Direct Fired Oxy-Fuel Combustor for sCO2 Power Cycles
Jacob Delimont, Ph.D.Aaron McClung, Ph.D.
Southwest Research Institute
Marc PortnoffLalit Chordia, Ph.D.Thar Energy L.L.C.
Work supported by US DOE under DE-FE002401
11/10/2016 2016 University Turbine Systems Research Workshop 1
HEAT SOURCE
PRECOOLER
LOW TEMP RECUPERATOR
HIGH TEMPRECUPERATOR
EXPANDER
COMPRESSOR
RE-COMPRESSOR
P6 P7a P8
P1
P2
P3
P5
P7b
P7
P4
P4b
COOLING OUT COOLING IN
P4a
Outline
• Phase I Overview– Background– Project Objectives– Phase I Progress
• Cycle Modeling• Chemical Kinetics• Preliminary Combustor Design• Bench-top Combustor Test
• Phase II Project Plan
11/3/2015 2015 University Turbine Systems Research Workshop 2
What is a sCO2 cycle?
• Closed Cycle– Working fluid is CO2
• Cycle Type– Vapor phase– Transcritical– Supercritical
• Supercritical CO2 has:– High fluid density– High heat capacity– Low viscosity
11/3/2015 2015 University Turbine Systems Research Workshop 3
HEATSOURCE
PRECOOLER
RECUPERATOR
EXPANDERCOMPRESSOR
P5 P6
P1
P2
P4
COOLING OUTCOOLING IN
P3
Recuperated ClosedBrayton Power Cycle
Why sCO2 Power Cycles?
• Offer +3 to +5 percentage points over supercritical steam for indirect coal fired applications
• High fluid densities lead to compact turbomachinery
• Efficient cycles require significant recuperation
11/3/2015 2015 University Turbine Systems Research Workshop 4
Third Generation 300 MWe S-CO2 Layout from Gibba, Hejzlar, and Driscoll, MIT-GFR-037, 2006
Why Oxy-Fuel Combustion?
11/10/2016 2016 University Turbine Systems Research Workshop 5
HEAT SOURCE
PRECOOLER
LOW TEMP RECUPERATOR
HIGH TEMPRECUPERATOR
EXPANDER
COMPRESSOR
RE-COMPRESSOR
P6 P7a P8
P1
P2
P3
P5
P7b
P7
P4
P4b
COOLING OUT COOLING IN
P4a
• Capture 99% of carbon dioxide• Higher turbine inlet temperatures
possible
CO2
Project Objectives
• Optimize the supercritical CO2 power cycle for direct fired oxy-combustion– Target plant conversion efficiency is 52% (LHV)
• Technology gap assessment for direct fired plant configurations
• Develop a high inlet temperature oxy-combustor suitable for the optimized cycle– Target fuels are Natural Gas and Syngas
11/3/2015 2015 University Turbine Systems Research Workshop 6
Outline
• Phase I Overview– Background– Project Objectives– Phase I Progress
• Cycle Modeling• Chemical Kinetics• Preliminary Combustor Design• Bench-top Combustor Test
• Phase II Project Plan
11/3/2015 2015 University Turbine Systems Research Workshop 7
DESIGN -SPECCOMPMAT C
DESIGN -SPECFUELFLOW
DESIGN -SPECFUELPRES
DESIGN -SPECO2PRES
DESIGN -SPECPINLET
DESIGN -SPECT INLET
DESIGN -SPECT MIXBAL
CALCU LATORCOOLT OWR
CALCU LATOREFFICIEN
CALCU LATORO2CALC
PIPELCMP
PIPLNCO2
H2OSEPER FUELCOM
W
MIXER
FUELWORK
O2PUMP
COMBUST
W
MIXER
WMIXER
COOLER
FMIX
REJ ECT HX
FSPLIT
HXLOWHXHIGH
MAINCOMP
RECOMP
EXPANDER
CO2
S15
WPIPELIN
T AKEOFF
H20
S21
S3
CH4IN
CH4PRE
FUELCW
ASUPOWW
O2PWORK
WFUELSYS
O2IN
O2PRE
INLET
S14
WMAINCOM
WRECOMP
WT URBINE
WCOOLW
WNETW
S7
S10
QRECOMPC
Q
S9
S11
MAIN
S6
QREJECT
Q
RECOMPRE
S2
S8
S1
OUT LET
De-watering and Cleanup
Fuel, Oxidizer, and Combustion
Oxy-Combustion Plant Model
11/3/2015 2015 University Turbine Systems Research Workshop 8
Condensation and Recompression Cycles
ThermalInput
Cooler
Low TemperatureRecuperater
High TemperatureRecuperater
Turbine
Compressor
Recompressor
ThermalInput
11/3/2015 2015 University Turbine Systems Research Workshop 9
ThermalInput
CoolerRecuperater
Turbine
Compressor
Pump
ThermalInput
Cycle Analysis Results
• Recompression cycle has highest efficiency – 53.4% at 200 bar, 56.7% at 300 bar
• Condensation cycle – 51.6% at 200 bar, 54.0% at 300 bar– Superior in all other metrics – Reduced recuperation (~ 50%)– Lower combustor inlet temperature– Higher power density (power output / flow rate)
• Both cycle configurations are compatible with an auto-ignition style combustor for 1200 C Turbine inlet temperatures.
