METHANOL PRODUCTION BY DIRECT OXIDATION OF METHANE IN A PLASMA REACTOR by RICK MOODAY, B.S. A DISSERTATION IN CHEMICAL ENGINEERING Submitted to the Graduate Faculty of Texas Tech University in Partial Fulfillment of the Requirements for the Degree of DOCTOR OF PHILOSOPHY Approved Accepted August, 1998
148
Embed
METHANOL PRODUCTION BY DIRECT OXIDATION OF METHANE …
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
METHANOL PRODUCTION BY DIRECT OXIDATION
OF METHANE IN A PLASMA REACTOR
by
RICK MOODAY, B.S.
A DISSERTATION
IN
CHEMICAL ENGINEERING
Submitted to the Graduate Faculty of Texas Tech University in
Partial Fulfillment of the Requirements for
the Degree of
DOCTOR OF PHILOSOPHY
Approved
Accepted
August, 1998
ACKNOWLEDGEMENTS
I would like to thank Phillips Petroleum Company for their constant support
during the course of my work. This research project was made possible by their generous
financial assistance. I would like to extend my gratitude to Dr. Uzi Mann, for his
guidance, creative ideas, encouragement, and unlimited patience. Dr. Mann ensured that
the project never strayed off the correct path and often contributed well beyond the call of
duty as my committee chairman. Sincere thanks are also extended to Dr. Richard W.
Tock, Dr. Dominick J. Casadonte, Jr., and Dr. Lynn L. Hatfield, for their active
participation in the project and technical guidance while serving on my committee. Dr.
Raghu S. Narayan was extremely kind and made me feel more than welcome in the
Chemical Engineering Department at Texas Tech.
I would like to extend special thanks to Dr. Robert M. Bethea. My experience
working for Dr. Bethea in the Unit Operations Lab re-introduced me to education and
chemical engineering after being away for a time. His example and guidance have been
invaluable and will continue to serve me well throughout my career.
Many people contributed to making this time a success for me. Dr. James Riggs
is a highly respected chemical engineer and faculty member at Texas Tech, but I
acknowledge him here for his ability to play the sport of golf I must admit that he taught
me much about course etiquette, scoring, foreign golf traditions ("Aussie Rules"), and
temperament. I will miss our morning outings but I will not forget the significance of a
"tainted par" or a "fully legitimate birdie."
11
Thanks must go out to a few of the people who helped me through the difficult
times during my pursuh of this degree. My good friends Ravishankar Sethuraman,
Mahesh Rege, Aashish Ahuja, Robert Ellis, Steve Tsai, Johnson Fung, Siva Natarajan,
Scott Hurowitz, Coby Crawford, and Mike Barham deserve special mention. They are
responsible for many good times at the office, on the golf course/football field, or at some
other establishment. Everyone should be so lucky to have friends like these.
Many thanks go out to Bob Spruill, Marybeth Abemathy, Tammy Low, and
Kathy Womble for their unfailing support. It was a pleasure to work with such a group of
professionals who always got the job done well, and on time.
I must express my deepest thanks to my family and friends for their unconditional
love and support throughout my life. Distance, no matter how great, has never cracked
the foimdation that their love gives me. When I am reunited wdth them, it is always as if
we never parted.
Lastly, and mostly, I would like to thank my parents. Loving, selfless, and
capable parents are and have always been my greatest earthly gift. They taught by
example and never discouraged my dreams or desires. They made my education their
priority and selflessly gave of themselves to that end. Only after I became a man did I
begin to understand the sacrifices that they made for me and my sisters. I hope to be as
strong as they one day, and I will consider myself to be successful only when I have
given to my children what they have given to me.
Ill
TABLE OF CONTENTS
ACKNOWLEDGEMENTS ii
ABSTRACT viii
LIST OF TABLES x
LIST OF FIGURES xi
CHAPTER
I. INTRODUCTION 1
n. TECHNICAL BACKGROUND . . . . 6
2.1 Background of Natural Gas and Methanol 6
2.2 ABrief History of Methanol Synthesis 7
2.3 Commercial Methanol Synthesis . . . 8
2.3.1 Synthesis Gas Preparation Techniques 8
2.3.2 Methanol Synthesis from Synthesis Gas 14
2.4 Potential Advances in Methanol Use . . 2 0
2.4.1 Methanol as Transportation Fuel 20
2.4.2 Methanol to Facilitate the use of Methane 22
HI. LITERATURE REVIEW 24
3.1 Homogeneous Partial Oxidation . . . 24
3.1.1 Effect of Reactor Walls, Additives, and Promoters 2 5
3.1.2 Kinetics and Kinetic Modeling 27
IV
3.2 Heterogeneous Catalytic Partial Oxidation 29
3.3 Methane Oxidation in the Liquid Phase 32
3.4 Methane Oxidation in Plasma Reactors 34
IV. PLASMA 36
4.1 Background . . . . . . 36
4.1.1 Plasma in Nature . . . . 37
4.1.2 Potential Applications for Plasma 38
4.2 Plasma Characteristics, Generation and Uses. 39
4.2.1 Plasma States . . . . . 39
4.2.2 Plasma Generation . . . . 41
4.3 ABrief Description of the Physics of Plasmas 45
4.3.1 Criteria for Plasma Occurrence 47
V. EXPERIMENTAL SYSTEM AND PROCEDURES 50
5.1 Experimental Approach 50
5.2 Experimental Apparatus . . . . 52
5.2.1 Feed System . . . . . 52
5.2.2 Plasma Generation System . . . 56
5.2.3 Reactor System . 61
5.2.4 Product Collection System . . . 62
5.2.5 Product Analysis . . . . 63
5.3 Experimental Procedures . . . . 65
5.3.1 System Preparation and Warm-up 66
5.3.2 Experimental Run Procedures and Checklist. 68
distances, 2 levels of oxygen concentration, and two levels of water concentration. The
oxygen concentrations tested were at the lower limit of what was reliably measurable by
92
T3 93
0.014
0.012
0.01
0.008
0.006
o «
^ o
J3 a S "
2 a 0.004
a 0.002
0
A
1
1 ' 1
y. \ 1 1
1 ^ 1
I 1
"
1 i
• Tv
ij • X
1 i • ^ ^ ^ " ^ - ^ ^
r—^——
!
