STOPS: The Sludge-to-Oil
Reactor System
PB86-175684
Battelle Pacific Northwest Labs., Richland, WA
Prepared for
Environmental Protection Agency, Cincinnati, OH
Mar 86
I I
U*. DuMrtment §f Commerce
Ttdniui Mvrmatim Stnrict
-------
PBdfc-175684
EPA/600/2-86/034
March 1986
STORS: THE SLUDGE-TO-OIL REACTOR SYSTEM
by
P.M. Molton, A.G. Fassbender, and M.D. Brown
Battelle-Northwest
Richland, Washington 99352
Cooperative Agreement No. CR-810690-01-0
Project Officer
Howard Wall
Wastewater Research Division
Water Engineering Research Laboratory
Cincinnati, Ohio 45268
WATER ENGINEERING RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
KWODDCEO BY
NATIONAL TECHNICAL
INFORMATION SERVICE
US DEPARTMENT 01 COIMEICE
SMMtfKll. *» 22161
-------
TECHNICAL REPORT DATA
'Please r< jJ Instructions on the reverse before completing
1 REPORT NO
EPA/600/2-86/034
4 TITLE AND SUBTITLE
STORS: The Sludge-to-Oil Reactor System
7 AUTHOR(S)
P. M. Molton, A. G. Fassbender, and D. M. Brown
3 RECIPIENT S ACCESSION NO
|S. REPORT DATE
March 1986
ifo Pfc RFORMlNG ORGANIZATION CODE
8 PCRFORMING ORGANIZATION REPORT-NO
9 PE RFORMING ORGANIZATION NAME AND ADDRESS
Battelle Pacific Northwest Laboratories
Richland
Washington 99352
12. SPONSORING AGENCY NAME ANO ADDRESS
Water Engineering Research Laboratory, Cin., OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 452b8
10 program element no
DU B-113. P.E.CAZB1B/B-54
11 CONTRACT GRANT NO
CR-810690
13 TYPE OF REPORT AND PERIOD COVERED
14. SPONSORrNG AGENCY CODE ~
EPA/600/14
15 SUPPLEMENTARY NOTES
Project Officer: Howard Wall (513) 569-7659
16 ABSTRACT
The SIudge-to-OiI Reactor System (STORS) continuously converted over 400
gallons (20 percent solids) of sewage sludge to oil during 100 hours of operation.
About 80 percent of the energy in the sludge was recovered as an oil and cnar.
The energy recovered was sufficient to make the STORS energy self-sufficient
assuming the char is burned to furnish the heat for the process. STORS, when
fully developed, appears to be another option competitive with incineration for
sludge disposal. One test performed by spiking the sludge with pesticides
indicated that some of the pesticides were totally destroyed and others were
partly destroyed during processing.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b IDENTIFIERS OPEN ENDED TERMS
COSATI I It'll] Ciroup
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19 SECURITY CLASS (This Report)
21 NO OF PACES
124
20 SECURITY CLASS litis page,
Unclassified
22 PRICE
EPA Form 2220-1 (R»v. 4-77) previous edition is obsolete
-------
DISCLAIMER
"The information in this document has been funded wholly or in part by the
United States Environmental Protection Agency under assistance agreement
number CR-810690 to Battelle-Northwest. It has been subject to the Agency's
peer and administrative review and also to a Battelle-Northwest review, and it
has been approved for publication as an EPA document. Mention of trade names
or commercial products does not constitute endorsement or recommendation for
-------
FOREWORD
The U.S. Environmental Protection Agency is charged by Congress with
protecting the Nation's land, air, and water systems. Under a mandate of
national environmental laws, the agency strives to formulate and implement
actions leading to a compatible balance between human activities and the
ability of natural systems to support and nurture life. The Clean Water Act,
the Safe Drinking Water Act, and the Toxic Substances Control Act are three of
the major Congressional laws that provide the framework for restoring and
maintaining the integrity of our Nation's water, for preserving and enhancing
the water we drink, and for protecting the environment from toxic substances.
These laws direct the EPA to perform research to define our environmental
problems, measure the impacts, and search for solutions.
The Water Engineering Research Laboratory is that component of EPA's
Research and Development program concerned with preventing, treating, and
managing municipal and industrial wastewater discharges; establishing prac-
tices to control and remove contaminants from drinking water and to prevent
its deterioration during storage and distribution; and assessing the nature
and controllability of releases of toxic substances to the air, water, and
land from manufacturing processes and subsequent product uses. This publica-
tion is one of the products of that research and provides a vital communica-
tion link between the researcher and the user community.
The report describes the development of a continuous process that will
convert municipal wastewater primary sludge into a useful nonpolluting product
having properties similar to No. 6 fuel oil and usable as boiler fuel. The
product is also a projection of a method that can be used to recover energy
from wastes and save fossil fuels. This work is one of the tools necessary to
assess how the ultimate disposal of municipal wastewater sludge can be done in
a manner that will make it a truly valuable and profitable resource.
Francis T. Mayo, Director
Water Engineering Research
Laboratory
i i i
-------
ABSTRACT
Direct, continuous thermochemical liquefaction of primary undigested
municipal sewage sludge wa* carried out to produce a heavy oil and char
product suitable for use as a boiler fuel. The process was highly energy-
efficient in that up to 73% of the energy content of the 20% solids feedstock
was recovered as combustible products (oil and char) that were essentially
dry. The oil prodr:t had a heating value of 80% to 90% that of diesel fuel.
TSese products are capable of supplying the energy requirements for dewaiering
and liquefaction. Thus a wastewater treatment plant based on the liquefaction
concept can, in principle, be energy-self-sufficient.
Wastewater from the process was highly biodegradable, based on a standard
BOD determination. The only other byproduct was a gas that is characteristi-
cally more than 95% carbon dioxide. The process is therefore relatively
nonpolluting.
The prototype continuous liquefaction reactor was operated for nore than
100 hours without any sign of corrosion or char buildup on the inside walls.
Complete feedstock conversion was achieved at 300°C with a nominal 1.5-hr.
residence time, or 5.8 kg/hr of solids throughput. An economi;; assessment
prepared for a conceptual commercial liquefaction reactor indicated a cost of
S45/dry tonne, which is highly competitive with incineration. Projected
capital construction costs for a liquefaction unit, however, were $6.1
million, which is much lower than that for an incineration plant. The process
therefore has considerable promise for many potential sites in the United
States at the present time.
This report was submitted in fulfillment of Contract No. 810690-01-0 by
8attelle-Northwest under the sponsorship of the U.S. Environmental Protection
Agency. This report covers the period September 1983 to June 1985, and work
was completed as of June 30, 1985.
iv
-------
CONTENTS
Foreword iii
Abstract iv
Figures vii
Tables viii
1. Introduction 1
2. Conclusions 2
3. Recommendations ..... 3
4. Materials and Methods 4
Reactor Design 4
Sludge Feedstock Preparation 12
Analytical Methods 12
Sludge Analysis 14
Oil Analysis 15
Char Analysis 16
Wastewater Analysis 17
5. Results and Discussion 19
Reactor Testing 19
Reactor Operating Experience 23
Discussion of Reactor Operation 26
Sludge Composition 28
Oil Characterization 32
Char Characterization 39
Product Yields and Reaction Conditions 43
Data Reduction 44
Liquefaction Chemistry . 48
Gas Analysis 51
Wastewater Characterization 53
6. Liquefaction Process Preliminary Economics 56
Introduction 5fi
Estimate Basis ..... 56
Cost Estimate Summary ...» 6U
7. Overall Achievement of Project Objectives 6r>
References 67
Appendix A: Reactor temperature and pressure profiles, runs 208-509 . 68
Appendix B: Detailed inorganic elemental analvse* . 23
Appendix C: Oil therm»l oh^Ivsis profiles (TG and DSC) 88
V
-------
FIGURES
Page
1 Flow and valve labeling schematic
2 Photograph of liquefaction reactor system . . .
3 Arrangement of thermocouples 1n reactor . . . .
4 Reaction temperature and pressure profile. Run Kb
Preceding page blank
vi i
-------
*
I
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Page
21
22
22
23
28
29
30
31
32
33
34
36
37
38
38
40
41
42
43
47
48
48
49
49
51
52
53
54
54
55
57
61
62
63
TABLES
Reactor testmq experience
tlenencal cot.uositio.i of Portland sluiiae . . .
Initial results with Portland secondary sludge
Starting Material ano proauct composition, .lun 3S • • • •
Sludge feedstock composition
Sludqe feedstock ultimate analysis
Sluoge feeustock normalized ultimate organic analysis . •
Major inorganic elements in sludge feedstock
Water, heat, and ash contents of oil products
Major elemental compositions of product oils
Summarized inorganic element composition of produce oils .
Volatilization of oils by thermogravimetry
Oil viscosities
Pour point values for skinned oils
Flash point values
Char composition
Char ash analysis
Major inorganic elements in char
Product yields and production conditions
Mass/energy yields of oil and char products
Feed/product average elemental and ash contents
Feed/product average organic element contents
Feed/product organic element ratios
Ratios of C,H,N,0 in products (feed as unity)
selected inorganic elei.-ents in feed and products
Reactor off-gas composition
Wastewater analysis
COD and tJOD of wastewater
Wastewater biodegradabi1lty
Dissolved solius ash content
Estimate bases for three STORS sizes
Purchased material cost estimate
Energy requirements and outputs of the sludge to oil
facilities
Cost estimate comparison for three different sizes of
facilities
vi i i
-------
SECTION I
INTRODUCTION
Disposal of sewage sludge is an increasing problem for many municipali-
ties in the United States. A current need exists for imp'ementaticn of
alternative disposal technologies, fit the same time, it is recognized that
alternative technologies need to be energy-efficient, ind that if possible,
some product of value should be recovered from sludge. Direct thermochemical
liquefaction has the capability of -leeting all of these requirements.
Thermochemical liquefaction investigations by various authors over many
years have shown that organic material can be converted to bitumen, heavy oil,
and distillate oils with combustion heats approaching those of petroleum
product*. An organic solvent liquefaction process investigated on behalf of
the U.S. Environmental Protection Agency (EPA) also proved capable of produc-
ing liquid fuels, but it was abandoned because of poor economics and the fact
that it required a dry feedstock that had already solved the disposal problem
(1).
The work reported here describes results from continuous liquefaction
experiments using primary sewage sludge (20% solids) as feedstock. The sludge
was converted to fuel oil and char products having 732 of the heat energy of
the dry feed organic material, or 80% of the heat value of diesel fuel.
Separation of products from the wastewater is envisioned as being sponta-
neous: coupled with the use of a wet feedstock, this makes the energy balance
of the direct liquefaction process particularly favorable compared with
incineration, for example.
At the beginning of this project, which was based on earlier autoclave
(batch process) work to make a synthetic asphalt (2), we did :^t know if a
continuous process would work, if useful products could be obtained, or if the
overall process would be environmentally acceptable. Answers to these ques-
tions are all positive, but further research is needed before the technology
can be efficiency commercialized.
-------
SECTION 2
CONCLUSIONS
The results of experimental studies on primary sewage sludge liquefaction
in a prototype continuous reactor have amply confirmed the feasibility of the
process on a continuous scale for the production of fuel oil. The major
objective of the project has therefore been met. New information has also
been obtained on the behavior of aqueous primary sewage sludge slurries at
elevated temperature and pressure. This information will be of value in
designing the next generation of liquefaction reactors, as it significantly
alters conventional assumptions about heat transfer, viscosity, corrosion,
erosion, and char deposition during the process. As in previous batch
studies, sludge liquefaction occurred rapidly above 265°C. Complete conver-
sion to products (oil, char, and gas) occurred within 1.5 hr at 300"C, the
shortest residence time achievable with the reactor in its present
configuration.
Inspection of the reactor after more than 100 hr of operation showed that
anticipated problems of char buildup on unscraped surfaces, corrosion, and
p'tugging did not occur under the conditions uied. This is a significant
departure from previous experience, as for example with the wood liquefaction
pilot plant at Albany, Oregon. Some erosion of teflon valve seals did occur,
presumably because of high-velocity sand and ash particles, but chis was
solved by inclusion of a steel wool particle trap in the line. We conclude
that corrosion, plugging, and erosion are not likely to be obstacles in
further development of the liquefaction process.
Product mass and energy yields were high, with typically up to 361 yield
of oil, 18% yield of char, and combined energy yields of 73% of feedstock
heating value. Spontaneous product separation from wastewater did not occur
in all experiments, requiring a solvent extraction for recovery of products.
Later experiments performed with higher pressures of inert gas (nitrogen) did
achieve spontaneous separation of a higher quality oil product from waste-
water. Further work needs to be done on product separation to confirm this
result.
BOD determinations were performed 'ising an unadapted microbial seed from
a local sewage treatment plant. The results suggest that the wastewater from
the process would be highly biodegradable using an aerobic digestion. Bypro-
duct gas was also a minor environmental pollution concern. Metal concentra-
tions were as much as 27-fold lower in the product oil tnan in the starting
sludge, but were more concentrated in the char. Any toxic organic materials
2
-------
produced in this process are likely to be concentrated in the oil, and would
be destroyed during combustion. Disposal of ash from char combustion remains
a potential problem.
The study performed a preliminary economic comparison of the liquefaction
process with conventional incineration. Liquefaction appears to be economi-
cally more attractive, both on an initial capital investment and an operating
cost basis, with, for example, a sludge disposal cost of $43/dry tonne for a
city of 1 million, with a capital equipment cost of only $6.1 million.
3
-------
SECTION 3
RECOMMENDATIONS
The project described in this report demonstrated the feasibility of the
liquefaction process on a continuous scale. The optimism about this process
expressed in our previous report was thus fully justified. We continue to
feel that liquefaction offers a cost-effective alternative to incineration —
one that would be environmentally acceptable in practice. However, the
behavior of the sludge at elevated temperature and pressure was unexpected.
Based on our experience with the prototype, further work should be performed
with a modified reactor, to refine the parameters for a pilot plant.
Further work should be done on product separation, based on the pressure
effect, to ensure optimum product recovery and quality. Solvent extraction is
unacceptable in a commercial process for sludge conversion to fuel. Sponta-
neous separation of oil and char was observed in some experiments and not in
others, but a detailed examination of separation was outside the scope of this
project.
Liquefaction has shown the potential for concentrating metals in the
inorganic fraction of the char, and also for destroying organic materials
through alkaline digestion. The potential of the liquefaction process as a
novel technique for concentrating heavy metals and rendering organic wastes
harmless should be studied further, and accurate data on mineral balances and
destruction rates should be obtained.
An additional research project should be performed using a smaller-
diameter reactor so that extended runs can be performed. With the use of
sludge "spiked" with various metals and organic nonradioactive tracers, the
work outlined here on process parameter refining, separation, and detoxifi-
cation potential can be performed in a cost-effective manner.
4
-------
SECTION 4
MATERIALS AND METHODS
REACTOR DESIGN
General
Specific engineering details of reactor design are proprietary, but the
following general description, together with the diagram of the system (Figure
1.) and photograph (Figure 2) should provide a clear picture of the continuous
liquefaction operation. The reactor is a 15 cm o.d. vertical stirred tube,
with sludge injection from the bottom by means of a metering injection pump.
The reactor is electrically heated, originally with a 13 kw ceramic heater
wrapped around the lower 1.3 m of the reactor (the heating zone). In the
original design, the middle 1.3 m was designated as the reaction zone and was
not independently heated, while the upper 1.3 m was the cooling zone. Water
cooling was used, and was achieved by means of coiled copper tube around the
upper part of the cooling zone, prior to product exit from the reactor into
the pressure let-down system. In practice, as indicated in the next section,
the hot digesting sludge behaved as a very low viscosity fluid, so that plug
flow was not achieved, and heat transfer occurred between the heating and
cooling system. It was therefore necessary to install additional electrical
heating capacity (8.4 kw) around the reaction zone and move the cooling off to
the side. The system as it stands is capable of processing 144 kg/d of dry
sludge (or 720 1/d of 20% solids sludge). This compares, for example, with
the 1 1 scale batch reactors used in previous work by us and others (e.g.,
ref. 1).
To our knowledge this was the first time that the liquefaction process
had been applied to primary sewage sludge, continuously, on this scale. Other
continuous reactors in existence have used substrates such as wood or garbage,
and hence have encountered somewhat different problems. Based on the experi-
ence with wood liquefaction at Albany, Oregon (3 ton/day pilot plant) and with
bench scale work supporting this plant, we had expected problems with heat
transfer, pumping a viscous slurry, corrosion, char build-up (coking) of the
reactor walls, slow reaction times and a requirement for high temperatures.
On the whole, the prototype reactor worked very well, although we were unable
to reach the full range of reaction conditions originally designed for, due to
unexpectedly low feed/reactant viscosities inside the reactor. Each of these
factors will be briefly discussed, together with some suggestions for improve-
ments or further work.
5
-------
Flow and Valve Labeling Schematic
FIGURE 1. Flow and Valve Labeling Schematic
-------
-------
Size Criteria
The size of the reactor system was determined by considering many fac-
tors, including the desired volumetric flow, height of the crane in the
Mechanical Development Laboratory (where the reactor was located), Washington
State pressure vessel regulations, construction material availabi1ity, and
Battelle's pressure system safety guidelines. For instance, to avoid strict
and costly State codes the inside diameter of the pressure vessels had to be
less than 6", the volume of each vessel less than 5 ft , and the thermal input
less than 60 Kw. In considering all of these factors, a six inch nominal,
double extra strong pipe (Schedule 160 XXS, wall thickness 0.864 in) emerged
as the optimum shell for the pressure vessels.
A reactor system that was about 100 times smaller than the target pilot
plant design of 3 tonnes/day was desired. This amounts to 30 kg/d. Since
8-12 hr shifts were anticipated, this increased the desired size to 60 kg per
8-12 hr day, or 7.5 kg/hr, corresponding to a volumetric flow rate of roughly
7 1/hr. The design residence times were to be 0.5 to 3 hr. This expanded the
required volume for the 3 hr residence time to 21 1. Dividing this value by
the required cross-sectional area of the 6 in pipe yielded a length of 1.7 m.
Based on this and the height constraints, the reactor heating zone,
residence zone and cooling zone were each set at a length of 1.22 m (4 ft),
giving the residence zone a gross volume of 0.014 m^ or 14.7 1. This deter-
mined the metering pump capacity range as between 4.9 and 29.4 1/hr (the
currently installed metering pump has a range of zero to 60 1/hr).
The injector and water storage vessel were sized to have essentially the
same volume as the reactor, each let-down vessel having half the volume of the
reactor.
The maximum heating system duty should have been equal to the enthalpy
change of the sludge between 20 and 350°C plus losses. The maximum enthalpy
change at the maximum flow rate was therefore expected to be approximately 12
Kw. A heating system rated at 13 Kw was specified initially.
It was assumed that the high viscosity of sludge would inhibit signifi-
cant backmixing within the cooling zone and that the desired high temperatures
could be reached. This turned out to be incorrect and heat loss by reflux was
much greater than anticipated. The knowledge gained from this experience has
led to design changes that will prevent this problem in the pilot plant.
Heat transfer
Scraped surface heat exchange was selected after consideration of immer-
sion heaters and steam injection. Immersion heating was not selected due to
concern over charring problems. Steam injection, in addition to being
costly, would not work at high temperatures. External electric-fired radiant
heaters were selected based on cost, reliability and accuracy of control.
The radiative heat transfer to the outside surface of the reactor is
essentially instantaneous. The inside film coefficient of the scraped surface
8
-------
sludge is the slowest part of the heat transfer. The value of the heat trans-
fer coefficient was estimated using equation 1.1:
hi = 2*(kpcnb/Pi)'5 1.1
Where: k = the thermal conductivity in W/m^°C
p = the density in g/m
c = the heat capacity in J/g C
n = the scraper rotation speed in rps
b = the number of blades
hi = the scraped surface inside coefficient W/nr°C
(Source: McCabe, 1967 (Reference 3)).
It is important to note that this inside heat transfer coefficient is indepen-
dent of viscosity. This is important because the viscosity of the alkaline
sludge as a function of temperature is unknown. Based on this correlation, it
was determined that there would be sufficient heat transfer area to heat the
sludge.
Hydraulic scraper system
A hydraulic motor was used inside the reactor to turn the scraper because
mechanical seals were too costly due to the high temperatures anticipated.
Early in the design stage it was decided to put the flange on the top of the
reactor, for ease of maintenance and removal of the scraper. With the flange
on top, the scraper could be removed and the reactor cleaned without disman-
tling the entire system.
The hydraulic system was designed to turn the scraper at an adjustable
rate. A backpressure valve loaded by the headspace pressure was used to keeo
the pressure in the hydraulic motor higher than the pressure in th« reactor.
One of the key constraints was finding a hydraulic motor that would fit inside
the reactor. Once a motor was identified, the system was sized based on the
oil flow required by the motor at the desired speed and pressure.
Gas handling Sub-system
The gas line connections were the most complicated part of the system.
The purpose of the gas lines was to control the reactor pressure and minimize
the loss of high pressure nitrogen and reactor gas during the let-down
operation.
