Presentation given by Richard Marsh of Cardiff University on "Phase 1 Project: Methane Oxycombustion in a Swirl Stabilised A Gas Turbine Burner" at the UKCCSRC Gas CCS Meeting, University of Sussex, 25 June 2014

1. PHASE 1 PROJECT:
METHANE OXYCOMBUSTION IN A
SWIRL STABILISED A GAS TURBINE BURNER.
Richard Marsh
Agustin Valera-Medina
Cardiff University – School of Engineering
UKCCSRC Gas CCS Meeting
25/06/2014

2. METHANE OXYCOMBUSTION IN SWIRL
STABILISED A GAS TURBINE BURNER.
• There has been a significant amount of interest in CO2 diluted
cycles. This has been championed by system developers
looking at advanced gas turbine cycles.
• The overall design aim would be to develop an oxyfuel GT,
using recycled CO2 as a moderator (rather than N2 as in air).
• Hence the project objective was to examine the relationship
between CO2 and N2 diluted swirl flames.
• Given that swirl burners involve a complex interaction
between chemical and fluid dynamic timescales, a systematic
study of the stability envelope was conducted.

3. CHEMKIN modelling

4. provides 0.8 ≤ Sg ≤ 1.5 for thermal output up to 50 kW
Generic Swirl Burner- Utilises different vane configurations to alter geometric swirl
number (28mm exit diameter)

5. Swirl stabilised oxyfuel flames diluted with:
(a) N2 and (b) CO2 at comparable equivalence ratios.
N2 and CO2 diluted flames at a thermal power of 4.3 kW and an equivalence ratio of
close to 0.5. On visual inspection the flame shapes are similar, but the CO2 diluted
flame appears to be wider which is supported by subsequent PIV analysis.

6. (Flashback video)
Burner stability envelopes
• The project aim was to quantify the differences in diluent
properties in swirl burners.
• There is very little research literature to confirm whether using
CO2 as a diluent will provide an operationally stable flame.
• CO2 is far more dense and has a greater heat capacity than N2.
• This could affect flame stability due to significantly different
chemical kinetics.
• Hence flashback and blowoff behaviour might be significantly
different.

7. 28mm Diameter burner
Flashback and Blowoff limits
Equivalence Ratio (based on CH4 to O2 by vol only)

8. Effect of oxygen concentration at Blowoff
E.R. = 1.0
E.R. = 0.5

9. 20mm diameter burner No diluent

10. 10
PIV Measurements
• Used Litron Nd:YAG laser PIV system
• Swirl number Sg =1.045
• Flow seeded using Aluminium Oxide
• Typical frame gap of 125 µsec.
• 1000 image pairs per condition
Four conditions investigated
1. 7 l/min CH4, 15 l/min 02, 51 l/min N2
2. 7 l/min CH4, 40 l/min 02, 40 l/min N2
3. 7 l/min CH4, 15 l/min 02, 36 l/min CO2
4. 7 l/min CH4, 40 l/min 02, 32 l/min CO2
Oxy combustion PIV

11. Conditions examined
Condition O2 (l/min) CH4
(l/min)
N2 (l/min) CO2
(l/min)
Total
(l/min)
O2
(%)
Stable 15 7 51 - 73 21 %
High oxy 40 7 40 - 87 46 %
Stable 15 7 - 36 58 26 %
High oxy 40 7 32 79 51 %

12. Axial velocities at stable operating point
N2 diluted CO2 diluted

13. 13
Oxy combustion PIV
• Test point 1 – Stable with N2 dilution.
• Presence of unstable flame front increases turbulent nature of velocities.
• Strong CRZ still visibly present

14. 14
Oxy combustion PIV
• Test point 2 – High O2 with N2 dilution.
• Condition is close to “blow-off”
• No evidence of CRZ
• Higher axial velocities, more “ordered” tangential velocities compared to TP 1.

15. 15
Oxy combustion PIV
• Test point 3 – Stable with CO2 dilution.
• Small CRZ is present
• Similar axial and tangential velocities to that seen in TP 1

16. 16
Oxy combustion PIV
• Test point 4 – High O2 with CO2 dilution.
• Condition is close to “blow-off”
• Little evidence of CRZ
• Similar axial and tangential velocities to N2 “blow-off” condition

17. Key Findings
• Achieving a stoichiometric methane oxyflame wasn’t
possible with a traditional swirl burner unless a diluent
was present.
• It is possible to operate without a diluent, but the
momentum required to sustain the recirculation zone is
only possible at very dilute (high oxygen) conditions.
• Replacing N2 with CO2 as a combustion diluent appears to:
– Widen the stability range of the swirl flame.
– Shift the operating points to lower volume flow rates.
• Velocity data shows that the criteria for stability appears
to be the same for both diluents.

18. Impact
• The data shows that despite very different
physical properties between the diluents,
under the conditions tested, stability is still
largely governed by velocity distributions in
the central recirculation zone.
• Thus for CO2 moderated flames, existing swirl
burner design rules can be modified.