understanding transient combustion phenomena in low-nox ... · overview of presentation 12...
TRANSCRIPT
![Page 1: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/1.jpg)
PI: Jacqueline O’Connor, Ph.D.Co-PI: Dom Santavicca, Ph.D.RE: Stephen Peluso, Ph.D.Graduate students: Dan Doleiden, Wyatt Culler, Adam Howie, John StrolloUndergraduates: Olivia Sekulich
Industry Partner: GE Global ResearchKeith McManus, Tony Dean, Fei Han
Mechanical and Nuclear EngineeringPennsylvania State Universitysites.psu.edu/rfdl/
Understanding Transient Combustion Phenomena in Low-NOx Gas Turbines Project DE-FE0025495, Oct. 2015 – Sept. 2018 (now Sept. 2019 with NCE)Program Monitor: Mark Freeman
1
![Page 2: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/2.jpg)
Overview of presentation
2
—Project motivation and approach
—Review of previous results
—Year 3 major results:
— Stability bifurcation during long-duration transients— Damping quantification— Local flame dynamics
—Conclusions and next steps
![Page 3: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/3.jpg)
Overview of presentation
3
—Project motivation and approach
—Review of previous results
—Year 3 major results:
— Stability bifurcation during long-duration transients— Damping quantification— Local flame dynamics
—Conclusions and next steps
![Page 4: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/4.jpg)
Objective of the program is to understand, quantify, and predict combustion instability during transient operation
4
— Two major deliverables for the program:
1. Fundamental understanding of flow and flame behavior during combustion transients and mechanisms for transition to instability
2. Development of a stability prediction or quantification framework
![Page 5: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/5.jpg)
The transients will be quantified using three different metrics: amplitude, timescale, and direction
5
Amplitude
0 0.2 0.4 0.6 0.8 10.5
0.6
0.7
0.8
0.9
1
Time
Eq
uiv
ale
nce
Ra
tio
Amplitude 1
Amplitude 2
Direction
0 0.2 0.4 0.6 0.8 10.5
0.6
0.7
0.8
0.9
1
Time
Eq
uiv
ale
nce
Ra
tio
Direction 1
Direction 2
Timescale
0 0.2 0.4 0.6 0.8 10.5
0.6
0.7
0.8
0.9
1
Time
Eq
uiv
ale
nce
Ra
tio
Timescale 1
Timescale 2
![Page 6: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/6.jpg)
Varying the transient timescales allows for different processes to equilibrate during the transient, changing the path
6
![Page 7: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/7.jpg)
Project Management Plan – progress to date
7
—Task 1 – Project management and planning
—Task 2 – Modification of current experimental facility with monitoring
diagnostics and new hardware for transient control
—Task 3 – Map combustor timescales at target operating points
—Task 4 – Design of transient experiments
—Task 5 – Fuel split transients (multi-nozzle combustor)
—Task 6 – Equivalence ratio transients (single- and multi-nozzle)
—Task 7 – Fuel composition transients (single- and multi-nozzle)
—Task 8 – Data analysis and determination of prediction/quantification
framework
![Page 8: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/8.jpg)
Three types of transients are being considered in both multi-nozzle and single-nozzle combustors
8
— Fuel-staging transients — Multi-nozzle only
— Equivalence ratio transients— Multi- and single-nozzle
— Fuel composition transients— Multi- and single-nozzle
![Page 9: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/9.jpg)
Experimental facilities include both a single-nozzle and multi-nozzle combustor, fuel splitting on multi-nozzle only
9
Five Swirled Nozzles
Combustor Can
Manifold
Premixed NG-air mixture
260 mm 30
0 m
m
Dump plate
Dump plate metal temperature
(K-type thermocouple)
Centerbody metal temperature
(K-type thermocouple)
Combustor pressure fluctuation
(PCB pressure transducer)
![Page 10: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/10.jpg)
Hardware modification focused on a valve with linear actuation to control fuel flow transients for fuel-splitting studies
10
Dump plate
Quartz combustor
Five nozzles
Control valve
Staging fuel enters combustor here
![Page 11: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/11.jpg)
Single-nozzle combustor is created by plugging four nozzles and using a smaller quartz liner with the same dump ratio
11
Five Swirled Nozzles
Combustor Can
Manifold
Premixed NG-air mixture
260 mm 30
0 m
m
Dump plate
![Page 12: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/12.jpg)
Overview of presentation
12
—Project motivation and approach
—Review of previous results
—Year 3 major results:
— Stability bifurcation during long-duration transients— Damping quantification— Local flame dynamics
—Conclusions and next steps
![Page 13: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/13.jpg)
Major Result #1: Fuel staging works both in axisymmetric and non-axisymmetric configurations
13
1
3
24
Culler, W., Chen, X., Peluso, S., Santavicca, D., Noble, D., O’Connor, J., (2018) “Comparison of Center Nozzle Staging to Outer Nozzle Staging in a Multi-Flame Combustor,” ASME Turbo Expo
![Page 14: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/14.jpg)
Major Result #2: While instability decay is smooth, instability onset takes longer and is intermittent – direction matters!
