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).