Upload
doanthuy
View
224
Download
0
Embed Size (px)
Citation preview
Tunable Diode Laser Sensors for Monitoring Combustion and Gasification Systems
Jay B. Jeffries*, Ronald K. Hanson*, and Kevin Whitty***High Temperature Gasdynamics Laboratory, Stanford University g e pe a u e Gasdy a cs abo a o y, S a o d U e s y
**Institute for Clean and Secure Energy, The University of UtahDoE/EPRI Workshop on Instrumentation June 2012
1. Fundamentals of TDL absorption sensing2. TDL sensing for coal gasification
T and H O sensing* T and H2O sensing* CO, CO2, CH4, H2O sensing**
3. Monitor of syngas heating value
*Sponsored by EPRI (Jose Marasigan & Jeff Phillips)**Sponsored by DoE NETL (Susan Maley)
My main message today is that:TDL Absorption is Practical in Harsh Environments
Utilizes economical, robust and portable TDL light sources and fiber optics Can yield multiple properties: species, T, P, V, & m in real-time over wide conditions
T to 8000K, P to 50 atm, V to 15km/sec, multiphase flows, overcoming strong
.T to 8000K, P to 50 atm, V to 15km/sec, multiphase flows, overcoming strong emission, scattering, vibration, and electrical interference
Demonstrated in harsh environments and large-scale systems: Aero-engine inlets, scramjets, pulse detonation engines, IC engines, gas turbines g j p g g g
arcjets, shock tunnels, coal-fired combustors, rocket motors, furnaces…. Potential use in control of practical energy systems
IC-Engines @ NissanCoal Gasifier @ U of UtahCoal-fired Utility Boiler
IC-Engines @ Nissan
J ff i Pitt b h C l C f 2011
2
Jeffries, Pittsburgh Coal Conf, 2011
Chao, Proc Comb Inst, 2011 Jeffries, SAE J. Eng, 2010
Absorption Fundamentals: SpeciesAbsorption of monochromatic light
• Scanned-wavelength line-of-sight direct absorptionB L b t l ti I• Beer-Lambert relation
• Spectral absorption coefficient
LnLkII
io
t exp)exp(
PPTTSk )()(absorbance
• Spectral absorption coefficient PPTTSk ii ),,()(
3
Absorption Fundamentals: Velocity
• Shifts & shape of contain information (T,V,P,i)
4
Absorption Fundamentals: Temperature
• T from ratio of absorption at two wavelengths
5
Absorption Fundamentals: Multiplexed
• Wavelength multiplexing is often utilizedWavelength multiplexing is often utilized• To monitor multiple parameters or species• To assess non-uniformity along line-of-sight
6
TDL Sensors Provide Access to a Wide Range of Combustion Species/Applications
S ll i h Small species such as NO, CO, CO2, and H2O have discrete rotational transitions in the vibrational bands
Larger molecules, e.g., hydrocarbon fuels, have blended features
77Two primary TDLAS sensor strategies
Two Absorption Measurement Techniques:Direct Absorption (DA)
Baseline
& Wavelength Modulation Spectroscopy (WMS)
Direct Absorption
Injection current tuning
0.25
0.50
0.75
Abs
orba
nce
D irect absorption lineshape
0.1
0.2
0.3 Direct Absorption Scan
r Int
ensi
ty S
igna
l Baselinefit
for Io
Gas sample
Io Iti 0.4 0.5 0.6 0.7
0.00
A
W avelength (re lative cm -1)
0.4 0.5 0.6 0.7 0.8 0.9 1.00.0La
ser
T ime(ms)
0.04
gnal8
l
WMS+ Injection current
modulation @f
i’ Lockin @1f, 2f
0.00
0.02
orm
aliz
ed 2
f si
2
4
6
WM
S S
igna
Direct absorption Method of choice hen applicable
@-0.02 WMS-2f lineshapeN
o
Wavelength (relative cm-1)0.4 0.5 0.6 0.7
0WMS Scan
Time (ms)0.4 0.5 0.6 0.7 0.8 0.9 1.0
Direct absorption: Method of choice when applicable WMS: More sensitive especially for small signals (near zero baseline)
WMS with TDLs improves noise rejection 1f-Normalized WMS-2f/1f: Provides Io without a baseline
8
High P,T Sensing Enabled by WMSSimulated Absorbance SpectraSimulated Absorbance Spectra
2500 K12% H2OL = 4 cm
High P, T challengesB d d bl d d t t hi h P
25 atm1 atm
Broad and blended spectra at high P Decreased absorbance at high T
Simulated WMS Spectra2500 K
Solution 1f-Normalized WMS-2f
25 atm
2500 K12% H2O
Recovers strong peaks No baseline Io needed! Also suppresses noise and pp
transmission losses
