Progress in Oxyfuel Combustion - Relcom - Reliable … in...replicates Schwarze Pumpe plant layout Analysis of rig data allows validation of boiler models TOTeM 41 “Optimisation

Progress in Oxyfuel

Combustion

G Hesselmann

TOTeM 41 “Optimisation of OXY/COAL/RFG Systems”

10th ~ 11th June 2014

Warsaw, Poland

Content

Early days

– Proof of concept

– High level feasibility studies

– Interim assessment

Taking the concept to demonstration

– Full scale burner demonstration

– Full cycle testing

Moving from rig to plant

– Heat transfer

– Engineering (FEED) studies

Summary

Concluding Remarks

A review of over two decades of oxyfuel progress from a boilermaker's perspective

2 TOTeM 41 “Optimisation of OXY/COAL/RFG Systems”

Proof of Concept

Demonstrated oxyfuel firing concept at 160kWt in

Renfrew

– CO2 typically 80 to 85%v/v dry; 95% max

– NOx reduces with flue gas recycle rate

– Early data on slagging and fouling effects (world-first by

“industry”)

– Early data on impact of oxyfuel on ash pozzolanic

activity (world-first)

– Smooth transition from air to oxyfuel firing

– Many practical lessons learned (e.g. need to use dry

primary FGR) that are universally applied today

Demonstrated oxyfuel firing concept at 2.5MWt in

Ijmuiden (IFRF #1 furnace)

– CO2 typically 82%

– Air in-leakage a key issue

– NOx reduces with flue gas recycle rate

– Turndown to 70% load demonstrated

– Many practical lessons learned

European Union collaborative project JOU2-CT92-0062 undertaken over period

1992~1995; 1st pilot-scale demonstration of oxyfuel in Europe by Doosan & IFRF


160kWt Renfrew Test Facility

FO.SPER Furnace (replaces IFRF #1)

Feasibility

Numerous high level feasibility studies undertaken

– Considered effect of coal rank (bituminous, sub-bituminous), plant location (coastal, inland), plant size

(150~700MWe), boiler type (natural circulation, supercritical), and new-build vs. retrofit

Similar findings obtained from many studies

– For a plant with 90% CO2 capture rate, the typical efficiency loss is 9~10%age points

• The penalty can be reduced by plant integration (e.g. to recover usable heat)

– The efficiency penalties for Oxyfuel and PCC (amine) capture technologies are comparable

• For oxyfuel the major losses arise from the electrical power to drive the compressors in the ASU & CPU

• For PCC the major losses arise from lost generation due to steam extraction from the turbine

• Both processes require electrical power for compression to pipeline pressure

– Levelised CO2 capture cost and electricity cost for Oxyfuel and PCC (amine) are comparable

– Key specific issues were identified relating to:

• Capture plant energy penalty and steam cycle matching

• Waste streams and emissions performance

• Utilities requirements

• Footprint and layout requirements

• Safety and operability

High level studies indicate costs & issues associated with oxyfuel……


……but they cannot answer the “hard questions”

Status after “early work”

Is it possible to design an oxyfuel fired boiler, and be confident that it will perform as we expect?

Combustion

■ Can a stable flame be achieved?

■ Can the burner be operated flexibly?

■ What is the burner’s performance?

■ Is the process sufficiently robust that its operation can be automated?

Thermal Performance

■ How does Furnace Exit Gas Temperature (FEGT) change?

■ How does the convective pass performance change?

■ Is modification of the heat transfer surfaces required?

■ Can existing thermal performance tools be used for oxyfuel fired plant design?

Combustion and furnace thermal performance are the main areas of uncertainty


Full-Scale Burner Demonstration

Design based on our current Mk III low NOX axial swirl

burner

– Proven design with over two decades of operational experience

in numerous coal-fired boilers worldwide

– Applicable to new build and retrofit coal-fired boilers.

Volumetric flow of the primary gas for oxyfuel firing

maintained as per air firing

– Coal transport considerations

– Oxygen content of the primary gas controlled to 21%v/v dry

– Safe operation of coal milling plant

Overall stoichiometric ratio controlled to ~1.2

– Maintain combustion efficiency

Flue gas recycle rate chosen on consideration of the

adiabatic flame temperature and furnace heat transfer

characteristics

Other OEM burner suppliers have adopted a similar

approach

– Most burner OEM suppliers are also looking at enhanced designs

for oxyfuel firing

The 40MWt OxyCoalTM burner was based on our existing low NOx air-fired burner


Full-Scale Burner Demonstration

A full scale 40MWt OxyCoal™ burner was successfully

demonstrated on air and oxyfuel firing, achieving safe

and stable operation across a wide operational envelope

Oxyfuel flame stability and flame shape was comparable

to air firing experience

Safe and smooth transitions between air and oxyfuel

operation were demonstrated

Realistic CO2 levels were achieved (in excess of 75%

v/v dry, and up to 85% v/v dry)