14
Culler, W., Chen, X., Samarasinghe, J., Peluso, S., Santavicca, D., O’Connor, J., (2018) “The effect of variable fuel staging transients on self-excited instabilities in a multiple-nozzle combustor,” Combustion and Flame, vol. 194, pg. 472-484
Instability Decay Instability Onset
![Page 15: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/15.jpg)
Major Result #3: Time-scale of a transient matters in the multi-nozzle combustor, and heat transfer likely plays a role
15
Short Timescales Long Timescales
![Page 16: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/16.jpg)
Major Result #4: Most significant difference between the single- and multi-nozzle instability is transient timescales
16
Single-Nozzle Transients
Multi-Nozzle Transients
Chen, X., Culler, W., Peluso, S., Santavicca, D., O’Connor, J., (2018) “Comparison of equivalence ratio transients on combustion instability in single-nozzle and multi-nozzle combustors,” ASME Turbo Expo
![Page 17: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/17.jpg)
Major findings and remaining questions
17
1. Fuel staging works both in axisymmetric and non-axisymmetric configurations— What is driving the differences between staging efficacy?
2. While instability decay is smooth, instability onset takes longer and is intermittent – direction matters!— How do we quantify the intermittency we see during onset?
3. Time-scale of a transient matters in the multi-nozzle combustor, and heat transfer likely plays a role— What are the flame dynamics occurring during the transition?
4. Most significant difference between the single- and multi-nozzle instability is transient timescales— How much of a role does flame/flow interaction play on the
transient physics?
![Page 18: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/18.jpg)
Overview of presentation
18
—Project motivation and approach
—Review of previous results
—Year 3 major results:
— Stability bifurcation during long-duration transients— Damping quantification— Local flame dynamics
—Conclusions and next steps
![Page 19: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/19.jpg)
Short- and long-duration transient behavior was different –long-duration timescales did not scale with the actuation time
19
Short Timescales Long Timescales
![Page 20: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/20.jpg)
The long-duration tests were done slowly enough; system responded quasi-steadily, allowing heat transfer to “keep up”
20
![Page 21: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/21.jpg)
This steady-state behavior allowed us to see the instability evolve in real time and identify the stability bifurcation point
21
1
3
24
![Page 22: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/22.jpg)
Overview of presentation
22
—Project motivation and approach
—Review of previous results
—Year 3 major results:
— Stability bifurcation during long-duration transients— Damping quantification— Local flame dynamics
—Conclusions and next steps
![Page 23: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/23.jpg)
Quantifying the thermoacoustic damping of the combustor is a way to understand the efficacy of fuel staging
23
Thermoacoustic system model
Heat release rate model
Van der Pol oscillator
Acoustic damping
Flamedriving
Thermoacoustic damping
Autocorrelation function
![Page 24: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/24.jpg)
Damping is a more reliable metric for quantifying suppression because of variations in the bifurcation equivalence ratio
24
1
3
24
CenterNozzle Nozzle 1 Nozzle 2
![Page 25: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/25.jpg)
Overview of presentation
25
—Project motivation and approach
—Review of previous results
—Year 3 major results:
— Stability bifurcation during long-duration transients— Damping quantification— Local flame dynamics
—Conclusions and next steps
![Page 26: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/26.jpg)
Local flame dynamics were measured with a high-speed OH-PLIF system at a rate of 10 kHz, imaging two of five flames
26
![Page 27: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/27.jpg)
OH-PLIF provides localized flame oscillation information, which we link back to the heat release rate oscillations
27
![Page 28: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/28.jpg)
The first goal of this analysis was to test the hypothesis about phase cancellation as the suppression mechanism with staging