The gas handling system consists of numerous needle valves, seven elec-
tric ball valves, a cascade of nitrogen supply tanks, a large gas receiving
tank and an outside vent connection. There are three parts to the gas connec-
tions. The high pressure line conveys high pressure gas from the let-down
vessel to the intermediate line. The needle valves act as orifices between
the high pressure line and the intermediate line. The intermediate line
conveys the gas to the low pressure let-down vessel. When both let-down
vessels are at the desired pressure or during let-down operations, the storage
tank access line is used.
9
-------
In the original design and operation of the system, the reactor pressure
was controlled by loading the dome of two back-pressure valves. The back-
pressure valves worked well until the teflon diaphragm failed. Severe erosion
of the diaphragm was caused by particles and vapor droplets passing through
the orifice at sonic velocities. Several measures were taken to reduce this
problem, including the installation of a dogleg liquid collector and stainless
steel wool in-line filter. The doglegs were drained during reacto" operation
by manual control of the needle- and hand-operated ball valve at the bottom of
the dogleg. The in-line filter has a needle valve at either end so that it
can be cleaned during reactor operation. Problems with leaking back-pressure
valves were still encountered. Finally, two needle valves in series were
substituted for the back-pressure valves.
Control of the gas was accomplished by using the 7 electric ball valves
and the needle valve on the line to the storage tank and the needle valves
between the high pressure line and the intermediate line.
Reactor Instrumentation
The reactor instrumentation consisted of 19 thermocouples, 3 pressure
gages and 4 pressure transducers. The pressure transducers were connected to
the intermediate line, reactor headspace, injector water feed line and storage
tank. The pressure gages were located on the storage tank, injector feed
line, and a connecting line from the reactor headspace and intermediate line.
The data was collected and stored in a Hewlett-Packard data acquisition
system with real time video display output. This information was used in
conjunction with the gages to operate and control the reactor system.
Prior to a run we found it useful to leave the 24 V power supply on over-
night so that the transducers and the power supply could stabilize. This
reduced the amount of calibration necessary and improved accuracy. However,
sometimes the pressure transducers would read low by about 25%. Light tapping
with a copper tube "recalibrated" the transducers to the correct value. This
observation is important because it helps to explain some of the low pressures
indicated by the data traces of the early runs.
On leaving the reactor vessel, the hot pressurized reaction products
passed into one of two presrure let-down vessels, after being cooled. A
complex gas handling system we*, used to effect product transfer to a collec-
tion drum, control gases evolved from the reaction and pass them to a gas
collection vessel, and maintain a safe condition in the reactor itself (by
reducing boiling).
Figure 3 shows the arrangement of thermocouples on and in the reactor (if6
was used to monitor the nominal run temperature, since it was at the midpoint,
inside the reactor).
Testing and modification of the reactor and its appurtenances was an
important part of the research program. To avoid redundancy, details of the
work are presented later in this report (Section 5).
10
-------
1 Cooling water outlet
2 Outer vessel wali-top o< reactor
3 Sludge-exit thermowell
4 Cooling water outlet
5 Outer vessel wall-top of insulation
6 Sludge-mid thermowell
7. Outer wall-lower thermowell
8 Sludge-lower thermowell
9 Product line (not shown)
10 Lower wall outer surface
11. Lower wall outer surface
12 Lower wall outer surface
13 Lower wall outer surface
14 Mid wall outer surface
15 Mid wall outer surface
16 Mid wall outer surface
17 Mid heater thermocouple
18 Mam ceramic heater thermocouple
19 Supplementary heater thermocouple
©—
©—
©—
H) ©
©
©
=0 ©
FIGURE 3. Arrangement of Thermocouples in Reactor
11
-------
SLUDGE FEEDSTOCK PREPARATION
The continuous liquefaction reactor was designed to operate with a 20%
solids sludge feedstock, made alkaline with 5% of anhydrous sodium carbonate.
The feedstock was required to have as high an organic content as possible, and
hence to be a primary rather than a secondary sludge. The first portion of
this task was to identify a local Washington state municipality with a supply
of 20% solids sludge, or failing this, to obtain enough of a more dilute
material for delivery to Battelle, where it would be dewatered. After inves-
tigation, the latter alternative was selected, as no readily available source
of primary sludge of 20% solids content was available.
Initial shake-down tests on the reactor system were performed with water,
with secondary sludge from Portland (67% ash, dry basis), and with secondary
City of Richland sludge mixed with peat moss (48% ash, combined dry basis) to
increase its organic content. Only after these tests were successfully com-
pleted was the primary settled and centrifuged Richland sludge passed into
the reactor.
Primary settled sewage sludge from the City of Richland, WA, was deliv-
ered by truck to our Chemical Engineering Laboratory (CEL), where it was
transferred to 55 gallon drums. This material was obtained in several
batches, to minimize decomposition between runs. As quickly as possible after
delivery, 1% polymer was stirred into a drum, and the contents transferred
to a stirred holding tank for dewatering by centrifugation. The centrifuge
was a Bird 6" continuous model capable of taking 6 gal/min. of 3-4% solids
sludge and converting it to 18-20% solids material. The centrifuged sludge
(20% solids) was mixed immediately with 5% by wet weight of solid anhydrous
sodium carbonate, using the reactor feedstock progressive cavity p>np in a
recycle mode to ensure complete mixing. In this state, we observed no
digestion or decomposition of the feedstock (which was at a pH too high for
normal digestion) even in small samples kept at ambient temperatures for
several weeks. The mixed sludge was a thick, black, homogeneous and
gelatinous slurry which showed no indication of phase separation. It was
re-mixed prior to injection into the reactor. Drums of feedstock were weighed
and labeled after preparation.
ANALYTICAL METHODS
Sample Preparation
Bulk products obtained from continuous reactor experiments were collected
in 55 gallon drums, clearly labeled with the date, project number, and run
conditions. We attempted to use one drum of sludge feedstock and to collect
one drum of products for each run; this corresponded to a rate of 12 gallons/3
hr (one reactor volume used for residence time basis), or an approximate run
time of 14 hr. at a nominal 3 hr residence time. Each drum collected was
thoroughly stirred, and a ca. 200 ml sample of the homogeneous product taken,
placed in a glass jar, and transferred to the laboratory. The mixed sample
was accurately weighed. The weight of the sample was obtained by weighing the
empty jar. The mixed product containing water, oil, and char was extracted by
dichloromethane in 3 x 100 ml aliquots, placed in a separatory funnel and
12
-------
shaken for about 5 min. The oil and char suspension/solution in dichlorometh-
ane was separated from the water phase. The water phase was evaporated and
the residue weighed, ashed, and weighed again to obtain total solids, organ-
ics, and ash. The char was separated from the oil solution by mild centri-
fugation on a bench centrifuge at 3000 rpm for 5 min. The oil solution was
poured off and evaporated at 20°C in a rotary evaporator. The char was air-
dried and then dried in a 110°C oven overnight. Both oil and char were
weighed to obtain yields.
Following collection of analytical samples from the vigorously stirred
product, immediately after completion of a run, these drums were stored at
ambient temperatures until the separation of oil, aqueous phase, and char
was deemed complete. Bulk products were extracted with d ichl oromethane,
separated from aqueous phases, combined, and solvent removed. Oil, char, and
aqueous phases were subjected to various analytical tests. As far as possi-
ble, standard methods and ASTM-approved procedures were used. Data obtained
was recorded either in dated and signed laboratory notebooks or in a project
file.
Results obtained corresponded to char and oil definitions as materials
insoluble, and soluble in dichloromethane, respectively. In some measure-
ments involving oil, a sample was physically skimmed off the surface of the
settled product, and water decanted to leave the char. The latter char
product contained considerable amounts of entrained oil and is therefore
different from the char obtained by extraction. When such samples were used,
they were carefully labeled as 'skimmed*. The symbol 's' following any sample
designation in this report is used to distinguish skimmed oils and chars from
extracted samples. Since in a commercial application of this technology we
would plan to use skimming and decantation rather than solvent extraction for
product separation, the comparison of results between skimmed and extracted
products is of some value.
Gases evolved during the course of a run were collected in a gas collec-
tion vessel rated to 500 psig. A mid-run gas sample was taken for each run,
and in some cases 2 samples were taken and analyzed by gas chromatography, in
duplicate. The major components were nitrogen (used as a purge gas and pres-
sure control for the reactor) and carbon dioxide (product from liquefaction).
A log was kept of gas pressure in the collection vessel so that it was possi-
ble to determine the actual mass of each gas produced during a run. Some
water and organic volatiles also condensed in the gas collection vessel and
were periodically drained off.
One unanticipated result was the collection of a tar in the gas pressure
relief system. This material was of low volatility and must have formed there
from volatiles. The problem was solved by extension of the system with a dog-
leg attachment, which was periodically pcrged with nitrogen. The organic
product was collected from one run and was designated as a 'dogleg sample'.
It was small in quantity, and significant only from its potential to clog the
relief valve arrangement.
13
-------
SLUDGE ANALYSIS
City of Richland primary sewage sludge used for the liquefaction experi-
ments was dewatered by centrifuging following addition of 1% polymer, and then
addition of 5% by wet weight of anhydrous sodium carbonate. This material was
analyzed for ash, water, fat, and heat content, and its elemental composition
determined, as follows:
Ash Content
Prepared sludge (4-5 g) was accurately weighed and placed in a pre-fired
porcelain crucible, which was then heated in an oven at 750°C for 1 hr.
After cooling to room temperature in a desiccator, final weight was measured
and the ash weight obtained by difference. The ash contents obtained are
reported in the Results section as % ash in the sample. Duplicate analyses
were performed.
Water Content
Water contents of sludges were obtained in a similar manner as for ash,
but with heating only to 110°C overnight in a vacuum oven, followed by
reweighing. Water content was obtained by difference and reported as % water
in the sample. Total solids content and organic solids content were derived
from water and ash contents by subtracting percent water from 100%, and by
subtracting percent ash from the percent solids thus obtained, respectively.
Duplicate analyses were performed.
Fat Content
An accurately weighed sample of dried sludge (obtained from the water
content analysis above) was extracted in a dried Soxhlet thimble for 24 hr
with dichloromethane. The extract was evaporated under reduced pressure at
40°C, and the last traces of solvent removed at 110°C in a vacuum oven for 5
min before reweighing. The weight of extract was obtained by subtracting the
weight of the empty flask from the weight of flask plus extract. Duplicates
were run, together with a control consisting of an empty Soxhlet thimble.
Heat Content
Heat contents of dried sludge samples were determined in duplicate by
combustion in pressurized oxygen in a Parr bomb calorimeter according to ASTM
D-3286-77, using benzoic acid as a standard.
Ultimate Elemental Analysis
Carbon, hydrogen, oxygen, and nitrogen contents of sludge samples were
obtained in duplicate using a Perkin Elmer 240B elemental analyzer system.
Oxygen was determined directly, not by difference. Sludge samples were dried
at 110°C prior to analysis. Inorganic element components, including phos-
phorus and sulfur contents of oils were determined by X-ray fluorescence
directly on the oil samples, according to the procedure described in Anal.
Chem., 55( 12 ), 191 1-1914(1983); Ref. 4), rather than by the Inductively
14
-------
Coupled Plasma technique originally proposed, for reasons of cost-effective-
ness, and a higher sensitivity for sulfur and phosphorus.
OIL ANALYSIS
Oil is defined as dichloromethane-soluble material, after char removal,
unless otherwise specified in the text as a "skimmed oil". Skimmed oil was
removed physically from the surface of some products for viscosity, ash, pour
point and other measurements. The following measurements were made on the
oils from the continuous liquefaction runs 208-509 or from combined oil from
runs 208-313:
* Water content
* Elemental analysis (C,H,N,0)
* X-ray fluorescence analysis (S,P, metals)
* Heat content (ASTf4-D3286-77)
* Ash content
* Fractional distillation
* Thermogravimetry (in air and nitrogen)
* Differential scanning calorimetry (in air and helium)
* Viscosity, flash point, pour point
A cetane value engine test (ASTM 0-613-79) was planned, but it proved
impracticable to separate char and oil from the combined bulk product, since
this was obtained before a method was found to achieve spontaneous char/oil
separation. The combined product was deemed too viscous for the test, having
apparently thickened during storage at outside ambient temperatures for 6
months.
Analyses were performed as described below:
Water Content
The method we planned to use (based on hydrolysis of 2,2-dimethoxypropane
to acetone and GC measurement of acetone) proved to be too erratic in practice
with these oils; we substituted water measurement on a Karl-Fischer apparatus
'Photovolt Corp, Aquatest IV), which gave much more reproducible results.
Elemental Analysis
This was performed exactly as for the sludge feedstock, using the tech-
niques already described (Elemental analyzer and X-ray fluorescence).
Heat Content
This was performed exactly as for the sludge feedstock, by ASTM 0-3286-
77.
Ash Content
This was also performed by the same technique already described for the
sludge feedstock.
15
-------
Fractional Distillation
Fractional distillation was performed on a composite oil from runs 208,
305, 306. and 313 (designated 'bulk oil* since it was exhaustively methanol
extracted from the whole product from these runs, and was intended for use In
the fuel quality test). The test was performed using the apparatus cf ASTM
D—1160.
Thermoqravimetry and Differential Scanning Calorimetry
To broaden the range of information available on distillation properties
of the products, thermogravimetric (TG) and differential scanning calorimetric
(DSC) analysis was performed on oils from individual runs. TG analyses were
performed between 30 and 630°C in air and in helium, while DSC was performed
over the same temperature range in air and nitrogen.
Viscosity
Oil viscosities were measured on a Brookfield LVT viscometer equipped
with a small-sample adapter, sample jacket, and a Brookfield EX-200 tempera-
ture-control lad water bath. All measurements were made at 50°C, using
spindles 18 and 34 and speeds of 0.3 - 60 rpm. This apparatus is limited to
viscosities of 10,000 cp or less.
Pour Point
Pour points of oils were determined according to ASTM D-97, involving
heating in an oil bath without stirring, for extracted oil samples. Skimmed
oils flowed more easily, and so their pour points were determined by the
method of cooling to the lowest temperature which still permitted flow (also
ASTM D-97).
Flash Point
Flash points were determined using the apparatus and technique described
in ASTM D-396-79.
CHAR ANALYSIS
Char yield, ash content, and elemental analysis were all performed using
the same techniques already described for the sludge and oil samples. One
additi 'al test - combustion - was performed on char derived from the extrac-
tion ot combined oil from runs 208-313.
Char Combustion Test
A char sample (about 10 g) was accurately weighed and heated in a cruci-
ble. Combustion behavior was observed, including ease of ignition, type of
burning (flaming or glowing combustion), and ease of sustained combustion.
Ash residue weight was determined. The purpose of this test was to permit
evaluation of the char as an energy source for the liquefaction process.
16
-------
GAS ANALYSIS
Gas samples obtained during a run were collected in gastight mylar bags
and transferred to the laboratory for duplicate analysis on a Carle 3114 gas
rhromatograph equipped with a data processing system. Because of the large
amount of diluent nitrogen present in the samples, the accuracy of analysis of
trace gases was reduced.
WASTEWATER ANALYSIS
Analysis of the wastewater from the liquefaction runs was to include BOD,
COD, suspended solids, evaporation residue, pH.and dichloromethane extraction
on selected samples, rather than on samples from every run. Some additional
data was obtained to better quantify the nature of the wastewater from the
liquefaction process.
Solvent Extraction
Wastewater samples from 3 selected runs (100 ml each, in duplicate) were
extracted in a separatory funnel with 3 x 50 ml aliquots of dichloromethane.
The weights of solvent-extractable material were obtained by evaporation of
the solvent under reduced pressure in a weighed round-bottom flask, reweighing
with the extract, and taking the difference.
Suspended Solids
Water samples (100 ml) were filtered through previously weighed and
thoroughly dried quartz microfiber filters under vacuum. After drying the
filter and collected insoluble material at 100°C/1 hr under vacuum, the
residue weight was determined. This corresponds to 'char' suspended in the
water.
Evaporation residue
The aqueous phases from solvent extraction were filtered and evaporated
under reduced pressure in a rotary evaporator, using a previously weighed, dry
round bottom flask. Evaporation residue was then determined after completely
drying the material and flask at 110°C for 1 hr under vacuum and reweighing.
pH Measurement
This was done using a standard pH meter and reference buffer for
calibration.
COD Measurement
The wastewaters from these experiments were heavily loaded with organ-
ics. After a series of trials, it was found that a 1:200 dilution gave the
best results. Standard COD determination technique was used: A 50.0 ml water
sample (diluted) was oxidized by 25.0 ml of potassium dichromate (0.250 N)
under 2 hr reflux. The solution was then back-titrated with ferrous ammonium
17
-------
sulfate hexahydrate to determine the amount of unreduced dichromate. Subtrac-
tion of results from a 50.0 ml water blank gave the results shown in the next
section.
BOD Measurement
After initial experiments, a 1:8000 and 1:12,000 dilution of the waste-
water was used for BOD determinations. A standard microbial seed was obtained
from the Richland sewage treatment plant and diluted by 100, 50, and 25-fold.
Combination of the seed dilutions with the two wastewater dilutions gave six
determinations for each of the 3 wastewater samples examined. The initial
dissolved oxygen concentration was determined with an oxygen electrode for
these samples and also for 3 standards consisting of glucose/glutamine solu-
tions with seed (same dilutions as above), and 3 blanks consisting of diluted
seed in water. Samples were incubated for 5 days at 20°C and the final
dissolved oxygen concentration determined. 'Biodegradabi 1 ity' of organic
material in wastewater was determined as the ratio of B0D:C0D expressed as a
percentage.
Dissolved Solids and Ash Contents
In addition to the above data aqueous evaporation residue (dissolved
solids), and ash contents were determined for runs 418 and 502-509 to enable
us to better determine mineral balance and overall mass distribution during
liquefaction.
Anaerobic Digestion
Two attempts were made to obtain measurable degradation of organic waste-
water components under anaerobic conditions. In the first, a 4 1 fermenter
was loaded with 2 1 water, 3.3 g of fresh, 20% solids sludge seed, and 10 ml
of Run #313 wastewater, sparged with nitrogen, and stirred for 1 week. In the
second, a series of bottles was set up according to the procedure of Shelton
and Tiedje (Ref. 5), using a volume of wastewater calculated to provide a
concentration of organic solids of 50 mg/1. Theoretical gas evolution from a
glucose control was 4.7 ml.
18
-------
SECTION 5
RESULTS AND DISCUSSION
REACTOR TESTING
A first sample of sludge was obtained from the Portland facility to use
in reactor testing. This material was fully digest3d secondary sludge and
contained 68% of ash. It was felt that if the reactor could be operated with
such low-organic material, there would be little difficulty with sludge with a
higher organics, lower ash content.
Fabrication of the reactor was complete in July, 1983. Hydro-testing at
the fabricators was performed on the complete reactor assembly, to 2.6 x 106
kg/nr (3750 psi) with no leaks being detected. The system was then disman-
tled, transported to our Materials Development Laboratory (MDL), and reassem-
bled, together with all power, water, and other supply lines and instrumenta-
tion. A repeat hydrostatic test was performed and the reactor system readied
for startup testing. A series of initial problems developed, as anticipated
in a prototype system, and were dealt with. These problems, and their solu-
tions, are listed in Table 1. The reactor assembly was then taken apart for a
detailed inspection and replacement of stirrer blades in a new configuration,
on the basis of what was learned from two testing runs with Portland sludge.
The composition of the Portland secondary digested sludge starting material is
shown in Table 2, while Table 3 gives experimental results from the liquefac-
tion run at 250°C.
The problem of achieving a reactor temperature of above 250°C was
believed due to our use of a "low viscosity" feedstock. After installation of
additional heating capacity, one additional reactor test run was performed on
November 11, 1983, with the purpose of determining if there would be the same
heating problem with a thick s^dge as with a free-flowing sludge (i.e.,
Portland secondary). Since a suitable source of primary sludge was not avail-
able at the time, a thicker feed wdi simulated by mixing peat moss into Rich-
land secondary sludge. The sludge/peat moss mixture had the composition
shown in Table 4. Starting material was preheated in the reactor overnight,
at 100°C. The next day a continuous run (the first), was begun, maintaining a
maximum temperature at the top of the heating zone of 320°C, with a residence
time of 16% of pump capacity (approx. 3 hr). The run was terminated after 3
hr of continuous flow, and sample products analyzed. (Because the amount of
material passed through was relatively small, no material balance was
attempted). The temperature and pressure condition readout from the pressure
transducers and thermocouples is shown in Figure 4, as an example. Similar
19
-------
500
400
ro
O
<-> 300
CD
LU
O
s-'
Q_
U w
— 200
100
.2500
RUN TESTS
DATE 11/22/83
PRESSURE
8
TIME(HRS)
->000
.1500 ^
CO
LT>
Q_
(XI
en
1000 Co
CO
LU
Od
n_
.500
J I L
8
i«
8
a
FIGURE 4. Reaction Temperature and Pressure Profile, Run RS
-------
data from other runs is given in Appendix A1-A8. The analytical results on
the products are also shown in Table 4. Oil was produced during this run,
together with some char. The experiment was designated as Run RS, and all
samples labeled with this prefix.