1
i 1
> •
0 0.05 0.1 0.15 0.2 0.25
Methane Conversion (mole reacted/mole fed)
0.3
• Preliminary Phase Experiments • Phase I Experiments (3/8 inch Injection)
A Phase I Experiements (5/8 inch Injection) X Phase I Experiments (1.5 inch Injection)
Figure 6.8. Preliminary Phase and Phase I-Methanol Selectivity Versus Methane Conversion
93
the apparatus. The upper water level tested was near the upper limit of experimental
feasibility (i.e., more water would have made h impossible to complete the experiment
effectively). Table 6.3 shows the experimental parameters from Phase II runs.
Table 6.3. Phase II Experimental Parameters
Experiment
0213 0215 0217 0218 0225 0226
0226-2 0226-3
Injection Distance (inches)
0.375 0.375 0.375 0.375 0.250 0.250 0.250 0.250
Ar/CH4 Ratio
7.22 7.02 8.02 7.98 8.02 7.98 7.78 7.82
CH4/O2 Ratio
4.75 4.75 4.75 4.44 4.75 4.44 4.44 4.75
CH4/H2O Ratio
1.04 0.89 1.90 1.93 1.67 1.96 0.96 0.95
Methane Conversion
(%)
0.453 0.756 2.035 3.161 4.769 6.062 2.450 4.505
Methanol Selectivity
(%)
3.715 1.826 1.301 0.854 0.785 0.578 0.990 0.596
6.3.1 Injection Distance
From inspection of the methane conversion attained in Phase II runs, injection
location behavior parallels the behavior observed in earlier phases. More distant
injection resulted in lower methane conversions and conversion levels were in the desired
range (<10%). The locations selected for methane injection were appropriate for this
experimental apparatus and these experimental conditions.
6.3.2 Overall Performance of Phase II Experiments
The selectivity levels obtained in Phase II experiments are clearly higher than
those of earlier phases. Figure 6.9 shows the comparison of Phase II data with both the
94
0.04
T3 0.035
93
13 C/2
§
J3
u (J 3
2
J3 • « - •
(L)
s "o B
0.03
0.025
0.02
0.015
^ 0.01
0.005
0 0.05 0.1 0.15 0.2 0.25
Methane Conversion (mole reacted/mole fed)
0.3
• Preliminary Phase Experiments • Phase I Experiments (injection at 3/8 inches)
A Phase I Experiments (injection at 5/8 inches) X Phase I Experiments (injection at 1.5 inches)
D Phase II Experiments (injection at 1/4 inches) • Phase II Experiments (injection at 3/8 inches)
Figure 6.9. Preliminary Phase, Phase I, and Phase II-Methanol Selectivity Versus Methane Conversion
95
Phase I data and Preliminary Phase data. Progression of the experimental data away from
the origin is desired. The optimum trend in data would be high conversion and high
selectivity (i.e., progress up and to the right on the Figure 6.9). The data from Phase II
experiments shows a definite increase in the methanol selectivity over Preliminary Phase
and Phase I data. The data exhibh considerable scatter, but the Phase II data confirm
system performance improvements, especially at the 3/8 inch injection distance.
The data from this research indicate that the best methane-to-oxygen ratio for methanol
production from methane for this apparatus is approximately 4.5 to 1. A higher methane-
to-oxygen ratio may result in performance improvements but this apparatus is unable to
operate at higher ratios.
The data from this research indicate that the best water-to-oxygen ratio for
improved methanol selectivity is approximately 5 to 1. This is a surprisingly high
concentration of water. Methane conversion is favored by lower water levels because
water dilutes the system and interferes with the reactions between oxygen and methane.
High water content has been shown to favor methanol selectivity, while low water
content favors methane conversion. Data from investigations into the effects of water
content also illustrate the competition between methanol selectivity and methane
conversion.
Water has been shown to participate favorably in the production of methanol from
methane by increasing the number of pathways towards methanol production (Badani et
al., 1995). This research supports the claim that high levels of water in the plasma
stream, relative to oxygen, improve methanol selectivity. Little has been written about
96
how water participates in other important reactions occurring in this type of plasma
reactor.
For Phase II experiments, oxygen and water levels are at or near the experimental
apparatus limhs. It is not possible to make significant changes in these two parameters
without system redesign and reconstmction. Further refinement of reactant levels could
continue, resulting in slight improvements in system performance. Small refinements
will not solve this synthesis problem. Large-scale (i.e., order of magnitude)
improvements must occur for this approach to become feasible and it is unlikely that
ftirther modification of these parameters will produce significant improvement unless
changes in system design incorporated. For this reason. Phase III experiments will not
attempt further refinement of reactant concentration levels but concentrate on
investigation of a novel approach aimed at improving system performance.
6.4 Phase III Experiments
It has been theorized by other researchers that water contributes to methanol
production by increasing the number of pathways that lead to its production. Apparently,
the presence of water and radicals generated from water improved the likelihood that
methanol would result from free-radical reactions. This idea of increasing the number of
methanol synthesis pathways led to a new approach for this research. For Phase I and II
experiments, oxygen (along with argon and water) passed through the plasma region and
then reacted with the methane stream. The methane stream has always consisted of pure
methane. If oxygen were present in the methane stream, it would not be "activated" by
the plasma and could react at a lower energy state than the more energetic oxygen in the
97
plasma stream. It seems plausible that oxygen introduced by way of the cooler methane
stream would exhibit different reactivity and could potemially improve the pathways to
methanol. Phase III experiments divided the oxygen into the two reactant streams to
discover if this created measurable improvemems in methanol production.
A single experimem was conducted in which all of the oxygen was placed into the
methane stream and isolated from the plasma conditions. In light of the other planned
experiments, an experiment of this type was warranted to obtain some perspective about
system reactivity when oxygen was only presem in the methane stream. This approach
was not expected to produce significant conversion of methane since oxygen would not
be energized in the plasma stream. The oxygen was well-mixed with the methane and the
low energy state of the oxygen should have prevented appreciable reaction with methane.