11/3/2015 2015 University Turbine Systems Research Workshop 10
Outline
• Phase I Overview– Background– Project Objectives– Phase I Progress
• Cycle Modeling• Chemical Kinetics• Preliminary Combustor Design• Bench-top Combustor Test
• Phase II Project Plan
11/3/2015 2015 University Turbine Systems Research Workshop 11
Kinetic Model: Motivation
• The fundamental size of the combustor is governed by the timescale of chemical reactions
• The chemical reaction kinetics determine how fast fuel oxidation occurs– A detailed chemical kinetic model is required to
size the combustor – A reduced chemical kinetic model is required for
detailed flow-field design in CFD
11/3/2015 2015 University Turbine Systems Research Workshop 12
Kinetics Knowledge Base
11/10/2016 2016 University Turbine Systems Research Workshop 13
CO2 concentration
PressureCurrent Application
P up to 200 barxCO2 up to 0.96 (mostly as diluent)
Well-Developed MechanismsP up to 20 bar
xCO2 < 0.10 (mostly as product)Sparse data at low pressure, high CO2
Sparse data at high pressure, low CO2
Knowledge front
No data available at conditions relevant to this application.
Mechanism Selection• Primary selection criterion is accurate prediction of the
overall reaction time scales– Drives the combustor design– More important than other details such as peak
concentration values• USC-II is the clear choice based on this criterion
– Most accurate in highest pressure flamespeed and autoignition validation comparisons
• USC-II also had good to adequate performance in low pressure CO2 studies
• USC-II predictions should carry +/- 50% uncertainty in this application
11/3/2015 2015 University Turbine Systems Research Workshop 14
Reduced Order Model
• For incorporation into a CFD model a reduced order model was developed
• Equations based on Arrhenius rate equation were tuned to match USC-II model predictions– Match autoignition delay– Match residual CO levels– Overall time to complete reaction
11/3/2015 2015 University Turbine Systems Research Workshop 15
Outline
• Phase I Overview– Background– Project Objectives– Phase I Progress
• Cycle Modeling• Chemical Kinetics• Preliminary Combustor Design• Bench-top Combustor Test
• Phase II Project Plan
11/3/2015 2015 University Turbine Systems Research Workshop 16
Mixing vs. Kinetics Time Scales
• Time scale of reaction kinetics is much smaller than physical mixing time scales
• Combustion size and length governed by physical mixing
• Use of CFD with finite rate chemistry to model this
11/10/2016 2016 University Turbine Systems Research Workshop 17
Time (ms)
Tem
pera
ture
(K)
Kinetics Models
Initial Combustor Concept
11/10/2016 2016 University Turbine Systems Research Workshop 18
CFD Model Setup
• ANSYS CFX 16.2• Unstructured mesh
– Boundary layer and injection region refinement
– 4 million elements– Mesh sizes from 2 to 17
million elements for independence study
• Finite rate chemistry– Extrapolated reduced order
equations
11/10/2016 2016 University Turbine Systems Research Workshop 19
Temperature in 45° Clocked Case
11/10/2016 2016 University Turbine Systems Research Workshop 20
1000
1100
1200
1300
1400
1500
1600
1700
1800
1900
2000
0.00 0.20 0.40 0.60 0.80 1.00
Tem
pera
ture
(K)
Axial Distance (m)
Ave Temperature (K)
Max Temperature (K)
Change Injection Spacing
11/10/2016 2016 University Turbine Systems Research Workshop 21
4 Hole 45°Clocked
4 Hole 11.25°Clocked
4 Hole Aligned
• Injection oxygen and fuel need not be at same location
• Auto-ignition allows even small concentrations of fuel+oxidizer to react
Final Design: Fuel Injection 24in Upstream
• Fuel well mixed throughout combustor before oxygen• Allows hydrocarbon “cracking” before oxygen injection• Cooler max temperatures• Very good mixing at outlet• Very low unburnt fuel percentage
11/10/2016 2016 University Turbine Systems Research Workshop 22
1000
1100
1200
1300
1400
1500
1600
0.00 0.50 1.00 1.50 2.00
Tem
pera
ture
(K)
Distance (m)
Ave Temperature (K)
Max Temperature (K)
Preliminary Mechanical Design
• Thermal design– Thermal
containment using refractory insulating layer
– Cooling CO2
• Mechanical design– Utilizes stainless
steel ANSI pipe and flanges
11/10/2016 2016 University Turbine Systems Research Workshop 23
Hot Inlet Air
Combustion Processes
Cooling CO2
Outline
• Phase I Overview– Background– Project Objectives– Phase I Progress
• Cycle Modeling• Chemical Kinetics• Preliminary Combustor Design• Bench-top Combustor Test
• Phase II Project Plan
11/3/2015 2015 University Turbine Systems Research Workshop 24
Bench-top Combustor Test
• Small bench top test to study proof of concept, autoignition delay, and chemical kinetics
• Once through type system– 200 bar pressure
• Electric heaters used to set inlet temperature• Jet in cross flow type fuel and oxidizer
injection
11/10/2016 2016 University Turbine Systems Research Workshop 25
Test Stand Loop Design
11/10/2016 2016 University Turbine Systems Research Workshop 26
Oxy-fuel Test Reactor
• Machined from Haynes 230 bar stock
• Instrumentation standoff tubes welded to main combustor
• Two stage pre-heater to achieve 925°C combustor inlet
11/10/2016 2016 University Turbine Systems Research Workshop 27
Gas Sampling Cooling Jacket
WeldFitting
Green: 1.25” O.D. Reactor Body – 12” longOrange: Dynamic pressure sensorYellow: ThermocoupleRed: Gas SamplingPurple: Oxygen & Methane/Ethane injectionGray: Reactor inlet and outlet (1/2”)
Haynes 2301.25” O.D.0.188” I.D.
