Post on 21-Jan-2018
Analytical Solutions by the SeaThe 60th Annual Symposium of the Analysis DivisionGalveston Island Convention Center, Galveston, Texas, USA; 26th – 30th April 2015
Session 10 / Paper No.2Maintaining TDL measurement accuracy in high temperature
applications
Jean-Nic Adami, Mettler-Toledo AG
Analytical Solutions by the SeaThe 60th Annual Symposium of the Analysis DivisionGalveston Island Convention Center, Galveston, Texas, USA; 26th – 30th April 2015
Slide 2
Outline
TDL technology benefits
Path lengths effects
Influence of process temperature
External Temperature error
Spectral Temperature
Experimental setup
Conclusion
Analytical Solutions by the SeaThe 60th Annual Symposium of the Analysis DivisionGalveston Island Convention Center, Galveston, Texas, USA; 26th – 30th April 2015
Slide 3
TDL's now fully established
High selectivity for the target gas species
In-situ analysis
Non-contacting
Fast response time
Dynamic measurement of process
Minimal calibration/validation requirements
Zero drift free
Analytical Solutions by the SeaThe 60th Annual Symposium of the Analysis DivisionGalveston Island Convention Center, Galveston, Texas, USA; 26th – 30th April 2015
Slide 4
Cross-stack TDL's
Originally designed for long path length applications, i.e. combustion, DeNOx
Process Applications tend to be more challenging
Installation requirements in small pipes can be detrimental to measurement quality
• Challenge to keep alignment with varying temperatures
Analytical Solutions by the SeaThe 60th Annual Symposium of the Analysis DivisionGalveston Island Convention Center, Galveston, Texas, USA; 26th – 30th April 2015
Slide 5
Effect of long path lengths
Concentration measurement over path length is averaged
Combustion control: ability to exclude leakage errors with over-the-burner location
Higher impact of dust load in process gas
Beam steering effects can lead to loss of transmission
But: what is the effective gas temperature over whole path? Temperature sensor is a point measurement Location of sensor very relevant for accuracy
Analytical Solutions by the SeaThe 60th Annual Symposium of the Analysis DivisionGalveston Island Convention Center, Galveston, Texas, USA; 26th – 30th April 2015
Slide 6
Folded path probe design
Single-port installation
No alignment user-side
Suitable for small pipes (2" and above)
Concept is expandable to different process interfaces
More adapted to control applications in process
Analytical Solutions by the SeaThe 60th Annual Symposium of the Analysis DivisionGalveston Island Convention Center, Galveston, Texas, USA; 26th – 30th April 2015
Slide 7
Effect of short effective path lengths
Wider choice of installation points
Collimated beam Improved SNR Dust tolerant
Detrimental to detection limit
Purge flow and process gas speed must be balanced
Possible T gradients over pipe length
Analytical Solutions by the SeaThe 60th Annual Symposium of the Analysis DivisionGalveston Island Convention Center, Galveston, Texas, USA; 26th – 30th April 2015
Slide 8
Lambert Beer
• I0 : Initial intensity
• I : Signal intensity after crossing the gas
• S : Line strength
• L : Optical path length
• G : Line profile
• n : density concentration
][exp ),,()( nLGS TPT ×××−= νoI
I=τ
Analytical Solutions by the SeaThe 60th Annual Symposium of the Analysis DivisionGalveston Island Convention Center, Galveston, Texas, USA; 26th – 30th April 2015
Slide 9
Temperature influence
Analytical Solutions by the SeaThe 60th Annual Symposium of the Analysis DivisionGalveston Island Convention Center, Galveston, Texas, USA; 26th – 30th April 2015
Slide 10
Temperature influence
Analytical Solutions by the SeaThe 60th Annual Symposium of the Analysis DivisionGalveston Island Convention Center, Galveston, Texas, USA; 26th – 30th April 2015
Slide 11
Sources of temperature errors
External temperature sensor too far away
Varying process gas temperature over path length Wall effects on small ID
pipes
Over the burner location
Filter on probe type process adaptions
Analytical Solutions by the SeaThe 60th Annual Symposium of the Analysis DivisionGalveston Island Convention Center, Galveston, Texas, USA; 26th – 30th April 2015
Slide 12
Probe style TDL with filter
No process side purging
Filter for dust prevention
Automated filter blow-back
Low beam steering
Fixed path length
Analytical Solutions by the SeaThe 60th Annual Symposium of the Analysis DivisionGalveston Island Convention Center, Galveston, Texas, USA; 26th – 30th April 2015
Slide 13
Probe style TDL with filter
Measurement-free zone (wall thickness) -100mm (standard)-200mm-300mm
Process window
Optical path lengths (effective) -200mm-400mm-800mm
Instrument side purge
Exchangeable metal filter -40 microns-100 microns-200 microns
Analytical Solutions by the SeaThe 60th Annual Symposium of the Analysis DivisionGalveston Island Convention Center, Galveston, Texas, USA; 26th – 30th April 2015
Slide 14
Impact of temperature error
(% Relative error on % Oxygen in air reading at ambient pressure and 1m OPL)
T∆T= -3°C
∆T= -5°C
∆T= -15°C
∆T= -30°C
20°C 1.39 2.34 7.22 15.12100°C 1.24 2.08 6.4 13.38200°C 1.07 1.79 5.52 11.47300°C 0.94 1.58 4.83 9.99500°C 0.76 1.27 3.88 7.96
Analytical Solutions by the SeaThe 60th Annual Symposium of the Analysis DivisionGalveston Island Convention Center, Galveston, Texas, USA; 26th – 30th April 2015
Slide 15
External temperature error
(Oxygen concentration in air measured in DN100 pipe with filter probe during temperature ramp to 225°C)
30% error on measurement ∆T approx. 40°C
Analytical Solutions by the SeaThe 60th Annual Symposium of the Analysis DivisionGalveston Island Convention Center, Galveston, Texas, USA; 26th – 30th April 2015
Slide 16
Spectral temperature
S1 S2
( )( )
( )( ) ( )
−−
−=
0
0
2
0
102
01
2
1 11exp
TTk
hc
TS
TS
TS
TSEE
Analytical Solutions by the SeaThe 60th Annual Symposium of the Analysis DivisionGalveston Island Convention Center, Galveston, Texas, USA; 26th – 30th April 2015
Slide 17
Spectral temperature measurement
Analytical Solutions by the SeaThe 60th Annual Symposium of the Analysis DivisionGalveston Island Convention Center, Galveston, Texas, USA; 26th – 30th April 2015
Slide 18
Conclusion
Best measurement performance with TDL's can be obtained when effective process gas temperature is taken into account External temperature sensor location
Expected temperature gradients over long path lengths
Expected wall effects on short path TDL's
Best option is to use spectral temperature measurement to achieve full freedom from above sources of error