Post on 24-Jun-2020
DEVELOPMENT OF ROBUST TDLAS SENSORS FOR
COMBUSTION PRODUCTS AT HIGH PRESSURE AND
TEMPERATURE IN ENERGY SYSTEMS
A DISSERTATION
SUBMITTED TO THE DEPARTMENT OF MECHANICAL ENGINEERING
AND THE COMMITTEE ON GRADUATE STUDIES
OF STANFORD UNIVERSITY
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
Ritobrata Sur
September 2014
http://creativecommons.org/licenses/by-nc/3.0/us/
This dissertation is online at: http://purl.stanford.edu/bq161vy0151
© 2014 by Ritobrata Sur. All Rights Reserved.
Re-distributed by Stanford University under license with the author.
This work is licensed under a Creative Commons Attribution-Noncommercial 3.0 United States License.
ii
I certify that I have read this dissertation and that, in my opinion, it is fully adequatein scope and quality as a dissertation for the degree of Doctor of Philosophy.
Ronald Hanson, Primary Adviser
I certify that I have read this dissertation and that, in my opinion, it is fully adequatein scope and quality as a dissertation for the degree of Doctor of Philosophy.
Chris Edwards
I certify that I have read this dissertation and that, in my opinion, it is fully adequatein scope and quality as a dissertation for the degree of Doctor of Philosophy.
Jay Jeffries
Approved for the Stanford University Committee on Graduate Studies.
Patricia J. Gumport, Vice Provost for Graduate Education
This signature page was generated electronically upon submission of this dissertation in electronic format. An original signed hard copy of the signature page is on file inUniversity Archives.
iii
iv
v
Abstract
Energy extraction by combustion of fossil fuels leads to the release of greenhouse gases
and other harmful emissions. It has been a priority for combustion research in the recent
few decades to mitigate these emissions and harness the energy contained in these fuels
more efficiently. Coal is among the most abundant and widely distributed sources of
fossil fuel in the world. Integrated gasification combined cycle (IGCC) is one of the
cleanest methods of extracting energy from coal, when combined with carbon capture
and storage. The gasifier, the cornerstone of this technology, produces synthesis gas (also
known as syngas) via partial oxidation of coal. Continuous monitoring of the syngas
composition is imperative to the success of this technology as it indicates the extent of
reaction in the gasifier, the heating value of the output syngas and hence the overall
health of the gasification system. The primary objective of the research presented in this
dissertation is the development of robust, in-situ sensors that can reliably monitor
concentrations of CO, CO2, CH4 and H2O in gasifiers.
Tunable diode laser absorption spectroscopy (TDLAS) sensors for detection of CO, CO2,
CH4 and H2O at elevated pressures in mixtures of syngas were developed and tested.
Wavelength modulation spectroscopy (WMS) with 1f-normalized 2f detection was
employed. Fiber-coupled DFB diode lasers operating at 2325, 2017, 2290 and 1352 nm
were used for simultaneously measuring CO, CO2, CH4 and H2O, respectively. Criteria
for the selection of transitions were developed, and transitions were selected to optimize
the signal and minimize interference from other species. To enable quantitative WMS
measurements, the collision-broadening coefficients of the selected transitions were
determined for collisions with possible syngas components, namely CO, CO2, CH4, H2O,
N2 and H2. Sample measurements were performed for each species in gas cells at a
vi
temperature of 25 °C and pressures up to 20 atm. To validate the sensor performance, the
composition of synthetic syngas was determined by the absorption sensor and compared
to the known values. A method of estimating the lower heating value and Wobbe index of
the syngas mixture from these measurements was also demonstrated.
The sensors demonstrated in a sample cell were then deployed in a pilot-scale (1
ton/day), high-pressure (up to 18 atm), entrained-flow, oxygen-blown, slagging coal
gasifier at the University of Utah. Measurements of species mole fraction with 3-second
time resolution were taken at the pre- and post-filtration stages of the gasifier synthesis
gas output flow. Although particulate scattering makes pre-filter measurements more
difficult, this location avoids the time delay of flow through the filtration devices. With
the measured species and known N2 concentrations, the H2 content was obtained via
balance. The lower heating value and the Wobbe index of the gas mixture were estimated
using the measured gas composition. The sensors demonstrated here show promise for
monitoring and control of the gasification process.
The sensors were further improved using a scanned-wavelength modulation spectroscopy
technique and was demonstrated for the first time in the product stream of an
engineering-scale (50 tons/day coal throughput) transport reactor gasifier (15 ton/hr
syngas production). A robust optical design was created to counter various challenges
including beam steering, loss in beam intensity due to particulate scattering and wide
dynamic range in transmission in a typical gasifier environment. In addition, due to the
unavailability of low-loss, high-strength fibers and combiners at the wide range of
operating wavelengths, an extensive optical design was required for enabling such a
group of sensors to operate simultaneously. The results from the measurements during
the gasification process, starting from the propane heat-up till the full-scale gasification
process, reveals interesting dynamic behavior not observable by extractive measurement
techniques. These sensors show high-bandwidth detection in a gasifier and thereby
eliminate the need for surrogate indicators that can monitor the transient performance of
the gasifier.
Methane, a more potent greenhouse gas than CO2, is often produced as an intermediate
product of hydrocarbon combustion processes. Hence, a more sensitive CW laser
vii
absorption diagnostic (50 times stronger than the fiber-coupled CH4 sensor at 2290 nm,
which was designed for detection of higher concentrations at high pressure) for in-situ
measurement of methane mole fraction at high temperatures was also developed. The
selected transitions for the diagnostic are a cluster of lines near 3148.8 cm-1 from the R-
branch of the ν3 band of the CH4 absorption spectrum. The selected transitions have 2-3
times more sensitivity to CH4 concentration than the P-branch in the 3.3 micron region,
lower interference from major interfering intermediate species in most hydrocarbon
reactions, and applicability over a wide range of pressures and temperatures. Absorption
cross-sections for a broad collection of hydrocarbons were simulated to evaluate
interference absorption, and were generally found to be negligible near 3148.8 cm-1.
However, minor interferences from hot bands of C2H2 and C2H4 were observed and
characterized experimentally, revealing a weak dependence on wavelength. To eliminate
such interferences, a two-color on-line and off-line measurement scheme was proposed to
determine CH4 concentration. The colors selected, i.e. for on-line (3148.81 cm-1) and off-
line (3148.66 cm-1), were characterized between 0.2 to 4 atm and 500 K to 2100 K by
absorption coefficient measurements in a shock tube. Minimum detectable levels of CH4
in shock tube experiments were reported for this range of temperatures and pressures. An
example measurement was shown for sensitive detection of CH4 in a shock tube chemical
kinetics experiment.
viii
Acknowledgement
I would like to express my gratitude towards my advisor, Professor Ronald K. Hanson,
for his support, guidance and motivation over the last five years of my life. He taught me
the virtues that make a true scholar – dedication, attention to detail and the importance of
asking the right questions. I admire his perseverance to advance the cause of science for
over four decades.
I would like to thank Dr. Jay B. Jeffries for his mentorship and clairvoyance in difficult
and strenuous situations such as field trips. I was fortunate to tap into his expertise on
lasers, optics and spectroscopy. I am grateful to Dr. David Davidson for his support
throughout my PhD career especially regarding the shock tube and other laboratory
facilities.
I am thankful to Professor Christopher F. Edwards for everything from ME370B, to
participating on my reading committee and providing valuable advice on how to be an
effective communicator. I would also like to thank Professor Cappelli for teaching me
Physical Gas Dynamics and participating in my oral examination committee. I also want
to express my gratitude towards Professor Franklin M. Orr, Jr. for participating in my
oral examination committee as the chair.
My journey from the suburbs of Calcutta to the commencement stage at Stanford
University was a long and arduous one. But, I was truly blessed with the company of
numerous friends, teachers and wise strangers who I have met along the way. I would
like to thank them all. Stanford is an extraordinary place and I am fortunate to spend five
years of my life studying here. My Stanford experience consists of a blend of world class
courses, motivated colleagues, state of the art research facilities and engaging workshops
and clubs which makes it, deservedly, one of the best academic institutions in the world.
ix
I must convey my deepest gratitude to my fellow researchers at the Hanson group for the
countless times they have rescued me out of a difficult situation. In particular, I would
like to thank Kai Sun, who has been with me in nearly all the field trips and I am blessed
to find a partner so dedicated and passionate about his research. I would also like to thank
Shengkai Wang whose prodigious talent was instrumental in the development of the
methane diagnostics. In addition, I would like to thank Zekai Hong, Xing Chao, Wei Ren,
Mitchell Spearrin, Matthew Campbell, Sijie Li, Brian Lam, Christopher Goldenstein,
Christopher Strand, Ian Schultz, Sean Gates, Ivo Stranic, Marcel Nations Martin, Yangye
Zhu, Tom Parise and Luke Zaczek. I greatly appreciate the help that I have received on
the productive field campaigns from Professor Kevin Whitty, Randy Pummill, Dave
Wagner, Dave Wagner Jr., Travis Waind, Scott Machovec, Justin Anthony, Tommy
Clarke, John Northington and John Socha. I am also grateful to my friends outside the
Hanson group, in particular – Manuel Lopez, Matthew Hoffman, Sayak Banerjee,
Shiladitya Mukherjee, Surya Deb, Arnab Roy, Partha Saha, Jon Connolly and Arnab
Mukherjee who helped me keep my sanity in check over these years.
Finally, I would like to recognize that without the support of my family, it would be
impossible for me to reach this milestone. I derive a lot of my interest in upholding the
pillars of science from my great-uncle, Gopinath Sur, who was an amazing scholar but
lost his vision in his later days. My parents, Dr. Abhi Sur and Mrs. Poly Sur, and my
sister, Somolekha Sur, have been a constant in my life, providing comfort, direction and
meaning to my existence. I am also quite fortunate to have my uncle, Dr. Biswajit Sur
and his family live in the Bay Area, because without them, my efforts at getting a PhD
might have abruptly ended four years ago.
x
Table of Contents
Abstract .......................................................................................................... v
Acknowledgement ...................................................................................... viii
List of Tables .............................................................................................. xiii
List of Figures ............................................................................................. xiv
Chapter 1 Introduction ................................................................................. 1
1.1 Background and Motivation ...................................................................................1
1.2 Overview of dissertation ..........................................................................................3
Chapter 2 Development of multi-species laser absorption sensors for in-situ monitoring of syngas composition ........................................................ 4
2.1 Introduction ..............................................................................................................4
2.2 Sensor principle .......................................................................................................5
2.3 Laser wavelength modulation characterization and optical system intensity modulation characterization .........................................................................................8
2.3.1 Laser wavelength modulation characterization..............................................8
2.3.2 Optical system intensity modulation ................................................................9
2.4 Effects of collision broadening on modulation optimization ..............................10
2.5 Selection of transitions ...........................................................................................15
2.5.1 Carbon monoxide ............................................................................................15
2.5.2 Carbon dioxide ................................................................................................16
2.5.3 Methane ...........................................................................................................17
2.5.4 Water ...............................................................................................................18
2.6 Measurement of spectral line parameters ............................................................19
2.7 Sample WMS measurements of species in N2 at elevated pressure ...................23
2.8 Sample WMS measurements of species in syngas mixture at different pressures ........................................................................................................................24
2.8.1 Carbon monoxide ...........................................................................................24
2.8.2 Carbon dioxide ................................................................................................25
2.8.3 Methane ...........................................................................................................26
2.9 Summary of the laboratory validation experiments ...........................................27
xi
2.10 Calculation of LHV and Wobbe Index of syngas mixture ...............................29
2.11 Conclusions ...........................................................................................................31
Chapter 3 Application of TDLAS-based sensors for in-situ measurement of syngas composition in a pressurized, oxygen-blown, entrained-flow coal gasifier ......................................................................... 32
3.1 Introduction ............................................................................................................32
3.2 Apparatus ................................................................................................................32
3.2.1 Entrained-flow gasifier and sampling locations ..........................................34
3.2.2 System operation.............................................................................................38
3.2.3 Lasers and control system ...............................................................................39
3.2.4 Free space beam multiplexing ........................................................................40
3.2.5. Optical access to the syngas ...........................................................................41
3.3 Results and Discussion ...........................................................................................43
3.3.1 Field validation of sensor performance .........................................................43
3.3.2 Simultaneous time-resolved multi-species concentration measurements ...44
3.3.3 Estimation of lower heating value and Wobbe index ...................................47
3.4 Conclusions .............................................................................................................51
Chapter 4 Application of scanned wavelength modulation spectroscopy sensor for simultaneous measurement of CO, CO2, CH4 and H2O in a high-pressure syngas output stream from an engineering-scale transport reactor gasifier ............................................................................ 53
4.1 Introduction ............................................................................................................53
4.2 Sensor apparatus ....................................................................................................56
4.2.1 Transport Reactor Gasifier at Wilsonville, Alabama ..................................56
4.2.2 Multi-laser beam multiplexing over 5 m path...............................................60
4.3 Results .....................................................................................................................62
4.3.1 Propane burner heat-up ..................................................................................62
4.3.2 Pulsed coal addition during heat-up ..............................................................63
4.3.3 Failed attempts at gasification ........................................................................65
4.3.4 Onset of final gasification phase .....................................................................66
4.3.5 Pressurization and gasifier stabilization phase .............................................66
4.3.6 Steady-state conditions ....................................................................................67
xii
4.3.7 Gasifier shutdown ............................................................................................69
4.4 Conclusions .............................................................................................................70
Chapter 5 Development of high-sensitivity interference-free diagnostic for measurement of methane in shock tubes ............................................ 71
5.1 Introduction ............................................................................................................71
5.2 Sensor design and selection of CH4 transitions ...................................................72
5.3 Measurement of absorption coefficient in Argon ................................................76
5.3.1 Experimental set up .........................................................................................76
5.3.2 Measured absorption coefficients...................................................................78
5.4 C3H8 pyrolysis: A demonstration of the method .................................................82
5.5. Conclusions ............................................................................................................84
Chapter 6 Summary and future opportunities ........................................ 85
6.1 Summary .................................................................................................................85
6.2 Future opportunities ..............................................................................................87
6.2.1 Additional minor species measurements in the fluidized-bed coal gasifier in NCCC ....................................................................................................................87
6.2.2 Extension of the methane sensor for higher pressure measurements in shock tubes ................................................................................................................87
6.2.3 Higher sensitivity measurement of CH4 species time-history in shock tubes using cavity enhanced absorption spectroscopy (CEAS) / WMS .........................87
Appendix ...................................................................................................... 88
A.1 Description of PSDF, National Carbon Capture Center (NCCC) gasifier ......88
A.1.1 KBR Transport reactor gasifier ....................................................................89
A.1.2 Particulate control device (PCD) ...................................................................91
A.2 Window assembly design drawings for optical access in NCCC, Alabama .....93
A.2.1 Parts list (All parts SS316) .............................................................................93
A.3 Components of Optical heads ...............................................................................98
A.4 Fiber optics coupling assembly for CH4 sensor .................................................99
A.5 List of studied species with absorption coefficients found to be less than 5% that of CH4 at 3148.81 cm-1 ........................................................................................100
References .................................................................................................. 101
xiii
List of Tables
Table 2.1 Parameters for simulating the scanned-WMS lineshape .................................. 11
Table 2.2 Selected Modulation depth for the sensors ....................................................... 14
Table 2.3 Measured spectroscopic parameters of the selected transitions ........................ 22
Table 2.4 Percentage compositions of components in the mixtures used for the sensor
validation experiments ........................................................................................ 29
Table 3.1 Gasifier specifications....................................................................................... 35
Table 4.1 Operating conditions for the scanned WMS sensors ........................................ 54
Table 4.2 Typical conditions at the gasifier exhaust ......................................................... 56
xiv
List of Figures
Figure 1.1 A typical integrated gasification combined cycle process diagram [4] ............. 1
Figure 2.1 Example of laser wavelength response to the injection-current modulation at 1
kHz. The blue circles are the relative frequency measured by a solid etalon with
0.02cm-1 FSR. The measured modulation depth is 0.101cm-1 and initial phase of
the wavelength modulation is -2.1363 radian. The red line represents the best-fit
sine function which is used to model the variation of the laser frequency tuning
due to modulation. ................................................................................................ 9
Figure 2.2 Transmission of a thin cavity built by two parallel surfaces (cavity length:
2mm, surface reflectivity: 2%) ........................................................................... 10
Figure 2.3 (a) Simulated absorbance spectrum of the fictitious molecule characterized by
a single transition of CO2 near 4957 cm-1 at 5 and 12 atm; (b) 1f-normalized
WMS-2f spectrum of the fictitious molecule near 4957 cm-1 at 5 and 12 atm. . 12
Figure 2.4 WMS X2f components of fictitious molecule characterized by a single
transition of CO2 near 4957 cm-1 at (a) 5 and (b) 12 atm. .................................. 12
Figure 2.5 Spectral lineshape blending at elevated pressures for CO2 in the P-branch of
20012 ← 00001 band. Numbers in the plot indicate the rotational quantum
number of the lower energy state. ...................................................................... 13
Figure 2.6 Variation of the modulation index for the largest WMS-2f signal due to the
blended absorbance profile of CO2 near 4957 cm-1 at 296K. ............................. 13
Figure 2.7 Variation of 1f normalized WMS-2f magnitude at the selected peak
wavelength with modulation depth at 10 kHz modulation frequency for the laser
operating at 2017 nm for CO2 detection at different pressures. ......................... 14
Figure 2.8 Overview of the absorption spectra of CO, CO2, CH4 and H2O at 15 atm and
350 K .................................................................................................................. 15
xv
Figure 2.9 (a) Absorbance spectrum of CO near 4301 cm-1; (b) 1f-normalized WMS-2f
spectrum of water near 4301 cm-1. ..................................................................... 16
Figure 2.10 (a) Absorbance spectrum of CO2 near 4957 cm-1; (b) 1f-normalized WMS-2f
spectrum of CO2 near 4957 cm-1. ....................................................................... 17
Figure 2.11 (a) Absorbance spectrum of CH4 near 4367 cm-1; (b) 1f-normalized WMS-2f
spectrum of CH4 near 4367 cm-1. ....................................................................... 18
Figure 2.12 Absorbance spectrum of CH4 near 4367 cm-1 showing the presence of
interfering species near the CH4 absorption band used. ..................................... 18
Figure 2.13 (a) Absorbance spectrum of water near 7394 cm-1; (b) 1f-normalized WMS-
2f spectrum of water near 7394 cm-1. ................................................................. 19
Figure 2.14 (a) Sample direct absorbance spectra of CO2 showing agreement with a Voigt
fit; (b) Measured collisional FWHM for 5% CO2 in H2 at different pressures. . 20
Figure 2.15 Sample WMS spectra for (a) CH4, (b) CO and (c) CO2 in N2 at 10 atm. ...... 23
Figure 2.16 Sample WMS spectra for CO in a sample syngas mixture at (a) 5 atm, (b) 10
atm, (c) 15 atm and (d) 20 atm. .......................................................................... 25
Figure 2.17 Sample WMS spectra for CO2 in a sample syngas mixture at (a) 5 atm, (b) 10
atm, (c) 15 atm and (d) 20 atm. .......................................................................... 26
Figure 2.18 Sample WMS spectra for CH4 in a sample syngas mixture at (a) 5 atm, (b) 10
atm, (c) 15 atm and (d) 20 atm. .......................................................................... 27
Figure 2.19 Comparison of the known and the measured values of (a) CH4, (b) CO and
(c) CO2 mole fractions in various syngas mixtures. ........................................... 28
Figure 2.20 Comparison of known and measured (a) LHV in MJ/kg C; (b) Wobbe index
in MJ/Nm3. ......................................................................................................... 31
Figure 3.1 University of Utah Gasification Research Facility. ......................................... 33
Figure 3.2 Schematic of entrained-flow gasification research facility. ............................ 34
Figure 3.3 Schematic to the pilot-scale, entrained-flow coal-gasifier at University of Utah
............................................................................................................................ 35
Figure 3.4 Schematic of the entrained flow gasifier at the Institute for Clean and Secure
Energy at the University of Utah. The locations 1-4 are identified in the
diagram. The sections are shown by green arrows. ............................................ 37
xvi
Figure 3.5 Schematic to measurement locations at University of Utah: (1) the reactor-
core, (2) pre-quench, (3) post-quench, (4) after clean-up ................................... 38
Figure 3.6 Schematic illustrating the components of the sensor including the controller
and the DAQ in relation to the syngas exhaust pipe .......................................... 40
Figure 3.7 Free-space quadruple beam multiplexing on a single detector ....................... 41
Figure 3.8 Photograph of the optical access assembly in Location 3 ............................... 42
Figure 3.9 (a) Schematic and (b) Photograph of the optical access assembly in Location
4; time delay between the shorter (CO and H2O) and the longer (CH4 and CO2)
path was less than 0.1 seconds. .......................................................................... 42
Figure 3.10 Sample WMS spectra for (a) CO, (b) CO2, (c) CH4 and (d) H2O measured in
the gasifier at 11 atm. ......................................................................................... 44
Figure 3.11 Time lag observed for traditional sensors vs. laser based in-situ sensors at (a)
Location 3 and (b) Location 4. ........................................................................... 46
Figure 3.12 Time-resolved (dry basis except H2O) mole fractions of (a) CH4, (b) CO, (c)
CO2 and (d) H2O and comparison with GC measurements (except H2O) in
Location 3 at about 11 atm. ................................................................................ 47
Figure 3.13 Multispecies measurements at Location 3 at 11 atm; Box 1: varying CH4
addition, Box 2: varying coal slurry / O2 feed ratios. ......................................... 48
Figure 3.14 LHV of the syngas; Box 1: varying CH4 addition, Box 2: varying coal slurry /
O2 feed ratios. ..................................................................................................... 49
Figure 3.15(a) Variation of the measured CO and CO2 mole fractions and (b) Variation of
the inferred LHV with trends in oxygen - coal slurry feed ratio [kg/m3] and (c)
Variation of the inferred Wobbe index with trends in oxygen - coal slurry feed
ratio [kg/m3]. ....................................................................................................... 50
Figure 3.16(a) Multispecies measurements at Location 4 at 11 atm, and (b) Measured
LHV and Wobbe index of the syngas. ................................................................ 51
Figure 4.1 Illustration of the time demultiplexed collection of multi-laser signal (not
actual data) .......................................................................................................... 55
Figure 4.2 Sample fit of the scanned 1f-normalized WMS-2f spectrum for 8% CO2 at 8
atm and 580K during coal feed .......................................................................... 56
Figure 4.3 Location of the optical access path in relation to the Gasifier ........................ 58
xvii
Figure 4.4 Schematic of the sensor setup in the syngas output flow channel. .................. 59
Figure 4.5 Nitrogen purged multiple laser enclosure. The cables on the right connect the
lasers to the controller. The conduit on the left convey the 2 m optical fibers and
the DB-25 cable for controlling the control motor. ............................................ 59
Figure 4.6 Multi-beam multiplexing hardware used for the gasifier sensors ................... 61
Figure 4.7 Variation of measured signal strength with gasifier operating pressure ......... 61
Figure 4.8 Simultaneous multi-species measurements during propane burner heat-up
phase at 60 psig. The propane burner is ignited at time 55.5 hrs, and peaks in the
CO and CH4 are observed at 69.9, 74.7, and 75.1 hrs when the fuel flow is
increased. ............................................................................................................ 63
Figure 4.9 Pulsed fluctuations in CO, CO2 and H2O levels during initial parts of the coal-
fed heat-up phase. ............................................................................................... 64
Figure 4.10 Pulsed fluctuations in species concentration observed at a later time during
the coal-fed heat-up phase. ................................................................................. 64
Figure 4.11 Failed attempts at gasification due to coal feed problems as captured through
the multi-species TDLAS measurements. The local gas conditions are 122 psig
and 600K. ........................................................................................................... 65
Figure 4.12 A period of gasifier run showing three important sections of the gasifier run
............................................................................................................................ 66
Figure 4.13 Stepped increasing pattern in CO and CH4 levels during gasifier
stabilization/pressurization process. ................................................................... 67
Figure 4.14 Oscillatory behavior of CO and CH4 mole fractions in the early phase of
steady state operations ........................................................................................ 68
Figure 4.15 Oscillatory behavior of measured species showing correlation of CO2 and
H2O with measured temperature fluctuations and an anti-correlation with CO
and CH4 measurements ....................................................................................... 68
Figure 4.16 Measurement of multi-species mole fractions from four days before the
shutdown until the end. Operating conditions: 220 psig, 630 K. At about 898
hrs, the GC sampling line was blocked and maintenance to clear the line
produced a fast change in the GC reading at that time. The gasifier feed was
xviii
unstable due to blockage in the coal feed line resulting in sharp changes in all
the species concentrations at around 879 and 928 hrs. ....................................... 69
Figure 5.1 Absorption coefficient simulations of CH4 based on HITRAN 2012 at 1 atm
and 900 K and 1400 K. The red arrow indicates the selected cluster of
transitions. The hot bands, which are more pronounced at 1400 K, are marked
by the orange circle. ........................................................................................... 73
Figure 5.2 (a) Linestrengths of common combustion intermediates from HITRAN 2012
near selected CH4 transitions, (b) Fourier transform infrared spectroscopy survey
of some common combustion species at 773K and 1 atm (c) Expanded view of
absorption coefficent of some species with 1-3% interference; all plots are for
773K and 1 atm. ................................................................................................. 74
Figure 5.3 Absorbances of CH4 and possible interfering species as measured by scanned
direct absorption method for a path length of 14.13 cm and location of the on-
line and off-line shocks in wavenumber. ............................................................ 76
Figure 5.4 Shock tube apparatus showing driver and driven sections (top), launch of the
incident shock wave (middle), and reflection of the shock wave from the
endwall. .............................................................................................................. 77
Figure 5.5 Sample time-resolved trace of measured voltage signal during the absorption
coefficient measurements before and after a shock for 1% CH4 in Ar mixture. 79
Figure 5.6 (a) Surface plot of absorption coefficient of CH4 in Ar at 3148.81 cm-1 (νon).