9
WMS-2f/1f Accounts for Non-Absorption LossesPitch Lens Fixed WMS 2f/1f
0 061392 nm, Partially Blocking Beam 1392 nm, Vibrating Pitch Lens
Modulated TDL near 1392nm
Pitch LensDetector
Fixed WMS-2f/1fAmbient H2O (T=296 K, 60% RH) L=29.5 cm, ~6% absorbance)
0.6
ude
0.00
0.02
0.04
0.06
2f M
agni
tude
2f Magnitude
0.6
ude
0.00
0.02
0.04
0.06
2f M
agni
tude
2f Magnitude
0 06
0.09
0.12
agni
tude
0.0
0.2
0.4
1f M
agni
t
2f/1f
1f Magnitude
0 06
0.09
0.12
Mag
nitu
de
0.0
0.2
0.4
1f M
agni
tu
1f Magnitude
Demonstrate normalized WMS-2f/1f in laboratory air
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.350.00
0.03
0.06
2f/1
f Ma
Time (s)
2f/1f
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.70.00
0.03
0.06
2f/1
f M
Time (s)
2f/1f
Demonstrate normalized WMS-2f/1f in laboratory air 2f/1f unchanged when beam attenuated (e.g., scattering losses) 2f/1f unchanged when optical alignment is spoiled by vibration
WMS-2f/1f signals free of window fouling or particulate scatteringWMS has other advantages too
10
Sensing with Large Transmission Losses from Scattering Enabled by WMS
Transmission of laser light at non-absorption wavelengths
M t i d t li b f ti l t filt i Measurement in syngas product line before particulate filtering Particulate loading increases with pressure (99.9% loss at 150psig)
Varies with gasifier performance, fuel, temperature, etc.
11
Solution: Stanford’s 1f-normalized WMS-2f scheme
What might we measure in syngas?
Vision and Goals for TDL Sensing in IGCC
X
Oxygen or Steam
Sensor for T & SyngasComposition Fuel/air Ratio
GasifierX
Oxygen or Steam
Sensor for T & SyngasComposition Fuel/air Ratio
GasifierX
Oxygen/CoalRatio Control
Composition
3 4
Fuel/air Ratio Control1Reactor
Core2
Syngas
XOxygen/CoalRatio Control
Composition
3 4
Fuel/air Ratio Control1Reactor
CoreReactor
Core2
SyngasX
Coal
QuenchSyngasCleanup Gas TurbineX
Coal
QuenchQuenchSyngasCleanup Gas Turbine
Sensor for control signals to optimize gasifier output and gas turbine input Two flow parameters considered: gas temperature and heating value
Vision:Goals:
Gas temperature determined by ratio of H2O measurements Measurements of CO, CH4, CO2, and H2O provide heating value
H2 determined by gas balance as other species ignored
1212
2 y g p g Four measurement stations considered: spanning reactor core to products
Oxygen-blown, Down-fired, Entrained-flow Coal Gasification Facility at the University of Utah
Rated to 450 psig current data to 200 psig
Pilot scale gasifier
current data to 200 psig Rated to 3100 °F Coal throughput: 1 ton/day Overall dimensions
5.1 m (17’) tall0.76 m (30’’) OD
Reactor dimensions1.5 m (60’’) long 0.20 m (8’’) ID
Four measurement campaigns to test Stanford TDL sensors: Four measurement campaigns to test Stanford TDL sensors: Aug. 2010, Dec. 2010, Aug. 2011, May 2012
Ideal facility for instrumentation testing:R id t iti f 1 t fl t 20 t ifi ti diti
13
Rapid transition from 1 atm flame to 20 atm gasification conditions Reactor kept hot with 1 atm natural gas flame between runs
Sensor Setup in Utah Gasifier: T and H2O
Two reactor locations tested Position 1: Reactor core
P iti 2 Q h l ti
N2 Purge
Gasifier reactor
N2 Purge
Slag
1 Position 2: Quench location
12.5cm
CollimatorIrisFocusing mirror
Optical filter
1
20cm
Sapphire window
2
Detector34cm
Liquid H2Oquench
14
Sensor Setup in Utah Gasifier: T and H2O
Two reactor locations tested Position 1: Reactor core
Hi h t T
N2 Purge
Gasifier reactor
N2 Purge
Slag
1 Highest T Largest scattering losses Emission interference
12.5cm
CollimatorIrisFocusing mirror
Optical filter
1
Time limited by slag flow Successful measurements
demonstrated
20cm
Sapphire window
2
Detector34cm
Liquid H2Oquench
Laser current
Diode Lasers1352, 1347 nm
Laser
1×2 fibercombiner
15Control Room
Diode Lasers1392, 1469 nmDAQ PC
Laser controller 1×2 fiber
combiner
Temperature in Reactor Core
