Life cycle modelling of fossil fuel power generation with post-combustion CO2 capture

International Journal of Greenhouse Gas Control 4 (2010) 289–300

Contents lists available at ScienceDirect

International Journal of Greenhouse Gas Control journal homepage: www.elsevier.com/locate/ijggc

Life cycle modelling of fossil fuel power generation with post-combustion CO2 capture Anna Korre *, Zhenggang Nie, Sevket Durucan Department of Earth Science and Engineering, Imperial College London, Exhibition Road, London SW7 2AZ, United Kingdom

A R T I C L E I N F O

A B S T R A C T

Article history: Received 5 May 2009 Received in revised form 17 August 2009 Accepted 18 August 2009 Available online 22 September 2009

Due to its compatibility with the current energy infrastructures and the potential to reduce CO2 emissions significantly, CO2 capture and geological storage is recognised as one of the main options in the portfolio of greenhouse gas mitigation technologies being developed worldwide. The CO2 capture technologies offer a number of alternatives, which involve different energy consumption rates and subsequent environmental impacts. While the main objective of this technology is to minimise the atmospheric greenhouse gas emissions, it is also important to ensure that CO2 capture and storage does not aggravate other environmental concerns. This requires a holistic and system-wide environmental assessment rather than focusing on the greenhouse gases only. Life Cycle Assessment meets this criteria as it not only tracks energy and non-energy-related greenhouse gas releases but also tracks various other environmental releases, such as solid wastes, toxic substances and common air pollutants, as well as the consumption of other resources, such as water, minerals and land use. This paper presents the principles of the CO2 capture and storage LCA model developed at Imperial College and uses the pulverised coal post-combustion capture example to demonstrate the methodology in detail. At first, the LCA models developed for the coal combustion system and the chemical absorption CO2 capture system are presented together with examples of relevant model applications. Next, the two models are applied to a plant with post-combustion CO2 capture, in order to compare the life cycle environmental performance of systems with and without CO2 capture. The LCA results for the alternative post-combustion CO2 capture methods (including MEA, K+/PZ, and KS-1) have shown that, compared to plants without capture, the alternative CO2 capture methods can achieve approximately 80% reduction in global warming potential without a significant increase in other life cycle impact categories. The results have also shown that, of all the solvent options modelled, KS-1 performed the best in most impact categories. ß 2009 Elsevier Ltd. All rights reserved.

Keywords: Life Cycle Assessment Fossil fuel power generation CO2 capture

1. Introduction The world relies heavily on carbon-containing fuels and virtually all the energy produced from carbon-containing fuels is generated by directly burning fuels in air. Approximately 86% of global energy utilised and about 75% of current anthropogenic CO2 emissions originate from fossil fuels. The total emissions from fossil fuel use and flaring of natural gas were 24 GtCO2 per year (6.6 GtC) in 2001 (IPCC, 2001, 2005), which is considered to be the dominating contributor to the global warming. Carbon dioxide capture and storage (CCS) is a process consisting of separation of CO2 from industrial and energy-related sources, transport to a geological storage location, and long-term isolation from the atmosphere. Three alternative approaches can integrate CO2 capture technologies with power generation systems: post-

* Corresponding author. Tel.: +44 020 7594 7372; fax: +44 020 7594 7354. E-mail address: [email protected] (A. Korre). 1750-5836/$ – see front matter ß 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.ijggc.2009.08.005

combustion, pre-combustion and oxy-fuel combustion. The choice of specific capture method for a power generation system is determined largely by the power plant conditions. Geological storage of CO2 can be operated in a variety of geological settings in sedimentary basins, such as depleted oil or gas fields, oil fields, unminable coal seams, and saline aquifers, where CO2 can remain underground by a number of different trapping mechanisms. The CO2 capture technologies require significant amounts of energy during their implementation. For example, the increase in fuel consumption per kWh for plants that capture 90% CO2 by using best current technology ranges from 24 to 40% for new supercritical pulverised coal plants, 11–22% for Natural Gas Combined Cycle (NGCC) plants, and 14–25% for coal-fired Integrated Gasification Combined Cycle (IGCC) systems compared to similar plants without CCS facilities (IPCC, 2005). Furthermore, the increased consumption of fuels and chemicals, such as monoethanolamine (MEA) and limestone used for CO2 capture and flue gas desulphurisation, results in an increase in both onsite and upstream GHG emissions and other environmental