The dots represent the measured data points. (b) Surface plot of absorption
coefficient of CH4 in Ar at 3148.66 cm-1 (νoff), (c) Deviation of measurements
and the fitted equation for kon at different temperatures and pressures. All
samples lie within 5% of the fitted equation. (d) Deviation of measurements and
the fitted equation for koff at different temperatures and pressures as a percentage
of differential absorbance. All measurements lie within 2.5%. ......................... 80
Figure 5.7 Location of the on-line and off-line measurement frequencies optimized for
CH4 detection at 0.2 - 1.5 atm with respect to high pressure (15 atm) CH4
spectrum. ............................................................................................................ 81
Figure 5.8 Minimum detectivity of CH4 under different pressure and temperature
conditions at 0.001 absorbance noise (SNR = 1) over a 14.13 cm path ............. 82
xix
Figure 5.9 (a) Absorbance time history [on-line (αon) and off-line (αoff)], (b) CH4 time
history of 1% C3H8 pyrolysis in Ar at 1763K, 1.64 atm obtained using the two-
color scheme, and (c) Measured residual interference absorbance and
comparison with simulated absorbance from LLNL and USC Mech II
mechanisms. ....................................................................................................... 83
xx
1
Chapter 1 Introduction
1.1 Background and Motivation
Extraction of energy from fossil fuel combustion results in the emission of greenhouse
gases such as CO2 and CH4 and other environmentally hazardous gases including CO,
SO2, H2S and NOx. Public awareness and legislation have led to a policy of reduction of
greenhouse gas and other harmful gas emissions worldwide [1]. Coal is among the most
widely used fuels for generation of electricity. Integrated gasification combined cycle
(IGCC) is one of the cleanest methods of extraction of energy from coal when combined
with carbon and sulfur sequestration and storage [2,3]. A typical IGCC process diagram
is shown in Figure 1.1.
Figure 1.1 A typical integrated gasification combined cycle process diagram [4]
2
The gasifier is one of the core components of such a system. To effectively operate and
control the performance and the output power of the IGCC system, continuous
monitoring of the output gas composition of the gasifier is crucial [5]. In addition to
predicting the heating value of the input gas mixture to the gas turbine, it indicates
important performance parameters, e.g., conversion efficiency, gasifier core reaction
temperature, and hence the overall health of the gasification system.
Conventionally, an extractive method can be used to analyze the components of the
syngas with industrially available sensors. However, often the sample extraction and
preparation (depressurization, cooling, dehumidification and filtration of particulates)
significantly delays the time response and a faster control variable is required. Solution of
this syngas analysis problem has been recognized as a crucial requirement for the
improvement of control and instrumentation of gasifiers by the U.S. Department of
Energy [5]. The primary focus of the research presented in this dissertation is the
development of a robust high-pressure gas sensor for in-situ detection of the syngas
components (e.g. CO, CO2, CH4 and H2O) in harsh, particulate-laden environments based
on laser absorption. Tunable diode laser absorption spectroscopy (TDLAS) with
wavelength-modulation spectroscopy (WMS) offers an effective in-situ method of
measuring concentration of different species with high bandwidth in harsh environments.
The details of this technique are discussed in Chapter 2.
Subsequently, a high-temperature sensor with higher sensitivity was developed for
measurement of methane, an important combustion intermediate produced during
combustion processes. Note that this is a different higher-sensitivity sensor than the one
mentioned before, designed specifically for shock tube applications. The objective of this
study was to develop a tool for better understanding a wide range of combustion
processes and for inferring chemical reaction rates, which are otherwise difficult to
measure directly. This sensor can also be used as an in-situ diagnostic for methane in any
combustion-based, high-temperature environment. Most of the previous efforts in the
literature were primarily focused towards atmospheric monitoring of methane at room
temperature [6–14] or combustion exhaust applications [15–18]. The details of this
sensor are presented in Chapter 5.
3
1.2 Overview of dissertation
This dissertation is arranged into six chapters:
Chapter 2: Development of multi-species laser absorption sensors for in-situ
monitoring of syngas composition. In this chapter, the criteria for selection of
the absorption transitions used for sensing CO, CO2, CH4 and H2O at pressures up
to 20 atm are discussed. The sensors were used to measure the syngas
composition and infer the lower heating value in a high pressure laboratory cell.
Chapter 3: Application of TDLAS-based sensors for in-situ measurement of
syngas composition in a pressurized, oxygen-blown, entrained-flow coal
gasifier. The results of the field measurement of the exhaust composition of an
entrained-flow gasification facility at the Institute for Clean and Secure Energy,
University of Utah are presented in this chapter.
Chapter 4: Application of scanned-wavelength-modulation spectroscopy
sensor for simultaneous measurement of CO, CO2, CH4 and H2O in a high-
pressure syngas output stream from an engineering-scale transport reactor
gasifier. This chapter discusses the design details and the results of the field
campaign for sensing CO, CO2, CH4 and H2O in the high-pressure (< 20 atm) and
moderate-temperature (~ 600 K) outlet syngas stream at the National Carbon
Capture Center operated for US Department of Energy by Southern Company in
Wilsonville, Alabama.
Chapter 5: Development of a high-sensitivity, interference-free diagnostic for
measurement of methane in shock tubes. This chapter describes the details of
the selection criteria for the CH4 transition and a demonstration of the sensor
performance in a shock tube experiment.
Chapter 6: Summary and opportunities for future work. The major
contributions of the dissertation are summarized and future work is proposed.
4
Chapter 2 Development of multi-species laser
absorption sensors for in-situ monitoring of
syngas composition
The contents of this chapter have been published in Applied Physics B [16] under the title "Multi-species laser absorption sensors for in situ monitoring of syngas composition" and presented in the 8th National Combustion Meeting [19].
2.1 Introduction
Tunable diode lasers (TDLs) offer promise for providing a sensor with desirable
characteristics. The advantages of the TDL-based sensors are fast response, non-intrusive
nature, and sensitive species-specific detection capabilities. TDL sensors have been
demonstrated previously to monitor gases in different combustion systems [20–25], for
environmental monitoring [26–28], and other energy conversion devices [29]. Our initial
work in coal gasifiers to monitor the temperature and water concentration has been
discussed by Sun et al. [30]. Here we present the design rules and validation testing of the
gas composition sensors utilized in that work. Other related literature includes the
detection of HCl in atmospheric pressure syngas [31].
Some of the challenges of developing an optical sensor unique to gasification
environments are:
1. A particulate-laden environment with extremely low light transmission [32].
2. The high pressure of gasification processes [3] produces collision broadening of the
absorption transition leading to a decreased peak signal and an absence of the non-
absorbing baseline typically used with direct absorption spectroscopy.
These problems were overcome by the use of a 1f-normalized WMS-2f technique that
has been demonstrated previously to be effective in high-pressure and noisy
5
environments [33,34]. In this method, the injection laser current of the diode laser is
modulated sinusoidally, producing a simultaneous variation in laser output intensity and
frequency. The signal transmitted through the absorption medium is analyzed at the
integer multiples of the modulation frequency — hence the terminology 1f, 2f, and so on.
This technique is well-known for noise rejection in the detection of trace species [21–
23,25–28,35] and has been used at Stanford University in noisy environments [20,36–
38]. This normalization technique accounts for the variations in non-absorption losses of
the transmitted laser intensity [33,34,39].
The current sensor uses four lasers for detection of CO, CO2, CH4, and H2O at the center
frequencies of 4301, 4957, and 4367 (Nanoplus), and 7394 (NEL) cm-1; i.e. 2325, 2017,
2290, and 1352 nm, respectively. The remainder of the gas is assumed to be H2, thus
accounting for the major species in the syngas. With this information, the heating value
and the Wobbe index of the syngas can be monitored as a part of a real-time control loop.
This chapter describes the sensor design and validates performance in a laboratory
environment with known gas composition.
2.2 Sensor principle
A great deal of work has been done in the past [33–35,40–42] to develop an accurate
WMS model for the large modulation depths needed for absorption sensing at elevated
pressure. A summary is presented here to guide the reader and define the notation.
The laser is modulated by sinusoidally varying injection current at angular frequency ω =
2πf which results in an intensity and frequency response as follows:
(2.1)
(2.2)
where is the frequency of light at time t, is the center frequency, is the
modulation depth, ψ is the initial phase of the frequency modulation, is the unabsorbed
beam intensity, is the average intensity at the center frequency, i1 is the linear (1f)
221100 2coscos1)( titiItI
tat cos)(
a
0I
0I
6
intensity modulation amplitude, i2 is the second-order (2f) intensity modulation amplitude
and ψ1 and ψ2 are the initial phases of the 1st- and 2nd- order intensity modulation. For
the lasers used currently, the higher-order intensity modulation terms were found to be
negligible. But if faced with a highly non-linear DFB diode laser performance, such
effects could be taken into account as shown by Sun et al. [42].
From the Beer-Lambert law, it is known that the transmission coefficient (τ) of a
monochromatic light beam at frequency is governed by the relation:
(2.3)
where I0 is the incident beam intensity, I is the transmitted beam intensity and α is the
spectral absorbance for a pressure P, path length L, mole fraction of the ith absorbing
species xi, transition linestrength Si,j and lineshape function of the jth transition, as
defined by the expression
ji
jijii SxLP,
,, (2.4)
The above expression assumes uniform gas composition and temperature along the laser
line of sight (LOS). For an isolated transition, the total area under the absorbance curve is
given by:
jii SLxPd ,
(2.5)
The lineshape function is approximated by a Voigt function characterized by the
collision-broadened full-width at half-maximum (FWHM), [cm-1] and the Doppler
FWHM, [cm-1]. The collisional FWHM is given by the expression:
(2.6)
The dependence of the collision-broadening factor on temperature can be modeled as a
power law expression:
eI
I
0
ji ,
ji ,
ic,
d
j
ijjic xP 2,
7
jn
refrefijij T
TTT
(2.7)
where Tref is the reference temperature (chosen to be 296 K) and nj is the temperature-
dependence index.
Due to the sinusoidal modulation at a frequency ω of the laser wavelength, the resulting
transmission is also periodic with a period ω. Therefore, it can be expressed as a Fourier
series expansion as follows:
0
cosk
k tkHt (2.8)
where Hk is the kth Fourier coefficient of the expansion and can be expressed as:
dkaHk
k cos cos1
1
0
(2.9)
By combining equations (2), (3) and (8) we get the transmitted laser intensity:
022110 cos 2coscos1)(
kk tkHtitiItI (2.10)
No assumptions were made about the optical depth in this expression, and thus it can be
used with any level of absorbance. A crucial part of the WMS technique involves
extraction of the harmonics of the above signal at integral multiples of the modulation
frequency ω. This is achieved using a digital lock-in amplifier, where the signal obtained
from the detector is first multiplied by or , after which a low-pass filter
for the X and Y components, respectively, can yield the nf signal. The resulting equations
for the X and Y components of the 1f and 2f signals are given by (as previously shown by
Rieker et al. [34]):
2cos
2cos
22 2312
12
0110
1 HHiH
HiHIG
X f (2.11)
2sin
2sin
22 2132
102
10
1 HHi
HH
iIG
Y f (2.12)
tncos tnsin
8
SymmetricAsymmetric
f HHHi
HHiIG
X 22042
13110
2 2cos22
cos22
(2.13)
SymmetricAsymmetric
f HHi
HHiIG
Y 2sin22
sin22 20
42113
102 (2.14)
where G is the overall electro-optic gain in the signal that includes the attenuation due to
beam extinction by particles. The 1f signal magnitude is given by
(2.15)
The 2f by 1f normalization scheme with background subtraction used in this work is
identical to the one used by Rieker et al. [34]:
(2.16)
This normalization scheme cancels the factor G that accounts for the majority of random
fluctuations in the average signal due to laser noise and non-absorption transmission
losses, resulting in a robust sensor applicable in harsh environments.
2.3 Laser wavelength modulation characterization and optical
system intensity modulation characterization
2.3.1 Laser wavelength modulation characterization
To model an accurate WMS simulation, the absorption is described by the spectroscopic
parameters including linestrength and collisional broadening coefficients as described in
Section 2.2, and the parameters that describe the injection current-tuning behavior of the
specific TDL. These parameters include a [cm-1], the modulation depth, and ψ, the initial
phase of the wavelength modulation. Figure 2.1 shows an example laser characterization
result of a DFB laser near 1352nm. The red line shows the best-fit sine function which is
21
211 fff YXR
2
1
2
1
2
2
1
2
1
21/2
bgf
f
rawf
f
bgf
f
rawf
fff R
Y
R
Y
R
X
R
XS
9
used to model the variation of the laser frequency tuning due to modulation. The method
to evaluate the parameters that quantify the injection-current tuning characteristics of the
TDLs has been discussed in detail previously, and the reader is referred to the
literature [33,43].
Figure 2.1 Example of laser wavelength response to the injection-current modulation at 1 kHz. The blue
circles are the relative frequency measured by a solid etalon with 0.02cm-1 FSR. The measured modulation
depth is 0.101cm-1 and initial phase of the wavelength modulation is -2.1363 radian. The red line represents
the best-fit sine function which is used to model the variation of the laser frequency tuning due to
modulation.
2.3.2 Optical system intensity modulation
The light transmission can vary as a function of laser wavelength due to the optical
components along the line of sight. Most often such variation is due to interference from
partial reflection at parallel optical surfaces, commonly called "etalon" effects. At some
wavelengths, the interference is constructive, resulting in a transmission close to the
unity, and at other wavelengths, this interference can be destructive, resulting in a lower
transmission (see Figure 2.2). To minimize this interference, some surfaces of the optical
components are slightly wedged, but it is difficult to avoid all such interference in the
entire optical system (especially for components with thin parallel surfaces). For
example, many IR-detectors are protected with flat windows in front of the active area
that can result in reflective interference when the laser wavelength is tuned. As this
0 0.2 0.4 0.6 0.8 1
x 10-3
-0.1
-0.05
0
0.05
0.1a= 0.101cm-1; phi= -2.1363rad
Time [s]
Fre
quen
cy [c
m-1
]
characterized wavelength response result
10
interference is wavelength dependent, it can produce background signals at the harmonics
of the laser modulation frequency. Even larger interference can be produced by other
optical components such as measurement volume windows (non-wedged), optical filters,
or optical amplifiers.
Figure 2.2 Transmission of a thin cavity built by two parallel surfaces (cavity length: 2mm, surface
reflectivity: 2%)
As a result, it will be more accurate to evaluate the intensity modulation characteristics
( mi and m ) of the entire optical system rather than simply the laser. The parameters mi
and m are extracted from the WMS background signals for a measurement with no
absorbers in the l as described in [30]. In that work, it was illustrated that for practical
measurements, when some external interference (e.g., etalons) cannot be easily
eliminated, or if the system includes optical components that are spectrally sensitive,
characterizing the overall system can be more relevant than characterizing only the laser.
2.4 Effects of collision broadening on modulation optimization
The WMS signal depends on the absorbance, the lineshape, and the range of the
wavelength modulation. Thus, optimization of the modulation parameters must include
consideration of collision broadening. First, we consider an absorber with a single
isolated transition. This fictitious species is approximated by the P24 transition of the
11
20012 ← 00001 band of CO2 near 4957 cm-1. Simulating a scanned-WMS lineshape
requires the parameters for both the transition and the TDL wavelength modulation as
given in Table 2.1.
Table 2.1 Parameters for simulating the scanned-WMS lineshape
Molecular Parameters (Eqn. 4 and 7) Laser Parameters (Eqn. 1 and 2)
Line center [cm-1] 4957.08 Modulation depth [cm-1] 0.5
Line Strength [cm-2atm-1] 0.024 i1 0.199
2γself (296K) [cm-1atm-1] 0.196 i2 0.0012
nself 0.73 ψ1 3.44
2γN2 (296K) [cm-1atm-1] 0.143 ψ2 3.07
nN2 0.72
The wavelength-scanned direct-absorption simulation of the single isolated transition is
shown in Figure 2.3(a) for 10% absorber in air at 5 and 12 atm. All the simulations were
computed using the Humlicek [44] Voigt lineshape algorithm. For pressures above 0.5
atm and T < 500 K, collision broadening dominates over Doppler broadening, resulting in
a predominantly Lorentzian nature for the lineshape. When collision-broadening is
dominant, the peak absorbance for constant mole fraction of the collision-broadened
Voigt lineshape is nearly pressure independent, while the increase in linewidth scales
linearly with pressure. The 1f-normalized WMS-2f lineshape is shown in Figure 2.3(b).
One of the interesting changes observed in the 12 atm, 2f spectra is the transformation of
a three-lobed WMS lineshape into a two-lobed structure. The disappearance of the first
lobe is due to the nature of symmetric and asymmetric terms in equations 13 and 14.
Plots of these asymmetric and symmetric parts of the X2f component of the spectra at 5
and 12 atm are shown in Figure 2.4(a) and Figure 2.4(b), respectively. The appearance or
disappearance of the first peak in the 1f-normalized WMS-2f spectrum (Figure 2.3(b)) is
governed by the relative magnitudes and shapes of the lobes in the symmetric and the
asymmetric parts of the X2f signal denoted by SL1 and ASL1 in Figure 2.4(a) and Figure
2.4(b). At lower pressures (say 5 atm), SL1 is prominent and higher in magnitude than
ASL1. But with increase of pressure, the SL1 flattens out much faster in comparison to
ASL1, evolving into structures comparable in magnitude and therefore cancelling each
12
other. Hence, the absolute value of the 1f-normalized WMS-2f signal morphs into a two-
lobed structure as shown in Figure 2.3(b).
Figure 2.3 (a) Simulated absorbance spectrum of the fictitious molecule characterized by a single transition of CO2 near 4957 cm-1 at 5 and 12 atm; (b) 1f-normalized WMS-2f spectrum of the fictitious molecule near 4957 cm-1 at 5 and 12 atm.
Figure 2.4 WMS X2f components of fictitious molecule characterized by a single transition of CO2 near 4957 cm-1 at (a) 5 and (b) 12 atm.
Modulation index, defined as m = a/HWHM, of around 2.2 was reported to maximize the
2f signal at lower pressures in Reid and Labrie [40] and other authors [35,43]. The
increase of pressure, which blends transitions as shown in Figure 2.5 for CO2, changes
the overall shape of the absorption feature and thus the WMS signal. This results in a
different modulation index for maximum WMS signal; an example of which for CO2
absorption in the 2 µm band is shown in Figure 2.6. This variation is naturally dependent
upon the separation of the transitions and the efficiency of the broadening collisions,
4952 4956 4960 49640.00
0.02
0.04
0.06
0.08 10% target absorber in airT = 296 KL = 7.8 cm
Abs
orba
nce
Frequency [cm-1]
5 atm 12 atm
(a)
4956 4958 49600.00
0.05
0.10
1f-n
orm
aliz
ed W
MS
-2f m
agni
tude
[a.u
.]
Frequency [cm-1]
5 atm 12 atm
10% target absorber in airT = 296 K
a = 0.5 cm-1
L = 7.8 cm
(b)
4955 4956 4957 4958 4959 4960
-0.005
0.000
0.005
0.010
0.015
ASL1
X2f a
nd it
s co
mpo
nen
ts [
a.u.
]
Frequency [cm-1]
X2f
X2f - asymmetric part
X2f - symmetric part
P = 5 atm(a)
Line of symmetry
SL1
4955 4956 4957 4958 4959 4960
-0.002
0.000
0.002
0.004
SL1X
2f an
d its
com
pone
nts
[a.u
.]