40MWt OxyCoal™ burner turndown proven from 100%

load to 40% load – a comparable turndown to Doosan

Power Systems’ commercially available air firing low

NOx axial swirl burners

NOx and SO2 is significantly lower under oxyfuel firing

compared to air firing

Combustion efficiency under air and oxyfuel conditions,

as expressed by CIA and CO, is comparable

Successful demonstration of a full-scale burner is a key step towards commercial

implementation of oxyfuel technology


Air Firing

Oxyfuel Firing

Full-Cycle Testing

Doosan joined the Technology Partnership

for the Oxyfuel Pilot Plant (OxPP) project

– Agreement between Vattenfall Europe

Technology Research GmbH and Doosan

signed in December 2010

– The purpose of the pilot plant was to validate

engineering, to investigate and better understand

the dynamics of oxyfuel combustion, and to

demonstrate the capture technology

The 30MWt OxyCoal™ burner demonstrated good flame stability over a wide range

of operating conditions, while maintaining low levels of NOx and CO


Air Firing Oxy Firing

Automatic control logic developed, demonstrated

safe and smooth transitions between air and oxy

firing, and vice versa.

Stable anchored flame maintained at all stages of

the transition

300 hours operation on air firing.

2500 hours operation on oxy firing.

Steady oxy firing operation for extended periods -

a requirement for parallel test measurements.

Combustion performance optimised to achieve set

targets.

– Excess O2 < 3 vol% (wet)

– NOX < 250 mg/Nm3 (air) < 800 mg/Nm3 (oxy)

– CO < 50 mg/Nm3 (air) < 100 mg/Nm3 (oxy)

Moving from rig to plant

Large scale testing has focused on combustion performance, and has provided strong

experience on

– Emissions (NOx, CO, SO2, etc.)

– Flame length

– Flame stability

– Turndown

– Operability (air to oxyfuel transition, etc.)

– As a result of large scale testing we have a good understanding of how a burner will perform under air

and oxyfuel firing conditions, and how it will respond to changing process parameters (e.g. recycle rate)

Large scale testing has provided significant intangible benefits

– Health & safety

– Confidence in the oxyfuel process

– Resolution of unforeseen problems

Large scale testing can, potentially, provide information on heat transfer

– But there are issues……

In moving to plant, it is important to understand what test rigs can (and cannot) tell us


10

Heat Transfer - Limitations of Test Facilities

Test furnaces (CCTF, OxPP) were designed to

demonstrate combustion performance, but (in

principle) it should be possible to extract

thermal performance data as well.

However…….

Test furnaces cannot adequately replicate the

radiation processes in utility plant

Specific issues include

– Realistic mean beam lengths

– Estimation of extinction coefficient

– Pendant (radiant) superheaters

– Volumetric utilisation of the furnace

Use large scale testing to generate data to

support the verification of predictive models

– Verification is not validation, some uncertainty

expected to remain

Plant scale demonstration is needed to verify thermal performance on oxyfuel fired boilers Triatomic Gas Emissivity Comparison

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

0 5 10 15 20 25 30

Mean Beam Length (m)

Ga

s E

mis

siv

ity

(-)

Utility Boiler Furnaces

Large Test

Facilities

Air Firing

Oxyfuel Firing


Heat Transfer – Importance of Flue Gas Recycle (FGR)

•Increased recycle

flow leads to:

Greater mass per

unit heat input →

lower adiabatic flame

temperature and less

radiant heat transfer

Greater mass flow

through boiler →

higher gas velocity

and more convective

heat transfer

Note – flue gas

recycle is one of the

“standard”

approaches to reheat

steam temperature

control in large utility

boilers

Recycle flue gas flow rate can be used to vary radiant and convective heat transfer

Source: IFRF Report F98/Y/1


Heat Transfer - Predictive Approaches

Simple semi-empirical models (e.g. stirred furnace models)

– Robust, easy to use, fast

– Require empirically derived factors

– Unreliable when extrapolated beyond validated experience

Zone models (e.g. Doosan’s proprietary “HotGen” model)

– Robust, easy to use, fast, based on sound theoretical principles, optimised for furnace design

– Some inputs (e.g. dirt factors) empirically derived by “calibrating” model to plant

– Can be used beyond validated experience (with care)

Computational Fluid Dynamics

– Can be difficult to converge (not robust), difficult to use (need experts), slow

– Some inputs (e.g. dirt factors) empirically derived by “calibrating” model to plant

– Can be used beyond validated experience (with care)

Models can predict the impacts, but require reliable inputs

– It is challenging for models to predict soot formation in flames (impact on flame luminosity, gas

extinction coefficient) and flame shape/length (heat release profile)

– Full-scale burner testing provides information to support the use of engineering models to predict

furnace performance

Engineering models for thermal performance prediction


Heat Transfer – Furnace Thermal Performance (HotGen)

HotGen uses Hottel’s zone method to predict furnace thermal performance

Predicts overall and local performance (FEGT, heat to walls, heat to platens,

heat flux profiles, etc.)