290

A. Korre et al. / International Journal of Greenhouse Gas Control 4 (2010) 289–300

impacts per kWh generated relative to plants without CO2 capture. The holistic perspective offered by Life Cycle Assessment (LCA) can help decision makers to quantify the trade-offs inherent in any change to the power production systems and ensure that a reduction in GHG emissions does not result in increases in other environmental impacts. The other strength of LCA is that the International Organization for Standardization (ISO) has developed the ISO 14040 series of LCA standards, which provide guidance on setting appropriate system boundaries, reliable data collection, evaluating environmental impacts, interpreting results, and reporting in a transparent manner. This offers an excellent starting point for the development of measurement protocols for greenhouse gases (GHGs) and other environmental impacts (Brady, 2000). Finally, under the Kyoto Protocol, three flexible mechanisms (Emissions Trading Scheme (ETS), Joint Implementation (JI) and the Clean Development Mechanism (CDM)) were developed to help emitters in developed countries to meet their GHG emission targets. As a credible and internationally accepted tool, LCA offers the means to include CO2 capture projects into the CDM framework and help the participants of flexible mechanisms to assess their proposed CO2 capture projects and verify their emission reductions from a life cycle perspective. Previous LCA studies (Akai et al., 1997; Fiaschi et al., 2000; Doctor et al., 2001; CISS, 2003; Lombardi, 2003; Corti and Lombardi, 2004; Pehnt and Henkel, 2009; Koornneef et al., 2008) have investigated power generation plants with alternative CO2 capture systems and concluded that CO2 capture can reduce CO2 emissions by around 80% throughout the life cycle of power generation. However, with respect to other environmental impacts, such as abiotic resource depletion, acidification, human toxicity, etc., previous studies report a fairly wide range of results, as they focus on specific CO2 capture cases only. Moreover, previous studies and commercial LCA software (e.g. TEAM, GaBi 4, Ecoinvent, SimaPro 7, ETH-ESU 96, and U.S. LCI Database) have a very rigid approach to system boundaries and do not recognise the importance of the level of detail that should be included in the Life Cycle Inventory (LCI) data. So far, the LCI data considered for power generation have all been at plant level (or gate-to-gate data) rather than being at the level of unit processes inside the plants. The gateto-gate data used in previous studies imply that the electricity generation systems have been largely simplified to a single blackbox with constants and linear coefficients used as inputs and outputs, covering a broad range of technological and geographical differences, in which the actual variability of process parameters and operating conditions are implicitly neglected. The LCI models at plant level restrict the flexibility of the model and cannot account for the technical and geographical differences in power generation with CCS, which can be configured with alternative CO2 capture routes; use a variety of CO2 capture technologies, and be located in different geographical areas that determine the emission control options, waste treatment methods, as well as the type of fuels used. In addition, plant level analysis limits the capacity of the model to quantify the trade-offs inherent in any change to the power production systems. Consequently, these simplifications limit the possibility of tracing emissions back to individual unit processes and restrict one’s ability to eliminate highly polluting options by design. The main objective of the research described in this paper was to develop a complete and dynamic LCA model which includes fossil fuel power generation, CO2 capture, transport and storage, quantifies the environmental impacts at the highest level of detail, and allows for the assessment of technical and geographical differences between the alternative CO2 capture and storage technologies considered. The LCI database developed models the inputs and outputs of the processes at component or unit process