TABLE 1. REACTOR TESTING EXPERIENCE
Item
Problem
Solution
1 Sludge injection piston would
not move smoothly.
2 Data Acquisition System (DAS)
inoperative.
3 Insufficiently high temperature
achievable.
4 Excessive wall temperature
reached (950°C); unexpected
because of thermocouple within
15 cm and wall thickness 2.5 cm.
5 Thermowell failed at 106 kg/m^
(1450 psi) and 250°C.
Maximum reactor temperature of
250°C achievable instead of
design 370°C.
Scraper assembly bent.
Steel piston replaced by nylon.
Return faulty board with system
to manufacturer for correction.
Drain water and inject sludge
slurry directly.
Battelle materials experts
consulted; reactor re-annealed
in place; thermocouple re-sited.
Emergency shutdown. Hydraulic
pump had been installed back-
wards, breaking thermowell.
Pump direction reversed and
thermowell replaced.
Caused by convection currents,
heat transfer to cooling coils,
rapid reduction of sludge vis-
cosity on heating. Alleviated
by removal of central scraper
blades, resiting cooling coils,
adding extra 8.3 kw heating.
Result of hydraulic motor
reversal. Blades replaced.
21
-------
TABLE 2. ELEMENTAL COMPOSITION OF PORTLAND SLUDGE
Element, %
C H N
0*
S
P Si
Fe
A1
Ca Na
%
23.3 3.7 2.2
14.1
0.3
1.8 17.5
3.6
4.8
2.6 1.5
* Determined directly; not by difference.
TABLE 3. INITIAL RESULTS WITH PORTLAND SECONDARY SLUDGE
SIudqe
Product
Moisture (%)
68.2
60
Ash (wet, %)
21.6
23
Ash (dry, %)
68
58
Heat Content*
5300 +/- 400
8100
(cal/g)
pH
7.3
Distillate (ml)
37-38 °C
_
0.5
80-90+
-
4.8
90-95
—
1.8
* Moisture, ash-free
+ Contained dibutyl
phthalate (plasticizer)
22
-------
TABLE 4. STARTING MATERIAL AND PRODUCT COMPOSITIONS. RUN RS
Materi al
C
H
N
Ash
Richland sec. sludge 21.3 3.0 2.3
22.6 3.1 2.6
Peat moss
48.9 5.1 0.7
47.6 5.0 0.6
26.4 3.1 1.3
26.7 3.1 1.2
27.2 3.2 0.2
29.3 3.5 0.0
32.5 8.1 1.0
30.7 8.1 0.8
38.4 6.2 2.0
37.2 6.1 1.2
76.2 10.1 2.2
73.5 10.1 2.1
Sludge + Peat Moss*
51.0
49.8
45.5
45.7
Sludge + Peat Moss
Char from LDV #1
Char from LDV #2
Oil from LDV #2
-0-
-0-
+ LDV = Let-down vessel.
* Percentage of dichloromethane extractable material in the
starting material was less than 1% (i.e., fats and oils).
Percentages of dichloromethane-soluble materials in char
from LDV#1 and LDV#2 were 24% and 37%, respectively.
REACTOR OPERATING EXPERIENCE
Tne first full experimental run with centrifuged primary Richland sludge
was begun on Wednesday, February 8 at a nominal 305°C and 1.5 hr residence
time (designated Run 208). The term "nominal" is used because the temperature
varied at different parts of the reaction zone, and residence time depends on
the density of water at reaction temperature, definition of reaction zone,
etc. Thermocouple #6 (Figure 3) was used as the temperature set point, in the
center of the sludge just above the main heaters. Temperature varied somewhat
during the run, as shown in the profile (Appendix A-1). Residence time was
defined as (time x reactor vo1.)/(vol. product), or 190 1 of total product in
6 hr from a reactor volume of 45.4 1.
Using the above definitions for temperature and residence time, the run
was completely successful with no problems, and with total feed conversion to
product.
Based on this result, with total feedstock conversion at 300°C, it
appeared that our original design assumption of 350°C was conservative, and
that on a commercial design a much lower operating temperature could be used.
However, we continued to try to meet the original design criteria.
23
-------
On February 9, the reactor was started up again to start a secono run
with Richland sludge. We quickly experienced a problem with a blockage in the
sludge injection to the injector from the drum. This was cleared (into the
line) by reversing the valve on the injection point from the sludge pump.
(Preheating the sludge did not appear to work well, because unless there is
continual stirring, sludge dries out on the drum walls and either chars or
dries, forming lumps).
The blockage was cleared, and the reactor started heating without any
more problems. We could not get above 305°C even with no throughput, so this
was the maximum temperature with the reactor in its configuration at that
time. Having decided that we could not reach 320°C, we decided to try for the
lower residence time, while monitoring the letdown vessel valve temperature.
We collected about 20 1 of product at 300°C and 0.75 hr nominal residence
time, before we decided that the valve was getting too hot. This was not
considered as "Run #2" because it was run at 300°C (not a set temperature) and
only 20 1 were collected, with no gas analysis possible.
It was also at this point that the plug from the injector, which should
have passed through without problem, plugged the injection line into the reac-
tor. The reactor was shut down, as we felt we had no chance of clearing this
safely while running. This was also a good point to inspect the reactor, re-
place valve seals, and install additional heating capacity. This was done by
rewiring the existing upper wall heaters and adding additional reactor
insulation.
On March 5, the reactor was started up again, at 14:00, following sludge
and reactor preparation. Appendix A2-A3 shows the temperature and pressure
profiles for this run, which was the longest one completed to date (29 hr).
During the run, two sets of conditions were maintained: 275°C/3 hr (EPA 305)
and 305°C/3 hr (EPA 306), and two drums of products were collected. Product
collected during changeover from one condition to another, and while heating
and cooling the reactor, were collected separately in a combined "intermediate
products" drum and were not included in the steady-state products collected.
After this long run, a planned shutdown took place. The reactor was
still behaving well. A further run, EPA 313 (275°C/1.5 hr) was completed
successfully on March 13 (Appendix A-4).
Run 418 (Appendix A5) was begun on April 18 at 8:45 a.m. and terminated
at 23:20 after filling one 55-gallon drum with products. Operational temper-
ature was 275°C with a 1.5 hr residence time. No problems were encountered.
Run 502 (Appendix A6) was begun on May 2, under the same nominal reaction
conditions as Run 418, and continued for about 14 hr before a plug developed
in the bottom of the reactor and the run was terminated. This was the first
time a reactor plug caused a shutdown. Run 507 (Appendix A7) was the contin-
uation of Run 502, and proceeded similarly, except that an emergency shutdown
had to be performed due to a ruptured diaphragm in the hydraulic system. Nei-
ther of these events was serious in itself, but required shutdown for correc-
tion. The run was completed on May 9 (Run 509) (Appendix A8), after a total
of 7 samples had been collected. Completion of this run also completed Task
1.
24
-------
Comments:
Runs 208-313 in the continuous reactor demonstrated the essentially
complete conversion of sludge to oil, char, water-soluble organics, and gas at
temperatures as low as 275°C, with energy yields of char and oil combinod of
up to 73% (as shown by the yield calculations on p. 44). However, the spon-
taneous separation of oil and char from the wastewater which was observed
during autoclave experiments did not always occur with the continuous reactor
products. After discussion of the various possible reasons for this, it was
decided to perform subsequent runs with an increased nitrogen overpressure, to
determine whether this would improve separation. Run 418 was the first in
which this was attempted, with good success. In Runs 502-509, we endeavored
to establish the effect on a reproducible basis, again successfully. Product
separation occurred spontaneously within 24 hr in most cases, although some
oil was still entrained with the char. The wastewater was a transparent light
brown. Because samples were taken mere frequently than normal, although there
was at least one letdown vessel volume of product collected between samples,
equilibrium conditions were not attained at each sampling step. Also, in
sample 507-6A, the product remained in the letdown vessel much longer than
normal at the end of a run.
Although it was our original intention to obtain mass balances in bulk,
through relating weight of sludge input to weights of products out, the fact
that we did not have plug flow as anticipated originally made this impossible
to achieve. The sludge/water/gas/product mixture in the reactor at tempera-
tures above 275°C behaved as a liquid of very low viscosity. A calculation
based on gas production during Run 208 and shown below indicated that the
dissolved and undissolved gas concentration in the reactor corresponded to
about 1.3 times the reactor volume at STP:
Gas Production Rate Calculation:
Logbook entry: 40.9 psi at time 1, temp. = 316°K
83.2 psi at time 2+20 min
Reactor volume: 45.42 1 (12 gal)
Gas tank volume: 492.05 1 (130 gal)
Gas produced in 20 min = (83.2/40.9) x 492.05 1 at 316°K
= 998.9 1 at 316°K
= 2997 1/hr at 573°K and 1 atm.
= 40.25 1 at 573°K (300°C) and 135 atm.
Sludge flow: 45.42 1/1.5 hr, = 30.28 1/hr
Gas volume = 40.25/30.28, = 1.33 x Sludge volume.
Coupled with the fact the reacting sludge no longer behaved as a thermoplastic
fluid, this means that effectively the reactor was a stirred tank. Hence,
25
-------
with the volumes of sludge passed through and with time and funding
constraints, steady state conditions were not achieved and samples obtained
cannot be directly related to input sludge. To obtain steady state conditions
with systems such as this, it is our experience that between 5-10 full reactor
volumes have to be passed through under one set of reaction conditions, where-
as we had planned for allowing one 'intermediate' reactor volume of products
between samples for every change of conditions. Further work with a smaller
diameter reactor would allow us to obtain steady state mass balance results.
For present purposes, small (ca. 200 ml) samples were taken from the drums of
thoroughly stirred collected products, and were used for analysis. Bulk
products were separated as described in the Materials and Methods section.
DISCUSSION OF REACTOR OPERATION
Corrosion and char build-up
Operation of a liquefaction reactor with sewage sludge at high pressure,
temperature, and under alkaline conditions required careful selection of
reactor material, to avoid the problems of stress corrosion cracking caused by
chlorides and alkali. Another potential problem with reactor material
occurred when hydrogen in excess of anticipated values was produced during the
first runs. Hydrogen can cause embrittlement of steel under certain condi-
tions. In this case we determined that the amounts were still too small
for this to become a problem. However, potential corrosion was a concern, and
the reactor walls were carefully checked, together with the scraper assembly,
whenever the reactor was opened. At no time was any sign of corrosion noted,
even on surfaces which were not scraped. Also, there was no build-up of char
or coke on any heated surface, during the approximately 100 hours of reactor
operation. Our tentative conclusion is that corrosion and coking should not
present a problem in commercial practice, provided that reactor materials are
carefully selected.
Erosion
One problem which did occur to some extent was that of erosion of teflon
seals on the pressure relief valves, caused by high velocity mineral particles
entrapped in the out-flowing gas. It is possible that these particles helped
inside the reactor to prevent coking, by their abrasive action. They were a
nuisance in the gas lines until a steel wool trap was used to filter them
out. Future designs will need to be evolved that can take care of sand
particles traveling at high velocity.
Heat transfer
1. Reactor heating - According to the original design, plug flow of
sludge and products would occur inside the reactor due to the high viscosity,
narrow diameter, and slow upward movement of the feed. Horizontal heat trans-
fer would occur through the scraped stirrer blades, which would cause effec-
tive transverse mixing between the wall and the cooler center of the reactor.
On passing upward through the heating zone and the reaction zone, the products
would be cooled by the water cooling coils wrapped around the top third of the
reactor. This whole theory was based on an assumption of a relatively high
26
-------
viscosity of feed and products, and seemed reasonable even though little was
(and is) known about sludge behavior at elevated temperatures. In fact,
repeated discussions between the project staff and experienced sanitary engi-
neers failed to raise any concerns that the s1 ige would not behave as expec-
ted. Reactor design was reviewed by our industrial sponsor. Battelle safety
engineers, and others prior to construction. In view of our experience to the
contrary, the actual behavior of hot sludge inside the reactor should be taken
as a contribution to sludge engineering technology. The material behaved as a
very low viscosity fluid as soon as reaction temperature was reached (about
265°C), and further designs for liquefaction reactors should be based on this
fact. Total mixing occurred between incoming sludge and products, and was
easily observed on the thermocouple readouts, as temperature excursions
between upper and lower parts of the reactor were frequent and significant.
In some cases, the heaters had to be turned off for several minutes while the
reactor continued to heat up without additional heat input; in other cases,
cooling occurred with heaters on maximum. Control of the reactor under these
circumstances was a matter of operator skill. However, there was no danger of
uncontrollable overheating as the reactor had been designed completely fail-
safe, and the liquefaction reactions take place endothermically.
Because of the vertical reactor configuration, there was little that
could be done to modify the reactor so as to achieve its original design
temperature of 350°C. By putting in additional heating capacity, moving the
cooling coils over to the side, removing the central stirrer blades, and
adding wall insulation to the top of the reactor, we were able to raise the
maximum achievable temperature to, in one case, 320&C, although normal oper-
ating maximum was 305°C. The mobility of the reactor contents caused a
designed plug flo* to become a stirred tank reactor. This did not interfere
with achievement of project objectives, but raises design questions for the
future. Use of a narrow diameter, tilted or horizontal reactor would minimize
vertical mixing and heat transfer problems, but would increase overall reactor
length. This question needs to be studied in more detail.
2. Product cooling - As a result of the mixing of reactants vertically in
the reactor, cooling of products prior to entry into the let-down vessels
became a problem. The pressure valve between the reactor and the let-down
vessel was rated at below 100°C because of the plastic internal seal. The
connection between the top of the reactor and the entrance to the valve was a
1.8 m length of 5 cm o.d. steel tube. This was wrapped with copper cooling
coils, but the cooling capacity was inadequate to maintain the 100 degree
maximum outlet temperature required, except at low flow rates. Originally,
residence time in the reactor was defined as the time required for a particle
to pass into and out of the reaction zone - one third of the reactor length.
Because of mixing, the whole reactor had to be redefined as the reaction zone,
so that a 3 hr. residence time by the original definition became a 9 hr.
residence time by the new, and realistic definition. As it turned out, a 1.5
hr. residence time was quite adequate for complete liquefaction to occur
(corresponding to 0.5 hr. by the original definition). This was an excellent
result in that the liquefaction was complete in a third of the minimum time
previously considered likely, but it raised problems with sludge flow and
product cooling. Future designs will require much more cooling capacity and
27
-------
higher flow rates through the reactor, or replacement of the let-down valves
with more temperature resistant (and expensive) varieties.
SLUDGE COMPOSITION
The sludge starting material was extensively analyzed so that product
compositions could be related back to the feed. The following tables (Tables
5-8) present the results of sludge analysis. More detailed analytical data on
trace elements in the feedstock is given in the Appendix (B-l). A discussion
of the effects of liquefaction on feed and product comp-sitions and chemistry
follows the data on major product analysis (oil and char; p. 43).
TABLE 5: SLUDGE FEEDSTOCK COMPOSITION
Run # Water Solids Organics Ash Fat Heat Value*
X % % % % Measured Average
cal/g
208
79.27
20.73
65.26
34.74
9.61
3189
-
-
305
76.42
23.58
66.07
33.93
2.25
2961
2943
2952
306
81.16
18.84
64.33
35.67
3.67
3295
3285
3290
313
80.22
19.78
GS.37
34.63
3.80
3356
3293
3325
418 #1
79.23
18.77
67.95
32.05
3.51
3040
3063
3052
418 #2
79.78
20.22
65.92
34.08
4.14
3224
3288
3256
502-1
79.79
20.21
62.26
37.74
2.76
2872
3004
2838
502-2
Same '
feed as 502-
-1
502-3A
79.41
20.59
61.33
38.67
2.73
2725
3013
2869
507-4
79.82
20.18
62.81
37.19
2.76
3109
—
—
507-5
Same f
feed as 507-
-4
507-6A
79.33
20.67
62.81
37.19
2.94
2944
2948
2946
509-6
80.60
19.40
65.19
34.81
3.39
3216
3144
3180
Blank
—
m,
—
-
0.00
-
-
-
m
—
-
-
0.04
-
-
-
Benzoic acid used as standard: 6333(measured)
6318 (actual)
* To convert cal/g to BTU/lb dry sludge, multiply by 1.8
28
-------
TABLE 6: SLUDGE FEEDSTOCK ULTIMATE ANALYSIS
Run # C H N 0
Z Z Z Z
208 38.80. 39.0 5.51. 5.78 1.72, 2.02 24.61, 20.03
305 29.74. 31.38 3.93. 4.50 1.05. 1.45 30.98. 28.61
306 33.98, 35.01 4.48, 4.78 1.62, 1.89 29.03. 29.58
313 33.20, 33.05 5.05, 5.10 1.86, 1.84 29.35, 30.76
418 Drum #1 32.05, 33.80 4.66, 4.82 0.90, 1.35 27.71, 25.57
Drum #2 31.24, 31.94 4.43, 4.46 1.14, 1.20 27.64. 25.71
502-1 23.76. 29.66 3.52, 4.31 0.11, 0.73 23.53, 26.59
502-2 Same feed as 502-1
502-3A 22.64, 27.66 4.58, 4.44 1.00, 1.04 26.06, 26.88
507-4 34.12, 30.44 5.08, 4.58 1.01, 1.07 27.52. 23.87
507-5 Same feed as 507-4
507-6A 28.62. 29.76 4.81, 4.48 1.08, 0.89 28.32, 27.18
509-6 32.42, 30.92 4.77, 4.78 1.20, 1.07 27.10, 28.37
Cellulose 41.87, 42.12 6.21, 6.23 -0- -0- 45.51, 45.38
Cellulose,
actual 41.62 5.82 -0- 46.20
Oxygen determined directly, not by difference.
29
-------
TABLE 7. SLUDGE FEEDSTOCK NORMALIZED ULTIMATE ORGANIC ANALYSIS*
Run #
C
H
N
0
Ash
Z
or
+
%
I
7.
208
59.66
8.66
2.87
34.20
34.74
(38.94)
(5.65)
(1.87)
(22.32)
305
46.25
6.39
1.89
45.10
33.93
(30.56)
(4.22)
(1.25)
(29.80)
306
53.63
7.20
2.74
45.66
35.67
(34.50)
(4.63)
(1.76)
(29.31)
313
50.68
7.77
2.83
45.98
34.63
(33.13)
(5.08)
(1.85)
(30.06)
418
48.20
6.86
1.72
39.83
33.07
(32.26)
(4.59)
(1.15)
(26.66)
502-1
42.90
6.30
0.67
40.25
37.74
(26.71)
(3.92)
(0.42)
(25.06)
502-3A
41.01
7.35
1.66
43.16
38.67
(25.15)
(4.51)
(1.02)
(26.47)
507-4
51.39
7.69
1.66
40.92
37.19
(32.28)
(4.83)
(1.04)
(25.70)
507-6A
46.47
7.40
0.72
44.18
37.19
(29.19)
(4.65)
(0.45)
(27.75)
509-6
48.58
7.33
1.75
42.55
34.81
(31.67)
(4.78)
(1.14)
(27.74)
Overal1
average:
48.88
7.30
1.85
42.17
35.76
(31.44)
(4.69)
(1.20)
(27.09)
* Adjrsted to averaged, dry ash-free organic basis.
Ash-included basis values in parentheses.
30
-------
TABLE 8. MAJOR INORGANIC ELEMENTS IN SLUDGE FEEDSTOCK
Elemental Concentration,
%, +/- probable error
Run*
A1
Si
P
S
CI
K
Ca
Ti
Fe
208
0.705*
2.69
1.254
0.905
0.147
0.309
4.57
0.187
1.188
0.089
0.15
0.071
0.048
0.009
0.016
0.23
0.010
0.059
305
0.425
2.60
0.947
0.744
0.099
0.164
1.881
0.102
0.908
0.078
0.14
0.050
0.040
0.007
0.009
0.094
0.005
0.045
306
0.352
2.03
0.845
0.612
0.070
0.150
1.979
0.106
0.783
0.072
0.11
0.050
0.033
0.005
0.008
0.009
0.006
0.039
313
0.139
1.573
0.768
0.608
0.071
0.144
2.11
0.096
0.660
0.071
0.093
0.047
0.033
0.006
0.008
0.11
0.005
0.033
418
0.674
2.11
0.792
0.687
0.093
0.169
2.25
0.113
0.848
#1
0.091
0.12
0.050
0.038
0.007
0.009
0.11
0.006
0.042
418
0.534
1.615
0.801
0.709
0.113
0.173
2.09
0.112
0.827
#2
0.079
0.095
0.040
0.039
0.007
0.009
0.10
0.006
0.041
502
0.309
1.90
0.749
0.663
0.107
0.168
1.932
0.109
0.793
-3A
0.075
0.11
0.046
0.036
0.007
0.009
0.097
0.006
0.040
507
0.233
1.671
0.700
0.664
0.110
0.168
1.921
0.111
0.722
-4
0.070
0.096
0.043
0.036
0.007
0.n09
0.096
0.006
0.036
509
0.425
2.28
0.791
0.655
0.107
0.187
2.27
0.115
0.818
-6
0.078
0.13
0.048
0.036
0.007
0.010
0.11
0.006
0.041
* Not all sludge samples analyzed for inorganics;
t IAEA Soil-5, 1633A Flyash, and NBS Coal used as standards -
complete analysis for comparison is given in Table 9;
* Ti source values, except Ti, Fe, used Zr source.