This particular experiment produced some interesting results. Table 6.4 shows
experimental parameters from Phase III mns.
Table 6.4. Phase III Experimental Parameters
Experiment
0330 0331 0403
0403-2 0405 0406
0406-2
Injection Distance (inches)
0.250 0.250 0.250 0.250 0.250 0.250 0.250
Ar/CH4 Ratio
7.22 6.88 6.88 6.88 7.22 6.19 6.88
CH4/O2 Ratio
4.75 4.71 4.05 4.05 3.48 5.53 5.53
CH4/H2O Ratio
1.01 1.13 1.54 1.02 0.95 1.64 1.07
Methane Conversion
(%)
1.679 1.036 2.189 1.614 1.142 2.311 2.056
Methanol Selectivity
(%)
1.349 1.534 0.345 0.744 0.766 0.198 0.225
98
6.4.1 Oxygen with Methane Stream
The first experimental mn of Phase III (mn 0330) was conducted whh all of the
oxygen placed in the methane stream. Appreciable conversion of methane was not
anticipated in this mn. This flow configuration did indeed achieve measurable methane
conversion and respectable methanol selectivity. The run replicated the experimental
conditions of a Phase II experimental mn (Phase II, mn 0213), with the noted exception
that oxygen was included with the methane stream. That particular mn (0213) had
achieved the highest methanol selectivity of any experimental mn.
Including oxygen in the methane stream reduced methane conversion, relative to
the earlier run of Phase II. This reduction in conversion was expected because a primary
reactant has been isolated from the highly reactive conditions generated in the plasma. (It
should be noted that placement of oxygen in the methane stream induces greater
penetration of the methane stream into the plasma zone because of the mixing effects
discussed above). The high conversion and selectivity of this run implies that one, or a
combination, of the following must be tme.
1. The argon and water in the plasma stream must, to a certain extent, be capable of
producing species that (a) survive long enough to contact the methane stream and (b)
are effective in oxidizing methane.
2. The methane stream, now well-mixed with oxygen and more reactive, is significantly
easier to convert.
3. The high energy of this microwave generation system "leaks" down the waveguide
"sleeve" (which houses the quartz tube) and contributes to methane conversion.
99
4. The methane stream penetrates considerably ftirther into the plasma zone than
expected. (This factor probably has limited impact since oxygen makes up less than
2.5% of the total gas flow.)
The implications of this particular experimental run were not investigated ftirther. This
experimental mn reinforces the ideas discussed earlier about extremely complicated
energy and mixing/flow considerations involved in this high pressure (~ 1 atmosphere)
plasma reaction system.
6.4.2 Oxygen Divided into Plasma and Methane Streams
There were 6 experimental runs in which oxygen was divided equally into both
the plasma stream and the methane stream. A variety of experimental conditions were
tested in the ranges that have been identified as optimum for this apparatus. The overall
performance results of this approach are shown in Figure 6.10. It is apparent that this
approach did not result in the order of magnitude performance improvements that were
desired. For reference, this figure includes the results from Phase II.
In general, Phase III experiments do not approach the performance of Phase II
experiments. As discussed above, quicker mixing should resuh from the impinging flow
approach, enabling reactions to proceed farther before conditions degrade. The theory of
improving the pathways to methanol by placing oxygen in both streams has not been
supported. Data from this research indicate that the presence of oxygen in both streams is
counterproductive to methanol production.
Several factors may contribute to the decreased performance of Phase III
experiments.
100
0.04
0 0.01 0.02 0.03 0.04 0.05 0.06
Methane Conversion (mole reacted/mole fed)
0.07
• Phase II Experiments (injection at 1/4 inches) • Phase II Experiments (injection at 3/8 inches) A Phase III Experiments (injection at 1/4 inches;pt O oxygen in methane stream)
Figure 6.10. Phase II and Phase III-Methanol Selectivity Versus Methane Conversion
101
1. The presence of oxygen with methane means that reactive species (intermediates)
have a higher probability of being consumed by oxygen at the mixing point.
2. Any methanol that is produced will contact oxygen with greater frequency because
oxygen is present throughout the system. Methanol is easily oxidized to carbon oxide
products in the presence of oxygen.
3. The presence of oxygen mixed with methane anywhere in a highly oxidizing
environment should contribute to greater amounts of CO and CO2 produced because
these products are favored by thermodynamics.
These and other factors may make it imperative that a pure methane stream be mixed
with a plasma activated stream, or combination of streams, to maximize methanol
production in this and other plasma reactors.
6.5 Overall Performance Evaluation
Figure 6.11 shows the selectivity/conversion performance of all of the
experimental phases of this research. The progression towards higher methanol
selectivity is illustrated and performance improvements are apparent from the
Preliminary Phase through Phase II. A new approach to methanol synthesis is attempted
in Phase III, but fails to result in performance improvements. The ability of this system
to produce methanol directly from methane has been demonstrated. This system
compares favorably with other more advanced, low-pressure plasma systems.
102
0 0.05 0.1 0.15 0.2 0.25
Methane Conversion (mole reacted/mole fed)
0.3
• Preliminary Phase Experiments
• Phase I Experiments (injection at 3/8 inches)
A Phase I Experiments (injection at 5/8 inches)
X Phase I Experiments (injection at 1.5 inches)
D Phase II Experiments (injection at 1/4 inches)
• Phase II Experiments (injectionat 3/8 inches)
A Phase III Experiments (injection at 1/4 inches, pt O oxygen in methane stream)
Figure 6.11. Preliminary Phase, Phase I, II, and III-Methanol Selectivity Versus Methane Conversion (all phases)
103
6.5.1 Mixing and Flow Analysis
The rationale behind the impinging flow design of this system was to improve the
mixing characteristics in the oxidation region with the goal of improving reactor
performance. A simple Reynolds Number calculation was performed to gain some
insight into typical flow conditions existing in the three streams (plasma stream, methane
stream, combined stream) during experimentation. Reynolds Number calculations for
streams during run 0213 are shown in Table 6.5. The calculation assumes the lowest
possible viscosity for the streams.