Furnace volume divided into discrete blocks – “zones” – Temperature and physical properties are implicitly assumed to be uniform within each

zone

– Smaller zones better justify this assumption

Energy balance solved for enthalpy for each zone – Radiation, convection, heat release in zone

– Gas absorption (extinction) coefficient defines fraction of radiant heat absorbed within

zone vs. fraction passing through (optical thickness)

Monte Carlo approach used to calculate radiant heat transfer between zones – Discrete packages of energy from each zone (of random strength and direction) are

tracked through furnace volume

Sound theoretical basis – Implicitly handles thermal radiation issues (furnace size, impact of flyash, etc.)

– All inputs have physical significance (model is truly predictive, unlike simpler semi-

empirical models)

– Can accommodate oxyfuel firing

Robust – Can simulate all the main technologies (wall firing, downshot firing, tangential firing)

– Short run times with no convergence problems (unlike CFD)

Absorbed Heat Flux

to Furnace Side Wall

(kW/m2)


Heat Transfer – Boiler Thermal Performance (SteamGen)

Predicts furnace and boiler thermal performance

Overall Furnace Performance – FEGT, heat to walls,

heat to platens

– Uses simple empirical furnace model (HotGen is considerably

more sophisticated)

Overall Boiler Performance – bank absorptions, gas

temperature at bank exit

– Model accommodates spray, split rear pass, and FGR steam

temperature control methods

– Detailed representation of plant – banks, cavities, screens,

enclosures, sling tubes, etc.

– Accurately describes gas and steam paths (network)

Design and performance modes

Performance mode

– Calibrate model to plant data and derive bank performance

factors

– Feed back plant operating experience into design process

– Quantify operating experience for novel scenario’s, e.g. oxyfuel

operation

Design mode

– Specify factors and predict performance

SteamGen model

replicates Schwarze

Pumpe plant layout

Analysis of rig data allows validation of boiler models


Heat Transfer – Using “Flame” Information for Furnace Performance

Heat release & gas extinction coefficient

– Flame length and shape

• Similar for air and oxyfuel firing, implying similar heat release

– Flame temperature & luminosity

• Air firing has much higher flame temperature (~200°C), leading to increased radiation to furnace walls

• Air firing has a brighter flame (note reduced exposure time for air firing photo!), also leading to increased

radiation to furnace walls

• But strong dependency on FGR rate; at lower FGR rates flame luminosity is similar

• Based on observations, a reasoned judgement can be made on the gas extinction coefficients

Flame effects are more dominant in single burner furnaces

Large test rig experience can help in specifying inputs for furnace models


Heat Transfer - Limitations of Large Test Facilities (Revisit)

CIUDEN’s es.CO2 test facility addresses some of the issues raised for heat transfer

testing of radiant furnaces

Specific issues include

– Realistic mean beam lengths (test furnaces are an order of magnitude smaller)

– Pendant (radiant) superheaters (test furnaces generally do not have pendants, and avoid radiant surfaces)

– Volumetric utilisation of the furnace (combustion zone in single burner furnaces vs. multi-burner furnaces)

Schwarze Pumpe

Single burner

Convective SH

Small furnace

CIUDEN

Multi-burner (4)

Radiant SH

Small furnace

Plant

Multi-burner (36)

Radiant SH

Large furnace TOTeM 41 “Optimisation of OXY/COAL/RFG Systems”

Heat Transfer - OxPP vs. es.CO2

CIUDEN‘s es.CO2 test facility complements Vattenfall‘s OxPP test facility

Vattenfall

Oxyfuel Pilot Plant (OxPP)

– 30MWt thermal input

– 1 x 30MWt burner

– Down fired

– Pre-dried lignite

– Radiant natural circulation furnace

– Convective superheater

– Convective economiser

– Spray attemporation

– Demonstration of near full-scale burner

CIUDEN

Technology Development Centre for CO2

Capture (es.CO2)

– 20MWt thermal input

– 4 x 5MWt burners

– Opposed wall fired

– Pre-dried bituminous coal

– Radiant natural circulation furnace

– Radiant + Convective superheaters

– Convective economiser

– Spray attemporation

– Investigation of burner interaction

– Investigation of furnace heat transfer in a

realistic arrangement

– Air and oxyfuel testing at es.CO2 is a key

activity in the RELCOM project TOTeM 41 “Optimisation of OXY/COAL/RFG Systems”

