www.cranfield.ac.uk
Corrosion Life Model
Development for
Superheater Materials
under Oxy-fuel
Combustion
Conditions Nigel Simms
Adnan Syed
Tanvir Hussain
John Oakey
Outline
• Introduction
• Oxy-combustion systems
• Fireside corrosion environments
• Existing models
• Laboratory fireside corrosion testing
• Deposit recoat method
• Dimensional metrology
• Alloy performances – effect of temperature, SOx, deposition flux
• Development of alkali-iron tri-sulfate damage peak
• Coating performances – effect of deposition flux
• Model developments
• Summary
Conventional Power Plant Lifetime Extension
Air –fired conventional pulverised fuel
power system
Typical Power Station Boiler System
REHEAT CYCLE
Modification to Rankine
cycle (thermodynamic
cycle)
Remove the vapor for
reheat at constant pressure
in the boiler
& return it to the turbine for
continued expansion to
condenser pressure
Purpose
· Raises the temperature of the
steam generated above the
saturation level.
· Minimizes moisture in the last
stages of a turbine to avoid blade
erosion.
Conventional Power Plant Lifetime Extension
Oxy-firing with recycled flue gases
Typical Power Station Boiler System
REHEAT CYCLE
Modification to Rankine
cycle (thermodynamic
cycle)
Remove the vapor for
reheat at constant pressure
in the boiler
& return it to the turbine for
continued expansion to
condenser pressure
Purpose
· Raises the temperature of the
steam generated above the
saturation level.
· Minimizes moisture in the last
stages of a turbine to avoid blade
erosion.
B A
Oxygen
Oxy-fuel firing
Plant options
• Purity of oxygen
• Location of flue gas recycle take-off point
• Before any gas clean-up
• After particle removal
• After flue gas desulphurisation
• Further gas clean-up before recycle
• Dry primary combustion gas stream
• Dry secondary or tertiary combustion gas streams
Have direct effect on component operating conditions
Coal preparation
Air separation
CO2 compression
Condensation & CO2 separation
Steam turbine
CO2 storage
Sulphur
Electricity
Oxygen Combustor
Particle removal
Flue gas desulphurisation
Fly ash
Bottom ash
Dryer
Primary
Secondary
c b a1
2
Flow Diagram for
Component Life Modelling
Component
Specification
Operating
Conditions
Fuel Spec. &
Reactor System
Thermal Model Aerodynamic
Model
Thermochemical
Model
Transport & Deposition
Models
Mechanical
Property Data
Corrosion &
Erosion/Corrosion
Models
Life Predictions
· Contaminant effects
· Operating condition effects
Gas flow rate
P & T distributions
Inlet & outlet
gas P & T
Contaminant
levels & species
Particle
deposition
flux
Component design &
life criteria
Alloy
specificationDeposition flux &
composition
Metal surface
temperature
Component
Geometry
Damage
rates
Component
Specification
Operating
Conditions
Fuel Spec. &
Reactor System
Thermal Model Aerodynamic
Model
Thermochemical
Model
Transport & Deposition
Models
Mechanical
Property Data
Corrosion &
Erosion/Corrosion
Models
Life Predictions
· Contaminant effects
· Operating condition effects
Gas flow rate
P & T distributions
Inlet & outlet
gas P & T
Contaminant
levels & species
Particle
deposition
flux
Component design &
life criteria
Alloy
specificationDeposition flux &
composition
Metal surface
temperature
Component
Geometry
Damage
rates
Plant operating conditions for
heat exchanger tubes
• Fuel: coal / biomass
• Oxidant: air / oxygen
• Gas stream characteristics:
• Gaseous species – e.g. SO2, HCl, O2, CO2,
H2O, NOX, N2
• Vapour species – e.g. Na, K
• Particles
• From ash in fuel
• Condensed vapour species
• Gas temperature
• Heat exchanger characteristics:
• Water / steam temperature (& pressure)
• Metal temperature (& heat flux)
• Deposit
• rate of formation
• composition
Gas compositions
• UK vs S. American coal
• air vs oxy-fired (with hot recycled flue gas)
0.001
0.010
0.100
1.000
10.000
100.000
H2O CO2 O2 N2 Ar SO2 HCl
Vo
lum
e %
Gas species
UK coal, air-fired
S Amer, air-fired
UK coal, oxy-fired
S Amer, oxy-fired
Deposition on
Superheater / Reheater
Tubing
Vapour species
Condensation
onto solid particles
& aerosolsSolid particles &
aerosols
Condensation into
deposit
Coarse particles
stick
Thermophoresis of
fine particles
Water /
steam
Corrosion
Heat transfer
SOx
HCl
Vapours, SOx &
HCl diffuse in
porous deposits
Vapour species
Condensation
onto solid particles
& aerosolsSolid particles &
aerosols
Condensation into
deposit
Coarse particles
stick
Thermophoresis of
fine particles
Water /
steam
Corrosion
Heat transfer
SOx
HCl
Vapours, SOx &
HCl diffuse in
porous deposits
Deposition mechanisms:
Particles:
• Direct inertial impaction
• Thermophoresis
• Eddy diffusion
• Brownian
Vapour:
• Direct condensation
• Condensation on particles
Fuel derived deposit compositions (coal / biomass)
Deposit compositions:
• Si-Al-O compounds
• can fix Na, K if particle temperatures high enough
• Al only if coal co-fired
• Ca/Mg carbonates / sulphates / chlorides
• Na / K sulphates / chlorides
• Fe sulphates / chlorides / oxides / sulphides
• Phosphates – from biomass
Important factors
• Minerals in fuels
• Balance between elements
Corrosion aggravated by:
• Low melting point deposits
• High chloride deposits
Phase diagram for alkali sulphates – chlorides
(Lindberg, 2007).