28
outer = 0.70, middle = 0.86 overall = 0.73
outer = 0.70, middle = 0.70 overall = 0.70
outer = 0.70, middle = 0.86 overall = 0.73
outer = 0.70, middle = 0.70 overall = 0.70
Unstable:
Stable:
Samarasinghe, J., et al. (2017) “The Effect of Fuel Staging on the Structure and Instability Characteristics of Swirl-Stabilized Flames in a Lean Premixed Multinozzle Can Combustor,” Journal of Engineering for Gas Turbines and Power, 139(12), pg. 121504
![Page 29: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/29.jpg)
We considered three cases: unstable, center-nozzle staged, and right-nozzle staged, two nozzles with different efficacy
29
1
3
24
CenterNozzle Nozzle 1
![Page 30: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/30.jpg)
The unstable case displayed significant flame oscillations, where the adjacent branches oscillated in-phase
30
![Page 31: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/31.jpg)
Center nozzle staging showed almost no oscillations, while Nozzle 1 staging still had some coherent oscillations
31
Nozzle 1 Staging Center Staging
![Page 32: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/32.jpg)
We quantify the flame branch oscillations using the lateral movement of the flame and the phase between the branches
32
![Page 33: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/33.jpg)
Overview of presentation
33
—Project motivation and approach
—Review of previous results
—Year 3 major results:
— Stability bifurcation during long-duration transients— Damping quantification— Local flame dynamics
—Conclusions and next steps
![Page 34: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/34.jpg)
Wrap-up and Questions
34
— Key findings to date— Implemented a number of new quantification metrics for stability
in multi-nozzle combustors— Quantified the impact of transient timescales on combustor
behavior for multi- and single-nozzle combustors— Began investigating local flame oscillations as a way to understand
the effects of staging and flame dynamics in multi-nozzle systems
— Next steps— Understand the role of intermittency in the dynamics of transient
systems – need to quantify it as well— Sensitivity of flame behavior to fuel composition with blends of
natural gas + hydrogen
![Page 35: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/35.jpg)
Acknowledgements
35
— Penn State: Dom Santavicca, Bryan Quay, Janith Samarasinghe, Wyatt Culler, Dan Doleiden, Adam Howie, John Strollo, Xiaoling Chen, Jackson Lee, Steve Peluso, Ankit Tyagi, Olivia Sekulich
— GE Global Research: Keith McManus, Tony Dean, Janith Samarasinghe, Fei Han
— DOE/NETL: Mark Freeman
— College of Engineering Instrumentation Grant Program, Mechanical and Nuclear Engineering at Penn State
![Page 36: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/36.jpg)
Questions?
36
Understanding Transient Combustion Phenomena in Low-NOx Gas Turbines Project DE-FE0025495, Oct. 2015 – Sept. 2018Program Monitor: Mark FreemanPI: Jacqueline O’Connor, Ph.D.Co-PI: Dom Santavicca, Ph.D.RE: Stephen Peluso, Ph.D.Graduate students: Wyatt Culler, Dan Doleiden, Adam Howie, John StrolloUndergraduates: Olivia SekulichIndustry Partner: GE Global ResearchKeith McManus, Tony Dean, Fei Han
Mechanical and Nuclear EngineeringPennsylvania State Universitysites.psu.edu/rfdl/
![Page 37: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/37.jpg)
Backup slides
37
![Page 38: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/38.jpg)
The convective time depends on the amount of additional
fuel added.
38
Center Staging Calculated Convective Time
0.75 75 ms
0.80 37 ms
0.85 25 ms
![Page 39: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/39.jpg)
A general logistic regression is used to obtain the time
constants.
39
Used to model the growth (or decay) of systems that have a lower and upper asymptoteGrowth is initially exponential before leveling offLogistic fits are symmetric, but can be made general to change where the maximum growth rate is Using a more general logistic regression:
𝑃′ 𝑡 =𝐴 − 𝐵
(1 + 𝑏𝑒𝑘(𝑡−𝑡0)+ 𝐵
A: Instability AmplitudeB: Stable Amplitudek: Logistic Rate (negative is growth, positive is decay)t0: Time offsetb: Pre-exponential factor, typically taken as 1
![Page 40: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/40.jpg)
A general logistic regression is used to obtain the time
constants.
