Progress in Energy and Combustion Science

Progress in Energy and Combustion Science 36 (2010) 581e625
Contents lists available at ScienceDirect
Progress in Energy and Combustion Science

journal homepage: www.elsevier.com/locate/pecs
Oxy-fuel combustion of solid fuels

Maja B. Toftegaard a, b, Jacob Brix a, Peter A. Jensen a, Peter Glarborg a, Anker D. Jensen a, *
a
Department of Chemical and Biochemical Engineering, Technical University of Denmark, DK-2800 Kgs. Lyngby, Denmark
b
DONG Energy, Kraftvaerksvej 53, DK-7000 Fredericia, Denmark
a r t i c l e i n f o a b s t r a c t
Article history: Oxy-fuel combustion is suggested as one of the possible, promising technologies for capturing CO2 from
Received 15 May 2009 power plants. The concept of oxy-fuel combustion is removal of nitrogen from the oxidizer to carry out
Accepted 10 February 2010 the combustion process in oxygen and, in most concepts, recycled flue gas to lower the flame
Available online 31 March 2010
temperature. The flue gas produced thus consists primarily of carbon dioxide and water. Much research
on the different aspects of an oxy-fuel power plant has been performed during the last decade. Focus has
Keywords:
mainly been on retrofits of existing pulverized-coal-fired power plant units. Green-field plants which
Carbon capture and storage
provide additional options for improvement of process economics are however likewise investigated. Of
Oxy-fuel combustion
Coal
particular interest is the change of the combustion process induced by the exchange of carbon dioxide
Biomass and water vapor for nitrogen as diluent. This paper reviews the published knowledge on the oxy-fuel
Emissions process and focuses particularly on the combustion fundamentals, i.e. flame temperatures and heat
transfer, ignition and burnout, emissions, and fly ash characteristics. Knowledge is currently available
regarding both an entire oxy-fuel power plant and the combustion fundamentals. However, several
questions remain unanswered and more research and pilot plant testing of heat transfer profiles,
emission levels, the optimum oxygen excess and inlet oxygen concentration levels, high and low-
temperature fire-side corrosion, ash quality, plant operability, and models to predict NOx and SO3
formation is required.
Ó 2010 Elsevier Ltd. All rights reserved.
Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 582
1.1. Carbon capture and storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 582
1.2. Carbon storage technologies overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 583
1.3. Carbon capture technologies overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 583
2. The oxy-fuel combustion technology and retrofit implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 584
2.1. Process overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 584
2.2. CO2 purity requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 585
2.3. CO2 processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587
2.3.1. Compression step . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587
2.3.2. Removal of water and non-condensable gas species . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587
2.4. Air separation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589
2.5. Flue gas recirculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589
2.5.1. Positioning of recycle streams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589
2.5.2. Oxygen addition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591
2.6. Operation of conventional flue gas cleaning equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591
2.6.1. Desulphurization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591
2.6.2. NOx removal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 592
2.6.3. Particulate removal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 592
2.6.4. Potential improvements for a green-field plant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 592
* Corresponding author. Tel.: þ45 4525 2841; fax: þ45 4588 2258.
E-mail address: aj@kt.dtu.dk (A.D. Jensen).
0360-1285/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.pecs.2010.02.001
582 M.B. Toftegaard et al. / Progress in Energy and Combustion Science 36 (2010) 581e625
2.7. Boiler and steam cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593

2.7.1. Burner operation and flame stabilization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593
2.7.2. Heat uptake . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593
2.7.3. Oxygen excess . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 594
2.7.4. Fire-side corrosion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 594
2.7.5. Aspects regarding green-field plants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 594
2.8. Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 594
3. Oxy-fuel combustion fundamentals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595
3.1. Research groups and experimental facilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595
3.2. Heat and mass transfer effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595
3.2.1. Flame and gas phase temperatures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595
3.2.2. Radiative and convective heat transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 598
3.3. The combustion process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 600
3.3.1. Devolatilization and ignition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 600
3.3.2. Volatile and char burnout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 601
3.4. Gaseous pollutants emissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 604
3.4.1. NOx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 604
3.4.2. SOx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 612
3.4.3. Trace elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615
3.5. Ash and deposition chemistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615
3.5.1. Particle formation mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615
3.5.2. The effects of gas composition on particle formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615
3.5.3. Ash quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 617
3.5.4. Depositions, slagging, and fouling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 618
3.6. Oxy-fuel combustion of biomass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 618
3.6.1. The combustion process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 618
3.6.2. Emissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 619
3.6.3. Ash and corrosion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 619
3.7. Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 620
4. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .621
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 622
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 622
1. Introduction world will depend on fossil fuels for many years to come. A rapid
move away from fossil fuels could result in great conflicts concerning
The world, and especially the developing countries such as water and land use between biomass for energy production, food
China and India, is facing an increasing growth in the demand for production, and forestation [7] as well as in serious disruption to the
electrical power [1,2]. New power plants are thus being constructed global economy [8]. The latter is mainly caused by the long lifetime
at a considerable rate in order to keep up with this demand [1e3]. of the energy supply infrastructure. In the transitional period,
The majority of the recently constructed and planned power plants, technologies are sought which will enable the continuous usage of
on a world-wide basis, are coal-fired [1,2]. Coal is a cheaper and fossil fuels but at the same time eliminate the emission of CO2.
more abundant resource than other fossil fuels such as oil and
natural gas while at the same time being a very reliable fuel for 1.1. Carbon capture and storage
power production [4,5].
In the developed countries an increasing part of the energy Since power plants constitute large point sources of CO2
consumption is being produced from renewable sources of energy; emission the main focus is related to their operation. Currently,
wind, biomass, solar, hydro power, etc. [1]. The main purpose of the
shift from a fossil fuel based production to renewable energy is to 400
CH
1800
4
decrease the emission of greenhouse gases. Especially the emission CO
2
of CO2 from the combustion of fossil fuels has gained great focus in N2O
recent years in connection with the discussions of global warming.
CO2 (ppm), N2O (ppb)
350 1400
Since the beginning of the industrialization in the late part of the
CH (ppb)
18th century the amount of CO2 in the atmosphere has increased

sharply from about 280 to 380 ppm [6], see Fig. 1.
4
Table 1 lists the current and projected CO2 emissions, in Gton

300 1000
carbon per year, from power generation (both electricity and heat)
[1]. Both the emissions and the coal share of the emissions are seen
to increase toward 2030 for the world as a whole. Even though the
CO2 emissions are seen to increase within Europe the percentage
250 600
increase is much less pronounced than for the rest of the world and 0 500 1000 1500 2000
the coal share of the emissions is expected to decrease. Despite the Year
fact that the ultimate goal for most countries is to phase out all fossil
Fig. 1. Development in the concentrations of important long-lived greenhouse gases in
fuels in heat and power production as well as in the transport sector, the atmosphere over the last 2000 years. The increases in concentrations since about
the share of renewable energy sources increases only slowly and the year 1750 are attributed to human activities in the industrial era. Data taken from [6].
M.B. Toftegaard et al. / Progress in Energy and Combustion Science 36 (2010) 581e625 583
Table 1 combination with enhanced oil recovery (EOR) differs between

Estimated CO2 emissions from power generation (Gton C/year). The numbers in authors and ranges from only 10s of years [20,19] to permanent
parenthesis indicate the percentage coal share of the emissions. Data taken from [1].
disposal [7].
Region 2005 2015 2030 Because of the limited retention times and the great risks of
World 3.0 (72) 4.0 (74) 5.1 (74) explosive release of CO2 back into the atmosphere and/or an
European Union 0.38 (70) 0.39 (66) 0.42 (61) alteration of the ocean chemistry in the near vicinity of the disposal
sites [7] ocean disposal is regarded a less attractive storage solution.
several possible technologies are being investigated which will A comparison of the estimated CO2 emissions from power
enable the so called Carbon Capture and Storage (CCS) from power production, Table 1, and the estimated storage capacities in EOR
plants [5,8e14]. Both researchers in universities and other research and saline aquifers, Table 2, yields between 75 and 6000 years of
institutions, most manufacturers of boilers and other power plant storage on a world-wide basis (2.5 Gton C/year stored). This
related equipment, and many power companies are active. CCS will calculation is based on the fact that due to small size and remote
act as a complimentary technology to the ongoing work related to location of many utility plants only a limited fraction of these
increasing fuel efficiency and the change toward fuels with lower emissions can be captured and stored cost-effectively. Baes et al.
fossil carbon content, e.g. natural gas and/or biomass. As indicated [18] estimate this fraction to be around 50%. CCS is generally not
by the term CCS, the elimination of CO2 emissions include two anticipated as a permanent solution to the elimination of anthro-
consecutive operations: pogenic CO2 emissions from electricity and heat generation. The
lower limit of 75 years of storage capacity should thus be sufficient
1. Capture of CO2 from the power plant flue gas in order for the industry to change almost entirely toward renew-
2. Storage of the CO2 (incl. transport to storage site) able sources of energy.
The estimated cost of separation, capture, and compression of 1.3. Carbon capture technologies overview
CO2 (point 1) from power plants or other point sources accounts for
around 75% of the total cost of a geologic sequestration process The identified technologies for carbon capture can be divided
[11,15e17]. into four main categories [5,11,12,23,27e30], described briefly
below. Fig. 2 shows the main operations concerned with the post-,
1.2. Carbon storage technologies overview pre-, and oxy-fuel combustion technologies.
Post-combustion capture: CO2 is separated from the flue gas of
The disposal technology should ensure a complete elimination conventional pulverized-coal-fired power plants. The separation is
of the CO2 from the earth's carbon cycle in order to stabilize the CO2 typically performed via chemical absorption with monoethanol-
concentration in the atmosphere. Two types of disposal are amine (MEA) or a sterically hindered amine (KS-1) [23,31e35].
defined: sequestration (permanent disposal) or storage (disposal Amine absorption is a proven technology in the process industry
for a significant time period) [7]. These terms are often inter- [23,34,36,37]. The demonstrated scale of operation is, however,
changed in the sense that time periods of more than the order of significantly smaller than the typical size of power plants [34] and
10,000 years are considered permanent. Possible storage methods serious penalties to the plant efficiency exist at the current state of
suggested include injection in e.g. depleted oil and gas reservoirs, development [5,8,12,16,34,38e43]. The anticipated drop in the net
coal beds, deep saline aquifers, etc. [7,11,17e24]. The estimated efficiency of the power plant is about 10e14% points [41,34]. Some
storage potential for the suggested options is given in Table 2. current research projects investigate the possibility of developing
When CO2 is injected below the caprock in oil and gas reservoirs
as well as deep saline aquifers it is first trapped by static and
hydrodynamic mechanisms. Secondary trapping mechanisms
begin operating over time and act to immobilize the CO2 in the
reservoir, thereby significantly limiting the risk of leakage
[7,15,24e26]. This type of storage is considered secure even in the
initial injection phase where the secondary trapping mechanisms
contribute only minimally [7].
The large storage potential in deep aquifers without structural
traps is only obtainable if the traps are not required for secure
storage during the initial phases [19]. Even without this storage
volume the remaining sites offer storage capability for potentially
the next many hundred of years [19,22], see more below. According
to Table 2 the estimated retention time in the underground storage
sites is 105e106 years. The retention time for storage in
Table 2
Estimated storage capacities and retention times for CO2 in different types of sinks
on a world-wide basis. Data taken from [7,17e22].
Sink Storage capacity Retention

(Gton C) time (years)
Enhanced Oil Recovery (EOR) 20e65 10e106
Deep aquifers with structural traps 30e650 105e106
Deep aquifers without structural traps w14,000 105e106
Depleted oil and gas wells 130e500 105e106
Coalbeds 40e260 105e106
Fig. 2. Possible, overall plant configurations for the three main categories of carbon
Ocean disposal 400e1200 500e1000
capture technologies. Adopted from [10].
more efficient absorbents [35]. More on the technology can be seen Emerging technologies: Technologies such as membrane sepa-
in [5,8,13,29,44e47]. ration, chemical looping combustion, carbonationecalcination
The chilled ammonia process in which an aqueous solution of cycles, enzyme-based systems, ionic liquids, mineralization, etc.
ammonia constitutes the absorbent has shown promising reduc- impose the possibility to drastically reduce the cost of electricity
tions in energy consumption in laboratory studies, up to 50%, and the energy penalty concerned with carbon capture from power
compared to the MEA process [42]. The process benefits from low plants. The papers by Eide et al. [70], Abu-Khader [28], Hossain and
operating temperatures and precipitation of ammonium bicar- de Lasa [71], and Figueroa et al. [14] provide broad overviews of
bonate (NH4HCO3) yielding a higher CO2 loading of the absorbent. these technologies and their current state of development.
Retrofit to existing plants for both process types is considered The choice of technology will depend on several factors. First and
relatively simple since the capture unit can be added downstream foremost the economy and the expected development in plant
of the boiler and flue gas cleaning systems without any significant efficiency is of importance. The maturity, expected availability,
changes to the original plant [8,45]. There are, however, strict operating flexibility, retrofit or green-plant built, local circum-
requirements for removal of SO2 and NO2 from the flue gas prior to stances, utilities preferences, etc. will likewise have to be taken into
the CO2 capture since these components react irreversibly with the account. No general acceptance of superiority of one of the
absorbent leading to its degradation. presented technologies over the others exists. Several techno-
Pre-combustion capture: Also termed fuel decarbonisation. The economic studies also indicate that with the current knowledge on
process is typically suggested to be used in connection with Inte- the technologies no significant difference in cost within the limits of
grated Gasification Combined Cycle (IGCC) power plants where it is precision of the applied cost estimates can be determined between
termed IGCCeCCS. Coal gasification is applied to obtain a gas amine absorption capture, coal-based IGCC type capture, and
(syngas) containing CO, CO2, and H2. The CO is transformed into CO2 oxy-fuel combustion capture [5,8,22,57,66,67,72].
by the water-gas shift reaction and can then be separated from the Because of the large changes induced in the power plant by the
remaining hydrogen containing gas before this is combusted in a gas implementation of oxy-fuel combustion, more research is needed to
turbine. Alternatively, H2 can be separated from the syngas and fully clarify the impacts of the introduction of this technology. Many
the CO combusted in an O2/CO2 atmosphere [48]. Some laboratory scale investigations of the technology have been per-
techno-economic calculations [11,30,36,49,50] show that IGCC has formed within the last two decades and it is generally accepted that it
promising process economics and plant efficiency characteristics. is possible to burn coal and natural gas in an O2/CO2 atmosphere. On
However, high capital costs are associated with plant construction the other hand, it is likewise recognized that much work still remains
and IGCC plants are generally much more complicated systems than in obtaining sufficient insight into the effects on e.g. emissions,
suspension-fired boilers [51,37]. Only few electricity producing IGCC residual products such as fly ash, flue gas cleaning, heat transfer, etc.
units exist [29,50,52e54], none of which are equipped with CCS. As In 2005, Wall and coworkers [4] published a literature review on
a consequence of the few plants and limited operating experience the oxy-fuel combustion technology. The work was updated in the
along with the highly integrated nature of the plants compared to broader CCS review by Wall [5] in 2007. The reviews focused mainly
the more matured, conventional pulverized-coal-fired power on combustion fundamentals, overviews of research groups and
plants, the demonstrated availability for IGCC is significantly less their experimental facilities, techno-economic assessments of the
(80e85% versus w96%, respectively) [5,30,37,50,52,55]. IGCCeCCS is technology, and research needs.
not a viable option for retrofit of existing pf plants [30,51,56,57]. The amount of literature on the oxy-fuel technology has
Oxy-fuel combustion: By eliminating molecular nitrogen from increased drastically over the latter years and significant new
the combustion medium the flue gas will consist mainly of CO2 and information is thus now available. The objective of the present
water. The plant configuration typically suggested involves flue gas review has been to summarize the current knowledge status on the
recirculation to the burners to control the flame temperature to oxy-fuel combustion technology. The current review has two
within the acceptable limits of the boiler materials. Implementa- focuses. (1) The possible advantages and challenges associated with
tion of the oxy-fuel combustion technology in existing pulverized- retrofitting of existing pulverized-coal-fired power plants to the
coal-fired power plants will induce a larger change of the plant oxy-fuel combustion technology as well as considerations
configuration when comparing to the post-combustion absorption regarding green-field plants. (2) The reported results from labora-
processes mentioned above. This is mainly due to the fact that the tory- and semi-technical scale experiments regarding the
combustion chemistry is altered by substituting recycled flue gas combustion process fundamentals, including the flue gas compo-
(mainly CO2 and water) for nitrogen in the oxidizer. Several of the sition and residual products.
earlier techno-economic assessment studies indicate that oxy-fuel
combustion should be the most energy and cost efficient of the 2. The oxy-fuel combustion technology and retrofit
carbon capture technologies [9,16,38,58e63]. This conclusion is implications
mainly based on assumptions of greater boiler efficiency caused by
a smaller flue gas volume and the reduced need for flue gas 2.1. Process overview
cleaning, i.e. deNOx and desulphurization, including the derived
decrease in capital and operating costs. It is suggested that SOx and In open literature, oxy-fuel combustion with recirculation of flue
NOx can be stored along with CO2 in the geospheric sinks gas was proposed almost simultaneously by Horn and Steinberg
[8,12,64,65]. Typically, no experimental validation of these [58] and Abraham et al. [60] in the early eighties. Abraham et al.
assumptions has been performed. Whether co-storage of SOx and proposed the process as a possible mean to produce large amounts
NOx is politically acceptable is, however, questionable. of CO2 for Enhanced Oil Recovery (EOR) whereas Horn and Stein-
The main disadvantage of the oxy-fuel combustion technology berg had in mind the reduction of environmental impacts from the
is the need for almost pure oxygen. The available large-scale use of fossil fuels in energy generation. As such, the technology
technology for air separation is based on cryogenic distillation received renewed interest in the mid-90s in connection with the
which will impose a very large energy penalty on the plant [65]. re-emerging discussions of global warming caused by increased
The expected efficiency drop is about 7e11 percent points, or about CO2 levels in the atmosphere [12].
15e30% of the generated electricity (net power output), depending Oxy-combustion can in principle be applied to any type of fuel
on the initial plant efficiency [5,8,12,16,27,29,43,58,59,66e69]. utilized for thermal power production. The research interests have
Fig. 3. Possible configuration of an oxy-fuel power plant. ASU: Air Separation Unit, SCR: Selective Catalytic Reduction reactor (deNOx), ESP: Electrostatic Precipitator, FGD: Flue Gas
Desulphurization. Energy inputs and low-temperature-heat outputs new to the plant in case of a retrofit are indicated.
mainly been focused on coal and natural gas since these are the most concentration of 3% in the dry recycle stream. The recycle ratio for
abundant fuels. For these specific fuels the technology is typically the dry flue gas is 0.75. As a preliminary assumption, the condenser
termed oxy-coal and oxy-natural gas combustion, respectively. will remove 100% of the water in the flue gas. The CO2 capture
Application of CCS through oxy-combustion of biomass or blends of efficiency is w92%. The remainder of CO2 is vented to the atmo-
coal and biomass will result in a possible mean of extracting CO2 sphere together with the non-condensable gas species in the
from the atmosphere and thereby possibly inverting the presumed exhaust gas from the CO2 cleaning unit. Generally, between 90 and
anthropogenic caused changes to the climate [11,23,30]. 95% capture efficiency is expected for oxy-fuel plants [30].
As indicated in Section 1.3 a shift from conventional air-firing to The survey of the implications of retrofitting for oxy-fuel
oxy-fuel combustion in a power plant will induce multiple changes combustion in the next sections will be starting at the point of
to the plant configuration. Fig. 3 provides a sketch of a coal-fired delivery of the sequestration-ready CO2. The discussion will
oxy-fuel plant with indications of the major process steps and the proceed with the auxiliary units new to the plant and end up with
necessary energy inputs and low-temperature-heat outputs new to the original parts of the power plant. In order to limit the survey,
the plant when retrofitting an existing coal-fired unit. The sketch the discussions, when relevant, will refer to plants utilizing bitu-
covers the original state-of-the-art plant with boiler, coal mills, and minous and/or sub-bituminous coals. Specific aspects regarding
flue gas cleaning equipment. The final processing of the CO2 stream, lignite fired plants are excluded.
i.e. the removal of water and the non-condensible gases like O2, N2,
Ar, etc. to meet the requirements regarding purity of the CO2
2.2. CO2 purity requirements
stream, as well as the air separation unit (ASU) and the compres-
sion step for the CO2 stream before it is transported to the storage
The exact requirements to the quality of the CO2 stream for
site are new to the plant. The discussion in the coming sections is
different storage scenarios are not yet fully clarified [30,73e75].
based on the assumption that an underground geological forma-
tion, a saline aquifer, is used as storage site. The type of storage will
have an impact on the oxy-fuel process, especially the flue gas Table 4
purification units, through the quality criteria for the CO2 stream. Properties for El Cerrejon bituminous coal (Colombian).
Table 3 provides the results of our calculation on the approxi- LHV, as received (MJ/kg) 27.09
mate daily flows in and out of a 500 MWe oxy-fuel combustion Moisture, as received (wt%) 5.0
power plant. The calculations are based on a state-of-the-art Proximate analysis (wt%, dry)
electrical efficiency of 46% of the conventional plant (supercritical Ash 10.1
steam parameters, LHV basis) [49] with a 10 percent point decrease Volatile 36.7
due to oxy-fuel operation. The used reference coal is a high-volatile Fixed carbon (by difference) 53.2
bituminous coal from Colombia see Table 4 for its properties. The Ultimate analysis (wt%, daf)
oxygen is assumed to have a purity of 100%, in order to simplify the C 80.70
calculations. The combustion is performed in a mixture of 30% O2 in H 5.41
N 1.69
CO2 with an oxygen excess of 10%, corresponding to an oxygen
S 0.73
O (by difference) 11.47
Table 3 Ash composition (wt%, dry)

Approximate mass streams in a 500 MWe oxy-fuel combustion power plant with an Al 10.89
electric efficiency of 36% on a net heating value (LHV) basis. Ca 1.58
Fe 5.05
Stream Mass Flow (ton/day) K 1.78
Oxygen in 9700 Mg 1.39
Coal in 4400 Na 0.59
Waste water (from condenser) 2000 P 0.08
Exhaust gas 1100 Si 26.73
CO2 for sequestration 10,300 Ti 0.53
N2 from ASU 31,800 O (by difference) 51.37
Table 5
Suggested CO2 quality specifications from different sources.
Parameter Modest quality, aquifer storage High quality, on-shore storage U.S. Specifications Saline formation
Anheden et al. [76] Lee and Miller [77] Fout [78]

Pressure 110 bar 110 bar e 150 bar
Temperature 50 C 50 C <50 C e
CO2 >96 vol% >96 vol% >95% Not limitedb
H2O <500 ppm <50 ppm <480 ppmv 150 ppmv
N2, Ar <4 vol%a e <4% Not limitedb
O2 <4 vol%a <100 ppm <10 ppm <100 ppmv
SO2 <200 mg/Nm3 <50 mg/Nm3 e <3 vol%
H2S e e <10e200 ppm <1.3 vol%
NOx e e e Uncertain
NH3 e e e Not limited
CO e e e Not limited
CH4 e e e <0.8 vol%
HC's e e <5% <5 vol%
H2 e e e Uncertain
Glycol e e <0.04 ppmv e
a
Sum of N2, O2, and Ar should be <4 vol%.
b
No limit but the impacts on compression power and equipment cost need to be considered.
However, Table 5 provides a number of suggestions for purity SO2 receives the largest amount of interest with respect to the
requirements found in the literature. effect of contaminants on the structure of storage formations. A
Some authors lay down different criteria for different storage typical assumption regards the possibility of co-storing the SO2
sites [76,78], the differences mainly being associated with the together with the CO2 because of very similar physical and chemical
content of water, oxygen and SO2. Lee and Miller [77] comment on properties at supercritical conditions [10,58,65,67,75,79]. However,
the individual limits with respect to e.g. the minimum miscibility even small amounts of SO2 may cause problems due to the risk of
pressure (CO2, hydrocarbons, and N2), the risk of corrosion (O2, calcium sulphate formation and thus a decreasing porosity of the
water), as well as materials (temperature), operations (glycol), and reservoir rock [26,75,79,80]. Oxygen could likewise lead to the
safety (H2S). Anheden [76] likewise identifies the aspects which formation of precipitations [74]. On the other hand, if the concerns
should be taken into account when determining the individual regarding SO2 in the storage formations prove to be insignificant
limits. These aspects come down to operational issues, storage there would obviously be a possible economic benefit from
integrity, environmental aspects during the full lifetime of the combined capture and storage of CO2 and SO2 [80]. The current state
capture and storage chain, health and safety aspects, legal aspects, of flue gas cleaning on modern power plants involves w85%
and economic considerations. The authors state that the require- removal of NOx, w98% removal of SO2, and w99.8% removal of
ments arising from exposure limits in case of leakage to air put the particulates [68,30]. The fact that only a finite percentage of the SO2
strictest restrictions to the process and are, at the same time, the is removed with the current best available technology elucidates
easiest to quantify. The exact requirements will most probably be the importance of identifying the correct purity demands through
determined for each individual case of capture and storage. e.g. field tests since these may have a significant impact on the plant
Jordal and coworkers [10,74] considered the optimum specifi- configuration, operating conditions and operating costs [43,64].
cations with respect to technical and economical considerations. Besides the chemical effects of impurities in the CO2 on the reservoir
Not surprisingly, the optima differed. For economic reasons, the rock, CO2 itself has the potential to alter the mechanical properties
preferable option is to co-store as many of the impurities; SOx, NOx, of the rock [26,17]. Especially calcite (CaCO3) precipitation which
non-condensable gas species, and water, as possible. This will can cement the reservoir around the injection well and render
reduce the plant investment and operating costs of the process. The further injection impossible should be taken into consideration.
disadvantage could be the requirement for more expensive mate- With respect to transportation the greatest concern involves the
rials in e.g. compressors and pipelines to withstand the potentially water content in the CO2 stream [27,64,73,75]. In the presence of
severely corrosive environment. There is, however, an economic water, CO2 can cause so called sweet corrosion [26,17]. Water vapor
optimum for the non-condensables (N2, O2, NO, CO, H2, CH4, Ar, and CO2 in the presence of liquid water can likewise form solid
etc.) since co-storage of these species will increase energy and ice-like crystals known as hydrates [64,73,79,80]. Concurrent pres-
reservoir size requirements as well as capital and operating ence of both water and SO2 (incl. H2S) in the CO2 stream will increase
expenses in the transport chain with an amount proportional to the risk of sulfuric acid corrosion. If the flue gas is dehydrated to a dew
their concentrations [24,73,74]. At the same time the non- point 5 C below the temperature required for transport conditions,
condensible gas species will entail an energy and capital penalty the sulphur dioxide will behave almost as carbon dioxide in the
when removed from the CO2 stream [73,74]. Technically, there are supercritical state and the two gases should not cause any corrosion
two general issues which should be considered. First, the purity problems [26,79,80]. Others report no risk of corrosion at a dew point
requirements for transport and storage with respect to corrosion of less than 60 C [66]. A requirement for a very low water content in
and the risk of structural changes within the storage formations the CO2 is thus present. This can most likely not be achieved by
caused by impurities in the CO2 stream. Secondly, the limitations to condensation alone and drying by e.g. absorption in a recyclable
the present best available technology for flue gas cleaning, i.e. dehydrant (triethylene glycol) in combination with the last
particle removal, water condensation, dehydration, SOx removal compression step, see Section 2.3, will be necessary [43,65,74,79].
and removal of non-condensible gas species, and how to minimize Some researchers believe that all the limiting factors regarding
the loss of CO2 to the atmosphere during the purification process. It purity of the CO2 stream arise from compression and trans-
is obvious that the technical considerations will set both the lower portation requirements [67,73,74]. In this respect, the demand of
and upper limits to the purity requirements. dryness is crucial.
Another aspect regarding the concentrated CO2 stream is the and flue gas passages will thus be a trade off with the power
legislative classification. This will depend on the content of consumption of the compression unit. Air-fired boilers operate at
contaminants such as H2S, sulphur oxides, NOx, hydrocarbons, etc. negative gauge pressure for obvious safety reasons. To avoid air
Significant quantities of these elements could mean that the stream ingress in oxy-fuel boilers one could consider operation at positive
should be regarded as a hazardous waste [7,19,64,73,74,80]. This gauge pressure. It is, however, questionable whether the avoidance
could eventually lead to difficulties with respect to locating suitable of air ingress can counterbalance the enhanced risk of severe CO
storage sites or a large penalty concerned with the purification. and CO2 poisoning of the boiler surroundings and the necessary
safety systems to avoid casualties [83]. The amount of air leaking
into the boiler and flue gas ducts is believed to constitute around 3%
2.3. CO2 processing of the flue gas mass flow (1% in boiler and 2% in ESP) [5,40,67,94] for
a new-build plant. As the ingress of air typically increases over time,
Assuming that CO2 transportation should be in pipelines, the older boilers can have air in-leakage rates of 8e16% [65,67].
CO2 must be conditioned to both pipeline and reservoir specifica- Significant air ingress will reduce the CO2 concentration in the flue
tions, as described in the above section. Moreover, the stream must gas and thus result in increased costs for CO2 cleaning since the
be compressed to a high enough pressure to overcome the frictional non-condensable gases most likely must be removed before (or
and static pressure drops and deliver the CO2 at the storage site during) compression. According to Tan et al. [12] the air ingress
without risking flashing of gas anywhere in the process [73,75]. level should be limited to about 3% in large-scale plants to obtain
The typically suggested conditions of the CO2 stream at the CO2 concentrations in the flue gas which will enable an economical
delivery point includes the following ranges of pressures and treatment of the CO2 stream.
temperatures; 80e200 bar and 0e50 C [8,11,25,27,41,56,67,74, This additional need for clean-up, besides air separation, is
81e84]. There is, however, a preference for 100e110 bar and potentially a significant drawback of the oxy-fuel technology
a temperature above the critical value (31.1 C), see Fig. 4. At compared to e.g. post-combustion capture with amine absorption
these conditions the CO2 is in the supercritical state, even though which has only one purification step [95]. When the oxy-fuel
the presence of impurities will increase the critical pressure of technology was first proposed purification of the CO2 was not
the mixture compared to that of pure CO2 [67]. discussed and has most likely not been considered necessary.
The purity of the oxygen for combustion will likewise be
2.3.1. Compression step a trade off between the ease of liquefaction of the flue gas and the
Compression of gases is a highly energy demanding process. power requirement for the air separation [61,62,67]. Nakayama
This step will induce one of the larger penalties to the plant effi- et al. [61] performed an investigation regarding the combined
ciency in oxy-fuel operation when comparing to conventional air- power requirement for air separation and CO2 liquefaction under
firing. Generally, w2e3% point decreases in electrical efficiency, the assumption of no air ingress into the system. Fig. 5 shows the
directly associated with the compression step, have been reported relationship between purity of feed oxygen and power
[5,16,50,66,67]. Factors such as compressor efficiency, the amount consumption for a plant with a gross capacity of 1000 MWe.
of impurities in the CO2-rich stream, and opportunities for inte- According to the figure, an oxygen purity of 97.5% will require the
gration with the remaining plant will influence the power lowest overall power consumption. Purities greater than this leads
consumption [65,85]. to a significant increase in the air separation energy consumption
Since no specific new problems are expected with respect to the which the decreased consumption in the liquefaction step cannot
compression of CO2, no demonstration programs are considered to counterbalance. Others report optimum purities of 95%
be required [66]. [8,43,62,66,67,83,95] taking into account the effect of air ingress.
At an oxygen purity of 95% obtained by cryogenic distillation, the
2.3.2. Removal of water and non-condensable gas species
Many existing boilers were not designed to be leak tight and
300
may be difficult to retrofit to oxy-fuel operation where air ingress
should be avoided. It is generally reported very difficult to totally
avoid air ingress into even small laboratory and pilot scale burners
[12,16,63,86e91] and in real size plants it will be almost impossible
[62,83,92,93]. The amount of work put into tightening the boiler
200 CO Liquefaction
2
Power (MW)
( = 90 %)
100
Air Separation
0
90 92.5 95 97.5 100
Purity of Feed Oxygen (%)
Fig. 5. Theoretical relationship between the purity of oxygen fed to the burners and
the power consumption for air separation and CO2 liquefaction in a 1000 MWe (gross)
Fig. 4. Phase diagram for CO2. plant. Data taken from [61].
impurities such as N2, O2, and Ar [85]. The thermodynamic proper-