level and aims at generating reliable and precise LCI data in a consistent and transparent manner with a clearly arranged and flexible structure for long-term strategic energy system planning and decision-making. The following sections present the principles of the LCA model developed and use the pulverised coal postcombustion capture with chemical absorption example to demonstrate the methodology developed in detail. 2. Life Cycle Assessment methodology and its application in carbon dioxide capture and storage Life Cycle Assessment is a compilation and evaluation of the inputs and outputs and the potential environmental impacts of a product system throughout its entire life cycle, from raw material extraction and acquisition, through energy and material production and manufacturing, to use and end of life treatment and final disposal (ISO 14040, 2006). In order to deal with the complexity of LCA, the International Standards Organization (ISO) established a methodological framework for performing an LCA study, which comprises four phases, including Goal and Scope, Life Cycle Inventory (LCI) analysis, Life Cycle Impact Assessment (LCIA) and Interpretation. Goal and Scope definition states the aim of an intended LCA study, the system boundary, the functional unit, competing systems considered, and the breadth and depth of (or level of detail) the LCA study in relation to this aim. Life Cycle Inventory Analysis is the phase to quantify the input/output relationships and to prepare an inventory of input/output data for all component processes involved in the life cycle of the system(s) under study. The input/output flows for a unit process to be quantified include economic and environmental flows (Fig. 1). For every unit process, LCI generates a unit process table (matrix), illustrated in Eq. (1), in which the variables represent the changes of economic flows or environmental interventions, relating to the functional unit. The total process along the life cycle of a product can be represented as a set of column vectors (Fig. 1, Eq. (2)) (Guine´e et al., 2001). The objective of Life Cycle Impact Assessment is to understand and evaluate the magnitude and significance of the potential environmental impacts of a product system (ISO 14040, 2006). In this phase, impact categories (e.g. global warming, acidification, and human toxicity), category indicators, and characterisation factors are defined, and LCI results are assigned to categories and then converted into category indicators via characterisation factors. Characterisation factors can convert environmental flows into environmental impacts. There are two characterisation approaches: the midpoint method (e.g. Guine´e et al., 2001) and the endpoint method (e.g. Goedkoop and Spriensma, 2001). The midpoint approach aims at quantitative modelling at a point before the end of the cause–effect chain (fate, exposure, effect and damage) and uses potential impact midpoint indicators (such as global warming potential, acidification, etc.) to reflect the relative environmental importance of an emission or resource extraction. The endpoint method, in addition to the midpoint potential impacts, finally models the complete cause–effect chain including the final potential environmental damages to human health, ecosystems and resources. In view of the ongoing discussions on appropriate characterisation factors to be used for the endpoint categories (van Zelm et al., 2009; Goedkoop et al., 2009), the sensitivity of the results to the modelling assumptions and the fact that there are established ISO standards for the midpoint potential impact categories, it was decided to use the midpoint method of characterisation in this study. Interpretation is the phase in which the findings of either the inventory analysis or the impact assessment, or both, are analysed in relation to the defined Goal and Scope in order to deliver conclusions, explain limitations and provide recommendations (ISO 14040, 2006).

A. Korre et al. / International Journal of Greenhouse Gas Control 4 (2010) 289–300

291

Fig. 1. A schematic representation of environmental interventions and economic flows used in LCA (after Guine´e et al., 2001).

One of the objectives of the research presented in this paper was to develop a comprehensive LCI database for the analysis of power generation with alternative CO2 capture and storage options in a consistent and transparent manner. The underlying principle applied in developing this methodology can be summarised as: 1. Transparency: To show precisely how life cycle impacts are calculated and the extent to which the inputs/outputs of any unit process have been quantified. 2. Comprehensiveness: To identify all of the inputs/outputs that may give rise to significant environmental impacts. 3. Consistency of methodology: To present models and assumptions so as to allow for valid comparisons between technological or operational options for each unit process. The system boundaries of LCA in power generation with CO2 capture, a generalised outline of which is presented in Fig. 2, covers power generation, alternative CO2 capture options, and upstream processes such as extraction and production of fossil fuels, raw materials production, as well as gas compression, transport and storage. In this paper, however, only the LCA of post-combustion CO2 capture technology is presented. The functional unit was selected as 1 MWh of electricity generated, considering that this is the product of the power generation system. In the case of the upstream processes, the LCI data used in this study are based on material from the literature. Power generation with CO2 capture, transport and storage involves a set of inter-linked component unit processes and the relationships between them. In the methodology the authors developed, the systems were broken down or modularised into

subsystems or component unit processes connected by flows of intermediate products or emissions as illustrated in Fig. 3 for the post-combustion CCS system. The purpose of modularisation was to make complex systems more easily understood and more accurately modelled. Through modularisation, the LCI models developed quantify flows of materials, natural resources, energy, intermediate products and emissions at component unit process level. This approach makes sure that the technical, spatial and temporal differences that exist between different industrial sites and operations can be accounted for by modifying certain parameters of the component unit processes as necessary. Furthermore, modularisation allows plant operators and designers to model and compare different technical and engineering scenarios from a life cycle perspective. Ultimately, modularisation eliminates the limitations introduced by the linear input/output coefficients used by conventional LCI models. The flexible structure of the LCI database provided through modularisation enables the practitioner to choose component unit processes so that different technological options can be considered without the need for redesign or loss of information (Durucan et al., 2006). The models reported in the literature focus on the LCA study of existing plants and do not offer methods for novel systems that are not commercially operated. Therefore, the modularisation methodology presented in this paper provides the option to conduct LCA by configuring virtual systems of power generation with CO2 capture, based on the best available techniques and component unit processes of the system. As presented in Table 1, this research employed the CML 2001 baseline impact categories, category indicators, and characterisation methods for the LCIA (Guine´e et al., 2001).