Frequency [cm-1]
X2f
X2f - asymmetric part
X2f - symmetric part
P = 12 atm(b)
Line of symmetryASL1
13
which in turn depends on the molecule, band, etc. Even with a single transition there is a
slight change in the optimum modulation index with pressure, since the magnitude of 2.2
tracks the peak of the term H2 in equation 13 and the contributions from the other terms
rise in importance with increase in pressure. Thus a complete simulation is required to
find the optimum modulation depth for each chemical species targeted. At high pressures,
the narrower the feature is, the larger is its WMS signal. Narrowness of the feature often
becomes a more important factor than the absolute peak absorbance magnitude.
Figure 2.5 Spectral lineshape blending at elevated pressures for CO2 in the P-branch of 20012 ← 00001 band. Numbers in the plot indicate the rotational quantum number of the lower energy state.
Figure 2.6 Variation of the modulation index for the largest WMS-2f signal due to the blended absorbance profile of CO2 near 4957 cm-1 at 296K.
The magnitude of the 1f-normalized WMS-2f signal varies strongly with the modulation
depth as discussed above. To select the optimum modulation depth for use over a range
of pressures, the peak 1f-normalized WMS-2f signals were plotted versus modulation
4950 4955 4960 49650.00
0.04
0.08
0.12
0.16
0.2022
27
23
2625
10% CO2 in air
T = 296 KL = 7.8 cm
Abs
orb
ance
Frequency [cm-1]
5 atm 10 atm
24
P branch (20012 00001 band)
0 2 4 6 8 10 12 14 16 180.0
0.4
0.8
1.2
1.6
2.0
2.4
2.8
3.2
3.6
4.0
Op
timum
mod
ula
tion
inde
x (a
/HW
HM
)
Pressure [atm]
CO2 molecule (Pressure blended)
Fictitious single isolated transition
14
depth at different pressures. Generally, at higher-pressure, the WMS signals become
smaller due to the reduction in spectral curvature. Therefore, the modulation depth was
optimized to enhance the higher pressure signals. As an example, Figure 2.7 shows the
case for the 2.0 µm laser for detection of CO2 in a sample syngas-like mixture containing
30% CO, 30% CO2, 15% H2, 15% N2, and 10% H2O. The modulation depth selected
optimized the signal strength at 20 atm while retaining as much signal strength as
possible at lower pressure. For CO, the optimum modulation depth could not be reached
as a result of the limitation of the laser tuning characteristics. The selected modulation
depths for these sensors are listed in Table 2.2.
Table 2.2 Selected Modulation depth for the sensors
Species Selected modulation depth [cm-1]
CO 0.58
CO2 0.76
CH4 1.02
H2O 1.08
Figure 2.7 Variation of 1f normalized WMS-2f magnitude at the selected peak wavelength with modulation
depth at 10 kHz modulation frequency for the laser operating at 2017 nm for CO2 detection at different
pressures.
0.2 0.4 0.6 0.8 1.0 1.2 1.40.00
0.04
0.08
0.12
0.16
0.20
1f -
nor
mal
ized
WM
S-2
f pea
k m
agn
itude
[a.u
.]
Modulation depth [cm-1]
10 atm 15 atm 20 atm
Selectedmodulation
depth
T = 350 KL = 7.8 cm30% CO
2 in syngas
15
2.5 Selection of transitions
An overview of absorption spectra of CO, CO2, CH4 and H2O at 15 atm, 350 K and a
path length of 20 cm is shown in Figure 2.8. First we consider the line selection for CO
which has an absorption spectrum that behaves nearly like an isolated single transition for
the pressure range less than 25 atm studied here. CO2 is considered next as it provides an
example of an absorption spectrum that is severely blended at 15 atm. Finally we
consider CH4 and H2O, which have irregularly structured absorption spectra.
1000 1200 1400 1600 1800 2000 2200 2400
0.01
0.1
1
10
Abs
orba
nce
Wavelength [nm]
10% CO2
10% CO
10% H2O
1% CH4
P = 15 atmT = 350 KL = 20 cm
Figure 2.8 Overview of the absorption spectra of CO, CO2, CH4 and H2O at 15 atm and 350 K
2.5.1 Carbon monoxide
Successful measurements of carbon monoxide were performed by Chao et al. [43] in
combustion exhaust at atmospheric pressure. The current work uses the same line (R11)
in a different pressure domain. Chao et al. showed that at low concentrations of CO and
high concentrations of H2O, this line possesses the potential for a sensor with an excellent
detection limit. At higher pressures, the R branch of CO in Figure 2.9(a) still retains its
nearly resolved structure, in contrast to CO2, shown in Figure 2.5, because the line
spacing of the CO transitions is about 2.3 times that of CO2. In this case, the sensitivity of
WMS signal in the R-branch of CO at higher pressures is much less impacted by line
blending. Additionally, the P-branch of this CO band has a significant amount of
interference absorption from H2O making it unsuitable for CO sensing in gasifiers. A plot
16
of the absorbance profile of the R-branch of CO in the 2.3 m band is displayed in Figure
2.9(a) and the WMS 2f/1f spectra in the vicinity of the selected transition are shown in
Figure 2.9(b). The temperature sensitivity was calculated from the change in the absolute
peak absorbance divided by the temperature change. In the 350-400 K range, this
absorbance peak temperature sensitivity was K-1 in a 30% mixture with air at 300
– 400 K. Thus, a variation of 10K results in less than a 1% change in the absorption
signal, and this transition can provide measurements of CO that are quite insensitive to
variation in gas temperature for the range studied here.
Figure 2.9 (a) Absorbance spectrum of CO near 4301 cm-1; (b) 1f-normalized WMS-2f spectrum of water near 4301 cm-1.
2.5.2 Carbon dioxide
The simulated absorbance spectra of 20% CO2 in air at 1 and 15 atm and 350 K, shown in
Figure 2.10(a), reveals that the R-branch of the 2 µm absorption band of CO2 is stronger
than the P-branch. Despite this fact, the sensor transition was selected from the P-branch,
since collision-broadening leads to the blending of all distinct features in the R-branch at
pressures as low as 15 atm (at 300K). The line spacing in the P-branch is larger than that
of the R-branch, and as a result there are a few distinct features observable at 15 atm that
can be utilized for measurements. Figure 2.10(b) displays the WMS spectra near the
selected transition. This transition is primarily chosen to maximize the WMS signal.
Additionally, at the gasifier exhaust temperatures of interest (300 - 450 K), spectral
interferences from the other species are negligible. The low temperature sensitivity
3101
4300 43500.0
0.2
0.4
0.6
0.8
1.0(a)
Abs
orba
nce
Frequency [cm-1]
1 atm 15 atm
20% CO in airT = 350 KL = 7.8 cm
Selectedtransition
R11 @ 4300.7 cm-1
4300 43040.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50(b)
1f -
no
rma
lized
WM
S-2
f ma
gnitu
de
[a.u
.]
Frequency [cm-1]
10% CO 20% CO 30% CO
Balance: Air
modulation depth: 0.58 cm-1
T = 350K, P = 15 atm
Selected WMS peak
17
( K-1 in a 30% mixture with air) of the selected line avoids unnecessary variation
of the absorption signal with mixture temperature.
Figure 2.10 (a) Absorbance spectrum of CO2 near 4957 cm-1; (b) 1f-normalized WMS-2f spectrum of CO2 near 4957 cm-1.
2.5.3 Methane
Due to a generally low methane concentration in syngas, special attention was paid to the
strength of the selected transition. From Figure 2.11(a), it can be seen that the selected
transition is the strongest and the sharpest feature in its vicinity. This has led to a larger
WMS-2f/1f magnitude in comparison to its neighbors (Figure 2.11(b)). Another
important limiting aspect of this selection was presence of spectral interference from
other species in the 2.25 µm band (Figure 2.12). Stronger CH4 absorption features are
present in the same band at frequencies lower than 4350 cm-1. But due to the presence of
the extremely strong absorption band of CO, shown in Figure 2.12, these stronger CH4
transitions were not suitable for CH4 sensing in syngas mixtures. Another major
interfering species in this region is NH3. The peak absorbance of CH4 near 4367 cm-1 is
about 14 times that of NH3 as shown in Figure 2.12. Therefore, interference from NH3
was also minimized by the transition selection. In addition, the low temperature
sensitivity of the selected transition ( K-1 for 1% CH4 in air) minimized the
variation of the absorption signal with temperature in the range 350-400 K.
4104
4940 4960 4980 50000.0
0.1
0.2
0.3
0.4
0.5
0.6(a)
Abs
orba
nce
Frequency [cm-1]
1 atm 15 atm
Selected transition
P24 @ 4957 cm-1
20% CO2 in air
T = 350KL = 7.8 cm
4956 4958 4960 4962 49640.00
0.02
0.04
0.06
0.08
0.10
0.12(b)
1f -
no
rmal
ize
d W
MS
-2f m
ag
nitu
de [a
.u.]
Frequency [cm-1]
15% CO2
20% CO2
25% CO2
20% CO2
35% CO2
Balance: Air
modulation depth: 0.76 cm-1
T = 350 K, L = 7.8 cm
Selected WMS peak
4102
18
Figure 2.11 (a) Absorbance spectrum of CH4 near 4367 cm-1; (b) 1f-normalized WMS-2f spectrum of CH4
near 4367 cm-1.
Figure 2.12 Absorbance spectrum of CH4 near 4367 cm-1 showing the presence of interfering species near the CH4 absorption band used.
2.5.4 Water
Simulations of absorbance spectra based on the HITRAN database [45] for 10% H2O in
air near 7400 cm-1 at 400 K and pressures of 1 and 15 atm in air are shown in Figure
2.13(a). The simulations show that at elevated pressures, the transitions distinctly
resolved at lower pressures become blended into a few continuous features. However, the
water transition near 7394 cm-1 has some distinctly noticeable features that led to its
selection. In particular, this transition is much narrower and more isolated than the
surrounding transitions. The 1f-normalized WMS-2f spectrum for water in the vicinity of
the selected transition is shown in Figure 2.13(b). The magnitude of the WMS spectra of
this peak is larger than that of the neighbors. In addition, the peak-magnitude sensitivity
to mole fraction is considerably larger than that of the neighbors as well. Finally, due to
4350 4400 4450 45000.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14(a)
Abs
orba
nce
Frequency [cm-1]
1 atm 15 atm
1% CH4 in air
T = 350 KL = 7.8 cm
Selected transition
@ 4367 cm-1
4360 4370 4380 43900.00
0.01
0.02
0.03
0.04
0.05
0.06(b)
1f n
orm
aliz
ed
WM
S-2
f mag
nitu
de
[a.u
.]
Frequency [cm-1]
0.5% CH4
1.0% CH4
1.5% CH4
Balance: Air
modulation depth: 1.0 cm-1
T = 350 KL = 7.8 cm
Selected WMS peak
4320 4350 4380 44100.0
0.1
0.2
0.3
Abs
orba
nce
Frequency [cm-1]
1% CH4
10% H2O
30% CO 40% CO
2
0.2% NH3Selected
transition
@ 4367 cm-1
inaccessiblestrongertransition
Balance: AirT = 350 KP = 15 atm
19
its narrow and isolated lineshape, dependence on the broadening parameters of the
neighbors is also minimal. There also was no interference absorption expected from any
other major syngas component in this region. Again, this selected line has minimal
sensitivity to temperature in the 350-400 K range, with a peak absorbance variation of
K-1 in a 10% mixture with air.
Figure 2.13 (a) Absorbance spectrum of water near 7394 cm-1; (b) 1f-normalized WMS-2f spectrum of water near 7394 cm-1.
2.6 Measurement of spectral line parameters
One of the most important steps in the development of a sensor to be employed in a
multi-component mixture environment is knowledge of the transition line strengths and
collision-broadening parameters of the measured species as well as those for the other
major species in the mixture. For the four species concerned, the broadening parameters
with all the other species, assumed here to be CO, CO2, H2, H2O, and N2, were measured.
Due to the relatively low mole fraction of CH4 in syngas, collision broadening of the
other measured species broadened by CH4 was not investigated.
3101
7360 7380 7400 7420 74400.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
Abs
orba
nce
Frequency [cm-1]
1 atm 15 atm
Selectedtransition
@ 7394 cm-1 10% H2O in air
T = 400KL =7.8cm
(a)
7380 7390 7400 74100.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50(b)
1f-n
orm
aliz
ed
WM
S-2
f mag
nitu
de [a
.u.]
Frequency [cm-1]
5% H2O
10% H2O
15% H2O
Balance: Air
modulation depth : 1.1 cm-1
T = 400 K P = 15 atm
SelectedWMS peak
20
Figure 2.14 (a) Sample direct absorbance spectra of CO2 showing agreement with a Voigt fit; (b) Measured collisional FWHM for 5% CO2 in H2 at different pressures.
The spectroscopic parameters were measured by wavelength-scanned direct absorption as
a function of pressure and temperature in a static cell of length 9.9 cm for CO, CO2 and
CH4, and another static cell of 76.2 cm for H2O, following the method of Arroyo et
al. [46]. The cell was first filled by a known quantity of pure gas and the lineshapes of the
selected transitions were acquired, e.g. for CO2 in Figure 2.14(a). The linewidth of the
single-mode diode laser (< 0.0002 cm-1) is neglected. The best Voigt fit of the measured
lineshape was used to compute the integrated absorbance, which is proportional to the
line strength and absorbing-species partial pressure, as discussed above. The linear slope
of integrated absorbance versus pressure provided the line strength.
0.00
0.05
0.10
0.15
0.20
4956.6 4956.8 4957.0 4957.2 4957.4 4957.6
-202
Abs
orba
nce
Measurement Voigt fit
5% CO2 in H
2
T = 421 KP = 549 torrL = 9.9 cm
Res
idua
l [%
]
Frequency [cm-1]
(a)
0.1 0.2 0.3 0.4 0.5 0.60.00
0.02
0.04
0.06
0.08
0.10
0.12(b)
308K 362K 421K 529K 640K
Col
lisio
nal F
WH
M [
cm-1]
Pressure [atm]
21
The collision-broadening coefficients were determined from the Lorentzian-halfwidth
parameter from Voigt fits of the lineshape with the line strength fixed. The self-
broadening was first determined over a range of pressures. Then, binary mixtures of the
absorbing gas with a collision partner were studied and the absorption lineshape fit with
only the width contribution of the collision partner allowed to vary. The slope of the
fitted width versus pressure gives the broadening coefficient, the results of which are
given in Table 2.3. A sample set of measurements for CO2 broadening in H2 is shown in
Figure 2.14(b). The uncertainty in the measured line strengths and broadening (except
water) was estimated to be within 1% and 2%, respectively, in the measured
temperature range of 300 – 700 K. For water broadening, this uncertainty was larger
due to the reduced range of binary mixture pressures and was estimated to be less than
5%.
22
Table 2.3 Measured spectroscopic parameters of the selected transitions
23
2.7 Sample WMS measurements of species in N2 at elevated
pressure
The first validation experiments of the sensor involved the measurement of high-pressure
1f-normalized WMS-2f spectra for a binary mixture (in this case in N2), for comparison
with spectra simulated using the measured spectral database. As observed from Figure
2.15, the WMS spectra at higher pressures show very good agreement with the
simulations. These measurements confirm that other high-pressure phenomena that were
not considered, such as line mixing and other non-Lorentzian effects, are not very
important at these pressures and hence could be neglected in this work.
Figure 2.15 Sample WMS spectra for (a) CH4, (b) CO and (c) CO2 in N2 at 10 atm.
4365 4366 4367 4368 4369 43700.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07(a)
1f -
nor
ma
lize
d W
MS
-2f m
agn
itud
e [a
.u.]
Frequency [cm-1]
0.5 % CH4 in N
2
T = 298 K, P = 10 atm
L = 23 cm, a = 1.0 cm-1
4298 4300 4302 43040.0
0.2
0.4
0.6
0.8
1.0
1.2(b)
1f -
nor
ma
lized
WM
S-2
f mag
nitu
de
[a.u
.]
Frequency [cm-1]
3% CO in N2
T = 298 K, P = 10 atm
L = 100.5 cm, a = 0.58 cm-1
4956.5 4957.0 4957.5 4958.00.000.010.020.030.040.050.060.070.080.090.100.110.12(c)
1f -
nor
ma
lized
WM
S-2
f mag
nitu
de
[a.u
.]
Frequency [cm-1]
2% CO2 in N
2
T = 298 K, P = 10 atm
L = 100.5 cm, a = 0.76 cm-1
24
2.8 Sample WMS measurements of species in syngas mixture at
different pressures
After verification of the spectra to pressures of 20 atm, the sensor was tested in different
synthetic syngas mixtures of varying concentrations as a function of pressure. A
standard mixture was procured (Praxair) containing 25% CO, 25% H2, 0.6% CH4 and
balance (49.4%) CO2. This gas mixture was then combined with varying amounts of
one of the components. Wide spectral range wavelength scans of WMS lineshape with
frequency for each of the components were made with these mixtures for different
pressures in a room temperature, high-pressure cell, with a path length of 23 cm. Water
vapor was not included in these room temperature measurements as the vapor pressure
of water is about 0.03 atm and a maximum mixture composition of 0.2% H2O can be
produced for a mixture of 20 atm, which is much lower than the target concentration of
the sensor's domain of measurement.
2.8.1 Carbon monoxide
The CO sensor has the advantage of largest absorbance and a well-isolated feature. This
leads to a very high WMS signal level at all the pressures investigated, i.e. 5, 10, 15 and
20 atm as shown in Figure 2.16(a), (b), (c) and (d), respectively. The good agreement
between the simulations and the measurements especially for the peak near 4300.7 cm-1 is
evident from these figures.
25
Figure 2.16 Sample WMS spectra for CO in a sample syngas mixture at (a) 5 atm, (b) 10 atm, (c) 15 atm and (d) 20 atm.
2.8.2 Carbon dioxide
The CO2 absorption spectrum is severely blended at pressures greater than 5 atm. This
has a negative influence on the WMS signal strength as discussed above. Sample
measured and simulated 1f-normalized WMS-2f spectra for CO2 in a typical syngas
mixture at 25°C and at pressures of 5, 10, 15 and 20 atm are shown in Figure 2.17 (a),
(b), (c) and (d). As seen from Figure 2.17 (d), the second lobe has completely
disappeared at 20 atm. This is a result of a relatively featureless absorbance signature for
CO2 at high pressures. Despite this, the peak near 4957 cm-1 has a consistent agreement at
all pressures with the simulations, thus showing suitability for CO2 detection in syngas
flows.
4300 4301 4302 43030
1
2
3
4
5(a) P = 5 atm35% CO in syngas mixtureT = 298 KL = 23 cm
1f n
orm
aliz
ed W
MS
-2f m
agni
tude
[a.u
.]
Frequency [cm-1]
4300 4301 4302 43030.0
0.5
1.0
1.5
2.0
2.5
3.0(b)P = 10 atm35% CO in syngas mixtureT = 298 KL = 23 cm
1f -
nor
mal
ize
d W
MS
-2f m
agn
itud
e [a
.u.]
Frequency [cm-1]
4300 4301 4302 43030.0
0.5
1.0
1.5(c)P = 15 atm35% CO in syngas mixtureT = 298 KL = 23 cm
1f -
nor
ma
lized
WM
S-2
f mag
nitu
de
[a.u
.]
Frequency [cm-1]
4300 4301 4302 43030.0
0.2
0.4
0.6
0.8(d)P = 20 atm35% CO in syngas mixtureT = 298 KL = 23 cm
Frequency [cm-1]1f -
no
rma
lize
d W
MS
-2f m
ag
nitu
de
[a.u
.]
26
Figure 2.17 Sample WMS spectra for CO2 in a sample syngas mixture at (a) 5 atm, (b) 10 atm, (c) 15 atm and (d) 20 atm.
2.8.3 Methane
1f-normalized WMS-2f spectra were measured for CH4 at a temperature of 25°C and
pressures of 5, 10, 15 and 20 atm as shown in Figure 2.18(a), (b), (c) and (d),
respectively. In spite of having multiple lines in the region and a relatively more densely
spaced spectral structure, the agreement between the simulation and the data is quite
reasonable. These results serve to verify the accuracy of the mixture broadening values
and the spectral modeling used.
4956.5 4957.0 4957.5 4958.00.0
0.5
1.0
1.5
(a)P = 5 atm44% CO
2 in syngas mixture
T = 298 KL = 23 cm
1f -
nor
mal
ized
WM
S-2
f mag
nitu
de [a
.u.]
Frequency [cm-1]4956.5 4957.0 4957.5 4958.0
0.0
0.1
0.2
0.3
0.4(b) P = 10 atm44% CO
2 in syngas mixture
T = 298 K, L = 23 cm
1f -
nor
ma
lize
d W
MS
-2f m
ag
nitu
de [a
.u.]
Frequency [cm-1]
4956.5 4957.0 4957.5 4958.00.00
0.05
0.10
0.15
0.20(c) P = 15 atm44% CO
2 in syngas mixture
T = 298 KL = 23 cm
Frequency [cm-1]1f -
nor
mal
ize
d W
MS
-2f m
agni
tude
[a.u
.]
4956.5 4957.0 4957.5 4958.00.00
0.02
0.04
0.06
0.08
0.10(d) P = 20 atm44% CO
2 in syngas mixture
T = 298 KL = 23 cm
1f -
no
rma
lized
WM
S-2
f ma
gnitu
de
[a.u
.]
Frequency [cm-1]
27
Figure 2.18 Sample WMS spectra for CH4 in a sample syngas mixture at (a) 5 atm, (b) 10 atm, (c) 15 atm and (d) 20 atm.
2.9 Summary of the laboratory validation experiments
The gas composition of the major components of syngas determined by the sensor is
compared with known values in Figure 2.19. The measurements were repeated three
times at the same mole fraction of CH4 and two times for CO with different mixture
compositions. These measurements consistently fell within the uncertainty of the sensor
from the known value. The known mixtures were prepared by volumetric addition of
varying amounts of one of the components with the base mixture of 25% CO, 25% H2,
0.6% CH4 and balance (49.4%) CO2. The mixture compositions used for the validation
experiments are listed in Table 2.4. Measured data points agree with the known values
within 4%, 4% and 8% (1%, 2% and 0.05% of total) of the measurements of the mole
fraction of CO, CO2 and CH4, respectively. These measurements were done as a function
of pressure and the general trend reflects a small increase of the difference between
4366 4367 4368 43690.00
0.01
0.02
0.03
0.04
0.05
0.06(a)
1f -
nor
mal
ized
WM
S-2
f mag
nitu
de [a
.u.]