Table 6.5. Reynolds Numbers for streams during Run 0213
Reynolds Number of the Plasma Stream (0213) (assumed to be 400°C)
60
Reynolds Number of the Methane Stream (0213) (assumed to be 250°C)
35
Reynolds Number of the Combined Stream (0213) (assumed to be 250°C)
16
It is apparent that the flow conditions present in each stream are well within the
laminar regime. The benefits of the impinging flow design used in this reactor system
should be accentuated when streams in laminar flow are mixed. Although, it is not
known whether other researchers operated under laminar or turbulent flow conditions, an
impinging flow design will always improve mixing of the plasma and methane streams at
the mixing point.
6.5.2 Material Balances
The credibility of this experimental research demands that overall material
balances are presented. It is appropriate that balance closure be presented here to support
104
findings. Although significant experimental error are present, carbon and hydrogen
element balances closed to ±25% for all experimental runs. Only 1 experimental run
exceeded 20% error (20.3%) in the carbon balance, and none exceeded 20% error in the
hydrogen balance. Of the 36 mns, 7 exceeded ±10% carbon element balance error and 5
exceeded ±10% hydrogen element balance error. All other experimental mns were
within ±10% of closure for carbon and hydrogen balances. The inability to address
oxygen balance closure is a major concem in direct methanol synthesis from methane.
This point is addressed in greater detail below. This research was conducted using
existing department equipment which necessitated some important design and
experimental limitations. The major and minor sources of error will be discussed briefly
at the end of this chapter.
6.6 Sources of Error
In any experimental research project, serious consideration of the sources of error
must occur. It is not uncommon for experimental results to contain error margins of 50-
100%. Minimizing sources of error can be an expensive proposition as more accurate
and precise equipment is purchased, experiments become more numerous, and project
duration is extended. The limited budget available for this project made it impossible to
remove some considerable error sources that were present. The major sources of error in
this experiment are listed below.
1. Small errors, or non-closure, in the oxygen balance of this chemical system (methanol
production by direct oxidation of methane) have been shown to produce large errors
105
in system performance parameters, namely methanol selectivity (Helton, 1991). The
inability of this system to even measure oxygen balance closures must be considered
the major source of error in this experiment.
2. The analysis method used in this research required that columns be interchanged
between gas and liquid analysis. Removing, replacing, and re-conditioning analytical
GC columns can create major inconsistencies in performance. Optimally, two
different gas chromatographs, with the appropriate detectors, should be used in the
analysis. This would maximize the reliability of the columns and the analysis.
3. Although every effort was made to minimize the effects of transporting samples to
the analysis system, off-line analysis must be considered another major source of
error. All activities having to do with sample handling (see Chapter V) will introduce
errors. On-line analysis would minimize these sample handling sources of error.
4. Inability to maintain constant mixing and flow characteristics must be considered a
potentially major source of error. From run to run, slight changes in flow might
create significant changes in the mixing region. As discussed above, small changes in
this mixing region may impact reaction characteristics to a greater extent than the
other experimental parameters under investigation.
Other sources of error are listed below. These are not considered "minor" but
probably have less impact than those presented above.
1. The apparatus was vented to the atmosphere for simplicity. The pressure in the
reactor was approximately 0.9 atmospheres. This corresponds to the normal local
atmospheric pressure where the research was conducted. Changes in atmospheric
pressure did occur between experiments and must have affected system performance.
106
2. Accumulation of liquid product in the system. Efforts were made to minimize this
factor but, no matter how long the system "warmed-up," water balances exhibited
significant error.
3. Inability to precisely control water temperature in the water saturation unit made it
very difficult to closely control the amount of water placed into the plasma stream. It
was possible to maintain water temperature within a few degrees, but significant
changes in water content occur within a few degrees. Consequently, it was possible
to maintain water content within a certain range, but not at a precise point.
4. The errors inherent in regulators, flow controllers (rotameters), pressure gauges,
thermometers, timers, and GC calibration cannot be disregarded or ignored.
Additionally, the equipment used to constmct this apparatus was, by no means, new.
5. Concentration levels in the GC calibration gases are advertised to be ±5%.
6. Reactor/Feed gases have appreciable concentration tolerances (±5%).
107
CHAPTER VII
CONCLUSIONS AND RECOMMENDATIONS
This investigation into methanol production via an experimental plasma-based
reactor operated at atmospheric pressure leads to the following conclusions:
1. Methanol production by direct oxidation of methane in a plasma reactor operated at
high pressure (approximately I atmosphere) has been demonstrated.
2. Low oxygen concentrations, relative to the concentration of methane, contribute to
increased selectivity for methanol in this system.
3. High water concentrations, relative to the concentration of oxygen, contribute to
increased selectivity for methanol in this system.
4. High pressure plasma reactor operation creates high temperature gradients, inducing
significant velocity fluctuations in plasma streams. Complicated flow and mixing
phenomena result from the substantial temperature increases in these systems.
Reactor configuration and stream mixing techniques had significant impact on system
performance. These should be primary considerations when designing a high
pressure (~1 atmosphere) plasma reactor system.
5. The high peak power microwave system used in this research was able to ignite and
sustain a plasma in an argon environment without difficulty. Problems sustaining a
plasma experienced by other researchers could be alleviated by microwave systems
with higher peak power. This type of system is ideal for plasma studies using air.
because of the high peak power of the system and its ability to maintain the plasma.
108
6. Placing oxygen in both streams (plasma and methane) had a detrimental effect on
methanol production, relative to methanol production with oxygen present in the
plasma stream only.
It is not possible to make a conclusion about the effects of higher pressure on methanol
production. The higher temperatures associated with plasma at high pressure should
induce less selective oxidation of methane. Methanol production in this system is
somewhat less than other low pressure systems, but still comparable. This unit was not
operated at low pressure. Detailed low and high pressure investigations will have to
occur before a conclusion about pressure effects can be made.
Significant improvements in plasma reactor performance must occur before this
approach can become useful for methanol production by direct oxidation of methane.
Investigation into reactor configuration, mixing techniques, plasma generation
techniques, and high pressure operation are critical to progress in this area.