Fig. 2. System boundaries of LCA in power generation with CCS.

292

A. Korre et al. / International Journal of Greenhouse Gas Control 4 (2010) 289–300

Fig. 3. The level of detail involved in the LCA of a post-combustion CCS system.

3. Life cycle modelling of power generation with postcombustion carbon dioxide capture Post-combustion systems separate CO2 from the flue gases produced by combustion of a primary fuel (coal, natural gas, oil or biomass) in air. The fraction of CO2 present in the flue gas streams is typically 3–15% by volume, and the other main constituent in flue gas is nitrogen. Due to the low concentration and low pressure of CO2 in the flue gas, chemical absorption CO2 capture methods are conveniently applicable to post-combustion systems. Main systems of reference for post-combustion capture are the current installed capacity of 2261 GWe of coal, natural gas and oil power plants worldwide and, in particular, the 155 GWe of supercritical pulverised coal-fired plants and the 339 GWe of Natural Gas Combined Cycle (NGCC) plants, both representing the types of high efficiency power plant technology where CO2 capture can be best applied (IPCC, 2005). Carbon dioxide chemical or physical absorption systems are welldeveloped technologies that have been applied to numerous commercial processes, including natural gas treatment and ammonia production. However, when capturing CO2 from power plants the improvement of solvent performance is desirable to reduce equipment size and the energy required for sorbent regeneration. As shown in Fig. 4, a post-combustion CO2 capture system includes the component unit processes of coal combustion, particulate matter (PM) removal, desulphurisation, CO2 capture and CO2 compression. The specifications of the base case conventional power plant without CO2 capture and the same power plant with alternative post-combustion CO2 capture methods used in this study is presented in Table 2. In this section, the coal combustion and chemical absorption CO2 capture LCI models are presented as examples to demonstrate the methodology used in quantifying the inputs and outputs at unit process level. The Interpretation of LCI and LCIA results are discussed in detail.

3.1. Coal combustion LCI model development and LCIA results The coal combustion LCI model calculates mass flows (including coal, CO2, N2, O2, and H2O), energy consumption and heat (steam) output using stoichiometric reactions and engineering models. These methods are based on chemical and physical principles and estimate emissions (CO, SO2, SO3, NOx, HCl, HF, PM and trace metals) using the US EPA AP-42 emission factors (US EPA, 1998). In order to characterise the technological and geographical differences that exist between the coal combustion processes considered, the LCI model developed allows for the use of six types of boilers and different coal characteristics as input. A schematic representation of the coal combustion LCI model is shown in Fig. 5. A fixed plant size of 500 MW was used throughout the case studies reported. Plant gross efficiency of new conventional power plants at subcritical steam conditions is around 40% (IEA, 2004). Advanced supercritical steam plants can attain efficiencies that exceed 45% and, as further developments take place, the ultrasupercritical plants may ultimately achieve efficiencies of 50–55% (IEA, 2003). This research considered the internationally quoted efficiency ratings from the literature, at the same time, allowance was made for geographical differences in gross efficiencies of existing plants and the average regional gross efficiencies of subcritical plants in various regions across the world were taken as reference. As shown in Table 3, the average regional gross efficiency of individual subcritical plants ranges from 27 to 40%. Boiler operation conditions used in this research include availability and capacity utilisation rate of the plant, which also influence the plant efficiency. Modern power plants, both subcritical and supercritical, are capable of achieving a very high availability, typically ranging from 88 to 92%, and the majority of the lost availability is attributed to planned outages of about 4 weeks a year (IEA, 1995). Therefore, this research assumed that the availability of a power plant is 90%. The average capacity utilisation rate of OECD power plants is only about 50%. This rate is higher in developing countries as they are unlikely to have any surplus generation

A. Korre et al. / International Journal of Greenhouse Gas Control 4 (2010) 289–300

293

Table 1 Life cycle impact categories and characterisation factors (compiled after Guine´e et al., 2001 and US EPA, 2006). Definition Impact category Global warming Refers to the impact of anthropogenic emissions which enhance the radiative forcing of the atmosphere, causing the temperature at the earth’s surface to rise.