-------
OIL COMPOSITION
Oil is defined as dichloromethane-soluble material, after char removal,
unless otherwise specified as a "skimmed oil".
A cetane value engine test (ASTM D—613—79) was planned, but it proved
impracticable to separate char and oil from the combined bulk product, since
this was obtained before a method was found to achieve spontaneous char/oil
separation. The combined product was deemed too viscous for the test, having
apparently thickened during storage. However, in commercial practice, this
product would be burned as a boiler fuel in a sewage treatment plant rather
than used as a diesel oil. For diesel oil use, refining of the crude product
will be needed.
Water and Ash Content, and Heating Values of Product Oils
Water contents of oils generated from sewage sludge are shown in Table 9,
together with heat and ash contents and adjusted heat values.
TABLE 9: WATER, HEAT, AND ASH CONTENTS OF OIL PRODUCTS
Run #
Water
Heat Value
Ash Content
Adj. Heat Value
(%)
(cal/g)
(%)
(cal/g, ash-free)
Measured
Ave.
Measured
Ave.
Measured
Ave.
208(s)
ND
8741; 8912
8827
ND
208
7.94; 8.78
8.36
8940
8001
8347
8.1; 6.6
7.4
9909
8100
305
9.37; 10.10
9.73
7835
7768
7802
5.7; 5.7
5.7
9926
306
4.46; 5.8b
5.16
8569
8601
8585
6.3; 6.3
6.3
9696
313
7.68; 9.17
8.43
8716
8667
8692
3.2; 3.5
3.4
9858
418
5.69; 5.39
5.54
8809
8814
8812
0.93 0.65
0.47
8854
502-1
0.53; 0.33
0.43
8627
8621
8624
ND
—
—
502-2
3.88; 3.06
3.47
8553
8530
8542
5.7; 5.5
5.6
9436
502-3A
ND
—
8951
8989
8970
3.7; 3.5
3.6
—
507-4
ND
—
8098
8153
8176
2.3; 1.9
2.1
—
507-5
12.1; 12.1
12.1
8686
8686
8686
14.7; 15.
8 15.4
11984
507-6A
9.10; 8.70
8.90
8193
8146
8170
3.4; 3.8
3.6
9337
509-6
ND
—
7674
7681
7678
15.9; 16.
6 16.3
—
509-7A
5.02; 5.10
5.06
8500
—
—
1.7; 1.7
1.7
9116
509-7B
ND
—
8487
8491
8489
ND
—
—
ND - Not Determined
* Ash and water contents not determined on every sample. Values were
determined after heating to 750°C for 8 hr.
+ - High ash contents in #507—5 and 509-6 may be due to char inclusion.
Heat content determined according to ASTM-D-3286-77 throughout.
32
-------
The heating values of oil products, determined by bomb calorimetry, were
all very high. The overall average value was 8460 cal/g, with a range of 7678
to 8970 cal/g, on an as-recovered basis. The individual values are all too
close together to indicate any effect of reaction conditions, and apparently
the same product was obtained throughout. These values compare with n-octane
(11,410 cal/g), being between 67.3 - 78.6% of the n octane value, as recov-
ered. On a dry, ash-free basis, for example, this translates to a value of
86.8% of n-octane for run 208 oil. Water content of the oils varied widely
between less than 1% to over 12%, while ash content was between 2.1 to over
16%. These variations almost certainly represent differences in product
homogeneity rather than bona fide changes due to reaction parameters.
Elemental Analysis
Organic elements (C, H, N, 0) were determined in duplicate on a Perkin-
Elmer elemental analyzer. Results are shown in Table 10, and are for as-
extracted oils (including small amounts of ash and water).
TABLE 10: MAJOR ELEMENTAL COMPOSITION OF PRODUCT OILS
C H N n
Run #: % % % %
208(s)
74.28,
74.30
10.99,
10.54
1.18,
0.81
ND
208(2)
75.77,
76.57
10.43,
10.07
2.02,
1.96
9.29,
9.31
305
71.47,
71.06
10.03,
9.50
5.10,
2.67
11.32.
10.33
306
69.51,
71.74
9.99,
9.84
2.45,
2.68
9.47,
9.38
313
73.31,
73.31
10.45,
10.03
2.82,
3.20
10.76,
11.11
418
74.83,
74.69
10.31,
10.08
3.36,
2.82
8.46,
8.96
502-1
78.57,
72.74
10.18,
10.62
1.78,
2.19
11.42,
11.21
502-2
73.07,
72.83
10.83,
11.00
2.85*
16.65,
15.95
502-3A
73.84,
74.68
10.36,
11.49
2.09,
2.80
12.37,
12.01
507-4
67.05,
67.85
10.41,
11.05
1.34,
2.05
12.43,
12.75
507-5
72.90,
73.29
10.76,
11.04
1.83,
1.84
ND
507-6A
72.11,
73.19
10.34,
9.92
1.99,
2.15
9.87,
9.60
509-6
72.32,
72.25
10.64,
10.05
2.02,
2.58
ND
509-7A
72.24,
71.04
10.89,
10.09
2.44,
0.77
9.03,
9.62
509-7B
73.94,
73.75
10.69,
10.19
2.94,
2.40
ND
Cellulose
41.87,
42.12
6.21,
6.23
-0-
45.51,
45.38
Cellulose,
44.44
6.22
-0-
49.34
actual
ND = Not determined
* bingle determination due to small sample.
Overall averages: C, 72.95%; H, 10.43%; N, 2.31%; 0, 10.97%
33
-------
Heavy metal, phosphorus and sulfur contents of oils were determined by
X-ray fluorescence spectrometry, directly on the oil samples Summary results
are shown in Table 11; corolete data are given in Appendix B3.
TABLE 11: SUMMARIZED INORGANIC ELEMENTAL COMPOSITION OF PRODUCT OILS
Element (ppm): A1 Si P S CI K Ca Fe Cu Zn
Run #:
208
1530
417
518
5870
656
15
163
561
48
108
305
2370
2060
960
8550
1860
181
1604
1966
112
183
306
2460
530
350
6270
22200
46
411
777
15
43
313
1530
410
550
5830
247
72
372
478
13
42
418
4650
280
510
9900
6380
39
85
782
7
59
502-1
1680
220
950
11140
285
116
53
352
6
17
502-2
1750
340
660
8790
6630
54
42
137
6
20
502-3A
2060
320
960
11950
563
9
22
79
4
5
507-4
2050
1330
690
8950
2330
129
41
135
4
11
507-5
1640
1770
770
10680
2700
82
41
321
4
6
507-6
2040
620
220
5710
35700
42
22
269
3
7
509-6
2860
260
750
9830
6980
65
25
114
7
12
509-7A
1950
2180
232
5450
18280
59
20
131
4
9
509-7B
1920
390
1000
11270
1024
8
45
58
1
13
Fractional distillation
A 250 ml sample of bulk oil was obtained by exhaustive methanol extrac-
tion of the bulk product from runs 208-313 containing char and water, followed
by removal of most of the methanol under vacuum on a rotary evaporator.
Considerable difficulty was encountered with sudden foaming during the
evaporation, particularly towards the end when water was being removed. This
difficulty continued when we attempted a vacuum fractional distillation. The
following fractions were obtained without difficulty:
Fraction 1 - Clear white, 14 ml.
Fraction 2 - Cloudy white, 10 ml. (slight yellowing overnight).
Fraction 3 - Pale yellow aqueous (15 ml) under yellow oil (2.5 ml).
Fraction 4 - Same two phase product as #3, pot temperature increasing
rapidly from 115 to 192°C., while distillation slowed
(2 ml).
Fractions 1 and 2 were mostly methanol, while fractions 3 and 4 were mostly
water with some light distillate. Gas chromatography/mass spectrometry anal-
ysis of these fractions was performed, but apart from methanol and water,
the composition was too complex to permit identification of individual
compounds.
34
-------
The oil sample had been produced some 8-9 months prior to the experiment,
had been stored in a mixed dichloromethane/water environment in 5 gallon drums
at ambient temperatures outside, and then had been dewatered, had solvent
removed, and extracted again with methanol to extract oil into solution and
away from residual char. The resulting material was much different from a
freshly prepared oil, being viscous and tarry. Presumably the combination of
conditions resulted in continued increase in molecular weight during storage.
Foaming made vacuum distillation impossible, and made careful control essen-
tial throughout the atmospheric distillation as well. There was some smoking
during collection of fraction 3, and non-uniform bubbling during collection of
#4. A further attempt at vacuum distillation was made on the cooled residual.
At a temperature of 247°C under 34 mm Hg, the material appeared to decompose.
The residual from this resembled a hard wax or plastic material rather than a
coke, and consisted of 80% of the original material. This is all consistent
with a polymerized product, and is in complete contrast to the results from
our previous project (2) on conversion of sewage sludge to a synthetic asphalt
material by liquefaction and fractional distillation. In this case, a
freshly- prepared oil was vacuum distilled to yield a light oil and an
asphalt. The light oil remained stable at room temperature after 3 years. The
conclusion to be drawn from this is that the fresh product oil from future
experiments should be decanted and either burned or fractionally vacuum
distilled within a month of production, for best results. In light of this
unanticipated result, there seemed to be no point in attempting to proceed
with an engine test. This should be performed with a fresher sample. Addi-
tional work on oil aging under various storage conditions would also be very
useful. The fact that mass spectrometric (GC/MS) analysis of distillate
fractions of the 'old' oil derived from methanol extraction of the bulk
product showed a highly complex result, an estimated 1000+ compounds being
present, supports the hypothesis of continued reaction of oil components after
production.
Differential Scanning Calorimetry and Thermoqravimetry
Distillation characteristics of the product were determined by DSC and
TG. Table 12 shows the temperatures for 50% loss of volatiles from the oils,
together with the final residue percentage, in air and in nitrogen. The lower
temperature for 50% weight loss in helium versus in air is attributed to the
higher diffusion rate or organic molecules into a helium atmosphere, while the
lower final readings encountered for the runs in air are likely due to actual
combustion of the residual material.
The actual TG and DSC curves are presented in Appendix C. They vary
between a continually increasing rate of volatile loss in inert gas due to
distillation (e.g., run 208), to an almost constant rate of volatilization
(e.g., runs 305, 306, 313, 502), to examples where there are significant
discontinuities in volatilization rate (such as in runs 502-3A and 509-7B).
The curves which were produced by heating oils in air have the same general
shape as those obtained in nitrogen, but with 'humps' corresponding to oxi-
dation of components. Examples of this latter case are runs 208 and 502-2
(Appendix C). The DSC results are useful to indicate whether a 'hump' is due
to oxidation or to endothermic volatilization of a component, since the pres-
ence of a valley or a peak, respectively, on the appropriate chart provides
35
-------
TABLE 12. VOLATILIZATION OF CIL BY THERMOGRAVIMETRY
50% Wt. loss temp in: Residue at final temp.
Air Helium Air Helium
Run # % at °C % at °C
208
320
250
14.24
530
10.63
530
305
420
400
25.55
520
25.13
520
306
310
290
2.80
620
11.32
590
313
400
380
8.19
620
10.98
620
418
3n0
310
4.00
630
19.00
620
502-1
410
'60
7.37
620
16.95
620
502-2
350
c70
4.60
620
12.20
620
502-3A
275
270
0.80
620
9.33
620
507-4
410
280
8.31
620
13.14
620
507-5
410
310
8.92
620
15.18
620
507-6A
360
290
3.04
620
12.64
620
509-6
350
280
4.03
620
12.21
620
509-7A
385
280
10.33
620
13.24
620
509-7B
290
260
0.40
620
10.50
620
a clear demonstration. Since DSC measures the rate of heat absorption or
evolution rather than a cumulative loss in mass, it is sometimes easier to
interpret than the equivalent TG result. However, DSC results with these oils
are quite variable. For example, oil from run 313 shows a significant exo-
therm in air, beginning at 270°C and continuing to the end of the measurement
at 390°C. This was not present in nitrogen, and is therefore due to oxida-
tion. A similar result was obtained with oil from run 502-1. Other oils gave
almost identical curves in air and in nitrogen, while that from run 507-4 has
a series of endotherms from 70-150°C (perhaps due to loss of volatiles and
water). Oil from 509-7A shows two sharp oxidation peaks at 350°C which were
not present in the ostensibly identical sample from 7B. It is apparent from
these results that, although general conclusions can be drawn on oil quality,
there is still much to be learned about the variations in properties resulting
from different production, extraction, and aging conditions. This fact is
brought out even more emphatically by the fractional vacuum distillation
results.
The TG results were far different from fractional distillation. For
example, run 208 oil left only a 10.6% residue after heating to 530°C in
helium, and only a 14.2% residue in air. This was an average result. Appar-
ently the bulk oil had undergone some changes during storage outside for
6 months. The sample given as an example was 162 days old at the time of
analysis. (The1age of oils at the time of TG and DSC analysis can be derived
from the difference between the run number (Run 208 for example was performed
on 2-08-84, run number being the date), and the date on the TG or DSC chart in
Appendix C. Where relevant, ages of oils at the time of analysis are included
in the text).
36
-------
Oil Viscosity
Attempts to determine viscosity of oils from runs 208-418 directly were
unsuccessful; apparently the dichloromethane extraction process results in a
more viscous product than is obtained by skimming the oil from the surface of
the aqueous phase. This may be due to removal of the last traces of water
from the oils, loss of volatiles during removal of the dichloromethane solvent
from the extract, or an increase in molecular weight of the oil due to heatinq
during extraction and solvent removal. Viscosities from runs 208-418 were all
above 10000 cp (i.e., above the measurable limit with the apparatus) while
viscosities determined on skimmed oils from runs 502-509 were generally below
1000 cp. Pour point data therefore substitutes for viscosity measurement for
the earlier runs. There are three reasons why we believe the viscosities of
oils from later runs are much lower than those from the first runs: Different
reactor conditions, oil viscosities measured much earlier (the ages of the
oils listed in Table 13 were all between 79-88 d at the time of viscosity
measurement, while earlier runs were all above 100 days prior to analysis)
and the presence of water in the oil. For the latter reason, results on
samples 502-1, 507-5, and 509-7B are not included here, as they are probably
highly aqueous.
Viscosities of oils from runs 502-509 are shown in Table 13, and were
determined on a Brookfield viscometer using spindle numbers and speeds indi-
cated, at 50°C after a 20 min equilibrium period. It was difficult to obtain
constant readings, suggesting that these oils may have some non-Newtonian
fluid properties.
Based on this viscosity data, the oils may be too unstable for use if
stored for longer than 2 months (60 d), unless the increase in viscosity can
be retarded by use of an antioxidant. However, use in a sludge treatment
plant would be expected to be within the 60 d period. An alterr tive would be
to perform a fractional distillation immediately after production, if foaming
problems can be overcome.
TABLE 13. OIL VISCOSITIES
Run #:
Spindle #
0.3
0.6
1.5
Speed
3
6
12
30
60
502-2
18
240
190
160
165
127
94
63
42
502-3A
18
240
105
138
183
150
100
46
33
507-4
18
—
1438
1170
725
393
245
—
-
34
-
—
-
-
—
55
50
45
507-6A
18
490
295
198
146
112
89
67
—
509-6
18
315
300
182
141
275
—
-
—
34
-
500
320
280
230
245
292
305
509-7A
18
1350
990
796
774
-
—
—
-
34
5200
3200
1960
1620
1400
1225
756
451
37
-------
Pour point
1. Extracted oils - Approximately 5 g of oils from runs 208, 306,
502-3A, and 507-5 were heated in an oil bath. Product 502-3A became soft at
107°C and the other samples at 111°C, but none became pourable. Softening
appeared to continue to 125°C, but at 130°C the oils became thicker. On
continued heating to 155°C, the thickening continued, and on cooling to room
temperature the originally soft products became hard, although still plastic.
A repeat of this experiment with stirring (although ASTM D-97 does not permit
stirring) gave identical results. Larger samples of extracted oils used for
drying and ashing appeared to melt in the oven between 105-115°C.
2. Skimmed oils. Results of pour point tests according to ASTM D-97
(where oils are cooled to the lowest temperature which still permits flow)
were as follows (Table 14):
TABLE 14. POUR POINT VALUES FOR SKIMMED OILS
Run #
Pour Point
502-3A
-2.7°C
509-6
-2.7°C
509-7A
-2.5°C
Flash Point
Corrected flash points of extracted oils according to ASTM D-97 were as
follows (Table 15):
TABLE 15. FLASH POINT VALUES
Run #
Flash Point, °C.
208
182.3
305
179.6
306
168.4
313
174.1
38
-------
CHAR CHARACTERIZATION
Yield, ash content, major element, and heavy metal analysis were
determined for the product chers. One additional datum was obtained for each
char - heat content - in addition to the above, to allow the calculation of
approximate it balances. The char burn test was carried out on the smaller
scale altern. ive of the two alternatives outlined in the original proposal as
there was insufficient material for a full-scale test. The following table
(Table 16) provides the data obtained from char organic element, ash content,
and heat value analysis. Table 17 gives data on the major inorganic elements
present in chars. Additional data on minor inorganic components is presented
in Appendix B3.
Char produced during these experiments yielded few surprises, in contrast
to the results with some oil products. Liquefaction converted sludge with a
heating value of about 3000 cal/g to a char with heating values ranging from
1100-3100 cal/g. Ash contents ranged from 43 to 80%. Yields of char were
between 14-36% This variation is probably due to incomplete extraction of the
oil from the char. In fact, the distinction between char and oil may be more
of a continuum than a sharp difference: In experiments where oil and char did
not separate from the water phase, this is believed to be due to oil absorp-
tion onto char, the average density of the combination being near to one.
Char Burn Test
The char sample for this test was obtained from ca. 1 1 of combined pro-
duct from runs #208, 305, 306, 313, following a 72 hr methanol extraction in a
Soxhlet apparatus. The residue after this exhaustive extraction was designa-
ted as 'char'. It was black, and had a slight but offensive odor reminiscent
of burned rubber. It was inert to a lighted match at room temperature, but
ignited readily when heated over a gas flame. The flame was yellow and smoky,
and gradually died out on removing the heat source. Re-application of the
flame led to repeat ignition. This was repeated several times. The char be-
came plastic when hot, and very tarry. Removal of the sample to a copper
gauze (to allow air flow through the sample) and heating led to re-ignition,
and smoldering combustion when all of the volatiles had been consumed. The
bulk of the sample remained black after burning had ceased. The residue was
then transferred to an oven and ashed at 750°C for 4 hr., leaving a light-
colored ash. The composition of this ash is shown in Table 17.
Numerical results: Original char weight = 27.62 g.
Weight after burning = 16.84 g.
Weight after ashing = 10.02 g.
Hence, ash content of combined char was 36.28%, or, by difference, 63.72%
organic material, of which 61% burned readily, and the remainder burned at
750°C. An analysis of the inorganic components of char ash is shown in Table
18.
39
-------
TABLE 16. CHAR COMPOSITION
C H N 0 Ash Heat Value
Run # % % % % % cal/g
208
30.88
3.82
1.26
9.74
58.6
2416
31.94
3.57
1.57
9.99
59.4
2394
305
13.16
1.67
1.05
11.58
79.6
1178
12.02
1.46
-0-
12.08
78.5
1194
306
13.51
1.64
-0-
16.36
78.7
1779
12.69
1.55
-0-
16.45
77.5
1758
313
32.76
4.22
2.17
24.16
55.2
3150
29.98
3.71
0.80
22.99
58.5
3185
418
26.62
3.67
0.17
22.42
50.20
2815
27.49
3.66
0.33
21.08
50.11
2847
502-1
19.96
2.84
-0-
27.09
48.79
1817
17.12
3.33
-0-
25.49
48.25
1789
502-2
16.78
3.17
-0-
27.28
44.98
1423
16.10
3.48
-0-
26.41
45.14
—
502-3A
15.11
3.18
-0-
26.06
52.39
1462
15.33
3.06
0.13
26.15
51.52
—
507-4
21.04
3.47
-0-
26.96
51.24
2026
19.23
3.26
0.16
27.73
51.56
2058
507-5
21.48
3.55
0.03
24.75
55.65
2348
22.04
3.59
0.21
22.46
55.99
2435
507-6A
29.05
4.30
-0-
22.69
47.85
3121
27.98
4.22
-0-
23.31
47.59
3154
509-6
16.52
2.88
0.08
24.03
55.88
1702
17.15
2.96
-0-
21.54
55.21
—
509-7B
24.47
0.36
0.86
22.35
42.60
2599
24.65
3.56
0.80
22.91
44 44
—
Average 21.35
3.09
0.37
21.70
55.59
2148
+ -0- value was a measured resu
Single heat values determined
sample.