Frequency [cm-1]
P = 5 atm0.47 % CH
4 in syngas mixture
T = 298 K, L = 23 cm
4366 4367 4368 43690.00
0.01
0.02
0.03
0.04
0.05
0.06(b)P = 10 atm0.47 % CH
4 in syngas mixture
T = 298 K, L = 23 cm
1f -
nor
ma
lized
WM
S-2
f mag
nitu
de [a
.u.]
Frequency [cm-1]
4366 4367 4368 43690.00
0.01
0.02
0.03
0.04
0.05
0.06(c) P = 15 atm0.47 % CH
4 in syngas mixture
T = 298 KL = 23 cm
1f -
nor
mal
ized
WM
S-2
f mag
nitu
de [a
.u.]
Frequency [cm-1]4366 4367 4368 4369
0.00
0.01
0.02
0.03
0.04
0.05(d)P = 20 atm0.47 % CH
4 in syngas mixture
T = 298 KL = 23 cm
Frequency [cm-1]1f -
nor
mal
ized
WM
S-2
f mag
nitu
de [a
.u.]
28
measured and known values with increasing pressure. The WMS signal becomes
increasingly sensitive to the broadening parameters at higher pressures and these
differences could be attributed to the uncertainties in the broadening coefficients. Some
part of the high-pressure differences might be also attributed to non-Lorentzian behavior
of gases at higher pressures such as line-mixing and finite duration of collisions.
Figure 2.19 Comparison of the known and the measured values of (a) CH4, (b) CO and (c) CO2 mole fractions in various syngas mixtures.
0 5 10 15 200.0
0.2
0.4
0.6
0.8
1.0(a) Measured (for 0.47%) Measured (for 0.59%) 0.48% 0.59%
CH
4 mo
le f
ract
ion
[%]
Pressure [atm]
T = 298 KL = 23 cm
0 5 10 15 200
10
20
30
40
50
60
70(b)
CO
mol
e f
ract
ion
[%]
Pressure [atm]
Measured (for 35%) 35% Measured (for 25%) 25% Measured (for 20.1%) 20.1%
T= 298 KL = 23 cm
0 5 10 15 20
40
60
80
100
(c)
CO
2 mol
e fr
actio
n [%
]
Pressure [atm]
Measured (at 41.0%) 41.0% Measured (at 44.0%) 44.0% Measured (at 49.6%) 49.6% Measured (at 60.0%) 60.0%
T = 298 KL = 23 cm
29
Table 2.4 Percentage compositions of components in the mixtures used for the sensor validation experiments
CO CO2 CH4 H2
25.0 49.6 0.6 24.8
35.0 44.0 0.5 20.5
20.1 41.0 0.5 38.4
20.1 60.0 0.5 19.4
2.10 Calculation of LHV and Wobbe Index of syngas mixture
The lower heating value (LHV) of a fuel is one of the most widely used properties to
compare the heat release from burning different fuels. When a mixture of one mole of
fuel and a stoichiometric amount of air enters a steady-flow reactor at a standard
reference state (1 atm and 25°C) and the products (assumed to be CO2, H2O (vapor
phase), N2) exit at the same standard reference state, the energy released is termed the
lower heating value. When characterizing the LHV output from a gasifier, the mass basis
is often chosen to be the mass of the reacted carbon present in syngas, to be indicative of
the efficiency of the conversion process from the parent solid coal.
In general, the syngas-like mixtures are primarily composed of CO, CO2, CH4, H2O and
H2 along with many trace species such as H2S, NH3, etc. The laser absorption sensors
described here can measure all the components except H2, which is assumed to be the
balance. This assumption provides a path to infer the lower heating value (in MJ/Kg C) of
the syngas as:
42
4422222
0,
0,
0,
0,
0,
0)(, 2
CHCOCOC
CHCHfCOfOHfCOCOfCOfHgOHfC xxxM
xHHHxHHxHLHV
(2.17)
where xi is the mole fraction, is the standard heat of formation and Mi is the molar
mass of the species i. The subscript C on LHV refers to the per kg C basis. The subscript
(g) indicates the gaseous phase of water. Here, the parameters were calculated from the
0,ifH
30
NIST-JANAF [47] tables and the bimolecular species H2 and O2 are reference species
and hence have a zero heat of formation ( ) at 25°C. However, this method of
obtaining the LHV is valid only for oxygen-blown gasifiers which have only low
concentrations of fuel nitrogen in the syngas stream. For air-blown systems or systems
with significant nitrogen purge, N2 must also be accounted for in the syngas mixture.
Another parameter of importance when dealing with modern gaseous fuels is the Wobbe
Index (WI), which is a measure of interchangeability of fuels. It is expressed as:
s
CHCHfCOfOHfCOCOfCOfHlOHf
s G
xHHHxHHxH
G
HHVWI
mol
Nm 0.024465
23
0,
0,
0,
0,
0,
0)(, 4422222
(2.18)
where Gs is the specific gravity of the gaseous fuel with respect to dry air at 25°C and 1
bar. The subscript (l) indicates the liquid phase of water.
The lower heating value was calculated per kilogram carbon basis for each of the syngas
mixtures, and compared to the known value in Figure 2.20(a). A maximum scatter of less
than 6% (rms error < 0.4%) was observed for all these measurements.
Similarly, generally good agreement was achieved between the measured and the known
values of the Wobbe Index as apparent from Figure 2.20(b). The maximum scatter
observed in this case was below 8% (rms error < 0.4%). The effective uncertainty in
these inferred values from the constituents was due to the cumulative uncertainty in each
of the measured components.
0fH
31
Figure 2.20 Comparison of known and measured (a) LHV in MJ/kg C; (b) Wobbe index in MJ/Nm3.
2.11 Conclusions
Multi-species laser-absorption sensors were designed, constructed and tested to monitor
the mole fraction of CO, CO2, CH4 and H2O in synthesis gas mixtures at pressures up to
20 atm and temperatures of 300 - 400 K. The sensor design was based on 1f-normalized
WMS-2f detection of infrared laser absorption. The line selection was optimized for
performance at high pressures and to suppress interference from typical syngas
composition. A database of collision-broadening coefficients was acquired for collisions
with the set of species (CO, CO2, H2, H2O, N2 and CH4) expected in syngas. The
performance of these sensors was evaluated at room temperature up to a pressure of 20
atm. The spectral simulations for the 1f-normalized WMS-2f signals showed agreement
with the measurements in both binary mixtures with N2 and in multi-species synthetic
syngas. The lower-heating value and the Wobbe index were calculated from the sensor
data and compared with the known values. The inferred values were within 6% for the
LHV and 8% for the Wobbe index over the entire pressure range. The sensor has the
potential to become a reliable and fast real-time monitor for gasifier product syngas
composition, with a promising future for new strategies of gasification control. The
sensor was tested successfully in the pilot-scale entrained-flow gasifier at the University
of Utah and engineering-scale transport reactor gasifier at National Carbon Capture
Center, Wilsonville, Alabama. The results of these field campaigns will be discussed in
the next two chapters.
0 4 8 12 16 20 24 28 32 36 400
4
8
12
16
20
24
28
32
36
40
1atm 5atm 10atm 15atm 20atm
Mea
sure
d LH
V [M
J/kg
C]
Known LHV [MJ/kg C]
(a)T = 298 KL =23 cm
0 4 8 120
4
8
12
1atm 5atm 10atm 15atm 20atm
Mea
sure
d W
obb
e In
dex
[MJ/
Nm
3]
Known Wobbe Index [MJ/Nm3]
T = 298 KL = 23 cm
(b)
32
Chapter 3 Application of TDLAS-based
sensors for in-situ measurement of syngas
composition in a pressurized, oxygen-blown,
entrained-flow coal gasifier
The contents of this chapter have been published in Applied Physics B [18] under the title "TDLAS-based sensors for in situ measurement of syngas composition in a pressurized, oxygen-blown, entrained flow coal gasifier" and presented in the 8th National Combustion Meeting [19].
3.1 Introduction
This chapter describes the results from a field measurement campaign conducted during
May 2012, where these sensors were installed on the syngas output stream from the
entrained-flow, oxygen-blown, slagging, pilot-scale coal gasifier at the Institute for Clean
and Secure Energy, University of Utah. The prototype sensor used here had four lasers
for detection of CO, CO2, CH4 and H2O tuned to transition frequencies near 4301, 4957,
and 4367 cm-1 (Nanoplus), and 7394 cm-1 (NEL) (2325, 2017, 2290, and 1352 nm,
respectively), respectively as discussed in Chapter 2. The remainder of the gas was
assumed to be H2, thus accounting for the major species in the syngas from oxygen-
blown gasifiers. (Note the Utah syngas also includes nitrogen from gas purging, which
must be separately determined.) With the laser-absorption sensor readings, the heating
value and the Wobbe index of the syngas could potentially be monitored as a part of a
real-time control loop.
3.2 Apparatus
The prototype diode laser absorption gas sensor for gas composition and heating value,
designed and tested in the laboratory at Stanford University as described in detail in
33
Chapter 2 and Sur et al. [16], was demonstrated for practical applicability in a pilot-scale,
pressurized, oxygen-blown, entrained-flow “Texaco-style” or “GE-style” gasifier [48]
located at the University of Utah’s Industrial Combustion and Gasification Research
Facility (Figure 3.1). The gasifier is located indoors in a dedicated laboratory building,
which offers excellent access for research and monitoring. The control room has
adequate space for the sensor electronics and control, and the reactors can be reached
with modest length (~30 m) signal cables. The research nature of this facility was ideal
for the proof-of-concept testing and investigation of the optimum engineering of optical
view ports for gasifier TDL sensors. The Utah gasifier provided a unique test
environment with high-pressure, high-temperature syngas with significant particle
loading for the test of TDL sensors, which was not available at Stanford. Fundamental
TDL absorption strategies could be tested in a realistic gasifier environment.
Figure 3.1 University of Utah Gasification Research Facility.
34
3.2.1 Entrained-flow gasifier and sampling locations
A schematic diagram of the entrained-flow gasification system is shown in Figure 3.2.
Technical details of the gasifier are presented in Table 3.1. The heart of the system is a
20-cm diameter, 1.5-m long down-fired refractory-lined reactor (Figure 3.3). An injector
positioned at the top of the reactor uses oxygen to atomize a water-based slurry of
pulverized (~70 micron) coal. Five B-type thermocouples flush with the inner wall of the
refractory along the length of the reactor monitor the reactor temperature.
Figure 3.2 Schematic of entrained-flow gasification research facility.
35
Figure 3.3 Schematic to the pilot-scale, entrained-flow coal-gasifier at University of Utah
Table 3.1 Gasifier specifications.
Specification Units Typical Max Units Typical Max
Pressure psig 250 425 atm 18 30
Temperature °F 2600 3100 °C 1425 1700
Coal feed rate lb/hr (dry) 75 135 t/day (dry) 0.8 1.5
Thermal input MMBtu/hr 1.0 1.7 kWth 300 500
Slurry flow rate gal/hr 15 30 liter/hr 55 115
Slurry solids content wt% 59 65 wt% 59 65
The gasifier [30] can be broken down into four main sections as shown in Figure 3.4 and
Figure 3.5:
Section 1. Reactor core: Slurry of micronized (mean particle size ~ 70 µm, maximum
particle size ~ 100-120 µm) coal in water, as well as pure oxygen, are fed through the
36
nozzle at the top of this section. The water in the slurry rapidly vaporizes upon
introduction to the hot reactor core (1300K-1700K). The coal undergoes devolatilization,
pyrolysis and finally gasification. Section 1 hosts most of the partial oxidation reactions.
This region is characterized by extremely high particulate density and slag formation. Six
pairs of opposing sample ports along the length of the gasifier allow optical access across
the flow. The fourth set of ports, approximately 0.7m downstream from the injector, were
used for measuring H2O as described in [30]. This position is referred to as Location 1
in Figure 3.4 and Figure 3.5. About 90% of the coal - raw syngas conversion takes place
before the flow reaches Location 1. The main reactions that occur in Section 1 [48] are:
:12
,12
,12
: ↔ 2
: ↔
: 2 ↔
A typical composition of the gas phase at Location 1 (assuming equilibrium at 11 atm
and 1500 K) is 40% CO, 20% CO2, 20% H2, and 20% H2O.
Section 2. Pre-quench section: Below the reactor core, several flat spray nozzles inject
water into the flow to cool the products to about 600K-1000K. This rapid cooling
quenches the gasifier reactions and causes liquid slag to solidify. The gas composition
"freezes" corresponding to equilibrium at an intermediate temperature. During normal
operation, four spray nozzles are used, but for these tests two opposing spray nozzles
were removed and the remaining two nozzles were pointed downwards at angle of
roughly 30 degrees. The two empty ports were used to provide optical access into this
region of transition between the hot reactor and the water quench. This position is
indicated as Location 2 in Figure 3.4 and Figure 3.5. By the time the flow reaches
Location 2, most of the coal particles have reacted and the gas composition is governed
by the water-gas shift reaction:
: ↔ ∆ 41.1 /
37
A typical composition of the gas phase at Location 2 (assuming equilibrium at 11 atm
and 900 K, with introduction of N2) is 25% CO, 25% CO2, 25% H2, 10% H2O and 15%
N2. As a result of the reduction of temperature from Location 1 to 2, the equilibrium of
the water-gas shift reaction (exothermic) moves towards right, yielding increased
amounts of CO2 and H2.
Section 3. Post-quench section: The syngas, after being quenched to a colder temperature
by the water sprays in the previous section, bubbles out through a column of liquid water
and is then conveyed through a pipe, about 7.8 cm in diameter. At this location, the
temperature of the gas is in the range 340K-400K. In reference [30] agreement of 2-line
TDLAS based temperature measurements with the thermocouple readings was
demonstrated. For the measurements here, the gas temperature was determined by
thermocouples placed in the flow stream.
Section 4. Post-filtration section: The syngas is passed through a high-pressure candle-
style filtration unit to remove the particulates (unreacted coal, soot or ash) from the flow.
The temperature does not drop significantly between Section 3 and this location. The
difference is around 20 K. The overall temperature range is 320K-380K.
Figure 3.4 Schematic of the entrained flow gasifier at the Institute for Clean and Secure Energy at the University of Utah. The locations 1-4 are identified in the diagram. The sections are shown by green arrows.
38
TDLAS measurements of H2O and temperature were performed at Locations 1 and 2 and
reported in [30] using amplified diode lasers. The low power and less-robust fiber-
coupling of the lasers available in the 2 - 2.3 µm region limited the multi-species
measurements reported here to Locations 3 and 4. However, the multi-species monitoring
of syngas energy content is most useful at the gasifier exit.
3.2.2 System operation
The gasifier was operated on pulverized coal or liquid ethanol during the day and idled
on natural gas at atmospheric pressure during the night. Before the first feed into the
system, the reactor was heated with natural gas for approximately three days to ensure
that the refractory was thoroughly heated and to allow the system to come to thermal
steady state.
Figure 3.5 Schematic to measurement locations at University of Utah: (1) the reactor-core, (2) pre-quench,
(3) post-quench, (4) after clean-up
To prepare for introduction of coal, the natural gas burner was removed and the slurry
injection lance was installed. After a final safety check of all systems, the feed pump was
39
turned on to begin feeding fuel to the reactor. For startup, either ethanol or isopropyl
alcohol was used. Alcohol is much more combustible than the coal slurry and is used to
establish a flame and heat the reactor to the target temperature. Once it was confirmed
that fuel was flowing through the injector, oxygen flow was initiated at a flow rate
corresponding to a stoichiometry of roughly 0.5. Presence of a flame was confirmed both
by UV flame detectors and by a rise in reactor temperature. Shortly after the temperature
began to rise, the pressure-control valve was closed and the system was allowed to
pressurize to the target pressure. When the pressure reaches approximately 60 psi the
fuel was switched from alcohol to coal slurry. Significant production of soot when
feeding alcohol (by alcohol pyrolysis) was observed. The gasifier was then pressurized to
a target pressure and stabilized for gasification.
3.2.3 Lasers and control system
As illustrated in Figure 3.6, the sensors use four lasers for detection of CO, CO2, CH4 and
H2O, which were mounted in 14-pin butterfly packages and their outputs were pigtail-
fiber coupled. The transmission of extended-NIR light beyond 2 µm wavelength through
silica-based fibers suffers increased wavelength-dependent loss. Thus, fiber lengths were
limited to less than 2 m for the lasers detecting CO, CO2 and CH4. The lasers, along with
the laser current/temperature controllers and the data acquisition (DAQ) system, were
located about 1 m away from the optical access windows in the syngas output from the
gasifier. The sensor system was remotely controlled from a room ~30m away from the
high-pressure gasifier.
40
Figure 3.6 Schematic illustrating the components of the sensor including the controller and the DAQ in relation to the syngas exhaust pipe
3.2.4 Free space beam multiplexing
The four lasers operated at wavelengths ranging from 1350 to 2325 nm, which is beyond
the bandwidth of typical fiber combiners. Thus an important sensor component was the
design of a multiplexer illustrated in Figure 3.7, which was used to combine the four
beams into a single path through the syngas. The four beams were launched in parallel
through the syngas and collected by an aspheric uncoated CaF2 lens of 50 mm focal
length. These closely spaced beams were then focused onto a single extended-InGaAs
detector (Thorlabs). The entire optical assembly was rigidly mounted on an optical
breadboard attached to the syngas output plumbing to ensure stability of the optical
components.
41
Figure 3.7 Free-space quadruple beam multiplexing on a single detector
3.2.5. Optical access to the syngas
Laser transmission across the syngas required windows that can withstand the elevated
pressure, temperature and corrosive gas contact. The optical access for the TDL light
consisted of 2.5 cm thick sapphire windows (diameter of 2.5 cm and 1.27 cm at Locations
3 and 4 respectively) sealed by Teflon gaskets. Figure 3.8 shows the window mounting at
location 3. Sapphire was chosen as the window material due to its superior IR
transmission, physical strength and resistance to chemical attack. The thickness of the
windows ensured safe operation for pressures up to 100 atm. To avoid water
condensation on the windows, the window housing was electrically heated to maintain
the window temperature (about 150°C) well above the water vapor saturation
temperature.
Section 3 was the pre-filtered stage of the quenched syngas exhaust, where the gas was
saturated with water vapor with a considerable amount of suspended particulate. In the
filtration stage, most of the particles were removed from the flow before it passed the
windows at Location 4. The four sensors utilized transitions with varying degrees of
sensitivity to their respective concentrations [16]. The CO and H2O absorption was
stronger than that of the CO2 and CH4 owing to the product of transition linestrength and
relative concentration in the syngas. The absorption path length could be increased for
more measurement sensitivity. However, with the increase in the optical path length, the
42
laser beam extinction due to the particulates in the flow also increased. Therefore, for
CO2 and CH4 the maximum optical path length in Location 3 was ~8 cm without severely
degrading the signal-to-noise ratio from scattering losses, as the transmitted laser
intensity approached the lower limit for the available detectors. Note that lasers with
higher output power (here ~2mW) could be used to increase this path length. However,
for measurements at Location 4, where the extinction due to the particulates was greatly
reduced, the CO2 and CH4 detection used a longer optical path (17.9 cm) (See Figure
3.9(a) and Figure 3.9(b)). The large optical depth for CO and H2O absorption using the
selected transitions would limit the dynamic range with this longer path length, and these
two laser sensors continued to use the shorter path length.
Figure 3.8 Photograph of the optical access assembly in Location 3
(a) (b)
Figure 3.9 (a) Schematic and (b) Photograph of the optical access assembly in Location 4; time delay between the shorter (CO and H2O) and the longer (CH4 and CO2) path was less than 0.1 seconds.
43
3.3 Results and Discussion
The sensor performance was evaluated during a field measurement campaign at the Utah
gasifier. These results can be divided into two segments: 1) Wavelength-scanned
measurements to confirm interference-free species measurements and 2) Simultaneous
time-resolved multi-species concentration WMS measurements at fixed wavelength to
determine energy content of the gas.
3.3.1 Field validation of sensor performance
The coal gasifier presents an extremely harsh environment that has the potential to
produce unknown interfering species at the selected laser wavelengths. To investigate the
possibility of interference, a scan over a wide wavelength range was made for each laser
at each selected transition. Comparison of measurement and simulation of the shape of
the WMS lineshape would reveal any significant interference absorption. For example,
earlier measurements identified interference by NH3 in the gas mixture, which led to an
alternate selection of a CH4 transition (used here) devoid of this NH3 interference. Such
measurements were performed for CO, CO2, CH4 and H2O and compared to the expected
WMS lineshape as shown in Figure 3.10. Agreement of the 1f-normalized WMS-2f
lineshapes with the simulated spectra confirms the identity of the absorbing species,
yields a high signal-to-noise ratio of the obtained signals and verifies the absence of
significant interference for any of the selected transitions.
44
4300 4301 4302 43030.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.161f
-nor
mal
ize
d W
MS
-2f m
agn
itud
e [a
.u.]
Frequency [cm-1]
Measured spectrum Simulated spectrum
GC : 17.5 % COTDLAS : 18% CO
P = 11 atm, T = 350 KL = 6.35cm
(a)
4956.0 4956.5 4957.0 4957.5 4958.0 4958.50.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
GC: 42 % CO2
TDLAS : 44 % CO2
P = 11 atm, T = 350KL = 6.35 cm
Measured spectrum Simulated spectrum
1f-
norm
aliz
ed W
MS
-2f m
agn
itude
[a.u
.]
Frequency [cm-1]
(b)
7393 7394 7395 7396 73970.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
GC: N/ATDLAS: 6% H
2O
P = 11 atm, T = 370 KL = 8.3 cm
Measured spectrum Simulated spectrum
1f N
orm
aliz
ed
WM
S-2
f mag
nitu
de [a
.u.]
Frequency [cm-1]
(d)
Figure 3.10 Sample WMS spectra for (a) CO, (b) CO2, (c) CH4 and (d) H2O measured in the gasifier at 11 atm.
3.3.2 Simultaneous time-resolved multi-species concentration measurements
After the verification of the sensor performance, time-resolved multi-species
concentrations were monitored over extended durations during a gasifier run, with a time
resolution at Location 3 of 0.33 Hz and at Location 4 of 0.26 Hz. The fixed wavelength
WMS sensors operated at a modulation frequency of 10 kHz with the WMS parameters
as listed in Table 2.2.
The operating conditions of this entrained flow gasifier (EFG) produced CH4 at mole
fractions less than 1% of the syngas. However, CH4 is an important minor species in
syngas and gasifiers operating at lower temperature than the Utah facility, or processing
4365 4366 4367 4368 4369 43700.00
0.03
0.06
0.09
0.12
0.15 Measured spectrum Simulated spectrum
1f-
no
rmal
ize
d W
MS
-2f m
ag
nitu
de
[a.u
.]