Stratospheric ozone depletion Refers to the thinning of the stratospheric ozone layer due to anthropogenic emissions, which causes a greater fraction of solar UV-B radiation to reach the earth’s surface, with potentially harmful impacts on human/ animal health, terrestrial and aquatic ecosystems, biochemical cycles and materials. Acidification Refers to the acidifying pollutants’ potential impacts on soil, groundwater, surface waters, biological organisms, ecosystems and materials. Eutrophication Refers to the potential impacts of excessively high environmental levels of macronutrients, which may cause an undesirable shift in species composition and elevated biomass production in both aquatic and terrestrial ecosystems. Photo-oxidant formation Refers to the formation of reactive chemical compounds, such as ozone, by the action of sunlight on certain primary air pollutants.

Ecotoxicity Refers to the potential impacts of toxic substances on aquatic, terrestrial and sediment ecosystems.

Relevant LCI data

Common characterisation factor

Carbon dioxide (CO2), nitrogen dioxide (NO2), methane (CH4), chlorofluorocarbons (CFCs), hydrochlorofluorocarbons (HCFCs), methyl bromide (CH3Br)

Global warming potential: converts LCI data to carbon dioxide (CO2) equivalents. Note: global warming potentials can be considered for 50, 100, or 500 years.

Chlorofluorocarbons (CFCs), hydrochlorofluorocarbons (HCFCs), halons, methyl bromide (CH3Br)

Ozone depleting potential: converts LCI data to trichlorofluoromethane (CFC-11) equivalents.

Sulphur oxides (SOx), nitrogen oxides (NOx), hydrochloric acid (HCl), hydroflouric acid (HF), ammonia (NH3)

Acidification Potential: Converts LCI data to hydrogen (H+) ion equivalents.

Phosphate (PO4), nitrogen oxide (NO), nitrogen dioxide (NO2), nitrates, ammonia (NH3)

Eutrophication potential: converts LCI data to phosphate (PO4) equivalents.

Non-methane hydrocarbon (NMHC)

Photochemical oxidant creation potential: converts LCI data to ethane (C2H6) equivalents.

Toxic chemicals with a reported lethal concentration to rodents or to fish.

LC50: converts LC50 data to equivalents.

Global

Global

Regional, local

Local

Local

Local

Human toxicity Covers the potential impacts on human health of toxic substances present in the environment.

Global, regional, local Total releases to air, water, and soil.

LC50: converts LC50 data to equivalents.

Depletion of abiotic resources Refers to the depletion of natural resources (including energy resources) which are regarded as non-living.

Scale

Global, regional, local Quantity of minerals used, quantity of fossil fuels used.

capacity due to rapid growth in energy demand in these countries (IEA, 1999). On the other hand, CO2 capture only makes sense for new power plants because of their high thermal efficiency (IEA, 2004) and new coal-fired power plants are normally intended for base load operation. Based on this analysis, it was assumed that plants with CCS are for base load operation and the utilisation rate is 75 and 80% for industrialised countries and developing countries, respectively. As the efficiencies of coal combustion based plants do not fall off significantly at part load (IEA, 1995), it is assumed that there is no efficiency loss for power plants at part load. In many parts of the world, the choice of coal is limited to that produced locally. This results in a geographical difference in emissions from power generation. For instance, in China coal-fired power plants almost entirely depend on indigenous coals with an average sulphur content of 1.3%. On the other hand, most power plants in Germany fire on lignite. In order to represent the

Resource depletion potential: converts LCI data to a ratio of quantity of resource used versus quantity of resource left in reserve.

characteristics of the coal at a regional level, the model developed considers the composition and heating value of coal when calculating the mass flow of coal, air requirements and environmental emissions. Trace metals are also emitted during coal combustion. The quantity of any given metal emitted depends on the physical and chemical properties of the metal, the concentration of the metal in the coal, and the combustion conditions (US EPA, 1998). Some trace metals become concentrated in certain particle streams from a combustor (e.g., bottom ash, collector ash, and flue gas particulate) while others do not. Generally, trace element partitioning behaviours are classified into three groups (Lee and Lesley, 1992):  Group 1: Elements that are approximately equally concentrated in the fly ash and bottom ash, or show little or no small particle enrichment.