ties of the CO2 stream fed to the purification unit will be affected by
its content of impurities (N2, O2, Ar, NOx, and SO2). Possible ranges of
impurity concentrations are: Ar (0e5%), N2 (0e15%), O2 (0e7%), and
SO2 (0e1.5%). The changes are mainly affecting the dew and bubble
points, heat capacity as well as the enthalpy and entropy. This again
will have an impact on the operating conditions, energy consump-
tion, configuration, and separation performance of the purification
system [85]. Generally, removal of SO2 in a flash or distillation system
is not preferable since it has a large, negative impact on product
purity and energy requirements for even small impurity levels [85]
since the thermodynamic properties are very close to that of CO2.
The basic principle in the flash and distillation units is to utilize
the differences in boiling point for the different species to separate
them. Fig. 7 shows a comparison between the performances of the
two purification systems. The figure reveals that for the same
recovery rate and feed compositions the distillation process
requires less energy and produces a slightly higher-purity CO2
stream than the flash process. This has likewise been reported by
Aspelund and Jordal [73,74].
Fig. 6. Possible process scheme for CO2 processing and purification, i.e. removal of
non-condensable gas species, in an oxy-fuel power plant. Adapted from [85]. The typical CO2 recovery for oxy-fuel plants is reported at w90%
[12,30,40,43,67]. 100% capture was expected theoretically in one of
the earlier studies of the technology [59], under the assumption
remaining impurities present will be argon (3e4 mol%) and that the entire flue gas stream could be sequestered in the ocean.
nitrogen (1e2 mol%) [67,68]. The almost complete avoidance of A possibility exists that the initial flue gas cleaning and the CO2
nitrogen is beneficial with respect to also reducing NOx emissions. processing steps can be combined in one process unit. This is
Fig. 6 illustrates in more detail suggestions to the CO2 processing interesting in the case that desulphurization is neglected within the
and purification steps in an oxy-fuel plant. It is possible that an recycle loop. Air Products Inc. has proposed a process that removes
initial dehydration of the flue gas is performed in combination with both non-condensable impurities as well as Hg, essentially all SOx
the flue gas cleaning step before the initial compression in order to
reduce the total flue gas volume and thus costs and equipment size.
After condensation of water vapor, the CO2 stream will have
a purity of about 70e95% depending on oxygen purity, oxygen
excess, and air ingress into the boiler [4,8,16,40,43,67,91,92,96]. The
condensation process will likewise serve to remove water-soluble
gas phase components as well as other species with higher boiling
points than CO2 [74]. Some CO2 will dissolve in the water and cause
a loss in the capture efficiency if not regenerated [74]. If the major
part of the flue gas water content is removed at relatively high
temperatures and low pressures the loss of CO2 can be kept low [74].
The flue gas processing and purifications system will produce
three streams as indicated in Fig. 6; The CO2 stream for compres-
sion and sequestration, an exhaust gas containing mainly the non-
condensable gas species (Ar, N2, O2, CO, NOx) and some CO2
depending on technology, see more below, and a waste water
stream originating from the condensed water.
The exhaust gas stream, around 10e20% of the dry flue gas by
volume, is vented to the air. Its composition depends on the chosen
cleaning process but is assumed to be about half CO2 and half N2,
O2, and Ar. Aspelund and Jordal [73] state that the loss of CO2 to the
exhaust stream for a one column system will be 1:1, i.e. 1 mol of CO2
per mole of non-condensable gas.
Depending on the plant configuration, especially whether
a desulphurization plant is installed, the waste water stream could
be very acidic (sulfuric acid, hydrocloric acid, etc.) and thus
extremely corrosive (pH 1.5e3) [10,67,92,97]. The flue gas purifi-
cation unit should thus be designed acid-proof [82,10] as well as the
flue gas pathway [10] due to the risk of low-temperature acid
corrosion. Additionally, a fraction of the CO2 (due to its solubility in
water) and a large part of the particles remaining after the ESP will
be captured in the waste water stream. This stream therefore
requires further processing before disposal. This is expected to
occur with methods already commercially available [10]. Fig. 7. Performance comparison between 2-stage flash and distillation column based on
The gas purification unit could e.g. be a flash [40,10] or a distil- simulations. The CO2 recovery rate is 92.15% and the feed compositions (in mol%) to the
lation unit [27,74,92] since CO2 forms non-azeotropic mixtures with purification units are CO2: 76, O2: 6, N2: 15, Ar: 2.5, and SO2: 0.5. Adopted from [85].
and about 90% of the NOx through the production of waste water operate at 60e100% load [82,43] and show rather poor efficiencies
containing sulfuric and nitric acid [98]. below 80%. In combination, this gives rise to some operational
This compression and purification process has been initially difficulties, which should be investigated further.
tested in mid-2008 [98]. The tests have confirmed the potential of The introduction of a liquid oxygen storage could provide
the process. However, further tests are necessary to gain further a solution in which the difference between demand and production
understanding of the chemical mechanisms and the kinetics [98]. during positive load changes in the boiler is taken from the storage
This is expected to be performed at the Vattenfall Schwarze Pumpe and vice versa during decreases in the boiler load [43,92,99]. The
pilot plant [99]. The process should likewise be evaluated against storage facility also makes it possible to shift to air-firing mode
well proven deNOx and desulphurization technologies [67,43]. within a relatively short amount of time without risking boiler trips
in case of unexpected ASU shut-down. Even without the load
2.4. Air separation changing challenges, it may be beneficial to decouple the ASU and
power plant by introducing oxygen storage. In that way, the liquid
The air separation unit provides the oxygen for the combustion. oxygen could provide a means of energy storage in periods with
The current state of technology development necessitates the high electricity production from e.g. wind mills. The safety issues
application of a cryogenic distillation unit. No other mature tech- related to a liquid oxygen storage on a power plant site should be
nologies exist which can be applied for the necessary size of assessed thoroughly however in order to evaluate its applicability.
operation [40,59,65,79,80,82,93,100], see Fig. 8. At the same time, Preliminary studies by Xu et al. [43], suggest to install a tank with
because of mechanical size limitations it will be necessary to build a 500 tonne capacity for liquid oxygen at 2.5 bar (about 175 C),
more cryogenic air separation units to operate in parallel. The corresponding to approximately 1.5 h of operation of their
largest cryogenic ASU plant currently operating (2008) produces 413 MWe, net, oxy-fuel reference plant.
4500 t/d, but design studies for a 7000 t/d have been performed For economic reasons it should be considered whether the
[65]. As seen from Table 3 the oxygen requirement for a 500 MWe massive amount of almost pure nitrogen, w31,800 t/d at low
plant is between 9000 and 10,000 t/d, indicating the need for 2e3 temperature, could be applied for a purpose in the plant or
parallel ASU plants. elsewhere.
Cryogenic distillation consumes high amounts of energy in the
form of electricity. Preliminary calculations indicate that this 2.5. Flue gas recirculation
operation alone requires about 60% of the power consumption for
carbon capture and reduces the overall efficiency of the power The oxygen for combustion is mixed with recycled flue gas to
plant by about 7e9 percentage points [8,10,16,40,43,50,67]. Major moderate flame temperatures and obtain a boiler heat transfer
improvements are thus required to yield a more efficient oxygen profile similar to that of air-firing in the case of retrofit. Typically,
production. It is not likely to find these improvements within the between 60 and 80% of the flue gas is recycled to the boiler
cryogenic technology [62,65,67] and thus alternative technologies [4,5,8,10,43,66e68,79,80,101,102].
like e.g. ion transport membranes (ITMs) should be investigated As shown in Fig. 3 two recycle streams will be necessary. The
further. However, it is possible that some synergies can be obtained primary recycle is used for fuel transport and corresponds to the
from an integration of the ASU's with the power plant, i.e. by using primary air in conventional units. It will be about 20% of the total
the low-temperature heat from the ASU intercoolers for e.g. feed amount of combustion air (RFG þ O2) [67]. The secondary stream
water heating [93]. will constitute the equivalent to the secondary, tertiary, and over-
The availability of the oxygen plant is crucial to the power plant fire (if necessary) air flows when oxygen has been added.
operation. Typical availabilities of 98e99% and only limited time for A study performed by Kather et al. [68] determined that the
maintenance are reported from suppliers [65,67]. Another impor- temperature of the recirculated flue gas should be between 200 and
tant requirement is the ability to follow the load changes of the 350 C in order to take into account the operation of an ESP and flue
power plant. According to Xu et al. [43] the maximum ramp rate for gas fan, among others.
an ASU is 3%/min. The boiler, however, can generally be operated at
a ramp of up to 6%/min. At the same time, ASU's are only able to 2.5.1. Positioning of recycle streams
2.5.1.1. Primary recycle. A general consensus exists that the
primary recycle stream must be cooled, scrubbed, dried and then
reheated to about 250 C or more before entering the mills
[66e68,83,101]. The obvious choice is thus as indicated in Fig. 3 to
take this stream after the first condensation step during the flue gas
processing.
The reheating is necessary in order for the stream to be able to
dry the coal and carry the moisture at the typical mill exit
temperature of 60e90 C [67,68]. If the drying capacity of the
primary recycle gas is too small the wet coals will clog the mills.
Scrubbing of the flue gas ensures removal of SO2 which would
induce corrosion in the mills due to condensation of sulfuric acid at
temperatures below 150e160 C [103]. Xu et al. [43] suggest adding
a bypass from upstream the ESP outlet to increase the temperature
of the recycle streams approximately 5 K in order to avoid moisture
condensation along the recycling ductwork.
Fig. 8. Oxygen supply technologies. Selection as a function of quantity (tonnes per 2.5.1.2. Secondary recycle. Several options exist for the position
day) and required purity. Cylinders and Liquid: Continuous delivery from manufac-
turers in either cylinders or for own storage vessel. VSA: Vacuum Swing Adsorption.
where the secondary recycle stream is taken. The dashed lines in
Future plants could be either cryogenic or based on other technologies, such as e.g. Fig. 3 indicate these possibilities. Depending on the actual position
membranes. Adapted from [65]. the heat and water contents along with the concentration of
pollutants (NOx, particles, SOx, etc.) will vary. Even though several the configuration of an oxy-fuel power plant. In conclusion, the
reports on oxy-fuel combustion conclude that both deNOx and secondary recycle stream will most likely have to be taken down-
desulphurization can be eliminated due to co-sequestration of the stream of some sort of desulphurization process whether it be
impurities with the CO2 [60,61,67] this is not necessarily possible in a conventional wet FGD plant or another process.
practice, see Section 2.2. Most probably, the power plant should be The water content in the secondary recycle could potentially
able to run in air-firing mode without capture, at least during start- have an impact on the combustion characteristics and the general
up and shut-down, and thus also needs the flue gas cleaning plant performance and efficiency [102]. This is mainly a consequence
equipment to obey the legislation regarding emissions. There is, of the differences in thermodynamic properties, e.g. heat capacity,
however, the possibility of by-passing the SCR and FGD plants between CO2 and water vapor. The effect of wet and dry recycle
during oxy-fuel operation. It will still be necessary to remove the streams on in-furnace temperatures, ignition, emissions, etc. are
particulates to avoid accumulation of solids in the boiler and to discussed in more detail in Section 3. This section will focus on the
prevent the flue gas recirculation fan and gas passages from plant configuration aspects of a wet versus a dry flue gas recycle.
unnecessary wear due to erosion. Nakayama et al. [61] proposed five different plant configurations
Even though oxy-fuel combustion provides the opportunity of covering both wet and dry recycle, see Fig. 9. The configurations
highly reduced emissions of NOx [4,12,16], it is doubtful if the differ in the position of the recycle path and the oxygen preheater.
legislations concerning emissions can be fulfilled without some Changing the position of these units will have an impact on the size
degree of NOx removal for all future power plants. During start-up of both the preheater, the electrostatic precipitator (ESP), and the
and shut-down which most likely will occur at air-firing conditions flue gas cooling unit. The choice of a wet versus a dry recycle will
NOx removal is likewise regarded necessary. It is possible that the influence the gas recirculation fan (GRF) as will the possibility of
SCR plant can be run at reduced efficiency during oxy-fuel recirculating fly ash to the boiler (system A). In relation to the gas
combustion and that operating costs thus can be reduced. The recirculation fan, a dry recycle is most favourable since it induces
SCR is typically positioned at high-dust conditions. By passing as less wear of the fan. On the other hand, a dry recycle will require
large an amount as possible of the flue gas around the SCR it is
possible to significantly enhance the catalyst lifetime. Even though
ammonia is not added to the flue gas, its passage through the SCR
catalyst will reduce the catalyst lifetime. It is likely that the extra
cost related to the bypass will be counterbalanced by the reduced
cost for change of catalyst may be even with an overall saving. An
alternative option is the production of a concentrated liquid waste
water stream containing both NOx and SOx [66] which can be
treated to avoid pollutant emissions to the air, see Section 2.3.2.
The acid gases such as sulphur oxides, HCl, and HF are not
reduced through the same mechanisms as NOx in the boiler.
Recirculating the flue gas before desulfurization will result in up to
an about 3e4 times increase of the sulphur (and chloride) levels in
the boiler [4], see Table 6. Since both sulphur and chloride are
potentially corrosive this significant increase can cause problems in
the boiler components when firing medium to high sulphur con-
taining coals [43]. Significantly increased levels will likewise have
a negative impact on a wet desulphurization plant. The change
corresponds to firing coals with excessive amounts of S, Cl, F, etc.
Sulphur retention in fly ash is another area of concern. Signifi-
cantly increased retention can lead to problems utilizing the fly ash
for cement and concrete production. The upper limit for sulphur in
fly ash used for concrete is 3 wt% measured as SO3 [104]. The
limiting value for sulphur in fly ash used in cement production is
not regulated by standards but rather determined individually
between the cement manufacturer and the fly ash producer.
However, the limit is usually not less strict than for concrete
purposes. Utilization of fly ash from power plants is becoming
increasingly important [105]. The potential risk of contaminating
the fly ash with excessive amounts of sulphur is thus very critical to
Table 6
Example of the effect of recycle strategy on SO2 concentration in the flue gas, based
on 1000 ppmv without recycle. Data taken from [4].
Fraction of total flue gas recycled Sulphur concentration in flue gas

0.7 3110
0.6 2370
0.5 1920
0.4 1650
0.3 1390
Fig. 9. Different, possible configurations for oxy-fuel combustion with either dry or
0.2 1230
wet recirculated flue gas. The suggested configurations do not include process units for
0.1 1080
removal of NOx and SO2. ASU: Air Separation Unit; GRF: Gas Recirculation Fan, ESP:
0 1000
Electrostatic Precipitator. Adopted from [61].
approximately four times larger capacities of the preheater, cooling, corresponding to the excess amount, approximately 3 vol%, will be
and drying units compared to the wet recycle case. present in the primary recycle.
Nakayama et al. [61] concluded that a wet recycle taken after
the ESP and oxygen preheater (system B) was the most promising 2.5.2.2. Secondary stream. If no oxygen is added to the primary
configuration. The evaluation is based solely on theoretical recycle the entire amount should be mixed into the secondary
considerations and no experiments were performed to test the stream. Due to the risk of spontaneous ignition of dust in the
effect of wet and dry recycle on the combustion process. The recycle stream there will most probably be an upper limit to the
suggested configurations include only particle removal for flue gas concentration of oxygen in this stream as well. If this limit is
cleaning, not deNOx nor desulfurization since these were not exceeded it will be necessary to inject oxygen either directly
considered necessary to obtain the required purity of the CO2 through the burners or the over-fire ports [83].
stream. The investigation likewise neglects the fact that some of Mixing of the oxygen stream into the recycled flue gas cannot be
the recycled flue gas should be used in the coal mills as transport considered unproblematic with the amounts relevant for a power
and drying gas. plant. It is crucial that the resulting oxidant stream has a homoge-
Dillon et al. [66] likewise assess different recycle arrangements neous composition in order to maintain stable burner operation
considering both a primary and secondary recycle stream. All of and prevent safety hazards due to injection of large amounts of
their five arrangements did, however, have the same conditions for pure oxygen to the furnace. Generally, it is very important to
the primary recycle, the only difference being the type of equip- minimize the risk of having pure oxygen present together with
ment for reheating. The authors arrived at a wet, reheated combustibles anywhere in the system. It is questionable if nozzle
secondary recycle as being the most advantageous option. injection of one stream into the other without e.g. a static mixer
As described in Section 2.2, mixtures of CO2 and H2O are known will be satisfactory. In the Schwarze Pumpe pilot plant static mixers
from the oil and gas industry to cause potential corrosion problems are used to ensure proper mixing of oxygen and flue gas [106,94].
in piping. Corrosion in the flue gas recirculation ducts in case of Some researchers are addressing alternative equipment which will
temperatures below the water dew point is thus an issue that allow proper mixing in the flue gas ducts [91].
should be addressed when designing an oxy-fuel plant. Table 7 summarizes the advantages and disadvantages associ-
For security and availability reasons, Hellfritsch et al. [82] ated with the different options for oxygen introduction to the
suggest that the flue gas recirculation should consist of multiple boiler. Either way the oxygen addition to the boiler is performed it
independent paths (e.g. 4 of 25% each). is most likely that new (or modified) burners will need to be
In relation to retrofit of an existing power plant the exact developed to obtain efficient and stable combustion in an oxy-fuel
benefits and disadvantages concerned with each of the possible plant [67,107e109].
plant configurations regarding secondary recycle position and its
water content should be clarified in order to choose the most 2.6. Operation of conventional flue gas cleaning equipment
optimal solution. For green-field plants it is possible that a specific
concept could prove to be generally accepted. In case it is necessary to include both desulfurization and deNOx
in an oxy-fuel combustion power plant the operation and perfor-
2.5.2. Oxygen addition mance of these units in atmospheres with high CO2 concentrations
2.5.2.1. Primary stream. Fig. 3 indicates that no oxygen is added to should be clarified. These aspects of oxy-fuel combustion seem to
the primary recycle before entering the mills. This is mainly due to be an overlooked area in the research since most investigations
safety reasons [83]. The highest possible temperature in the coal neglect these basic units in the plant configuration. Most of those
mills is 120 C. This limit is set according to materials who actually do include these units [4,10,63,80,82] assume that
considerations. In N2 there is no risk of explosions in the mills if the they can operate with exactly the same or even better settings,
oxygen concentration is below 12%. The use of air during operation throughput, and economics as in a conventional plant.
in conventional plants can thus be considered a safety risk and is
handled through safety precausions. Due to a suggested, inhibitory 2.6.1. Desulphurization
effect of CO2 on explosions it is possible that the O2 level in the The largest uncertainties regarding operation concerns the
mixture sent through the mills can be raised to above 21% without desulfurization plant. Below, the series of reactions occurring
inducing a risk of explosion [67]. No data are available however. during the process in a wet plant are provided. First, SO2 is dis-
Even though it is theoretically possible to have relatively high solved in the aqueous phase producing hydrogen sulfite, (1).
oxygen concentrations in the primary recycle without risking an Simultaneously, limestone (CaCO3) is dissolved releasing CO2, (2)
explosion, some authors advise against it [67]. In practical power and (3). The hydrogen sulfite is oxidized to sulphate in the irre-
plant operation, mills are continually brought in and out of versible reaction given by (4). Finally, gypsum (CaSO4$2H2O) is
operation to meet the changes in electricity demands. This produced by the reaction between calcium and sulphate ions, (5).
provides a possibility of transient mismatches between recycle
and oxygen flows and thus significant variations in the oxygen SO2 þ H2 O#HSO
3 þH
þ
(1)
concentration. Additionally, equipment failures in control valves,
recycle fans, etc. are considered severe security risks [67]. The risk CaCO3 þ 2Hþ #Ca2þ þ HCO
3 þH
þ
(2)
of obtaining significantly higher oxygen concentrations than the
safe limit in the coal mills and induce an explosion from ignition of
Ca2þ þ HCO þ
3 þ H #Ca
2þ
þ H2 O þ CO2 (3)
coal dust is thus profound.
With respect to NOx formation, it is likely that limiting the
oxygen concentration in the primary oxidant stream is beneficial HSO 2
3 þ 0:5O2 /SO4 þ H
þ
(4)
since a fuel-rich flame core will be formed. The only concern could
thus be flame stabilization. Some preliminary results obtained by
SO2
4 þ Ca
2þ
þ 2H2 O#CaSO4 $2H2 O (5)
IFRF [88,12] suggest, however, that satisfactory ignition and flame
stabilization is obtainable when all oxygen is introduced with the Since the wet desulfurization process rely on limestone emitting
secondary oxidant stream. It should be noted that oxygen CO2 during the conversion equilibrium limitations could arise due
to the high CO2 concentration in the flue gas. Likewise, the oxida- Table 8
tion of HSO 2
3 to SO4 , (4), in the desulfurisation plant is performed
Selected quality requirements for gypsum produced by wet desulphurization of flue
gas [112].
with air in conventional units. Application of air to oxy-fuel
desulfurisation would induce an undesirable dilution of the CO2 Parameter Value [wt%]
with nitrogen [68,101,110]. Instead, a portion of the secondary Purity >95
oxidant stream or alternatively pure oxygen could be directed to Residual limestone <4
Moisture <10
the desulfurisation unit. The consequences of each of the three
possibilities should be investigated. An other alternative could be to
construct the desulphurization plant in such a way that the SO2 the dew point, especially during low load operation, severe clogging
absorption and the oxidation occur in two separate units of the catalyst must be expected. Additionally, the catalyst acts as an
[68,76,94,110]. In that case air can be used as oxidizer without the oxidizer and will convert part of the SO2 content of the flue gas to SO3
risk of reducing the CO2 concentration in the flue gas. increasing the risk of ammonium bisulfate formation [113].
In the case of a wet recycle the flue gas entering the flue gas
desulphurization unit will contain significantly higher amounts of 2.6.3. Particulate removal
water vapor than in a conventional plant. Even in the case of a dry As Suriyawong et al. [114] state, the collection efficiency of
recycle, the water vapor concentration will increase due to less particulate removal devices is size dependent. Since oxy-fuel
dilution. The increase in water concentration could lead to an combustion shows the ability of altering the size distribution of
increase of the equilibrium temperature of the gypsum/limestone ash particles, see Section 3.5, the performance of the electrostatic
suspension and thereby influence the kinetics of the desulphuri- precipitator (ESP), which is the most common particulate removal
zation process. A higher temperature of the suspension could thus system in large-scale power plants, will be subject to changes. For
lead to a reduction in the degree of desulphurization due to both ultrafine (<100 nm) and sub-micrometer (100 nm < dp < 1 mm)
a decrease of the solubility of SO2. sized ash particles the penetration (1 e collection efficiency)
Studies performed by Air Liquide Inc. together with The Babcock depends on particle size due to different charging effects and these
& Wilcox Company in the 30 MWth test facility located in Alliance, particles are thus not effectively captured.
Ohio [91,51] revealed no noticeable change in the performance of A potential change in the size distribution of the fly ash from
a conventional wet FGD scrubber. Similar conclusions were obtained oxy-fuel combustion is not the only aspect of relevance for ESP
by Oryshchyn et al. [111] from laboratory-scale experiments. performance. The flue gas composition could likewise impose
Due to the fact that only few results on desulphurization in oxy- a change in the ion production rate within the ESP and thus lead to
fuel atmospheres have been published, this is an area of the oxy- a change in collection efficiency. Suriyawong et al. [114] have found
fuel process which needs further attention. that for ESP's with positive coronas it is necessary to increase the
The gypsum produced should likewise meet certain quality applied voltage to obtain similar collection efficiencies in oxy-fuel
requirements, see Table 8. The high partial pressure of CO2 in the plants as for air-fired plants. However, when negative coronas are
gas phase could limit the dissolution of the limestone in the solu- used, there is only an insignificant effect of the ESP performance
tion and thus potentially increase the amount of residual limestone from the change in flue gas composition in the oxy-fuel plant.
in the gypsum. This could cause problems in the further utilization Based on equilibrium calculations Schnurrer et al. [115] conclude
of the gypsum and would influence process economics negatively. that the increased SO3 content in the flue gas is expected to improve
the performance of the ESP. This is caused by lower ash resistivity as
2.6.2. NOx removal a result of deposition of sulphate species on the fly ash.
The concerns regarding removal of NOx from the flue gas stream
are much less than for desulphurization. However, it will still be 2.6.4. Potential improvements for a green-field plant
necessary to verify the performance of e.g. an SCR unit, which is the For a green-field plant there could be an incentive for applying
typical choice for large power plants, if the NOx reduction obtained a hot ESP as the first flue gas cleaning unit downstream of the
during the oxy-fuel combustion is not adequate. An increase in the boiler. The hot ESP is generally considered too costly for air-blown
CO2 concentration above the catalyst is not considered to have an units. Removal of the particulates as the first step will, however,
effect on performance. The possibly increased levels of both SO2 and yield the possibility of applying a low-dust SCR downstream of the
SO3 in the flue gas do, however, impose a risk for reduced ESP with low catalyst degradation compared to the more
performance. SO3 in the flue gas is known to form sticky and conventional high-dust option. The secondary recycle can then
corrosive ammonium bisulfate when NH3 is added [113]. Increasing potentially be taken in between the ESP and SCR, which addi-
the SO3 level will likewise increase the dew-point temperature for tionally reduces the necessary size of the SCR unit. In order to
ammonium bisulfate. If the SCR operating temperature falls below control the sulphur accumulation in the system and thus prevent
Table 7
Advantages and disadvantages associated with the different options for oxygen addition to recirculated flue gas.
Option Advantage(s) Disadvantage(s)

Before coal mills (1) Sufficient space in recirculation duct to ensure adequate (1) Risk of zones with very high-oxygen concentrations if
mixing mixing is inadequate and thereby increased risk of
(2) The stream to the mills will be drier than without oxygen combustion/explosions in mills
(3) With oxygen in primary stream burner design is expected to (2) Primary stream will be cooled due to the low temperature
be closer to that for conventional air burners of the oxygen
After mills, before burners (1) No risk of explosion in mills (1) Significant risks are associated with the injection of nearly
(2) Possible to obtain the same oxygen concentration in the pure oxygen into a high temperature stream of coal and
burner as is the case for operation on air flue gas
Directly in burners (either (1) No risk of explosion in mills or flue gas ducts (1) Burner design has to be re-thinked
pure oxygen lance or with (2) No expenses for mixing in primary stream (2) Limited capacity for coal drying in mills
the secondary stream) (3) Oxygen lance provides increased control of the mixing of
oxygen and fuel in the near-burner zone
both corrosion and ammonium bisulfate degradation of the SCR to obtain stable combustion. With this approach the total volu-
catalyst due to high SO3 levels, a desulphurization unit prior to the metric flow through the burners and boiler is reduced compared to
SCR should be installed. air-firing because of the overall lower amount of ‘inerts’. In order to
The major advantage concerned with the change of sequence of keep the coal particles in suspension, it has been suggested that the
the flue gas cleaning units is to reduce the size of the equipment due primary flow must have the same linear velocity as in air-firing
to the fact that the flue gas volume is reduced with about 60e80% [83], typically 17 m/s as minimum [116]. Since the density of CO2
after the secondary recycle has been taken. In the suggested likewise is higher than for N2, see Table 9, the mass flow ratio of
configuration, the desulphurization unit will, however, not be primary to secondary flows through the burner inlets is increased.
reduced in size. A disadvantage is that if a wet FGD (the conventional This could create an imbalance in the burner aerodynamics
choice) is installed before the SCR, additional cooling and reheating compared to air-firing and thus have an impact on ignition, flame
will be necessary. These operations will reduce the plant efficiency. shape, and mixing [116]. It is possible that the minimum velocity
will depend on the density of the primary stream. A lower value
2.7. Boiler and steam cycle than used in air-firing could then prove to be adequate for oxy-fuel
combustion. This would have to be tested in larger-scale setups.
The oxy-fuel boiler is subject to relatively extensive changes Besides increasing the oxygen concentration to obtain stable
regarding chemistry compared to the case of air-firing. The two combustion, other suggestions have been made:
main, combustion related properties that change during oxy-fuel Liu and Okazaki [127] suggested to stabilize the flame by
combustion are [5]: recirculating additional heat with the flue gas while maintaining
an air-like composition of the oxidant. In connection with the
Gas radiative properties application of Ion Transfer Membranes (ITM) for air separation
Gas heat capacity they likewise suggested heat recirculation to obtain stable
combustion with oxygen concentrations down to about 15 vol% in
Substituting CO2 (þwater vapor) for N2 in the oxidant results in the burner inlet. Utilization of ITM membranes will require rela-
a larger specific heat capacity compared to air, see Table 9. Firing tively low oxygen concentrations in the oxidant in order to achieve
with the same ratio between oxygen and ’inerts' will thus result in sufficient driving force in the air separation. The authors claimed
a lower flame temperature [118e122] and difficulties in stabilizing that the recirculation of heat should not affect the plant efficiency
the flame [12,123]. In order to obtain a stable flame and the same negatively. There has been no reports of experiments with this
adiabatic flame temperature in oxy-fuel experiments as in air approach in larger scale.
combustion overall oxygen concentrations between 25 and 42 vol% Another alternative approach is taken by Toporov et al. [107,108]
have been reported necessary [87,97,116,118,120e122,124e126]. who work on changing the burner design in order to obtain stable
The span is somewhat dependent on the coal type and the typical flames at low oxygen concentrations in the O2/CO2 oxidant mixture.
values reported for bituminous coals are 28e35%. This approach is likewise to enable the use of ITM membranes for
CO2 and water vapor are radiating species as opposed to N2. The air separation instead of a cryogenic unit and thus reduce the
radiative heat flux in the flame zone originating from the gas phase penalty on the plant efficiency from about 8 to between 3 and 5%
is thus expected to increase in oxy-fuel combustion compared to [108]. In order to obtain stable operation their aim is to provide the
air-firing at comparable gas phase temperatures. The radiative heat necessary heat to compensate for the higher heat capacity of the
flux from solid particles (soot, char, and ash) depends on a number oxidant and the heat used for the endothermic gasification reac-
of parameters. Radiation from ash particles depends mainly on the tions taking place in the near-burner region from the post-flame
particle temperature whereas radiation from soot and char is zone. This is done by promoting under-stoichiometric conditions
influenced by the mixing characteristics of the flame, local O2, CO2, near the burner region (l ¼ 0.6) and having a strong internal
and H2O concentrations, and temperature. recirculation zone to lead hot combustion products back into the
In coal flames radiation from soot can dominate the heat burner quarl. A burner design was obtained which allowed stable
transfer in the radiative section. The formation rate of soot can be operation with just below 21% oxygen in the oxidant and an overall
very sensitive to the mixing of fuel and oxidizer in diffusion flames. stoichiometry of l ¼ 1.3. The drawback of this concept is an increase
The near-burner flow dynamics are expected to change in oxy-fuel in the recirculation rate of flue gas. However, this is believed to be
combustion compared to air-firing due to the changes in flow rates fully offset by the reduction in power demands for the air separa-
through the burners if conventional burners for air-firing are used. tion process [107,108].
Soot formation and hence the radiation intensity is thus expected
to change as well. 2.7.2. Heat uptake
Existing boilers have been carefully designed to match the
2.7.1. Burner operation and flame stabilization radiative and convective heat transfer properties of the air-fired
Increasing the oxygen concentration to achieve similar adiabatic combustion process. Especially important is the distribution of
flame temperatures to that of air operation is the obvious solution heat transfer between the furnace chamber (evaporation) and
Table 9
Properties of gases at 1123 C and atmospheric pressure [116,117].
H2O O2 N2 CO2 Ratio CO2/N2