40
Original Equation: 𝑃′ 𝑡 =𝐴−𝐵
(1+𝑒𝑘(𝑡−𝑡0)+ 𝐵
Fractional Amplitude Characteristic Time:𝐴 − 𝐵
𝑅=
𝐴 − 𝐵
(1 + 𝑒𝑘(𝑡−𝑡0)+ 𝐵
R is the amplitude fractionSolving for 𝑡 − 𝑡0
𝑡 − 𝑡0 = ln 𝑅 +𝐵
𝐴 − 𝐵− 1
1
𝑘
The absolute value of this should be proportional to a time constantThis will not be dependent on valve actuation time
Half-Maximum Time for b=1 (neglecting zero offset B)Given by 𝑡0:
𝑃 𝑡0 =𝐴 − 𝐵
1 + 𝑒𝑘(𝑡0−𝑡0)=
𝐴 − 𝐵
2Will depend on valve actuation time
![Page 41: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/41.jpg)
Both staging strategies, i.e. increasing overall equivalence
ratio and keeping overall equivalence ratio constant resulted
in successfully suppressing instabilities
41
Unstable
Stable
![Page 42: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/42.jpg)
Differences in staging efficacy near the bifurcation point typically relate back to the level of intermittency
42
Sharp transition to coherent phase relationship
In-phase relationship between heat release rate oscillations and pressure
Continued, but somewhat intermittent, growth in instability amplitude after coherence switch
![Page 43: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/43.jpg)
Images of forced flames can be decomposed into mean, RMS and phase components to understand instability mechanisms
43
𝐹𝑖𝑙𝑡𝑒𝑟𝑒𝑑 𝐼𝑚𝑎𝑔𝑒 = 𝑴𝒆𝒂𝒏 𝑰𝒎𝒂𝒈𝒆 + 2 ∗ 𝑹𝑴𝑺 𝑰𝒎𝒂𝒈𝒆 ∗ cos(2𝜋𝑓𝑡 + 𝑷𝒉𝒂𝒔𝒆 𝑰𝒎𝒂𝒈𝒆)
Mean, RMS, and phase images are analyzed at different test conditions to determine the effects of fuel staging on time-averaged and phase-averaged flame structure
![Page 44: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/44.jpg)
φglobal= 0.65 (prms = 0.024 psi)Stable – not staged
φglobal = 0.70 (prms = 0.21 psi)Unstable – not staged
φglobal = 0.71 (prms = 0.20 psi)Marginally stable – staged
φglobal = 0.73 (prms = 0.027 psi)Stable – staged
Flame structure does not change significantly with additional staging, though center flame has higher heat release
44
![Page 45: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/45.jpg)
Line-of-sight chemiluminescence images are acquired at 5°increments around the combustor to create tomographic image
45
High-speed camera-intensifier
setup
Multi-Nozzle Combustor
180° circular track
Optical Table
![Page 46: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/46.jpg)
Different flame structures are observed between stable un-staged and stable staged cases through tomographic imaging
46
Max. Intensity0
10 mm 25 mm 50 mm 90 mm
No fuel staging-Global=0.65
Fuel staging-Global=0.7
![Page 47: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/47.jpg)
Heat release rate RMS levels are suppressed with staging, though signature is visible even at highest staging amount
47
φglobal = 0.70 (prms = 0.21 psi)Unstable – not staged
φglobal = 0.71 (prms = 0.20 psi)Marginally stable – staged
φglobal = 0.73 (prms = 0.027 psi)Stable – staged
φglobal= 0.65 (prms = 0.024 psi)Stable – not staged
![Page 48: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/48.jpg)
Phase of oscillations seems to indicate phase shift in oscillations during staging, possible suppression mechanism
48
φglobal = 0.70 (prms = 0.21 psi)
φglobal = 0.71 (prms = 0.20 psi) φglobal = 0.73 (prms = 0.027 psi)
φglobal= 0.65 (prms = 0.024 psi)
![Page 49: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/49.jpg)
Task 4: Test matrix for initial transient testing considers step-change transients to determine natural time-scales of system
49
φoverall = 0.70 (prms = 0.21 psi)
φoverall = 0.73 (prms = 0.027 psi)
φcenter=0.7
φcenter=0.85
What is the time delay?
![Page 50: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/50.jpg)
Impulse transients are executed using a fast-acting proportional control valve
50
φcenter=0.7
Φcenter>0.7
Time
0 seconds 4 seconds2 seconds 6 seconds
High-speed CH* recording start
High-speed CH* recording end
Feedback Voltage
Scaled Valve Pressure Difference
Control Signal
![Page 51: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/51.jpg)
Both the fluctuation in CH* (blue) and pressure (red) track each other through the transient event.
