Development in Oxyfuel Combustion Technologies...

Development in Oxyfuel Combustion Technologies forOxyfuel Combustion Technologies for

Coal Fired Power Plants with CCS(Part 3: Flue Gas Processing and CO2 Processing Unit)(Part 3: Flue Gas Processing and CO2 Processing Unit)

Stanley SantosStanley SantosIEA Greenhouse Gas R&D Programme

Cheltenham, UK

Instituto de Inginieria UNAM28th March 201228 March 2012

Oxyfuel Perspective…Oxyfuel Perspective…

What is the best a to manage the nonWhat is the best way to manage the non-CO2 components to meet any CO2

specification requirement and maximize some of the implicit advantages i.e. p gefficiency and zero emissions ?

22

What are the impurities ?

Coming from coalH2O Other Impurities to consider…PMSOxH

Other Impurities to consider…• O2, N2, Ar, water from air ingress• NH3 from SCR/SNCR• Other trace elements

HgHCl, HF

Coming from oxygenComing from oxygenO2N2Ar

Coming from coal and oxygenNOxN2OCO

The world leader in gases for industry, health and the environment 3

CO

What are the locations to manage impurities ?

Air Separation Unit (ASU)

Coal handling and preparation

Boiler

Environmental Island

CO2 Compression and Purification Unit (CO2 CPU)


What are the locations to manage impurities ?

PM H2O SOx NOx HClHF

Hg N2

ArO2 CO

ASU

Coal handling & preparation

BoilerBoiler

EnvironmentalEnvironmental Island

CO2 CPU2

2 impurities are key for their impact on LCOE (CAPEX and OPEX) : H O and SOx


OPEX) : H2O and SOx

What are the best locations to manage impurities ?

PM H2O SOx NOx HClHF

Hg N2

ArO2 CO

ASU

C l h dliCoal handling & preparation

B ilBoiler

E i t lEnvironmental Island

CO CPUCO2 CPU

Environmental island functions could be significantly reduced


Environmental island functions could be significantly reducedCO2 CPU becomes the main emission control equipment

Flue Gas Processing UnitFlue Gas Processing Unit

7

Flue Gas Processing Units(Summary of Results @ Schwarze Pumpe Pilot Plant)

Test on Electrostatic Precipitator (ESP)• Test on Electrostatic Precipitator (ESP)• Completed and Results better than expected as

compared to air firing.• Test on Flue Gas Desulphurisation (FGD)Test on Flue Gas Desulphurisation (FGD)

• Different design – with external oxidation tank.SO2 l ( hi d t 99 7% l)• SO2 removal (achieved up to 99.7% removal)

• SO3 removal (achieved ~65% reduction)

8

Process overviewThe Oxyfuel pilot plant Schwarze Pumpe

Confidentiality - Medium (C2)

9 | 2nd Oxyfuel Combustion Conference | 2011.09.05

Experiences from the pilot plantImpact on slurry chemistry

• No major differences have been encountered.• No negative impact of higher CO2 partial pressure on reaction system.

- CO2 seems to enhance the limestone/chalk dissolution- Degassing of dissolved CO2 in the external reaction tank.

• Slightly lower pH during Oxyfuel operation• Higher moisture content led to water condensation in pre-cooler during Oxyfuel conditions.

Water condensation could be prevented by reduction of cooling duty to avoid local sub coolingWater condensation could be prevented by reduction of cooling duty to avoid local sub-cooling.

2141540 1510

200

220

240

1500

1650

1800

166

1143 1123

140

160

180

1050

1200

1350Oxyfuel Air-firing

Samples taken during operation withouthydrocyclone Actual

98,8 94,25

60

80

100

120

4 0

600

750

900

hydrocyclone. Actualvalues (especiallyChloride) are expected to be higher.5,89

0,456,3 4,55

220 248

0

20

40

60

0

150

300

450



pH Gypsum [%]

Carbonate [%]

HCO3 [mg/l]

Nitrate [mg/l]

Chloride [mg/l]

Desity[mg/cm³]

Experiences from the pilot plantImpact of the operating conditions on the removal rate – Slurry flow rate



Experiences from the pilot plantPerformance parameters – SO3 removal rate

• Early measurements indicated low removal rate for SO3 (approx. 40%)

• Probably caused of aerosol formation by “shock cooling” due to the local surfaceProbably caused of aerosol formation by shock cooling due to the local surface temperature (Formation H2SO4 aerosols from SO3 → only marginally removable from flue gas)

• After modifications good removal rate could be achieved.