Density (r) [kg/m3] 0.157 0.278 0.244 0.383 1.6
Thermal conductivity (k) [W/m K] 0.136 0.087 0.082 0.097 1.2
Specific heat capacity (cp) [kJ/kmol K] 45.67 36.08 34.18 57.83 1.7
Specific heat capacity (cp) [kJ/kg K] 2.53 1.00 1.22 1.31 1.1
Heat sink (rcp) [kJ/m3 K] 0.397 0.278 0.298 0.502 1.7
Dynamic viscosity (m) [kg/m s] 5.02e-05 5.81e-05 4.88e-05 5.02e-05 1.0
Kinematic viscosity (n) [m2/s] 3.20e-04 2.09e-04 2.00e-04 1.31e-04 0.7
Mass diffusivity of O2 in X ðDO2 =X Þ [m2/s] e e 1.7e-04 1.3e-04 0.8
convective part (superheating) but the maximum heat flux to the 2.7.4. Fire-side corrosion
furnace walls is also important. The change in the composition of the combustion medium in
The choice of recycle ratio, i.e. the fraction of the total flue gas oxy-fuel combustion could potentially give rise to increased risk of
flow which is recirculated to the boiler, and thus the oxygen corrosion. The increased levels of CO2, H2O, and SO2 are expected to
concentration in the oxidant affects the heat transfer by controlling change the corrosion potential of the flue gas. Both aspects
the flame temperature and the volume flow through the boiler [65]. regarding metal surface oxidation and corrosion under deposits
The radiative heat transfer is mainly determined from the flame due to changes in deposit compositions are important to investi-
temperature and the gas radiative properties, whereas the gate [129e131]. Regarding deposits, especially the co-existence of
convective heat transfer is determined mainly by the Reynolds and sulphates and carbonates within the deposits could be critical to
Prandtl numbers, the thermal conductivity, and the temperature of the boiler tube surfaces [131]. The aspect of carburization of Cr-
the flue gas passing the superheater banks. Fig. 10 shows the containing steels due to high partial pressures of CO2 could like-
connection between the recycle ratio and the adiabatic flame wise constitute a challenge to materials lifetimes due to brittleness.
temperature and flue gas volume. In this respect, the flue gas Pirón Abellán et al. [132] studied the oxidation of martensitic
volume is a rough indicator for the convective heat transfer. It is steels (Cr-containing alloys) in different CO2 and H2O-containing
seen that recycle ratios of either 0.61 or 0.76 yield conditions cor- atmospheres at temperatures between 550 C and 700 C in a labo-
responding to air-firing (with l ¼ 1.15), although only for one of the ratory setup. They found significant carburization in high CO2 con-
parameters. It is thus not possible to simultaneously obtain both an taining atmospheres. They expect the carburization to be governed
adiabatic flame temperature and a volumetric flue gas flow through primarily by the permeability of the protective oxide layer toward
the boiler similar to that of air combustion. Section 3.2 will assess CO2. However, the authors show that the presence of water vapor will
the subject of heat transfer in more detail. limit the transport of CO2 through the oxide layer and thus the extent
of carburization. The exposure atmospheres contained no O2 and the
2.7.3. Oxygen excess consequence of the presence of excess oxygen on the obtained results
It is important to note that excess oxygen for oxy-fuel will need to be examined to fully clarify the effect of the change from
combustion carries a much greater penalty than in the air-firing air-firing to oxy-fuel combustion on fire-side corrosion.
case and that the power consumption in the ASU is directly
proportional to the amount of oxygen lead to the boiler [95]. A 2.7.5. Aspects regarding green-field plants
reduction in the oxygen excess would thus benefit both the ASU The above discussions are based on the case of a retrofitted
and CO2 compression train power demands [66,68,81,95]. In conventional air-fired plant to oxy-fuel conditions. For a green-field
modern power plants the air excess is typically 15e20% plant it will be possible to refine the boiler design. The adoption of
(l ¼ 1.15e1.2) [68,103,128] corresponding to about 3 vol% O2 in the conditions similar to those of air-firing is not necessarily the most
dry flue gas. The excess is predominantly determined by the optimal for an oxy-fuel combustion plant. A major target will be
uncertainty in the coal mass flow to each burner. By introducing a significant reduction or even complete elimination of the rate of
more efficient monitoring of the coal flow to each burner and thus external flue gas recirculation. Instead, control of the flame
a better control of the oxidizer and fuel flows, it should be possible temperature can be obtained from internal recirculation in the
to reduce the required oxygen excess during oxy-fuel combustion boiler [10]. This will likewise have a positive impact on the resi-
to no more than 10% [68]. At 10% excess oxygen and an oxidizer dence time and thus the burnout of the char particles. The possible
containing 30% O2 the dry flue gas O2 concentration will likewise be avoidance of the external recirculation will reduce the size of the
about 3 vol%. boiler and thus the boiler capital cost significantly [10]. At the same
Theoretical considerations regarding the second law efficiency time, efficiency loss due to thermal radiation to the environment
of gas-fired power plants equipped with either oxy-fuel combus- and the electricity demand for the flue gas recirculation fans will be
tion or post-combustion capture have been conducted by Simpson reduced.
and Simon [95]. Their computations showed that oxy-fuel Maintaining the external recycle is likewise an option. Instead of
combustion becomes more favourable with respect to efficiency matching the flame temperature to air-like conditions it would be
than post-combustion systems as the oxygen excess level is possible to shift heat transfer to higher temperature throughout the
decreased. entire boiler, leading to less entropy loss. Higher combustion
temperature may lead to more efficient combustion with better
burnout at reduced oxygen excess levels. Higher oxygen concen-
tration will at the same time require less flue gas recirculation and
1.4 thus a smaller volume flow through the boiler.
Normalized Tad and Vfg
1.2
2.8. Summary
1
The present section has discussed the effects of retrofitting
0.8
a suspension-fired power plant to oxy-fuel operation. A number
0.6 of new process units are necessary; an air separation facility to
provide the combustion process with almost pure oxygen, ducts
0.4 and fans for external recirculation of flue gas to the burners in
Tad,1/Tad,0 (K/K)
0.2 order to control flame temperatures, and a CO2 processing facility
Vfg,1/Vfg,0
including a flue gas condensation unit, compressors, and a puri-
0 fication unit to remove non-condensable species from the
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 product stream.
Recycle Ratio The review of the literature has shown that even though the cost
Fig. 10. Theoretical considerations on the effect of flue gas recycle ratio on adiabatic
and efficiency penalties associated with the technology are signifi-
flame temperature, Tad, and flue gas volume, Vfg. The reference values, Tad, 0 and Vfg, 0, cant, oxy-fuel combustion of pulverized coal is technically and
are for air combustion. Data taken from [88]. economically feasible for retrofitting of existing power plants. At the
same time, oxy-fuel combustion for CO2 recovery and sequestration reported in the open literature. The assessment will focus on
is a competitive power generation technology in relation to post- chemical aspects connected to the boiler and other issues directly
combustion capture with amines. However, a number of critical related to the combustion, e.g. convective and radiative heat
aspects regarding the technology and plant configuration have been transfer, corrosion, and emissions.
identified, as listed below. These are the issues which should be
investigated in further detail to ensure that the technology is in fact 3.1. Research groups and experimental facilities
a useful alternative to air-firing and post-combustion capture.
Plant availability: The risk of lowering the availability of the Tables 10 and 11 provide an overview of the different groups
plant due to the introduction of the additional auxiliary equipment active within oxy-fuel combustion research. The groups have been
could be crucial to implementation. divided according to the scale and type of their experiments, i.e.
Plant efficiency: With the currently available cryogenic tech- whether they run combustion experiments with flue gas recircu-
nology for large-scale oxygen production the efficiency penalty for lation in semi-technical scale or in once-through, laboratory-scale
an oxy-fuel power plant is significant. Besides alternative air reactors.
separation technologies, it is necessary to investigate additional Included in the tables is information on the most important
measures that will allow the improvement of the electrical parameters for the experiments; maximum reactor temperature,
efficiency. Reduced oxygen surplus during combustion and a high oxygen concentration, fuel type, and oxygen excess, as well as an
steam cycle efficiency would have a positive effect on the efficiency. interpretation of the aim of the research performed by each group.
Load changes: The relatively low gradients for load change in the
air separation unit compared to those for the thermal cycle could 3.2. Heat and mass transfer effects
provide a problem with respect to operation of the plant in a system
with a large amount of decentralized, non-constant electricity The present section covers aspects of the effects of operating
sources, e.g. wind mills. In a world with an increasing fraction of the parameters such as oxidizer composition and oxygen excess levels
electricity produced by variable sources, central power plants will on flame temperature and heat transfer in the boiler.
be required to run with frequent changes in the load to adjust to the As described in Section 2.7 the differences in the radiative and
demand for electricity. Additionally, the inability of current air thermo-physical properties of CO2 and N2, see Table 9, affect the
separation units to run at low load without major penalties to the combustion process through alterations of the temperature and
efficiency will require several oxygen plants running in parallel, as mass transfer properties of the gas phase. As a consequence, the
well as some sort of storage/buffer capacity. heat uptake in a boiler will likewise change compared to that of
Enhanced purification: The potential need for two purification a conventional air-fired plant.
steps, i.e. both the air separation unit and the pre-compression
purification due to leakage of air into the flue gas ducts, could be 3.2.1. Flame and gas phase temperatures
a major disadvantage to the process. Post-combustion, in compar- Croiset et al. [120] report the results of combustion of two
ison, only requires the amine absorption step. The exact require- different coals in air and different mixtures of O2 and CO2 in the
ments for the final CO2 quality are still to be determined and they 0.3 MW oxy-fuel combustion facility at CANMET. The coals tested
will dictate the plant configuration with respect to the different were a Canadian western sub-bituminous coal (Highvale) and a US
purification units. eastern bituminous coal. Temperatures along the centre-line have
Flue gas recycle: The optimal positioning of the flue gas recycle been measured with a suction pyrometer for air-firing and 21, 28,
withdrawal points has not been identified. The choice could have 35, and 42 vol% O2 in CO2, see Fig. 11. The experiments have been
a large impact on both operation, economics, and plant mainte- performed with an oxygen excess between 1.46 (21% O2) and 1.3
nance related to corrosion issues. (35% O2) for the Highvale coal. For the Eastern bituminous coal the
Combustion process: The significant changes to the combustion oxygen excess was between 1.2 (28% O2) and 1.1 (42% O2), corre-
process, including the potential effects on heat uptake, fly ash sponding to 5% oxygen in the flue gas (dry). The data in the lower
quality and fire-side corrosion likewise provide a challenge. figure show that for the 21% O2 case a direct substitution of N2 for
Several of the above aspects are difficult to examine either CO2 in the oxidant leads to a considerably lower temperature in the
theoretically or during small-scale tests. The information and flame zone, a decrease of up to about 300 C, caused by the higher
experience obtained from the operation of a larger-scale demon- specific heat capacity of CO2 compared to N2. By increasing the
stration plant will greatly improve the knowledge of the process. oxygen concentration it is possible to obtain flame and gas phase
The majority of issues regarding the combustion process is, on the temperature profiles almost similar to those obtained from
other hand, possible to clarify in smaller scale experiments. Much conventional combustion in air. The necessary oxygen concentra-
work has already been performed in this area including heat uptake tion is somewhat dependent on the type of coal, w35% for the
and burner stability measurements as well as aspects regarding coal bituminous coal and w31% for the sub-bituminous coal. Based on
particle ignition, burnout, flame propagation, radiating properties of additional experiments the authors conclude that up to 5% nitrogen
the flame, boiler efficiency, and changes in emission levels. In fact, in the combustion medium has no significant effect on flame
most of the literature related to oxy-fuel combustion is concerned temperatures. Comparable results regarding the effect of oxygen
with the combustion chemistry and emissions. concentration on flame temperature have been reported by Wang
In addition to the changes in the combustion, issues related to et al. [118], Liu et al. [121], and Tan et al. [122], among others.
corrosion are also very relevant for the evaluation of the process. Experiments in a 1.2 MW burner at Ishikawajima-Harima
Because of the increased concentration of CO2 and perhaps SO2 in Heavy Industries Co. Ltd. (IHI) [119] explored the effect of wet
the flue gas, the risk of enhanced high temperature corrosion (16 wt% H2O) and dry primary recycle (for coal transportation) on
compared to conventional air-firing is present. furnace temperature and flame stability, see Fig. 12. During the
experiments the flow rate and composition of the secondary
3. Oxy-fuel combustion fundamentals recycle was kept constant. It was found that drying the flue gas
increased the gas temperature near the burner by about 150 C for
The following subsections summarize the work performed on the same volumetric gas flow rate and thus helped stabilizing the
oxy-fuel combustion fundamentals by different researchers and flame by improving ignition stability. The heat capacity of water
596
Table 10
Oxy-fuel combustion research groups working in pilot scale and semi-technical scale with flue-gas recirculation.
Group Ref. Test unit Unit description Inlet O2 Recycle: Coal type Aim of work (interpretation of the authors)
[MWth] [vol%] wet/dry
M.B. Toftegaard et al. / Progress in Energy and Combustion Science 36 (2010) 581e625
Vattenfall [133,94,106,134, 30 Down-fired boiler, single burner Max 40 Dry PFGR/ Lignite, Bituminous, Schwarze Pumpe pilot plant. Investigate the entire
92,99] wet SFGR Biomass technology chain of an oxy-fuel combustion power plant in
an industrial relevant pilot scale.
B&W and Air [91,51,131,135] 30 B&W Clean Environment e e Bituminous, Sub- Verification of process parameters for full-scale design.
Liquide Development Facility bituminous, Lignite
ANL [118,86] 3 Tower furnace, single swirl burner 22e40 Wet/dry Sub-bituminous Determine practical feasibility of the oxy-fuel process from
combustion and heat transfer data.
IFRF [88,12] 2.5 Horizontal furnace, air staged swirl 21e26 Wet Bituminous Implementation of oxy-firing
burner
B&W [63] 1.5 Pilot scale boiler, air staged 21e25 Dry Sub-bituminous Feasibility of oxy-firing. Process performance, pollutant
combustion system formation.
IHI [87,119,136e139, 1.2 Vertical, down-fired furnace, swirl 27 Wet/dry 5 Bituminous Implementation of oxy-firing. Direct injection of O2. Ignition
116,115,128] burner characteristics. Effect of load changes. Ash impacts.
E.ON [140e142] 1.0 Combustion Test Facility Max 24 Wet 3 Bituminous Combustion characteristics, coal ash transformation and
deposition, corrosion (50 h tests).
IVD [143e148] 0.5 Vertical furnace, swirl burner 27e30 (dry Wet Bituminous, Lignite Identification of variations in combustion behaviour. Impact
basis) on facility components other than furnace. Continuous
operation for w100 h.
CANMET [120,97,89,149,150, 0.3 Vertical combustor, down-fired swirl 28e42 Wet/dry Bituminous, Emissions from oxy-fuel combustion. Modelling tool
9,151,122] burner Sub-bituminous, validation. Burner and boiler design.
Lignite, NG
Doosan Babcock [72,152,98] 0.16 Vertical furnace, down-fired low-NOx e Dry 6 Bituminous NOx Reduction Test Facility (NTRF). Demonstration of oxy-
burner fuel combustion. Supply of flue gas for test of novel flue gas
cleaning process.
Chalmers [123,90,153e156] 0.10 Down-fired furnace, swirl burner 25e29 Dry Lignite, Propane Cause of NOx reduction. Gas-phase reactions. Firing
strategies for NOx reduction. Radiation.
CIRCE [157] 0.10 Bubbling Fluidized Bed combustor 20e40 Wet 2 Sub-bituminous, Technical feasibility of oxy-fuel combustion of blends of coal
wood, olive and biomass.
residues
TU Dresden [101,110] 0.05 Vertical reactor, down-fired burner 20e30 e Lignite Investigation of combustion related parameters. Validation
of CFD model. Production of flue gas for flue gas cleaning
test pilot.
Table 11
Overview of oxy-fuel combustion research groups working in once-through laboratory-scale reactors.
Group Ref. Test unit Unit description Max T Inlet O2 l Fuel type Aim of work (interpretation of the authors)
[kWth] [ C] [vol%]
Tokyo [158,127,159] e Flat flame burner 1177 21 0.7e1.2 Anthracite/CH4 Cause of NOx reduction. Investigate concept of heat
recirculation.
Nagoya [96,160,161] e Quasi-1D EFR 1300 20e100 0.7e3.3 3 Bituminous, Semi- Influence of l, O2 concentration, temperature, and type of
M.B. Toftegaard et al. / Progress in Energy and Combustion Science 36 (2010) 581e625
Anthracite coal with and without initial NOx.
Nagoya [109] 145 Bench-scale vertical furnace > 3000 85e88 0.9e1.1 Sub-bituminous Verify lab-scale results for high-oxygen concentration oxy-
fuel combustion with respect to NOx emission and fuel
burnout. Burner design.
Leeds [162,121] 20 Down-fired coal combustor 1350 21, 30 1.2 7 Bituminous Comparison between air- and oxy-firing. Sensitivity to
different fuels.
IVD [163,164,143e148,165] 20 EFR 1400 21, 27, 35 1.15 4 Bituminous, Lignite Comparison between air- and oxy-firing. Effectiveness of
air-staging. Sulphur emissions formation.
Sandia [166,126,167,168] e Optical EFR facility 1730 6e36 e Bituminous, Sub-bituminous Ignition devolatilization, and combustion rate of pulverized
coal particles.
Sandia [169] e Multi-Fuel Combustor (MFC) 1350 12e36 6.67 Bituminous/NG NOx formation mechanism.
TU Munich [170,171] 200 Side-fired furnace e Up to 50 0.3e4 NG Proof-of-concept for controlled staging in oxy-fuel
combustion.
DTU [172e174] e Laminar Flow Reactor 1527 0.05e0.4 0.25e8 CH4 Investigate chemical effects of high concentrations of CO2.
Dependence of presence of NH3 and NO.
BYU [124,175,176] e Multi-fuel Flow Reactor (MFR), 1627 21e37 0.75e1.05 2 Bituminous, Sub- NOx formation and reduction.
laminar bituminous
TU Aachen [177] 25 Gas Combustion Chamber 900 14e21 1.15 CH4 Stabilization of flameless combustion with low O2
concentrations.
TU Aachen [107,108,178,179] 100 Lab Scale Oxy-coal Test Rig 1250 19e30 1.3 2 Lignite, 3 Bituminous Burner design for stable oxy-fuel combustion at low O2
concentrations. NOx emissions comparison between
flameless and flame combustion.
CSIC [180] e Entrained-Flow Reactor 1100 21e35 0.75e4 2 Bituminous, Anthracite, Ignition and burnout of coal/biomass blends.
Semi-anthracite, Biomass
Northeastern Uni. [181] 4.2 Laminar-flow Drop-tube 1327 20e100 e Bituminous, Lignite, Synthetic Comparison between air- and oxy-firing. Volatile flame and
Furnace char char burning phenomena. Effect of particle size.
Huazhong [182,183,125] e Drop-Tube Furnace 1500 20e40 e 5 Bituminous Ash formation. Comparison between air- and oxy-firing.
ECN [184] e Lab-scale Combustion 1450 25e28 e Coal, 2 Biomasses Coal/biomass blends. Effect of different burner zone
Simulator residence times on burnout, emissions, and fouling.
Utah (UU) [185] 29 U Furnace 1550 21 1.15 3 Bituminous, Sub- Fate of char-N during both air- and oxy-fuel firing.
bituminous, Lignite
Utah (UU) [186] 100 Oxy-Fuel Combustor 1088 21e30 1.15 Bituminous Provide data for validation of coal-jet ignition model.
Washington Uni. [117,114] e Tubular furnace 1200 20e50 e Sub-bituminous Understand effect of oxy-fuel combustion on sub-
in St. Louis micrometer particle formation and mercury speciation.
Pennsylvania [187] e Drop-Tube Reactor 1550 21e30 1.25 2 Bituminous Char burnout and CO emissions investigations.
State Uni.
597
1700 vapor is significantly lower than that of CO2, see Table 9. Taking
Air only this factor into account, the presence of water vapor in the
1600 28 % O2 recycle would act to increase the flame temperature, i.e. the
Flame temperature ( C)
35 % O opposite effect of what was observed. These findings thus suggest

o
2
1500 42 % O2 that factors such as radiation and endothermal radical formation
(O, OH, H, etc.) dominate the temperature effect of water vapor in
1400
the recycle.
1300 Injection of pure oxygen through the centre of the burner like-
wise increased the gas temperature in the near-burner area
1200 compared to a similar experiment with the same O2/CO2 ratio in the
oxidant but without separate oxygen injection.
1100
3.2.2. Radiative and convective heat transfer
1000 As mentioned in Section 2.7, it is essential to match the flame
Air
and heat transfer characteristics for oxy-fuel combustion to those
1600 21 % O2
for air-firing in case of a boiler retrofit. In order not to reduce plant
Flame temperature ( C)
28 % O
efficiency or induce operational difficulties, the ratio between
o
2
1500 35 % O2
radiative and convective heat transfer should be maintained similar
1400 to that of air-firing as well [12].
1300 3.2.2.1. Overall heat transfer. Experiments performed by IFRF

[88,12] have shown that a recycle ratio of 0.58 (26 vol% O2) yields
1200 radiative and convective heat transfer rates similar to those of air-
firing in their experimental setup. The observed, optimum recycle
1100 rate is marginally lower than the theoretically determined value, see
Fig. 10, which is caused by some degree of air leaking into the boiler
1000
0.5 1 1.5 2 2.5 3 3.5 and the fact that pure CO2 was used for fuel transportation [88].
Axial distance from burner (m) Investigations of the possible differences between wet and dry
recycle has been performed by the Argonne National Laboratory
Fig. 11. Top: Centre-line axial temperatures for U.S. eastern bituminous coal burned in [86], see Fig. 13. The experiments showed that the optimal
different O2/CO2 mixtures. The oxygen excess yields a dry flue gas O2 concentration of
(CO2 þ H2O)/O2 molar ratio is different in each case. The ratio
5%. Bottom: Centre-line axial temperatures for Canadian western sub-bituminous coal
(Highvale) coal burned in different O2/CO2 mixtures. The oxygen excess was between yielding the same overall heat uptake changed from 3.25 (equal to
1.46 (21% O2) and 1.3 (35% O2) corresponding to approximately 8% O2 in the dry flue a recycle ratio of 0.68 or 23.8 vol% O2 at the burner inlet) for the wet
gas. Both set of data are obtained in the CANMET 0.3 MW combuster operated at case to 2.6 (recycle ratio of 0.63 or 27 vol% O2 at the burner inlet) for
a firing rate of 0.21 MW. Data taken from [120].
the dry case. The optimum recycle ratios were likewise validated at
reduced load and with a bituminous coal. Wall et al. [128] saw
a similar difference between wet and dry recycle. Their values were
28% and 35% for wet and dry recycle, respectively. The difference
suggests that the absolute concentration of water vapor in the
boiler has a significant effect on overall heat transfer efficiency. The
Dry recycle
Gas temperature ( C)
Air combustion exact reason for this influence is not clear and modelling could
o
1600 O2/CO2 combustion

provide indications to the influence from different factors such as
O2 injection
1400 radiation, thermo-chemical properties of the gases, etc.
1200 3.2.2.2. Radiative heat transfer. The main contributor to the heat
transfer from a flame from conventional fuels is thermal radiation
1000 [4,5]. The radiating species considered for pulverized-coal-fired
systems are H2O, CO2, CO, SO2, soot, char, and ash particles
1800
800 [65,86,128,188].
Oxy−fuel with O2 injectionL/D
Gas temperature ( C)
Wet recycle
The presence of higher concentrations of CO2, H2O, and SO2 in
o
1600 Dry recycle

oxy-fuel combustion will increase the non-luminous radiation [65].
1400 Payne et al. [86] and Khare et al. [116] estimated gas emissivities in
both air- and oxy-fuel atmospheres. Even though their estimates
1200 are not in exact agreement, they generally find a difference of 0.1
between the two types of environments, e.g. from 0.45 in air to 0.55
1000 in oxy-fuel [116,128].
Because of the higher concentrations of H2O and CO2 in the
800
0 5 10 15 20 25 furnace and their higher gas emissivities compared to nitrogen, the
L/D radiative heat transfer in the boiler will exceed that of conventional
air-firing for the same adiabatic flame temperature [4,5,128,155]. In
Fig. 12. In-flame gas temperature profiles along burner axis in the IHI 1.2 MW burner order to obtain the required heat transfer profile in both the radi-
for a low-volatile bituminous coal (coal A). The overall oxygen concentration in the ative and convective passes a slightly lower oxygen concentration
combustion gas is 27 vol% (30 vol% in the secondary stream). Top: Effect of pure oxygen
injecting through burner in the case of a dry primary recycle stream. Bottom: Effect of
than required to reach a comparable adiabatic flame temperature
wet and dry primary recycle stream in the case of direct oxygen injection. Wet flue gas should thus be chosen [4,5,128]. Typical results suggests a decrease
contains 16 wt% H2O. Data taken from [119]. of the oxygen concentration of 2e3 vol%. The use of the adiabatic
60 200
Dry recycle Air
Radiative Heat Flux (kW/m2)

22 % O2
160
Heat Extraction (% input)
26 % O2
40 120
Furnace Total
Cooling Panel
Superheater Bank 80
Refractory Walls
20
40
0
0 2 4 6 8 10 12
60
0 x/Do
Wet recycle
Fig. 14. The effect of recycle ratio on the net radiative heat flux along the IFRF 2.5 MW
Heat Extraction (% input)
furnace operated with wet flue gas recirculation. Results obtained for a bituminous
coal during air combustion and oxy-fuel combustion with oxygen concentrations of
40 22% and 26% at the inlet. These oxygen concentrations correspond to recycle ratios of
0.73 and 0.58, respectively. Data taken from [88].
overall radiation from a solid-fuel flame is not known and could

thus be the subject of future research.
20 Shaddix and Molina [168] measured the soot cloud size for
a bituminous and a sub-bituminous coal in both N2 and CO2-based
combustion media. Fig. 15 show the results for the bituminous coal.
Generally, oxy-fuel combustion leads to larger soot cloud formation
compared to the nitrogen based experiments at equivalent oxygen
0
1 2 3 4 concentrations, confirming the results from Andersson et al.
(CO2 + H2O)/O2 Molar Ratio [153,155]. For increasing oxygen partial pressure the soot cloud size
decreases for both types of combustion atmospheres.
Fig. 13. Heat extraction in different furnace sections of a 3 MWth setup as function of
The effect of in-flame particulates on flame radiative properties
the (CO2 þ H2O)/O2 molar ratio in the flue gas for dry recycle and wet recycle. For the is mainly a function of their total mass and temperature. The
dry case the recycled flue gas contains 5% residual H2O due to lack of efficiency in the absolute amount of ash does not change between air- and oxy-
water removal system. Solid symbols correspond to experiments in air. The coal tested firing operation for equal fuel input. The amount of soot in the
is sub-bituminous (Black Thunder). Data taken from [86].
flame can, on the other hand, change drastically with changes in
the near-burner flow dynamics. Changes in the burnout rate of
flame temperature as a rough indicator of radiative heat transfer char can likewise affect the location of heat transfer by radiation in
characteristics is thus altered compared to air-firing [88,12]. the flame zone.
Fig. 14 shows the radiative heat transfer profiles for three Besides experiments on propane, Andersson et al. [188] con-
different coal flames; the air-firing case, as well as oxy-fuel flames ducted radiation intensity measurements in air- and oxy-fuel flames
with overall oxygen concentrations in the oxidant of 22 and 26 vol burning lignite. They found that the shift in combustion medium
%. The experiments are performed in the IFRF 2.5 MWth furnace. It is resulted in only a negligible increase in the total radiation intensity
seen that 26% O2 in the oxidant yields a profile very similar to that at comparable gas phase temperatures. Due to gas-particle overlaps
of air-firing, as expected. At the same time, the inlet oxygen and the fact that particle radiation dominates in solid-fuel flames the
concentration of 26% was reported to yield in-flame gas composi- effect of increased gas emissivity is limited. Regarding the gas phase,
tion trends, combustion performance, flame length, and flame the authors concluded, however, that operating with a wet flue gas
stability comparable to normal air operation. recycle would have a much larger impact on radiation intensity in
Experiments on oxy-fuel combustion of propane [153,155] have large-scale boilers than the increase in CO2 partial pressure. This
revealed an up to 30e60% increase in the flame radiation intensity observation is in line with that of Nozaki et al. [119].
(27% O2) at peak levels compared to air-firing. At the same time, the
gas temperature levels are generally slightly lower for the oxy-fuel 3.2.2.3. Convective heat transfer. The increase in both density and
case. The authors observed that the increased gas emissivity due to heat capacity of the flue gas from increased concentrations of CO2
enhanced CO2 concentrations could not account for the entire and water vapor will increase the heat transfer in the convective
difference. Further investigations showed that the in-flame soot section of the boiler [4,5] (for the same volumetric flow and
volume fraction increased when shifting to oxy-fuel conditions entering temperature as during air-firing). However, in order to
[153]. Due to the lower volumetric flow rates and higher fuel obtain an adequate temperature and heat uptake in the radiative
concentration compared to oxidant in the oxy-fuel combustion the section the flue gas flow is reduced due to the necessary increase in
concentration and residence time of soot precursors in the near- oxygen concentration. The resulting heat transfer rate in the
burner region are increased. The formation of soot will thus be convective section is not necessarily reduced in comparison with
promoted [155]. These mechanisms for a gaseous fuel most prob- air-fired conditions as a consequence of the reduced flue gas flow
ably also apply to volatiles combustion in coal-fired operation. The rate. The convective heat transfer is a function of the Reynolds
relative effect of an enhanced soot formation from volatiles on number, the Prandtl number, and the thermal conductivity of the
400 12
N
2
soot cloud size ( m)
CO 10
2
300
8
(ms)
200 6
i
4
100 open symbols = N2