Transmission at 50 psig 0.13% dropping to 0.02% at 150 psig Normalization scheme successful Very strong optical emission - optical filtering scheme successful
Optical access tube successfully stayed open in presence of flowing slag’p y y p p g g Later unsuccessful with different coal (and different atomizer)
Temperature in Reactor Core
Reactor CoreIncrease O2
DetectorFiber to Control Room
Normalization scheme successful with low transmission (< 0.02%) TDL sensor time response can capture flow changes TDL sensor time response can capture flow changes
17
Sensor Setup in Utah Gasifier: T and H2O
Two reactor locations tested Position 2: Quench location
M d t fl k
N2 Purge
Gasifier reactor
N2 Purge
Slag
1 Modest purge flow keeps
windows clean Lower T – different line pair 12.5cm
CollimatorIrisFocusing mirror
Optical filter
1
Amplifier available Increase power x10
Successful measurements
20cm
Sapphire window
2
even with 10-5 attenuationDetector
34cm
Liquid H2Oquench
Laser current
Diode Lasers1352, 1347 nm
Laser
1×2 fibercombiner
Fiber amplifier
18Control Room
Diode Lasers1392, 1469 nmDAQ PC
Laser controller 1×2 fiber
combiner
Temperature @ Quench Location
Normalized WMS accounts for varying transmission (10-3 at 160 psig)y g ( p g) Measured T at reactor pressures of 90, 120 and 160 psig stable Measured T at 200 psig identifies potential fuel/O2 input instabilities
Temperature @ Quench Location
Location 2, P ~ 200psigTransmission~ 7 X 10-6Transmission 7 X 10
Different gasifier conditions, different coal, more particulate scattering High SNR, time-resolved measurements of T using fiber amplifier
20
g , g p Less than 10-5 of the laser light transmitted
Sensor Setup in Utah Gasifier: Syngas Composition
CO, CO2, and CH4 lasers use lasers 2-2.3 m
21
Fiber technology less available TDLs controlled remotely but located near measurement
Sensor Setup in Utah Gasifier: Syngas Composition
Syngas can by pass sensor location for window maintenance Syngas can by-pass sensor location for window maintenance Similar setup before and after particulate filter (similar results) Multiple-lasers directed through one window
Rapid (10 Hz) switching from one species to another Time-resolution ~1/3 second
22
TDL Sensor Measured Syngas Composition
CH4 addition Vary O2/coal feed rates
Laser absorption measurements of CO, CO2, H2O and CH4 over 1 hour CH4 added to syngas to test sensor response and vary gas composition
23
Gasifier feed rates changed to test sensor response
Syngas Composition Including N2 and H2
N2 in flow from gas purges – determined by metering and GC data Assume the rest of the syngas is H2
24
Enables determination of lower heating value (LHV)
Time-Resolved Monitor of Syngas LHV
One hour time record of syngas lower heating value (LHV) CO, CO2, CH4 and H2O from TDL sensor and N2 from facility data
Assume balance of syngas H
25
Assume balance of syngas H2
LHV contribution of small concentrations of H2S and NH3
are estimated to be less than 2% (accounted as H2)
Summary
A novel modulation strategy enables measurements in high pressure environments with extinction by scattering Scheme validated for extinction as large as 105 Scheme validated for extinction as large as 10
Sensor demonstration measurements made in four locations of a pilot-scale, entrained-flow, coal gasifier Time-resolved measurements capture small changes in gasifier Time-resolved measurements capture small changes in gasifier
operating conditions Current work focused on sensor validation and demonstrationNe t StepsNext Steps: Transition sensor to real-time for continuous unattended monitoring Add H2S and NH3 to sensor suite Package next-generation sensor for industrial-scale applications (test Utah?) Find suitable industrial-scale demonstration opportunities
A k l d t St f d PhD t d t K i S d Rit SAcknowledgements: Stanford PhD students Kai Sun and Rito SurProfessor Kevin Whitty at University of UtahSusan Maley @DoE; Jeff Phillips and Jose Marasigan @EPRI26