Based on the experimental and operational insight gained from this study, the
following recommendations are made:
1. High pressure plasma studies should be continue in this area. High pressure
conversion is cmcial if this type of plasma reactor technology is to operate
commercially. High Pressure plasma reactor systems should be developed that can
accurately evaluate many factors, including power input, effects of variable rates of
quench, effects of using air as a reactant, effects of impurities on system performance,
pressure effects, effects of reactant concentration levels, etc. The systems should
109
possess high reliability and minimize contributions from experimental and analysis
errors.
2. Studies into innovative reactor configurations and novel mixing techniques should be
undertaken. The physical and chemical interactions that occur in high pressure
plasma reactors of this type are very complicated. Optimizing performance in high
pressure plasma reactors may require complicated reactors and creative flow systems.
For example, combination of a plasma energized stream with another oxygen
containing stream(s), followed by subsequent mixing with the methane stream, or
incorporation of a methanol selective catalyst into the plasma system.
3. Serious consideration should be given to the energy source for the generation of
plasma. This and other approaches use pulsed systems to create the plasma. It is
recommended that a continuous source be considered for plasma generation.
Conditions created by pulsed systems are highly variable because energy is added to
the system in short, but powerful bursts, after which, the energy is completely shut off
(hence the term "pulsed"). If a set of condhions exists, under which methanol
production is favorable, then those conditions should be established and maintained.
Continuous sources will create a more "steady-state" energy environment. (Note:
Considerable difficuhies exist in igniting and maintaining a plasma, from a
continuous energy source, in any environment at atmospheric pressure.)
110
BIBLIOGRAPHY
Ammann, P. R. and R. S. Timmins, "Chemical Reactions During Rapid Quenching of Oxygen-Nitrogen Mixtures from Very High Temperatures." AICHEJ., 12(5), 956(1966).
Andmshkevich, T. V., V. V. Popovskii, and G. K. Boreskov, "Catalytic Properties of the Metal Oxides of Period IV of the Periodic Table with Respect to Oxidation Reactions," Kinet Katai, 6(5), 860(1965).
Badani, M. V., J. Huang, S. L. Suib, J. B. Harrision, and M. Kablaoui, "Selecting Oxygen Source for Partial Oxidation of Methane to Methanol Using Microwave Plasmas," Res Chem. Intermed, 21(6), 621(1995).
Baddour, R. F. and R. S. Timmins, "Introduction," in The Application of Plasma to Chemical Processing, R. F. Baddour and R. S. Timmins Eds., p. v, The MIT Press, Cambridge, MA (1967).
Bissett, L., Chem. Eng, 84(21), 155(1977).
Bittencourt, J. A., Fundamentals of Plasma Physics, Pergamon Press, Elmsford, NY (1986).
Bova, B.,The Fourth State of Matter: Plasma Dynamics and Tomorrow's Technology, St. Martin's Press, NY, (1971).
Brinkman, N. D., E. E. Ecklund, and R. J. Nichols, " Fuel Methanol: A Decade of Progress," PT-36, Soc of Automotive Engineers Warrendale Pa., (1989).
Burch, R., G. D. Squire, and S. C. Tsang, "Direct Conversion of Methane into Methanol," ' J Chem. Soc Faraday Trans I 85(10), 3561(1989).
Casey P S and KFoger, "Selective Oxidation of Methane to Methanol," iVa/wra/Ga^
' Conversion II, H. E. Curry-Hyde and R. F. Howe, Eds., Elsevier Science B. V.,
387(1994).
Chen, W. and H. H. Kung, Eds. Methanol Production and Use Mercel Dekker NY
(1994).
Chou, T. C, and L. F. Albright, "Partial Oxidation of Methane in Glas^and Metal Tubular Reactors,"/«^. Eng Chem. Process Des Dev.. 17(4), 454(1978).
Ill
Chun, J., "Direct Oxidation of Methane to Methanol," Ph.D. Dissertation, Department of Chemical Engineering, Texas A&M University (1992).
Chun, J., and R. G. Anthony, "Catalytic Oxidations of Methane to Methanol," Ind. Eng Chem. Res, 32,259(1993).
Chung, J. S., R. Miranda, and C. O. Bennett, "Mechanism of Partial Oxidation of Methanol over M0O3," J. Catai, 114,398(1988).
Dixon, C. N., and M. A. Abraham, "Conversion of Methane to Methanol by Catalytic Supercritical Water Oxidation," J. of Supercritical FL, 5, 269(1992).
Dowden. D. A., C. R. Schnell, and G. T. Walker, "Creation of Complex Catalysts," Proc 4th Int. Cong on Catalysis, 201, (1968).
Droege, M. W., L. M. Hair, W. J. Pitz, and C. K. Westbrook, "The Thermal Gas Phase Reactions of Methane and Oxygen: A Comparison of Modal Calculations and Experimental Results," AICHE Spring Meeting 1989, Houston Tx. Paper No. 52E.
Durante, V. A., D. W. Walker, W. H. Seitzer, and J. E. Lyons, "Vapor Phase Hydroxylation of Methane," 1989 International Chemical Congress of Pacific Basin Societies, Preprints of3B Symposium on Methane Activation, Conversion, and Utilization, Honolulu, Hawaii, (1989).
Dybkjaer, I., P. E. Hojlund Neilson, J. B. Hansen, "Synthesis of Methanol over Cu-based Catalyst," 74^^ AICHE Annual Meeting New Orleans, Nov 8-12, 1981.
Eliasson, B. and U. Kogelschatz, "Nonequilibrium Volume Plasma Chemical Processing," IEEE Trans On Plasma Sci, 19(6), 1063(1991).
Feng, W., "Reactor Design for Methane Oxidation to Methanol," M. S. Thesis Department of Chemical Engineering, Louisiana State University, (1993).
Frank-Kamenetskii, D. A., Plasma: The Fourth State of Matter, Plenum Press, NY, (1972), translated by J. Norwood Jr.
Fukuoka, N., K. Omata, and K. Fujimoto, "Effect of Additives on Parial Oxidation of Methane," 1989 International Chemical Congress of Pacific Basin Societies Preprints of SB Symposium on Methane Activation, Conversion, and Utilization, Honolulu, Hawaii, (1989).
