SCO2 Oxy-Combustion Working Group

SCO₂ Oxy-Combustion Working Group
Meeting Minutes
November 12, 2019

Sixth Meeting of the SCO₂ Oxy-Combustion Working Group

At the first meeting, computational modeling results were presented by Dr. Pete Strakey of NETL.  Dr. Jacob Delimont from Southwest Research Institute gave a presentation at the second meeting on recent efforts to design, build, and test an oxycombustor in the existing Sunshot loop at Southwest Reasearch Institute. At the third meeting which took place in April we had two presenters: Dr. Wenting Sun from Georgia Tech and Dr. Subith Vasu from the University of Central Florida, who both presented progress associated with University Turbine Systems Research awards to investigate autoignition delay of syngas and natural gas in sCO₂ oxycombustion conditions. Last August we heard a presentation from Dr. Lee Shunn from Cascade Technologies, Inc. At the UTSR Workshop in Daytona Beach last November, we had a session to review all of the ongoing NETL-funded projects in sCO₂ oxycombustion. In July, we had a presentation from Dr. Matthias Ihme from Stanford University titled “Thermostructural Analysis and Numerical Modeling for Scalar Mixing and Combustion at Supercritical Conditions”.
At this sixth meeting, Dr. Wenting Sun joined us from Georgia Tech to discuss some of the final findings from the UTSR project that just came to a close at the end of September. The title of today’s talk is “Investigation of Methane and Syngas Autoignition for Supercritical CO₂ Power Cycles". 
As a reminder the mission of this working group is to promote a technical dialogue on the subject of oxy-combustion for direct SCO₂ power cycles to identify priority technology gaps and track progress on addressing technical challenges.

Presentation

• Dr. Wenting Sun – Georgia Tech
• Title: “Investigation of Methane and Syngas Autoignition for Supercritical CO₂ Power Cycles”

Presentation Highlights:

• Kinetic challenges exist for SCO₂-fuel-O₂ mixtures
  • Different kinetic mechanisms deviate with increases in pressure
  • Existing kinetic mechanisms differ by a factor of three for a given set of conditions.  GRI 3.0 predicts significantly shorter autoignition delay.
  • Need experiments in region of interest
• Why should we care about combustion kinetics?
  • Different kinetic models yield very different results as evidenced by temperature contours generated through CFD modeling.
  • Must have validated kinetic models for combustor design.
• The project used Georgia Tech’s high-pressure shock tube for experiments up to 300 bar.
  • Shock tube is 6” diameter and 69 feet long,
• Experimental challenges
  • Large Cp of CO₂
    • Requires a strong shock wave
    • Non-ideal thick boundary layer formation which means diameter of shock tube must be large
  • High experimental cost – requires large amount of high-pressure gas
  • Challenges of data interpretation
    • P&T not constant
    • Data interpretation is not trivial
• At beginning of project, data not available in high P/high CO₂ mole fraction region
• Now, data is available in this regime.
• Sidewall and endwall pressures and emissions (OH*) were measured simultaneously.
• Methane/Oxygen/CO₂ or Argon
  • Aramco 2.0 has the best match with experiments in the low temperature regime where kinetic models begin to differ.
  • Reaction pathways were analyzed across each mechanism to determine why GRI, Aramco, and FFCM-1 yield different results.
  • Both pressure and temperature have independent effects on ignition delay time, but Aramco seems to match experiments well across most conditions tested.
• Syngas/Oxygen/CO₂ or Argon
  • All models work well for syngas autoignition.
  • The diluent used appears to have negligible effect on ignition delay time.
  • Sensitivity analysis was performed to determine if diluent should have an effect at the reaction level.
    • CO concentration increases with time when CO₂ is used as diluent
    • CO concentration decreases with time when Ar is used as diluent
  • The H mole fraction was much higher in Ar mixtures because H is consumed in CO₂ mixtures
• Real Gas Effect
  • At high T, real gas behaves like ideal gas
  • At low T (near the critical point), real gas effect is important

Conclusions – CH₄

• Most kinetic models can predict autoignitions reasonably well at SCO₂ conditions.
  • GRI 3.0 is rejected
• CH₃O₂ is important at high pressure and low temperature conditions.
• CO₂ has negligible chemical effect at high pressure conditions.

Conclusions - Syngas

• Most kinetic models can predict H₂/CO autoignitions reasonably well at SCO₂ conditions.
• CO₂ has chemical effect on elementary reactions. Its effect on autoignition delays is washed out (within uncertainty of experiment).

Final Thoughts

• Conclusions are limited to autoignition chemistry.
• Thermal effect (heat capacity) is eliminated in both experiments and simulations.
• Real gas effect is not important in combustor if inflow is hot.
• Real gas effect is only important when temperature is near critical point.

• Q&A followed the presentation

Future Meeting

• The next SCO₂ Oxy-Combustion Working Group Meeting will be in January 2020
  • Subith Vasu of University of Central Florida will be presenting
• If interested in presenting at a future meeting, contact Seth Lawson
• See NETL website netl.doe.gov for presentation and meeting minutes from today’s meeting, and notices for future meetings.

Other Topics

• Turbo Expo Panel, June 2020, London
  • Panel on oxy-combustion for SCO₂ power cycles
  • Requesting volunteer panelists – if interested contact Seth Lawson (seth.lawson@netl.doe.gov)