$ There was insufficient sample
run 509-7A.
t.
in some cases due to lack of
for analysis of product from
40
-------
TABLE 17. CHAR ASH ANALYSIS
Source Element 7, + Source Element PPM +
A1
1.80
0.19
Ag
U
23.0
2.7
Si
10.25
0.54
Sr
311.
22.
P
2.63
0.15
Y
10.9
1.3
S
0.112
0.064
Zr
1780.
120.
CI
1.091
0.058
Nb
5.18
0.94
K
0.649
0.034
Mo
245.
17.
Ca
9.24
0.46
Th
13.0
2.6
PPM +
V
<54.
Am
Ru
<5.0
Cr
941.
53.
Rh
<4.7
Mn
422.
27.
Pd
<4.0
Fe
4.46
0.22
Ag
97.6
7.4
Ni
598.
31.
Cd
<5.7
Cu
812.
91.
In
<6.0
Zn
7970.
400.
Sn
263.
19.
Gd
14.9
2.5
Sb
14.0
4.0
Se
6.25
0.88
Te
<9.3
Pb
270.
14.
I
13.7
5.7
As
<4. 5
Cs
<11.
Br
13.2
1.0
Ba
1430.
100.
Rb
17.2
1.2
La
38.4
9.6
Ce
<18.
41
-------
TABLE 18. MAJOR INORGANIC ELEMENTS IN CHAR
R^ri 208 305 306 313 "418
Element (%)
+/-
+/-
+/-
+/-
+/-
AT
2.60
0.18
2.93
0.21
3.02
0.27
1.35
0.13
0.59
0.095
Si
11.18
0.57
14.63
0.75
16.67
0.86
9.70
0.50
3.09
0.17
P
3.13
0.17
3.85
0.21
3.96
0.22
2.26
0.12
0.99
0.061
S
0.909
0.051
0.963
0.055
0.838
0.054
1.057
0.057
0.526
0.031
CI
0.069
0.000
0.059
0.009
0.040
0.011
0.084
0.008
0.135
0.009
K
0.307
0.016
0.520
0.027
0.561
0.030
0.309
0.016
0.222
0.012
Ca
7.78
0.39
11.47
0.57
11.52
0.58
7.81
0.39
2.77
0.14
Ti
0.526
0.027
0.544
0.028
0.381
0.021
0.331
0.017
0.151
0.008
Fe
3.24
0.16
5.58
0.28
4.64
0.23
3.34
0.17
1.14
0.057
-------
Throughout this test, the char combustion was reminiscent of the behavior
of powdered coal, with the exception of the tarring. We concluded that the
char would be an acceptable fuel source, although problems could develop
because of the high ash content and relatively low heat value, and the high
temperature required for complete combustion. Further study of char burning
needs to be done before this product can be used as a fuel source for the
reactor.
The combustion test with a char sample showed that it will burn, an', so
can be used as a source of fuel for the liquefaction reactor. Heavy metals in
the sludge were concentrated in the char, as expected. The oxygen content was
quite high (ca. 45%), which explains the relatively low heating value compared
to oil. In practice, since the char and oil would be used to provide fuel in
a water treatment plant, to run the reactor and dewatering equipment, they
would probably be recombined after separation from the wastewater. A discus-
sion of the relative yields of char and oil as a function of reaction condi-
tions is therefore unnecessary.
PRODUCT YIELDS AND PRODUCTION CONDITIONS
Weight yields of oils and chars extracted from the dichloromethane-
extracted analytical samples prepared as described previously are given in
Table 19.
TABLE 19: PRODUCT YIELDS AND PRODUCTION CONDITIONS
Run #
Conditions*:
% Feed
Weight (
in g) of
r: Yield (in
%) of:
(°C)
(mi n)
(psi)
sol ids
Sample
Char
Oil
Char
Oil
208
305
90
1700
20.7
233.85
11.45
4.95
23.65
10.23
305
275
180
1700
23.6
215.83
5.75
10.28
11.29
20.18
306
305
180
1750
18.8
190.53
5.40
8.92
15.08
24.90
313
275
90
1750
19.8
221.63
5.27
7.42
12.01
16.91
418
275
90
2000
20.2
237.46
11.59
8.46
24.14
17.62
502-1
275
105
2150
20.2
232.46
2.67
4.62
5.68
9.83
502-2
275
114
2150
20.2
236.61
2.00
7.51
4.18
15.71
502-3A
275
200
2150
20.6
260.56
3.17
4.78
5.91
8.91
507-4
275
165
1650
20.7
250.66
2.41
8.99
4.65
17.35
507-5
275
152
1250
20.7
211.92
12.69
8.60
28.97
19.63
507-6A
275
261
2150
20.7
241.06
8.08
18.09
16.22
36.30
509-6
275
252
2050
19.4
175.84
1.64
4.78
4.81
14.01
509-7A
275
57
2050
19.4
190.34
—
6.25
—
16.90
509-7B
275
60
2060
19.4
230.61
1.66
3.18
3.70
7.10
* Nominal conditions for temperature, pressure and residence time; actual
values vary somewhat - see reaction profiles (Appendices A1-A8).
+ Yields calculated from feed solids content, on a dry weight basis.
43
-------
DATA REDUCTION
Results obtained by analysis of oil, char, wastewater, and gas products
have been discussed in the context of the reactor conditions used for produc-
tion, and the overall goals of the project. In the following section, data
obtained by various methods are correlated, reduced, and compared to give a
presentation of the efficiency of continuous sludge liquefaction. This will
serve as a basis for discussion of the results and for the economic analysis
which follows.
The major numerical items of interest for the following discussion are
product yields in terms of mass and energy, to answer the question whether
continuous liquefaction works, and if so, if it converts sludge efficiently.
In Table 20, empirical data presented elsewhere in this report are assembled
and mass and energy yield calculations performed. Bearing in mind that the
theoretical oil and char mass yield from sewage sludge is between 60-70% of
the starting material weight (due to loss of water and carbon dioxide from the
feed during liquefaction chemistry), actual yields varied between 18-70% based
on organic material. At the same time, energy yields varied between 22-79%
(ignoring one result over 100%). The process did in fact work, based on these
results, and under certain conditions did so efficiently.
The simple arithmetic given in Table 20 represents the major portion of
data reduction for this project. The following example calculation of mass and
energy yields of oil and char from Run 306 is provided to facilitate interpre-
tation of Table 19 and Table 20:
Sample Yield Calculation: (sample containing Na2C03)
Feed sample weight = 190.53 g; Solids content = 18.84%;
Original weight of feed in sample = 190.53 x 18.84/100
= 35.90 g;
Weight of char extracted = 5.40 g
Weight of oil extracted = 8.92 g
Hence, weight yield of char = (5.40 x 100)/35.90 = 15.04%
and, weight yield of oil = (8.92 x 100)/35.90 = 24.85%
(includes ash);
Ash content of dry feed = 35.67%, so organic content = 64.33%
Weight of organics in dry feed = 35.90 x 0.6433, = 23.09 g
Ash content of char = 78.10%, of oil = 6.3%. By a similar
calculation,
Ash-free weight of char = 1.18 g; of oil = 8.36 g
Hence, weight yield of char = (1.18 x 100)/23.09, = 5.11%
weight yield of oil = (8.36 x 100)/23.09, = 36.21%
And, total oil + char organic yield = 41.32%;
44
-------
Heat content of feed = Weight of feed x unit heat value (g)
= 35.90 g x 3290 cal/g
= 118,111 cal
Similarly, heat content of char = 5.40 g x 1768 cal/g
= 9,547 cal
heat content of oil = 8.92 g x 8585 cal/g
= 76,578 cal
Therefore, energy yield in char = (9,547 x 100)/118,111, = 8.08%
energy yield in oil = (76,578 x 100)/118,111 = 64.84%
Total energy yield in oil and char = 8.08 + 64.84 = 72.92%
A total of 8 continuous runs was performed during this project, including
5 full-length runs and 3 runs where conditions were changed during the run.
In our original concept, steady state conditions would have been achieved
during each run, whereupon a sample of product would be taken. For the set of
conditions initially proposed, there would have been 9 runs (3 runs at 250,
300, and 350°C and at 3 residence times of 0.5, 1.5, and 3 hr at each temper-
ature) based on a residence time definition of one-third of the total reactor
volume. The objective was to find a set of conditions where above a 60%
energy recovery in product oil and char was achieved. In practice this was
achieved with run #306 and again with run 418, and was closely approached
on run 305 (58% energy recovery). Rather than trying to improve yields in
subsequent runs, we decided to concentrate on improving product separation, as
this had been poor or nonexistent in the first experiments, and good product
separation from wastewater was considered vital to the future development of
the process. The remaining runs were therefore designed to test the effect of
pressure on product separation rather than to maximize yield. Most of the
work was performed within a fairly narrow temperature range of 275-305°C, with
residence times in the whole reactor volume of 1.5 to 3 hr. Longer residence
times would have been practicable, although shorter times would have exceeded
the capacity of the metering pump and the cooling system.
The first result of note was that below 250°C there was no noticeable
reaction of the sludge; a preliminary run at 250°C yielded no oil although the
solids had been digested in the alkaline medium. Above 265°C a rapid endo-
thermic reaction set in, temporarily exceeding the capacity of the heaters.
At 275°C, where all but two of the runs were performed, feed conversion to
products was total. Beyond this point a discussion of relative efficiencies of
conversion at 275 or 300°C and 1.5 or 3 hr residence time is of questionable
value, as each yield is based upon a single experiment. In the last 3 runs,
steady state conditions probably were not achieved because of mixing inside
the reactor. From the data obtained, it made little difference to product
yields whether the reaction was run at a low or high temperature or for 1.5 or
3 hr. This indicates that the conversion of sludge to oil occurred in less
than the lowest residence time (i.e., in less than 1.5 hr).
According to results obtained using autoclaves, a 3 hr residence time in
the reactor was considered to be the minimum to obtain liquefaction and oil/
char formation, at 320°C. The reactor was designed around these parameters.
45
-------
It was quickly learned that complete reaction occurred at the lowest practical
residence time and within a few degrees of the temperature where the reaction
started (265°C.) This is an excellent result in terms of design and process
economics, because of the facts that equipment can be run at lower tempera-
tures, pressures, and residence times than previously thought possible, with
corresponding decreases in the power, thickness, and size of the equipment.
It is important to distinguish between the processes of sludge disso-
lution to soluble materials, oil and char formation through continued reaction
of water-soluble materials, and product separation from water after reaction
and cooling. From examination of the temperature/pressure profiles during a
run, it was apparent that within 10-15 min of the feedstock reaching 265°C,
the endotherm for the reaction was overcome by the heaters and temperatures
inside the reactor recommenced rising. This is believed to correspond to the
sludge dissolution and degradation phase, forming water-soluble, low molecular
weight products through hydrolytic reactions. After this, these products
recombined through Aldol reactions, condensations, cyclizations, and related
reactions to form oil (water-insoluble) products, and eventually char and high
molecular weight tars. (This process has been extensively investigated by us
using a range of feedstocks, in the laboratory). Following this, the products
exit from the reactor through the let-down process into the collection drum,
where spontaneous separation of water, char and oil phases occurs. In several
of the earlier runs, such separation either did not occur, or was very slow.
In considering the possible reasons for this, we thought of two possi-
bilities: Incomplete reaction, with a large proportion of residual water-
soluble compounds, or emulsification caused by too rapid mixing as a result of
gas bubble formation. Chemical analysis of the products eliminated the first
possibility. Raising the gas overpressure, it was felt, would relieve the
second problem by preventing local boiling, although it would not affect gas
formation through liquefaction reactions. In fact, an increased nitrogen
overpressure in runs 418 and 502-509 resulted in rapid product separation from
wastewater, spontaneously, within a day. The separation was good (the water
phase was transparent, although light brown), but further work needs to be
done on the actual combination of conditions leading to best separation as
well as to the best and most rapid liquefaction. These conditions may not be
identical. Also, since oils from the first runs were isolated through solvent
extraction, it was felt best to continue doing this to maintain continuity of
analysis. By visual observation and viscosity and pour point measurements it
was apparent that oils obtained in the high-overpressure runs by spontaneous
separation, decantation, and drying were superior to those obtained at lower
pressure, through extraction. Whether this is due to reaction conditions or
continued reaction of the products during the dewatering/solvent extraction is
not clear. We have previously noted a rapid viscosity increase in oils heated
above 100°C, which is apparently not due to water loss, and which could ex-
plain the high viscosity of the extracted oils. Alternatively, retention of
water in the decanted products could also explain their relatively low
viscosity. Further work with decanted oils would resolve this question.
46
-------
TABLE 20. MASS/ENERGY YIELDS OF OIL AND CHAR PRODUCTS
RuiTT: 258 355 555 JT3 ITS 55?T 55F? 55T-T*—WPl 557^5 S61-6K 509-6 50TT*—509^
Simple wt.
233.85
215.83
190.53
221.63
237.46
232.46
236.61
260.56
250.66
211.92
241.06
175.84
190.34
230.61
wt. char
11.45
5.75
5.40
5.27
11.59
2.67
2.00
3.17
2.41
12.69
8.08
1.64
-0-
1.66
wt. oil
4.95
10.28
8.92
7.42
8.46
4.62
7.51
4.78
8.99
8.60
18.09
4.78
6.25
3.18
Z feed solids
20.73
23.58
18.84
19.78
20.22
20.21
20.21
20.59
20.18
20.18
20.67
19.40
19.40
19.40
wt. Feed solids
48.48
50.89
35.90
43.84
48.01
46.98
47.82
53.65
50.58
42.77
49.83
34.11
36.93
44. 74
Z Feed Organic*
65.26
66.07
64.33
65.37
65.92
62.26
62.26
61.33
62.81
62.81
62.81
65.19
65.19
65.19
Wt. Feed organic*
31.64
33.62
23.09
28.66
31.65
29.25
29.77
32.90
31.77
26.86
31.30
22.24
24.07
29. 17
Wt. yteld char**
23.62
11.30
15.04
12.02
24.14
5.68
4.18
5.91
4.76
29.67
16.22
4.81
-0-
3.71
Wt. yield oil**
10.21
20.20
24.85
16.93
17.62
9.83
15. 70
8.91
17.77
20.11
36.30
14.01
16.92
7.11
Ash content char
(Z)
59.00
79.05
78.10
- 56.85
50.16
48.52
45.06
51.96
51.40
55.82
47.72
55.55
—
43.52
Ash content oil
(Z)
7.4
S.7
6.3
3.4
0.47
NO
5.6
3.6
2.1
15.4
3.6
16.3
1.7
NO
Wt. char organic*
4.69
1.20
1.18
2.27
5. 78
1.37
1.10
1.52
1.20
5.61
3.93
0. 73
-0-
0.94
Wt. oil orgamcs
4.58
9.69
8.36
7.17
8.42
—
7.09
4.61
8.80
7.28
17.44
4.00
6.14
»•**
Orjjajic char yield
14.82
3.57
5.11
7.92
18.26
4.68
3.69
4.62
3.78
20.89
12.56
3.28
-0-
3.22
Organic oil yield
(Z)
14.48
28.82
36.21
25.02
26.60
—
23.82
14.01
27.70
27.10
55.72
17.99
25.51
(H
Combined organic
yield
29.30
32.39
41.32
32.94
44.86
—
27.51
18.63
31.48
47.99
68.28
21.27
25.51
M.
Heat content feed
3189
2952
3290
3325
3256
2938
2938
2869
3109
3109
2946
3180
3180
3180
Heat content char
2405
1186
1768
3168
2833
1803
1423
1462
2042
2392
3138
1702
2599
Heat content oil
8347
7802
8585
8692
8812
8624
8542
8970
8176
8686
8170
767R
8 SOU
8489
Energy yield char
17.81
4.54
8.08
11.45
20.93
3.49
2.03
3.01
3.13
22.83
17.27
2.57
-0-
3.03
Energy yield oil
26.73
53.39
64.84
44.24
47.69
28.87
45.66
27.86
46.74
56.18
100.7
33.84
45.21
18.97
Total energy
yield
44.54
57.93
72.92
55.69
68.62
32.36
47.69
30.87
49.87
79.01
117.97
36.41
45.24
22.00
* Assumes H-0 constant.
Includes Ish
Ash-free
Excludes orgenics In gas and wastewater.
-------
LIQUEFACTION CHEMISTRY
Table 21 shows overall average feed and product compositions taken from
the experimental data together with ash contents both experimentally deter-
mined and by difference. The results are internally consistent, within exper-
imental error limits.
TABLE 21. FEED/PRODUCT AVERAGE ELEMENTAL AND ASH CONTENTS
C
H
N
0
Ash,%
Material
%
%
%
%
By Expt.
By Oiff.
Sludge
31.51
4.68
1.24
27.05
35.52
35.52
Oil
72.95
10.43
2.31
10.97
6.46
3.34
4.38*
Char
21.35
3.09
0.37
21.70
55.57
53.49
* Value with two highest (anomalous) results not included.
Table 22 shows the same average values on an organic basis, after
correction for ash content:
TABLE 22. FEED/PRODUCT AVERAGE ORGANIC ELEMENT CONTENTS
Material
C
H
N
0
%
%
%
%
Sludge
48.87
7.26
1.92
41.95
Oil
75.47
10.79
2.39
11.35
Char
45.90
6.64
0.80
46.66
(Value of ash determined by difference used in calculation).
Elemental ratios determined by division by appropriate atomic weights are
shown in Table 23:
48
-------
TABLE 23. FEED/PRODUCT ORGANIC ELEMENT RATIOS
Material
C
H
N
0
Sludge
4.07
7.20
0.14
2.62
Oil
6.28
10.70
0.17
0.71
Char
3.82
6.59
0.057
2.92
Representing sludge values as unity, the ratios of major organic elements
in oil and char products to feedstock are obtained in Table 24.
TABLE 24. RATIOS OF C.H.N.O IN PRODUCTS (FEED AS UNITY)
Material C H N 0
Oil 1.54 1.49 1.21 0.27
Char 0.94 0.92 0.41 1.11
Differences in composition between the oil and char products are readily
apparent. Carbon, hydrogen and nitrogen contents are all increased in the oil
relative to the feed, while oxygen is significantly reduced (by a factor of
approximately 4). In the char, excepting nitrogen content, all of the other
elements are present in similar amounts to the values in the feed. This
raises the possibility that the char does in fact represent unreacted feed-
stock rather than polymerized oil.
Using the ratios calculated above, it is possible to represent the con-
version of sludge to products in the form of an equation:
Sludge ~Oil + Char (neglecting wastewater and gas)
Dividing by the lowest value in Table 24 (above) gives the elemental
ratios in numbers of unity or above (N = 0.057 becomes N= 1.0). Rounding off
to the nearest whole number gives the net chemical change as:
C71H126N2°46 ~ C110H188N3°13 + C67H115N051
49
-------
From Table 19, the average ratio of oil/char yields on an organic basis
is 1:3.54. Using this ratio and combining oil and char as one product gives
the modified equation:
C71H126N2°46
¦*~C101H172N2.5°21
Assuming a 100% recovery of carbon in oil and char, the theoretical
yield of oil and char from sludge becomes
which calculates out to a value of 71%. Actually, of course, most of the
oxygen is lost as carbon dioxide, and so carbon cannot be assumed constant.
If the above equation is modified to reflect loss of water and carbon dioxide
as follows:
which represents the actual experimentally determined change quite accurately,
then the theoretical yield becomes 1523.2 ('molecular weight' of product)/
2475.5 ('molecular weight' of feedstock), or 61.5%. The highest combined
organic yield obtained experimentally (excluding run 507-6A, which may be
anomalous) was 48% from run 507-5. A conversion of sludge to recoverable
products of 48/61.5, or 78% of theoretical was obtained, therefore, the
remainder being lost in the form of aqueous organic soluble compounds.
Changes in Inorganic Element Distribution During Liquefaction
Because the liquefaction process is an alkaline digestion, it has the
potential for extensive degradation of toxic organic materials which may be
present in the feed; preliminary laboratory work has demonstrated this for
chloroform, lindane, and a 2,4,5-T analog (2,4,5-trichlorophenylpropionic
acid). Also, hydroxides of most heavy metals are insoluble, and so should be
precipitated from solution in the water portion of the feed during the lique-
faction, providing a method for cleaning the wastewater of high levels of
some toxic ions. This principle was borne out in practice, as shown by the
data in Table 25, where selected inorganic element concentrations from 3 runs
are compared between sludge feed, char and oil products.
C101H179N3°65
*-C101H172N3021
C101H179N3°65
^ ^*81^171^3^21 ^ CO2 + 4H2O,
50
-------
TABLE 25. SELECTED INORGANIC ELEMENTS IN FEED AND PRODUCTS
Material: Sludge Char Oil
Run #: 306 418 502-1 306 418 306 418 502-1
Element:
Cr
541
103.1
97.6
1107
220
19.5
13.9
11.7
Ni
20.1
27.9
26.9
308
73.1
9.3
9.6
8.0
Cu
309
339
302
2100
785
15.3
6.6
6.3
Zn
616
651
662
2500
1347
43.6
59.0
17.3
Pb
58.8
53.7
52.4
322
157
2.0
0.9
2.1
Ag
29.6
28.7
18.1
165
70.2
3.5
4.2
2.2
Cd
3.2
5.0
2.8
15.2
7.0
3.6
4.7
2.3
Sr
66.9
78.5
52.7
229
130
9.0
37.3
27.2
Sb
4.8
6.9
4.3
11.2
11.0
5.3
5.5
3.1
Te
5.1
8.3
4.8
4.4
14.0
5.7
6.9
3.3
Ba
369
319
309
1900
520
14.1
13.1
5.7
* Typical data taken from Run 306, 418, and 502-1 results.