112
Geletii, Y. V., and A. E. Shilov, "Catalytic Oxidation Alkanes by Molecular Oxidation. Oxidation of Methane in the Presence of Platinum Salts and Heteropoly Acids." Kinet. Katal, 24(2), 486(1983).
Gray, B. F., J. F. Griffiths, G. A. Foulds, B. G. Charleton, and G. S. Walker, "The Relevance of Thermokinetic Interactions and Numerical Modeling to the Homogeneous Partial Oxidation of Methane," Ind. Eng. Chem. Res., 33, 1126(1994).
Hargreaves, J. S. J., and G. J. Hutchings, "Control of Product Selectivity in the Partial Oxidation of Methane," Nature 348, 428(1990).
Hellund, E. J., The Plasma State, Reinhold Publishing Corp., New York, (1961).
Helton, T. E., "Methanol and Carbon Monoxide Production from Natural Gas," Ph.D. Dissertation, Department of Chemical Engineering, Texas A&M University, Research Advisor: R. G. Anthony (1991).
Huang, G., M. V. Badani, S. L. Suib, J. B. Harrison, and M. Kablaoui, "Partial Oxidation of Methane to Methanol through Microwave Plasmas. Reactor Design to Control Free Radical Reactions," J Phys Chem., 98, 206(1994).
Huang, G. and S. L. Suib, "Dimerization of Methane through Microwave Plasmas," J. Phys C/ze;w., 97,9403(1993).
Hunter, N. R., H. D. Cesser, L. A. Morton, and P. S. Yarlagadda, D.P.C. Fung, "Methanol Formation at High Pressure by the Catalyzed Oxidation of Nattiral Gas and by the Sensitized Oxidation of Methane," Appl Catal, 57(1), 45(1990).
Kaliaguine, S. L., B. N. Shelimov, and V. B. Kazansky, "Reactions of Methane and Ethane with Hole Centers O"," J Catal, 55, 384(1978).
Kao, L. C , A. C. Houston, and A. Sen, "Low-TemperaUire, Palladium(II)-Catalyzed, Solution-Phase Oxidation of Methane to a Methanol Derivative," J Am. Chem. Soc, 113, 700(1991).
Konig, G., German OflFen, (German Patent Number) 3101024 (1982).
Krause T R and J. E. Heh, "Chemical Detoxification of Trichloroethylene and 1,1,1-Trichloroethane in a Microwave Discharge Plasma Reactor at Atmospheric Pressure," in Emerging Technologies in Hazardous Waste Management III, ACS Symp Series No 518, D. W. Tedder and F. G. Pohland Eds., p. 393, Washington, DC (1993).
113
LaDue, D. E., "Microwave-Induced Plasma Destmction of Trichlotoethylene," M. S Thesis Department of Chemical Engineering Texas Tech University, RQSQaich Advisor: U. Mann (1993).
Lee, C. C, "Plasma Systems," in Standard Handbook of Hazardous Waste Treatment and Disposal, H. M. Freeman Ed., p. 8.169, McGraw-Hill, New York (1988).
Liu, H. F., R. S. Liu, K. Y. Liew, R. E. Johnson, and J. H. Lungsford, "Partial Oxidation of Methane by Nitrous Oxides over Molybdenum on Silica," J Am Chem Soc 106,4121(1984).
Lodeng, R., O. A. Lindvag, P. Soraker, P. T. Rotemnd, and O. T. Onsager, "Experimental and Modeling Study of the Selective Homogeneous Gas Phase Oxidation of Methane to Methanol," Ind Eng Chem. Res, 34(4), 1044(1995).
Lunsford, J. H., "Catalytic Conversion of Methane to Methanol, Formaldehyde and Higher Hydrocarbons," J. Am. Chem. Soc Preprint, 33(3), 357(1988).
McHugh, M., and R. N. Occhiogrosso, "Critical Mixture Oxidation of Cumene," Chemical Engineering Science, 42, 10(1987).
NRC, National Research Council, "Plasma Processing of Materials; Scientific Opportunities and Technological Challenges," National Academy Press, Washington DC (1991).
Olah, G. A., G. Klopman, and R. H. Schlosberg, "Chemistry in Super Acids. III. '' Protonation of Alkanes and the Intermediacy of Alkanonium Ions, Pentacoordinated Carbon Cations of CHs Type. Hydrogen Exchange, Protolytic Cleavage, Hydrogen Abstraction, and Polycondensation of Methane, Ethane, 2.2-Dimethylpropane (Neopentane), and 2,2,3,3-Tetramethylbutane in FSOjH-SbFj ("Magic Acid") Solution'' ,"7. Am. Chem. Soc, 91, 3261(1969).
Onsager, O. T., P. Soraker, R. Lodeng, "Experimental Investigation and Computer Simulation of the Homogeneous Gas Phase Oxidation of Methane to Methanol," 1989 International Chemical Congress of Pacific Basin Societies Preprints of SB Symposium on Methane Activation, Conversion, and Utilization, Honolulu, Hawaii, (1989).
Oumghar, A., J. C. Legrand, A. M. Diamy, and N. Turillion, "Methane Conversion by an Air Microwave Plasma," P/a5/wfl[ C/2g/w. and Plasma Proc, 15(1), 87(1996)
114
Periana, R. A., D. H. Taube, E. R. Evitt, D. G. Loffler, P. R. Wentrcek, G. Voss, and T. Masuda, "A Novel, High Yield System for the Oxidation of Methane to Methanol," Natural Gas Conversion U, H. E. Curry-Hyde and R. F. Howe, Eds., Elsevier Science B. V., 533(1994).
Razier, Y. P., Gas Discharge Physics, pp. 4, 300, Springler-Verlag, New York, (1991).
Rytz, D. W., and A. Baiker, "Partial Oxidation of Methane to Methanol in a Flow Reactor at Elevated Pressure," Ind Eng Chem. Res, 30, 2287(1991).
Satterfield, C. N., Heterogeneous Catalysis in Practice, McGraw-Hill, New York, 1980.
Savage, P. E., R. Li, and J. T. Santini Jr., "Methane to Methanol in Supercritical Water," J Supercritical FI, 1, 135(1994).
