gravimetric method.
Metal ions - Std Method 129A using atomic adsorption with Perkin-Elmer
Model 403 spectrophotometer.
Volatile Acids - by gas chromatographic method of S. M. Aronovic, et a],
Tappi 54 . 1963 (1971) and modified by St. Regis Paper Company.
Total Solids - Microwave drying method according to Kemeny & Chazin Tappi 56,
No. 8, 81 (1973). Modified by St. Regis Paper Company.
Sulfur, Carbon and Hydrogen - Rapid combustion method using a Leco induction
furnace. Method modified by St. Regis Paper Company.
pH Titration - Method developed by St. Regis Paper Company'using a Sargent
recording titrator.
Density - Calibrated pycnometer method. TAPPI Standard T625ts-63.
Volatile Matter and Ash - ASTM D271.
Nitrogen Surface Area - Brunauer, Emmettj and Teller. "Adsorption of Gases in
Multi-Molecular Layers." Proceedings of American Chemical Society ,
February 1938.
Carbon Dioxide Surface Area - Dubinin, M. M. "The Potential Theory of
Adsorption of Gases and Vapors for Adsorbents with Energetically Non-
Uniform Surfaces." Chemical Reviews 60: 235-241 1960.
84
-------
.
TECHNICAL REPORT DATA 1
(flease read Instructions on the reverse before completing) \
EPA-600/2-78-191
4. TITLE AND SUBTITLE
On-Site Production of Activated Carbon
from Kraft Black Liquor
/. Auiriun(s)
V. D. Del Bagno, R. L. Miller, J. J. Watkins
9. PtHHOhlvlirMU OR"AN ZATION NAME AND ADDRESS
St. Regis Paper Company
Cantonment, Florida 32533
12. SPONSORING AGENCY NAME AND ADDRESS
Industrial Environmental Research Lab. - Cinn, OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, OH 45268
3. RECIPIENT'S ACCESSION-NO.
5. REPORT DATE I
July 1978 issuing date I
6. PERFORMING ORGANIZATION CODE 1
8. PERFORMING ORGANIZATION REPORT NO. 1
10. PROGRAM ELEMENT NO. I
1BB610 I
11. CONTRACT/GRANT NO. I
12040 EJU I
13. TYPE OF REPORT AND PERIOD COVERED I
Final Report 11/71-9/75
14. SPONSORING AGENCY CODE 1
EPA/600/12 I
15. SUPPLEMENTARY NOTES Two prior reports have been published under Grant 12040 EJU; J
EPA-R2-73-164, Kraft Pulping Effluent Treatment and Reuse - State of the Art, 2/73 I
EPA-660/2-75-004, Activated Carbon Treatment of Unbleached Kraft Effluent for Reuse, 4/75J
A pilot plant was designed and constructed to produce char via the St. Regis
hydropyrolysis kraft chemical recovery process and to produce activated carbon from
the char. This report includes discussion of laboratory and prepilot work, the pilot
plant, and presents operating results.
After a period of optimizing feed pretreatment, temperature and pressure con-
ditions in the hydropyrolysis section of the plant, about 22,000 kg of char was pro-
duced for activation. The char was converted to a high quality granular activated
carbon having properties which compared favorably with commercially available carbons.
The concept of on-site production of activated carbon and the use of such carbon for
local effluent treatment has been demonstrated to be technically sound.
The cost of on-site production of hydropyrolysis activated carbon in quantities
required for local effluent treatment is not competitive with commercially available
carbons. To become competitive, the plant would have to supply carbon for effluent
treatment at several locations. To achieve significantly lower costs would require
entry into market production.
17. KEY WORDS AND DOCUMENT ANALYSIS j
a. DESCRIPTORS
Activated Carbon, Activated carbon treat-
ment, Pyrolysis, Pulp mills, Black liquors
Industrial w.aste treatment, Pilot plants,
Cost estimates
13. DISTRIBUTION STATEMENT
Release to public
b. IDENTIFIERS/OPEN ENDED TERMS
Activated carbon pro-
duction, Wastewater
treatment, Color removal,
Tertiary treatment,
Hydropyrolysis ,
Activated carbon
regeneration
19. SECUHiTY CLASS (This Report}
Unclassified
20. SECURITY CLASS (This page}
Unclassified
c. COSATI Field/Group
68D
21. NO. OF PAGES (
95 |
22. PRICE !
EPA Form 2220-1 (9-73)
OUSGPO: 1978 — 757-140/1405 Region 5-i I
85
------- |