+ X-ray analysis not done on char from run 502-1.
As expected, concentration effects depend to some extent on the solu-
bility product for the element hydroxide in question. Chromium was concen-
trated about twofold in the char, zinc 2-4 fold, and lead 3-6 fold. Cadmium
levels were too low in the feed sludge to permit any estimate of concentration
effects in char, and mercury was subject to interference with germanium in the
X-ray analysis and so was not reported. However, the trend is clear, and
represents a useful bonus of the liquefaction process in removing heavy metals
from water passing through the process. Assuming that most of the metal
content in the oils is due to particles of ash and char carried through the
extraction, it would be worthwhile to perform further research on liquefaction
of sludges heavily contaminated with toxic heavy metals. By "spiking" a
sludge with a known amount of a specific metal ion, such as cadmium, we could
determine the resulting distribution of cadmium in the products. The experi-
ment would also provide much needed data on residence times and distributions
inside the reactor.
GAS ANALYSIS
Table 26 lists gas analysis results obtained. The absence of hydrogen in
gas samples obtained after 3-06-84 is puzzling, and we have no explanation for
it. There was no known change of procedure which could account for this
anomaly.
51
-------
TABLE 26. REACTOR OFF-GAS COMPOSITION
Gas:
Hydrogen Carbon Ethylene Ethane Oxygen Nitrogen Methane
Run #: dioxide
1122*
1.58
(13.63)+
1122
0.98
(8.98)
208
4.75
(10.45)
305
1.38
(19.94)
306
1.40
(17.03)
313
ND
(-0-)
418
ND
(-0-)
502
ND
(-0-)
507
ND
(-0-)
509
ND
(-0-)
9.50 0.15
(81.97) (1.29)
9.49 0.12
(86.98) (1.10)
39.4 0.24
(86.74) (0.53)
5.19 0.16
(75.05) (2.30)
6.52 0.13
(79.62) (1.53)
3.63 0.05
(98.26) (1.33)
7.41 0.03
(98.40) (0.33)
1.95 0.04
(98.40) (0.33)
3.83 0.05
(97.50) (1.15)
1.83 0.10
(90.40) (4.70)
0.24
0.69
(2.07)
-
0.20
0.07
(1.83)
-
0.22
0.06
(0.48)
-
0.15
2.05
(2.23)
-
0.13
1.66
(1.59)
-
ND
1.26
(-0-)
-
ND
0.03
(-0-)
-
ND
0.94
(-0-)
-
ND
2.05
(-0-)
-
0.10
0.07
(4.90)
-
87.72
0.12
-
(1.04)
89.03
0.12
-
(1.10)
54.52
0.80
-
(1.76)
91.02
0.05
-
(0.69)
90.13
0.05
-
(0.56)
95.05
0.03
-
(0.68)
91.68
0.02
-
(0.31)
96.05
0.05
-
(2.20)
94.79
ND
-
(-0-)
96.34
ND
-
(-0-)
* Richland primary sewage sludge + peat moss
+ Normalized values, minus oxygen and nitrogen from air and purge gas,
given in parenthesis. In run 208, carbon monoxide was also noted at
0.04% (0.08% normalized), and also in Run 418 (0.04%, 0.57% normalized).
ND = not detectable.
A typical gas production rate was back-calculated as a percentage of feed
carbon, according to the calculation shown for Run 306 below. Gas production
accounted for 13% of carbon input principally as carbon dioxide. Variation in
this was not significant enough to warrant a separate calculation of mass for
each sample. Instead, a fixed value of 13% was used in calculating balances.
Storage tank pressure at 9:00 = 39.3 psi (2.67 atm)
at 9:15 = 59.0 psi (4.01 atm)
Volume of storage tank = 504 1
Gas produced in 15 min = (4.01 - 2.67) x 504 1 at 17°C, = 675.36 1
Volume of gas produced per hr at STP = 2543 1
Gas composition = 90% nitrogen, 8% carbon dioxide, 2% hydrogen
Gas actually produced = 0.1 x 2543 1 (excluding nitrogen), = 254.3 1
Average molecular weight of carbon dioxide/nitrogen mixture = 35.60
52
-------
Gram-moles/hr of gas produced = 11.35,
= 404.2 g/hr
At 20% solids, 3 hr residence time, sludge solids input
= 46.5/5 x 1/3 kg/hr, = 3.10 kg/hr solids
Average carbon content of sludge solids = 34.50%
Average carbon content of gas = 33.74%
Carbon input in feed = 1069.5 g/hr
Carbon output in gas = 136.4 g/hr
Hence, % carbon in gas as a function of feed = 12.75%
The major component of the by-product gas was carbon dioxide, excluding
the nitrogen used for pressure regulation. In practice, the gas produced in
liquefaction would be used for pressure control. Some hydrogen was also
generated in the early stages of a run, but in quantities insufficient to be
significant. Entrapped oil and mineral particles and volatile organic com-
pounds were also collected in the 'dogleg' portion of the let-down system
and the gas storage tank, and would need to be collected and recycled to the
product storage vessel. This would also be essential for environmental
reasons, due to the obnoxious odor of the gas. After this was done, the gas
could be vented to atmosphere without adverse effect on the environment.
WASTEWATER CHARACTERIZATION
Analysis of the wastewater from the liquefaction runs was to include BOD,
COD, suspended solids, evaporation residue, pH.and dichloromethane extraction
on selected samples, rather than on samples from every run. Techniques of
analysis have already been described in the Materials and Methods Section.
Results of these analyses are shown below:
TABLE 27. WASTEWATER ANALYSIS
Run# Conditions pH: Solvent Suspended Evaporation
(°C/hr.) Crude Filtered Extracted soluble solids residue
305 275/3 8.10 8.18 8.60 1.7* 0.19 —+
8.11 8.12 8.62 1.9 0.20 87.6
313 275/1.5 8.40 8.40 8.70 1.7 0.20 81.3
8.39 8.40 8.72 1.7 0.22 77.5
502-2 275/1.5 8.40 8.36 8.62 1.3 0.11 74.8
8.39 8.40 8.73 1.4 0.05 85.1
67.1
* Grams/1 iter
+ Sample bumped and was lost.
53
-------
Results of COD and BOD analyses are shown in Table 28. Biodegradabi 1 ity of
organic material in wastewater was determined as the ratio of BOD:COD
expressed as a percentage (Table 29).
TABLE 28. COD AND BOD OF WASTEWATER
Run
#
Trial #
COD
BOD
1
2
l
2 3
4
5
6
305
50.6
54.6
38.56
40.08 38.64
40.92
40.68
42.12
313
61.2
62.4
39.44
38.32 37.28
41.04
34.44
36.72
502-
-2
39.8
41.9
29.60
28.24 27.44
26.64
26.5Z
27.96
Gl-
-Gd
243.5 260.
,0 260.
,5
Values in g/1 except for the glucose/glutamine standard (Gl-Ga),
which is in mg/1.
Aqueous evaporation residue (dissolved solids), and ash contents were
determined for runs 418 and 502-509 to enable us to better determine mineral
balance and overall mass distribution during liquefaction. These results are
reported in Table 30.
TABLE 29. WASTEWATER BIODEGRADABILITY
Run # Run conditions Av. COD Av. BOD S.D.* % Biodeg.+
305 275/3 hr. 52.6 40.17 2.02 76.4%
313 275/1.5 hr. 61.8 37.87 2.09 61.3
502-2 275/1.5 hr. 40.8 27.73 1.04 68.0
* S.D. = standard deviation (+/-)
+ Percent biodegradabi1ity = BOD x 100/C0D
54
-------
TABLE 30. DISSOLVED SOLIDS ASH CONTENT
Run # Solids Ash, % Organics, %
(g/1) 1 2 Average (average)
418
77.8
57.42
57.25
57.34
42.64
502-1
79.7
61.95
62.40
62.18
37.82
502-2
67.1
66.96
67.12
67.04
32.96
502-3A
65.5
64.69
65.35
65.02
34.98
507-4
77.0
59.03
59.45
59.24
40.76
507-5
87.3
60.12
60.31
60.22
39.78
507-6A
71.6
57.87
58.09
57.98
42.02
509-6
75.1
62.36
—
—
37.64
509-7B
73.7
53.79
-
-
46.21
There was insufficient sample for an ash or residue determination on sample
7A. Organic content is by difference.
Anaerobic Digestion
The results of this test were negative. One of the two controls evolved
3.7 ml of gas while the others did not produce any, even after 8 weeks at
37°C. Adaptation of organisms to the largely phenolic and non-acidogenic
organic compounds present in these wastewaters would probably be required to
obtain a positive result.
Wastewater Treatability
At the beginning of this project, one of the main concerns was the ques-
tion of wastewater treatability. This concern has been dispelled by the
results obtained. Although COD was very high, the aerobic oxidation rate by a
standard bacterial culture was also very high, corresponding approximately to
that of glucose. Anaerobic digestion results were less encouraging, which is
to be expected in the light of the organic water-soluble products from lique-
faction: These are mostly phenols and low-molecular weight alicyclic and
aliphatic compounds, more suitable to attack by aerobic than anaerobic orga-
nisms. From the results, we conclude that there should be no problem in
recycling the wastewater from a liquefaction unit back through with the pri-
mary sludge input into the plant. The additional loading will be negligible.
The recycle liquor from the STOR process will result in additional
indirect costs for the sewage treatment plant. These costs are included in
Table 35 in the economics section of the report.
55
-------
SECTION 6
LIQUEFACTION PROCESS PRELIMINARY ECONOMICS
INTRODUCTION
A preliminary economic evaluation of the sludge-to-oil process is devel-
oped for small towns of 10,000 people, medium cities of 100,000 people and
large metropolitan areas of 1,000,000 people. The first part of this chapter
describes the basis for the cost estimate. The second part describes the
material, component and labor costs. The third part examines alternate sludge
disposal costs and compares them to the estimated costs of the sludge-to-oil
system.
ESTIMATE BASIS
The bases of the analysis are shown in Table 31. Sludge production is
generally on the order of 0.07 to 0.11 kilograms per capita per day. A sludge
production value of 0.08 kilograms per capita per day (dry basis) is used to
estimate the sludge production. It is assumed that the baseline sludge con-
sists of 65% organic material and 35Z inorganic material. The wet sludge,
when mixed with 5Z Na^COj, would be 25Z solids. It would have a density of
ca. 1.1 and occupy a volume of 3.82, 38.2 and 382 m . The sludge input to
each size sludge-to-oil system would therefore be 76.2Z water, 12.4Z organics
and 11.4Z inorganics. At this point, the similarities between the three sizes
ends. Due to the differences in size, the basic design of each plant is quite
different.
Design Basis for a Town of 10,000
Heating 4.2 tonnes per day of the 25Z solids reactor feed to 300°C
requires 1.26 x 10 calories or 61 Kw (210,000 Btu/hour). In one year this
plant would use 535,000 KwH (1.8 x 10 Btu). If the energy were supplied by
electricity at a cost of $0.05 per KwH this amounts to $27,000 per year. If
the energy were supplied by natural gas at $4.5 per 10" Btu, the annual cost
would be $8,200. This demonstrates that heat recovery equipment would have to
be quite inexpensive to be justifiable in a plant this small. One possible
system would be to jacket the high temperature portion of the cooler and the
preheater before the reactor. This jacket would then be filled with water and
kept at a pressure where the water would boil in the cooler and condense in
the preheater. This would require a little extra piping and a good pump, but
would probably not add more than $10,000 to the cost. This heat recovery
system would likely recover 50Z of the heat input. Based on this heat
56
-------
TABLE 31. ESTIMATE BASES FOR THREE STORS SIZES
Small Town City Metropolis
Design Item Pop. 10,000 Pop. 100,000 Pop. 1,000,000
Sludge Production* 0.8 8.0 80
(Dry basis)
Sludge Organics 0.5 5.0 50
(Dry basis)
Wet Sludge 4.0 40 400
(20Z Solids)
Sodium Carbonate Added 0.2 2.0 20
Reactor Feed 4.2 42 420
Oil Product 0.2 2.0 20
Char 0.12 1.2 12
Process Wastewater 3.2 32 320
Solids in Gas and Wastewater 0.68 6.8 68
* All units are tonnes/day.
57
-------
recovery scheme we assume that a net temperature rise of 150°C would be
supplied by the primary heater. For a small facility such as this, the
product oil could probably not be used to fuel the process. Thus, the product
would have to be incinerated on the site or transported away. This adds to
the overall cost and does not allow the energy value of the oil to be
considered a revenue.
Material and Component Costs for a Small Town
The material and component costs assume a 30 min residence time in the
reactor high temperature zone. The volumetric flow rate of the sludge is
about 3.82 m /d or 0.002652L nr/min. In 30 min, this is equal to an
accumulated volume of 0.0796 m . The reactor pipe size used for the process
development unit, 6" schedule XXS pipe, has a volume of 0.012152 m /m of
length. If this pipe were used for an 0.8 dry tonne per day facility, 6.55 m
of pipe would be required for the high temperature zone. When this is added
to an equal amount of pipe for the preheating zone of the reactor, the length
of pipe needed for the whole reactor amounts to 13.1 m. This could best be
assembled into two 6.55 m long -segments. The pipe alone would weigh 1.05
tonnes. At a rough cost of r7/kg for chrome-moly steel, this amounts to
$7,350 for the pipe.
Design Basis for a Medium City of 100,000
In a medium sized city the energy flows are large enough to warrant the
use of heat recovery. A fairly detailed and proprietary design has been
developed for a city of this size. In this design, a Dowtherm boiler is fired
with the oil produced from the sludge. Vaporized Dowtherm is used as the heat
transfer fluid to the high temperature portion of the reactor. Boiling and
condensing water in the jacketed cooling and preheating sections would be used
to recover heat. About 50% of the heat could be recovered in this manner.
This size plant is too small to consider a high pressure steam boiler as the
main heat transfer system.
Materials and Component Costs for a Medium City
The material and component costs assume a 30 min residence time in the
reactor higk temperature zone. The volumetric flow rate of the sludge is
about 38.2 m /d or 0.02652 m /min. In 30 min, this is equal to an accumulated
volume of 0.7958 m . The reactor pipe size used for the process development
unit, 6" schedule XXS pipe, has a volume of 0.012152 m /m of length. If this
pipe were used for an 8 dry tonne per day facility, 65.5 m of pipe would be
required for the high temperature zone. When this is added to an equal amount
of pipe for the preheating zone of the reactor, the length of pipe needed for
the whole reactor amounts to 131 m. This could best be assembled into twelve
11 m long segments. The pipe alone would weigh 10.5 tonnes. At a rough cost
of $7/kg, this amounts to $73,500 for the pipe.
If 61 cm (24" schedule 160) pipe were used, (0.189 m^/m) about 4.2 m of
pipe would be required for the high temperature zone. The entire reactor
would require only 1 segment about 9 m long. This would weigh about 7 tonnes
and cost about $49,000. The higher cost of the larger fittings would be
58
-------
offset by the smaller number required. In addition, the savings on the number
of seals and motors for the scraper system would be large. However, heat
transfer would be more difficult. Clearly, the selection of the reactor
diameter is an important consideration in the actual design phase. This cost
analysis uses the 61 cm (24") diameter pipe as the reactor vessel.
Another important consideration is the selection of the heat transfer
fluid. Plants this size and smaller will use heat transfer fluids such as
Dowtherm. However, large plants would benefit by using steam because of its
better heat transfer properties.
Design Basis for a Large Metropolitan Area of 1,000,000
The major design feature of a large liquefaction system would be the use
of a high temperature, high pressure stoker-fed steam boiler. This boiler
could co-fire coal and could be equipped with scrubbers and other pollution
control equipment. The basis design of the reactor itself can also be
changed. In the smaller systems using electric, gas or heat transfer fluid for
primary heating, the sludge was inside the pressure vessel and the heat has to
be transferred through the thick-walled vessel. With a high pressure steam
line, the steam and the sludge will both be inside the pressure vessel. This
means that the pipe through which the sludge flows can be thinner and made
from a high cost alloy such as Inconel. The sludge would be agitated inside
the alloy pipe, the steam would be outside the alloy pipe and the steam and
sludge would be enclosed by a large insulated carbon steel pressure vessel.
The carbon steel vessel would be very thick in order to withstand the
pressure. However, carbon steel is inexpensive and can readily be used in
steam service whereas it cannot be used directly with sludge. The condensing
steam on the alloy surface combined with a relatively thin alloy wall and
scraped surface agitation of the sludge inside the pipe will result in high
heat transfer rates. This heat transfer scheme will have heat transfer rates
at least an order of magnitude greater than smaller systems using Dowtherm
vapor and transferring heat through the pressure vessel. What this means is
that the sludge can be heated in a highly efficient heat transfer device and
then allowed to react in holding vessels. The bulk of the residence volume
would be provided by these holding vessels. The holding vessels could be made
of 24 inch nominal schedule 160 chrome-moly pipe or carbon steel pipe lined
with stainless steel. The holding vessels would be maintained at temperature
by steam trace heating on the exterior of the pipe. The holding vessels would
not have scrapers.
Materials and Component Cost for a Large Metropolitan Area
The heating system for this large plant would be different from the pre-
vious smaller plants. The preheater would be similar in design but the high
temperature heater would not be sized for the entire residence time. The high
temperature heater would contain about one tenth the total residence volume.
This large plant would be constructed from 24 inch or larger schedule 160
pipe. The heater, preheater and primary cooler would each be about 5 m long.
About 40 m of pipe would be required for the high temperature residence zone.
The entire reactor would require 7 segments; three 5 m long scraper surface
heat transfer vessels and 4 eight ml holding vessels. This combines to a
59
-------
total of 47 m of pipe arid weight of about 38 tonnes. The cost of the pipe
alone, assuming the use of carbon steel costing $1.0 kg, would be about
$38,000. This does not include the cost of lining the pipe with stainless
steel or the cost of the fittings. However, the use of carbon steel pipe with
a stainless steel liner made possible by the improved heat transfer
consideration, will result in lower purchase and fabrication costs. Unlike
chrome-moly steel, carbon steel pressure vessels do not require annealing
after being welded. This is an important consideration given the size and
weight of a 24 inch schedule 160 pipe. In addition, the improved heat
transfer system eliminates the need for multiple scraper systems. Scrapers
would be needed on the preheater and heater only and not on the residence
vessels.
COST ESTIMATE SUMMARY
Table 32 provides a simplified check list of items included in the esti-
mate of the fixed capital cost of the liquefaction process system for each
size facility. The first part of the material cost list in Table 32 describes
the costs incurred to obtain a building on-site to house the reactor system.
The rest of table outlines the costs of each of the major components of the
liquefaction system. The costs for the tanks were taken from reference 7. A
present dollar conversion factor of 2.76 was used to update the costs. The
cost of the reactor, a jacketed, high pressure, stirred kettle, was also taken
from reference 7. Other values are based on our experience in purchasing
components for the process development unit at Battelle-Northwest. The esti-
mates presented in this chapter are preliminary budget estimates.
Table 33 shows a thermal analysis of the overall process and estimates
the amount and value of excess oil for each size facility. The data used to
determine the weights of char and oil produced by the process are as follows:
The produced char and oil amounted to 15% and 25% of the weight of the dry
feed sludge. The average heating values of the char and oil are 1.770 and
8,590 cal/g respectively. These figures combine to give the annual energy
production. The energy required by the process was determined by assuming a
net 150°C heat input to the process and an overall furnace and heating system
efficiency of 65%. It was also assumed that all of the char would be burned
first and then the oil would be consumed. We have assumed that the energy
from the oil would be sold at a value of $4/10 Btu. It is noteworthy that
the energy recovered from this process is more than three times the
theoretical amount needed to heat the sludge a net of 150°C. This provides
considerable flexibility to the design of the furnace in that a high
efficiency unit may not be required.
The summary table of capital and operating costs is shown in Table 34.
This summary cost estimate is based on our best judgment at this time. The
final result is a process that completely disposes of sludge for ca. $45/dry
tonne for a large city. This amounts to roughly $ll/tonne of 25% solids sludge
and is almost competitive with ocean dumping and is less costly than inciner-
ation. Smaller plants are less economically feasible. (A more detailed com-
parison of liquefaction with other disposal methods will be highly site-
specific and would be made on a case-by-case basis.)