Inconel 625¼” O.D.
0.120” I.D.
CO2Flow
750°C1020°C
10”20”
3” O.D. Tool SteelHeated via 4 knuckle
band heaters (not shown)
Haynes 230½” O.D. x 0.093” I.D.
Heated via ceramic furnace heaters (not
shown)
To Reactor
Inlet
• Water jacketed gas sampling
Fuel and Oxygen Injector Design
• Precise sapphire orifice set into stainless steel mount
• Orifice constriction placed close to the combustor
• Mounted inside welded in place standoff
11/10/2016 2016 University Turbine Systems Research Workshop 28
OrificeBody
Orifice Guide316 Stainless Tube
1/8” x 0.035”
Standoff 0.5” OD x 0.135” ID
0.020” Support
Lip
Combustor Test Stand
11/10/2016 2016 University Turbine Systems Research Workshop 29
Test Stand Assembly
• Testing at Thar’s facility in Pittsburg, PA
• Outdoors with remote operation
11/10/2016 2016 University Turbine Systems Research Workshop 30
Instrumentation
• Thermocouples in combustion zone
• Dynamic pressure transducers
• Three gas sampling ports– Optical emission
spectroscopy (OES) to analysis chemical makeup
11/10/2016 2016 University Turbine Systems Research Workshop 31
OES• Optical emission
spectroscopy (OES)• Utilizes a plasma
generator to identify chemical species
• Requires rapid thermal quenching of sample to halt chemical reactions
• SwRI has experience using OES for gas species analysis
11/10/2016 2016 University Turbine Systems Research Workshop 32
Test Stand Operation
• Shake down tests– Observed auto-ignition
combustion during shakedowns at full pressure and 80% temperature
• Component failures– Backpressure control valve– Mass flow controllers
• Viton rubber does not mix with sCO2
11/10/2016 2016 University Turbine Systems Research Workshop 33
Outline
• Phase I Overview– Background– Project Objectives– Phase I Progress
• Cycle Modeling• Chemical Kinetics• Preliminary Combustor Design• Bench-top Combustor Test
• Phase II Project Plan
11/3/2015 2015 University Turbine Systems Research Workshop 34
Phase II
• Complete detailed design• Fabricate combustor and test loop• Shake down and commission • Test combustor• Phase II duration: 3.5 years• Partnered with Thar Energy, Georgia Tech, UCF
and GE Global Research
11/10/2016 2016 University Turbine Systems Research Workshop 35
Detailed Combustor Design
• Develop more detailed and accurate combustion kinetic mechanisms
• Utilize CFD to study combustion flow field
• Detailed thermal and mechanical design
• Final design for manufacturing
11/10/2016 2016 University Turbine Systems Research Workshop 36
Hot Inlet Air
Combustion Processes
Cooling CO2
Combustor Integration withSunshot Test Hardware
11/10/2016 2016 University Turbine Systems Research Workshop 37
Combustor Integration• Utilize existing Sunshot
hardware• Install oxy-fuel combustor
– Demonstrate a direct fired oxy-combustor in a closed Brayton cycle
– Evaluate combustor performance
– Evaluate flue gas cleanup– Indirect heater allows for
various combustor inlet conditions to be studied
11/10/2016 2016 University Turbine Systems Research Workshop
Oxy-Combustor added downstream of indirect heater
Add flue gas cleanup and water separation
38
Planned Test Measurements
• Multiple OES sample locations• Temperature measurements• High speed pressure measurement for acoustic
phenomena• Study water dropout and separation• Possible measurements
– Optical access for advanced diagnostics– Materials sample testing
11/10/2016 2016 University Turbine Systems Research Workshop 39
QUESTIONS?
11/10/2016 2016 University Turbine Systems Research Workshop 40
ThermalInput