51
![Page 52: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/52.jpg)
The growth/decay time of the instability reflects a natural time-scale of the system, and is dependent on staging level
52
Center-nozzle: φ=0.8 Center-nozzle: φ=0.85
Stable Unstable
Unstable Stable
![Page 53: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/53.jpg)
The growth/decay time of the instability reflects a natural time-scale of the system, and is dependent on staging level
53
Center-nozzle: φ=0.8 Center-nozzle: φ=0.85
Stable Unstable
Unstable Stable
![Page 54: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/54.jpg)
Box-and-whisker plots provide a useful way to visualize ensemble data.
54
Median (second quartile)
Median between minimum and median of whole data set (first quartile)
Median between maximum and median of whole data set (third quartile)
Minimum value
Maximum value
![Page 55: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/55.jpg)
Box-and-whisker plots provide a useful way to visualize ensemble data.
55
Inner Quartile Range (IQR)
Minimum value (outlier)<1.5 IQR from median
Maximum value (outlier)>1.5 IQR from median
Median-1.5*IQR(or minimum point if no outliers)
Median+1.5*IQR(or maximum point if no outliers)
![Page 56: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/56.jpg)
The pressures before and after the transient mirror the steady-state test results, showing high repeatability
56
Unstable to Stable Stable to Unstable
before after before after
![Page 57: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/57.jpg)
The characteristic decay time depends on staging amplitude, but the characteristic rise time does not appear to
57
![Page 58: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/58.jpg)
The functional form of the growth and decay profiles can help illuminate some of the physics involved in the processes
58
Model 1: Damped linear oscillator decaying at a single frequency
Model 2: General Logistic Growth/Decay
𝑃′ 𝑡 = 𝐴𝑒−𝜆𝑡 + 𝐴𝐵 ∗ 1 − 𝑒−𝜆𝑡 ∗ sin(𝜔𝑡 + Φ)
𝑃′ 𝑡 = 𝐴𝐵𝑒−𝜆𝑡 + 𝐵 ∗ 1 − 𝑒−𝜆𝑡 ∗ sin(𝜔𝑡 + Φ)
𝑃′ 𝑡 =𝐴 − 𝐵
(1 + 𝑒𝑘 𝑡−𝑡0 )+ 𝐵
![Page 59: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/59.jpg)
A tomographic reconstruction technique is used to obtain the 3-D chemiluminescence distribution
90 180 270 360
100
200
300
400
500100 200 300 400
50
100
150
200
250
300
350
400
450
(b)
s
θ90 180 270 360
100
200
300
400
500100 200 300 400
50
100
150
200
250
300
350
400
450
(c)
y
x
(d)
z
(a) Each line-of-sight image is divided into pixel wide horizontal bins
(b) The horizontal bins from each image acquired around the combustor are combined into an array
(c) This array is input into a filtered backprojection algorithm which reconstructs a 2-D cross section of the flame
(d) The cross sections at every axial location are stacked to obtain a 3-D matrix of the flame’s chemiluminescence distribution
(a)
s
z
![Page 60: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/60.jpg)
Horizontal 2-D slices of the 3-D image illustrate the flame structure at different points downstream of the dump plane
Location of slice
Horizontal slice
High IntensityLow Intensity
![Page 61: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/61.jpg)
Time-averaged FSD images
61
φ = 0.65, all nozzles fueled equally φ = 0.70, all nozzles fueled equally
Staged - φouter = 0.67, φmiddle = 0.82 Staged - φouter = 0.70, φmiddle = 0.85
![Page 62: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/62.jpg)
Engine load is typically varied by either varying fuel staging or the equivalence ratio of certain fuel nozzles
62Source: Davis and Black, “Dry Low NOx Combution Systems for GE Heavy-Duty Gas Turbines”
![Page 63: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/63.jpg)
Stability limits of certain operating points are already known, current work focuses on mapping instability with fuel splits
63
Stable Unstable Poor Stabilization Cannot Achieve Condition