Frequency [cm-1]
GC: 1.25% CH4
TDLAS: 1.3% CH4
P = 11 atm, T = 350 KL = 17.9 cm
(c)
45
biomass instead of coal, produce larger CH4 mole fractions. To test the CH4 sensor over a
wider dynamic range, controlled amounts of CH4 were injected into the flow. The
location of injection was sufficiently upstream of the measurement location to allow for
uniform mixing with the product syngas. The variation of the CH4 injection was made in
discrete time-steps to facilitate understanding of the sensor time response. The measured
CH4 mole fraction along with data from a GC (Varian CP-4900, repeatability < 0.5%
measured value) and a NDIR gas analyzer (California Analytical Instruments, Inc. (CAI),
Model ZRE) are displayed in Figure 3.11. The gas sampled for these analyzers was
further downstream of the optical access location, and the gas sample was depressurized
and dehumidified, which resulted in a lag in response relative to the laser-based in situ
sensors. As shown in Figure 3.11, there is a time lag of 4.2 minutes from Location 3 and
75 seconds from Location 4. Such delays can range up to 30 minutes in commercial scale
facilities [49]. In addition, the time resolution of the gas sampling suppressed any
variation in the syngas flow faster than 5 minutes. After these experiments were
complete, span-gas calibration of the NDIR analyzer found that a faulty sensor
calibration led to high readings. When the NDIR is properly calibrated, its results agree
with the GC. Thus, the agreement between laser and GC measurements suggests the
laser sensor provides good accuracy as well as much improved time resolution.
46
Figure 3.11 Time lag observed for traditional sensors vs. laser based in-situ sensors at (a) Location 3 and (b) Location 4.
Figure 3.12 displays the simultaneous measurements of CH4, CO, CO2 and H2O in
Location 3. For a comparison with the GC measurements (which can only analyze dry
gases), the water vapor mole fraction measured by using the TDLAS sensor was
subtracted from the rest of the TDLAS based measurements to convert them into an
equivalent dry basis. The measured data shows agreement of the laser sensor with the GC
measurements.
0 10 20 30 40 50 60 700123456789
101112 TDLAS measured data
GC measured data NDIR analyzer data
CH
4 mol
e fr
actio
n [%
]
Time [min]
Section 3P = 10 - 12 atmT = 350 - 370 K
(a)
0 5 10 15 20 25 30 35 40 45 500
2
4
6
8
10
12
14
CH
4 mol
e fr
actio
n [%
]
Time [min]
TDLAS measured data GC measured data NDIR analyzer data
Section 4P = 9 - 11 atm
T ~ 350 K
(b)
47
Figure 3.12 Time-resolved (dry basis except H2O) mole fractions of (a) CH4, (b) CO, (c) CO2 and (d) H2O and comparison with GC measurements (except H2O) in Location 3 at about 11 atm.
3.3.3 Estimation of lower heating value and Wobbe index
In the first part of the measurement as denoted by Region 1 in Figure 3.13, the amount of
doped methane was varied in discrete steps. The staircase-like pattern in CH4
measurements shows the laser sensor could successfully capture the rapid changes in
methane levels. At the same time, mole fractions of CO, CO2 and H2O were also
monitored as shown. Due to the purging of gasifier instrumentation, the syngas had a
nitrogen mole fraction of ~ 0.15, and the laser-sensor measured gas composition was
corrected for this N2 using the GC data. The variation in N2 mole fraction was less than ±
2% as shown in Figure 3.13. The H2 concentration was then inferred from the measured
major syngas components (CO, CO2, CH4, H2O, and N2).
0 10 20 30 40 50 60 70
0
1
2
3
4
5
6
7
8 Stanford measured data GC measured data
CH
4 m
ole
frac
tion
[%]
Time [min]
(a)
0 10 20 30 40 50 60 700
10
20
30
40
50 (b)
CO
mol
e fr
actio
n [%
]
Time [min]
Stanford measured data GC measured data
0 10 20 30 40 50 60 700
5
10
15
20
25
30
35
40
45
50(c)
CO
2 m
ole
frac
tion
[%]
Time [min]
Stanford measured data GC measured data
0 10 20 30 40 50 60 700
2
4
6
8
10
12
14 (d)
H2
O m
olef
ract
ion
[%]
Time [min]
48
Figure 3.13 Multispecies measurements at Location 3 at 11 atm; Box 1: varying CH4 addition, Box 2: varying coal slurry / O2 feed ratios.
The lower-heating value (LHV) of a fuel is a widely used parameter to compare the heat
release from the burning of different fuels. LHV is the amount of energy released when a
specific amount of fuel is burnt to completion at 25°C and 1 atm and the combustion
products are returned to the same temperature and pressure condition with water in its
fully vaporized state.
In general, gasifier syngas is primarily composed of CO, CO2, CH4, H2O and H2 along
with many trace species such as H2S, NH3, etc. The laser absorption sensors described
here can measure all primary components except H2, which is assumed to be the balance.
This assumption provides a path to infer the lower heating value (in MJ/Nm3) of the
syngas as:
24 ,, HCHCOi
ii LHVxLHV (3.1)
where xi is the mole fraction and iLHV is the lower heating value of the species i (volume
basis). Note the assumption of a H2 balance for our measurement is valid only for
oxygen-blown gasifiers which have low concentrations of fuel nitrogen in the syngas
stream. For air-blown systems or systems with significant nitrogen purge, N2 must also
be accounted for in the syngas mixture. Note, also for high sulfur coals, the heating value
of the H2S must also be included. The same goes for high nitrogen content coals, where
the NH3 content in the product stream must also be considered in the heating value.
0 10 20 30 40 50 60 700
5
10
15
20
25
30
35
40
45
50
2
CH4
H2O
CON
2
CO2
Mol
e fr
actio
n [%
]
Time [min]
H2
Section 3P = 10 - 12 atmT = 350 - 370 K
1
49
Another parameter of importance, when dealing with modern gaseous fuels is the Wobbe
Index (WI), which is a measure of interchangeability of fuels. It is expressed as:
s
OHHCHCOi
ii
s G
HHVx
G
HHVWI
224 ,,, (3.2)
where is the higher heating value of the species i and Gs is the specific gravity of
the gaseous fuel with respect to dry air at 25°C and 1 atm.
Using equations (3.1) and (3.2), the lower heating value and the Wobbe index of the
syngas were estimated. Comparison of the LHV of the syngas mixture inferred using the
major species compositions from the GC and the TDLAS-based measurements are shown
in Figure 3.14.
Figure 3.14 LHV of the syngas; Box 1: varying CH4 addition, Box 2: varying coal slurry / O2 feed ratios.
In the second part of the experiments as indicated by the region 2 in Figure 3.13, the ratio
of oxygen feed rate [kg/hr] to the coal slurry feed rate [m3/hr] was varied. These data
were collected to explore the use of TDLAS sensors to provide a control signal for an
optimum coal-to-oxygen flow rate ratio, crucial to the efficiency of the gasification
system. The CO and CO2 mole fractions in the region 2 along with the oxygen to coal
slurry feed ratio are plotted in Figure 3.15. With an increase/decrease of oxygen-to-coal
slurry feed rate, the extent of oxidation of coal increases, thereby increasing/decreasing
the product CO2 to CO ratio. An increase in oxygen feed rate triggers a rise in the core
0 10 20 30 40 50 60 704
5
6
7
8
9
Variation of O
2 - slurry
feed ratio
Stanford measured data GC measured data
Low
er h
eatin
g va
lue
[MJ/
Nm
3 ]
Time [min]
Section 3P = 10 - 12 atmT = 350 - 370 K
Methane dopedregion
1 2
50
reactor temperature (which may be desired on some occasions for the stability of the
reactor) but reduces the exergy content of the product syngas as manifested in a reduced
LHV and Wobbe index. Therefore, with these fast and robust TDLAS measurements of
the product syngas composition, the TDL sensor could provide a control variable for the
energy content of the syngas output.
Figure 3.15(a) Variation of the measured CO and CO2 mole fractions and (b) Variation of the inferred LHV with trends in oxygen - coal slurry feed ratio [kg/m3] and (c) Variation of the inferred Wobbe index with trends in oxygen - coal slurry feed ratio [kg/m3].
Similar measurements were also performed after the syngas filtration at Location 4 and
are shown in Figure 3.16. The decrease in non-absorption scattering losses at Location 4
improves the SNR of the measured species signals. But this improvement comes at the
loss of time response due to residence time in the filtration unit (~3 minutes). For
20
30
40
50
Mea
sure
d m
ole
frac
tion
[%]
CO CO
2
(a)
35 40 45 50 55 60
200
400
600
800
1000
1200
1400
CO2 rise
CO fall
O2/C ratio
fall
O2/C ratio
rise
CO2 fall
O2 to C slurry feed
ratio [kg / m3]
O2 t
o C
slu
rry
feed
rat
io [
kg /
m3 ]
CO rise
Time [min]
2400 2800 3200 36004
5
6
7
8
Variation of O
2 - slurry
feed ratio
Stanford measured data GC measured data
Low
er
hea
ting
va
lue
[MJ/
Nm
3]
Time [min]
Section 3P = 140 - 160 psigT = 350 - 370 K
2
4
6
8
10
12
O2 to C slurry feed
ratio [lbs / gal]
(b)
O2 t
o C
slu
rry
feed
rat
io [l
bs /
gal]
2400 2800 3200 36004
5
6
7
8
9
10
Variation of O
2 - slurry
feed ratio
Stanford measured data GC measured data
Wob
be I
ndex
[M
J/N
m3 ]
Time [min]
Section 3P = 140 - 160 psigT = 350 - 370 K
2
4
6
8
10
12
O2 to C slurry feed
ratio [lbs / gal]
(c)
O2 to
C s
lurr
y fe
ed r
atio
[lb
s /
gal]
51
example, although the CH4 feed was changed abruptly, the change in observed CH4
concentration in Location 4 was gradual as compared to the rapid response in Location 3.
Figure 3.16(a) Multispecies measurements at Location 4 at 11 atm, and (b) Measured LHV and Wobbe index of the syngas.
The LHV and Wobbe index of the gas mixture were also estimated as shown in Figure
3.16(b). The energy content of the syngas has an uncertainty of ± 5% as estimated by
propagation of errors from the mole fraction determinations — H2O (0.5%), CO (1%),
CO2 (2%), CH4 (0.05%), H2 (2.5%) and N2 (1%) [16]. As expected, these trends closely
follow each other closely. The measurements from the GC were not available during this
measurement sequence. The Location 4 offers improved species measurements requiring
minimal maintenance of the optical access at the cost of reduced time response. But as
these sensors mature, gasification systems may be optimized to minimize the time lag
between Locations 3 and 4.
3.4 Conclusions
The prototype TDLAS-based syngas composition sensor was first developed in the
laboratory at Stanford, and then performance-tested in a pilot-scale gasifier at the
University of Utah. Simultaneous multi-species measurements of CO, CO2, CH4 and H2O
mole fractions were then conducted in the gasifier at various operating conditions with a
time resolution of ~3s. These sensors were shown to be free of interference from other
species by the measurement of 1f-normalized WMS-2f lineshapes. A reliable solution for
multi-wavelength infrared optical access to the gasifier syngas product flow was
0 5 10 15 20 25 30 35 40 45 500
10
20
30
40(a)
CH4
COCO
2
N2
H2O
H2
Mol
e fr
actio
n [%
]
Time [min]
Section 4P = 9 - 11 atmT ~ 350 K
0 5 10 15 20 25 30 35 40 45 505.5
6.0
6.5
7.0
7.5
8.0
8.5
9.0
9.5
Section 4P = 9 - 11 atmT ~ 350 K
Low
er
He
atin
g V
alue
/ W
obbe
Ind
ex [
MJ/
Nm
3 ]
Time [mins]
Lower Heating Value Wobbe Index
(b)
52
described. The major species composition of the syngas (for O2-blown systems with
significant but nearly constant amount of N2) was determined from the measured CO,
CO2, H2O, a separate monitor for N2 (GC) and an assumption that the balance is H2.
When the important minor species CH4 was included in the sensor suite, the energy
content of the syngas could be determined within ± 5%. The LHV and Wobbe Index
determined in-situ in the wet syngas flow were in good agreement with dry values
determined from gas sampling and GC analysis. The trends in the rise and fall of the CH4,
CO and CO2 mole fractions correlate accurately with the physical changes in the gasifier
after minimal time lag. Thus the prototype sensor demonstrated here has good promise
for improved control of gasifier syngas quality.
53
Chapter 4 Application of scanned wavelength
modulation spectroscopy sensor for
simultaneous measurement of CO, CO2, CH4
and H2O in a high-pressure syngas output
stream from an engineering-scale transport
reactor gasifier
The contents of this chapter will be submitted to Fuel [18] under the title "Application of scanned wavelength modulation spectroscopy sensor for simultaneous measurement of CO, CO2, CH4 and H2O in a high-pressure syngas output stream from an engineering-scale transport reactor gasifier" and presented in the SPIE DSS14 Micro-Nanotechnology Sensors, Systems, and Applications Conference [50].
4.1 Introduction
The multi-species TDLAS sensors for coal gasifier applications have been optimized
following a series of previous field-test campaigns: (1) A 2011 campaign performed H2O
and temperature measurements in the syngas products from a pilot-scale (1 ton/day)
oxygen-blown down-fired gasifier at the University of Utah [30]. (2) A 2012 campaign
to Utah used a second-generation sensor for CO, CH4, CO2, and H2O in-situ syngas
measurements as discussed in Chapter 3. (3) A 2012/2013 campaign performed H2O and
temperature measurements in syngas from an engineering-scale (30,000 lb/hr syngas)
fluidized bed transfer gasifier at the National Center for Carbon Capture (NCCC) [49,51].
In this chapter, measurements using this optimized (fourth-generation) sensor for CO,
CH4, CO2, and H2O are reported from the 2014 measurements in the syngas output flow
from the NCCC gasifier. The novel contributions of this work are:
54
1. New optical design to combine multiple fiber-coupled laser-beams at wavelengths
beyond 2 µm onto a single optical path of about 5 meters.
2. New improved alignment strategies involving higher DOF stages and direct multi-
beam capture on the detector, resulting in reasonable signal levels throughout the 54-day
period after the initial cold alignment with remote access only to the sealed optical head.
3. Time-multiplexed system for multi-species time-resolved gas composition with 0.2 s
time resolution.
4. Large dynamic range sensor (for varying laser transmission) with remote detector gain
management for handling more than 300-fold change in signal strength for transmission
through 20 cm of gas effluent from startup to stable gasification.
5. Measurements that capture transient changes in product gas effluent as the process
changed from reactor heating, to both unstable (non-sustained) gasification and stable
gasification.
6. Measurements that indicate dynamic changes in synthesis gas products from the fuel
feed system (during stable gasification). These results show interesting dynamic behavior
of a practical transport reactor coal gasifier performance, unobservable by the
conventional extractive measurement approaches.
Table 4.1 Operating conditions for the scanned WMS sensors
Species Center frequency Scan range (200 Hz)
Modulation depth
(10 kHz) CO 4300.7 cm-1 6.63 cm-1 0.717 cm-1 CO2 4957.1 cm-1 4.37 cm-1 0.376 cm-1 CH4 4367.0 cm-1 4.42 cm-1 0.907 cm-1 H2O 7393.8 cm-1 6.77 cm-1 0.825 cm-1
This work utilized scanned-wavelength modulation spectroscopy (WMS) strategy for
measurement. In this method, the laser is modulated by sinusoidally varying injection
current which results in a time-varying intensity and frequency response. Specifically for
scanned WMS, a linear sum of a faster sinusoidal modulation signal and a much slower
sawtooth (or other similar waveform, e.g. triangle wave, sine wave, etc.) signal is fed to
55
the laser. The frequency of the lock-in amplifier is set corresponding to the faster
modulation signal. The slower ramp scan leads to a time-varying mean wavelength
resulting in a scan over wavelength space. This technique improves the SNR and gives
greater confidence over parameters including broadening [52]. For this current work, a
modulation frequency of 10 kHz is used on top of a linear ramp with a scanning
frequency of 200Hz for each of the four lasers operating at 2326, 2017, 2290 and 1352
nm for CO, CO2, CH4 and H2O detection operating sequentially in a time-demultiplexed
mode as illustrated in Figure 4.1. The operating WMS parameters for the four sensors are
listed in Table 4.1.
Individual chunks (time blocks of 5 ms) of laser-specific signals are then extracted,
analyzed and compared with simulated signals to obtain the individual species
concentration. A sample fit for the WMS waveform for CO2 detection is shown in Figure
4.2. The simulations show agreement with experimental lineshapes to within 4%. The
high signal-to-noise ratio at high signal attenuation increases confidence in our measured
species concentrations.
Figure 4.1 Illustration of the time demultiplexed collection of multi-laser signal (not actual data)
56
4956.5 4957.0 4957.5 4958.0
0.02
0.04
0.06
0.08
0.10
0.12
P = 8 atmT = 580 KX = 8%L = 20 cm
1f -
no
rma
lized
WM
S-2
f sig
nal
Wavenumber [cm-1]
Experiment Simulation
Figure 4.2 Sample fit of the scanned 1f-normalized WMS-2f spectrum for 8% CO2 at 8 atm and 580K
during coal feed
4.2 Sensor apparatus
4.2.1 Transport Reactor Gasifier at Wilsonville, Alabama
The in-situ multi-species sensors were used to measure CO, CO2, CH4 and H2O in an
engineering-scale transport reactor gasifier at the National Carbon Capture Center
operated for US Department of Energy by Southern Company in Wilsonville, Alabama.
Previously, a water vapor sensor was evaluated in this gasifier using telecom lasers as
reported in [49]. A detailed description of the transport reactor gasifier can be found in
the Appendix. The typical operating conditions at the gasifier exhaust measurement
location are listed in Table 4.2.
Table 4.2 Typical conditions at the gasifier exhaust
Property Value
Temperature 600 K Pressure 16 atm Path length (pipe diameter) 20 cm H2O mole fraction 0.06 - 0.12 CO mole fraction 0.08 - 0.12 CO2 mole fraction 0.06 - 0.10 H2 mole fraction 0.06 - 0.10 CH4 mole fraction 0.005 - 0.010 Trace species mole fraction (H2S, NH3, etc) < 0.01 Flow rate 12,500 kg/hr Flow velocity at measurement location 10 -15 m/s
57
The optical access in the syngas flow stream was set up approximately 30 m downstream
of the particulate control device (PCD) as illustrated in Figure 4.3. The syngas output
line indicated in the schematic is situated on the fifth floor of the gasifier structure and
several hundred feet from the base of the reactor. The data acquisition system and the
laser current and temperature controllers were housed in an instrument shelter about 100
feet away from the measurement site on the same floor.
The windows and the laser delivery and collection optics required for the sensor were
mounted on a flange attached on the syngas output line via a series of flanges, adapters
and bleed rings as shown in Figure 4.4. The optical setup incorporated several safety
features. The ball valve pair (one of them is for redundancy requirement) on each side
enabled isolation of the flow channel in case of a window failure. The windows, made of
sapphire, were 1” thick with 2° wedge angle, which were mounted onto standard 900lb
ANSI flanges and pressure tested to 1600 psig at 400°F. On each side of the syngas flow
channel, a pair of windows were mounted with a pressure and temperature monitor
between each window pair. In case of a window failure, the pressure and temperature in
between the two windows would rise, triggering an automatic shutoff of the ball valves.
The robust silica-based fibers (Thorlabs SM-2000) available at the extended NIR laser
wavelengths are lossy at 2.3 µm and hence short fibers (2 m) had to be used for light
delivery. Therefore, the lasers were kept in a N2-purged box (Figure 4.5) about 1.5 m
away from the laser pitch-side optical head. This serves two purposes: (i) isolate the
electronics that may be considered an ignition hazard in case of a gasifier leak, and (ii)
carry away the excess heat generated by the lasers. The lasers were controlled by the ILX
Lightwave LDC-3908 laser diode controller kept in the instrument shelter via 100 foot
controller cables.
58
Figure 4.3 Location of the optical access path in relation to the Gasifier
59
Figure 4.4 Schematic of the sensor setup in the syngas output flow channel.
Figure 4.5 Nitrogen purged multiple laser enclosure. The cables on the right connect the lasers to the
controller. The conduit on the left convey the 2 m optical fibers and the DB-25 cable for controlling the
control motor.
60
To enable simultaneous acquisition of the four laser signals, the four laser beams were
focused on a single 2mm x 2mm active area TE-cooled Hamamatsu detector after
travelling through 5 m of highly non uniform optical path. This was made possible
through intensive optical engineering as described in the following section.
4.2.2 Multi-laser beam multiplexing over 5 m path
Multi-beam multiplexing was achieved via a multi-core custom fiber bundle (four fiber
bundle) that consisted of a single Corning SMF-28e+ fiber (for the 1352 nm laser) and
three Thorlabs SM-2000 fibers (for >2 µm lasers). This custom bundling was done by
Neptec Optical Solutions, Fremont, CA. The four beams were launched using a 75 mm
plano-convex CaF2 lens attached to a telescopic mount for adjustment of the distance to
the fiber end as illustrated in Figure 4.6. The focal length of the lens was selected to
reduce the separation distance of the beams at a distance. The beam diameters however
were kept comparatively large to reduce beam steering artifacts. Then these overlapping
beams were collected on the large area detector by a 50 mm plano-convex CaF2 lens. The
detector had a switchable gain remotely controlled by a pair of gears connected to a
microprocessor controlled servo-motor. The pitch and catch side optics were mounted on
a remotely controllable pitch-yaw mount. There was also a manually controllable
miniature (because of space restrictions) x-y stage to accommodate flexibility in the x-y
location of the pitch/catch mounts. After the initial manual, cold alignment process, the
angular control of the pitch/catch optics was done fully remotely. This assembly was able
to maintain alignment for all four lasers from before the leak-test cycle until after the
gasifier shutdown, for a period of 54 days, enduring changes of over two orders of
magnitude in signal transmission as shown in Figure 4.7.