10 15 20 25 30 35 2 filled symbols = CO2
oxygen concentration (vol%)
0
Fig. 15. Variation of characteristic soot cloud size during devolatilization of Pittsburgh
coal (high-volatile bituminious) for different gas compositions in the Sandia optical 14
EFR facility. Error bars represent 95% statistical error. Data taken from [168].
12
flue gas. All of the above will change with the change in flue gas
composition and the flue gas temperature associated with the shift 10
(ms)
from air-firing to oxy-fuel combustion. Whether the resulting heat
8
transfer rate in the convective section will match that of the air-
d
firing case is dependent on the specific case. Simulations would 6
be a valuable tool in the evaluation.
4
3.3. The combustion process
2
3.3.1. Devolatilization and ignition 0
Volatiles constitute a major fraction of the combustible matter of 10 15 20 25 30 35
most coals used for power generation. The combustion of volatiles oxygen concentration (vol%)
releases heat which is important for ignition, local stoichiometries,
Fig. 16. Variation of ignition delay (si) and devolatilization (sd) times for Pittsburgh
and pollutant emissions [172]. The differences in properties for N2
coal burned in different gas mixtures in the optical EFR at Sandia National Laboratories.
and CO2 presented in Table 9 suggest that devolatilization and The data is based on images collected with an ICCD camera. Vertical bars represent the
ignition in oxy-fuel combustion will occur at different rates than for statistical error. Ignition delay was defined as the time elapsed from particle injection
combustion in air if CO2 is substituted directly for N2. into the reactor to the time when 50% of the images correspond to devolatilization.
Shaddix and Molina [126,168] studied the effects of the pres- Data taken from [168].
ence of CO2 and enhanced oxygen concentrations on the devola-

tilization and ignition of pulverized Pittsburgh (high-volatile kinetic control for increasing oxygen concentrations. At a low
bituminous) and Black Thunder (sub-bituminous) coals in oxygen partial pressure and high temperature, oxygen transport
a laminar flow reactor. Fig. 16 shows estimated values of particle through the gas film surrounding the particles will be rate limiting.
devolatilization and ignition times for the Pittsburgh coal at At increasing oxygen partial pressures and constant temperature
1700 K in different gas mixtures. From the data it is concluded that the kinetic rate will remain almost constant whereas the rate of gas
the exchange of CO2 for N2 does not significantly affect the film transport will increase linearly with the pressure.
devolatilization time for this particular type of experiment. From The differences in particle ignition times observed in Fig. 16 is
a consideration of heating an inert particle, the authors reason a consequence of the differences in transport properties of the
that for a fixed gas phase temperature the only factor that influ- surrounding gas, the combustion heat release, and the reactivity of
ences the initial heating of a particle is the thermal conductivity the local fuel-oxidizer mixture [126,168]. Generally, increasing the
[126,168]. As seen from Table 9 the ratio between the thermal mixture reactivity and heat release will decrease the ignition time,
conductivity of CO2 and that of N2 is close to one at 1400 K. Hence, whereas an increase in the product of density and heat capacity
the initial heating profile of particles in each atmosphere will (the heat sink, rCp) leads to an increase in the ignition time. The
effectively be similar. Since coal devolatilization is an endothermic heat sink for CO2 is substantially larger than for N2, see Table 9,
process with a rate that is strongly dependent on particle leading to an increase in ignition time for similar oxygen levels.
temperature and heating rate the specific conditions of the Kiga et al. [136], Kimura et al. [87], and Liu et al. [121] have observed
reported experiments suggest that there should be no difference similar delays in the ignition of coal particles when burning in an
between the different atmospheres, as observed. The result is, atmosphere of 21% O2 in CO2. Increasing the oxygen concentration
however, only valid for the case of equal gas phase temperatures, will increase the characteristic reaction rate of the local mixture
regardless of the gas phase environment. For a practical boiler this and thus decrease the ignition time [126,168,189]. Changing from
assumption is not necessarily satisfied. 24 to 36% O2 is seen to have a large, absolute effect on the oxy-fuel
As the oxygen concentration increases the devolatilization as mixture. Comparing the results for air- and oxy-fuel combustion
well as the ignition occur more rapidly. In earlier work on air reveals that it should be possible to obtain similar devolatilization
combustion Murphy and Shaddix [166] proposed the increase in and ignition properties for oxy-fuel combustion as for conventional
the devolatilization rate with increasing oxygen concentration to be air-firing.
the result of (1) a closer proximity of the volatiles flame to the coal Suda et al. [190] measured the flame propagation velocity with
particle, and (2) a higher temperature of the volatile flame. respect to different coal concentrations, coal types, and ambient gas
Furthermore, modelling considerations suggested increasing compositions in a microgravity facility. Fig. 17 shows the observed
1.5 2400
Flame propagation velocity (m/s)
Coal A (N2/O2) char
Coal B (N2/O2) soot
bulk gas
Coal C (N2/O2) 2200
Temperature (K)
Open symbols: N
1 Coal A (CO2/O2) 2
Filled symbols: CO
2
Coal C (CO /O )
2 2
2000
0.5
1800
0 1600
0 1 2 3 4 10 15 20 25 30 35
3 Oxygen Concentration (vol%)
Coal concentration (kg/m )
Fig. 18. Variation of characteristic soot cloud and char combustion temperature,
Fig. 17. Effect of coal concentration on the flame propagation velocity for three
during devolatilization and subsequent char combustion in different O2/N2 and O2/CO2
different coals; A, B, and C. Measurements are performed in a microgravity facility. The
mixtures. The fuel is a Pittsburgh coal (bituminous) and the experiments are per-
coals differ in their volatile matter contents. Coal A has the highest amount (46.0 wt%),
formed in the Sandia EFR facility. Error bars represent 95% statistical error. Data taken
followed by coal B (36.2 wt%) and then C (32.9 wt%), all on an “as received basis”. Coal
from [168].
particle diameters: 53e63 mm, O2 concentration: 40% by volume. Data taken from
[190].
of the burning particle could alter the reaction equilibrium and
slow down the burning rate [117]. However, that conclusion is
effect of each parameter. The coal type is seen to have a significant
based on a simplified analysis.
effect on the flame propagation speed in both enriched air- and
oxy-fuel environments. Each case operates with 40 vol% O2 in the
3.3.2.1. Burnout times. Bejarano and Levendis [181] have con-
oxidizer. The difference in coal type is mainly a matter of the
ducted a series of experiments with non-intrusive optical multi-
volatile content of each coal. The speed is seen to increase with
colour pyrometry measurements of volatile flame and char burning
increasing volatile matter content. For a given coal in a given
phenomena in single particle environments. Fig. 19 illustrates
environment an optimum in the propagation speed exists with
average temperatures and burnout times measured for a high-
respect to the coal concentration. The optimum concentration is
volatile bituminous coal. The data indicate an effect of
seen to be approximately similar for all data series. According to the
exchanging N2 for CO2 and the effect of varying the O2 concentra-
authors, this particle concentration corresponds to an inter-particle
tion in the oxidizer. The lower diffusivity of O2 in CO2 compared to
distance roughly equal to the flame radius of a single, burning
N2 is the major cause of reducing the burning rate during char
particle [190].
combustion [167]. Generally, it is seen that an oxygen mole fraction
of 0.3e0.35 yields similar temperature and burnout data for oxy-
3.3.2. Volatile and char burnout
fuel combustion compared to combustion in air. By comparing
Besides the impact from exchanging N2 with CO2 on the devo-
the measured and calculated burnout times shown in the figure,
latilization and ignition of coal particles the burnout of both char
Bejarano and Levendis [181] conclude that the char combustion for
and volatiles is expected to be affected.
both air- and oxy-fuel combustion occurs in Zone II.
The coal burning process is typically limited by both chemical
Wang et al. [118] have shown a significant increase in char
kinetics and external diffusion (Zone II conditions). At 1400 K the
burnout rate when increasing the oxygen concentration in O2/CO2
diffusivity of O2 in CO2 is only about 75% of that in N2. The lower
combustion from 21 to 29%. Similar results were found by Naredi
diffusivity will thus reduce both the burning and heat generation
and Pisupati [187] from drop-tube experiments on bituminous
rates for the coal particles [169]. At the same time, the diffusivity is
coals with 21% and 30% oxygen in the oxidant. Tan et al. [12] refer to
a strong function of temperature. Hence, the effect of the enhanced
studies performed by Takano and coworkers in the IHI 1.2 MW
heat capacity of CO2 and the associated decrease in flame temper-
swirl burner which gave similar results. Moreover, a marked
ature for similar oxygen concentration in oxy-fuel and air
reduction in unburned carbon in the fly ash with increasing oxygen
combustion will further reduce the burning rate. Fig. 18 shows
level was observed.
trends in ambient gas phase as well as soot cloud and char
In general, the reported investigations have shown that
combustion temperatures for different combustion media obtained
increasing the oxygen concentration in the oxidant for oxy-fuel
by Shaddix and Molina [168] in their EFR facility. At equal oxygen
combustion can compensate for the larger specific heat of CO2
concentrations and gas phase temperatures combustion in O2/N2
compared to N2. Comparable burnout times are thus achievable.
mixtures generally yield higher temperatures of both soot clouds
and char than in O2/CO2 mixtures.
3.3.2.2. Gasification reactions. Both CO2 and H2O may contribute to
The consumption rate of volatiles in oxy-fuel combustion is
the burnout of char particles through gasification reactions, (6) and
likewise expected to be slower than in air due to the lower diffu-
(7), when these species are present in high concentrations at high
sivity of small hydrocarbons in CO2 compared to N2 [122,126,168].
temperature [191,166] and the oxygen excess, l, is significantly
In retrofit respects, burnout in oxy-fuel combustion with
lower than 1.
comparable adiabatic flame temperatures as in air is expected to
improve. This is because of the higher oxygen partial pressure
C þ CO2 #2CO (6)
experienced by the burning fuel, possible gasification by CO2 (and
H2O), and longer residence times due to the lower gas volumetric
flows [5,128]. Other researchers claim that excess CO2 in the vicinity
C þ H2 O#CO þ H2 (7)
Bituminous Coal Volatiles Bituminous Coal Chars

3200 3200
3000 3000
2800 2800
Temperature(K)
Temperature(K)
2600 2600
2400 2400
2200 N 2200 N
2 2
CO CO
2000 2 2000 2
100 % O 100 % O
2 2
1800 1800
Oxygen (mole fraction)
70
40
60
Burnout Time (ms)
Burnout Time (ms)

30 50
40
20 30
20
10
10
0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
Oxygen (mole fraction) Oxygen (mole fraction)
Fig. 19. Average particle temperatures (top row) and burnout times (bottom row) for a high-volatile-A bituminous Pittsburgh #8 coal (45e53 mm) burning in O2/N2 and O2/CO2 in
a laminar-flow drop-tube reactor at Tfurnace ¼ 1400 K. The temperatures shown correspond to w50% burnout times for each phase. Error bars correspond to one standard deviation,
s. Data taken from [181].
In low-temperature regions (400e900 C) gasification does not At fuel-lean conditions the burnout asymptotically approach
play a role. This is due to a much lower rate of reaction for gasifi- a value of 100%. It is seen that the burnout of the high-volatile coal
cation with CO2 compared to O2-combustion at these conditions generally is a little lower than for the low-volatile coal. This is
[192]. contrary to what was expected since the high-volatile coal should
Results presented by Wall et al. [139,128] indicate that the be more reactive. However, there may be significant uncertainties
volatile yield is higher during oxy-fuel combustion than in air. They in the experimentally determined burnout, particularly because of
attribute it to gasification. However, Borrego and Alvarez [193] alkali metal volatilization and possible ash carbonization from CO2.
found the opposite trend during their experiments. For both coals the overall lowest burnout is, as expected, observed
Due to the conflicting results, the possible influence of gasifi- for oxy-fuel combustion with a low oxygen content.
cation reactions during oxy-fuel combustion is an area which needs
further investigations.
3.3.2.3. Effect of coal properties. Liu et al. [162] investigated the

burnout efficiency compared to combustion in air for a range of
bituminous coals in their 20 kW swirl burner, see Fig. 20. Even
though the burnout efficiency differs between the samples it is
always higher for combustion in 30% O2/70% CO2 than in air.
Generally, the increase in burnout efficiency is between 0.5 and 2%.
An examination of the coal characteristics has not revealed any
obvious correlation between the burnout efficiency and the prox-
imate or ultimate analysis. The reason for the differences within
each coal type is suggested to be the higher oxygen concentration
combined with similar temperatures for the oxy-fuel tests. The
authors stress that gasification and longer residence times of the
burning particles in the combustion chamber could also contribute
to the higher burnout efficiency observed [162].
Fig. 21 shows the results of the work performed on burnout Fig. 20. Effect of combustion media on coal burnout efficiency for seven different
characteristics of different coals by Arias et al. [180]. The burnout of bituminous coals. Results obtained in a 20 kW down-fired swirl burner. The coals are
both coals show similar trends with changes in the stoichiometry. burned at l ¼ 1.2 in air and a mixture of 30% O2 in CO2 (oxy). Data taken from [162].
100 the results given in Fig. 22 show only small or insignificant differ-
DI ences between the investigated particle size ranges.
90 The burnout times for the char particles are seen to depend on
Burnout (%)
the particle size. As expected, the larger the particles the longer the
burnout times. The effect of particle size is seen to be more
80
significant in an oxy-fuel environment. A plausible explanation
could be the lower diffusivity in CO2 compared to N2. Comparable
70 results have been obtained by Huang et al. [189] from experiments
in a TG thermal analyser.
100
DA 3.3.2.5. CO. At high partial pressure and high temperature CO2 can
90 dissociate into CO and O2 through the strongly endothermic reac-
Burnout (%)
tion, (8), [172].

79 % N2 − 21 % O2
80
79 % CO − 21 % O
2 2 CO2 #CO þ 0:5O2 (8)
70 70 % CO − 30 % O
2 2 In the flame zone in oxy-fuel combustion both of the above
65 % CO2 − 35 % O2
conditions are present. Gasification of carbon, reaction (6) and (7),
60 could likewise contribute to increase in the production of CO during
0.5 1 1.5 2 oxy-fuel combustion. However, among researchers there is
a difference in opinion on whether it is the thermal dissociation,
Fig. 21. Effect of gas atmosphere and stoichiometry on burnout of a low- and a high- [175], or the gasification reactions, [149], that play the dominant
volatile coal (DI and DA, respectively). The experiments are performed in an entrained- role in the significant increase in CO concentration in the flame
flow reactor. Data taken from [180]. compared to air-firing.
An even more important pathway to the increase in CO
concentrations in the flame zone of oxy-fuel flames could be the
3.3.2.4. Particle size effects. An investigation of the effect of particle reaction between CO2 and H radicals, see reaction (9). The increased
size (45e53 and 75e90 mm) on volatiles and char temperatures has level of CO2 induces a possible alteration of the composition of the
been performed by Bejarano and Levendis [181]. Given that the O/H radical pool during oxy-fuel combustion due to the competi-
actual data and the measurement uncertainty has not been stated, tion between O2 and CO2 for H radicals, (9) and (10) [172,194].
Bituminous Coal Volatiles Bituminous Coal Chars

3200 3200
3000 3000
2800 2800
Temperature(K)
Temperature(K)
2600 2600
2400 2400
2200 2200
2000 2000
1800 1800
45−53 m N 45−53 m N
2 2
45−53 m CO2 120 45−53 m CO2
40
75−90 m N 75−90 m N
Burnout Time (ms)
Burnout Time (ms)
2 100 2
75−90 m CO 75−90 m CO
2 2
30 80
20 60
40
10
20
0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
Oxygen (mole fraction) Oxygen (mole fraction)
Fig. 22. Trends in average temperatures (top row) and burnout times (bottom row) for a high-volatile-A bituminous Pittsburgh #8 coal of two different size cuts (45e53 and
75e90 mm) burning in O2/N2 and O2/CO2 in a laminar-flow drop-tube reactor at Tfurnace ¼ 1400 K. The temperatures shown corresponds to ca. 50% burnout. Data taken from [181].
3.3.2.6. Summary. The difference in the thermo-chemical prop-

H þ CO2 #CO þ OH (9) erties of N2 and CO2 affects the combustion process. At 21% O2 in
CO2 devolatilization, ignition, and burnout proceeds at a lower
H þ O2 #O þ OH (10) rate than in air. However, increasing the oxygen concentration in
the oxy-fuel environment to about 30% to obtain the same
Due to the high partial pressure of CO2 in oxy-fuel combustion the
adiabatic flame temperature as in air yields similar devolatiliza-
ratio of OH to both O and H will increase and the total amount of
tion, ignition, and combustion rates as seen in air combustion. At
radicals is expected to decrease compared to air-blown
comparable adiabatic flame temperatures oxy-fuel combustion is
combustion. Besides the reaction with H radicals, reaction of CO2
reported to yield improved burnout of char. This is most likely
with CH2 radicals will likewise contribute to enhance the concen-
due to longer residence times and higher partial pressures of
tration of CO in the flame zone [172,194e196].
oxygen in the vicinity of the burning particles. Even though
Whether the emission of CO from oxy-fuel combustion is larger
contribution of gasification by CO2 and/or H2O to the increased
than from air-firing has been subject to investigation. Because of its
burnout is suggested several times in literature, it is questionable
severe toxicity it is important that oxy-fuel combustion does not
if this effect is of significant importance for the conversion of the
lead to an increased CO emission. Wang et al. [118] report no
fuel.
significant difference in the CO concentration levels both within the
CO levels in the flame zone are generally reported to increase
latter part of the flame zone and in the exhaust from a 3 MWth
significantly in oxy-fuel combustion compared to air-firing. Even
horizontal furnace for both air- and oxy-fuel combustion experi-
though the high CO2 levels prevent CO from being oxidized at high
ments with comparable flame temperatures. A five time increase in
temperatures complete conversion is expected when excess oxygen
the CO level within the flame was expected however based on
is present during cool-down of the flue gas.
a modelling study. For an oxy-fuel experiment with an air-like
composition of the oxidant the CO concentration in the flame
3.4. Gaseous pollutants emissions
zone was expected to be slightly larger than for air combustion
based on modelling. The observed values for the oxy-fuel experi-
The following subsections discuss the reported knowledge on
ment were lower than calculated. All experiments yielded full
the most important gaseous emissions besides CO from oxy-fuel
burnout of CO before the exhaust.
combustion; NOx, SO2, and SO3. Trace elements in the gaseous
Experiments performed in the IFRF 2.5 MWth furnace by
phase, e.g. Hg, Cd, As, and Se, are likewise treated. No results have
Woycenko et al. [88] show significantly increased CO levels within
been published regarding emissions of HCl and HF.
the flame zone. Still, the combustion of CO has completed before
the furnace exit and no significant CO emission is observed. 3.4.1. NOx
Similar results were obtained by Scheffknecht and coworkers 3.4.1.1. NOx formation. In conventional air combustion the gener-
[164,148] and Liu et al. [121] even when the excess oxygen level ally accepted pathways for NOx formation are the following three
was slightly higher for the latter oxy-fuel test. Changing the mechanisms [197e199], see e.g. Fig. 23:
oxygen concentration from 30 to 21% increased the CO emission Thermal: Thermal NO formation results from N2 and O2 reacting
from 34 to 200 ppmv, due to delayed ignition and lower peak at high temperatures (above 1500 C) to form NO. The mechanism
temperature. involves three reactions, known as the extended Zeldovich
Experiments on natural gas combustion performed by Glarborg mechanism.
and coworkers [172e174] indicate that the high levels of CO2 in
oxy-fuel atmospheres prevent complete oxidation of fuel (CO) to
N2 þ O#NO þ N (11)
CO2 at high temperatures even when excess oxygen is present.
However, the effect is most pronounced at fuel-rich or stoichio-
metric conditions. There is no indication that high CO2 levels O2 þ N#NO þ O (12)
influence CO oxidation at low temperatures when oxygen is in
excess and hence there should be no increased risk of high CO N þ OH#NO þ H (13)
emissions from a gas-fired oxy-fuel plant if mixing of fuel and
Prompt: Prompt NO is formed when hydrocarbon radicals in
oxidant is adequate.
fuel-rich zones attack molecular nitrogen to form cyanide species,
which subsequently form NO when oxidized. These reactions can
take place at temperatures lower than is required for thermal NO
formation.
Fuel: Fuel NO is derived from nitrogen in the fuel reacting
through either volatile-N or char-N. Nitrogen released with the
volatiles further decomposes into cyanide and amine species. These
intermediate species may react to produce N2 or NO, depending on
the conditions. Char-N reacts through heterogeneous reactions and
intermediate CN species to eventually produce NO or N2. Detailed
understanding of the reaction pathways for the conversion of char-
N has not been established and is still an area of active research. The
split between NO and N2 on an overall level depends on factors
such as the nitrogen content of the coal, its rank and volatility, as
well as the stoichiometry.
It is generally assumed that up to 20% of the total NOx formed from
pulverized coal combustion in air is due to thermal NOx and about
80e100% is derived from fuel-N while the prompt NOx mechanism is
Fig. 23. The overall mechanism of NO formation and reduction. Vol-N is an interme- negligble, depending on the quantity of fuel-bound nitrogen species
diate gaseous compound, e.g. HCN or NH3. [199]. The low level of molecular nitrogen in oxy-fuel combustion will
suppress the formation of thermal and prompt NOx and hence the dominant mechanism in reducing NOx emissions and that it
potentially lower the overall NOx emission rate. accounts for 50e80% of the decrease in NO observed in oxy-fuel
The general conclusion in published literature is that the combustion. Reactions of NOx with N-volatiles were estimated to
amount of NOx emitted from an oxy-fuel plant can be reduced to contribute 10e50% to the NO reduction. The reduction of NO on
somewhere between one-third and half of that from combustion char in oxy-fuel combustion was found to be of minor importance,
in air [51,63,87,88,91,97,119,120,128,162]. However, the application but their results may be biased by the choice of a char with a low
of an oxygen concentration higher than 21% to obtain adequate reactivity (formed from anthracite).
flame temperatures could result in an enhancement of fuel-NOx Effect of the low N2 concentration: At a very low concentration of
formation [122]. N2 in the oxidizer, the thermal NOx formation mechanism can be
The potential for reducing the NOx emissions from a power plant disregarded since the oxy-fuel combustion atmosphere will be
considerably compared to air-firing [4,200] has been one of the key oversaturated with NO at all times [123,156]. For this reason the
drivers in oxy-fuel combustion research, particularly in USA [5] and Zeldovich reactions will be reversed and serve to convert NO to N2.
thus, NOx chemistry has been one of the most heavily investigated A modelling study using detailed chemical kinetics on oxy-fuel
areas within the oxy-fuel combustion technology. Normann et al. combustion of lignite performed by Andersson et al. [90] indicates
[201] have recently published a review on emission control of that while the formation of NO from fuel-N is the same or even
nitrogen oxides from oxy-fuel combustion. slightly higher for oxy-fuel combustion compared to combustion in
The potential for a significant decrease in NOx formation is air, the destruction mechanisms at oxy-fuel conditions are
mainly important for oxy-fuel power plants with a configuration of enhanced compared to air conditions, leading to an overall reduc-
the flue gas clean-up train which releases the impurities, including tion in the NO emission. At high temperatures (>1500 C), the
NOx, to the atmosphere. If instead a CO2 cleaning process is chosen near-elimination of molecular nitrogen in oxy-fuel combustion
which captures NOx and produces e.g. nitric acid the NOx formation facilitates reduction of NO by the Zeldovich mechanism. In order to
rate is of less importance although it will affect process economics. obtain an effective reduction of NO by the reverse Zeldovich
In any case, if NOx formation can be reduced without significant mechanism, a sub-stoichiometric combustion and/or an insignifi-
negative consequence to the remaining combustion process it cant amount of air ingress into the furnace is required (air-staging
is desirable. to promote long enough residence time) [156]. At the same time,
applying a wet flue gas recycle will increase the NOx reduction
3.4.1.2. Mechanism for the reduction of NOx during oxy-fuel potential [156].
combustion. The majority of homogeneous NOx chemistry is Effect of the increased CO2 concentration: Only few studies have
taking place in the devolatilization and near-flame zone. The been reported on the impact of high CO2 concentrations on the
implications of oxy-fuel combustion on the homogeneous NOx nitrogen chemistry in oxy-fuel combustion. Mendiara and Glarborg
formation and destruction are therefore likely to be alterations in [174,173] investigated the implications of the alterations to the
equilibria and reaction pathways caused by higher concentrations of radical pool as described in Section 3.3.2 on NO formation and
CO2 and NOx in this part of the furnace. If the recycled flue gas is wet destruction. They considered oxidation of methane doped with
the higher partial pressure of H2O will likewise interfere through its either NH3 (used to simulate volatile-N) [174] or NO [173] in
influence on radical formation and destruction. The enhanced levels a laminar flow reactor. According to their results, CO2 acts to
of CO2 and NOx are also expected to affect the heterogeneous promote NO formation from volatile-N under fuel-rich conditions
formation and reduction of NO. In addition to chemical effects, whereas it inhibits NO at both stoichiometric and fuel-lean
changes in the mixing patterns between fuel and oxidizer are likely conditions. At low oxygen concentrations formation of NO is fav-
to affect the nitrogen chemistry. In the following, these mechanisms oured by the increase in OH-concentration. At conditions with
for reduction of NOx in oxy-fuel combustion are discussed. oxygen excess, the reduced formation of NO is caused by the
Effect of the increased NO concentrations (reburning): Recircula- limitation of the O/H radical pool, particularly O. In reburning,
tion of NO with the flue gas through the burner would be expected a high concentration of CO2 had only a small impact on the NO
to lead to a considerable reduction of the nitrogen oxides emission reduction efficiency under reducing conditions, while at stoichio-
through processes similar to those of reburning. Reburning is metric conditions CO2 slightly enhanced removal of NO.
a chemically complex process in which nitric oxide is abated using Okazaki and Ando [158] pointed to the increased importance of
fuel as reducing agent [202]. The process involves partial oxidation reduction of NOx on char surfaces due to reaction with CO. The CO
of the reburning fuel under fuel-rich conditions, reduction of NO by levels increase because of the high CO2 concentration as described
reaction with fuel fragments, and subsequent conversion of the in Section 3.3.2. The promoting effect of CO on NOx reduction on
intermediate nitrogenous species. char is well documented [199,203e206], but this mechanism is
The reduction of NO may involve the following types of reaction: mostly important at fluidized bed conditions [198], i.e. at
temperatures well below those characteristic of pulverized fuel
1. NOx is reduced to cyanide and amine intermediates by reaction combustion.
with hydrocarbon radicals, formed from the volatiles released A change in selectivity of the release of char-N to NO has been
from the coal particles in the early flame zone. suggested by Park et al. [207]. They investigated nitrogen release
2. NOx is converted to N2 through reactions with other reactive from pulverized bituminous coal char during the reaction with O2,
nitrogen species (XN), such as cyanides and amines. The XN CO2, and H2O using Ar as carrier gas in a fixed bed flow reactor.
species may be formed either from from the hydrocarbon-NO Their experiments showed that increasing the concentration of CO2
reactions or from release of N-volatiles in the early flame. in an H2O and O2 free environment reduced the char-N to NO
3. NOx is converted to N2 by heterogeneous reactions on char release ratio. In contrast to gasification with H2O, no HCN or NH3
or soot. were measured when CO2 was the reactant. These experiments
thus suggest that the increased levels of CO2 during oxy-fuel char
From experiments performed in a small-scale laboratory combustion may suppress the formation of NO and its precursors.
reactor, Okazaki and Ando [158] evaluated the relative impact of Effect of changes to the flame and fuel/oxidizer mixing pattern:
these mechanisms for NO reduction. They concluded that reburn Implementation of oxy-fuel combustion may bring about changes
type reactions, i.e. reactions of recycled NOx with hydrocarbons, is in the flame structure and the fuel/oxidizer mixing pattern. These
changes will depend on the burner geometry and the degree of different operating conditions and design aspects. Results from
recirculation and the composition of the flue gas. both once-through laboratory reactors and setups with flue gas
Mackrory et al. [124,175] identify the following mechanisms recirculation are included. In each section a further subdivision is
that may enhance NOx reduction in oxy-fuel combustion: made, separating the different types of experimental setups, i.e. EFR
type experiments from swirling flames. The underlying mecha-
Less secondary oxidizer entrainment into the burner's recir- nisms of NOx formation and destruction do not change between
culation zone due to a more detached flame, i.e. reduced setup types. However, scaling effects and the fact that mixing of fuel
oxygen availability and limited initial NO formation. and oxidizer in larger burners is crucial to NOx formation will
Temperature increase in the fuel-rich zone will increase the influence the absolute levels of NOx emitted. A direct comparison of
rate of NO destruction. At higher temperatures the conversion small laboratory-scale setups and larger swirl burners with respect
of volatile-N proceeds faster toward N2 than NO. to NOx concentrations and emission rates is thus not possible. A
Reduced NO formation from char since more fuel-N is clear example of this is the experiments performed by Hasatani
released with the volatiles. et al. [161,109]. They burned a sub-bituminous Indonesian coal in
Indirect effects through changes in reaction rates (combus- an oxy-fuel environment with oxygen concentrations between 85
tion) and temperatures from the enhanced importance of and 88 vol% in both an entrained-flow reactor [161] and in
gasification reactions. a 145 kWth vertical furnace [109]. They observed NOx concentra-
tions of 4700 ppm and 1500 ppm, respectively, for otherwise
Their overall conclusion is that faster NO destruction in oxy-fuel similar combustion conditions.
combustion appears to be at least partially due to the higher CO and
NO concentrations. 3.4.1.4. Effects of changes in the oxygen concentration, oxygen excess,
flue gas recycling ratio, and gas phase temperature. Experiments on
3.4.1.3. Summary of reported, experimental results. Table 12 NOx emissions from oxy-fuel combustion of a high-volatile bitu-
provides a summary of the reported results on NOx emission minous coal in the IFRF 2.5 MWth furnace were performed in the
from oxy-fuel combustion in setups with flue gas recirculation early nineties by Woycenko et al. [88]. As shown in Fig. 24 they
compared with that from air-firing. Most of the experiments yield observed considerably higher NOx concentrations during oxy-fuel
a decrease in the NOx emission rate during oxy-fuel combustion. combustion with 26 vol% O2 in the oxidant than in the air-firing
However, some experiments show an increase which is suggested case. However, the total mass of NOx formed per energy input of
to be caused by the fact that the mechanism of fuel-NOx formation coal [mg/MJ], also termed the emission rate, was significantly lower
is very sensitive to the method with which oxygen and fuel is mixed in the oxy-fuel case which was assigned to the near-elimination of
in the flame [122,124,151]. As oxy-fuel burners are typically adop- thermal NOx formation and reburning of recirculated NOx.
ted from air-firing principles the resulting fluid dynamics of the Further experiments showed that an increase of the concen-
flame can be disadvantageous with respect to NOx emissions even tration of oxygen in the oxidant, and thus the flame temperature,
though the flames are stable. yielded an increase in both the flue gas NOx concentration and the
In the following sections the reported results on NOx emissions emission rate. The increase in oxygen concentration was performed
from oxy-fuel combustion are divided according to the effects of by decreasing the flue gas recycle ratio. Similar trends of reduced
Table 12
Summary of NOx emission results from experiments in pilot scale and semi-technical scale with flue-gas recirculation.
Author(s) Fuel input Inlet O2, Emission Conversion Conclusion(s)