Sazonov, B. A., and V. V. Popovskii, "Analytic Activity of Metal Oxides and the Energy of the Oxygen Bond," Kinet. Katal, 9(2), 312(1968).
Serafin, J. G., and C. M. Friend, "Evidence of Formation of Gaseous Methyl Radicals in the Decomposition of Methoxide on Oxygen-Precovered Mo(l 10)," J. Am. Chem. Soc, 111(24), 8967(1989).
Seyfert, W., Kinetischen Untersuchungen zur Methanolsythese im vrbesserte treibstahlreactor unter Lohen Drucken, T. H. Darmstadt, 1984.
Smith, J. M., and H. C. Van Ness, Introduction to Chemical Engineering Thermodynamics S^^ Edition, McGraw-Hill, Inc., New York, 1959.
Spencer, N. D., "Partial Oxidationof Methane to Formaldehyde by Means of Molecular Oxygen," J. Catal, 109, 187(1988).
Thomas, W. J., and S. Portalski, "Thermodynamics in Methanol Synthesis," Industrial and Engineering Chemistry, 50(6), 968(1958).
Vardanyan, I. A., and S. Yan, "Mechanism of the Thermal Oxidation of Methane," Kinet Katal, 22(4), S45(\9S\).
Walker, G. S., J. A. Lapszewicz, and G. A. Foulds, " Partial Oxidation of Methane to Methanol-Comparison of Heterogeneous catalyst and Homogeneous Gas Phase Reactions," Catalysis Today, 21, 519(1994).
115
War Department Technical Manual TM 11-1524, "Radio Sets SCR-584-A and SCR-584-B Service Manual: Theory, Trouble Shooting, and Repair," United States Government Printing Office, Washington D. C, 1946.
Webley, P. A., and J. W. Tester, "Fundamental Kinetics of Methane Oxidation in Supercritical Water," Energy & Fuels, 5, 411(1991).
Yarlagadda, P. S., L. A. Morton, N. R. Hunter, and H. D. Cesser, "Direct Conversion of Methane to Methanol in a Flow Reactor," Ind. Eng. Chem. Res, 27, 256(1988).
Zanetti, R. J., "Plasma: Warming Up to New CPI Applications," Chemical Engineering, Dec. 26, 1983.
Zhen, K. J., M. M. Khan, C. H. Mak, K. B. Lewis, and G. A. Somorjai, "Parital Oxidation of Methane with Nitrous Oxide over V205-Si02 Catalysts," J. Catal., 94, ^ 501(1985).
16
APPENDIX A
CALIBRATION OF EXPERIMENTAL
EQUIPMENT AND INFORMATION ON SYSTEM HARDWARE
The apparatus consisted of many different mechanical devices and analytical
instmments which required calibration for reliable operation. This consists primarily of
rotameters that controlled the flow rate of reactant stream and GC calibration.
Calibration of the GC for analysis of the gas product took place before each experimental
mn and details are presented in Appendix B, which consists of sample calculations.
Specific equipment and supply information is also presented in this section.
A. 1 Calibration of Gas Rotameters
Three gas rotameters were used to control the flow of reactant gases (argon,
methane, and oxygen). Methane and oxygen flows were controlled by Brooks rotameters
(Brooks R-2-15-AAA, and Brooks Flow Controller 8744A). Argon was controlled by
Brooks rotameter (Brooks 2-65). Back pressure on the rotameter was set at the regulators
on the cylinders to approximately 40 psi. Downstream of the rotameters, pressure was
maintained constant at 938.6 torr. This level was selected to me approximately 5 psi
above normal local atmospheric pressure. The temperature was not controlled in the vent
hood, but did not ever vary by more than 1.5° from 22°C. Several flow measurements
(50) were collected with a bubblemeter and stopwatch for each flow rate represented on
each calibration plot. Volumetric flows were converted to molar flows assuming the
Ideal Gas Law applies. The average of the flows was plotted and a trendline applied.
117
Calibration curves for argon, methane, and oxygen are shown in Figures A.l, A.2, and
A.3.
Calibration for methanol product was obtained by preparation of standard
solutions for multiple injection into the GC. As expected, large variations were observed.
It was decided to make a large number of injections to minimize instmment error. Four
different concentrations were tested. The calibration for methanol is shown in Figure
A.4.
As mentioned above, calibration for the gas samples occurred before every
experiment. Two different calibration gases (Scott Specialty Gases Mix 234 and Mix
216) were injected prior to experimental runs. The output from these calibration runs
was used to analyze the data for that particular experiment. Frequent changing of the GC
columns and re-conditioning necessitated this daily re-calibration. This method should
have minimized errors. The concentration of components present in the reactor effluent
gas stream was calculated using the output from the daily calibration injections. Details
of how this was accomplished are presented in the sample calculation section (Appendix
B).
118
40 1
o o o X! C
I J3
93
a
Flowmeter reading
Figure A. 1. Argon Rotameter Calibration Plot (back pressure of 40 psig, downstream pressure of 5 psig, and 22°C).
119
Flowmeter readmg
Figure A.2. Methane Rotameter Calibration Plot (back pressure of 40 psig, downstream pressure of 5 psig, and 22°C).
120
o o o X
c
I 93
30 40 50 60
Flowmeter Reading
Figure A.3. Oxygen Rotameter Calibration Plot (back pressure of 40 psig, downstream pressure of 5 psig, and 22°C).
121
10000000
1000000
y)
o
93
100000
10000
1000
100
10 100 1000
Concentration (ppm)
10000
1
1
1-i 1
- 1 h-
1
—
i i 1 ; i ! 1
i
in42^\
1 !
! u
l/r ! i' B 1—^—
I
; 1
I !
1 \ \ Jr
1 i /
1
1 i
\X\ Jr' 1
1 1—1-1 1 1
1
i !
1
-U— — h -! !
i 1 1
1 1 1
1 I ' i Mil 1 J I I I ' '
<1 T 1 I ' l l ! \^^ • M I i
1 j II
1 1 1
: ; 1 , i
1 i 1
i 1 j 1
i 1 1 j
1
1 1
i
- 1 1 1 ' 1
1 i l l 1 i : 1 1 ' ' 1
1 " 100000
Figure A.4. Methanol Calibration Plot (multiple injections at four different concentrations).