61
Figure 4.6 Multi-beam multiplexing hardware used for the gasifier sensors
2 4 6 8 10 12 14 160.1
1
10
100
GasificationPropane / Coal heat-up
Lower limit
Tra
nsm
itted
sig
nal [
%]
Gasifier operating pressure [atm]
Upper limit
Mean
Figure 4.7 Variation of measured signal strength with gasifier operating pressure
62
4.3 Results
During the R13 run of the gasifier, lasting for a total duration of 40 days, CO, CO2, CH4
and H2O concentration data were collected during the following phases of gasifier
operation: i. propane burner heat-up, ii. pulsed coal addition phase, iii. failed attempts at
gasification, iv. onset of final gasification phase, v. pressurization phase, vi. steady-state
running condition, and vii. final shutdown. All these events reveal unique features of the
gasifier operation via the measured simultaneous multi-species data showing agreement
with the measurements by a gas chromatograph in parallel, but revealing much more
transient variations in the species concentration, which were otherwise averaged-out by
the conventional measurement systems like the gas chromatograph (GC). These
measurement results are described in the following sections:
4.3.1 Propane burner heat-up
After the initial alignment was made, the pitch and catch side optical setup was kept
pressure sealed with N2. A steady N2 purge flow rate was maintained to keep syngas from
entering the optical access tube. Following initial pressurized leak checks, the gasifier
had to be heated up by burning propane in air. The sensors started measuring the four
species right from the time of ignition as shown in Figure 4.8. Since the gasifier started at
cold operating conditions, the initial water concentration was low. This can be attributed
to possible condensation (or surface adsorption) in cold spots along the long path of the
flow of gas. But as time passed, the water concentration steadily increased, stabilizing
towards equilibrium concentration of propane combustion. The mean concentration of
CO2 has a similar trend with the GC measurements; however it shows fluctuations in CO2
and water not captured by the GC. These transients can also be seen in the measured
temperature values. There is a 20-minute delay in the measured GC data which was
adjusted for one-to-one comparison in Figure 4.8. In addition, during the adjustments in
fuel flow rates, sharp spikes were observed in CO and CH4 measurements accompanied
by a sharp fall in CO2 and H2O concentration, indicating incomplete combustion during
these transition events.
63
56 60 64 68 72 76 800
2
4
6
8
GC CO2
CO CH4
H2O
Mol
e fr
actio
n [%
]
Time [hrs]
CO2
350
400
450
500
550
600
Tem
pera
ture
[K]
Figure 4.8 Simultaneous multi-species measurements during propane burner heat-up phase at 60 psig. The
propane burner is ignited at time 55.5 hrs, and peaks in the CO and CH4 are observed at 69.9, 74.7, and
75.1 hrs when the fuel flow is increased.
4.3.2 Pulsed coal addition during heat-up
After the gasifier is stabilized at elevated temperature with the propane flame, the coal is
gradually fed to the gasifier in small pulses in addition to the propane flame to raise the
temperature of the gasifier further. These pulses manifest in small fluctuations of CO,
CO2 and H2O mole fractions, as shown in Figure 4.9, with a time period of about 2
minutes. These fluctuations are not observable by the GC measurements, revealing the
low effective bandwidth characteristic of the extractive sampling process discussed
before. However, the mean values of the GC measurements agree with the TDLAS
measurements to within 1% of absolute mole fraction scale.
64
87.2 87.4 87.6 87.8 88.00
2
4
6
8
10
Mol
e F
ract
ions
[%]
H2O
CH4
CO
Time [hrs]
CO2
Figure 4.9 Pulsed fluctuations in CO, CO2 and H2O levels during initial parts of the coal-fed heat-up phase.
113.84 113.88 113.92 113.96 114.00
0
5
10
15
Mol
e fr
actio
n [%
]
Time [hrs]
60
80
100
120
140
GC : CH4CH
4
GC : COCO
GC : CO2
CO2
Pre
ssur
e [p
sig]
H2O
Figure 4.10 Pulsed fluctuations in species concentration observed at a later time during the coal-fed heat-up
phase.
As the temperature of the reactor becomes hot enough to sustain coal oxidation by itself,
the propane flame is turned off. As the flame is turned off, the fluctuations in coal feed
becomes much larger in amplitude (Figure 4.10) as coal is fed using manual control in a
batch process because the required coal feed rate is less than the minimum steady flow
rate.
65
4.3.3 Failed attempts at gasification
92.2 92.3 92.4 92.5 92.6 92.7 92.80
2
4
6
8
10
12
14
Fai
led2nd attempt
at gasification
Fai
led
Mol
e F
ract
ions
[%, r
ough
cal
cula
tions
]
Time [hrs]
H2O
CO2
CH4
CO
1st attemptat gasification
Figure 4.11 Failed attempts at gasification due to coal feed problems as captured through the multi-species
TDLAS measurements. The local gas conditions are 122 psig and 600K.
A major advantage of the TDLAS in-situ gas sensors is their fast time response. During
the first attempt at gasification, the coal feed had malfunctioned due to some blockage in
the hoppers. As a result, that attempt at gasification was unsuccessful. This transient
phenomenon was captured by the TDLAS sensors as seen from Figure 4.11. The brief
transition to the gasification phase is signaled by the plummeting CO2 and H2O levels,
accompanied by a steep rise in CO and CH4 mole fractions. The GC measurements, even
after adjustment of the time delay, was unable to capture this whole event. The peaks and
the troughs of the rise and fall of the measured species are smeared completely, giving a
very inaccurate documentation of the event. This shows that the TDLAS sensors can be a
valuable asset to assist in operation and control of gasifiers.
66
4.3.4 Onset of final gasification phase
114.0 114.5 115.0 115.5 116.00
5
10
15Pressurization
Mol
e fr
actio
n [%
]
Time [hrs]
Heat-up
On
set o
fga
sific
atio
n120
140
160
180
200
220
GC : CH4CH
4
GC : COCO
GC : CO2
CO2
Pre
ssur
e [p
sig]
H2O
Figure 4.12 A period of gasifier run showing three important sections of the gasifier run
A period of measurements showing important sections of the gasifier run is shown in
Figure 4.12. After solving the coal feed blockage issues mentioned in the previous
section, the coal gasifier was switched to gasification mode. This can be observed from
the sharp increase in CO and CH4 mole fractions and the decrease in the CO2 and H2O
mole fractions. It can be seen that the big pulses in CO, CO2 and H2O, characteristic of
the heat-up phase diminish right after the onset of gasification phase because the coal
feeder is within its standard operation range
4.3.5 Pressurization and gasifier stabilization phase
After the gasification began, the gasification system had to be stabilized at a higher
pressure operating condition conducive to efficient gasification. During this process, the
coal feed rate was also increased, leading to a stair-step-like rising pattern in CO and CH4
mole fractions as shown in Figure 4.13. The CO2 and H2O mole fractions remained
nearly constant during this process. The strong correlation between CO and CH4 mole
fractions exists because of the following interdependent chemical reactions:
67
: →
: 2 →
115.7 115.8 115.9 116.0 116.1
2
4
Mol
e f
ract
ion
[%]
Time [hrs]
120
140
160
180
200
220
GC : CH4CH
4
GC : CO
CO
Pre
ssur
e [p
sig]
Figure 4.13 Stepped increasing pattern in CO and CH4 levels during gasifier stabilization/pressurization
process.
4.3.6 Steady-state conditions
After pressurization and stabilization at a pressure of about 210 psig, the gasifier
operation was held steady at the same nominal operating conditions. During the early
stages of this phase, correlated oscillations in CO and CH4 were observed. But the
oscillations in H2O or CO2 were much smaller and hard to characterize, as shown in
Figure 4.14. But at a later time, even when the gasifier had run for 38 days, these
oscillations did not die down. Instead, more pronounced oscillations were observed in
CO2 and H2O that showed remarkable correlation with the measured temperature
fluctuations as shown in Figure 4.15. Interestingly, these fluctuations were anti-correlated
with the CO and CH4 concentrations indicating a direct manifestation of the fluctuating
coal-to-oxygen feed ratio.
68
121.8 122.0 122.2 122.4 122.6 122.80
3
6
9
12
15
GC: CH4
GC: CO2
GC: CO
CO2
CH4
H2O
Mol
e F
ract
ion
[%]
Time [hrs]
CO
190
195
200
205
210
215
220
Pre
ssur
e [p
sig]
Figure 4.14 Oscillatory behavior of CO and CH4 mole fractions in the early phase of steady state operations
930.3 931.0 931.7 932.4 933.1 933.80
2
4
6
8
10
12GC CO
GC CH4
TDLAS H2O
TDLAS CH4
TDLAS CO
GC CO2
Mol
e fr
actio
n [%
]
Time [hrs]
TDLAS CO2
1690
1700
1710
1720
1730
1740
1750
Tem
pera
ture
at t
he g
asifi
er e
xit [
oF
]
Temperature at the gasifier exit
Figure 4.15 Oscillatory behavior of measured species showing correlation of CO2 and H2O with measured
temperature fluctuations and an anti-correlation with CO and CH4 measurements
69
4.3.7 Gasifier shutdown
880 900 920 940 9600
3
6
9
12GC CO
GC CH4
TDLAS H2O
TDLAS CH4
TDLAS CO
GC CO2
Mol
e fr
actio
n [%
]
Time [hrs]
TDLAS CO2
Gas
ifie
r sh
utd
ow
n
Figure 4.16 Measurement of multi-species mole fractions from four days before the shutdown until the end.
Operating conditions: 220 psig, 630 K. At about 898 hrs, the GC sampling line was blocked and
maintenance to clear the line produced a fast change in the GC reading at that time. The gasifier feed was
unstable due to blockage in the coal feed line resulting in sharp changes in all the species concentrations at
around 879 and 928 hrs.
Multi-species TDLAS measurements for the last four days of the gasifier run are shown
in Figure 4.16. The transients in the species mole fractions were well captured by the
TDLAS sensors. The continuous measurements of the species concentrations were
interrupted by the sharp changes in the species mole fractions at around 879 and 928 hrs
due to blockage in the coal feed line and a maintenance operation performed to remove
blockage on the GC sampling line on a separate occasion, near 898 hrs. The moving
averages of these measurements agree within ±2% of the absolute mole fraction of the
GC measurements. Near the end, very large fluctuations in the species concentrations are
observed as expected from such a transient shutdown process.
70
4.4 Conclusions
Simultaneous in-situ scanned 1f-normalized WMS-2f based measurements of CO, CO2,
CH4, and H2O mole fractions were demonstrated for the first time in an engineering-scale
transport reactor gasifier. This work describes the optical engineering solutions necessary
to couple the multiple lasers on a single detector while achieving the required safety
standards of operating an optical access in the hot, toxic, corrosive, and pressurized
gasifier exhaust flow. The measurements were reported for all important sections of the
gasifier operation including propane heat-up, pulsed coal feed, failed gasification
attempts, onset of stable gasification, pressurization, steady-state operation, and shut
down during the R13 run of the NCCC gasifier lasting 40 days. Interesting dynamic
behaviors were observed via the TDLAS measurements which were otherwise
unobservable by conventional extractive sampling techniques with associated delays of
about 20 minutes.
71
Chapter 5 Development of high-sensitivity
interference-free diagnostic for measurement
of methane in shock tubes
The contents of this chapter have been submitted to Journal of Quantitative Spectroscopy and Radiative Transfer [53] under the title "High-sensitivity interference-free diagnostic for measurement of methane in shock tubes".
5.1 Introduction
CH4 is a stable and simple hydrocarbon that is commonly used as a fuel and can be
produced as a major intermediate species during combustion of other hydrocarbons.
Highly sensitive, high-bandwidth measurements of CH4 mole fraction would provide
important data in studies of reaction pathways and reaction rates in combustion kinetics.
Numerous articles have been reported in the literature for designing CH4 sensors by
utilizing the strongest transitions in the ν3 rovibrational band of the CH4 absorption
spectrum [6–14,54]. However, most of these sensors were designed for room
temperature applications. Pyun et al. [54] designed a DFG laser (generating light at
2938.24–2938.01 cm−1) sensor for high-temperature (1000-2000 K) CH4 over a pressure
range of 1.3 - 5.4 atm, for fast detection in a shock tube using a two-color peak-minus-
valley absorption subtraction technique. The absorption transitions were in the P-branch
of the CH4 ν3 band. This differential measurement strategy was based on the fact that
most of the absorbance from interfering species varies weakly over such small
wavelength ranges. By implementing their technique, interference-subtracted
measurements were possible, but the amount of interference from other hydrocarbon
species was significant, leading to increased uncertainty and reduced sensitivity for the
CH4 mole fraction determination. Other previous work included use of semiconductor
diode lasers at 1.65 and 2.29 μm [15–18,55–62], mid-IR difference-frequency-generation
(DFG) systems at 3.2–3.6 μm [6–14], interband cascade lasers (ICLs) at 3.3 μm [63,64],
72
a liquid-He-cooled, lead-salt diode laser at 7.4 μm [65] and a quantum cascade laser at
8.1 μm [66,67]. Some of these studies [15–18] were designed for temperatures suitable
for combustion exhaust applications (< 1200 K). However, none of these studies
addressed the possibility of interference from the commonly encountered intermediate
products of combustion at elevated temperatures (> 1200 K).
In this chapter, we present an improved sensor design based on transitions of CH4 in the
R-branch with significantly higher sensitivity and low interference from 35 common
combustion intermediate species/radicals. The currently selected transitions completely
avoid the strong absorption band due to the C-H stretch common to nearly all
hydrocarbons near 3.3 µm. The current work also demonstrates the use of a recently
developed DFB diode laser operating at 3.176 µm (Nanoplus [68]). Lastly, a sample
high-bandwidth and high-SNR measurement of CH4 is demonstrated during the pyrolysis
of C3H8 in a shock tube kinetics experiment.
5.2 Sensor design and selection of CH4 transitions
A cluster of CH4 ν3 1A1 → 1F2 R(14) A, F and E symmetry transitions near 3148.8 cm-1
was selected for the sensor. The selected transitions show the best performance from the
standpoints of interference and peak absorbance. Figure 5.1 shows simulations based on
the HITRAN 2012 [45] database for the R-branch of the ν3 band of CH4 at 900 K and
1400 K at 1 atm. Clearly, the selected region has a peak absorption coefficient among the
top four of the entire band for both of these temperatures, which is the approximate
temperature range of interest for the intended shock tube application of combustion
73
0
2
4
6
3000 3050 3100 3150 32000.0
0.5
1.0
1.5 1400 K
Abs
orpt
ion
coef
ficie
nt [c
m-1 a
tm-1]
900 K
Hot bands
Hot bands
R-branch
Q-branch
Q-branch
Frequency [cm-1]
R-branch
Figure 5.1 Absorption coefficient simulations of CH4 based on HITRAN 2012 at 1 atm and 900 K and 1400 K. The red arrow indicates the selected cluster of transitions. The hot bands, which are more pronounced at 1400 K, are marked by the orange circle.
chemistry studies. Figure 5.1 also shows the presence of hot bands in this region, which
becomes more distinct at higher temperatures.
Figure 5.2(a) shows the linestrengths of common combustion intermediates (computed
using HITRAN 2012 [45]) at 1400 K that absorb near this region. There are OH
transitions that interfere with the CH4 cluster at 3140 cm-1 and there is some interference
from the C2H2 bands at the CH4 clusters above 3157 cm-1. In addition, possible
interference from 25 other combustion species was studied using Fourier transform
infrared spectroscopy (FTIR) by Klingbeil et al [69,70]. Figure 5.2(b) shows the most
significant results of that survey at 773 K. This survey was used to select a suitable
region with minimum interference (absorption coefficient less than 5% that of CH4 at
3148.81 cm-1) from a large number of species. Out of the species studied, C2H4 produced
the highest interference (4% of the peak CH4 absorption coefficient at 773 K, also
discussed later) and some other species, shown in Figure 5.2(c), indicate 1-3% (but
spectrally flat) interference. A list of studied interfering species with low interference at
3148.81 cm-1 is given in Appendix A.5. Most previous work on high-temperature
diagnostics of methane [54,71,72] used CH4 transitions lower than 3000 cm-1. Clearly
from Figure 5.2(b), it can be seen that the interference from the higher hydrocarbons
(e.g., ethane, n-pentane and n-heptane) dominate the P-branch of the ν3 band. For a
74
2900 3000 3100 3200 3300 3400 35000.0
0.1
0.2
0.3
0.4
3135 3140 3145 3150 31550.00
0.05
0.10
0.15 C
2H
2
OH H
2O
CH4
Line
stre
ngth
at 1
400
K [
cm-2at
m-1]
Frequency [cm-1]
Selected region
{
(a)
Figure 5.2 (a) Linestrengths of common combustion intermediates from HITRAN 2012 near selected CH4
transitions, (b) Fourier transform infrared spectroscopy survey of some common combustion species at
773K and 1 atm (c) Expanded view of absorption coefficent of some species with 1-3% interference; all
plots are for 773K and 1 atm.
75
sensor designed to measure only CH4 in a combustion product environment, the reduced
interference near the selected absorption window provides a clear advantage over any line
previously studied in the P-branch (including the work using HeNe lasers [71]). In
addition, the R-branch peak is about 3 times stronger than the peak of the P-branch.
Although the interference near the selected transitions was minimal, it was found upon
further investigation that there is a low-but-measurable amount of interference at 3148.81
cm-1 from the hot bands of C2H2 and C2H4 that are not recorded in HITRAN 2012 [45] or
the FTIR surveys [69,70]. However, the spectrum of C2H2 and C2H4, measured by a
scanned direct-absorption method, shown in Figure 5.3, reveal flat absorption spectra in
sharp contrast to the structured CH4 spectrum. In such experiments where the interference
spectra is known to be flat with frequency, interference absorption can be subtracted
using a two-color technique, where one absorption measurement is performed at the
frequency corresponding to the peak methane absorption coefficient at 0.5-1.5 atm and
1000-2000 K (on-line, at 3148.81 cm-1), and another at a frequency well away from the
absorption peak (off-line, at 3148.66 cm-1). The methane mole fraction can then be
calculated using following relations:
expT (5.1)
offon
offonCH kkPL
1
4 (5.2)
where Tν is the fractional transmission at frequency ν, αν is the absorbance, P is the
pressure, L is the path length, k is the absorption coefficient of methane, and “on” and
“off” refer to on-line and off-line frequencies, respectively. It is important to note that
although the above method does not require accurate knowledge of the absolute
absorption coefficient of the interfering species, it does assume that their absorption
coefficients are equal at the two selected frequency values. The off-line frequency, νoff
was selected to be 3148.66 cm-1 to maximize sensitivity to CH4 while keeping the
variation in C2H2 and C2H4 absorption coefficients to a minimum (less than 0.5% of that
of CH4 peak absorption coefficient).
76
The next step in the development of the sensor was the measurement of kon and koff. Due
to the overlap of the symmetry multiplets of CH4 lines, there is an increased line mixing
observed at relatively low pressures (< 1 atm) in the absorption spectrum of R branch as
reported by Pine et al. [73] and Grigoriev et al [74]. Due to this and the presence of hot
bands, attempts at determination of the individual line spectroscopic parameters was
found to be forbiddingly difficult especially at higher temperatures where the deviations
were found to be large. To characterize these transitions for use in shock tube
experiments, absorption coefficient measurements were made with a 1% CH4 mixture in
Ar over a range of 0.2 – 4 atm and 500 – 2000 K in a shock tube.
Figure 5.3 Absorbances of CH4 and possible interfering species as measured by scanned direct absorption
method for a path length of 14.13 cm and location of the on-line and off-line shocks in wavenumber.
5.3 Measurement of absorption coefficient in Argon
5.3.1 Experimental set up
I. Shock tube
Absorption cross section measurements were made behind the incident and reflected
shock waves in a kinetics shock tube at Stanford. A detailed description of the shock
3148.5 3148.6 3148.7 3148.8 3148.90.0
0.1
0.2
0.3
0.4
0.5
0.6
Offl
ine
Abs
orba
nce
Wavenumber [cm-1]
1% C2H
2 in Ar, 2 atm, 1460 K
1% C2H
4 in Ar, 2 atm, 1200 K
1% CH4 in Ar, 1.1 atm, 1100 K
Onl
ine
77
tube operation for chemical kinetics experiments can be found in previous Stanford
studies [20,75–81]. Here we give a brief description of the shock tube method.
Shock tubes are a nearly ideal apparatus to study high-temperature gas properties due to
the ability to generate well-controlled temperatures and pressures. A gas-driven shock
tube consists of a long tube with driver and driven sections separated by a diaphragm
(typically a plastic film, e.g. Lexan) as seen in Figure 5.4. A pressure difference between
the sections causes the diaphragm to burst, and a shock wave is launched into the lower
pressure gas in the driven section, with a nearly instantaneous increase in the gas
temperature and pressure behind it. When the shock wave reaches the endwall, it is
reflected back toward the driver section, stagnating the gas and further raising its
temperature and pressure. The temperature increases are well-known from standard gas
dynamics relations and the measured shock speed. Accurate determinations of the
temperature and pressure (+/- 0.75%) are typically achieved, derived from incident shock
speed measurements. Uniform test times in the current experiments were approximately
2 ms long. Laser absorption measurements are done in the stagnated high-temperature
and pressure region behind the reflected shock wave, through a pair of ZnSe windows
located 2 cm away from the endwall, across the full width of the shock tube diameter
(14.13 cm).
Figure 5.4 Shock tube apparatus showing driver and driven sections (top), launch of the incident shock
wave (middle), and reflection of the shock wave from the endwall.
78
II. DFB diode laser at 3150 cm-1and ZBLAN fiber coupling
The DFB laser diode from Nanoplus GmBH was operated near 74 mA at 65°C to
maintain the laser frequency at 3148.81 cm-1. To effectively maintain the temperature of
the diode, a two-stage TEC control was implemented. The laser output was coupled into a
ZBLAN fiber to facilitate easier alignment and better beam quality management in the
shock tube (as previously demonstrated in harsh environments by Spearrin et al. [82]). A
9 µm core single mode ZBLAN fiber (1m long, IR Photonics S009S20FFP, NA = 0.2)
was used because of its low loss characteristics at 3150 cm-1. In the fiber-coupling
arrangement, the collimated laser light was focused by a microscope objective of 6 mm
focal length. The beam (< 2 mm in diameter) was launched from the other end of the
fiber and into the shock tube by a 12 mm focal length ZnSe AR-coated lens.
5.3.2 Measured absorption coefficients
The absorption coefficients of CH4 were measured at a pressure range of 0.2 – 4 atm
behind reflected shock waves in a 1% CH4 /Ar mixture (supplied by Praxair 99.9%). The
laser frequency was centered at 3148.81 cm-1 for on-line and 3148.66 cm-1 for off-line
measurements, set using a Bristol 721 wavelength meter (νuncertainty = ± 0.0035cm-1). This
procedure helped to mitigate uncertainty in the absorption coefficients due to drifts in
laser wavelength. A sample time-resolved voltage trace during the absorption coefficient
measurement is shown in Figure 5.5. The unattenuated laser signal (I0) was measured
with the shock tube test section in vacuum. The laser signal after absorption by CH4
follows the Beer-Lambert relation:
LPxTPkII
CH4),(log
0
(5.3)
where αν is the absorbance and kν is the absorption coefficient in cm-1atm-1 at frequency ν,
L is the path length in the shock tube in cm, P is the total pressure in atm, T is the gas
temperature in K and xCH4 is the mole fraction of CH4. From Figure 5.5, it can be seen
that the absorbance of CH4 varies with changes in the gas temperature and pressure
conditions across the incident and reflected shock waves. Small dips in signal at the
transitions are artifacts due to beam-steering effects related to passage of the shock wave.