Note: temperatures in Kelvin.
Sensitivity of SO2 vs HCl to changes in cereal co-product (CCP) or typical
wheat straw co-firing with two coals compared to example biomass
0
200
400
600
800
1000
1200
0 200 400 600 800 1000 1200 1400 1600
HC
l (v
pm
)
SOx (vpm)
UK coal + straw (wheat, typical)
UK coal + CCP
South American coal + straw (wheat, typical)
South American coal + CCP
Biomass (examples)
100 % S.Am. coal
100 % UK coal
Increase in co-firing
100% wheat straw
100% CCP
Simms (2011), in Power Plant Life Management and
Performance Improvement (adapted from Natesan et al., 2003)
Effect of metal temperatures on corrosion rates in
conventional pulverized-fuel- fired power systems
(iron-based alloys)
Overall characteristic
bell-shaped fireside /
hot corrosion peak
Temperature
Corrosion
damage Stable
deposit
Deposit
unstable
Oxidation damage
Simms, 2011
Temperature
Corrosion
damageStable
deposit
Deposit
unstable
Oxidation damage
Temperature
Corrosion
damageStable
deposit
Deposit
unstable
Oxidation damage
High temperature corrosion mechanisms
Corrosive deposits
• Sulphate deposits
• Pyro-sulphates (e.g., K2S2O7)
• Alkali-iron tri-sulphates (e.g., K3Fe(SO4)3)
• Mixed sulphates (e.g., K,Na,Fe)XSO4
• Chloride deposits – mixed
• Carbonates – mixed
• Sulphate – chloride – carbonate ‘soup’
• Molten vs sticky vs solid
Deposit instability
• Vapour condensation
dewpoints
• SO3 needed to stabilise some
sulphate phases
• SO3/SO2 balance favoured at
lower temperatures
• Other phases more stable with
change in deposit temperature
Possible corrosion mechanisms in sodium chloride /
sulphate dominated deposits
Chloride based
mechanism Sulphate based
mechanism
0
10
20
30
40
50
60
70
80
90
600 610 620 630 640 650 660 670 680 690
Metal temperature (°C)
Meta
l lo
ss (
µm
/100
0 h
ou
rs)
0.12
0.2
0.3
0.4
0.5
0.58
Superheater – austenitic eqn
(modified PE8)
0
20
40
60
80
100
120
600 610 620 630 640 650 660 670 680 690
Metal temperature (°C)
Meta
l lo
ss
(µ
m/1
000 h
ou
rs)
800
1000
1200
Effect of Cl%
Effect of Tg
r = A.B.(Tg/G)m.((Tm-C)/M)n.(Cl%-D)
•CEGB eqn
•UK coal only
•Cl% represents fuel composition
effect
•Other factors needed for other coals ?
Biomass ? & oxy-firing ?
Steps in model developments –
completing all the links
Fuel
composition
Gas
temperature Deposit
compositions
& flux
Gas
compositions
Metal
temperature
Corrosion
rate
Comparison of outputs for current models
• 18 % Cr steel
• Daw Mill coal
• Three current model versions
• PE8 modified (CEGB)
• Re-analysed PE8 data
• Version 3 of new linked combustion-deposition-corrosion models
Corrosion degradation sequence
• The incubation stage corresponds to the presence of protective oxides
• Propagation occurs when protective oxides no longer form
• The incubation period and hence component life is reduced as corrosion becomes more aggressive.