15

average 12,83

10

12,5

mg/

Nm

³, dr

y]

65%

11,55

13,98

5

7,5

once

ntra

tion

[m -65%

2,96

6,71

0

2,5SO3

co

average 4,57



0SO3 after FGD SO3 upstream FGD

FGC System

• Flue Gas Condenser operates very similarly tooperates very similarly to typical flue gas condensers used in various biomassused in various biomass fired CHP plant in Sweden.

• Potential improvement for heat recovery has been identified by various equipment manufacturers

13 | Lars Strömberg, IEAGHG OCC2 Australia | 2011.09.12

CO2 Processing UnitCO2 Processing Unit

14

Recti-fication

Compressor building

fication

Activated carbon filter

Analysis ycontainer

CO2-tanks(2x180 m³)

Trailer docking station

Overview of Development of CPU th l t 5CPU over the last 5 years...• Recognition of the NOx and SOx reaction by Air• Recognition of the NOx and SOx reaction by Air

Products (presented during GHGT Conference –June 2006)June 2006)• This has led to the rapid technology development among

the industrial gas producersthe industrial gas producers.• Identification of potential impact of Hg to the

operation of the CPUoperation of the CPU.• Development of the use of impure CO2 as

refrigerant driven mostly by reducing energy penalty.

16

• Work on further recovery of CO2 in the vent of CPU.

Ai P d t O F l CO C t d P ifi ti Air Products Oxy-Fuel CO2 Capture and Purification –with Air Products PRISM® Membrane

Product CO2

Boiler Steam Cycle

Commercial

Mature Tech, Needs Data

Needs Engineering Data

Heat Recovery CO2 Compression

Commercial

Raw Flue Gas

SourCompression &AP Acid Gas C l

TSA UnitMercury Removal

Auto‐Refrigerated N2, Ar, O2RemovalPColumn

CondensateCollection

Process

O2 and CO2 Rich

Major Utilities

Cooling Water

Offgas[To Atmosphere]

Process Condensate

OptionalAPCI PRISM® Membrane

O2 and CO2 Rich[To Boiler]

Cooling WaterElectric Power

Expander

17

[ p ]

NOx SO2 Reactions in the CO22 2Compression System

We realised that SO2, NOx and Hg can be removed in the CO2We realised that SO2, NOx and Hg can be removed in the CO2compression process, in the presence of water and oxygen. SO2 is converted to Sulphuric Acid, NO2 converted to Nitric Acid:Acid:

– NO + ½ O2 = NO2 (1) Slow– 2 NO2 = N2O4 (2) Fast– 2 NO2 + H2O = HNO2 + HNO3 (3) Slow2 NO2 H2O HNO2 HNO3 (3) Slow– 3 HNO2 = HNO3 + 2 NO + H2O (4) Fast– NO2 + SO2 = NO + SO3 (5) Fast– SO3 + H2O = H2SO4 (6) FastSO3 + H2O H2SO4 (6) Fast

Rate increases with Pressure to the 3rd power– only feasible at elevated pressure

No Nitric Acid is formed until all the SO2 is convertedPressure, reactor design and residence times, are important.