(MW) Oxy (%) (mg/MJ) ratio (%)
Payne et al. [86] 3 Dry: 28 Air: 560 e NOx formation is reduced by approximately 70% for dry recycle and 80% for
wet: 24 Oxy, wet recycle compared to air-firing.
dry: 160
Oxy,
wet: 95
Woycenko 2.1 26 Air: 320 Air: 30 NOx formation is lower in oxy-coal combustion with flue-gas recirculation
et al. [88] Oxy: 50e150 Oxy: 5e14 than in air case.
Châtel-Pélage 1.5 25 Air: 100e175a e Potential for drastic NOx reduction of up to 60e70%
et al. [63] Oxy: 25e55
Kimura and 1.2 27 Air: 340 Air: 30e33 NOx conversion ratio in oxy-fuel combustion is significantly lower than that
coworkers [87,119] Oxy: <90 Oxy: <8 in normal air combustion because the recycled NOx is rapidly reduced to
HCN or NH3 in the combustion zone (reburning).
Scheffknecht and 0.5 27e30 Air: 90e350b e NOx emission rate increases for increasing burner stoichiometry and is
coworkers [144] Oxy: 55e325 always higher for air combustion than oxy-fuel combustion for comparable
operation conditions.
Croiset and 0.21 28, 35, 42 Air: 340 Air: 35c High NOx concentration inside the furnace but lower NOx emission rate with
coworkers [120,97] Oxy: 100e210 Oxy: 10e22 the flue gas than for air-firing.
Chui et al. [150,151] 0.21 28 Air: 110 Air: 14 NOx formation is strongly dependent on burner design, i.e. the near-burner
Oxy: 62e150 Oxy: 8e19 flow field. NOx emission from oxy-fuel combustion can change from lower
to higher than in the air-firing case.
Tan et al. [122] 0.21 35 Air: 211e269 Air: 21e31 NOx emissions dependent on coal type. Lignite, sub-bituminous, and
Oxy: 68e233 Oxy: 8e24 bituminous coals tested. Specially designed burner for oxy-firing
significantly reduces the NOx emission rate.
Andersson et al. [90] 0.1 25, 27, 29 Air: 150 Air: 24 The reduction in NO emissions for oxy-fuel combustion is due to an
Oxy: 40e50 Oxy: 7e8 increased destruction of formed and recycled NO compared to air-firing. The
conversion of fuel-N to NO is similar or even slightly higher during oxy-fuel
combustion than in air.
a
The interval reflects the variation between l ¼ 0.68 and 1.2 for the air tests and l ¼ 0.85e1.05 for oxy-fuel tests.
b
The interval reflects the variation between lburner ¼ 0.75 and 1.15 for both the air- and oxy-fuel tests.
c
Conversion made using coal HHV.
reduction of NOx by reducing agents, such as hydrocarbons from

pyrolysis and resident char. They have performed a series of
experiments in their entrained-flow reactor analysing the influence
of oxygen concentration, oxygen excess, and gas phase temperature
to clarify the importance of each parameter on NOx formation and
destruction. Fig. 25 shows, as expected, that the NOx concentration
generally increases with increasing oxygen concentration, partly
due to decreased dilution, and shows a peak at near stoichiometric
conditions. The NOx emission rate (bottom row), however,
increases approximately linearly near l ¼ 1 when changing from
fuel-rich to fuel-lean conditions. At high oxygen excess for all
oxidant compositions the NOx emission rates approach asymptotic
values. Increasing the gas phase temperature results in both
increasing NOx concentrations and emission rates for all stoichio-
metric values. As accentuated in Fig. 26 for a temperature of 1273 K,
it is likewise seen that the emission rate yields a slight peak at
around 50% O2 in the oxidizer, however most significantly for fuel-
lean conditions. Hu et al. [96] state that this phenomenon is
Fig. 24. NO emissions and flue gas concentrations for air-firing and oxy-fuel
a consequence of the concentration of reducing agents increasing
combustion (oxy) with and without low-NOx burner technology. The fuel, a high-
volatile bituminous coal (Göttelborn), is burned in the IFRF 2.5 MWth furnace. The proportionally to the increase in O2 concentration (decreasing
combustion is performed at an overall stoichiometry of l ¼ 1.15 and the inlet oxygen recycle ratio). At low oxygen concentrations (<50 vol%) they thus
concentration in the oxy-fuel experiments is 26%, corresponding to a recirculation assume oxidation to dominate and NOx emissions to increase with
ratio of 0.58. The NOx concentration is in the dry flue gas and is normalized to 0% O2. increasing O2 concentration. For O2 concentrations above 50%,
Data taken from [88].
reduction by reducing species should play the dominant role and
NOx emissions decrease with increasing O2 concentration [96].
The concentration of reducing agents does undoubtedly
NOx emission rates during oxy-fuel combustion compared to air- contribute to the behaviour in Fig. 26 but other factors may likewise
firing have been reported for different fuels in laminar flow reac- contribute. As the oxygen concentration is increased the reaction
tors [174], entrained-flow reactors [96,160,161,169], and swirling between char and oxygen proceeds faster which will result in
flames [72,90,97,120,121,164]. higher particle temperatures. The fractional conversion of char-N to
According to Hu et al. [161] there are two opposing factors NO has been reported to increase initially with temperature [199]
influencing NOx formation and reduction in coal combustion; the until the point where the oxygen-char reaction approaches the
oxidation of fuel-N by oxygen and other oxidizing species, and the diffusion limited regime. A further increase in particle temperature
20 % O2 / 80 % CO2 50 % O2 / 50 % CO2 80 % O2 / 20 % CO2
5000 1123 K
1273 K
1423 K
4000 1573 K
NOx (ppm)
3000
2000
1000
0
12
NO (mg N / g coal fed)
4
x
0
0.5 1 1.5 2 2.5 3 1 1.5 2 2.5 3 1 1.5 2 2.5 3
Fig. 25. NOx emission versus stoichiometric ratio and temperature at different O2 concentrations for CO2-based oxidizers. Experiments performed on a bituminous coal in an
entrained-flow reactor with a coal flow rate of up to 180 g/h. The NOx concentrations in ppm are given as measured. Data taken from [96].
O +N O + CO 1600
2 2 2 2
5000
Gas temp. ( C)
1400
4000
1200
NO (ppm)
3000
1000
x
2000
800
1000
1000
NO (ppm)
0 800
12
x
600
NOx (mg N / g coal fed)
400
8
200
20% O
2
HCN (ppm)
50% O
4 2 100
80% O2
Pure O2
10
0
0.5 1 1.5 2 2.5 0.5 1 1.5 2 2.5 3
λ λ
1
Fig. 26. NOx emissions as a function of stoichiometric ratio at different O2 concen- Air
1000
Oxy−fuel
trations for N2 and CO2-based inlet gases at 1273 K. Experiments performed on NH3 (ppm) 800 Oxy−fuel with O2 injection
a bituminous coal in an entrained-flow reactor with a coal flow rate of up to 180 g/h.
The NOx concentrations in ppm are given as measured. Data taken from [96]. 600
400
will mainly increase the rate of NO reduction by char [199], in fact
200
what is observed from the data in the figure.
Nozaki et al. [119] showed that air- and oxy-fuel combustion of 0
0 5 10 15 20 25
a low-volatile bituminous coal yielded similar NOx profiles in their L/D
1.2 MW burner with flue gas recirculation, even though NOx was
recirculated to the burner during oxy-fuel firing. The experiments Fig. 27. Gas temperature, NOx, HCN, and NH3 concentrations along the burner axis for
were performed with fixed firing rate and excess oxygen conditions air combustion, oxy-fuel combustion, and oxy-fuel combustion with direct oxygen
injection through the burner in the IHI 1.2 MW furnace. O2 concentration is 0% in
at the burner. Measurements of HCN and NH3 along the burner axis
primary stream, 30% in secondary stream, and thus 27% overall. The coal is low-volatile
showed significantly increased levels of NH3 in oxy-fuel combus- bituminous (Coal A) and the recycle streams are dry. Data taken from [119].
tion compared to air-firing, see Fig. 27. The authors concluded that
NOx recycled with the flue gas is reduced primarily to NH3 and N2 in
the early part of the flame. Experiments with direct oxygen injec-
700
tion in the burner revealed a significant increase in the HCN
Air
concentration in the flame zone. Nozaki et al. [119] assume this to
28 % O2/CO2
be caused by increased devolatilization arising from a higher flame 600
NO emission rate (mg/MJ)
35 % O2/CO2
temperature compared to the oxy-fuel case without direct oxygen
injection. Air-firing and oxy-fuel combustion without oxygen 28 % O2, dry
500
injection yielded similar HCN profiles even though the flame 35 % O , dry
2
temperatures were significantly different. Based on the described 42 % O2, dry
400
observation, oxy-fuel combustion leads to higher concentrations of
the intermediate cyano and amine species in the NOx formation
mechanism in the flame zone compared to air-firing. At the same 300
time, oxy-fuel combustion yields similar concentrations of NOx in
the exhaust gas compared to air-firing and thus reduced NOx 200
x
emission rates due to the lower volume of flue gas.

100
3.4.1.5. Effect of flue gas composition. Most laboratory-scale
experiments on oxy-fuel combustion are performed in once-
through reactors, where the recirculated flue gas is simulated by
0
0.5 1 1.5 2 2.5 3 3.5 4
pure CO2. It is relevant to compare the results of once-through Axial distance from burner (m )
experiments to those with flue gas recirculation to clarify
whether there are significant differences which should be Fig. 28. Centre-line axial NOx concentration (shown as the emission rate) for a US
accounted for. eastern bituminous coal burned in the CANMET 0.3 MWth vertical combustor facility at
The experiments reported by Croiset and coworkers [97] yield a firing rate of 0.21 MW with 5 vol% excess oxygen in the dry flue gas. Comparison of
air combustion as well as oxy-fuel experiments with flue gas recycle (open symbols)
a 40e50% decrease in NOx emission rates when flue gas is recycled, and once-through runs (O2/CO2, filled symbols). Oxy-fuel experiments with flue gas
compared to once-through runs, see Fig. 28. The experiments were recycle are provided for the case with a dry recycle stream (O2, dry). The recirculation
performed in the CANMET 0.3 MWth vertical combustor facility at ratios are 0.63 (28% O2), 0.51 (35% O2), and 0.40 (42% O2). Data taken from [97].
a firing rate of 0.21 MW with 5 vol% excess oxygen in the dry flue gas. concentrations are markedly higher for the dry recycle compared to
Oxygen concentrations in the inlet of 28, 35, and 42% were tested. the wet. This is likewise attributable to the differences in temper-
The results reported in the literature generally show that ature profiles and hence the effect on the homogeneous gas phase
combustion in air yields the highest NOx emissions, oxy-fuel chemistry dominating NOx formation in the flame zone.
combustion based on synthetic gas mixtures (CO2 þ O2) yields Initial NOx Concentration: As indicated in Section 3.4.1.2 it is
lower emission rates at comparable conditions, whereas oxy-fuel likely that high NOx concentrations at the onset of both devolati-
combustion with recirculation of flue gas yields the lowest emis- lization and char combustion will limit the formation of further NO
sion rates. in the flame.
Wet versus Dry Recycle: Payne et al. [86] observed that the Liu et al. [162] determined the difference in reduction efficiency
reduction efficiency of NOx through reburning in a 3 MWth oxy-fuel for experiments on a bituminous coal with 500 and 1000 ppmv NOx
sub-bituminous coal flame appeared to be higher when a wet in the oxidant O2/CO2 mixture in their 20 kW swirl-stabilized once-
instead of a dry recycle was applied. The reduction was 80% and through burner setup. They saw that the change in reduction effi-
70%, respectively, compared to results obtained for air-blown ciency between the two levels of NOx was less than 1% and thus
combustion. During the experiments the flue gas oxygen concen- within the experimental uncertainty. For their specific operating
tration and the firing rate were fixed. conditions, the NOx concentration in the oxidant was thus not
Nozaki et al. [119] likewise investigated the effect of wet and dry a limiting parameter in the reduction process.
flue gas recycle for the primary flow, i.e. the flow for coal trans- Based on experiments in air, Spinti and Pershing [185] have
portation, in their 1.2 MWth burner. The experiments were shown that the apparent conversion of char-N to NOx is a strong
performed with fixed firing rate and excess oxygen at the burner. function of the initial NOx level at the onset of char combustion in
Fig. 29 shows measurements of gas phase temperature as well as the range of 0e900 ppm NOx on a dry basis. Likewise, the excess
NOx, HCN, and NH3 concentrations along the burner axis. The NOx oxygen level affected the conversion rate, see Fig. 30. To remove
concentration in both experiments peaks at the same level, the effect of volatile-N to NOx conversion during the experiments
although the formation of NOx is shifted further down the reactor in the authors burned premade chars in methane. The char particles
the case of a wet, primary recycle. This can be attributed to the were withdrawn from a pulverized-coal flame just after devola-
significant reduction in flame temperature in the first measure- tilization was terminated. For the conditions tested, increasing the
ment point. At the same time, both the HCN and NH3 initial NOx concentration had a greater effect on the reduction of
the apparent conversion of char-N to NOx than decreasing the
oxygen excess level. In air combustion which is the case for the
described experiments, a high initial NOx level translates to a high
1600 conversion of volatile-N to NOx. This will have a negative impact
on the process through an increase of the overall NOx emission
Gas temp. ( C)
1400 rate. In oxy-fuel combustion a high initial NOx concentration at the

base of the flame will most likely almost entirely originate from
1200
the recirculated flue gas. The apparent conversion of char-N to NOx
1000 during oxy-fuel combustion could thus decrease as the concen-
tration of NOx in the oxidant increases, e.g. due to a reduced
800 recycle ratio.
Fig. 31 shows computations of the total NOx emission as
1000
NO (ppmv)
a function of the conversion of volatile-N to NOx and the fraction of

800 coal-N released during devolatilization in air. It is seen that the
lowest overall emission rate is achieved when the volatile-N to NOx
600
x
400
Apparent Conversion of Char−N to NO (%)
200 60
2 % O2
x
HCN (ppm)
100 50 3 % O2
5 % O2
40 7 % O2
10
30
1
Wet recycle
1000 20
Dry recycle
NH (ppm)
800
600 10
3
400
0
200
0 −10
0 5 10 15 20 25 0 200 400 600 800 1000
L/D
Initial NOx (ppm, dry)
Fig. 29. Effect of wet versus dry recycle of primary gas on gas temperature, NOx, HCN,
and NH3 concentrations along the burner axis in the IHI 1.2 MW furnace. The setup is Fig. 30. Comparison of the effects of excess O2 and of initial gas phase NOx concen-
operated in oxy-fuel mode with direct O2 injection through the burner and with a low- tration on the conversion of char-N to NOx for an Illinois #6 coal char combusted in air
volatile bituminous coal (Coal A). Data taken from [119]. in a 29 kW furnace. Data taken from [185].
1200
Air
500 28 % O2, 100 % purity
NO emission rate (mg/MJ)

28 % O2, 90 % purity
1000 35 % O2, 100 % purity
Total NO (ppm)
35 % O2, 90 % purity
400
x
800
300
x
600
50 % of original N in char
40 % of original N in char 200
30 % of original N in char
400
0 0.2 0.4 0.6 0.8 1 0 0.5 1 1.5 2 2.5 3 3.5 4
Volatile−N to NO conversion Axial distance from burner (m )
x
Fig. 32. Centre-line axial NOx concentration (shown as the emission rate) for a US
Fig. 31. Estimated NOx emissions from Illinois #6 coal combusted in air. Results are
eastern bituminous coal. Comparison of air combustion and once-through (O2/CO2)
based on modelling. Data taken from [185].
oxy-fuel experiments with different oxygen purities in the CANMET 0.3 MWth vertical
combustor facility. The experiments were performed at a fixed firing rate of 0.21 MW
and with an excess oxygen level corresponding to 5 vol% O2 in the dried flue gas. Data
conversion is low and as much of the coal-N content as possible is
taken from [120].
released with the volatiles. The governing assumption applied in
the modelling was that char-N to NOx conversion decreases with
increasing initial NOx level during combustion. markedly reduces the reduction efficiency, down to the range of
Based on the observations above it seems that a high concen- 44e54% [162] as also shown by Scheffknecht and coworkers
tration of NOx in the recirculated flue gas could increase the [144,148,164] in comparable experiments.
reduction rate of NOx in the flame and thus lead to a further For a commercial oxy-fuel plant NOx would be present in all
reduction of the emission rate. oxidant streams. The use of over-fire air would then potentially
decrease the rate of reduction of NOx. On the other hand, a signifi-
3.4.1.6. Effect of oxygen purity and air penetration. Croiset et al. cant fraction of the NOx present in the recycled flue gas will be
[120] have determined the NOx emission rates from air-firing and
once-through oxy-fuel combustion experiments of a US eastern
bituminous coal with two different oxygen purities, 90 and 100%
(remaining is N2). The experiments were performed in the
CANMET 0.3 MW combustion facility at a firing rate of 0.21 MW
and with temperature profiles as shown in Fig. 11. As expected,
the NOx emission rate increases with decreasing oxygen purity,
see Fig. 32. The difference between data obtained with an oxygen
purity of 90% and 100% in the experiments with an oxygen
concentration of 28% versus those obtained with 35% oxygen at
the inlet is quite significant. There is, however, no obvious
explanation to the deviations.
The data indicate that a less clean inlet O2/CO2 mixture
(increased N2 content) or air entrainment near the burners may
give rise to increased NOx emission due to thermal NO formation.
3.4.1.7. Influence of oxidant staging. The use of oxidant staging in

oxy-fuel combustion as a primary measure to reduce NOx emissions
introduces a necessity for optimization due to the fact that the
oxidant stream will contain NOx.
Liu et al. [121,162] have investigated the effect of staging on the
reduction of NOx emissions from both air and oxy-fuel combustion Fig. 33. Effects of combustion medium, staging position, and oxidant stream chosen
in a once-through reactor. The authors focused on the effect of for NO recirculation (doping) on the reduction efficiency of recycled NO. Experiments
staging and NOx recycling position on the reduction of recycled are performed on a bituminous coal, Polish Blend, in a 20 kW swirl-stabilized once-
through burner setup. Five percent of NO in N2 was added and mixed to the different
NOx. Fig. 33 shows their results for the bituminous coal, Polish
streams of oxidants to simulate the NOx recycle. The concentration of the simulated
Blend. It is obvious that both the combustion media, air or oxy-fuel, recycled NO was in the range of 500e700 ppmv. 1st, 2nd, and 3rd refer to primary,
the operating mode, with or without staging, and the NO recycling secondary, and tertiary oxidant streams. The recycling locations (injection of the
location influences the reduction efficiency. Focusing on the posi- tertiary oxidant stream), Level 1 and 2, are 570 and 880 mm downstream of the burner,
tion where NOx is recycled there is no significant difference respectively. CM1: 1st & 2nd e Air, l ¼ lburner ¼ 1.20; CM2: 1st & 2nd e Air, l ¼ 1.20,
lburner ¼ 0.80, Level 1; CM3: 1st & 2nd e Air, l ¼ 1.20, lburner ¼ 0.80, Level 2; CM4: 1st
between the primary and secondary oxidant streams, which both & 2nd e 30% O2/70% CO2, l ¼ lburner ¼ 1.20; CM5: 1st & 2nd e 30% O2/70% CO2,
are introduced through the burner. The tertiary stream constitutes l ¼ 1.20, lburner ¼ 0.80, Level 1; CM6: 1st & 2nd e 30% O2/70% CO2, l ¼ 1.20,
the staging stream. Recycling of NOx through the tertiary stream lburner ¼ 0.80, Level 2. Data taken from [162].
introduced through the burners and thus experience fuel-rich experiments by Hu et al. [160,161] for two different coals. For the
conditions for a longer period of time. The optimum degree of high-volatile coal, Prima, the figure demonstrates no significant
over-fire oxidant still remains to be determined and would, at the differences between the two species with respect to the stoichio-
same time, be dependent on the burner design. metric ratio. However, for the medium-volatile coal, Datong, the
reduction efficiency in the fuel-lean region is markedly lower for
3.4.1.8. Coal properties. The effect of coal properties on the NO2 than for NO. There are some uncertainties regarding the
formation and reduction of NOx in oxy-fuel combustion has been specific mechanism for the reduction of NO2 in the flame [160,161]
investigated by several researchers. In entrained-flow reactor type even though suggestions to direct reduction via free radicals or
experiments both Hu et al. [161] and Shaddix and Molina [169] through initial conversion to NO have been given. The results by Hu
have observed increasing NOx emission with decreasing rank of et al. [160,161] indicate that the volatile matter in the coal has
the coals. The coals investigated range from semi-anthracite over a dominant effect.
high-volatile bituminous to sub-bituminous. Shaddix and Molina Fig. 36 shows burner axis profiles of gas temperature, as well as
[169] attributed the trend to the higher volatile content, higher char NOx, HCN, and NH3 concentrations for two different coals burned in
combustion temperature, and the lower fuel-N content for the air and oxy-fuel environments (O2 concentration of 27 vol% at inlet)
lower ranking coals. in a 1.2 MWth furnace [119]. The coals differ in their volatiles
Mackrory and Tree [175] have observed that the difference content. It is obvious from the peaks in the temperature profiles
between the NOx emissions in air- and oxy-fuel combustion in their that the low-volatile coal (A) ignites further downstream than the
laminar flow reactor increases with the rank of the coal. Their medium-volatile coal (B). The gas phase temperature just outside
observations are based on experiments performed with one sub- the burner mouth (L/D ¼ 0) for the two coals thus differs by
bituminous coal and two high-volatile bituminous coals, Illinois # approximately 200 C. For each coal, the NOx concentration profile
6 and Pittsburgh # 8. is very similar regardless of the combustion environment. It must
Seven different bituminous coals were investigated in a 20 kW thus be the coal properties determining the NOx formation, with
swirl-stabilized once-through burner setup by Liu et al. [162]. The the low-volatile coal producing more NOx. This observation is
authors measured the coal-N to NO conversion ratio between contrary to the results presented above. It is likewise seen that the
oxy-fuel combustion with 30% O2 in CO2 and in air. Their data coal volatile content has an impact on the HCN and NH3 profiles.
showed no clear correlation between the conversion ratio and the They report that the coal with the higher volatiles content forms
coal rank, volatile matter, nitrogen content, etc. Instead, as shown significantly more HCN during combustion than the low-volatile
in Fig. 34, there seems to be a dependence of the char reactivity. coal which on the other hand produces a very high concentration
Increasing reactivity leads to a relatively larger conversion of coal-N of ammonia in the early flame zone during oxy-fuel combustion.
to NO at oxy-fuel conditions. This could be attributed to the The limited investigations of the influence of coal properties on
observation of Liu et al. [162] that reactive coals produced slightly NOx formation during oxy-fuel combustion indicate that the rela-
higher temperatures in the combustion zone. According to Liu et al. tive production of NOx at oxy-fuel conditions compared to when
this will promote NOx formation at the fuel-lean conditions at combustion is performed in air increases with increasing reactivity
which the tests have been run. For the less reactive coals more of the coal char. The effect could be caused by a greater sensitivity
unburnt char exists in the flame zone which may promote reduc- to char combustion temperature of the NOx formation mechanism
tion of the previously formed NOx. in an oxy-fuel environment than in air.
A large portion of the recycled NOx in an oxy-fuel plant could be
NO2 due to oxidation of NO in the downstream processes [160,161].
Fig. 35 shows the results of interchanging NO for NO2 in the
synthetic, recycled flue gas applied in the entrained-flow reactor 1
Prima
0.8
Reduction ratio
0.6
between oxy−fuel and air combustion
1
Ratio of coal−N to NO conversion
UK2 0.4
0.9
COL
0.2
US
0.8 0
UK1
Datong
0.8
Reduction ratio
0.7
AUS
0.6
0.6 POL
0.4
NO
0.5 0.2
SA NO2
0
0.4 0.5 1 1.5 2 2.5
92 93 94 95 96 97 98 99
Burnout efficiency in air (%)
Fig. 35. Reduction ratios of recycled NO and recycled NO2 as a function of stoichi-
Fig. 34. The dependence of the ratio of coal-N to NOx conversion between coal ometry obtained in an entrained-flow reactor. The reduction ratio at a given stoichi-
combustion in 30% O2/70% CO2 and coal combustion in air on the coal burnout effi- ometry is obtained from the difference in NOx emission between an experiment with
ciency in air for seven different bituminous coals. Experiments are performed in and without recirculation of NOx. The operating conditions are: recycling ratio of 0.4;
a 20 kW swirl-stabilized once-through burner setup with an overall stoichiometric T ¼ 1373 K; fixed residence time of 2 s; concentration of NOx in the inlet gas of
value of 1.2. SA: South African, AUS: Australian, POL: Polish Blend, COL: Colombian, 1000 ppm for recycled-NOx experiments. Prima: high-volatile bituminous Indonesian
UK1(2): UK Bituminous 1 (or 2), US: US Blend. Data taken from [162]. coal; Datong: medium-volatile bituminous Chinese coal. Data taken from [161].
1600 down to 40 mg/MJ at oxy-fuel conditions and improved the

remaining combustion characteristics as well.
Gas temp. (oC)
1400 Due to the strong influence from the near-burner flow field on
NOx emissions from swirling flames it will be difficult to transfer
1200 the results from the changes to the specific burner configurations
obtained by Chui et al. [150,151] to a general trend for other oxy-
1000 fuel burners.
Based on experiments regarding char-N to NOx conversion and
800
modelling investigations, Spinti and Pershing [185] suggest the
1000 following measures in order to obtain the largest possible reduction
NO (ppm)
in NOx emissions. (i) a burner which yields a fuel-rich flame core,

800
(ii) high temperatures during devolatilization, and (iii) a char
x
600 oxidation zone with temperatures as low as possible in combina-

tion with low excess O2 and adequate residence time for char
400
burnout at the lower gas temperatures. This is not different from
200 the case of air-firing. However, because of the changes in mass flow
rates in the primary and secondary inlets, see Section 2.7.1, the
HCN (ppm)
100 optimal burner design for oxy-fuel combustion is not necessarily

similar to that developed for conventional air-firing.
10 3.4.2. SOx
Sulphur emissions and the effect on ash properties and boiler-
1 tube corrosion in oxy-fuel combustion from significantly increased
Coal A, Air levels of gas-phase-S in the boiler have received increasing atten-
1000
Coal B, Air tion in the recent years.
NH (ppm)
800 Coal A, O2/CO2 There have been contradictory observations on the SO2 emis-
600 Coal B, O2/CO2 sions from oxy-fuel combustion. Some researchers experimentally
3
400 show a decrease when comparing to combustion in air

[88,96,97,128,136,138] whereas others on the basis of either
200
experiments [121,113] or equilibrium calculations [149] report no
0 differences.
0 5 10 15 20 25
L/D Kiga et al. [136] reported oxy-fuel experiments performed in the
IHI 1.2 MWth combustion facility with recirculation of dry flue gas.
Fig. 36. Effect of coal volatiles content on gas temperature, NOx, HCN, and NH3 The investigations showed that the conversion of the coal sulphur
concentrations along the burner axis in a 1.2 MW furnace. Coal A, low-volatile bitu- content, measured as the amount of S in the outlet divided by the
minous, 26.1% volatile matter. Coal B, medium-volatile bituminous, 41.2% volatile
amount going in with the coal, decreased markedly in oxy-fuel
matter. O2 concentration is 0% in primary stream, 30% in secondary stream, and 27%
overall at inlet, l ¼ 1.2. The recirculated flue gas in the oxy-fuel experiments contains combustion compared to air-firing operation. Even though the
up to 200 ppm NOx. Data taken from [119]. authors were not able to close the mass balance for sulphur the
results indicated that the major part of the difference could be
attributed to removal of S through condensation as H2SO4 in low-
3.4.1.9. Effect of burner configuration for swirling flames. Fig. 24 temperature ducts and by retention in ash particles.
showed that exchanging a conventional burner for a low-NOx Zheng and Furimsky [149] modelled the formation of SO2 and
burner in oxy-fuel combustion yielded a 56% reduction in the NOx SO3 based on equilibrium considerations during both air- and oxy-
emission rate. Thus, NOx formation in oxy-fuel combustion fuel combustion. They found that practically all of the sulphur
depends on the burner configuration and especially the flow field in content in coals is released as SO2 and SO3 irrespectively of the
the near-burner zone, just as is the case during conventional air combustion medium, as CO2 plays an insignificant role for the
combustion. release compared to oxygen. With respect to recycle ratio and
Experiments performed by Chui et al. [150] in the CANMET oxygen excess, their simulations yielded an increased amount of
0.3 MW vertical combustor research facility show that NOx emis- SO3 with an increase in the amount of available oxygen.
sions are highly dependent on the burner swirl number, i.e. flow Generally, the disagreements observed between experimental
dynamics. Increasing the swirl number from 1 to 2 decreases the findings and the computations based on equilibrium modelling are
NOx formation by 34%. However, for the range of swirl numbers expected to be a consequence of computations reflecting condi-
applied in the experiments NOx emissions during oxy-fuel tions immediately after combustion and not accounting for changes
combustion was higher than for air-firing, 140e150 versus 110 mg/ occurring further downstream. Another plausible explanation is
MJ. This was true for both burner configurations tested. Because of the fact that some of the processes could be kinetically controlled,
safety limitations, the amount of oxygen mixed into the secondary e.g. the heterogeneous reactions controlling retention of S in ash
oxidizer stream was restricted to 28% on a dry basis. For the same and other deposits, a fact that is not captured in the simulations.
reason, the primary stream contained little or no oxygen. To obtain Table 13 provides a summary of the results on SO2 emissions
the necessary overall oxygen concentration during combustion from semi-technical and pilot scale experiments with flue gas
a portion of the total oxygen demand was injected as pure oxygen recirculation.
through a special annuli in the burners. The burners provided stable In the following sections the reported results on SO2 emissions
flames but the injection of pure oxygen caused the release of vola- from oxy-fuel combustion are divided according to the effects of
tiles and fuel-N from the Canadian sub-bituminous coal tested to different operating conditions and design aspects. The subject of
occur in the O2-rich portion of the flame. An improved burner design boiler-tube corrosion due to the presence of sulphur in the flue gas
subsequently tested by Chui et al. [151] reduced the NOx emission is likewise treated.
Table 13
Summary of SO2 emission results from experiments in pilot scale and semi-technical scale with flue-gas recirculation.
Author(s) Fuel input Inlet O2, Emission Conversion Conclusion(s)