• The end of incubation stage follows the local penetration of the salt through the scale and subsequent spreading along scale-alloy interface
Incubation Propagation
Dam
ag
e
Time
Corrosion model requirements
• Specific to:
• Components, e.g. superheater / reheater, evaporator
• Systems, e.g. pf combustion
• Predictive models for fireside corrosion damage as a function of:
• Gas composition (e.g. SOx, HCl, H2O)
• Deposit composition
• Na, K, Fe, Ca
• Sulphate vs chloride
• etc
• Deposition flux (mg/cm2/hour)
• Metal temperature
• Cumulative probability level required, e.g.: 4%,10% or 50% (i.e. median)
• Time
• Alloy compositions
• Validation corrosion data / model frameworks from plant data
• but note differences between plant data & laboratory data
• Gas-surface temperature difference vs isothermal
• Deposition governed by fuels, design + operating conditions
vs simulated deposits
• Component life criteria ?
Laboratory hot corrosion testing
Critical parameters
• Metal temperature
• Gas composition
• Deposit composition
• Deposition flux
• Coating / substrate composition
Deposit recoat technique
• Simple ‘simulation’ of deposition flux
• Allows control of deposit composition
• Needs multiple deposit recoats (ideally >5)
Controlled atmosphere furnaces
• Allows specific gas compositions to be tested
• Alumina lined reactors
• Exposure temperatures controlled to +/- 3-5°C
Samples manufactured from tubes / bars
• Machined
• Standard surface finish
• Precision for dimensional metrology
Schematic of a controlled
atmosphere furnace (oxy-
firing)
Scrubbers
DI
water
Gas B
(SO2/CO2/O2)
Gas A
(HCl/N2/CO2)
Alumina liner
Crucibles
holding
samples
VentMass flow
Controller 1
Mass flow
Controller 2
Mass flow
Controller 3
Gas C
(CO2)
Pump Stainless steel
reaction vessel
-5 C
+5 C
-5 C
Ho t
zone
Alumina tube
Pump
Hot water bath
Thermostat
Condensate
Sample Metrology &
Data Analysis (1)
Aim:
• To produce corrosion data in terms of metal loss
• To show variability of damage around samples
• To generate statistical data suitable for modelling purposes
• To overcome limitations of mass change methods
Method:
• Pre-exposure: contact metrology
• Post-exposure: measurements on cross-sections
e) original sample shape
c) microscope stage
sample
Dimensional Metrology: Technique
(Correction for systematic errors also required)
Range of shapes can be studied
b) d) image
internal damage
rad
ius
f) m
eta
l lo
ss
location
sam
ple
8x
diameter a)
g)
cumulative probability
met
al l
oss
corroded sample shape
Laboratory Gas Compositions
– Superheater/Reheater Environments
Gas
O2 CO2 H2O N2 SO2 HCl
% vol % vol % vol % vol vpm vpm
Gas 1 4 13.4 8.6 Bal 1150 325
Gas 3 4 59.1 30.9 4.4 6150 1700
Deposit No. Nominal deposit compositions (mole %)
Kaolinite* Na2SO4 K2SO4 KCl Fe2O3
D0 (bare) - - - - -
D1 - 37.5 37.5 25
D2 80 7.5 7.5 5
D3 80 7 7 1 5
D4 80 5 5 5 5
* Al2O3.2SiO2.2H2O
650°C – deposit D1 – oxy-fired gas environment
-500
-450
-400
-350
-300
-250
-200
-150
-100
-50
0
0 10 20 30 40 50 60 70 80 90 100
Chan
ge in
met
al (μ
m/1
000h
rs)
Cumulative Probability (%)
T92
347HFG
HR3C
625
-1400
-1200
-1000
-800
-600
-400
-200
0
0 20 40 60 80 100
Ch
an
ge i
n m
eta
l (μ
m/1
000h
rs)
Cumulative Probability (%)
600°C 650°C
700°C 750°C
Alloy 347HFG covered with deposit D1 exposed to 59% CO2/31%H2O/4%O2/6260vppm
SO2/1700vppm HCl for 1000h
Fireside corrosion –
temperature and metal loss
distributions
Fireside corrosion damage to alloys
exposed with alkali iron trisulfate
(deposit, D1)
(a) simulated air-fired combustion gases
(b) simulated oxy-fired combustion
gases (hot gas recycle option)
0
200
400
600
800
1000
1200
T92 347HFG HR3C Alloy 625
Me
tal l
oss
(µ
m)
Alloy
600 °C
650 °C
700 °C
0
200
400
600
800
1000
1200
T92 347HFG HR3C Alloy 625
Me
tal l
oss
(µ
m)
Alloy
600 °C
650 °C
700 °C
750 °C
Deposit D1 = 37.5 mole % Na2SO4 – 37.5 % K2SO4 – 25 %Fe2O3
Basis for fireside corrosion development
metal loss = kincub*tincubm + kprop*(t – tincub)n
Metal
loss