18

CO2 Compression and Purification System –2Removal of SO2, NOx and Hg

SO2 removal: 100% NOx removal: 90-99%1.02 bar 30°C67% CO28% H2O

30 bar to DriersSaturated 30°C76% CO28% H2O

25% InertsSOx

224% Inerts

BFW15 bar

Water

NOx BFW

Condensate

30 barcw

cw

cw

Dilute H2SO4

Dilute HNO3

19

2 4 HNO3

Hg

The effect of Pressure on SO2 and NO Conversion (1 l/ i 7 d 14 b )

D

NO Conversion (1 sl/min, 7 and 14 barg)Presented at the 9th International Conference on Greenhouse Gas Control Technologies (GHGT-9) “Purification of Oxyfuel-Derived

CO2”, Vince White, Laura Torrente-Murciano, David Sturgeon, and David Chadwick, Washington, D.C., November 2009

CMFC

D

From NRTF

CondensateSeparator

A B

Flue Gas CoolerReactor

Compressor &Receiver

7 bar g14 bar g

Inlet

After Compressor &

ReceiverInlet

After Compressor &

Receiver ConversionConversion

gg

(Point A) (Point C) (Point A) (Point C)

ppm SO2 900 20 98% 950 150 84%

20

ppm NOx 520 50 90% 390 120 68%

K L i f I i l C ll Key Learnings from Imperial College Work• NO oxidation rate confirmed to increase rapidly with increased pressure• At low temperature (~ 30‐50 °C), SO2 oxidation to SO3 / H2SO4 only occurs at an appreciable rate in presence of NO2 and a condensed phase

• Significant conversion can occur even with very small liquid phase volumes (i.e., very low L/G ratios)(i.e., very low L/G ratios)

– Speculation that condensation creates a fine mist with large interfacial surface area, and that reaction is occurring at the interfaceSpeculation that reactions are accelerated at low pH– Speculation that reactions are accelerated at low pH

• Recognized need to develop internal capability to explore broader range of parameters and to increase rate of data generation

21

DOE/NETL C ti A t DOE/NETL Cooperative Agreement: Air Products’ Sour Compression PDU

Side View of PDU

Acid Reactor (C102) • 1st campaign Jan 2010

• 2nd campaign April‐May 2010

22

2 campaign April May 2010

U.S. Department of Energy's National Energy Technology Laboratory under Award Number DE-NT0005309

DOE Project: Air Products’ Sour Compression PDU DOE Project: Air Products Sour Compression PDU –Key Results

700

Oxyfuel 22 January 2010

• For the overall process, total SO2 removal was 20‐

500

600

Condition DSOxConversion >99%NOx conversion >90%(NOx zero +/‐ 20 ppmv)

100 % (based on gas compositions).

• For the overall process 200

300

400

ppmv

[FLUGAS]NO.SCALED

[FLUGAS]SO2.SCALED

Reactor OutletReactor OutletReactor Outlet

For the overall process, total NOx removal was 60‐90 % (based on gas compositions)

0

100

20:38:24 20:52:48 21:07:12 21:21:36 21:36:00 21:50:24 22:04:48 22:19:12 22:33:36

Reactor Inlet Reactor Inlet Reactor Inlet Reactor Inlet

compositions).

• The effects of variations in the SO2/NOx feed ratio, column pressure, gas flowrate and liquid recirculation on the reactor performance were

‐100Time

flowrate and liquid recirculation on the reactor performance were explored. Process performance was most sensitive to SO2/NOx feed ratio, over the range of parameter values investigated.

23

• SO2 was removed from the flue gas through both sulfite and sulfate mechanisms.

Sour Compression Modelling• Concurrent with the technology demonstration projects lab experiments• Concurrent with the technology demonstration projects, lab experiments and ASPEN modelling were carried out to develop a robust reaction model

• Goals of modelling:l d l d b d– Single reaction model to describe reactions in compression train and

in column– Use literature reactions and kinetics wherever possible

S d l b d d ib l b d d d ib il– Same model can be used to describe lab data and to describe pilot data from multiple configurations / scales (Renfrew, Windsor, Schwarze Pumpe)

• Modelling timeline:– 2005: 1st reaction model developed from literature references (no

data available at this time))– October 2009: research program with Imperial College initiated – September 2010: startup of Air Products lab reactor system– July 2011: Final version of reaction model developed from lab data