(MW) Oxy (%) (mg/MJ) ratio
Woycenko et al. [88] 2.1 26 Air: 645 Air: 100% The conversion ratio of fuel-S into SO2 is lower for oxy-fuel combustion than air.
Oxy: 375e418 Oxy: 60e67%
Wall et al. [138,128] 1.2 27 Air: 150e416 Air: 63e83% Lower SO2 emission in oxy-fuel than in air, level dependent on coal type.
Oxy: 100e291 Oxy: 40e53% Increased SO3 concentrations increases acid dew point from w 133 to w 156 C.
Croiset and 0.21 28, 35 Air: 300e320 Air: 91% SO2 emissions independent of wet and dry recycle even though removal
coworkers Oxy: 260e275 Oxy: 56e65% through condensate was expected. Drop in conversion rate attributed to
[120,97] conversion to SO3 and/or retention in ash and deposits. Coal sulphur content
most important factor to determine SOx emission.
Tan et al. [122] 0.21 35 e e SO2 emission rates slightly lower for oxy-fuel combustion than in air even
though the concentration was 3e4 times higher in the oxy-fuel case.
Scheffknecht and 0.5 27e30 e Air: eOxy: 96e100% Negligible reduction of SO2 in the radiative section of a boiler. Higher SO2
coworkers concentrations yields higher H2S formation in fuel-rich zones. Calcium-rich
[144,145,147] coals show greater tendency to capture S.
3.4.2.1. Effects of changes in the oxygen concentration, oxygen excess, emissions from the combustion of a US eastern bituminous and
flue gas recycling ratio, and gas phase temperature. Hu et al. [96] a Canadian sub-bituminous (Highvale) coal in the 0.3 MW CANMET
focus on the effect of oxidizer composition, stoichiometry, and oxy-fuel combustion facility. Fig. 38 shows SO2 emission rates for
flame temperature in their investigations of SO2 emissions in an air-fired experiments, once-through experiments (synthetic flue
entrained-flow reactor with a coal feed rate up to 180 g/h. They gas denoted O2/CO2), and dry flue gas recycle (O2, dry) experiments.
encountered the highest conversion ratios at or near stoichiometric The data for the bituminous coal in the figure show similar emis-
conditions, see Fig. 37. The decrease in S conversion at fuel-rich sion rates for experiments with air and the synthetic O2/CO2
conditions is assumed to be due to retention in unburned coal and mixture. For the recycle experiments with 28 and 35 vol% O2,
formation of other, reduced S-containing species (e.g. H2S, COS, and however, the emission rates were slightly lower.
CS2). These species were not measured. The reduction in SO2 Comparing gas phase concentrations instead of emission rates
emissions at fuel-lean conditions compared to stoichiometric reveal that the measured SO2 concentrations in the burner during
conditions has not been explained but could partly be due to oxy-fuel combustion with a recycle fraction of about 60 vol% were
increased conversion to SO3. approximately 3 times higher than for air-firing [97]. The theoretical
According to Liu et al. [121] SO2 emissions are not significantly increase in the SO2 concentration was about a factor of 4. The authors
affected by the use of staging or changes in the combustion media. concluded that the conversion of sulphur into SO2 was independent
of oxygen concentration and that retention in ash and/or
3.4.2.2. Effect of the composition of the recirculated flue gas. Croiset further oxidation into SO3 could explain the difference. Ash analysis
and coworkers [120,97] reported experiments regarding SO2 showed that less than 3% of the initial S-content in the coal was
Air 50 % O2 / 50 % CO2 100 % O2
5000 1123 K
1273 K
1423 K
4000 1573 K
SO2 (ppm)
3000
2000
1000
0
12
SO2 (mg S / g coal fed)
0.5 1 1.5 2 2.5 3 1 1.5 2 2.5 3 1 1.5 2 2.5 3
Fig. 37. SO2 emissions versus stoichiometric ratio and temperature at different O2 concentrations for N2 and CO2-based oxidizers. Experiments are performed in an entrained-flow
reactor with a coal feed rate of up to 180 g/h. The SO2 concentrations in ppm are given as measured. Data taken from [96].
400
400 Air
28 % O2/CO2
SO emission rate (mg/MJ)
35 % O2/CO2
SO emission rate (mg/MJ)

28 % O2, dry 300
350 35 % O2, dry Air East. bit
42 % O2, dry Air Highvale
28 % O East. bit
2
200 28 % O2 Highvale
35 % O East. bit
2
300
35 % O Highvale
2
2
2
100
250
0.5 1 1.5 2 2.5 3 3.5 4 0
0.5 1 1.5 2 2.5 3 3.5 4
Axial distance from burner (m ) Axial distance from burner (m )
Fig. 38. Centre-line axial SO2 concentration (shown as the emission rate) for an
Fig. 39. Centre-line axial SO2 emission rate. Comparison of Highvale sub-bituminous
Eastern bituminous coal burned in the CANMET 0.3 MW vertical combustor research
and US eastern bituminous coals. The experimental setup is the CANMET 0.3 MW
facility at a firing rate of 0.21 MW with 5 vol% excess oxygen in the dry flue gas.
vertical combustor research facility operated in once-through mode at a firing rate of
Comparison of air combustion as well as oxy-fuel experiments with flue gas recycle
0.21 MW with 5 vol% excess oxygen in the dry flue gas. Data taken from [120].
(open symbols) and once-through runs (O2/CO2, filled symbols). Oxy-fuel experiments
with flue gas recycle are provided for the case with a dry recycle stream (O2, dry). The
recirculation ratios are 0.63 (28% O2), 0.51 (35% O2), and 0.40 (42% O2). Data taken
from [97]. Because of the potentially higher SO3 concentrations in the flue
gas, the acid dew point will increase accordingly
[115,138,145,147,148]. Increases of about 20e40 C from w130 to
present in the ash. However, the condensed water did have w160 C have been reported [113,138,145]. Schnurrer et al. [115]
a sulphate content of above 3000 mg/L and a pH between 2 and 3. found a similar trend from their thermo-chemical equilibrium
The high sulphate concentration in the condensate indicates high calculations. An earlier formation of sulfuric acid in the flue gas
SO3 levels in the flue gas. In the work by Mönckert et al. [147] the ducts will increase the risk of flue gas side corrosion in the low-
sulphur retention in calcium containing coals is shown to increase temperature parts considerably. Schnurrer et al. suggest using
with increasing SO2 concentration in the recycled flue gas. This is a dry flue gas recycle to decrease the concentration of water in the
due to the formation of CaSO4. flue gas and thus the amount of acid that can form.
Experimental runs with wet flue gas recycle were likewise Scheffknecht and coworkers [144,145,147,148] investigated the
performed by Croiset and Thambimuthu [97] but showed no impact of recirculation of SO2 (by doping the oxidant stream in
significant difference to the dry case. The authors, however, their 20 kW EFR) on the transformation of SO2 in the radiant
expected a lower emission rate for the latter case since some SO2 section of a boiler. They found only a negligible reduction in the
should be removed with the waste water during the flue gas concentration of the recycled SO2. However, high concentrations of
condensation. SO2 in the furnace was found to increase the concentration of H2S in
the fuel-rich regions for both air- and oxy-fuel firing conditions, see
3.4.2.3. Effect of coal sulphur content. Comparing the data in Fig. 39 Fig. 40, inducing a higher risk of corrosion. The H2S to SO2 ratio in
leads to the conclusion that the specific SO2 emissions are almost the oxy-fuel experiments were lower than in air even though the
exclusively dependent on the coal type, i.e. sulphur content. The absolute concentration was higher.
Eastern bituminous coal has a sulphur content of 0.96 wt% (dry
basis) whereas the sub-bituminous coal (Highvale) contains 3.4.2.5. In-boiler desulphurization. In a series of papers Liu and
0.24 wt% (dry basis) [120]. The relative differences due to changes coworkers [127,208,209] address the ability of oxy-fuel combustion
in combustion medium and oxygen concentration are thus negli-
gible in this context. 1400
Air 0 ppm
1200
Air 3000 ppm
3.4.2.4. The SO2 to SO3 conversion and sulphur-induced corrosion. In
1000 Oxy 0 ppm
H S (ppm)
later work from CANMET, Tan et al. [122] reported further results on
Oxy 3000 ppm
the conversion of SO2 to SO3 in the flue gas. They showed that the 800
conversion was about 5% whereas it is typically between 1 and 5% 600
2
in conventional air-firing systems [113,122,147] depending on the 400

combustion conditions and the sulphur content of the coal. These
values have likewise been found by Scheffknecht and coworkers 200
[145]. Klostermann [113] and Mönckert et al. [147] report that SO3 0
formation is promoted by both high oxygen and high water levels in 0 0.5 1 1.5 2 2.5 3
the flue gas. In the case of a flue gas recycle without prior SO2 Distance from Burner (m)
removal the amount of SO3 in the boiler can thus reach high values
Fig. 40. H2S concentration profiles along burner axis for air- and oxy-fuel combustion
(up to about 85 ppmv [147]) and thus increase the risk of sulphur experiments on lignite. The experiments were performed in an 20 kW entrained-flow
corrosion in the regions of the system which operate below the acid reactor with and without doping of the oxidizer with 3000 ppm SO2. Data taken from
dew point. [144].
to drastically reduce the emissions of SO2 from combustion of coal.

They suggest in-boiler desulphurization by injection of limestone.
The combination of both high SO2 and CO2 partial pressures should
ensure high reactivity of the limestone toward SO2. The authors
explain the increased reactivity by two factors: (1) high SO2 limits
CaSO4 decomposition and (2) high CO2 limits decarbonisation of
limestone before sulphation whereby direct sulphation of lime-
stone without decarbonisation is favoured. The direct sulphation
will minimize the diffusion resistance through the solid phase (no
pore clogging) and hence a larger part of each limestone particle
will participate in the desulphurization reaction.
Even though the experiments show significantly increased
reactivity of the limestone toward SO2 for a sulphur containing flue
gas recycle compared to conventional air-fired conditions, this
strategy is not likely to be adopted in power plants which produce
fly ash for cement and concrete production due to the associated
mixing of gypsum and fly ash.
3.4.2.6. Summary. The substitution of N2 by CO2 in oxy-fuel

combustion does not affect the release of sulphur from the coal Fig. 41. Mercury concentration measured at the furnace exit of a laminar reactor.
during combustion. However, the increased oxygen partial pressure Experiments are performed with a sub-bituminous coal in air and for two O2/CO2
mixtures with a fixed reactor temperature of 1200 C. The total Hg concentration is
necessary to maintain an appropriate flame temperature increases divided into the fraction consisting of metallic (elemental) Hg and the oxidized forms.
the formation rate of SO3. During operation with a flue gas recycle Data taken from [117].
without SO2 removal the SO2 and thus the SO3 levels in the boiler
and flue gas ducts increase significantly which will enhance the combustion atmosphere and the potential change in local particle
risk of sulphur-induced corrosion at both high and low temperature between oxy-fuel combustion and air-fired combus-
temperatures. High in-boiler concentrations of sulphur oxides can tion may thus have an effect on the ash formation mechanisms and
likewise increase retention of S in the fly ash. Increased retention hence the ash composition and quality.
will reduce the SO2 emission rate but could at the same time yield
problems with further utilization of the fly ash in cement and 3.5.1. Particle formation mechanisms
concrete production. Fig. 42 shows the mineral transformation and particle formation
pathways for both fine (sub-micrometer, mode around 0.1 mm) and
3.4.3. Trace elements coarse particles during coal combustion. There are two pathways
The subject of trace element emissions during oxy-fuel for the formation of sub-micrometer particles; (1) direct vapor-
combustion is generally seen to have drawn minor attention ization of volatile metals, e.g. Na, Pb, Cd, and Hg, which react in the
compared to other of the fundamental combustion issues gas phase and subsequently nucleate or condense on the surface of
investigated. Because of the scarcity of published results within this existing particles; (2) non-volatile species like e.g. silica oxides can
area of research none of the results described below have been be reduced to sub-oxides, e.g. as in reaction (14). These sub-oxides
confirmed by other research groups. have a lower melting point and are devolatilized and rapidly
According to the equilibrium calculations performed by Zheng reoxidized in the gas phase causing an oversaturated mixture.
and Furimsky [149] the Hg-, Cd-, As-, and Se-containing emissions
MeOx þ CO#MeOx1 þ CO2 (14)
are only insignificantly affected by the combustion medium.
However, in case of incomplete elimination of these species in the Particle formation occurs mainly through nucleation whereas
flue gas cleaning equipment situated before the flue gas recycle growth is dominated by condensation and collision mechanisms.
point elevated concentrations will occur in the boiler and thus in The sub-micrometer particle formation mechanisms are complex
the flue gas stream to be treated before sequestration. functions of such factors as coal type, combustion temperature, fuel
Suriyawong et al. [117] measured the mercury emission from to oxidizer ratio, and residence time. This is likewise the case for the
air- and oxy-fuel combustion of a sub-bituminous coal in a laminar resulting particle size distributions [117].
reactor and determined the ratio between the elemental and Coarse ash particles are formed from the non-volatile mineral
oxidized forms. Fig. 41 shows their results. inclusions in the coal which are not released as sub-oxides. These
The data lie within typically agreed intervals of sub-bituminous inclusions can coalesce and form a glassy matrix or remain in their
coals, i.e. 10e20% of the total mercury emissions are in the oxidized original state if the temperature does not exceed their melting
states with the remainder in the elemental state. The concentra- points [117].
tions measured are within the range of uncertainty of the mercury The main fraction of the fines will consist of spherical particles due
content in the coals which indicate that burning coal at oxy-fuel to their origin from gas phase species. On the other hand, the shapes
conditions does not have a significant effect on neither the vapor- of the coarse particles may vary from spherical or near-spherical to
ization nor the speciation at the furnace exit. irregular, depending mainly on the char particle temperatures during
At CANMET, ongoing research addresses mercury removal combustion, i.e. whether they have been melted.
technologies for oxy-fuel combustion [9].
3.5.2. The effects of gas composition on particle formation
3.5. Ash and deposition chemistry Combustion at oxy-fuel conditions with an air-like composition
of the oxidant will result in a lower adiabatic flame temperature
The transformations of coal mineral matter during combustion and hence a reduction of the coal burning rate compared to
are significantly affected by the temperature and the gas phase combustion in air, see Sections 2.7, 3.2, and 3.3. The vaporization of
composition surrounding the coal particles [125]. The change in both volatile metals and metal sub-oxides as well as the particle
Fig. 42. Mineral transformation and particle formation pathways during coal combustion [117].
formation rates will thus be significantly smaller, as observed by air and oxy-fuel combustion with 20% O2 in CO2 might alter the
Suriyawong et al. [117] and Sheng and coworkers [125,182,183]. particle fragmentation and coalescence mechanisms in addition to
From their investigations, Suriyawong et al. found that the ratio lowering the ash vaporization rate. However, the particle number
between fine and coarse particles shifted significantly toward fewer concentration in the less than 50 nm range does not change.
sub-micrometer particles in this particular oxy-fuel case. The mean
size of the fine particles likewise became smaller because the 6
particles have less time to grow when ignition and combustion is 10
delayed, see Fig. 43. The bottom part of the figure shows the Air
d(N)/d logDp (#/cm3)
20 % O2/80 % CO2
amount of various elements in the fine ash. It is generally observed
that a smaller amount of ash forming elements have vaporized 5
10
from the coal in the case of oxy-fuel conditions causing the lower
number concentration. At the same time, the ratios between the
different elements have changed. Particularly calcium and iron are 4
released to the gas phase to a much lesser extent than in air. Sur- 10
iyawong et al. [117] explain this phenomenon in the way that the
increased concentration of CO2 in the bulk gas will shift the sub-
oxides formation equilibrium toward the oxides, see reaction 3
10
(14), and thus reduce the formation of sub-micrometer particles 1 29.1 40.3 2 3
10 10 10
through this pathway. The reduced particle surface temperature is,
however, of significant importance. Sheng et al. [182] likewise Dp (nm)
determined the compositions of their ashes. They observed 800
Air
a similar shift in the ratio between volatile and less volatile species
20 % O2/80 % CO2
Element vaporized
(μg/g coal burned)
in the sub-micron ash for oxy-fuel combustion with an air-like 600

oxidant composition. Increasing the oxygen concentration dimin-
ished the difference between air and oxy-fuel combustion. Differ-
ences in the oxygen concentration and thus the combustion 400
temperature of the coal and char particles can alter the distribution
between the main phases slightly, as more or less of the included
200
minerals melt into a glass phase [183,125].
Increasing the O2 concentration in CO2 from 20 to 50% increases
the particle surface temperature during combustion from 1772 K to 0
2679 K [117]. As a result, the particle size distribution of the sub- Si Al Mg Fe Ca K Ti S Ba
micrometer-sized fraction of ash particles is shifted toward larger
Fig. 43. Top: Sub-micrometer particle size distribution from air and 20% O2/80% CO2
sizes, see Fig. 44. combustion of a sub-bituminous coal obtained in a laminar flow reactor with a fixed
Sheng et al. [183] observed the same trend for their Chinese coals. temperature of 1200 C. Bottom: Measured elemental compositions in sub-micro-
They likewise concluded that the temperature difference between meter-sized ash from the same experiments. Data taken from [117].
6 0
10 10
20 % O
2
25 % O −1
2
10
d(N)/d logDp (#/cm3)
5
10 40 % O
2
CO
50 % O2
−2 79 % CO2
10
4
10
0 % CO2
−3
10
3 79 % CO2
10
−1
10
2
CO
2
10 −2
1 2 3 10
10 10 10 0 % CO
Dp (nm) 2
−3
Fig. 44. Sub-micrometer particle size distributions for oxy-fuel combustion at 10
different O2 concentrations obtained from combustion of a sub-bituminous coal in
a laminar flow reactor with a fixed temperature of 1200 C. Data taken from [117]. 1
10
2
CO/CO
According to Suriyawong et al. [117] this is due to the fact that the 0
10 79 % CO2
nucleation and condensation mechanisms are competing. Below 0 % CO2
a critical concentration of small particles vaporized species tend to −1
10
nucleate. Above this limit condensation dominates leading to
growth in particle size instead of particle number concentration. −2
10
Comparing Figs. 43 and 44 leads to the conclusion that −4 −3 −2 −1 0
10 10 10 10 10
increasing the oxygen concentration in oxy-fuel combustion from
r (cm)
20% to between 25 and 40% will produce a particle size distribution
similar to that obtained in air for the small size fraction of ash Fig. 45. CO and CO2 partial pressures (atm) and the CO/CO2 ratio as a function of radius
particles. Experiments reported by Wall et al. [138,128] confirm this from the centre of a particle through the boundary layer. Vertical dotted line corre-
observation. sponds to the surface of the particle. The data are for a 100 mm char particle burning in
a furnace with walls at 1750 K and 0, 10, 20, 50, and 79 vol% CO2 in the bulk gas. The O2
A modelling study by Krishnamoorthy and Veranth [210]
concentration in the bulk gas is 20 vol% for all simulations. The remainder of the bulk
investigated the effect of particle size, bulk gas composition, and gas consist of 1% H2O and the difference is N2. Note that the position of the legends in
in-furnace temperature during oxy-fuel combustion on the vapor- the figure showing the CO/CO2 ratio has been corrected compared to the original figure
ization of metal sub-oxides from the burning char due to locally in [210].
reducing conditions within the particles. Fig. 45 shows the ratio
between CO and CO2 as a function of radius in a char particle
burning in atmospheres of varying CO2 concentrations. Wall et al. [138] report similar compositions and melting
The results indicate that increasing the CO2 level in the bulk gas behaviour of both fly and bottom ashes obtained from air and oxy-
significantly decreases the CO/CO2 ratio inside the particles even fuel combustion. The coal with the highest calcium content showed
though the absolute concentration of CO increases. Based on a slightly increased concentration of sulphur. During the experi-
equilibrium considerations considering reaction (14) a shift from ments emphasis was reported to be on generating the same
air-blown to oxy-fuel combustion should thus impose decreased temperature and heat transfer conditions, i.e. by operating with
vaporization of refractory oxides from the coal mineral phase 27% O2 on a wet basis in the oxidizer. From their experiments Sheng
during combustion. and Li [125] reported a small increase in the content of limestone
due to the higher partial pressure of CO2 at oxy-fuel conditions
3.5.3. Ash quality compared to air-firing.
To the power plants which sell their coal fly ash for cement or For a flue gas recycle without prior sulphur removal the sulphur
concrete production the quality of the ash in this respect is crucial. level in the boiler could increase with a factor of up to 4e5. This
Especially the sulphur content of the ash is a critical parameter, the significant increase will induce a risk of enhanced sulphur retention
limiting value for S is 3 wt% measured as SO3, as described in in the fly ash [12,115,128,145]. Lowering the flame temperature in
Section 2.5.1.2. Significantly increased sulphur retention compared the boiler will also increase the possibility of SO2 retention in ash
to the level which is normal in air-fired operation will make the fly [12]. Maier et al. [145] state that retention of sulphur in ash does not
ash useless as a substitute for cement in concrete production. This is take place in the radiant section of a boiler due to the temperature
due to sulphates reacting with the other components in the being above the 1150 C where sulphate salts are stable. However,
concrete matrix after hardening. The result is an expansion of the in the convective section the temperature will be lower and the ash
mineral phase and thus fatal cracks in the structure. will be able to capture sulphur compounds.
The results published by Farley [72] indicate that the fly ash Experiments by Croiset et al. [97] on a US eastern bituminous
from oxy-fuel combustion of coal should be “equally acceptable for coal have shown that the retention is generally low, i.e. only about
cement manufacturing as those arising from conventional air- 2e3% of the initial S-content in the coal is present in the ash. A later
firing”. However, the paper provides no reference to the specific investigation showed that the retention of sulphur in the fly ash is
restrictions on the fly ash composition applying in the UK by which very dependent on the alkaline and alkaline earth content of the
the above statement cannot be assumed applicable to the regula- ash [122]. For a sub-bituminous coal with a highly alkaline ash they
tions and conditions in other countries without further verification. observed that nearly 14% of the initial sulphur content of the coal
was retained in the ash. It is suggested that it is the SO3 in the flue oxidation of iron-bearing species can be seen for relatively low char
gas that is retained through sulphate salts. combustion temperatures and high CO concentrations inside the
Based on the above, a possible solution could be to restrict the burning particles. However, no direct comparison between air- and
power plants to operate on only low alkaline and/or low sulphur oxy-firing at the same heat transfer rates with respect to slagging has
bituminous and sub-bituminous coals and thereby limit the been found. It is thus unknown, whether incompletely oxidized iron
changes in the sulphur content of the fly ash from oxy-fuel will induce a problem in a full-scale boiler.
combustion. This is however not a desirable restriction for the Deposit sampling tests performed by Mönckert et al. [147]
plant owners as it will reduce the operability of the plants. indicate that besides sulphation, carbonization of deposit surfaces
The question of ash quality, especially the risk of increased occur. The most obvious reason is that the decomposition
sulphur retention, is still a potential risk for the application of oxy- temperature of carbonates will increase due to the high partial
fuel combustion as a carbon capture technology. The risk is largest pressure of carbon dioxide [4]. However, the implications of this
for the most frequently proposed process design where flue gas is observation is not clear.
recycled before the desulphurization plant. The investigations reported till now indicate that the changes to
depositions in an oxy-fuel plant compared to an air-fired unit will
3.5.4. Depositions, slagging, and fouling not be of essential significance to the plant operation.
Early studies on ash deposition, slagging, and fouling at the ANL
3 MWth pilot scale furnace referred by Payne et al. [86] showed that 3.6. Oxy-fuel combustion of biomass
there were no identifiable differences between air-firing and oxy-
fuel combustion. However, Glarborg and Bentzen [172] state that Carbon capture and storage from combustion of biomass has the
the high CO levels in the near-burner region could promote potential to reduce the CO2 emission to below zero, i.e. to extract
increased corrosion and slagging. Similarly, Schnurrer et al. [115] CO2 from the atmosphere and possibly limit the anticipated global
saw from their equilibrium calculations that the formation of warming [84,157,180,211e214]. The abbreviation BECS is used to
deposits potentially could occur at furnace wall temperatures denote the concept of Biomass Energy for Carbon Capture and
30e60 C higher than what is seen during conventional combus- Sequestration [213e215]. In principle, all CCS technologies sug-
tion in air. This may alter the position of deposits and thus poten- gested for fossil fired power systems can be applied to systems
tially necessitate a relocation of soot blowers, etc. At the same time, utilizing biomass fuels [216].
they expect oxy-fuel combustion to provide a higher propensity for This section contains a survey of the very limited amount of
slagging and fouling caused by increased amounts of molten and literature available on the use of biomass in oxy-fuel combustion.
solid alkali sulphates. The main type of biomass investigated is wood. This is a natural
Experiments conducted by Mitsui Babcock Energy, ltd on consequence of its abundance compared to other types of biomass,
a 160 kWth test facility in Renfrew, UK [72], likewise included such as straw and other annual plants, olive residues, or other
studies into the impact of oxy-fuel combustion on slagging and dedicated energy crops. Additionally, the transport and feeding of
fouling. Inserting a deposition probe into the combustion chamber wood is easier to perform than for herbaceous energy sources.
led to the observation of a slightly faster rate of deposition When utilizing biomass in oxy-fuel combustion it is important to
compared to air-firing. The temperature of the probe has not been consider whether there are any special issues regarding biomass
stated by the author. At the same time the deposition rate was compared to coal which could have a significant impact on the
largely independent of the operating parameters such as recycle combustion process and related phenomena. Of particular impor-
ratio during oxy-fuel combustion. Also, fouling occurred more tance will be the volatile matter content, the ash composition, i.e. the
rapidly during oxy-fuel than during conventional air combustion. content of Cl and alkali, and the change in fuel particle size. At
The impact was, however, small and the deposits were easy to the same time, in order to conclude on specific aspects related to the
remove. change of combustion atmosphere when co-firing coal and biomass
Wall et al. [138,128] collected ash samples from the deposits in it is necessary to compare the reported results with the behaviour of
the radiative and convective sections of the IHI 1.2 MW test facility. these types of fuel blends in conventional combustion.
SEM images showed no significant difference between samples
from air and oxy-fuel combustion. The furnace deposits, however, 3.6.1. The combustion process
contained significantly more sulphur compared to samples Arias et al. [180] have investigated the ignition behaviour of
collected from air combustion. The deposition rates for the different coal/biomass blends during oxy-fuel combustion and the effect of
combustion conditions were different. In the convective section oxy-fuel conditions on burnout. Their experiments are performed
(fouling) the coals tested generally showed increased fouling in an electrically heated entrained-flow reactor (EFR) at about
tendency during oxy-fuel combustion compared to air-firing. The 500 C for ignition tests and 1000 C for combustion tests. Both
slagging propensity of the different coals tested changed from fuels are ground and sieved to a particle size of 75e150 mm. Oxy-
being higher to less than that in air combustion. fuel combustion with 21, 30, and 35 vol% oxygen were compared
The composition of the deposits from both the radiative and with results obtained in air, see Fig. 46. As has previously been seen
convective sections showed no difference between air and oxy-fuel for pure coals, a delay in ignition (higher ignition temperature) is
conditions, except for the SO3 content which was higher (0.65 observed as the oxygen concentration in the oxy-fuel oxidizer is
(oxy) versus 0.2 (air) wt%) during oxy-fuel combustion for the coal reduced. From these experiments the relative effect of oxygen
with the highest sulphur content (0.88 wt%, dry). This particular coal concentration is nearly independent of the type of fuel.
does, at the same time, have the highest content of alkali and alkali In air there is a reduction in the ignition temperature when the
earth metals in the ash prior to combustion. This could be the bituminous coal is blended with the biomass, see the three most
explanation to the increased sulphur retention, see Section 3.5.3. left-hand-side bars in the figure for each case. The effect is,
A lower conversion of the iron species in the parent coal into however, much less pronounced in the case of oxy-fuel combustion
oxides can increase the amount of iron melting into glass silicates regardless of the oxygen concentration. This can be attributed to
and thus increase the slagging propensity of the coal. Untransformed the fact that even though the biomass has a high reactivity and
FeO-FeS phase with an eutectic temperature of 940 C will likewise a high volatile matter content the heating value is lower than for
increase the slagging propensity of the ash. This effect of reduced the coal. The heat released from the biomass during ignition is thus
profiles in the reactor caused by the introduction of the more reac-

tive biomass fuel.
Characteristics of wood chip, rice husk, and forest residues chars
obtained by pyrolysis in air and oxy-fuel atmospheres were
investigated by Borrego et al. [217]. They saw that the char char-
acteristics such as pore volume, morphology, optical texture,
specific surface area, and reactivity showed no significant differ-
ence between air and oxy-fuel combustion. The authors thus
concluded that the application of biomass in co-fired oxy-fuel
boilers (coal/biomass) should not constitute any specific difficulties
regarding the chars.
3.6.2. Emissions
In commercial-scale tests co-firing coal and biomass in air, large
decreases and moderate increases are occasionally observed for the
Fig. 46. Ignition temperatures of a high-volatile bituminous coal (DA) and blends of
NOx yields [218]. Tests by Robinson et al. [218] have shown that
the coal and Eucalyptus (RE) in air combustion and oxy-fuel combustion with different these changes most probably are caused by the biomass influencing
O2 concentrations at the inlet. The blending ratio is presumably given on a mass basis. fluid dynamics as well as temperature and stoichiometry profiles
Experiments are performed in an electrically heated entrained-flow reactor. Data taken during the combustion. No fundamental synergistic chemical
from [180].
effects could be observed. The coals tested were bituminous
whereas the biomass were red oak wood chips and switchgrass.
not substantial to enhance the heating of the coal particles and the Fryda et al. [184] have found that the NOx emission rate during
ignition temperature is thus not increased significantly. As previ- oxy-fuel combustion can be decreased when coal (Russian) and
ously described, more heat is needed to increase the temperature of biomass in the form of cocoa residues are co-fired in a lab-scale
the surrounding gases and hence the coal particles in the case of setup compared to tests with pure coal. Generally, using over-fire
oxy-fuel combustion than for combustion in air [180]. air (OFA) reduces the emission rate. The cocoa residues have
Results of burnout experiments for the same high-volatile coal a higher content of N compared to the pure coal (2.61 versus
and its blends with Eucalyptus are shown in Fig. 47. An obvious 0.87 wt%, dry basis) and thus has the fuel blend. The decrease in
difference in burnout degree is seen for all three cases when NOx emission rate can thus be explained by the increased volatile
changing the combustion atmosphere from air to CO2-based. In content and the related, lower amount of fixed carbon in the blend
oxy-fuel combustion with an air-like oxidizer composition the compared to the pure coal.
burnout degree decreases below that found in air. For oxygen Emissions of SOx from oxy-fuel combustion of pure biomass or
concentrations of 30 and 35% the burnout increases above that coal and biomass blends have not been reported in open literature.
found in air. However, the effect of blending biomass and coal is
seen to have only little impact on the burnout, i.e. there is only 3.6.3. Ash and corrosion
a minor improvement by an increasing biomass content in Robinson et al. [218] have investigated the ash deposition rates
the blend. from both individual fuels and their blends (wood chips, coal,
Because of the significantly higher volatile content in biomass switchgrass, and wheat straw) during conventional combustion in air.
compared to coal an improvement in the burnout degree of the fuel They observed that the particle capture efficiency (the ash deposition
blends could be expected. However, reported data do not confirm rate normalized by the fuel ash content and the size of the deposition
this. Arias et al. [180] explain the lack of improvement in burnout for probe) was significantly larger for straw than for wood chips and coal
the blends in terms of the changes in oxygen and temperature in the order: wood chips < coal < switchgrass wheat straw. The
order is closely related to the absolute amount of alkali in the fuels.
Blending coal and biomass, especially straw, yield deposition rates
which are lower than what would be expected based on the behaviour
of the unblended fuels, see the upper part of Fig. 48. Especially in the
case of co-firing coal and straw the particle capture efficiency is
markedly reduced compared to the predicted value (about 10%).
Sulphation of alkali chlorides released from the straw with sulphur
from the coal is expected to account for part of the observed effect
[218]. The main effect is due to incorporation of mainly potassium
from the biomass into alumina-silicates in the coal ash followed by
release of HCl, see (15).
ðAl2 O3 ÞðSiO2 Þ2 þ2KCl þ H2 O/ðK2 OÞðAl2 O3 ÞðSiO2 Þ2 þ2HCl