60
-------
TABLE 32. PURCHASED MATERIAL COST ESTIMATE
Town - 10.000 Persons
City - 100,000 Persons
Metropolis - 1,000,000 Persons
Size Or
Cost
Size Or
Cost
Size Or
Cost
Item
Capacity
Estimate
Capacity
Estimate
Capacity
Estimate
Land
Existing
.2 Ha
12.500
.5 Ha
100,000
Surveys
—
1.000
—
2.000
—
4,000
Fee and Permits
—
2.000
5,000
—
10,000
Site Preparation
4.000
6,000
20,000
Foundation
10mX15m
10mX15m
30m*30m
10 tonnes
15.000
40 tonnes
15,000
100 tonnes
75,000
Process Building
lOmXIOmXIOm
10mX15mX20m
20mX15mX20m
steel sheet sides
50,000
steel sheet sides
100,000
steel sheet sides
200.000
Crane
5 tonnes
12,000
10 tonnes
20,000
25 tonnes
40.000
Plumbing
water, nitrogen
water, nitrogen
water, nitrogen
compressed air
8,000
compressed air
10.000
compressed air
17,500
Ventilation
1000 cubic
3000 cubic
10,000 cubic
meters per hour
3,000
meters per hour
6,000
meters per hour
15.000
Fire Protection
System
per code
4.000
per code
8,000
per code
20.000
Lighting
indoor/outdoor
indoor/outdoor
indoor/outdoor
sodium vapor
4,000
sodium vapor
4.000
sodium vapor
8,000
Communications
public address
public address
public address
and warning
1,000
and warning
1,000
and warning
2.000
Sludge Preparation
Subsystem
15 1 iters/mi r.
7.500
150 liters/min
20.000
1500 liters/min
75,000
Sludge Injection
2000 psig
2000 psig
2000 psig
Subsystem
3 liters/mn
10.000
30 liters/min
30.000
300 liters/min
150,000
Reactor & Cooler
275-310°C
275-310'C
275-310°C
Subsystem
3 liters/min
200.000
30 liters/min
400,000
300 liters/min
2.200,000
Let Down Subsystem
3 liters/min
20,000
30 liters/min
50,000
300 liters/min
200.000
Product Recovery
5000 liters
50,000 liters
100 meters
Subsystem
3 liters/m1n
15,000
30 Hters/m1n
60,000
300 liters/min
200.000
Primary Heating
System
40 Meal/hour
30,000
400 Heal/hour
4.120,000
4.0 Gcal/hour
800.000
Ash Handling
200 kg/hr
1.000
2 tonnes/hr
7,000
10 tonne/hr
30,000
Process Integration
Piping
as needed
12,500
as needed
35,000
as needed
100,000
Instrumentation and
computerized with
computerized with
computerized with
Control
manual override
60,000
manual override
90.000
manual override
150,000
Electric Utility
Hookup
480 volt
10,000
480 volt
30,000
480 volt
50,000
Material Subtotal
470,000
1,031,500
4,466,500
61
-------
TABLE 33. ENERGY REQUIREMENTS AND OUTPUTS OF THE SLUDGE TO OIL FACILITIES
Town - 10,000
City - 100,000
Metropolis - 1,000,000
Operating
Element
Heat Value
Annual Energy
Kcal
Heat Value
Annual Energy
Kcal
Heat Value
Annual Energy
Kcal
Energy Input
of Sludge
3300
9.64E+08
3300
9.64E+09
3300
964E+10
Heat Used
In Process
3.54E+08
3.54E+09
3.54E+10
Energy
In Char
1770
7.76E+07
1770
7.76E+08
1770
7.76E+09
Energy
In Oil
8590
6.27E+08
8590
6.27E+09
8590
6.27E+10
Energy of
Excess Oil
6.27E+08
3.51E+09
3.51E+10
Amount of
Excess Oil
83 tonnes/yr
416 tonnes/yr
4160 tonnes/yr
Value Of
Excess Oil
$4.00/HMBtu
$55,672
S4.00/MMBtu
$556,719
62
-------
TABLE 34. COST ESTIMATE COMPARISON FOR THREE DIFFERENT SIZES OF FACILITIES
Town - 10.000
City -
100,000
Metropolis -
1.000,66(3
Capital
Percent of
Estimated
Percent of
Estimated
Percent of
Estimated
Cost Element
Capital Cost
Cost
Capital Cost
Cost
Capital Cost
Cost
STORS
Engineering Design
15
*100,000
13
$ 200.000
8
$ 500,000
Purchased Materials
68
$470,000
66
$1,031,500
69
$4,466,500
Construction Labor
11
$ 75,000
16
$ 250.000
15
$1,000,000
Construction Management
4
S 25,000
3
$ 40.000
2
$ 150,000
Start-Up Materials
1.02
$ 7,000
0.64
$ 10,000
3
$ 200,000
Indirect Capital Costs
1.46
$ 10,000
1.48
$ 23.000
3
$ 170.000
Capital Cost
100
$687,000
100
$1,554,500
100
$6,116,500
Operating
Percent of
Estimated
Percent of
Estimated
Percent of
Estimated
Cost Element
Operating Cost
Annual Cost
Operating Cost Annual Cost
Operating Cost
Annual Cost
STOSS
Oebt Service at 8".
20 years
39.9
$ 69,972
41
$158,329
35
$ 660.664
Maintenance Labor
and Supplies
14.2
$ 25,000
6
$ 25,000
5
$ 100,000
Operator Labor
22.8
$ 40,000
21
$ 80,000
13
$ 250,000
Fuel and Soda Ash
8.5
$ 15,000
26
$100,000
40
$ 750,000
Miscellaneous Supplies
14.2
$ 25,000
6
$ 25,000
5
$ 100,000
Indirect Maintenance
0.1
$ 200
0
$ 600
0
$ 3,600
Indirect Operiting Cost
0.2
$ 300
0
$ 1,000
0
$ 2,800
Annual Operating Cost
100
$175,472
100
$389,929
100
$1,867,064
Annual Operating Revenue
$ 0
$ 55,672
$ 556,719
Net Annual Cost
$175,472
$334,257
$1,310,345
Sludge Disposal Cost
S/T (Dry)
$ 601
$ 114
$ 45
INCREMENTAL ANNUAL
Estimated
Estimated
Estimated
TREATMENT PLANT COSTS
Annual Cost
Annual Cost
Annual Cost
Operating and Maintenance
$56,000
$130,000
$ 600,000
Amortized Capital
$37,000
$ 85,000
$ 400,000
Total Incremental Cost
$93,000
$215,000
$1,000,000
Incremental Sludge Disposal Cost S/T (Dry)
CO
en
w
74
34
63
-------
The indirect costs of the waste liquor recycle were derived from
reference 8 with some adjustment of the basis of calculation. Reference 8
shows a graph of incremental indirect O&M costs versus on gallons of recycle
liquor and on million gallons per day of capacity. The basis sludge in
reference 8 contained 4.5% solids whereas the basis sludge for the STORS
process contains 20% solids. The cost figures were determined by estimating
the gallons per minute of recycle liquor from the STORS process and adjusting
the flow rate for an estimated 9% downtime. The capital costs were estimated
by back calculating the total sewage flow into the treatment plant based on
the given figure of 1.1 tons of solids per million gallons per day of
wastewater. This total sewage flow figure was then reduced by a factor of
4.5%/20% to account for the difference in the water content in the sludge.
-------
SECTION 7
OVERALL ACHIEVEMENT OF PROJECT OBJECTIVES
This project was designed to evaluate a promising batch process for
sludge disposal on a continuous scale. Although some specific goals have not
been met, the overall result of the research was highly favorable: Liquefac-
tion of a municipal primary sludge was achieved on a continuous basis for over
100 hours total in a prototype reactor, without major problems in clogging,
char deposition, or corrosion. Process parameters were defined, and product
behavior and properties characterized. An overall yield of 73% of the dry
feed sludge energy content in the form of separable oil and char appears to
be readily achievable on a larger scale, with product oil having a heating
value of about 8800 cal/g. Wastewater was found to be readily treatable by
current practice, although heavily loaded with organic materials, and the
off-gas represents no environmental threat, being mostly carbon dioxide and of
low volume. A preliminary economic estimate based on realistic assumptions
and the experimental data reported here gives a sludge disposal cost of $45/
dry tonne for a city of one million people, with a capital cost of $6.1M. If
municipal garbage can also be fed to the system, economics become even more
favorable, and will permit application of the process to even smaller
communities.
To continue this work to the point where a pilot plant could be built,
additional research is needed. Some of the items which require further inves-
tigation are as follows:
* Optimum liquefaction parameters: Our initial estimates were too
pessimistic; complete conversion of sludge occurred under temperature, pres-
sure, and residence time conditions which were much milder than anticipated.
In particular, the effects of reduced residence times and alkali concentration
need further investigation.
* Material balances, both organic and inorganic. Mixing occurred to a
greater extent than anticipated, a problem which should not occur in the
modified reactor. There were indications that the liquefaction process may be
useful in destroying organic toxic materials and concentrating inorganic heavy
metals in the ash and char fraction and away from the wastewater.
* Product separation and stability. Initial product separation from
wastewater was not good, but this was corrected by modifying reactor condi-
tions, and good separation occurred rapidly and spontaneously in later runs.
The separation process requires further study to determine exact conditions.
65
-------
possibly leading to patentable information. Properties of separated oil, as
opposed to solvent-extracted oil, need to be examined. Limited results
obtained to date on separated oil indicate that it is much less viscous than
extracted oil. The variation of oil viscosity with storage time and the
effect of vacuum distillation to make a light (#2) product also need further
work.
* Addition of a comminution device should allow work on liquefaction of
other materials such as garbage. This would greatly extend the range of
application for the process.
* Longer duration runs need to be performed to obtain data on corrosion,
erosion, mineral settling, and char buildup. Although no problems were
encountered in 100 hr of operation of the reactor, this is insufficient to
identify long-term materials behavior problems.
Previous work by others (1) used a dried sludge which was liquefied in an
organic solvent with a nickel catalyst, or is an aqueous sludge slurry at
400°C, with hydrogen. The economic cost of either pre-drying or hydrogen
addition is prohibitive. Solvent recovery is also costly. The STORS process
uses only conventional dewatering of primary sludge to 20% solids, and a cheap
catalyst (sodium carbonate), with no hydrogenation, at a much lower tempera-
ture (275°C). It is consequently much more competitive economically. If a
higher quality oil is required, it is more cost-effective to upgrade the
STORS product than to apply hydrogenation to the whole system, including
water. Consequently we feel that the simpler the liquefaction process, the
more chance it has of commercialization.
In a society where sludge disposal is an ever-growing problem, the lique-
faction technique offers a flexible, safe, and economical alternative to
incineration and should be pursued further.
66
-------
REFERENCES
1. Kranick, Wilmer L., Conversion of Sewage Sludge to Oil by Hydroliquefac-
tion EPA 600/S2-84-010 USEPA, Cincinnati, Ohio, 1984, pp. 27, NTIS
PB84-133768.
2. Donovan, J. M., R. K. Miller, T. R. Batter, and R. P. Lottman. Physical
and Chemical Characteristics of Synthetic Asphalt Produced from Lique-
faction of Sewage Sludge. EPA-600/2-82-242, U. S. Environmental
Protection Agency, Cincinnati, Ohio, 1981, 57 pp., NTIS PB-82-119082.
3. McCabe, W. L., and J. C. Smith. 1967. Unit Operations of Chemical
Engineering, Second Edition, McGraw-Hill Book Company, New York, New York.
4. Sanders, R. W., K. B. Olsen, W. C. Weimer, and K. K. Nielson.
Multielement Analysis of Unweighed Oil Samples by X-ray Fluorescence
Spectrometry with Two Excitation Sources. Anal. Chem., 55(12):1911-1914,
1983.
5. Shelton, D. R., and J. M. Tiedje. General Method for Determining
Anaerobic Degradation Potential. Appl. Env. Microbiol., 47(4):850-857,
1984.
6. EPA. Sludge Treatment and Disposal; Process Design Manual, United States
Environmental Protection Agency, EPA 625/1—79—0117 1979.
7. Peters, M. S., and K. D. Timmerhaus, Plant Design and Economics for
Chemical Engineers. Second Edition, McGraw-Hill Book Company, New York,
New York.
8. Ewing, L. J., H. H. Alngren, and R. L. Culp. 1978. Effects of Thermal
Treatment of Sludge on Municipal Wastewater Treatment Costs.
EPA-600/2-78-073, Mun icipal Environmental Research Laboratory,
U.S. Environmental Protection Agency, Cincinnati, Ohio.
67
-------
500
RUN NUMBER—EPA20B
305C FOR 1.9KS
400
300
cr
oo
200
LONER THERNOVELL
UPPER THERNOVELL
100
1 1
. 1
.1.1
i 1
1 I
I I
I I
. 1
| |
8
CD
8
a
8 8
a
8
KJ
8
rfl
8
8
£
8
<£
m <
8
8
mm
8
dS
RUN T1ME-IIRS
Reaction Temperature and Pressure °rofile, Run 208
-------
vo
?sm
RUN NUHBER—EPA2O0
305C FOR l.SHRS
2000
- REACTOR HEAOSPACE
OO
0
&
0
1500
|
CD
~—«
i
1000
500
i 1 i 1 I 1 i 1 i 1 i 1
. 1 .
1
1 1 1 1 1
a
r2
o a 8 8 8 8
2 2 43 - * *
CD dl •-»
00*1
IS 00
8 8
uB tC
18.00
RUN TIME-MRS
Reaction Temperature and Pressure Profile, Run 208
-------
£488888888888
RUN T1ME-HRS
Reaction Temperature and Pressure Profile, Run 305
-------
500
RUN NUHOER—EPA305
275C FOR 3IRS
2 of 2
400 _
300 _
«~
LOO THERMOVELL
CENTER THERMOWELL
UPPER THERMOKELL
100 _
1
i 1
. 1
. 1
8
8
8
8
tr
i/l
iS
8
c3
8
8
kj
8
RUN TINE-HRS
8
cB
8
d»
8
c3
8
-------
2500
RUN NUMBER—EPA305
275C FOR 3JRS
2000
150C _
o
o
o
®
o
1000
QtfttAj
qP
o
0
0
0
°ffl
of ® <#
0
0
%
1 of 2
REACTOR HEAOSPACE
jO QWb
o- S, ^ 0
0 o
o °
500
1 °l
, I
. 1
i I
i 1
. 1
1 1
. 1
i I
. 1 .
8
8
8
8
8
8
8
8
8
8
i«
2
cB
dt
•¦4
(M
{3
8
rfi
RUN TIME-HRS
8
CNJ
Reaction Temperature and Pressure Profile, Run 305
-------
RUN TIME-URS
-------
500
RUN NUMBER—EPA306
305C FOR 3HRS
400 _
LOVER THERHOVELL
CENTER THERHOKELL
-------
2500
RUN NUMBER—EPA306
305C FOR 3HRS
2000
o 0 qfo0
O Ofli)V0 0 (PS?
>5p°-*
.v
: ° v
0) „
1000
%>
o
o
o o
- 0
°y
O 0
REACTOR HEADSPACE
I? 0 O 0°
o O o n°
^ //'/ft
IP
500 _
8
<3
8
dJ
8
d!
8
a
8
8
rU
I I I
I I I
8
8
8
irt
8
<£
8
8
ca
8
cK
RUN TIME-I1RS
Reaction Temperature and Pressure Profile, Run 306
-------
500
4C0
RUN Nl»M8ER- -EPA313
275C FOR 1.5HRS
300
u
CD
l
Ui
£ 200
100
•V\
LOWER THERHOWELL
w-'--.-. ^ *WV
o
o
o
~+ t.
CENTER TIERH0WELL
HP
UPPER THERHOWELL
<*b
\y* ++^
v
v ******* v
€»
o
o
% o
o
o
oa>
A
8
d>
§
dJ
8
8
—
RUN T1ME-URS
8
(it
s
8
i*
8
8
8
do
Reaction Temperature and Pressure Profile, Run 313
-------
2500
RUN NUMBER—EPA313
275C FOR 1.5HRS
2000
°°°®
o°o
* * rf>
0 oo
-------
5GC
| RUN NU<:8ER-EPr4:8
U
i
oo
4 CO |__
!
»
»/*j 7HERM0»r' •
TCP THERWutlL
o \f
o 1
°o«
\\
• <
• »
M
I ! '
1 J I— ) L
J j t__ L
_.j L a X
8
JS
t3
r'j
LJ
CO
L i
L)
CO
&
c"!
a
do
ifi
cl
PJ
..L.
u
s
rfs
CM
8
HU.I TI.'^-HRS
I
Reaction Temperature and Pressure Profile, Run 418
-------
25C(L
RUM NUMBER—EPA 418
i
f"
* Ifc'tfSMCE
^ ^ A\vv; o*
i * « * * ^ \« ** * **
l * * * : % ~ *
* * . « *
! /
;5&ol*_
j *
S\
\-
c.o !
Q.
2 $ iCccL
0$
e/> j
u» •
(r.
a.
i
•JoJ
L l_
s
i— -L .. 1 i_ . ^ _ L
ci c)
CJ 15
Kj rK
I-..-*.
--L,- .L - -
-L. x.-
CO
a
• j
. i_ _
i.- .4
C")
o
Q
LJ
O
8
1 f*»
33
1
P
c3
H
r"
do
d?
CJ
••
• -4
CJ
22:
cn
INJ
"V
cv
RU.M TIME-HHS
Reaction Temperature and Pressure Profile, Run 418
-------
500
400
<-> 300
CD
UJ
0
Q_
I— 200
I
100
TIME(HRS)
RUN- 502
_>000
.1500
CD
CO
Q_
OC
->000^
CO
UJ
oc
Q_
MIDDLE
TOP THERHOKELL
B0TT0M THERM0YELL
Reaction Temperature and Pressure Profile, Run 502
-------
<-> 300
CD
UJ
a
o
n_
200
TIME(MRS)
Reaction Temperature and Pressure Profile, Run 507
-------
500
400
<-> 300
O
LU
a
Q_
LU
I— 200
100
—2500
BOTTOM THERHOWELL
THERHOWELL
TOP THERHOKELL
_ 1500
_ 1000
8 8 8 8
3 H £ iS
TI ME(HRS)
CD
i—t
CO
a.
cc
rD
en
CO
LxJ
cc;
D.
Reaction Temperature and Pressure Profile, Run 509
-------
B-l. Trace Element Concentrations in Feedstock Sludges
oo
U>
WW 1
?w
305
I
306
313
418 11
AIR 1?
50?
507
509
SOIWCf
rr ramr wwrorwn
. ~
Ir
V
<».
<17.
<17.
<16.
??. 3
9.7
18.8
9.3
«W.
<16.
<17.
Cr
179.
1?.
8?.1
6.6
5A1.
?9.
80.5
6.8
103.1
8.?
96.5
7.8
97.6
1.2
11?.4
7.9
100.9
1.1
Hn
136.8
9.0
80.9
6i0
95.6
7.1
64.6
5.3
2.3
6.3
7?.0
5.8
69.8
5.5
74.4
5.5
81.3
6.0
HI
?5.1
?. A
17.3
1.9
?0.1
1.5
17.9
1.4
21.9
2.4
37.?
?.6
?6.9
?.1
22.1
1.6
?0. 8
1.6
Cu
690.
35.
480.
?4.
309.
16.
?86. «
15.
339.
17.
399.
?0.
30?.
15.
?90.
15.
?«6.
15.
In
M??.
56.
575.
?9.
616.
31.
516.
?6.
651.
33.
6R6.
34.
66?.
33.
65).
33.
77?.
19.
Ca
3.5/
0.12
4. ?8
0.58
1.9?
0.51
?.01
0.5?
3.?9
0.6?
?. 18
0.5)
?.?9
0.5?
?.04
0.50
?. 51
0 55
S«
A. 50
0.44
3.93
o.lio
?.?9
0.33
?.51
0.36
3.39
0.45
2.9?
0.34
?.64
0.35
3. ?4
0. 35
3.03
0.36
Ph
146.3
1.5
75.3
4.0
58.8
3.?
59.1
3.3
53.7
3.1
54.4
3.0
5?. 4
? 9
55.3
3.0
<9 0
3 ?
As
?.6
1.1
?.? 9
0.81
<1.3
1.59
0.75
?.63
0.86
?. 79
0.69
1.75
0.69
1 40
0.69
1. 77
0. 74
Br
t>. 93
0.79
7.58
0.55
5.67
0.47
6.6?
0.51
8.33
0.6?
9.04
0.59
6.93
0.51
6. 35
0.49
6. 76
0.51
Aq
Rt>
4.5?
0. 79
7.59
f
0.90
J.67
0.73
<1.5
5.01
0.50
3.67
0.68
3.38
0.71
4.31
0.7?
4. ?8
0.75
U
7.1
1.3
7.4
1. A
10.3
1.4
8.4
1. A
6.8
1.4
7.0
1.?
7.5
1.3
6.6
1.?
8.1
1.3
Sr
151.
11.
105.7
7.5
87.5
6.?
77.0
5.5
101.9
1.2
100.0
7.1
80.6
5.7
81.?
5.7
93.6
6.6
V
?.65
0.70
3.37
0.70
3.00
0.68
1.90
0.73
3.17
0.12
?.?1
0.59
3.3?
0.65
?. 69
0.63
?. 45
0.*6
It
51?.
30.
498.
35
80?.
56.
939.