Measurements include:• Flow rates• Dynamic pressure• Surface
temperatures• Global heat release• High-speed flame
imaging
Goals of Q1/Q2 testing:
— Quantify steady-state flame behavior and stability
— Develop methodologies for ensuring repeatability
![Page 64: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/64.jpg)
U = 22.5 m/s, Tin = 200°C, fully premixed, unforced
ϕ = 0.65 ϕ = 0.60 ϕ = 0.55 ϕ = 0.50 ϕ = 0.45
![Page 65: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/65.jpg)
ϕ = 0.60 ϕ = 0.50 ϕ = 0.48
Photographs of multi-nozzle flame at U = 25 m/s, Tin = 200°C
Instabilities may arise as a result of changes in flame shape and flame anchoring that occur with variation in equivalence ratio
![Page 66: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/66.jpg)
To further investigate the structure of the multi-nozzle flame, 3-D image sets were obtained at φ = 0.60 and φ = 0.48
φ = 0.60
φ = 0.48
φ = 0.60
φ = 0.48
![Page 67: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/67.jpg)
Stable Unstable Poor Stabilization Cannot Achieve Condition
Data from unforced and forced flames are available in a
range of operating conditions
67
Inlet temperature = 150°C
15 17.5 20 22.5 25 27.5 30
0.40
0.45
0.50
0.55
0.60
0.65
0.70
U (m/s)
φ
Inlet temperature = 100°C
15 17.5 20 22.5 25 27.5 30
0.40
0.45
0.50
0.55
0.60
0.65
0.70
φ
U (m/s)
Inlet temperature = 200°C
15 17.5 20 22.5 25 27.5 30
0.40
0.45
0.50
0.55
0.60
0.65
0.70
U (m/s)
φ
Inlet temperature = 250°C
15 17.5 20 22.5 25 27.5 30
0.40
0.45
0.50
0.55
0.60
0.65
0.70
U (m/s)
φ
![Page 68: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/68.jpg)
An inlet temperature of 200°C and an inlet velocity of 25 m/s
was chosen for the steady-state tests
68
Inlet temperature = 200°C
15 17.5 20 22.5 25 27.5 30
0.40
0.45
0.50
0.55
0.60
0.65
0.70
U (m/s)
φ
Based on the stability maps of fully premixed operation, this condition was chosen as it enables both transition in flame structure and transition to instability by varying fuel flow rate
Stable Unstable Poor Stabilization Cannot Achieve Condition
![Page 69: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/69.jpg)
Stability map for GE-15 single-nozzle experiment (TPM)
69
Tin = 100°C Tin = 200°C Tin = 275°C
φ 25 m/s 30 m/s 35 m/s 25 m/s 30 m/s 35 m/s 30 m/s 35 m/s
0.50 LBO LBO LBO LBO LBO LBO
0.525 LBO LBO LBO 57 Hz 72 Hz 81 Hz 97 Hz 106 Hz
0.55 LBO 68 Hz 84 Hz 78 Hz 86 Hz 111 Hz 105 Hz 116 Hz
0.60 71 Hz 96 Hz 117 Hz 94 Hz 110 Hz 124 Hz 111 Hz 125 Hz
0.65 89 Hz 114 Hz 124 Hz 101 Hz 115 Hz 125 Hz 117 Hz 127 Hz
0.70 104 Hz 117 Hz 129 Hz 112 Hz 118 Hz 128 Hz 129 Hz 145 Hz
![Page 70: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/70.jpg)
Stability map for GE-15 single-nozzle experiment (FPM)
70
Tin=100 C Tin=150 C Tin=200 C
φ 20 m/s 25 m/s 30 m/s 20 m/s 25 m/s 30 m/s 20 m/s 25 m/s 30 m/s
0.50 LBO LBO LBO LBO LBO LBO LBO
0.55 LBO
0.60
0.65
0.70
0.75
![Page 71: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/71.jpg)
Fuel injection strategy for staging
71
Two options for adding additional fuel to middle nozzle:
(1) Inject fuel at swirler technically premixed
(2) Inject fuel at air manifold with a choke fully premixed
The main difference between these then becomes the type of governing mechanisms during the unstable flame case
Poravee showed using acetone PLIF, that the fuel and air are well mixed at the nozzle exit of the GE-15 nozzle
![Page 72: Understanding Transient Combustion Phenomena in Low-NOx ... · Overview of presentation 12 —Project motivation and approach —Review of previous results —Year 3 major results:](https://reader036.vdocuments.site/reader036/viewer/2022062604/5fc0912a4ea686120a44d8c4/html5/thumbnails/72.jpg)
Fuel injection strategy for staging
72
S
Middle Nozzle
We currently don’t have a flowmeter that can accurately measure as low of a flow rate as we are planning to run
Backup option is to use a rotameter