(15)
This is a recognized method for reducing corrosion in biomass
fired boilers.
In the bottom part of Fig. 48 the capture efficiency is calculated
based on the amount of available alkali in the blend. It is clearly
Fig. 47. Burnout of a high-volatile bituminous coal (DA) and blends of the coal and seen that this is a better measure of the slagging and fouling
Eucalyptus (RE) at a stoichiometric value of 1.25. Experiments are performed in air and
tendency of the fuel blends.
oxy-fuel atmospheres with different oxygen concentrations at the inlet of the
entrained-flow reactor setup used. The blending ratio is presumably given on a mass Studies of deposition and fouling in oxy-fuel combustion per-
basis. Data taken from [180]. formed by Fryda et al. [184] have shown no significant change in
10 Heat and Mass Transfer: The differences in thermo-chemical

blends properties between CO2 and N2 cause changes in the heat and
More Deposition
parity
mass transfer rates within a boiler if CO2 is substituted directly for
8 N2 in the oxidizer. In order to obtain an adiabatic flame temperature
during combustion similar to that seen in air-firing the oxygen
concentration in the oxidizer should be increased to 27e35 vol%
6 depending on the type of fuel. Lower rank coals require lower
oxygen concentations than higher rank coals.
Measured Collection Efficiency (%)
CO2 and H2O are radiating gases whereas N2 is not. The radiative
4
heat transfer in oxy-fuel combustion is thus higher than in air
combustion for the same flame temperature. In order to match the
2 heat uptake profile in a retrofittet boiler the flame temperature
should thus be kept lower than during air operation.
Less Deposition
Combustion Process: Devolatilization and ignition of coal parti-
10
0 cles are affected by the change in oxidizer composition from air to
More Deposition blends oxy-fuel combustion. The rate of devolatilization is primarily
pure fuels determined by the surrounding gas temperature as the difference
8 in thermal conductivity of N2 and CO2 is rather low. Particle igni-
tion, on the other hand, is a strong function of both the transport
6 properties of the gas phase surrounding the particles as well as the
combustion heat release rate and the reactivity of the local fuel-
oxidizer mixture. Ignition times comparable to those observed
4 during air combustion can be obtained by increasing the oxidizer
oxygen concentration to 27e35 vol% and thus the flame
temperature. Burnout of volatiles and char are likewise affected by
2 the high CO2 and H2O concentrations in the flame. Especially the
Less Deposition lower diffusivity of both oxygen and small hydrocarbons in CO2
compared to N2 is an important factor. Generally, a slight
0 improvement in char burnout is reported in the literature when the
0 2 4 6 8 10
Interpolated Collision Efficiency (%) flame temperature matches that of air combustion. This improve-
ment is assumed to be caused primarily by the higher oxygen
Fig. 48. Comparisons of measured and predicted particle capture efficiencies for partial pressure experienced by the burning particles. Gasification
different blends of coal and biomasses (wood chips, switchgrass, and wheat straw). reactions and residence time differences are believed to be of minor
Experiments are performed in a 30 kW down-fired, turbulent flow facility. Top: The influence.
predicted capture efficiencies are based on an interpolation between the two pure
fuels and the fraction of the ash content originating from the coal. Bottom: The bias
Due to the high CO2 concentration in the combustion environ-
observed in the upper figure is removed by performing the interpolation based on the ment high levels of CO are expected in the near-flame zone. There is
fuel-available alkali content rather than total ash. Data taken from [218]. thus an increased risk of CO corrosion in this zone. Because of the
large oxygen partial pressure the CO is reported to oxidize before
deposition rate between pure coal and coal/biomass blends (cocoa leaving the larger of the test furnaces yielding similar CO emission
residues and wood chips). The coal to biomass ratio in the blends is rates as for air combustion.
80/20 on a mass basis. The coal/wood blends slightly increase the Emissions: NOx emissions from oxy-fuel combustion is probably
specific fouling factor, whereas blends with cocoa residues slightly the single most investigated area of this technology. Due to the very
decreases the fouling factor. This correlates well with the results low levels of molecular nitrogen a potential exists to reduce the
obtained by Robinson et al. [218]. emissions rate considerably compared to combustion in air through
In relation to the aspects of corrosion and ash quality during co- near-elimination of both thermal and prompt NOx formation.
combustion of coal and biomass at oxy-fuel conditions the sulphur Reductions by 70e80% have been reported. Different suggestions to
and alkali contents of the parent fuels could induce both advan- the mechanisms responsible for the reduction have been proposed.
tages and disadvantages. If the concentration of sulphur in the However, it is generally accepted that reburning reactions play
boiler during oxy-fuel combustion is enhanced there is a larger a major role. Experiments indicate that increasing the oxygen
potential to convert alkali chlorides released from the biomass to concentration and the oxygen excess yield higher emission rates.
sulphates before the flue gas meets the superheater banks and On the other hand, the use of oxidant staging, recycling of flue gas
deposition occurs. The corrosion potential could thus be reduced. before drying, increasing the partial pressure of NOx in the oxidant,
On the other hand, the higher alkali content in biomass ashes, and increasing the oxygen purity, and limiting the air ingress into the
especially straw, will potentially increase the sulphur retention in boiler work to decrease the emission rate.
the fly ash when co-firing with coal as described in Section 3.5.3. Sulphur oxides is the other major pollutant from coal-fired
combustion. The emission of these oxides is not reduced to the
3.7. Summary same degree as NOx. However, increased retention in ash due to the
higher partial pressure of SO2 in the boiler when flue gas is recir-
Below is a short summary of the main findings presented in culated before the desulphurization unit has been reported.
Section 3. Increased risk of low-temperature corrosion by sulfuric acid and
Research Groups: 26 research groups with different experi- a higher acid dew point are some of the major concerns regarding
mental equipment doing oxy-fuel combustion studies have been sulphur in oxy-fuel combustion.
identified. The experimental setups range from 4.2 kWth to Limited work has been reported on the emission of trace
30 MWth in size and include units both with and without flue gas elements from oxy-fuel combustion. The work performed has
recirculation. focused on Hg, Cd, As, and Se with Hg attracting the most attention.
No significant differences between air-firing and oxy-fuel boiler necessitates removal of non-condensable species from the
combustion have been reported at this point. CO2 stream prior to storage. The capture efficiency of an oxy-fuel
Ash and Depositions: The transformations of coal mineral matter power plant is thus expected to be around 90%. Two flue gas
during combustion are affected by the temperature and the gas recycle streams are necessary. The primary stream is used for coal
phase composition surrounding the coal particles. At low drying and transportation to the burners and should preferably be
combustion temperature (for below 30% O2 in CO2) a shift in the desulphurized and dry. Most process configuration suggestions
size distribution of the sub-mircometer-sized fly ash toward avoid addition of O2 to the primary stream. Instead, O2 is added
smaller particles can be observed. Increasing the O2 partial pressure either to the secondary flue gas recycle stream or directly through
provides a size distribution similar to that found in air combustion. special lances in the burners. The overall O2 concentration at the
The ash quality is crucial to its application in cement and concrete inlet to the boiler should be between 27 and 35 vol% in order to
production. Significantly increased sulphur retention could cause yield comparable adiabatic flame temperatures and heat transfer
problems in this respect. profiles as seen for conventional combustion in air. The exact level
The propensity for slagging and fouling in oxy-fuel combustion depends on the type of coal and the boiler design. Even though CO2
is likewise subject to research. At this point, the reported results has significantly different thermo-chemical properties compared to
suggest that only minor changes compared to air-firing will result N2 increasing the O2 concentration in oxy-fuel combustion yields
from the change in the combustion environment. similar devolatilization and ignition behaviour in the two
BECS: Utilization of renewable biomass fuels such as wood, environments. The burnout is reported to improve in oxy-fuel
straw, and other energy crops in thermal power plants with carbon combustion. The emission rate of CO is reported to be similar in
capture and storage is attracting increased attention. The combi- air and oxy-fuel combustion even though significantly increased CO
nation of CO2 neutral fuels with CCS opens a possibility of levels are seen within the flame zone in the oxy-fuel environment.
extracting CO2 from the atmosphere. Until now, only few experi- SO2 emissions are likewise reported to be of similar magnitude
ments on oxy-fuel combustion of biomasses have been reported. It regardless of flue gas composition. The NOx emission is, however,
is expected that further work on BECS will reveal comparable significantly reduced during oxy-fuel combustion compared to air-
changes to the combustion fundamentals as is the case when firing due to a near-elimination of thermal NO formation and to
biomass is introduced during combustion of coal in air. reburning of NOx passed through the burners with the recirculated
flue gas. Reductions of the emission rate by 70e80% have been
4. Conclusions reported. Only minor changes to the deposit build-up are expected
for oxy-fuel combustion compared to air-firing. However, the oxy-
The reduction of CO2 emissions from power plants has become an fuel environment is expected to lead to a significant increase in the
increasing important topic in the discussions of how to prevent SO3 concentration within the boiler increasing the risk of both high
global warming arising from antropogenic CO2 emissions. Three CCS and low-temperature corrosion as well as increased retention of
(Carbon Capture and Storage) technologies have been suggested for sulphur in the fly ash.
medium term application which will reduce the emissions to near A significant amount of new information on the oxy-fuel tech-
zero; Chemical absorption by amines (post-combustion capture), nology has been published since Wall and coworkers released their
Integrated gasification combined cycle plants (pre-combustion literature survey in 2005. At that time the following research needs
capture), and oxy-fuel combustion capture. Of these, the post- were concluded to require the most attention in order to gain
combustion and oxy-fuel technologies can be applied as retrofit a deeper fundamental understanding of the oxy-fuel technology:
solutions to the existing fleet of pulverized-coal-fired power plants.
Compared to post-combustion capture oxy-fuel combustion is less Match of heat transfer characteristics in retrofitted boilers.
mature and thus requires comprehensive fundamental research as The cost of electricity and the cost of CO2 avoided.
well as pilot and demonstration scale testing before commercial Combustion characteristics and emission levels (pilot plant
operation of an oxy-fuel power plant seems to be feasible. Currently, scale).
no full-scale oxy-fuel plants are in operation. The only announced Requirements for gas cleaning - especially SO2.
large demonstration plant (250 MWe) is planned to be put into
operation in 2015 by Vattenfall. Commercial operation does thus not Based on the fact that no full-scale plants are in operation, the
seem to be feasible before 2020. required large-scale investigations of heat transfer characteristics
The current literature review describes the available knowledge still remain. Additional work on both convective and radiative heat
on the oxy-fuel technology. The focus has been on both the changes transfer in laboratory-scale reactors has however been conducted.
induced to an existing, conventional air-fired plant if retrofitted to Several techno-economic assessments of the oxy-fuel tech-
oxy-fuel combustion and the research performed on the funda- nology exist and new are continuously being published. However,
mental combustion characteristics including heat transfer effects, the estimated costs of electricity and CO2 avoidance costs generally
ignition and burnout, emissions, ash quality, and deposit build-up. suffer from large uncertainties due to the lack of construction and
Both coal and the application of CO2 neutral fuels such as wood or operation experience from full-scale plants.
other types of biomass are included in the review. Few results regarding combustion characteristics and emission
Retrofitting an existing pulverized-coal-fired power plant to levels in pilot plant scale setups (>1 MWth) have been published
oxy-fuel operation requires a number of new process units e an air within the last five years. This area thus needs continued attention.
separation facility to provide the combustion process with almost Suggestions to the purity requirements for CO2 for sequestration
pure oxygen, ducts and fans for external recirculation of flue gas to and enhanced oil recovery have been published. No standards or
the burners in order to control flame temperatures, and a CO2 legislative requirements have however been found. Clarification of
processing facility producing a near-pure CO2 stream for storage. the necessary extent of cleaning is very important to plant
The retrofit to oxy-fuel operation will reduce the plant electrical configuration and eventually the process economics.
efficiency by about 10 percent points depending on the initial Besides the topics discussed above, this review has revealed that
efficiency. It is generally accepted that a 95% pure O2 stream a deeper understanding of the effect of a shift from air-blown
constitutes the optimum with respect to minimizing the combined combustion to oxy-fuel combustion on the following issues is
expense for air separation and CO2 cleaning. Leakage of air into the necessary:
Optimum oxygen excess and inlet oxygen concentration [19] Grimston MC, Karakoussis V, Fouquet R, van der Vorst R, Pearson P, Leach M.
The European and global potential of carbon dioxide sequestration in tack-
levels.
ling climate change. Clim Policy 2001;1(2):155e71.
Corrosion at both high and low temperatures. [20] Gunter WD, Wong S, Cheel DB, Sjostrom G. Large CO2 Sinks: their role in the
Ash quality, especially regarding the risk of enhanced sulphur mitigation of greenhouse gases from an international, national (Canadian)
retention. and provincial (Alberta) perspective. Appl Energy 1998;61(4):209e27.
[21] Metz B, Davidson O, Swart R, Pan J, editors. IPCC, 2001: climate change 2001:
Operability, i.e. start-up and shut-down, dynamics during mitigation. Contribution of working group III to the third assessment report
transients, the need for air-firing capability, etc. of the intergovernmental panel on climate change (IPCC). Cambridge, United
Models to predict NOx and SO3 formation. Kingdom and New York, NY, USA: Cambridge University Press; 2001.
[22] Davison J, Freund P, Smith A. Putting carbon back into the ground. Tech-
nology overview, the IEA greenhouse gas R&D programme; February 2001.
The oxy-fuel combustion process is still in the developing phase [23] Gibbins J, Chalmers H. Carbon capture and storage. Energy Policy 2008a;36
and much research is still required in order to fully clarify the (12):4317e22.
[24] Solomon S, Carpenter M, Flach TA. Intermediate storage of carbon dioxide in
consequences of its implementation in power plants. geological formations: a technical perspective. Int J Greenhouse Gas Control
2008;2(4):502e10.
[25] Bentham M, Kirby G. CO2 storage in saline aquifers. Oil Gas Sci Technol
Acknowledgements 2005;60(3):559e67.
[26] Cailly B, Le Thiez P, Egermann P, Audibert A, Vidal-Gilbert S, Longaygue X.
Geological storage of CO2: a state-of-the-art of injection processes and
The work has been funded by Energinet.dk and The Danish technologies. Oil Gas Sci Technol 2005;60(3):517e25.
Ministry of Science, Technology and Innovation (VTU) as well as the [27] Allam RJ, Spilsbury CG. A study of the extraction of CO2 from the flue gas of
companies DONG Energy and Vattenfall. The help by Stine Hansen, a 500 MW pulverised coal fired boiler. Energy Convers Manage 1992;33
(5e8):373e8.
Department of Chemical Engineering, Technical University of [28] Abu-Khader MM. Recent progress in CO2 capture/sequestration. A Review Energy
Denmark in producing the figures for this paper is greatly Sources Part A Recovery Utilization Environ Effects 2006;28(14):1261e79.
appreciated. [29] Fischedick M, Günster W, Fahlenkamp H, Meier H-J, Neumann F,
Oeljeklaus G, et al. Separation in power plants e do retrofits make sense in
existing plants? VGB PowerTech 2006;4:108e17 [in German].
[30] Pehnt M, Henkel J. Life cycle assessment of carbon dioxide capture and storage
References from lignite power plants. Int J Greenhouse Gas Control 2009;3(1):49e66.
[31] Sander MT, Mariz CL. The Fluor Daniel econamine FG process: Past experi-
[1] International Energy Agency (IEA). World energy outlook 2007 e China and ence and present day focus. Energy Convers Manage 1992;33(5e8):341e8.
India insights. IEA Publications; 2007. [32] Reddy S, Scherffius J, Freguia S, Roberts C. Fluor's Econamine FG PlusSM
[2] International Energy Agency (IEA). World energy outlook 2008. IEA Publi- technology e an enhanced amine-based CO2 capture process. In: Proceedings
cations; 2008. of the second national conference on carbon sequestration. Alexandria, VA,
[3] Kavouridis K, Koukouzas N. Coal and sustainable energy supply challenges USA: US Department of Energy National Technology Laboratory; May 2003.
and barriers. Energy Policy 2008;36(2):693e703. [33] Yagi Y, Mimura T, Iijima M, Ishida K, Yoshiyama R, Kamino T. Improvements of
[4] Buhre BJP, Elliott LK, Sheng CD, Gupta RP, Wall TF. Oxy-fuel combustion carbon dioxide capture technology from flue gas. In: Greenhouse gas control
technology for coal-fired power generation. Prog Energy Combust Sci technologies. Proceedings of the seventh international conference on
2005;31(4):283e307. greenhouse gas control technologies, 5e9 September 2004, Vancouver,
[5] Wall TF. Combustion processes for carbon capture. Proc Combust Inst Canada. Oxford, UK: Elsevier Ltd; 2005.
2007;31(1):31e47. [34] Wegerich S, Witt A, Huizeling E, Rode H. Untersuchungen zur Nachrüstung einer
[6] Solomon S, Qin D, Manning M, Chen Z, Marquis M, Averyt KB, et al., editors. CO2-Abscheidetechnologie für das neue E.ON Kraftwerk Maasvlakte 3. In: 39.
IPCC, 2007: climate change 2007: the physical science basis. Contribution of Kraftwerktechnisches Kolloquium. Dresden, Germany; October 11e12, 2007.
working group I to the fourth assessment report of the intergovernmental [35] Oexmann J, Kather A. Post-combustion CO2-capture from coal-fired power
panel on climate change. Cambridge, United Kingdom and New York, NY, plants e wet chemical absorption processes. VGB PowerTech 2009;89
USA: Cambridge University Press; 2007. (1/2):92e103 [in German].
[7] Bachu S. CO2 storage in geological media: role, means, status and barriers to [36] Descamps C, Bouallou C, Kanniche M. Efficiency of an integrated gasification
deployment. Prog Energy Combust Sci 2008;34(2):254e73. combined cycle (IGCC) power plant including CO2 removal. Energy 2008;33
[8] Davison J. Performance and costs of power plants with capture and storage of (6):874e81.
CO2. Energy 2007;32(7):1163e76. [37] Kanniche M, Gros-Bonnivard R, Jaud P, Valle-Marcos J, Amann J-M,
[9] Tan Y, Thambimuthu KV, Douglas MA, Mortazavi R. Oxy-fuel combustion Bouallou C. Pre-combustion, post-combustion and oxy-combustion in
research at the CANMET Energy Technology Center. In: Proc 2003 5th Int thermal power plant for CO2 capture. Appl Thermal Eng 2010;30(1):53e62.
Symp Coal Combust; 2003. p. 550e54. [38] Okawa M, Kimura N, Kiga T, Takano S, Arai K, Kato M. Trial design for a CO2
[10] Jordal K, Anheden M, Yan J, Strömberg L. Oxyfuel combustion for coal-fired recovery power plant by burning pulverized coal in O2/CO2. Energy Convers
power generation with CO2 capture e opportunities and challenges. In: Manage 1997;38:123e7.
The 7th International conference on greenhouse gas control technologies [39] Altmann H, Stamatelopoulos G-N. Steps towards the minimisation of CO2
(GHGT-7). Vancouver, Canada; September 2004. emissions from coal-fired power plants. In: Conference and exhibition for the
[11] Feron PHM, Hendriks CA. CO2 capture process principles and costs. Oil Gas Sci European power generation industry, POWER-GEN Europe 2005. Milan
Technol 2005;60(3):451e9. (Italy); June 28e30, 2005.
[12] Tan R, Corragio G, Santos S. Oxy-coal combustion with flue gas recycle for the [40] Singh D, Croiset E, Douglas PL, Douglas MA. Techno-economic study of CO2
power generation industry e a literature review. Study report, IFRF Doc. No. capture from an existing coal-fired power plant: MEA scrubbing vs. O2/CO2
G 23/y/1. Velsen Noord, The Netherlands: International Flame Research recycle combustion. Energy Convers Manage 2003;44(19):3073e91.
Foundation (IFRF); September 2005. [41] Ewert M. The significance of power stations with CO2 capture in planning
[13] Knudsen JN, Vilhelmsen P-J, Jensen JN, Biede O. First year operating experi- future generation portfolio. VGB PowerTech 2005;85(10):36e40 [in German].
ence with a 1 t/h CO2 absorption e pilot plant at Esbjerg coal-fired power [42] Unknown. Chilling news for carbon capture [[carbon dioxide capture
plant. VGB PowerTech 2007;87(3):57e61. process]]. Mod Power Syst 2006;26(12):17e8.
[14] Figueroa JD, Fout T, Plasynski S, McIlvried H, Srivastava RD. Advances in CO2 [43] Xu B, Stobbs RA, White V, Wall RA, Gibbins J, Iijima M, et al. Future CO2
capture technology e the U.S. department of energy's carbon sequestration capture technology options for the Canadian market. Technical Report Report
program. Int J Greenhouse Gas Control 2008;2(1):9e20. No. COAL R309 BERR/Pub URN 07/1251. Doosan Babcock Energy Limited;
[15] Knauss KG, Johnson JW, Steefel CI, Nitao JJ. Evaluation of the impact of CO2, March 2007.
aqueous fluid, and reservoir rock interactions on the geologic sequestration [44] Wheeldon J, Booras G, Holt N. Post-combustion CO2 capture from pulverized
of CO2, with special emphasis on economic implications. In: 1st national coal plants. In: 23rd Annual international Pittsburgh coal conference, PCC e
conference on carbon sequestration. NETL Publications; 2001. coal-energy, environment and sustainable development; 2006.
[16] Varagani RK, Châtel-Pélage F, Pranda P, Rostam-Abadi M, Lu Y, Bose AC. [45] Notz R, Asprion N, Clausen I, Hasse H. Selection and pilot plant tests of new
Performance simulation and cost assessment of oxy-combustion process for absorbents for post-combustion carbon dioxide capture. Chem Eng Res Des
CO2 capture from coal-fired power plants. In: The fourth annual conference 2007;85(A4):510e5.
on carbon sequestration, May 2e5. Alexandria, VA; 2005. [46] Lee S, Maken S, Park J-W, Song H-J, Park JJ, Shim J-G, et al. A study on the
[17] Gozalpour F, Ren SR, Tohidi B. CO2 EOR and storage in oil reservoirs. Oil Gas carbon dioxide recovery from 2 ton-CO2/day pilot plant at LNG based power
Sci Technol 2005;60(3):537e46. plant. Fuel 2008;87(8e9):1734e9.
[18] Baes Jr CF, Beall SE, Lee DW, Marland G. The collection, disposal, and storage [47] Iea Ghg. Improvement in power generation with post-combustion capture of
of carbon dioxide. In: Interactions of energy and climate, proceedings of an carbon dioxide. Report PH4/33, IEA greenhouse gas research and develop-
international workshop; 1980. p. 495e19. ment programme; 2004.
[48] Williams TC, Shaddix CR, Schefer RW. Effect of syngas composition and CO2- [77] Lee CW, Miller CA. Understanding the potential environmental impacts of
diluted oxygen on performance of a premixed swirl-stabilized combustor. oxy-fuel combustion. In: 3rd Workshop of the IEA GHG international oxy-
Combust Sci Technol 2008;180(1):64e88. combustion network. Yokohama, Japan; March 5e6, 2008.
[49] Damen K, van Troost M, Faaij A, Turkenburg W. A comparison of electricity [78] Fout T. Oxy-COMBUSTION: research, development and systems analysis. In:
and hydrogen production systems with CO2 capture and storage. Part A: 3rd Workshop of the IEA GHG international oxy-combustion network.
review and selection of promising conversion and capture technologies. Prog Yokohama, Japan; March 5e6, 2008.
Energy Combust Sci 2006;32(2):215e46. [79] Andersson K, Johnsson F, Strömberg L. Large scale CO2 capture e applying the
[50] Beér JM. High efficiency electric power generation: the environmental role. concept of O2/CO2 combustion to commercial process data. VGB PowerTech
Prog Energy Combust Sci 2007;33(2):107e34. 2003;83(10):1e5.
[51] McDonald DK, Flynn TJ, DeVault DJ, Varagani R, Levesque S, Castor W. [80] Andersson K, Johnsson F. Process evaluation of an 865 MWe lignite fired O2/
30 MWt clean environment development oxy-coal combustion test program. CO2 power plant. Energy Convers Manage 2006;47(18e19):3487e98.
In: The 33rd international technical conference on coal utilization and fuel [81] Kakaras E, Koumanakos A, Doukelis A, Giannakopoulos D, Vorrias I. Simu-
systems. Clearwater, Florida; 2008. lation of a greenfield oxyfuel lignite-fired power plant. Energy Convers
[52] Minchener AJ. Coal gasification for advanced power generation. Fuel 2005;84 Manage 2007;48(11):2879e87.
(17):2222e35. [82] Hellfritsch S, Gilli PG, Jentsch N. Concept for a lignite-fired power plant based
[53] Gericke B, Kuzmanovski P, Nassauer K. Conceptual solutions regarding the on the optimised oxyfuel process with CO2 recovery. VGB PowerTech
influence of the oxy-fuel-process on conventional power plants [Konzeptü- 2004;84(8):76e82.
berlegungen bezüglich der Auswirkungen des oxy-fuel-prozesses auf kon- [83] Zanganeh KE, Shafeen A. A novel process integration, optimization and
ventionelle Kraftwerksanlagen]. VGB PowerTech 2006;86(10):64e72 [in design approach for large-scale implementation of oxy-fired coal power
German]. plants with CO2 capture. Int J Greenhouse Gas Control 2007;1(1):47e54.
[54] Chen C, Rubin ES. CO2 control technology effects on IGCC plant performance [84] Uddin SN, Barreto L. Biomass-fired cogeneration systems with CO2 capture
and cost. Energy Policy 2009;37(3):915e24. and storage. Renewable Energy 2007;32(6):1006e19.
[55] Eide LI, Bailey DW. Precombustion decarbonisation processes. Oil Gas Sci [85] Li H, Yan J, Yan J, Anheden M. Impurity impacts on the purification process in
Technol 2005;60(3):475e84. oxy-fuel combustion based CO2 capture and storage system. Appl Energy
[56] Lambertz J, Ewers J. Clean coal power e the response of power plant engi- 2009;86(2):202e13.
neering to climate protection challenges. VGB PowerTech 2006;86(5):72e7 [86] Payne R, Chen SL, Wolsky AM, Richter WF. CO2 recovery via coal combustion in
[in German]. mixtures of oxygen and recycled flue gas. Combust Sci Technol 1989;67(1):1e16.
[57] Gibbins J, Chalmers H. Preparing for global rollout: a ‘developed country first’ [87] Kimura N, Omata K, Kiga T, Takano S, Shikisima S. The characteristics of
demonstration programme for rapid CCS deployment. Energy Policy 2008;36 pulverized coal combustion in O2/CO2 mixtures for CO2 recovery. Energy
(2):501e7. Convers Manage 1995;36(6e9):805e8.
[58] Horn FL, Steinberg M. Control of carbon dioxide emissions from a power [88] Woycenko DM, van de Kamp WL, Roberts PA. Combustion of pulverised coal
plant (and use in enhanced oil recovery). Fuel 1982;61(5):415e22. in a mixture of oxygen and recycled flue gas. Summary of the APG research
[59] Herzog H, Golomb D, Zemba S. Feasibility, modeling and economics of program, IFRF Doc. F98/Y/4. Ijmuiden, The Netherlands: International Flame
sequestering power plant CO2 emissions in the deep ocean. Environ Prog Research Foundation (IFRF); October 1995.
1991;10(1):64e74. [89] Tan Y, Douglas MA, Thambimuthu KV. CO2 capture using oxygen enhanced
[60] Abraham BM, Asbury JG, Lynch EP, Teotia APS. Coal-oxygen process provides combustion strategies for natural gas power plants. Fuel 2002;81(8):
CO2 for enhanced recovery. Oil Gas J 1982;80(11):68e70. 75. 1007e16.
[61] Nakayama S, Noguchi Y, Kiga T, Miyamae S, Maeda U, Kawai M, et al. [90] Andersson K, Normann F, Johnsson F, Leckner B. NO emission during oxy-fuel
Pulverized coal combustion in O2/CO2 mixtures on a power plant for CO2 combustion of lignite. Ind Eng Chem Res 2008a;47(6):1835e45.
recovery. Energy Convers Manage 1992;33(5e8):379e86. [91] Farzan H, McDonald DK, McCauley KJ, Varagani R, Prabhakar R, Periasamy C,
[62] Simmons M, Miracca I, Gerdes K. Oxyfuel technologies for CO2 capture: et al. Oxy-coal combustion pilot. In: 3rd Workshop of the IEA GHG interna-
a techno-economic overview. In: 7th international conference on greenhouse tional oxy-combustion network. Yokohama, Japan; March 5e6, 2008.
gas control technologies. Vancouver, Canada; September, 2004. [92] Strömberg L, Lindgren G, Jacoby H, Giering R, Anheden M, Burchhardt U, et al.
[63] Châtel-Pélage F, Varagani R, Pranda P, Perrin N, Farzan H, Vecci SJ, et al. Update on Vattenfall's 30 MWth oxyfuel pilot plant in Schwarze Pumpe.
Applications of oxygen for NOx control and CO2 capture in coal-fired power Energy Procedia 2009;1(1):581e9.
plants. Thermal Sci 2006;10(3):119e42. [93] Darde A, Prabhakar R, Tranier J-P, Perrin N. Air separation and flue gas
[64] Anheden M, Andersson A, Bernstone C, Eriksson S, Yan J, Liljemark S, et-al. compression and purification units for oxy-coal combustion systems. Energy
CO2 quality requirement for a system with CO2 capture, transport and Procedia 2009;1(1):527e34.
storage. In: The 7th International Conference on greenhouse gas control [94] Burchhardt U, Radunsky D. Erfahrungen aus der Planung und Genehmigung
technologies (GHGT7). Vancouver, Canada; September 5e9, 2004. der Oxyfuel-Forschungsanlage von Vattenfall. In: 39. Kraftwerktechnisches
[65] Anheden M, Yan J, De Smedt G. Denitrogenation (or oxy-fuel concepts). Oil Kolloquium. Dresden, Germany; October 11e12, 2007.
Gas Sci Technol 2005;60(3):485e95. [95] Simpson AP, Simon AJ. Second law comparison of oxy-fuel combustion and
[66] Dillon DJ, Panesar RS, Wall RA, Allam RJ, White V, Gibbins J, et al. Oxy- post-combustion carbon dioxide separation. Energy Convers Manage
combustion processes for CO2 capture from advanced supercritical PF and 2007;48(11):3034e45.
NGCC power plant. In: 7th international conference on greenhouse gas [96] Hu Y, Naito S, Kobayashi N, Hasatani M. CO2, NOx and SO2 emissions from the
control technologies. Vancouver, Canada; September, 2004. combustion of coal with high oxygen concentration gases. Fuel 2000;79
[67] Dillon DJ, White V, Allam RJ, Wall RA, Gibbins J. Oxy Combustion Processes (15):1925e32.
for CO2 capture from power plant. Engineering Investigation Report, 2005/9, [97] Croiset E, Thambimuthu KV. NOx and SO2 emissions from O2/CO2 recycle coal
IEA Greenhouse Gas Research and Development Programme; June 2005. combustion. Fuel 2001;80(14):2117e21.
[68] Kather A, Klostermann M, Hermsdorf C, Mieske K, Eggers R, Köpke D. Kon- [98] White V, Torrente-Murciano L, Sturgeon D, Chadwick D. Purification of
zept für ein 600 MWel Steinkohlekraftwerk mit CO2-Abtrennung auf Basis oxyfuel-derived CO2. Energy Procedia 2009;1(1):399e406.
des Oxyfuel-Prozesses. In: Kraftwerksbetrieb unter künftigen Rahmenbe- [99] Altmann H, Porsche T, Burchhardt U. Erfahrungen aus der Inbetriebnahme
dingungen, 38. Kraftwerkstechnisches Kolloquium. Dresden, Germany; und dem Versuchsbetrieb der Oxyfuel-Forschungsanlage von Vattenfall. VGB
October 24e25, 2006. PowerTech 2009;89(9):96e100.
[69] Rezvani S, Huang Y, McIlveen-Wright D, Hewitt N, Wang Y. Comparative [100] Smith AR, Klosek J. A review of air separation technologies and their inte-
assessment of sub-critical versus advanced super-critical oxyfuel fired pf gration with energy conversion processes. Fuel Processing Technol 2001;70
boilers with CO2 sequestration facilities. Fuel 2007;86(14):2134e43. (2):115e34.
[70] Eide LI, Anheden M, Lyngfelt A, Abanades C, Younes M, Clodic D, et al. Novel [101] Gonschorek S, Hellfritsch S, Weigl S, Gampe U. Entwicklungsstand des
capture processes. Oil Gas Sci Technol 2005;60(3):497e508. Oxyfuel-Prozesses für Braunkohlekraftwerke. In: Kraftwerkstechnisches
[71] Hossain MM, de Lasa HI. Chemical-looping combustion (CLC) for inherent Kolloquium, Zittau; 26e27 September, 2006.
CO2 separationsea review. Chem Eng Sci 2008;63(18):4433e51. [102] Vitalis B. Overview of oxy-combustion technology for utility coal-fired
[72] Farley M. Developing oxyfuel capture as a retrofit technology. Mod Power boilers. Advances in Materials Technology for Fossil Power Plants. In:
Syst 2006;26(4):20e2. Proceedings from the 5th International Conference; 2008. p. 968e81.
[73] Aspelund A, Jordal K. A study of the interface between CO2 capture and [103] Kather A, Klostermann M, Hermsdorf C, Mieske K, Eggers R, Köpke D.
transport. In: The 8th International Conference on Greenhouse Gas Control Steinkohlekraftwerk mit CO2-Abtrennung auf Basis des Oxyfuel-Prozesses. In:
Technologies (GHGT-8). Trondheim, Norway; 19e22 June, 2006. 39. Kraftwerktechnisches Kolloquium. Dresden, Germany; October 11e12, 2007.
[74] Aspelund A, Jordal K. Gas conditioning e the interface between CO2 capture [104] European Standard. EN 450e1:2005 Fly ash for concrete. Definition, speci-
and transport. Int J Greenhouse Gas Control 2007;1(3):343e54. fications and conformity criteria; 2005.
[75] Sass BM, Farzan H, Prabhakar R, Gerst J, Sminchak J, Bhargava M, et al. [105] Mauder R, Hugot A. Flugaschevermarktung in der Zukunft. VGB PowerTech
Considerations for treating impurities in oxy-combustion flue gas prior to 2008;88(11):62e6.
sequestration. Energy Procedia 2009;1(1):535e42. [106] Kluger F, Lysk S, Altmann H, Krohmer B, Stamatelopoulos G-N. 30 MWth
[76] Anheden M, Rydberg S, Yan J. Consideration for removal of non-CO2 Oxyfuel-Pilotanlage e Untersuchungsschwerpunkte und Auslegung des
components from CO2 rich flue gas of oxy-fuel combustion. In 3rd workshop Dampferzeugers. In: Kraftwerksbetrieb unter künftigen Rahmenbedingungen,
of the IEA GHG International oxy-combustion network. Yokohama, Japan; 38. Kraftwerkstechnisches Kolloquium. Dresden, Germany; October 24e25,
March 5e6, 2008. 2006.
[107] Toporov D, Förster M, Kneer R. How to burn pulverized coal in CO2 atmo- [136] Kiga T, Takano S, Kimura N, Omata K, Okawa M, Mori T, et al. Characteristics
sphere at low oxygen concentrations. VDI Berichte 1988;55e60:2007. of pulverized-coal combustion in the system of oxygen/recycled flue gas
[108] Toporov D, Bocian P, Heil P, Kellermann A, Stadler H, Tschunko S, et al. combustion. Energy Convers Manage 1997;38:129e34.
Detailed investigation of a pulverized fuel swirl flame in CO2/O2 atmosphere. [137] Watanabe S, Endo Y, Kiga T, Kimura N, Okawa M. Analysis of pulverized-coal
Combust Flame 2008;155(4):605e18. flames in the system of oxygen/recycled flue gas combustion. Am Soc Mech
[109] Nikzat H, Pak H, Fuse T, Hu Y, Ogyu K, Kobayashi N, et al. Characteristics of Engineers, Environ Control Division Publ EC 1997;5:239e46.
pulverized coal burner using a high-oxygen partial pressure. Chem Eng Res [138] Wall TF, Elliott L, Khare S, Liu Y, Yamada T, Tamura M, et al. Ash impacts in
Des 2004;82(1):99e104. oxy-fuel combustion. In: Proceedings: impacts of fuel quality on power
[110] Hellfritsch S, Gonschorek S, Vilhelm R, Löser J, Klemm M, Weigl S, et al. production. Utah, USA; 2006.
Entwicklungsstand des Oxyfuel-Prozesses für Braunkohlekraftwerke. In [139] Wall TF. Performance of PF boilers retrofitted with oxy-coal combustion:
Kraftwerksbetrieb unter künftigen Rahmenbedingungen, 38. In: Kraftwerk- understanding burnout, coal reactivity, burner operation, and furnace heat
stechnisches Kolloquium. Dresden, Germany; October 24e25, 2006. transfer. In: 3rd Workshop of the IEA GHG international oxy-combustion
[111] Oryshchyn D, Gerdemann S, Armstrong J, Ochs T, Summers C. Oxy-firing flue- network. Yokohama, Japan; March 5e6, 2008.
gas character and its effect on FGD e a process study for integrated pollutant [140] Goh B. 1 MWth oxyfuel combustion test facility. “Recent developments in
removal. In: 23rd Annual international Pittsburgh coal conference, PCC e CCS” coal research forum e combustion division meeting 17th April 2007.
coal-energy, environment and sustainable development; 2006. Imperial College London; 2007.
[112] EUROGYPSUM. Association of European gypsum industries. FGD Gypsum e [141] Goh B. Oxyfuel combustion pilot-scale testing. Oxyfuel combustion: oppor-
quality criteria and analysis methods, www.eurogypsum.org; April 2005. tunities and challengese7th of May 2008. University of Leeds; 2008a.
[113] Klostermann M. Efficiency Increase of the oxyfuel process by waste heat [142] Goh B.E.ON UK's pilot scale oxyfuel combustion experiences: development,
recovery considering the effects of flue gas treatment. In: 3rd workshop of testing and modelling. In: 3rd workshop of the IEA GHG international oxy-
the IEA GHG international oxy-combustion network. Yokohama, Japan; combustion network. Yokohama, Japan; March 5e6, 2008.
March 5e6, 2008. [143] Dhungel B, Mönckert P, Maier J, Scheffknecht G. Investigation of oxy-coal
[114] Suriyawong A, Hogan Jr CJ, Jiang J, Biswas P. Charged fraction and electrostatic combustion in semi-technical test facilities. In: 3rd international confer-
collection of ultrafine and submicrometer particles formed during O2eCO2 ence on clean coal technology. Italy, 2007.
coal combustion. Fuel 2008;87(6):673e82. [144] Maier J, Dhungel B, Mönckert P, Scheffknecht G. Combustion and emission
[115] Schnurrer S, Elliott L, Wall T, Liu Y. Influence of oxy-fuel environment on behavior under oxyfuel conditions. In: 39. Kraftwerktechnisches Kolloquium.
sulphur species in ash from pulverized coal combustion. In: Impacts of fuel Dresden, Germany; October 11e12, 2007 [In German].
quality on power production and the environment, September 29eOctober 3. [145] Maier J, Dhungel B, Mönckert P, Kull R, Scheffknecht G. Impact of recycled gas
Banff, Alberta, Canada: The Banff Centre; 2008. species (SO2, NO) on emission behaviour and fly ash quality during oxy-coal
[116] Khare SP, Wall TF, Farida AZ, Liu Y, Moghtaderi B, Gupta RP. Factors influ- combustion. In: Proceedings of the 33rd international technical conference
encing the ignition of flames from air-fired swirl pf burners retrofitted to oxy- on coal utilization and fuel systems. Clearwater, Florida; June 1e5, 2008.
fuel. Fuel 2008;87(7):1042e9. [146] Mönckert P, Reber D, Maier J, Scheffknecht G. Operation of a retrofitted
[117] Suriyawong A, Gamble M, Lee M-H, Axelbaum R, Biswas P. Submicrometer 0.5 MWth PF combustion facility under oxyfuel conditionsean experience
particle formation and mercury speciation under O2eCO2 coal combustion. report. In: The 32nd international technical conference on coal utilization
Energy Fuels 2006;20(6):2357e63. and fuel systems, the power of coal. Clearwater, Florida; 10e15 June, 2007.
[118] Wang CS, Berry GF, Chang KC, Wolsky AM. Combustion of pulverized coal [147] Mönckert P, Dhungel B, Kull R, Maier J. Impact of combustion conditions on
using waste carbon dioxide and oxygen. Combust Flame 1988;72(3):301e10. emission formation (SO2, NOx) and fly ash. In: 3rd workshop of the IEA GHG
[119] Nozaki T, Takano S-i, Kiga T, Omata K, Kimura N. Analysis of the flame formed international oxy-combustion network. Yokohama, Japan; March 5e6, 2008.
during oxidation of pulverized coal by an O2eCO2 mixture. Energy 1997;22 [148] Scheffknecht G, Maier J. Firing issues related to the oxyfuel process. VGB
(2e3):199e205. PowerTech 2008;88(11):91e7.
[120] Croiset E, Thambimuthu K, Palmer A. Coal combustion in O2/CO2 mixtures [149] Zheng L, Furimsky E. Assessment of coal combustion in O2 þ CO2 by equi-
compared with air. Can J Chem Eng 2000;78(2):402e7. librium calculations. Fuel Processing Technol 2003;81(1):23e34.
[121] Liu H, Zailani R, Gibbs BM. Comparisons of pulverized coal combustion in air [150] Chui EH, Douglas MA, Tan Y. Modeling of oxy-fuel combustion for a western
and in mixtures of O2/CO2. Fuel 2005;84(7e8):833e40. Canadian sub-bituminous coal. Fuel 2003;82(10):1201e10.
[122] Tan Y, Croiset E, Douglas MA, Thambimuthu KV. Combustion characteristics [151] Chui EH, Majeski AJ, Douglas MA, Tan Y, Thambimuthu KV. Numerical
of coal in a mixture of oxygen and recycled flue gas. Fuel 2006;85(4):507e12. investigation of oxy-coal combustion to evaluate burner and combustor
[123] Andersson K, Normann F, Johnsson F. Experiments and modeling on oxy-fuel design concepts. Energy 2004;29(9e10):1285e96.
combustion chemistry during lignite-firing. In: The 32nd international [152] White V. Purification of oxyfuel-derived CO2. In: 3rd workshop of the IEA GHG
technical conference on coal utilization and fuel systems. Clearwater, Florida; international oxy-combustion network. Yokohama, Japan; March 5e6, 2008.
10e15 July2007. [153] Andersson K, Johnsson F. Flame and radiation characteristics of gas-fired O2/
[124] Mackrory AJ, Lokare S, Baxter LL, Tree DR. An investigation of nitrogen CO2 combustion. Fuel 2007;86(5e6):656e68.
evolution in oxy-fuel combustion. In: The 32nd international technical [154] Leiser S, Schnell U, Scheffknecht G. A kinetic homogeneous reaction model
conference on coal utilization and fuel systems, the power of coal. Clear- for oxy-fuel combustion. In: 9th conference on energy for a clean environ-
water, Florida; 10e15 June, 2007. ment; 2007.
[125] Sheng C, Li Y. Experimental study of ash formation during pulverized coal [155] Andersson K, Johansson R, Johnsson F, Leckner B. Radiation intensity of
combustion in O2/CO2 mixtures. Fuel 2008;87(7):1297e305. propane-fired oxy-fuel flames: implications for soot formation. Energy Fuels
[126] Molina A, Shaddix CR. Ignition and devolatilization of pulverized bituminous 2008b;22(3):1535e41.
coal particles during oxygen/carbon dioxide coal combustion. Proc Combust [156] Normann F, Andersson K, Leckner B, Johnsson F. High-temperature reduction
Inst 2007;31(2):1905e12. of nitrogen oxides in oxy-fuel combustion. Fuel 2008;87(17):3579e85.
[127] Liu H, Okazaki K. Simultaneous easy CO2 recovery and drastic reduction of [157] Valero A, Romeo LM, Díez LI, Pérez A. OXY-CO-FIRING: a negative CO2
SOx and NOx in O2/CO2 coal combustion with heat recirculation. Fuel 2003;82 emission process. In: The 8th international conference on greenhouse gas
(11):1427e36. control technologies (GHGT-8). Trondheim, Norway; 19e22 June, 2006.
[128] Wall T, Liu Y, Spero C, Elliott L, Khare S, Rathnam R, et al. An overview on [158] Okazaki K, Ando T. NOx reduction mechanism in coal combustion with
oxyfuel coal combustioneState of the art research and technology develop- recycled CO2. Energy 1997;22(2e3):207e15.
ment. Chem Eng Res Des 2009;87(8):1003e16. [159] Okazaki K. Technical consideration and challenges of oxy-pulverized coal
[129] Bordenet B. Influence of novel cycle concepts on the high-temperature combustion. In: 3rd workshop of the IEA GHG international oxy-combustion
corrosion of power plants. Mater Corros 2008;59(5):361e6. network. Yokohama, Japan; March 5e6, 2008.
[130] Bordenet B, Kluger F. Thermodynamic modelling of the corrosive deposits in [160] Hu YQ, Kobayashi N, Hasatani M. The reduction of recycled-NOx in coal
oxy-fuel fired boilers. Mater Sci Forum 2008;595e598(1):261e9. combustion with O2/recycled flue gas under low recycling ratio. Fuel 2001;80
[131] Kung SC, Tanzosh JM, Mcdonald DK. Fireside corrosion study using B&W (13):1851e5.
clean environment development facility for oxy-coal combustion systems. [161] Hu YQ, Kobayashi N, Hasatani M. Effects of coal properties on recycled-NOx
Advances in materials technology for fossil power plants. In: Proceedings reduction in coal combustion with O2/recycled flue gas. Energy Convers
from the 5th International conference; 2008. p. 982e92. Manage 2003;44(14):2331e40.
[132] Abellán JPirón, Olszewski T, Penkalla HJ, Meier GH, Singheiser L, [162] Liu H, Zailani R, Gibbs BM. Pulverized coal combustion in air and in O2/CO2
Quadakkers WJ. Scale formation mechanisms of martensitic steels in high mixtures with NOx recycle. Fuel 2005b;84(16):2109e15.
CO2/H2O-containing gases simulating oxyfuel environments. Mater High [163] Maier J, Dhungel B, Mönckert P, Scheffknecht G. Coal combustion and
Temp 2009;26(1):63e72. emission behaviour under oxy-fuel combustion. In: Proceedings of the 31st
[133] Rauscher K. Clean coal in the energy mix from tomorrow. VGB PowerTech international technical conference on coal utilization and fuel systems.
2005;85(9):70e2. Clearwater, Florida; 21e26 May, 2006.
[134] Ritter R, Holling B, Altmann H, Biele M. Konzepte und Ausblick für eine CO2- [164] Dhungel B, Maier J, Scheffknecht G. Emission behaviour during oxy-coal
Anlage eines Oxyfuel-Kraftwerkes am Bespiel Schwarze Pumpe. In: 39. combustion in a 20 kW once through furnace. In: 9th conference on
Kraftwerktechnisches Kolloquium. Dresden, Germany; October 11e12, 2007. energy for a clean environment. Portugal; 2007.
[135] McCauley KJ, Farzan H, Alexander KC, McDonald DK, Varagani R, Prabhakar R, [165] Al-Makhadmeh L, Maier J, Scheffknecht G. Coal pyrolysis and char combustion
et al. Commercialization of oxy-coal combustion: applying results of a large under oxy-fuel conditions. In: The 34th international technical conference on
30 MWth pilot project. Energy Procedia 2009;1(1):439e46. coal utilization and fuel systems. Clearwater, Florida; May 31eJune 4, 2009.
[166] Murphy JJ, Shaddix CR. Combustion kinetics of coal chars in oxygen-enriched [196] Masri AR, Dibble RW, Barlow RS. Chemical kinetic effects in nonpremixed
environments. Combust Flame 2006;144(4):710e29. flames of H2/CO2 fuel. Combust Flame 1992;91(3e4):285e309.
[167] Shaddix CR, Molina A. Effect of O2 and High CO2 concentrations on PC char [197] Miller JA, Bowman CT. Mechanism and modeling of nitrogen chemistry in
burning rates during oxy-fuel combustion. In: The 33rd international tech- combustion. Prog Energy Combust Sci 1989;15:287e338.
nical conference on coal utilization and fuel systems, coal for the future. [198] Johnsson JE. Formation and reduction of nitrogen oxides in fluidized-bed
Clearwater, Florida; 1e5 June, 2008. combustion. Fuel 1994;73:1398e415.
[168] Shaddix CR, Molina A. Particle imaging of ignition and devolatilization of [199] Glarborg P, Jensen AD, Johnsson JE. Fuel nitrogen conversion in solid fuel
pulverized coal during oxy-fuel combustion. Proc Combust Inst 2009;32 fired systems. Prog Energy Combust Sci 2003;29(2):89e113.
(2):2091e8. [200] Turns SR. An introduction to combustion: concepts and applications. Inter-
[169] Shaddix CR, Molina A. Understanding the effects of O2 and CO2 on NOx formation national Editions. McGraw-Hill; 2000.
during oxy-coal combustion. In: 3rd Workshop of the IEA GHG international [201] Normann F, Andersson K, Leckner B, Johnsson F. Emission control of
oxy-combustion network. Yokohama, Japan; March 5e6, 2008. nitrogen oxides in the oxy-fuel process. Prog Energy Combust Sci 2009;35
[170] Becher V, Goanta A, Gleis S, Spliethoff H. Controlled staging with non- (5):385e97.
stoichiometric burners for oxyfuel processes. In: The 32nd international [202] Wendt JOL, Sternling CV, Matovich MA. Reduction of sulfur trioxide and nitrogen
technical conference on coal utilization and fuel systems, the power of coal. oxides by secondary fuel injection. Proc Combust Inst 1973;14(1):897e904.
Clearwater, Florida; 10e15 June, 2007. [203] Levy JM, Chan LK, Sarofim AF, Beer JM. NO/char reactions at pulverized coal
[171] Goanta A, Becher V, Bohn J-P, Gleis S, Spliethoff H. Controlled staging with flame conditions. Proc Combust Inst 1981;18:111e20.
non-stoichiometric burners for oxy-fuel processesenumerical validation. In: [204] Aarna I, Suuberg EM. A review of the kinetics of the nitric oxide-carbon
The 33rd international technical conference on coal utilization and fuel reaction. Fuel 1997;76(6):475e91.
systems, coal for the future. Clearwater, Florida; 1e5 June, 2008. [205] Aarna I, Suuberg EM. The role of carbon monoxide in the NO-carbon reaction.
[172] Glarborg P, Bentzen LLB. Chemical effects of a high CO2 concentration in oxy- Energy Fuels 1999;13(6):1145e53.
fuel combustion of methane. Energy Fuels 2008;22(1):291e6. [206] López D, Calo J. The NO-carbon reaction: the influence of potassium and CO
[173] Mendiara T, Glarborg P. Reburn chemistry in oxy-fuel combustion of on reactivity and populations of oxygen surface complexes. Energy Fuels
methane. Energy Fuels 2009a;23(7):3565e72. 2007;21(4):1872e7.
[174] Mendiara T, Glarborg P. Ammonia chemistry in oxy-fuel combustion of [207] Park D-C, Day SJ, Nelson PF. Nitrogen release during reaction of coal char
methane. Combust Flame 2009b;156:1937e49. with O2, CO2, and H2O. Proc Combust Inst 2005;30(2):2169e75.
[175] Mackrory AJ, Tree DR. NOx Destruction experiments and modeling in oxy-fuel [208] Liu H, Katagiri S, Kaneko U, Okazaki K. Sulfation behavior of limestone
combustion. In: The 33rd international technical conference on coal utilization under high CO2 concentration in O2/CO2 coal combustion. Fuel 2000;79
and fuel systems, coal for the future. Clearwater, Florida; 1e5 June, 2008. (8):945e53.
[176] Mackrory AJ, Tree DR. Predictions of NOx in a laboratory pulverized coal [209] Liu H, Katagiri S, Okazaki K. Drastic SOx removal and influences of various
combustor operating under air and oxy-fuel conditions. Combust Sci Technol factors in O2/CO2 pulverized coal combustion system. Energy Fuels 2001b;15
2009;181(11):1413e30. (2):403e12.
[177] Heil P, Stadler H, Förster M, Kneer R. Flammlose Verbrennung in einer O2/CO2 [210] Krishnamoorthy G, Veranth JM. Computational modeling of CO/CO2 ratio
Atmosphäre. In: 39. Kraftwerktechnisches Kolloquium. Dresden, Germany, inside single char particles during pulverized coal combustion. Energy Fuels
October 11e12, 2007. 2003;17(5):1367e71.
[178] Stadler H, Ristic D, Förster M, Schuster A, Kneer R, Scheffknecht G. NOx- [211] Keith DW, Rhodes JS. Bury, burn or both: a two-for-one deal on biomass
emissions from flameless coal combustion in air, Ar/O2 and CO2/O2. Proc carbon and energy. Climatic Change 2002;54(3):375e7.
Combust Inst 2009;32(2):3131e8. [212] Obersteiner M, Azar C, Kauppi P, Möllersten K, Moreira J, Nilsson S, et al.
[179] Engels S. Developments in oxy-combustion technology for power plants with Managing climate risk. Science 2001;294(5543):786e7.
CCS. 1st Young Researchers Forum Organised by IEA Greenhouse Gas R&D [213] Kraxner F, Nilsson S, Obersteiner M. Negative emissions from BioEnergy use,
Programme, Hamburg, Germany; December 8, 2006. carbon capture and sequestration (BECS) e the case of biomass production by
[180] Arias B, Pevida C, Rubiera F, Pis JJ. Effect of biomass blending on coal ignition sustainable forest management from semi-natural temperate forests.
and burnout during oxy-fuel combustion. Fuel 2008;87(12):2753e9. Biomass Bioenergy 2003;24(4e5):285e96.
[181] Bejarano PA, Levendis YA. Single-coal-particle combustion in O2/N2 and O2/ [214] Azar C, Lindgren K, Larson E, Mollersten K. Carbon capture and storage from
CO2 environments. Combust Flame 2008;153(1e2):270e87. fossil fuels and biomass e costs and potential role in stabilizing the atmo-
[182] Sheng C, Li Y, Liu X, Yao H, Xu M. Ash particle formation during O2/CO2 combustion sphere. Climatic Change 2006;74(1e3):47.
of pulverized coals. Fuel Processing Technol 2007a;88(11e12):1021e8. [215] Read P. Bioenergy with carbon storage e strengthening Kyoto through
[183] Sheng C, Lu Y, Gao X, Yao H. Fine ash formation during pulverized coal complementary action to address the threat of abrupt climate change.
combustionea comparison of O2/CO2 combustion versus air combustion. Renewable Energy Dev 2006;19(3):9e11.
Energy Fuels 2007b;21(2):435e40. [216] Rhodes JS, Keith DW. Biomass with capture: negative emissions within social
[184] Fryda L, Cieplik MK, Jacobs JM, van de Kamp WL. Study of oxyfuel combustion of and environmental constraints: an editorial comment. Climatic Change
coal and biomass in a lab-scale pulverised fuel combustor. In: 16th European 2008;87(3e4):321e8.
biomass conference and exhibition. Valencia, Spain; June 2e6, 2008. [217] Borrego AG, Garavaglia L, Kalkreuth WD. Characteristics of high heating rate
[185] Spinti JP, Pershing DW. The fate of char-N at pulverized coal conditions. biomass chars prepared under N2 and CO2 atmospheres. Int J Coal Geol
Combust Flame 2003;135(3):299e313. 2009;77(3e4):409e15.
[186] Zhang J, Eddings EG, Wendt JOL, Smith PJ. Model validation studies for [218] Robinson AL, Junker H, Buckley SG, Sclippa G, Baxter LL. Interactions between
pulverized coal jet ignition in O2/CO2 environments. In: 3rd workshop of the coal and biomass when cofiring. Proc Combust Inst 1998;27(1):1351e9.
IEA GHG international oxy-combustion network. Yokohama, Japan; March
5e6, 2008.
[187] Naredi P, Pisupati SV. Comparison of char burnout and CO emissions from
oxy-coal combustion with combustion in air: an experimental and numerical Nomenclature
study. In: The 33rd international technical conference on coal utilization and
fuel systems, coal for the future. Clearwater, Florida; 1e5 June, 2008. ASU: Air Separation Unit
[188] Andersson K, Johansson R, Hjartstam S, Johnsson F, Leckner B. Radiation inten- BECS: Biomass Energy for Carbon Capture and Sequestration
sity of lignite-fired oxy-fuel flames. Exp Thermal Fluid Sci 2008c;33(1):67e76. CCS: Carbon Capture and Storage
[189] Huang X, Jiang X, Han X, Wang H. Combustion characteristics of Fine- and EFR: Entrained-Flow Reactor
micro-pulverized coal in the mixture of O2/CO2. Energy Fuels 2008;22 EOR: Enhanced Oil Recovery
(6):3756e62. ESP: Electrostatic Precipitator
[190] Suda T, Masuko K, Sato J, Yamamoto A, Okazaki K. Effect of carbon dioxide on FGD: Flue Gas Desulphurization
flame propagation of pulverized coal clouds in CO2/O2 combustion. Fuel IGCC: Integrated Gasification Combined Cycle
2007;86(12e13):2008e15. ITM: Ion Transport Membrane (for air separation)
[191] Stanmore BR, Visona SP. The Contribution to char burnout from gasification LCV/LHV: Low Calorific/Heating Value
by H2O and CO2 during pulverized-coal flame combustion. Combust Flame RFG: Recirculated Flue Gas
1998;113(1e2):274e6. SCR: Selective Catalytic Reduction
[192] Varhegyi G, Szabo P, Jakab E, Till F, Richard J-R. Mathematical modeling of char
reactivity in AreO2 and CO2eO2 mixtures. Energy Fuels 1996;10(6):1208e14. Research groups
[193] Borrego AG, Alvarez D. Comparison of chars obtained under oxy-fuel and ANL: Argonne National Laboratory
conventional pulverized coal combustion atmospheres. Energy Fuels BYU: Brigham Young University (Utah, USA)
2007;21(6):3171e9. CANMET: Canada Centre for Mineral and Energy Technology
[194] Liu F, Guo H, Smallwood GJ. The chemical effect of CO2 replacement of N2 in DTU: Technical University of Denmark
air on the burning velocity of CH4 and H2 premixed flames. Combust Flame IFRF: International Flame Research Foundation
2003;133(4):495e7. IHI: Ishikawajima-Harima Heavy Industries
[195] Liu F, Guo H, Smallwood GJ, Gülder Ö L. The chemical effects of carbon IVD: Institute of Process Engineering and Power Plants Technology, University of
dioxide as an additive in an ethylene diffusion flame: implications for soot Stuttgart
and NOx formation. Combust Flame 2001;125(1e2):778e87. UU: University of Utah