66.
747.
*?.
635.
44.
5?9.
37.
493.
35.
514.
36.
Nb
<1.1
<1.?
<1.?
<1.3
1.59
0.65
1.51
0.53
<1.1
<1.1
<1.?
No
l?.A
1.3
11.6
r. 3
IA.7
1.6
15.6
1.7
1?. 3
1.5
19.1
1.7
1?.0
1.S
11.4
1.)
11.1
1.1
JW
Aq
M.?
?.8
?8.0
?. 4
?9.l
?.6
31.6
2.1
?8.7
3.0
?1.9
?.o
18.1
1.8
<5.?
20 1
2.0
u
5.0
1.5
<3.2
<3.1
<3.?
<5.0
CL
o
-s >
lO ~u
Of T3
3 m
-*• z
n o
«—«
m x
o> CD
3
3
3
t/»
to
-------
B-2. Trace Inorganic Components in Standards
XI IS
(I
IMA SOU S
I631A IIYASII
N8S COAl
106 SlUOGC IP*
|Af* SOU
,-s
~/-
»/-
• /-
• /-
4/-
X
Al
«.s»
0.47
14.41
0. 75
1.54
0. II
0.152
0.0/2
8.52
0.46
z
SI
29. J
1.5
24.5
1.2
2.91
0.15
2.01
0. II
29.0
1.5
X
t
<0.065
<0.056
0.078
0.011
0.845
0.01*0
<0.064
I
$
O.OI5
0.012
0.217
0.017
1.551
0.0/9
0.612
0.011
0.00/
0 012
X
CI
o.os 7
0.00ft
0.011
0. 141
0.008
0.0/0
o.oos
0 059
o oon
X
I
M9
0.12
2.27
0.12
0. 321
0.016
0.150
0.008
2.1/
0.12
I
CA
2.64
0.11
1.211
0.062
0.199
0.020
1.9/9
0.099
2.60
0.11
~/-
• /-
«/-
4/-
•/-
X
Al
1.62
0. 79
11.02
0.96
1.80
0.19
1.80
0.49
7.82
0 80
X
SI
26.7
1.4
22.5
1.2
2.77
0.21
2.06
0.25
26.6
1.4
X
r
tO.il
<0.16
<0.14
0. 78
0.11
<0.36
X
s
<0.16
0.199
0.081
1.390
0.084
0.480
0.052
<0. 16
X
ci
0.0/8
0.034
<0.066
0.116
0.016
0.062
0.017
0.006
0 0)6
X
K
1.81'
0.095
1.671
0.087
0.291
0.016
0.149
0.010
1.861
0 09/
X
CA
2.12
0.12
1.024
0.054
0.4S8
0.024
2.41
0.12
2.33
0 12
X
II
0.411
0.024
0.751
0.03S
0.107
0.006
0. 106
0.006
0.469
0 024
PPM
V
III.
25.
186.
11.
22.5
7.1
<17.
129.
25.
PPM
CR
14.
11.
222.
20.
24.5
1.9
541.
29.
IS.
13.
prM
MM
879.
47.
97.
15.
14.4
1.4
95.6
7.1
848
48.
X
n
4.SI
0.24
9.29
0.46
0.819
0.041
0./81
0.039
4.81
0.24
PPM
Ml
<10.
122.
10.
14.9
I.S
20.1
1.5
<10.
PPM
CU
;9.<
4.7
119.2
6.9
16.2
I.I
109.
16.
82.8
4.9
PPM
IN
390.
20.
221.
12.
15.5
1.9
616.
)l.
386.
20
PPM
C*
16.4
1.1
54.2
1.1
5.19
0.45
1.92
0.5l
16.3
1.4
PPM
St
<0.96
9 44
0.85
2.40
0.12
2.29
0 11
<0.9/
PPM
P8
ISI.S
• .0
69.1
4.1
7.54
0.91
58 8
1.2
151.9
8 1
PPM
AS
97.6
5.1
144.2
7.5
5.07
0.49
<1.1
95 1
5.2
PPM
SO
6.41
0.79
2.52
0. 79
19.9
I.I
5.67
0.47
6.89
0.//
~/-
«/•
~/-
4/-
41-
PPM
*1
|]|.>
9.1
140.1
10.0
21.2
1.7
1>67
0.71
1)0.9
9.1
PPM
u
8 2
2.7
11.6
1.1
<2.8
10.1
1.4
<5.4
PPM
s«
363.
25.
642.
59.
159.
II.
8/.5
6.2
361.
25.
PPM
V
24.7
1.9
92.0
6.6
10.42
0.94
1.00
0.68
25.8
I.
PPM
21
258.
IS.
111.
21.
38 6
2.8
802
56.
259.
18.
PPM
N9
9.7
0.97
29.7
2.1
1.95
0.64
<1.2
10 15
0.99
PPM
Ml)
4.91
0.97
15.1
2.7
2.66
0.68
14.7
I.S
1.5
0.95
4/-
«/-
4/-
»/-
<1.4
l/-
PPM
Ac
4.6
I.S
6.1
I.S
<1.0
29.6
2.6
TPM
CO
<1.1 '
<1.0
<1.1
<1.2
<1. 7
PPM
SN
12.1
2.5
9.4
2.5
<4.1
66 9
S.l
10 9
i. 4
PPM
SO
15.9
2.9
9.9
2.9
<4.8
<4.8
20.6
1.0
PPM
II
<5.1
<5.7
>.1
2.0
-------
B-3. Inorganic Element Content of Product Oils
Inn I
M
r
"711'
~5OT:1~
"sor-r
vr
oo
in
Suyrcc
I Icaurnl
l/l
tL
iL
~/-
~/-
~/-
»/•
~/-
It
Al
lb JO.
270.
<2370.
400.
<2460.
yaa.
• 1430.
290.
<4650.
530.
<1680.
150.
<1750.
140.
<2060.
140.
Si
411.
96.
?060.
210.
530.
130.
110.
110.
?80.
130.
<220.
340.
110.
320.
120.
P
S18.
62.
960.
120.
350.
100.
550.
• 7.
510.
120.
950.
120.
660.
100.
960.
120.
S
SB 70.
310.
6550.
450.
6270.
310.
4830.
110.
9900.
520.
II140.
500.
8790.
460.
11950.
620.
CI
656.
44.
1660.
110.
22200.
1100.
24/.
26.
6180.
330.
285.
12.
6610.
140.
561.
41.
K
15.2
4.0
161.
13.
46.2
5.5
>2.1
i.4
19.0
6.1
115.8
8.9
51.5
5.5
9.2
1.1
c*
163.4
9.7
1604.
62.
411.
22.
172.
20.
84.8
6.8
52.8
4.8
41.5
1.9
21.9
1.0
tr
It
41.4
rr
114.
10.
U.t
5.1
35.2
477
41.4
n
24.4
err1"
16.1
5.4
<1.8
V
10.?
3.0
16.6
4.6
13.4
3.1
12.5
2.7
22.2
5.1
27.8
4.6
17.8
1.7
9.2
1.0
Cr
11.3
2.4
46.7
4.5
19.5
1.7
14.3
2.2
11.9
1.6
11.7
1.1
8.2
2.6
6.2
2.1
Mn
3.5
1.5
12.1
2.4
<2.9
2.7
1.2
<4.0
<1.6
4.0
1.5
5.9
1.4
r*
561.
29.
1966.
99.
177.
19.
478.
24.
782.
40.
152.
18.
117.1
7.)
78.6
4.5
Co
<4.7
<5.5
<3.6
<2.1
4.1
2.1
• 2.6
<1.8
<1.8
Ml
14.1
1.4
19.2
1.4
9.31
0.04
S. 29
0.74
9.60
0.99
8.00
0.81
5.49
0.64
1.46
0. 59
Cm
46.4
3.0
111.7
5.6
15.3
1.1
11.01
0.90
6.61
0.71
6.12
0.61
5.94
0.57
1.75
0.52
In
106. 1
S.9
162.9
9.3
43.6
2.4
41.»
2.2
59.0
1.1
17.3
1.0
20.1
1.1
5.11
0.49
4.2
<4.4
O.I
<1.8
<4.1
**
U
<3.4
<2.0
• 2.3
<1.6
T. 14
~6.»"
" «».i
<1.7
<2.1
Sr
<1.6
6.62
0. 79
1.96
0. 54
1.79
0.42
<0.86
<0.76
<0.76
<0.47
f
<1.9
<1.0
<1.2
<0.85
6.28
0.64
<0.85
<0.90
<1.1
i
lr
<2.0
60.6
4.3
7.93
0.87
39.5
1.9
<0.89
9.04
0.80
12.7
1.0
6.75
0.76
Ut
5.5
>5.5
•«
<12.
46.5
4.4
14.1
4.1
27.1
1.8
13.1
4.5
1.7
2.1
7.2
1.1
<6.5
U
<15.
9.1
1.7
<10.
•.1
4.1
<11.
<8.1
<8.1
<8.1
c«
<16.
<7.7
<12.
<9.5
•M.
<6.7
<».S
<8.8
-------
Inorganic Element Content of Product Oils (continued)
»n arc ssr? virt wm 5ur7i TtwtTsmn
~/-
~/-
~/-
~/-
~/-
»/-
~/-
A1
-------
B-4. Trace Inorganic Components in Chars
Run t
Source Element 208 305 306 313 418 IAEA S01L-5
V
<47.
<55.
<72.
<37.
<21.
129.
25.
Cr
603.
34.
1398.
74.
1107.
65.
719.
39.
220.
14.
35.
13.
Mn
460.
26.
564.
33.
451.
32.
331.
20.
112.0
8.1
898.
48.
Ni
159.0
9.5
432.
23.
308.
18.
267.
14.
73.1
4.7
<10.
Cu
4160.
210.
3430.
170.
2100.
110.
2130.
110.
785.
40.
82.8
4.9
Zn
3970.
200.
4080.
200.
2500.
130.
3000.
150.
1347.
68.
386.
20.
Ga
15.7
2.0
13.4
2.1
10.8
2.3
7.6
1.5
5.2
1.0
16.8
1.4
Se
10. 76
0.87
14.7
1.1
11.4
1.1
12.72
0.90
5.89
0.67
<0.97
Pb
618.
31.
509.
26.
322.
17.
469.
24.
156.9
0.67
153.9
8.1
As
<5.0
<5.3
<5.7
<4.2
<2.8
95.1
5,2
Br
2.36
0.51
3.68
0.61
1.60
0.68
4.26
0.51
12.68
0.91
6.89
0.77
Rb
5.1
1.1
13.4
1.6
16.7
1.7
6.4
1.1
5.99
0.66
130.9
9.3
U
25.7
2.5
36.3
3.5
30.3
3.0
19.2
2.2
12.8
2.0
<5.4
Sr
369.
26.
396.
28.
379.
27.
239.
17.
129.5
9.2
361.
25.
Y
8.7
1.1
10.7
1.4
13.4
1.4
9.6
1.3
<2.0
25.8
2.0
Zr
607.
43.
2240.
160.
2610.
180.
3160.
220.
1227.
86.
259.
18.
Nb
4.61
0.71
4.76
0.93
5.25
0.85
3.23
0.78
1.94
0.85
10.15
0.99
Mo
72.7
5.2
140.
10.
160.
11.
116.9
8.5
34.6
3.1
3.50
0.95
Ag
166.
12.
161.
11.
165.
12.
187.
13.
70.2
5.9
<3.4
Cd
18.6
2.0
15.8
2.2
15.2
1.9
15.9
2.0
<7.0
<3.7
Sn
513.
36.
367.
26.
229.
16.
329.
23.
10.9
2.4
Sb
17.5
2.5
14.2
2.8
11.2
2.4
13.3
2.6
130.
10.
20.6
3.0
Te
<4.6
<5.3
<4.4
<5.1
<11.
<5.4
Cs
<6.0
<7.7
<6.3
<6.5
<14.
55.9
5.3
Ba
1790.
130.
920.
140.
1900.
130.
1215.
85.
520.
38.
644.
45.
La
11.8
4.4
19.2
5.4
10.5
4.3
15.8
4.9
<20.
31.5
5.4
Ce
29.1
5.4
23.2
6.3
20.7
5.1
27.7
5.8
<23.
59.5
7.2
-------
APPENDIX C
Oil Thermal Analysis Profiles (TG and DCS)
. - • - .w r**e. 22. a*' o»*
\22. 222
FKUN. 27. 36
*Oi 524. 45
V 2 CHfiHGE* 85.77
1A 1c a tk& :i&' :«.» zst.» ?ia? si a: *.c «ia? *« se :*•
*U£>S*GP?. "5"
D#*ti »4 c* ;s *:ke» :4»5:
J?- !N HEL2UV
"EWrERA'JRE -C
¦ *: 6.381? »g SCAN RATE* •!.£»' ft vr
s Jis
TROW, 26. 62
TO» 522. 55
¥7. X CHANCE. 86. 46
X
c
UJ
< sea?
—
X.R1
file* <>«??:. rc
da*e, 84 c :& ':me, :4, e*
J* BP IK. BP 230.B* 33^^ 3C.fi
-------
2C(- ¦:
• " t Z. 4C •«
SC^'< ®' "E: CS
7B.B
in. i
riLSt SM042. j/
B*/r? -3e 'ihe, :sh :e
m 23ktt Z7K.B
"EH^ERaTuRE (C)
2SZ
208 IN NITROGEN
*rr, ie. se -c
SCAN :'*T£j 22. 22 oma/m
V
r.v£' sxr?-. .>
ti "2" '3! *:ȣ.
*cMO£P.«*^P£ <:C)
RUN 208
it
zsz
89
-------
*35 In mr
• '» 13. 4928 SCAN w^TFi
¦V. W c5.a -i~
iaa. 02*
FROM. 27. 36
TO. 524. 47
¥T. Z CHANCE. 74. 4"*
O
* St* -
a*
a «• :a &' i
J£* f JLt* smm** rs
0*"fc» 84/a" 20 *K, 56.x:
305
lar * m * w s hie ma
TEMPERA" _.PE 5 «»g SCAN RATfjt 22. SS J«» ^ TC
>'f» »j4 z" 2e -:#£, ^r*i22
'» 0* "HP A' ma
TEMPERATURE «C>
RUN 305
90
-------
-------
303 aw
12. !6li •<) SCA* «?2. &' «• a •f
X
ur
•*~, e-: .t~ ?e :a. 4
33£ :n nr.
ITT. a. a«&» SCAH RATE, 20 n
X
o
TL"MPL*PATURe (C> T&
RUN 306
!V »
Jfh r/i ti shops. v.
B#Tt. b4 07/» 0B.8L'
92
-------
-------
313 IN ^ IP !»*# '3. 421 3C** ¦>' tj ?0. 22 <4«*p/»kf^
lS ** iar :i» .iff rk* r»» sear «:V* ri* oka
J£H FlLEt SM8Z-. *S TEMPERA 7\jft£ »tj SCAN SA*E» 22.22 n«a • i »*
L& 12 SB W.K blk
-------
95
-------
-------
-------
ls - ¦
3B.fi fi.fi :52.X 2E.fi 271 ¦ 33B.fi 3BB.fi 45Z.fi 51E.fi 57E.fi 532*
rrt&smie. re ^EmPERaTjre £¦ =-,« run 502
98
t
-------
"* ea -a.® ;i.w slv :slt.ra ac lia? ji«az a.«
•f 'L£* SM&j:.JA 'C(/pCfjj ~ _y£ r- 2SZ
*Et s4 C? 26 )» 3
RUN 502
99
-------
JS. 87X
FROM.
TO.
WT. X
1L c.
_*r~ 75
0.*'£i B*/er-' 2* TIME. 11. S5
TEMPERATURE CC>
t522 *2 IN UELIJV
.3123 -O SCAN «AT£. ?Z.Z2
ama —i~
PROH. 26. 96
TC. 523. 54
*T. X CHANGE. 96.03 \
U
* ji.a -
12. 22
SL3J M.ffl 152.22 21
jeh Fr:.£ts*r:z. ~c
-/*£• a* '37.'32 J8« -iS
ra 22 m &• «L 22
~rMrERA TJRE
RUN 502
ra.2E
26. 96
623. 52
CHANCE.
100
-------
, v * • . . . . . • »
3c.» :x.v tap rt# 3 :£.& ?*&• m* <*.*'
/ "V.aSl-jE^ - . 2S"
/ LV'ti 7~ '7S. T:h- :*,
RUN 502
101
-------
-32 IN a:»
SC*S -'*?
2C. .V - -
:» .r
•J
•< biLff
X
HI
-« si.ar
JSJV fJlEi SMB&ff. 'S -r »>prp^7jpr
64 r ;>4 *:k?» :?» e
53? IN HELI -V WT, 16. "4Zs -a SCAN »AT£*
sl» 52. w :52.» ?:*.»• 7% a ne.* star « r2 ar 5:4a? rais em #
T£MPE&A7L»E O TB
RUN 502
Jen riL£tSMe::. rc
DAT£» B4'C~ ?3 T!MEi 0fc< 41
102
-------
•' » ~i c -ia ik« :VK i* **= "1W ,:|LJ' s a v '
CUE, SMB33. 3* TMPta/ 7 JRE
a7/r*
RUN 502
103
-------
*4 Ir. * If?
• T, 22.2223 -<» SCA* **"h 22.22 m/->
c»C«, 20. *3
*C. 923.95
«T. : CHANG€. 91.96
a
« fx
is —
*• v? ?:a.a
jev - tl& stmet. rs
DAT£» 04/®" *24 TI^E. 13.5?
J~ #4 IN HELI ..
s.® ma? «*.a 4«*.« ma? no *ai a
TEMPERATURE »* -
104
-------
SEAf. a»"F, ;-a. &'
'A'.
i" "2.® "i* ;i« :ijf ?*.£• ri.® jsts
"f- -.".A 3P (-
•*£. S< B~ 3: "!ve. 28. J~
i» t-
5C *4 :• mtrc;, -.
a. *» -3
S£> • ->rEi 22. *;• .1.-1 «lr
** "is i« :;k* :i* zb.«' r-j.*
J£fc Ptl -1/ SmZH. 2s TEMPEP/ ~~r>£ r
OTE. »i'a--2r *!>»£. cs. y>
RUN 507
105
-------
53" 05 I*. *1*
! 0, «* "1-£* • **"-» .7 »V va •-»
ram rk# a <£.* yjlx ri ar ria?
Fli£,SM*:x v. TEMPER/
-------
107
-------
505 :s *:p
?34* •<> - «»• RA'ti i «?. ¦***
rwQNi 2e. 9*
TQi R3. 51
fcfT. X CHANCE. 95. B"7
O
* y ar -
M
4.B32 /'
ma? «2 .v \ jl& : :i a
J£H FlL€» SMJJ4. 727
G*TE» 84/87/25 TINE, 0B* *.4
509 #5 IN H£.I
e i* a? tf ** ai
TEMPER*'^RE te, 84/97/23 T3Mt, 12, bfc
7*» r* a 4* c QLi
TEMPERATJR£ CCV
RUN 509
n» o» it
108
-------
\
J
.4 "jM \a si.* -*jp -i-.v lis
'• ft i*f~ ' 'EClfPA 'jt .CJ
. i4 e- rr a.?-- RUN 509
jsn
109
-------
s.s a.» :».* ?:«.* rz.*' ma* a> s- a' b:a* ya.®
JF* rtL£, SMB2S. 7C TEMPERATURE TG
DA^Et 04/®** 23 TIME, 13.5?
RUN 507
110
-------
t
:i x :i.* ail* rtt! r-i* *«
ssr
r:i£tsmex-Sf *emP£Ra~ s*
a* W 2" TJX', ;2.?~
RUN 507
111
-------
539 7 mq sc/tN fwrE. x.x
I. !t — ~ • • • • —» -» ,1 ¦ • » ¦ --
CAT «« ISA. Ji .1 &i '1LJB Q2.2? «L X t'g.a 5*.A« STUB 'Til »
JT*- rfifrSMBeS. T TfcMPEPA *"URE (C) TC
3A"£» 34/B7/25 Tlwf, ;a. i?
5d5 • * I N I " W. 9.2411 *3 SCAN JtftTE> 2Z. 23 (teg/aitn
«la' * * r.i* "ia m« *tA« 5if.« n* ss*
je± *7l£* sm:*. *; tmpera1^ jPE (C- tC
jjate, w-'r '3 •:**« :-t«-»5
RUN 509
112
-------
-------
509 #7B IN AIR WT. 5.0823 mg SCAN WF. 20. 00 d-q/.i..-'
JEH r.'LE, S*a?6. TC TEMPERATURE CO TC
DATE. 84/07 '25 TIME. 11.12
509 #7B IN HELIUM WT, 12.618? .3 SCAf. RA"E. 22.30 dmg/mir.
JEH r:LE,SM017. TC TEMPERATURE (O
3ATE. 84/07/24 TIME. 07. 19
RUN 509
114
-------
as "ije ::*« s.« st* rt.u 3:ijb sts
.'¦.v -:„s,s.v?sun "tMPtSAT'jRE CO -SC
3>" a-l -3""5r ": -•£« '9.33
RUN 509
115
------- |