122
APPENDIX B
SAMPLE CALCULATIONS
This section describes how quantities were calculated from the measured
variables. All of the experimental data that was measured during each experimental mn
is given in Appendix C. All of the calculations that are required to obtain the methane
conversion and methanol selectivity will be outlined. These two quantities are the most
important for this research.
B. 1 Molar Reactor Feed Rates
The total molar feed rate is obtained by simply summing the molar feed rates of
all components. Streams are assumed to behave as ideal gases throughout the analysis.
total moles into reactor time
f moles oxygen 1 +
moles argon
time
moles water 1
r + S moles methane 1
time J (B.l)
[ time J 1 time J
The amount of water being evaporated into the feed stream deserves special
attention. Multiple experiments were conducted without plasma to discover the amount
of water that was being evaporated into the system. Early estimations were based on
assuming that feed gas leaving the water sattiration unit was sattirated with water the
temperature in the headspace of the unit. Experiments revealed that gas leaving the unit
was, on the average 90% saturated. It was assumed that this was due to the relatively
short exposure of the stream to the hot liquid. The molar flow rate of the water into the
123
reactor was calculated from the molar flow rate of the plasma stream that was passing
through the saturation unit and the temperature of the headspace above the unit. Since the
mole fraction of water in the stream is known (because of our assumption of 90% stream
saturation at the headspace temperature), a simple mass balance is used to calculate the
amount of water fed to the reactor. The results of that mass balance are below. Let Y be
the mole fraction of water in the stream,
moles water [ moles gas through unit
1 X S time Ume 1(1-Y) J (B.2)
B.2 Reactor Effluent
The amount of material leaving the reactor was calculated from the two effluent
streams, gas and liquid. The volumetric flowrate of gas leaving the reactor was measured
with a bubblemeter and stopwatch, as described in Chapter V. The ideal gas law was
used to obtain the molar flowrate of gas. This effluent molar flowrate is vital to the
results of this study.
moles gas out = i
time
volume of effluent
time ^ X S
atmospheric pressure R X Temperature
(B.3)
where R is the gas constant.
The liquid product flowrate is simply measured at the end of the experimem and
is applied to the entire duration of the mn. The liquid product consists of many products,
but is essentially water. All non-water components added together do not appmach 1%.
124
It is important to maintain units in moles because this is how the analysis system
measures concentration. Both detectors measure a response that is based on molar
amount (i.e., a flame ionization detector or a thermal conductivity detector will not
indicate triple the response for a compound that is three times heavier.)
moles water out f mass collected _ J s X <
time I time J [molecular weight
1 (B.4)
B.3 Calculation of Effluent Component Concentrations
The GC analysis of the liquid and gas products creates a response for each
detected component. The primary components of interest are CO, CO2, and methanol.
These three components contain most of the carbon that is converted in the reactor. Since
this system is unable to track oxygen, analysis of carbon-containing products must
provide us with the required information.
The carbon-containing components present in the gas phase (primarily CO, CO ,
and methane) produce a response from the GC. This response is compared to the
response from the known concentration of the standard calibration gas injected prior to
the experimental mn. Componem concentration (cone.) was calculated like this.
. f cone, (ppm) of standard | ^^^^ cone, (ppm) = {component area counts) x | j . ^ ^ ^rea counts of standard J
Liquid concentrations of methanol and other products were calculated in a similar
mamier. Other than methanol, there were few products in the liquid phase. Traces of
fomiic and acetic acid were detected and quantified to the highest extent possible.
125
standard solutions of these components were mixed and injected in the GC. Liquid
concentration was calculated in the following way.
liquid cone, (ppm) = {area coums} x j cone, of standard 1 [ area counts of standard J ^ ^^
Knowing the total molar flowrate of the liquid stream and the concentration of the
components in the stream, the molar flowrate of the componems is obtained.
moles of component /total molar flowrate! tJm^ ~ component cone, (ppm) x < ^""^ [ time (B.7)
This consideration does not take into account the contribution of the non-water
constituents in the liquid. These components will not significantly affect the calculation
of the total molar flows since their concentration is so small (<1000 ppm or 0.1 molar %).
B.4 Material Balances
Since all of the flowrate data is known, it is a simple matter to compare the
amounts entering and exiting the system. To compute a carbon balance, the amount
entering was calculated from the rotameter reading and the known feed composition. The
amount leaving was calculated from the effluent flowrate and the concentration. The
amount entering was compared to the amount exiting, revealing the mass balance closure.
Carbon balances were closed to within ±10% for more than 80% of the experimental
runs. Only 7 of the 36 experimental mns exceeded ±10% error in the carbon balance.
Only 5 of the 36 runs exceeded ±10% error in the hydrogen balance. One run exceeded
126
±20% in carbon balance error (20.4%) and no mns exceeded ±20% error in the hydrogen
balance error.
B.5 Calculation of the Conversion and Selectivity
These are the two most important quantities for this research. The high degree of
variation in the GC analysis made it impossible to use differences in methane
concentration as a basis for conversion. With conversion levels in the 1% regime and low
GC reproducibility, small errors in methane concentration measurement (±5%) can easily
obscure the results. It was necessary to use the observed carbon-containing products as
the basis for conversion. Methane conversion was calculated by
S all carbon - containing products not present in feed
conversion = i time
methane fed to reactor/ /time
(B.S)
The calculation for methanol selectivity was dependent of this definition of
conversion that was used out of necessity. In short, methanol selectivity is an expression
for the amount of methane that reacts to produce methanol divided by the total amount of
methane that reacts. It is defined in this research as
methanol produced
selectivity = time
I all carbon containing products not present in feed 'time J
(B.9)
These two quantities contain most of the valuable information about the
performance of this system and are the focal points of the analysis and the performance.
127
APPENDIX C
RAW EXPERIMENTAL DATA
This section contain all of the information that was recorder during the
experimental mns. All of the data used in the analysis of this research was generated
from the information in this section.
128
Table C.l. Preliminary Phase and Phase I Raw Experimental Data