FEED

Unit #1

125MWe

Downshot boiler firing domestic

anthracite and heavy fuel oil

In-service 1973

OEM was Babcock Hitachi KK, boiler

was built under license from Doosan

and is on our reference list

Steam Conditions

– Evaporation (tonne/h) 420

– Main Steam Pressure (bar) 128.5

– Main Steam Temperature (°C) 541

– Reheat Steam Pressure (bar) 30.9

– Reheat Steam Temperature (°C) 541

– Cycle Efficiency 36%

In 2012 Doosan completed a FEED study for the retrofit of Oxyfuel to KOSEP's

Young Dong #1 Power Station


FEED

Case 2 selected – lowest cost solution with acceptable technical risk


Feasibility

Cases

Retrofit Retrofit New Boiler

CASE 1 CASE 2 CASE 3A CASE 3B

Firing System Downshot-Fired

Circular Burners Wall-Fired Circular Burners + New Windbox

Mill + PF Ducts Vertical Mill + New PF Piping, New PAF

Plant Layout Maximise use of boiler island components

New PAH, Refurbished SAH New

Boilerhouse

Prim. Steelwork Existing New

FEED

•Case 2 - Conversion to wall firing

Wall firing configuration leads to better

volumetric utilisation of the lower furnace

Potential for steam temperature shortfall

Via modelling, a solution was identified to

ensure that the required design steam outlet

conditions were retained across the control

range for the boiler

•Thermal performance results

Heat exchange surfaces behave similarly for

Air and Oxyfuel firing

Design performance achieved across full

load range (final steam conditions achieved)

Improved heat flux distribution (lower peaks)

for oxyfuel firing

No requirement to change or modify plant

convective pressure parts

Detailed boiler thermal performance assessment of Young Dong “Case 2” using

Doosan in-house codes BWHOT (Furnace) and SteamGen (Convective Pass)


FEED

Successful OxyCoal™ burner testing in Renfrew demonstrates Doosan oxyfuel burner

capability for Young Dong

Optimal solution for Young Dong Unit 1 identified

– Conversion to opposed-wall fired arrangement

– Firing bituminous coal

– Keep existing furnace envelope

Thermal performance models show design steam conditions can be delivered across full

control range

Conversion to bituminous coal and oxyfuel firing is technically viable

However both the Young Dong retrofit and the Janschwalde new build projects are not

proceeding to construction

An oxyfuel solution can be found for even the most difficult of retrofit opportunities


Summary of Oxyfuel Development in Doosan

Over 20 years development has moved us from concept to reality


Proof of Concept Feasibility Full-Scale Burner Full Cycle FEED

160KWt testing of

oxyfuel concept;

one of first

industrial scale

tests in Europe;

many lessons

learned.

Numerous studies

for retrofit & new

build; performance

prediction & design

tool development;

underpinning R&D.

40MWt hard coal

OxyCoalTM burner

tested in Renfrew,

30MWt pre-dried

lignite OxyCoalTM

burner tested at

Schwarze Pumpe.

30MWt OxyCoalTM

burner tested at

Schwarze Pumpe

in full cycle plant

(ASU, boiler, CPU)

with fully automatic

control.

•100MWe Young

Dong retrofit

(KOSEP, Korea).

•250MWe

Janschwalde new-

build (Vattenfall,

Germany).

1992 2010 2012 Present Day

Concluding Remarks

Testing of Doosan’s OxyCoalTM burner has demonstrated we have a viable combustion system

– Good flame stability, including during air↔oxyfuel transitions and at reduced load

– Achievement of required NOx and CO emissions

– Flexible operation, automated control logic

Large scale burner testing and boiler OEM experience has taken us to the point where we can

be confident in the thermal design of an oxyfuel fired boiler

– We can mitigate any remaining uncertainty through operational measures (FGR rate, etc.)

Considerable progress has been made in the development of oxyfuel technology

– The process is technically viable

– The process is reasonably well understood

– The process has been demonstrated at pilot scale

– The process has been demonstrated at large scale (30MWt +)

– Most of the individual components are in commercial operation at the required scale

• ……..now all we have to do is to convince the public and the politicians to take action

Over two decades of development has got us to the point where we have advanced

the oxyfuel technology sufficiently to demonstrate it at utility scale…..


Thank you

Documents

Progress in Oxyfuel Combustion - Relcom - Reliable … in...replicates Schwarze Pumpe plant layout Analysis of rig data allows validation of boiler models TOTeM 41 “Optimisation