24

S C i M d l d t Sour Compression Model compared to experimental data

Compression Train

70%

80%

90%

100%

ts

Compression Train Configuration

30%

40%

50%

60%

Lab un

it result

SO2 removal

NO conversion

N removal

N2O yield

0%

10%

20%

0% 20% 40% 60% 80% 100%

N2O yield

45° line

70%

80%

90%

100%

Model prediction

40%

50%

60%

70%Lab un

it results

0%

10%

20%

30%

0% 20% 40% 60% 80% 100%

Column Configuration

25

0% 20% 40% 60% 80% 100%

Model Prediction

0/ 0 l G i f d ft O G50/50 Flue Gas Mix From Before and After OxPP FGD

26

0/ 0 l G i f d ft O G50/50 Flue Gas Mix From Before and After OxPP FGD

400

Inlet K432 Inlet Column Exit Column

350

400

NO

NOx

250

300

m, D

ry)

200

NO2(ppm

100

150

NO and

NO2

50

2

27

0

942 947 952 957 962 967

Time (minutes)

C t l f d t f ti d i Control of condensate formation during compression

New

Used• Need to control where condensate forms

• Temperature – Pressure –

Wh t h h th i t

Newpcomposition profile needs to be understood to prevent condensation, as acids will form

What happens when the incorrect valve material is selected

• First set of diaphragm compressor valves incorrectly specified with incompatible

And what

Usedmaterial

• Mode of operation caused condensation, hence acid And what

happens when the correct

l t i l

formation• Not all stainless steels are equal!

28

valve material is selected:

LICONOX™ Process

Bey/L/092009/Cottbus.pptLinde AG Engineering Division 30

S l h i A id M th dSulphuric Acid Method Activated Carbon Method

Auto-Refrigerated Inerts RemovalR l f i iti i i i i d• Removal of impurities minimises compression and transportation costs.

• O2 can be removed for EOR‐grade CO22 2

• CO2 capture rate of 90% with CO2 purity >95%• CO2 capture rate depends on raw CO2 purity which depends on air ingressdepends on air ingress

– Increases from zero at 25mol% to 90% at 75mol%

– Reducing air ingress increases CO2 capture rate

35

Can we improve on ~90% CO Capture?Can we improve on ~90% CO2 Capture?Vent stream is at pressure and is CO2 (and O2) rich

Flue Gas Vent 1 1 bar

Flue Gas Expander

Aluminium plate/fin exchanger

1.1 bar~25% CO2~25% O2~50% N2

Flue GasHeater -55°C

Driers

CO product

30 bar Raw CO2

CO2 product~96% CO2~4% Inerts‐60°C dp

36

30 bar Raw CO2Saturated 30°C75‐85% CO2

25%mol CO275% i t ( 15% O )

25%mol CO2

75% inerts75% inerts (~ 15% O2) (~ 19% O2)

72%mol CO2 98%mol CO2

76%mol CO224% inerts (~ 5 - 6% O2)

96%mol CO24% inerts (~ 0.95% O2)

72%mol CO2

28% inerts(~ 5 - 6% O2)

98%mol CO2

2% inerts(~ 0.6% O2)

25% l CO2 25%mol CO2

75% inerts (~ 15% O2)

25%mol CO2

75% inerts(~ 15% O2)

72%mol CO2 99.999%mol CO2

72%mol CO2

28% inerts(~ 5 - 6% O2)

99.95%mol CO2

0.05% inerts(~ .01% O2)

37

72%mol CO228% inerts(~ 5 - 6% O2)

99.999%mol CO2

0.001% inerts(~ .0005% O2)

( )

Autorefrigeration Process

J-T expansion of purified LCO2 for

Ventpurified LCO2 for refrigerationRaw CO2 partially

> 99.9% CO2

DryerDryerliquefied by boiling product

Compared to NH

yy

Hg

Compared to NH3refrigeration:Simpler process

PHX

Lower CAPEXHigher CO2recoveryLower power

Cold Box

DistillationColumn Reboiler

3838US Pat. 7,666,251

Air Products Oxy Fuel CO Capture and Purification with Air Air Products Oxy-Fuel CO2 Capture and Purification – with Air Products PRISM® Membrane