Time For aggressive deposits / susceptible
materials: tincub ~ 0; n=1
tincub
Deposit compositions
(mol%)
Deposits Kaolinitea Na2SO4 K2SO4 Fe2O3
100% Alkali (D1) 0 37.5 37.5 25
80% Alkali (D1b) 20 30 30 20
40% Alkali (D1c) 40 22.5 22.5 15
60% Alkali (D1d) 60 15 15 10
20% Alkali (D2) 80 7.5 7.5 5
a Al2O3.2SiO2.2H2O
Deposit comp = Simulated fly ash + x % [3(Na/K)2SO4 + Fe2O3]
Where x = 100, 80, 60, 40, 20
HVOF alloy 625
- change in metal vs
cumulative probability
(1000 h at 650 ºC)
-300
-250
-200
-150
-100
-50
0
0 10 20 30 40 50 60 70 80 90 100
Ch
an
ge i
n M
eta
l (µ
m)
Cumulative Probability (%)
100% Alkali
80% Alkali
60% Alkali
40% Alkali
20% Alkali
Incubation time as function of alkali flux
for HVOF alloy 625
Alloy 625
Flux of deposit D1* (µg/cm2/h)
* Deposit D1 = 37.5 mole % Na2SO4 – 37.5 % K2SO4 – 25 %Fe2O3
HVOF alloy Ni - 50Cr
- change in metal vs
cumulative probability
(1000 h at 650 ºC)
-300
-250
-200
-150
-100
-50
0
0 10 20 30 40 50 60 70 80 90 100
Ch
an
ge i
n M
eta
l (µ
m)
Cumulative Probability (%)
100% Alkali
80% Alkali
60% Alkali
40% Alkali
20% Alkali
Characteristics of alkali
iron trisulphate fireside
corrosion peak (1)
Temperature
Corrosion
damage
Oxidation damage
Oxidation of alloy - rate depends on:
• Alloy composition
• Gas composition
• Alloy surface preparation
• Metal surface temperature
Characteristics of alkali iron
trisulphate fireside corrosion
peak (2)
Temperature
Corrosion
damage
Oxidation damage
Deposit-induced alloy corrosion – increasing rate depends on:
• Alloy composition
• Deposit composition (e.g. liquid deposit formation)
• Gas composition (e.g. SO3 above that needed to stabilise
deposit)
• Deposition flux
• Alloy surface preparation
• Metal surface temperature (must be above that required for
formation of stable deposit) …..
Stable
deposit
Characteristics of alkali iron
trisulphate fireside corrosion
peak (3)
Temperature
Corrosion
damage
Oxidation damage
Deposit-induced alloy corrosion – decreasing rate depends on:
• Deposit instability (e.g. reduced gas SO3 levels)
• Metal surface temperature (above that needed for stable
deposits)
• Gas composition
• Alloy composition
• Deposition flux …..
Stable
deposit
Deposit
unstable
Characteristics of alkali iron
trisulphate fireside corrosion
peak (4)
Temperature
Corrosion
damage
Oxidation damage
Stable
deposit
Increasing
deposition flux
Deposit
unstable
Increasing
SO3
Schematic view -
S/H fireside corrosion
propagation for
18 Cr steel
Temperature
Med
ian
meta
l lo
ss
Oxy-fired – hot gas recycle
Air-fired
Increasing SOx
650–675°C 700–725°C
SO3 / SO2
Molten salt
corrosion
Summary
• Corrosion in pulverised fuel fired systems results from a complex set of
processes
• To correctly assess materials challenges need to look at whole process - and
competing reactions at each stage:
• Fuel composition (wide range of different coals/biomasses need considering)
• Combustion (air vs oxy-fired variants)
• Deposition (gas and particle transport)
• Corrosion
• Determine effect of changing exposure conditions (including time,
temperature) on:
–Damage levels
–Mechanisms
• Dimensional measurements (not mass changes)
–Metal losses, internal damage, ….
• Generate understanding of damage development
Beware changes in mechanisms with changes in exposure conditions
Summary
• For corrosion modelling
• Need quantitative metal loss distribution data on the effect of changes in
exposure conditions on damage levels and mechanisms
• Identify similarities and differences in data and changes in exposure
environments
• Care needed with laboratory testing
• Setting appropriate and well-controlled exposure conditions
• Using appropriate monitoring techniques
• Dimensional metrology
• Microstructural characterisation
• Accelerated testing
• Desirable to save time / money
• Needs good understanding of particular damage mechanism(s)
• Aim of test work
• ranking of materials or generation of appropriate data for modelling ?
• Coatings vs base alloys ? Performance vs overall costs ?