79
Figure 5.5 Sample time-resolved trace of measured voltage signal during the absorption coefficient
measurements before and after a shock for 1% CH4 in Ar mixture.
The logarithms of the absorbance values obtained were fitted with a 2nd-order
polynomial function of ln(P) and ln(T). The expression for the absorption coefficient of
CH4 in Ar at 3148.81 cm-1 (on-line) is given by:
211 )ˆ(1216.0ˆ59.26ˆ7089.095.81exp(][ pTpP
Catmcmk on
on
))ˆ(136.2ˆˆ1836.0 2TTp (5.4)
Where 0ln ˆ PPp and 0ln ˆ TTT , with P0 = 1 atm, T0 = 1 K, and Con = 6.562 cm-1.
Figure 5.6 (a) shows the surface plot of the fitted equation of CH4 absorption coefficient
and the measured data. Figure 5.6 (b) shows the deviations in the fitted equation from the
measurements. All the samples show an error of less than 5%. The complexity of the
fitted expression (quadratic in logarithm space) is needed to account for the non-linearity
introduced by the existence of hot bands and the line-mixing effect.
-0.2 0.0 0.2 0.40
1
2
3
4 Non-absorbed laser signal (I0)
Incident shock, 0.56 atm, 574 K
Vol
tage
[V]
Time [ms]
Reflected shock, 1.68 atm, 917 K
1% CH4 fill,
0.13 atm,298 K,14.13 cm path
80
Figure 5.6 (a) Surface plot of absorption coefficient of CH4 in Ar at 3148.81 cm-1 (νon). The dots represent
the measured data points. (b) Surface plot of absorption coefficient of CH4 in Ar at 3148.66 cm-1 (νoff), (c)
Deviation of measurements and the fitted equation for kon at different temperatures and pressures. All
samples lie within 5% of the fitted equation. (d) Deviation of measurements and the fitted equation for koff
at different temperatures and pressures as a percentage of differential absorbance. All measurements lie
within 2.5%.
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0-10
-5
0
5
10
600 900 1200 1500 1800 2100-10
-5
0
5
10
500 - 1000 K 1000 - 1500 K 1500 - 2100K
Err
or in
fit
[%
of
diff
eren
tial
abs
orba
nce]
Pressure [atm] 0 - 0.5 atm 0.5 - 1 atm 1 - 4 atm
Temperature [K]
(c)
600 900 1200 1500 1800 2100-5.0
-2.5
0.0
2.5
5.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0-5.0
-2.5
0.0
2.5
5.0
Err
or in
fit
[%
of
diff
eren
tial
abs
orba
nce]
Temperature [K]
0 - 0.5 atm 0.5 - 1 atm 1 - 4 atm
500 - 1000 K 1000 - 1500 K 1500 - 2100 K
Pressure [atm]
(d)
81
Similarly, the absorption coefficient of CH4 at 3148.66 cm-1(off-line) was measured and
characterized for a similar range of temperature and pressures. The best-fit equation is
given by:
211 )ˆ(09973.0ˆ491.5ˆ298.578.14exp(][ pTpP
Catmcmk off
off
))ˆ(5435.0ˆˆ5933.0 2TTp (5.5)
where Coff = 7.077 cm-1. This equation agrees with the measurements with an error of less
than 2.5% of the differential absorbance. It must be emphasized that the on-line and off-
line wavelengths selected are optimized for 0.2 – 1.5 atm measurements. From the
simulations shown in Figure 5.7, it can be seen that these wavelengths, when used with
the two-color technique (Equation 5.2), give reduced sensitivity to CH4 concentration at
high pressure. The off-line wavelength selection needs to be revisited (e.g., moved to a
lower frequency) to extend the sensor to operate in high pressure.
3147.0 3147.5 3148.0 3148.5 3149.0 3149.50.000
0.001
0.002
0.003
0.004
0.005
0.006
1 atm
Off-line
Abs
orba
nce
Frequency [cm-1]
On-line
15 atm
HITRAN'12 simulation100 ppm CH
4 in N
2
1400 K, L = 14.13 cm
Figure 5.7 Location of the on-line and off-line measurement frequencies optimized for CH4 detection at 0.2 - 1.5 atm with respect to high pressure (15 atm) CH4 spectrum.
From the measurements shown in Figure 5.6, the minimum detectivity for the designed
CH4 sensor is calculated for a path of 14.13 cm (the diameter of the shock tube utilized at
Stanford) as shown in Figure 5.8. Note that this figure will change depending on the
noise floor of the system. For our measurements, the direct absorption RMS noise floor
appeared to be at an absorbance of 0.001 at 14.13 cm. Moreover, if the interfering species
82
identified have minor contributions, using a purely on-line measurement scheme would
be sufficient. This will improve the SNR of the data. Other techniques, such as
wavelength modulation spectroscopy (WMS) [33,52], have the potential to lower the
noise floor another order of magnitude, enabling sub-ppm detection of CH4, and still be
indifferent to flat interference spectra. Alternatively, cavity-based techniques [76,83]
can increase the path length and hence decrease the lowest detectivity inversely
proportional to the effective path length achieved.
Figure 5.8 Minimum detectivity of CH4 under different pressure and temperature conditions at 0.001
absorbance noise (SNR = 1) over a 14.13 cm path
5.4 C3H8 pyrolysis: A demonstration of the method
The performance of the CH4 diagnostic is demonstrated by an example measurement of
CH4 time-history during the pyrolysis of 1% C3H8 in Ar at 1763K, 1.64 atm. The
absorbances shown in Figure 5.9(a) were obtained using the two-color scheme from two
nearly identical shock tube measurements (ΔTinit < 10 K, ΔPinit < 0.02 atm). From these
data, the CH4 time history was calculated using equations (5.1), (5.2), (5.4) and (5.5), as
shown in Figure 5.9(b), whereas the interference absorbances from C2H2 and C2H4, the
primary fragments of C3H8 decomposition, were eliminated. It must be noted that the off-
line absorption measurement primarily comprises contributions from C2H4, C2H2 and
CH4 itself. The residual interference absorbance shown in Figure 5.9(a) consists of the
750 1000 1250 1500 1750 20001
10
100
1000
Min
imum
det
ectiv
ity [p
pm]
Temperature [K]
0.5 atm 1 atm 2 atm 4 atm
83
component of the off-line measurement that is purely from the interfering species (after
the contribution from CH4 is removed). This shows that there is a significant amount of
interference from other species, making it essential to perform off-line measurements.
The measured residual interference absorbance is compared to the simulated interference
absorbance in Figure 5.9(c), assuming the interference is entirely from C2H2, C2H4 and
C3H8. It can be seen that the predicted interference absorption using the USC Mech II
mechanism gives a closer match than the LLNL C1-C4 mechanism. The relative
contributions of C2H2 and C2H4 in the residual absorbance are in the ratio 13:10 at 1000
µs as calculated using the USC Mech II mechanism and measured absorption
coefficients.
-200 0 200 400 600 800 1000
0.00
0.05
0.10
0.15
Residual interference
Off-line
On-line
Abs
orba
nce
Time [s]
1% C3H
8 / Ar
T = 1763 K, P = 1.64 atmL = 14.13 cm
(a)
-200 0 200 400 600 800 1000
0.0
0.1
0.2
0.3
0.4
0.5
0.6
CH
4 m
ole
frac
tion
[%]
Time [s]
Measurement LLNL C
1-C
4
USC Mech II
1% C3H
8 / Ar
T = 1763 K, P = 1.64 atm
(b)
0 200 400 6000.00
0.02
0.04
Measured interference USC Mech II simulated interference LLNL simulated interference
Residual interference
Abs
orba
nce
Time [s]
1% C3H
8 / Ar
T = 1763 K, P = 1.64 atmL = 14.13 cm
(c)
Figure 5.9 (a) Absorbance time history [on-line (αon) and off-line (αoff)], (b) CH4 time history of 1% C3H8
pyrolysis in Ar at 1763K, 1.64 atm obtained using the two-color scheme, and (c) Measured residual
interference absorbance and comparison with simulated absorbance from LLNL and USC Mech II
mechanisms.
84
The uncertainty of the current measurement, mainly due to the uncertainty in the
temperature-dependent CH4 absorption coefficient, is approximately ±5%. The
temperature profile is obtained from chemical kinetics simulations using the USC Mech
II [84] and the LLNL C1-C4 [85] mechanisms, and the temperature time-history is found
to be mechanism-insensitive (less than 20 K difference in the plateau temperature
between the two models). The measured CH4 time-history is compared with the
numerical simulations using the two mechanisms. Both mechanisms yield results that are
seen to agree qualitatively with the current measurement. However, the USC Mech II
model under-predicts the initial formation rate, and the LLNL model over-predicts the
quasi-steady state value of CH4. With the help of this newly developed sensitive CH4
diagnostic, further kinetics experiments can be designed to resolve the discrepancies
between the two kinetic models.
5.5. Conclusions
A set of CH4 absorption transitions from the ν3 band near 3148.81 cm-1 were chosen for a
sensitive interference-free diagnostic of CH4 at high temperatures for shock tube
applications. The selected feature shows versatility over a large temperature range and in
addition has minimal interference from 35 studied species. The on-line and off-line
absorption coefficients of CH4 at 3148.81 cm-1 and 3148.66 cm-1 were characterized to
within 5% uncertainty between 0.2 and 4 atm and 500 K and 2100 K by shock tube
measurements. Interference absorptions, primarily due to C2H4 and C2H2, were
eliminated using the proposed two-color scheme. Best-fit equations for the temperature
and pressure dependence of CH4 absorption coefficients both on-line and off-line are
presented. Projected minimum detectivity at an absorbance detection limit of 0.001 was
reported. This high-sensitivity interference-free CH4 diagnostic was demonstrated in a
shock tube C3H8 pyrolysis experiment, confirming its potential for future chemical
kinetics studies.
85
Chapter 6 Summary and future opportunities
6.1 Summary
To summarize, a multi-species sensor suite for measurement of CO, CO2, CH4 and H2O
mole fractions was developed for use in a syngas environment (up to 20 atm) and
deployed successfully for the first time in two gasifiers of different capacity and
operating principles. 1f-normalized WMS-2f detection strategy (fixed wavelength at U.
Utah and scanned wavelength at NCCC) was implemented for these TDLAS sensors.
Firstly, the line selection was optimized for performance at high pressures and to
suppress interference from typical syngas composition. Then, a database of collision-
broadening coefficients was created for collisions with the set of species (CO, CO2, H2,
H2O, N2 and CH4) expected in syngas. The performance of these sensors was evaluated at
room temperature up to a pressure of 20 atm. The spectral simulations for the 1f-
normalized WMS-2f signals were validated against the measurements in both binary
mixtures with N2 and in multi-species synthetic syngas. The lower-heating value and the
Wobbe index were calculated from the sensor data and compared with the known values.
The inferred values were within 6% for the LHV and 8% for the Wobbe index over the
entire pressure range.
The sensors developed were then deployed in a pilot-scale gasifier at the University of
Utah. Simultaneous multi-species measurements of CO, CO2, CH4 and H2O mole
fractions were conducted in the gasifier at various operating conditions with a time
resolution of ~3s. Through the measurement of 1f-normalized WMS-2f lineshapes, these
sensors were shown to be free of interference from other species. A first-generation
solution for multi-wavelength infrared optical access to the gasifier syngas product flow
was described in Chapter 2. The major species composition of the syngas (for O2-blown
systems with significant but nearly constant amount of N2) was determined from the
measured CO, CO2, H2O, a separate monitor for N2 (GC) and an assumption that the
balance is H2. When the important minor species CH4 was included in the sensor suite,
the energy content of the syngas could be determined within ± 5%. The LHV and Wobbe
Index determined in the wet syngas flow were in good agreement with dry values
86
determined from gas sampling and GC analysis. The trends in the rise and fall of the CH4,
CO and CO2 mole fractions correlate accurately with the physical changes in the gasifier
after minimal time lag.
Then, with an improved scanned 1f-normalized WMS-2f strategy, simultaneous in-situ
measurements of CO, CO2, CH4 and H2O mole fractions were demonstrated for the first
time in an engineering scale transport reactor gasifier. This work describes the optical
engineering solutions necessary to couple the multiple lasers on a detector while
achieving the required safety standards of operating an optical access in the hot, toxic,
corrosive and pressurized gasifier exhaust flow. The measurements were reported for all
important sections of the gasifier operation including propane heat-up, pulsed coal feed,
failed gasification attempts, onset of stable gasification, pressurization, steady state run
and shut down during the R13 run of the NCCC gasifier lasting 40 days. The interesting
dynamic behavior observed via the TDLAS measurements, otherwise unobservable by
conventional extractive sampling techniques with associated delays of about 20 minutes,
can be used to capture the gasifier behavior for better gasification control and operation.
A set of CH4 absorption transitions from the ν3 band near 3148.81 cm-1 were selected for
a sensitive interference-free diagnostic of CH4 at high temperatures for shock tube
applications. The selected feature shows versatility over a large temperature range and in
addition has minimal interference from 35 studied species. Interference absorptions,
primarily due to C2H4 and C2H2, were eliminated using the proposed two-color scheme.
The on-line and off-line absorption coefficients of CH4 at 3148.81 cm-1 and 3148.66 cm-1
were characterized to within 5% uncertainty between 0.2 and 4 atm and 500 K and 2100
K by shock tube measurements. Best-fit equations for the temperature and pressure
dependence of CH4 absorption coefficients both on-line and off-line are presented.
Projected minimum detectivity of 6-700 ppm over a temperature range of 750-2000 K at
an absorbance detection limit of 0.001 (1σ) was reported. Finally, this high-sensitivity
interference-free CH4 diagnostic was exhibited in a shock tube C3H8 pyrolysis
experiment, reaffirming its promise for chemical kinetics studies.
87
6.2 Future opportunities
6.2.1 Additional minor species measurements in the fluidized-bed coal
gasifier in NCCC
The success of CO, CO2, CH4 and H2O measurements described in chapter 4 indicates the
possibility of other trace species including NH3, H2S and SO2 in practical systems such as
the fluidized-bed coal gasifier in NCCC.
6.2.2 Extension of the methane sensor for higher pressure
measurements in shock tubes
Measurement of CH4 time history at higher pressures (5-20 atm) is necessary to develop
reaction mechanisms for several combustion processes such as JP-8 pyrolysis and
oxidation. Further study is required to revisit the selection of on-line and off-line
wavelengths and measure absorption coefficients for higher pressures.
6.2.3 Higher sensitivity measurement of CH4 species time-history in
shock tubes using cavity enhanced absorption spectroscopy (CEAS) /
WMS
The detection sensitivity of the some species in shock tubes can be limited by the
dimension of the shock tube. For example, sub-ppm detection sensitivity of CH4 at 1MHz
measurement bandwidth is very hard to achieve in a shocktube with an inner diameter of
~10cm. The combination of CEAS and WMS can potentially be used to increase the
detection sensitivity of CH4 in a shock tube by orders of magnitudes, allowing improved
understanding of reaction kinetics.
88
Appendix
A.1 Description of PSDF, National Carbon Capture Center
(NCCC) gasifier
Figure A.1. 1. Block Flow Diagram of KBR TRIG Coal-to-SNG Process [86]
The Power Systems Development Facility (PSDF) is a large-scale test center located at
Southern Company Service’s (SCS) Clean Coal Research Center in Wilsonville,
Alabama. The Transport Gasifier is based on KBR’s fluid catalytic cracking
89
process [86,87], used in petroleum refineries for over 60 years to upgrade heavy oils to
transportation fuels. It has been developed as an alternative to fluidized-bed reactors and
can operate as either a combustor or a gasifier. A block diagram is shown in Figure A.1.1.
The coal feed top size is normally less than 500 microns (0.02 inches), slightly coarser
than that fed to PC boilers and very much coarser than that fed to entrained flow
gasifiers. The reactor has the even temperature characteristics of a fluidized bed but
operates at velocities 10 to 20 times higher. The higher velocity increases turbulence and
promotes intimate mixing of the coal with the reacting gases, resulting in high carbon
conversion. Further, the high velocity reduces the reactor diameter for a given air mass
flow rate, reducing reactor size and lowering capital cost. The highly turbulent
atmosphere is ideal for processing caking coals. Its operating characteristics are also well
suited to processing high-ash coals and fuels with low heating values, neither of which
are processed economically in current entrained-flow gasifiers.
A.1.1 KBR Transport reactor gasifier
A schematic of the Transport Gasifier installed at the NCCC is presented in Figure A.1.2.
Coal enters at the top of the mixing zone, and the air (or oxygen) and steam enters at the
bottom. The oxygen is consumed in the lower region of the mixing zone by the carbon in
the circulating solids. Hence, when the coal mixes with the heated solids and
devolatilizes in the upper region of the mixing zone, the volatiles do not burn but remain
to supplement the calorific value of the product gas. Staging of the oxidant is used to
control the heat release pattern.
The product gas from the riser passes through a disengager and cyclone, where the bulk
of the solids are removed. The separated solids pass from the disengager to the standpipe
and then through a J-valve to be returned to the mixing zone.
The J-valve is a non-mechanical valve that enables the solids to flow against the pressure
gradient between the disengager and the higher pressure environment of the mixing zone.
The solids separated by the cyclone are transferred through a loop seal and pass to the
standpipe along a downcomer. The loop seal is also a non-mechanical valve that allows
solids to be transferred from the cyclone to the higher pressure in the main standpipe. The
90
inventory of circulating solids is controlled by cooling and removing solids from the foot
of the standpipe.
Figure A.1. 2. Schematic of the Transport Gasifier at Wilsonville, Alabama [88]
Under normal operation the dusty syngas leaves the cyclone in the range 870 to 980°C
(1600 to 1800°F). It is cooled to about 370 to 430°C (700 to 800°F) in a fire-tube boiler
using water from a steam drum operating at 30 to 50 bar (400 to 700 psig). The syngas
then passes through the high temperature high pressure (HTHP) filter where metallic
filter elements remove entrained solids (outlet dust concentration is less than 0.1 ppmw).
The captured solids are cooled further before being depressurized and removed from the
filter vessel. The cleaned syngas is also further cooled before being depressurized and
burned in the thermal oxidizer. The flue gases produced are discharged to the stack.
When fired with air, the Transport Gasifier normally operates at 980°C and 16 bar
(1,800°F and 235 psia) with a coal feed rate of 2,300 kg/hr (5,000 lb/hr). The velocity in
91
the riser section is 12 to 18 m/s (40 to 60 ft/s) and provides about 1 to 2 seconds of
contact time for each pass of solids through the system. The riser has an inside diameter
of 23 cm (9 inches) and the overall reactor height is 24 m (78 feet). The inner diameters
of the standpipe and the downcomer are 25 cm (10 inches).
A.1.2 Particulate control device (PCD)
Gas exiting the syngas cooler at about 700°F flows through PCD, a proprietary filter
(designed by a suitable vendor) that removes remaining particulate matter as fine ash.
Removing fine particulates from syngas is an integral part of any gasifier system as it can
foul or corrode downstream equipment, reducing performance or causing equipment
failure.
Figure A.1. 3 Sketch of the Particulate Control Device (PCD) [86]
Figure A.1.3 shows a simplified sketch of the PCD employed in the TRIG gasification
system. It uses rigid, barrier-type, filter elements to remove essentially all of the fine
particulates in the syngas stream. The inlet solids concentration in the syngas to the PCD
is about 20,000 ppmw and is reduced to less than 0.1 ppmw upon exit. A small amount of
high-pressure recycled syngas is used to pulse-clean filters as they accumulate particles
from the unfiltered syngas. Downstream of each filter element, a safeguard (fail-safe)
92
device is installed to protect downstream equipment from particulate-related damage in
the event of an isolated filter element failure. The particulate stream (fine ash) is
depressurized to atmospheric pressure and removed via a proprietary continuous fine ash
removal system.
The PCD is a critical component of the TRIG gasifier development as it ensures the
syngas produced is particulate-free, eliminating dirty water or grey water systems that are
a feature of most other commercially available gasification processes. The elimination of
grey water systems also implies unique heat integration and water recovery possibilities.
KBR has developed proprietary technologies around the core TRIG unit to maximize heat
and condensate recovery. These novel features are incorporated in present coal-to-SNG
process scheme. Additional information on PCD and details on filter elements used can
be found in [89].
93
A.2 Window assembly design drawings for optical access in
NCCC, Alabama
The window assembly was designed to mate with a 300 lb - 2" flange installed in the
syngas output line on one side and the optical head connection plate on the other. Careful
construction ensured a straight optical access from the pitch end to the catch end. The
Solidworks drawings are shown in Figures A.2.1, A.2.2 and A.2.3.
Figure A.2. 1 Window assembly with adapter flanges
A.2.1 Parts list (All parts SS316)
1. ANSI 300lb-2” to ANSI 900lb-2.5” adapter (2ea)
94
2. Window flanges 1 and 2. (2ea)
3. 900lb Bleed ring with 4 X ½” NPT holes. (2ea)
4. ANSI 150lb-2.5” to ANSI 900lb-2.5” adapter. (2ea)
5. 150lb Bleed ring with 4 X ½” NPT holes. (2ea)
6. Window assembly studs – 1”/8 threaded rod X 13” long (16ea) (+ nuts/washer
32ea )
7. Studs for optics head – 5/8”/11 treaded rod X 4.5” long (8ea) (+ nuts/washer 8ea)
8. Valve connection – 5/8”/11 X 5.5” long (bleed ring here?) (16ea)(+ nuts/washer
16 ea)
9. ½” inch NPT nipple X 3” long (8ea)
10. ½” inch NPT nipple X 3.5” long (8ea)
11. Spiral wound 900lb - 2.5” gasket (8ea)
12. Spiral wound 150lb - 2.5” gasket (4ea)
13. Spiral wound 300lb - 2” gasket (2ea)
14. Bleed ring 300lb with 4 X ½” NPT holes (2ea)
95
96
97
98
A.3 Components of Optical heads
Figure A.3. 1 Photograph showing components of the pitch side optical head
Figure A.3. 2 Photograph showing components of the catch side optical head
99
A.4 Fiber optics coupling assembly for CH4 sensor
Figure A.4.1 Photograph showing fiber coupling assembly for CH4 sensor
Figure A.4.2 Photograph showing an example of alignment of the laser beam through a shock tube
100
A.5 List of studied species with absorption coefficients found to
be less than 5% that of CH4 at 3148.81 cm-1
Hitran 2012 [45] at 1400 K: CO, CO2, H2O, C2H4, HO2, C2H6, C2H2, OH, H2CO, N2O, NO2, O3
Klingbeil et al. [69,70] at 773 K: Ethanol, Formaldehyde, Benzene, 2-methyl-propane, Ethylene, 1-butene, 2-methyl-2-pentene, Ethane, Toluene, m-xylene, Ethyl-benzene, o-xylene, 3-ethyl-toluene, 2-methyl-butane, 2-methyl-pentane, 3-methyl-hexane, 2,2,4-trimethyl-1-pentane, 2-methyl-2-butene, cis-2-pentene, 1-heptene, 2,2,4-trimethyl-1-pentene, n-pentane, n-heptane, n-dodecane.