Flue Gas Expander

Aluminium plate/fin exchanger

Membrane

To Boiler

Flue GasHeater ‐55°C

Membrane

Driers

CO productCO2 product~96% CO2~4% Inerts‐60°C dp30 bar Raw CO2

39

30 bar Raw CO2Saturated 30°C75‐85% CO2

Use of Membrane to recover CO2 and O2 at the ventO2 at the vent

Vent:7% CO293% inerts (~10% O2)

ProductProduct96% CO24% inerts (~0.75% O2)

40

VPSA (Vacuum Pressure Swing Adsorption) for Recovering CO2 from Cold Box VentRecovering CO2 from Cold Box Vent

CO2 –leanvent

90%

100%Cold box + VPSAVPSAVPSA

CO2 –richrecycle

vent

80%

90%

very

, %

Cold boxCold box

Vent

60%

70%

O2 R

ecov

Feed CO2

40%

50%

CO

40%0% 3% 6% 9% 12%

Air Ingress, %

4444

Cold Box

Vattenfall ‐ Janschwalde (PC ‐250MWe) ‐ ???KEPCO/KOSEP ‐ Yongdong (PC ‐ 100MWe)

FutureGen2 ‐ Illinois (PC ‐ 100MWe)

Alstom Schwarze Pumpe 2008 30MWth LigniteHitachi Babcock Schwarze Pumpe 2010 30MWth LigniteIHI Callide 2011 30MWeAlstom / AL Lacq 2009 30MWth Gas/Oil?

l l h l

Coal

Target :

( )Endesa/CIUDEN ‐ El Bierzo (CFB ‐ 300MWe) ‐ ???

3 Newly announced oxyfuel projects in China.Drax Power Plant Oxyfuel in UK

CIUDEN El Bierzo CFB Facility 2011 30MWth CoalEl Bierzo PC Facility 2011 20MWth CoalCIUDEN

2009 – Lacq – World’s first 30MWt

t fitt d O NG

2011 – Callide –World’s first 30MWe retrofitted Oxy‐coal

power plant

By 2014‐2018

Demonstration of 50– 300MWe full scale power plant

“Commercialised by 2020”

2008

World’s FIRST 30 MWtfull chain demonstration

retrofitted Oxy‐NG boiler

2011 – CIUDEN –World’s first 30MWt Oxy‐CFB Pilot Plant

scale power plant.

2003 ‐ 2005

2007

B&W CEDF (30MWt) large scale burner testing started

at Schwarze Pumpe Pilot Plant

By the end of 2010/2011, Users (i.e. Power Plant Operators)

1990 ‐ 1995

1998 – 2001

CANMETUS DOE Project / B&W / Air Liquide

Vattenfall (ENCAP ++)CS Energy / IHI Callide Project

Users (i.e. Power Plant Operators) will have 6 burner manufacturers fully demonstrating “Utility Size Large Scale Burners” which should give a high level of confidence

1980’s

ANL/Battelle/EERC completed the first industrial scale pilot plant

EC Joule Thermie Project ‐ IFRF / Doosan Babcock / Int’l Combustion

NEDO / IHI / Jcoal Project

First large scale 35MWt B&W CEDF 2008 30MWth Coal

g e a g e e o co de cetoward demonstration

industrial scale pilot plant Oxy‐Coal Burner Retrofit Test done by

International Combustion

B&W CEDF 2008 30MWth CoalAlstom Alstom CE 2010 15MWth CoalDoosan Babcock DBEL ‐MBTF 2009 40MWth Coal

Updated by S. Santos (01/12/10)

Thank youWebsite: http://www.ieaghg.org

Special Issue:

Oxyfuel Combustion – Working Toward Demonstration and

Commercialisation

September 2011

59

Development in Oxyfuel Combustion Technologies...

Documents

Transcript of Development in Oxyfuel Combustion Technologies...