Progress in Energy and Combustion Science

Uploaded by

Copyright:

Available Formats

Progress in Energy and Combustion Science

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Progress in Energy and Combustion Science

Uploaded by

Copyright:

Available Formats

Progress in Energy and Combustion Science 36 (2010) 581e625

Contents lists available at ScienceDirect

Progress in Energy and Combustion Science

Oxy-fuel combustion of solid fuels

2.7. Boiler and steam cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593

18th century the amount of CO2 in the atmosphere has increased

Table 1 lists the current and projected CO2 emissions, in Gton

Table 1 combination with enhanced oil recovery (EOR) differs between

Sink Storage capacity Retention

Table 3 Ash composition (wt%, dry)

Anheden et al. [76] Lee and Miller [77] Fout [78]

impurities such as N2, O2, and Ar [85]. The thermodynamic proper-

Fraction of total ﬂue gas recycled Sulphur concentration in ﬂue gas

Option Advantage(s) Disadvantage(s)

H2O O2 N2 CO2 Ratio CO2/N2

35 % O opposite effect of what was observed. These ﬁndings thus suggest

1300 3.2.2.1. Overall heat transfer. Experiments performed by IFRF

1600 O2/CO2 combustion

1600 Dry recycle

Radiative Heat Flux (kW/m2)

overall radiation from a solid-fuel ﬂame is not known and could

100 open symbols = N2

ence of CO2 and enhanced oxygen concentrations on the devola-

Bituminous Coal Volatiles Bituminous Coal Chars

Burnout Time (ms)

3.3.2.3. Effect of coal properties. Liu et al. [162] investigated the

tion, (8), [172].

Bituminous Coal Volatiles Bituminous Coal Chars

Burnout Time (ms)

3.3.2.6. Summary. The difference in the thermo-chemical prop-

Author(s) Fuel input Inlet O2, Emission Conversion Conclusion(s)

reduction of NOx by reducing agents, such as hydrocarbons from

20 % O2 / 80 % CO2 50 % O2 / 50 % CO2 80 % O2 / 20 % CO2

emission rates due to the lower volume of ﬂue gas.

1400 rate. In oxy-fuel combustion a high initial NOx concentration at the

a function of the conversion of volatile-N to NOx and the fraction of

NO emission rate (mg/MJ)

3.4.1.7. Inﬂuence of oxidant staging. The use of oxidant staging in

1600 down to 40 mg/MJ at oxy-fuel conditions and improved the

in NOx emissions. (i) a burner which yields a fuel-rich ﬂame core,

600 oxidation zone with temperatures as low as possible in combina-

100 optimal burner design for oxy-fuel combustion is not necessarily

400 show a decrease when comparing to combustion in air

Author(s) Fuel input Inlet O2, Emission Conversion Conclusion(s)

Air 50 % O2 / 50 % CO2 100 % O2

0.5 1 1.5 2 2.5 3 1 1.5 2 2.5 3 1 1.5 2 2.5 3

SO emission rate (mg/MJ)

in conventional air-ﬁring systems [113,122,147] depending on the 400

to drastically reduce the emissions of SO2 from combustion of coal.

3.4.2.6. Summary. The substitution of N2 by CO2 in oxy-fuel

in the sub-micron ash for oxy-fuel combustion with an air-like 600

proﬁles in the reactor caused by the introduction of the more reac-

ðAl2 O3 ÞðSiO2 Þ2 þ2KCl þ H2 O/ðK2 OÞðAl2 O3 ÞðSiO2 Þ2 þ2HCl

10 Heat and Mass Transfer: The differences in thermo-chemical

You might also like