101
References
1. A. 39 Countries, "Climate Change 2013: The Physical Science Basis," http://www.climatechange2013.org/.
2. M. M. Joshi and S. Lee, "Integrated Gasification Combined Cycle— A Review of IGCC Technology," Energy Sources 18, 537–568 (1996).
3. T. F. Wall, "Combustion processes for carbon capture," Proc. Combust. Inst. 31, 31–47 (2007).
4. D. Santavicca, "Center for advanced power generation," http://www.capg.psu.edu/.
5. S. J. Clayton, G. J. Stiegel, and J. G. Wimer, "US DOE report," US DOE Rep. DOE/FE-044, (2002).
6. K. P. Petrov, S. Waltman, U. Simon, R. F. Curl, F. K. Tittel, E. J. Dlugokencky, and L. Hollberg, "Detection of methane in air using diode-laser pumped difference-frequency generation near 3.2 μm," Appl. Phys. B Laser Opt. 61, 553–558 (1995).
7. K. P. Petrov, S. Waltman, E. J. Dlugokencky, M. Arbore, M. M. Fejer, F. K. Tittel, and L. W. Hollberg, "Precise measurement of methane in air using diode-pumped 3.4-μm difference-frequency generation in PPLN," Appl. Phys. B Lasers Opt. 64, 567–572 (1997).
8. D. Richter, D. G. Lancaster, R. F. Curl, W. Neu, and F. K. Tittel, "Compact mid-infrared trace gas sensor based on difference-frequency generation of two diode lasers in periodically poled LiNbO 3," Appl. Phys. B Lasers Opt. 67, 347–350 (1998).
9. L. Goldberg, J. Koplow, D. G. Lancaster, R. F. Curl, and F. K. Tittel, "Mid-infrared difference-frequency generation source pumped by a 1.1–1.5 µm dual-wavelength fiber amplifier for trace-gas detection," Opt. Lett. 23, 1517 (1998).
102
10. K. P. Petrov, A. T. Ryan, T. L. Patterson, L. Huang, S. J. Field, and D. J. Bamford, "Mid-infrared spectroscopic detection of trace gases using guided-wave difference-frequency generation," Appl. Phys. B Lasers Opt. 67, 357–361 (1998).
11. K. P. Petrov, A. T. Ryan, T. L. Patterson, L. Huang, S. J. Field, and D. J. Bamford, "Spectroscopic detection of methane by use of guided-wave diode-pumped difference-frequency generation," Opt. Lett. 23, 1052 (1998).
12. D. G. Lancaster, D. Richter, R. F. Curl, and F. K. Tittel, "Diode lasers, DFG and molecules," OSA-TOPS Adv. Semicond. Lasers Their Appl. 31, 43–46 (2000).
13. U. Gustafsson, J. Sandsten, and S. Svanberg, "Simultaneous detection of methane, oxygen and water vapour utilising near-infrared diode lasers in conjunction with difference-frequency generation," Appl. Phys. B 71, 853–857 (2000).
14. H. Y. Clark, L. Corner, W. Denzer, G. Hancock, A. Hutchinson, M. Islam, R. Peverall, and G. A. D. Ritchie, "Difference frequency generation in periodically poled lithium niobate and its use in the detection of atmospheric methane," Chem. Phys. Lett. 399, 102–108 (2004).
15. V. Ebert, T. Fernholz, C. Giesemann, H. Pitz, H. Teichert, J. Wolfrum, and H. Jaritz, "Simultaneous diode-laser-based in situ detection of multiple species and temperature in a gas-fired power plant," Proc. Combust. Inst. 28, 423–430 (2000).
16. R. Sur, K. Sun, J. B. Jeffries, and R. K. Hanson, "Multi-species laser absorption sensors for in situ monitoring of syngas composition," Appl. Phys. B 115, 9–24 (2013).
17. V. Nagali, S. I. Chou, D. S. Baer, R. K. Hanson, and J. Segall, "Tunable diode-laser absorption measurements of methane at elevated temperatures.," Appl. Opt. 35, 4026–32 (1996).
18. R. Sur, K. Sun, J. B. Jeffries, R. K. Hanson, R. J. Pummill, T. Waind, D. R. Wagner, and K. J. Whitty, "TDLAS-based sensors for in situ measurement of syngas composition in a pressurized, oxygen-blown, entrained flow coal gasifier," Appl. Phys. B 116, 33–42 (2013).
19. R. Sur, K. Sun, J. B. Jeffries, R. K. Hanson, R. J. Pummill, and K. J. Whitty, "TDLAS-based in-situ monitoring of syngas composition from an oxygen-blown, down-fired, coal gasifier," in 8th U. S. National Combustion Meeting (2013).
20. R. K. Hanson, "Applications of quantitative laser sensors to kinetics, propulsion and practical energy systems," Proc. Combust. Inst. 33, 1–40 (2011).
21. P. Kluczynski, J. Gustafsson, Å. M. Lindberg, O. Axner, M. Lindberg, and O. A. U, "Wavelength modulation absorption spectrometry ᎏ an extensive scrutiny of
103
the generation of signals," Spectrochim. Acta Part B At. Spectrosc. 56, 1277–1354 (2001).
22. H. Teichert, T. Fernholz, and V. Ebert, "Simultaneous In Situ Measurement of CO, H2O, and Gas Temperatures in a Full-Sized Coal-Fired Power Plant by Near-Infrared Diode Lasers," Appl. Opt. 42, 2043 (2003).
23. J. Wolfrum, "Lasers in combustion: From basic theory to practical devices," Symp. Combust. 27, 1–41 (1998).
24. X. Chao, J. B. Jeffries, and R. K. Hanson, "In situ absorption sensor for NO in combustion gases with a 5.2μm quantum-cascade laser," Proc. Combust. Inst. 33, 725–733 (2011).
25. V. Ebert, H. Teichert, P. Strauch, T. Kolb, H. Seifert, and J. Wolfrum, "Sensitive in situ detection of CO and O2 in a rotary kiln-based hazardous waste incinerator using 760nm and new 2.3μm diode lasers," Proc. Combust. Inst. 30, 1611–1618 (2005).
26. R. R. Skaggs, R. G. Daniel, A. W. Miziolek, K. L. McNesby, C. Herud, W. R. Bolt, and D. Horton, "Diode Laser Measurements of HF Concentrations Produced from Heptane/Air Pan Fires Extinguished by FE-36, FM-200, FE-36 Plus APP, or FM-200 Plus APP," Appl. Spectrosc. 53, 1143–1148 (1999).
27. T. Güllük, H. E. Wagner, and F. Slemr, "A high-frequency modulated tunable diode laser absorption spectrometer for measurements of CO2, CH4, N2O, and CO in air samples of a few cm3," Rev. Sci. Instrum. 68, 230 (1997).
28. T. Zenker, H. Fischer, C. Nikitas, U. Parchatka, G. W. Harris, D. Mihelcic, P. Müsgen, H. W. Pätz, M. Schultz, A. Volz-Thomas, R. Schmitt, T. Behmann, M. Weißenmayer, and J. P. Burrows, "Intercomparison of NO, NO2 , NO y , O 3 , and RO x measurements during the Oxidizing Capacity of the Tropospheric Atmosphere (OCTA) campaign 1993 at Izaña," J. Geophys. Res. 103, 13615 (1998).
29. R. Sur, T. J. Boucher, M. W. Renfro, and B. M. Cetegen, "In Situ Measurements of Water Vapor Partial Pressure and Temperature Dynamics in a PEM Fuel Cell," J. Electrochem. Soc. 157, B45 (2010).
30. K. Sun, R. Sur, X. Chao, J. B. Jeffries, R. K. Hanson, R. J. Pummill, and K. J. Whitty, "TDL absorption sensors for gas temperature and concentrations in a high-pressure entrained-flow coal gasifier," Proc. Combust. Inst. 34, 3593–3601 (2013).
31. P. Ortwein, W. Woiwode, S. Fleck, M. Eberhard, T. Kolb, S. Wagner, M. Gisi, and V. Ebert, "Absolute diode laser-based in situ detection of HCl in gasification processes," Exp. Fluids 49, 961–968 (2010).
104
32. J. B. Jeffries, A. Fahrland, W. Min, R. K. Hanson, D. Sweeney, D. Wagner, and K. J. Whitty, "September 2009," Pittsburgh Coal Conf. (2009).
33. H. Li, G. B. Rieker, X. Liu, J. B. Jeffries, and R. K. Hanson, "Extension of wavelength-modulation spectroscopy to large modulation depth for diode laser absorption measurements in high-pressure gases," Appl. Opt. 45, 1052 (2006).
34. G. B. Rieker, J. B. Jeffries, and R. K. Hanson, "Calibration-free wavelength-modulation spectroscopy for measurements of gas temperature and concentration in harsh environments.," Appl. Opt. 48, 5546–60 (2009).
35. P. Kluczynski and O. Axner, "Theoretical Description Based on Fourier Analysis of Wavelength-Modulation Spectrometry in Terms of Analytical and Background Signals," Appl. Opt. 38, 5803 (1999).
36. X. Chao, J. B. Jeffries, and R. K. Hanson, "Wavelength-modulation-spectroscopy for real-time, in situ NO detection in combustion gases with a 5.2 μm quantum-cascade laser," Appl. Phys. B 106, 987–997 (2011).
37. G. B. Rieker, J. B. Jeffries, and R. K. Hanson, "Measurements of high-pressure CO2 absorption near 2.0 μm and implications on tunable diode laser sensor design," Appl. Phys. B 94, 51–63 (2008).
38. G. B. Rieker, H. Li, X. Liu, J. B. Jeffries, R. K. Hanson, M. G. Allen, S. D. Wehe, P. a Mulhall, and H. S. Kindle, "A diode laser sensor for rapid, sensitive measurements of gas temperature and water vapour concentration at high temperatures and pressures," Meas. Sci. Technol. 18, 1195–1204 (2007).
39. T. Fernholz, H. Teichert, and V. Ebert, "Digital, phase-sensitive detection for in situ diode-laser spectroscopy under rapidly changing transmission conditions," Appl. Phys. B Lasers Opt. 75, 229–236 (2002).
40. J. Reid and D. Labrie, "Second-harmonic detection with tunable diode lasers - Comparison of experiment and theory," Appl. Phys. B Photophysics Laser Chem. 26, 203–210 (1981).
41. R. Arndt, "Analytical Line Shapes for Lorentzian Signals Broadened by Modulation," J. Appl. Phys. 36, 2522 (1965).
42. K. Sun, X. Chao, R. Sur, J. B. Jeffries, and R. K. Hanson, "Wavelength modulation diode laser absorption spectroscopy for high-pressure gas sensing," Appl. Phys. B 110, 497–508 (2012).
43. X. Chao, J. B. Jeffries, and R. K. Hanson, "Absorption sensor for CO in combustion gases using 2.3 µm tunable diode lasers," Meas. Sci. Technol. 20, 115201 (2009).
105
44. J. Humlíček, "Optimized computation of the voigt and complex probability functions," J. Quant. Spectrosc. Radiat. Transf. 27, 437–444 (1982).
45. L. S. Rothman, I. E. Gordon, Y. Babikov, A. Barbe, D. Chris Benner, P. F. Bernath, M. Birk, L. Bizzocchi, V. Boudon, L. R. Brown, A. Campargue, K. Chance, E. A. Cohen, L. H. Coudert, V. M. Devi, B. J. Drouin, A. Fayt, J.-M. Flaud, R. R. Gamache, J. J. Harrison, J.-M. Hartmann, C. Hill, J. T. Hodges, D. Jacquemart, A. Jolly, J. Lamouroux, R. J. Le Roy, G. Li, D. A. Long, O. M. Lyulin, C. J. Mackie, S. T. Massie, S. Mikhailenko, H. S. P. Müller, O. V. Naumenko, A. V. Nikitin, J. Orphal, V. Perevalov, A. Perrin, E. R. Polovtseva, C. Richard, M. A. H. Smith, E. Starikova, K. Sung, S. Tashkun, J. Tennyson, G. C. Toon, V. G. Tyuterev, and G. Wagner, "The HITRAN 2012 molecular spectroscopic database," J. Quant. Spectrosc. Radiat. Transf. 130, 4–50 (2013).
46. M. P. Arroyo and R. K. Hanson, "Absorption measurements of water-vapor concentration, temperature, and line-shape parameters using a tunable InGaAsP diode laser.," Appl. Opt. 32, 6104–16 (1993).
47. M. W. Chase Jr., "JANAF-NIST Thermochemical Tables," J. Phys. Chem. Ref. data Monogr. No. 9, (1998).
48. C. Higman and M. van der Burgt, Gasification (Elsevier Inc., 2003).
49. K. Sun, R. Sur, J. B. Jeffries, R. K. Hanson, T. Clark, J. Anthony, S. Machovec, and J. Northington, "Application of wavelength-scanned wavelength-modulation spectroscopy H2O absorption measurements in an engineering-scale high-pressure coal gasifier," Appl. Phys. B (2014).
50. J. B. Jeffries, R. Sur, K. Sun, and R. K. Hanson, "Laser-absorption sensing of gas composition of products from coal gasification," SPIE DSS14 Micro-Nanotechnology Sensors, Syst. Appl. Conf. (2014).
51. K. Sun, "Utilization of multiple harmonics of wavelength modulation absorption spectroscopy for practical gas sensing," PhD Diss. Dec, (2013).
52. K. Sun, X. Chao, R. Sur, C. S. Goldenstein, J. B. Jeffries, and R. K. Hanson, "Analysis of calibration-free wavelength-scanned wavelength modulation spectroscopy for practical gas sensing using tunable diode lasers," Meas. Sci. Technol. 24, 125203 (2013).
53. R. Sur, S. Wang, K. Sun, J. B. Jeffries, D. F. Davidson, and R. K. Hanson, "High-sensitivity interference-free diagnostic for measurement of methane in shock tubes," J. Quant. Spectrosc. Radiat. Transf. TBD, TBD (2014).
54. S. H. Pyun, J. Cho, D. F. Davidson, and R. K. Hanson, "Interference-free mid-IR laser absorption detection of methane," Meas. Sci. Technol. 22, 025303 (2011).
106
55. Y. Shimose, T. Okamoto, A. Maruyama, M. Aizawa, and H. Nagai, "Remote sensing of methane gas by differential absorption measurement using a wavelength tunable DFB LD," IEEE Photonics Technol. Lett. 3, 86–87 (1991).
56. K. Uehara and H. Tai, "Remote detection of methane with a 1.66-microm diode laser.," Appl. Opt. 31, 809–14 (1992).
57. V. Weldon, P. Phelan, and J. Hegarty, "Methane and carbon dioxide sensing using a DFB laser diode operating at 1.64 μm," Electron. Lett. 29, 560 (1993).
58. H. Tai, K. Yamamoto, M. Uchida, S. Osawa, and K. Uehara, "Long-distance simultaneous detection of methane and acetylene by using diode lasers coupled with optical fibers," IEEE Photonics Technol. Lett. 4, 804–807 (1992).
59. D. C. Hovde, J. A. Silver, and A. C. Stanton, "Measuring atmospheric methane and water vapour using near-infrared diode lasers," in SPIE’s International Symposium on Optical Sensing for Environmental Monitoring (International Society for Optics and Photonics, 1994), pp. 110–117.
60. S. Schäfer, M. Mashni, J. Sneider, A. Miklós, P. Hess, H. Pitz, K.-U. Pleban, and V. Ebert, "Sensitive detection of methane with a 1.65 μm diode laser by photoacoustic and absorption spectroscopy," Appl. Phys. B Lasers Opt. 66, 511–516 (1998).
61. G. Stewart, A. Mencaglia, and W. Philp, "Interferometric signals in fiber optic methane sensors with wavelength modulation of the DFB laser source," J. Light. Technol. 16, 43–53 (1998).
62. T. Iseki, H. Tai, and K. Kimura, "A portable remote methane sensor using a tunable diode laser," Meas. Sci. Technol. 11, 594–602 (2000).
63. D. M. Sonnenfroh, R. T. Wainner, M. G. Allen, and R. K. Varner, "Interband Cascade Laser-Based Sensor for Ambient CH4," Opt. Eng. 49, 111118 (2010).
64. M. G. Allen, D. M. Sonnenfroh, M. B. Frish, and K. Parameswaran, "Low cost absorption sensors for networked applications," in Conference on Lasers and Electro-Optics (CLEO) (2011), pp. 1–2.
65. J. Reid, R. L. Sinclair, W. B. Grant, and R. T. Menzies, "High sensitivity detection of trace gases at atmospheric pressure using tunable diode lasers," Opt. Quantum Electron. 17, 31–39 (1985).
66. A. A. Kosterev, R. F. Curl, F. K. Tittel, C. Gmachl, F. Capasso, D. L. Sivco, J. N. Baillargeon, A. L. Hutchinson, and A. Y. Cho, "Methane concentration and isotopic composition measurements with a mid-infrared quantum-cascade laser," Opt. Lett. 24, 1762 (1999).
107
67. G. Gagliardi, F. Tamassia, P. De Natale, C. Gmachl, F. Capasso, D. L. Sivco, J. N. Baillargeon, A. L. Hutchinson, and A. Y. Cho, "Sensitive detection of methane and nitrous oxide isotopomers using a continuous wave quantum cascade laser," Eur. Phys. J. D - At. Mol. Opt. Phys. 19, 327–331 (2002).
68. "Nanoplus Nanosystems and Technologies GmBH," .
69. A. E. Klingbeil, J. B. Jeffries, and R. K. Hanson, "Temperature-dependent mid-IR absorption spectra of gaseous hydrocarbons," J. Quant. Spectrosc. Radiat. Transf. 107, 407–420 (2007).
70. A. E. Klingbeil, "Mid-IR laser absorption diagnostics for hydrocarbon vapor sensing in harsh environments," Stanford University (2007).
71. A. E. Klingbeil, J. B. Jeffries, and R. K. Hanson, "Temperature- and pressure-dependent absorption cross sections of gaseous hydrocarbons at 3.39 µm," Meas. Sci. Technol. 17, 1950–1957 (2006).
72. W. Ren, K.-Y. Lam, S. H. Pyun, A. Farooq, D. F. Davidson, and R. K. Hanson, "Shock tube/laser absorption studies of the decomposition of methyl formate," Proc. Combust. Inst. 34, 453–461 (2013).
73. A. S. Pine, "N2 and Ar broadening and line mixing in the P and R branches of the ν3 band of CH4," J. Quant. Spectrosc. Radiat. Transf. 57, 157–176 (1997).
74. I. M. Grigoriev, N. N. Filippov, M. V Tonkov, T. Gabard, and R. Le Doucen, "Estimation of line parameters under line mixing effects : the band of CH4 in helium," J. Quant. Spectrosc. Radiat. Transf. 69, 189–204 (2001).
75. Z. Hong, K.-Y. Lam, R. Sur, S. Wang, D. F. Davidson, and R. K. Hanson, "On the rate constants of OH+HO2 and HO2+HO2: A comprehensive study of H2O2 thermal decomposition using multi-species laser absorption," Proc. Combust. Inst. 34, 565–571 (2013).
76. K. Sun, S. Wang, R. Sur, X. Chao, J. B. Jeffries, and R. K. Hanson, "Sensitive and rapid laser diagnostic for shock tube kinetics studies using cavity-enhanced absorption spectroscopy.," Opt. Express 22, 9291–300 (2014).
77. S. Wang, D. F. Davidson, and R. K. Hanson, "High-temperature laser absorption diagnostics for CH2O and CH3CHO and their application to shock tube kinetic studies," Combust. Flame 160, 1930–1938 (2013).
78. S. Wang, D. F. Davidson, and R. K. Hanson, "High temperature measurements for the rate constants of C1–C4 aldehydes with OH in a shock tube," Proc. Combust. Inst. (2014).
108
79. M. A. Oehlschlaeger, D. F. Davidson, and R. K. Hanson, "High-temperature UV absorption of methyl radicals behind shock waves," J. Quant. Spectrosc. Radiat. Transf. 92, 393–402 (2005).
80. I. Stranic and R. K. Hanson, "Laser absorption diagnostic for measuring acetylene concentrations in shock tubes," J. Quant. Spectrosc. Radiat. Transf. 142, 58–65 (2014).
81. D. F. Davidson and R. K. Hanson, "Recent advances in shock tube/laser diagnostic methods for improved chemical kinetics measurements," Shock Waves 19, 271–283 (2009).
82. R. M. Spearrin, C. S. Goldenstein, J. B. Jeffries, and R. K. Hanson, "Fiber-coupled 2.7 µm laser absorption sensor for CO2 in harsh combustion environments," Meas. Sci. Technol. 24, 055107 (2013).
83. K. Sun, S. Wang, R. Sur, X. Chao, J. B. Jeffries, and R. K. Hanson, "Time-resolved in situ detection of CO in a shock tube using cavity-enhanced absorption spectroscopy with a quantum-cascade laser near 4.6μm," Opt. Express Submitted, (2014).
84. H. Wang, X. You, A. V Joshi, S. Davis, A. Laskin, F. Egolfopoulos, and C. K. Law, "USC Mech Version II. High-Temperature Combustion Reaction Model of H2/CO/C1-C4 Compounds.," .
85. N. M. Marinov, W. J. Pitz, C. K. Westbrook, A. M. Vincitore, M. J. Castaldi, S. M. Senkan, and C. F. Melius, "Aromatic and Polycyclic Aromatic Hydrocarbon Formation in a Laminar Premixed n-Butane Flame," Combust. Flame 114, 192–213 (1998).
86. S. Ariyapadi, P. Shires, M. Bhargava, and D. Ebbern, "KBR’s Transport Gasifier (TRIGTM) – An Advanced Gasification Technology for SNG Production from Low-Rank Coals," in Twenty-Fifth Annual International Pittsburgh Coal Conference (2008).
87. "KBR technologies," http://www.kbr.com/Technologies/Refining/Fluid-Catalytic-Cracking/.
88. J. M. Wheeldon, National Carbon Capture Center: 2010 Report (2010).
89. R. A. Yongue, X. Guan, R. S. Dahlin, and C. E. Landham, "Update on the Hot Gas Filtration Testing at the Power Systems Development Facility," in 32nd International Technical Conference on Coal Utilization & Fuel Systems, Clearwater, Florida (2007).