A. Korre et al. / International Journal of Greenhouse Gas Control 4 (2010) 289–300

294

Fig. 4. Schematic of a post-combustion CO2 capture system in a power plant (after Stro¨mberg, 2006). Table 2 Specifications of the base case conventional power plant without CO2 capture and the same power plant with the alternative post-combustion CO2 capture methods assessed in this study. Base case:

Power plant without CO2 capture

Power plant capacity Power plant gross energy efficiency (LHV) Boiler type Chemical absorption CO2 capture technology CO2 capture rate Particulate matter removal method Particulate matter removal rate NOx removal method NOx removal rate (NH3 to NO ratio) SOx removal method SOx removal rate Coal type

500 MW 40% PC wall-fired, dry bottom N/A N/A

a

95%

Power plant with post-combustion CO2 capture

500 MW 40%a PC wall-fired, dry bottom MEA, K+/PZ or KS-1 95% Electrostatic precipitator 99.97% Selective catalyst NOx reduction 0.8 Limestone based flue gas desulphurisation 99% US Appalachian (bituminous)

Refers to efficiency before the CO2 capture process, same value as the gross energy efficiency of a conventional plant.

Fig. 5. The schematic representation of a coal combustion process LCI model.

 Group 2: Elements that are enriched in fly ash relative to bottom ash, or show increasing enrichment with decreasing particle size.  Group 3: Elements which are emitted in the gas phase.

Emissions of 19 trace metals are calculated using emission factor equations proposed by the US EPA (1998). The pulverised coal combustion LCI model was run for three different boilers (wall-fired dry bottom, wall-fired wet bottom, cyclone) using three

A. Korre et al. / International Journal of Greenhouse Gas Control 4 (2010) 289–300 Table 3 Average regional efficiencies for coal-fired power plants (after IEA, 2004). Regions

Efficiency (%)

Africa Australia/New Zealand China Central and South America Eastern Europe India Japan Middle East Mexico Other Asia South Korea USA/Canada Western Europe

35 38 35 35 27 28 38 40 33 36 36 39

Note: Based on gross efficiency (LHV basis).

different coal types, the characteristics of which are listed in Table 4. Fig. 6 presents the LCI results for a wall-fired dry bottom pulverised coal (PC) boiler burning bituminous coal for 1 MWh electricity generated and demonstrates the level of detail achieved in the inputs and outputs. Fig. 7 presents the LCIA results of a combustion process using these different coals as inputs and with different types of boilers used. Figs. 6 and 7 illustrate the results for the combustion unit process only, not including the downstream emission control processes. The LCIA results show that, in general, coal type is the dominant factor in determining the environmental impact potential. The global warming potential (GWP) is mainly influenced by the type of coal used, and lignite combustion has the highest value in GWP due to the low heating value of lignites. Five combustion alternatives have the same value in depletion of abiotic resources since the boilers used had the same energy efficiency and consume same amount of energy for the generation of 1 MWh electricity. Human toxicity and ecotoxicity are mainly caused by trace metals in particulate matters or bottom ash, which depend on the type of coal burnt, and sub-bituminous coal combustion has the highest

295

Table 4 Chemical composition and heating values of the three different coal types used in the coal combustion LCI model runs. Coal type

Carbon (C) Hydrogen (H) Oxygen (O) Chlorines (Cl) Sulphur (S) Nitrogen (N) Ash/inerts Moisture (H2O) Higher heating value (MJ/kg) Lower heating value (MJ/kg)

Bituminous, Appalachian (% by weight)

Sub-bituminous, Powder River Basin (% by weight)

Lignite, North Dakota (% by weight)

71.40 4.62 6.09 0.07 0.64 1.42 9.79 5.63 32.07

48.00 3.40 11.00 0.03 0.05 0.62 6.40 30.00 18.64

39.70 2.70 11.40 0.02 0.07 0.60 7.80 37.00 15.34

29.19

17.90

14.72

value in these impact categories because of its highest particulate matter emission factor and low heating value. Sub-bituminous coal combustion and lignite combustion have the lowest values in photo-oxidant formation and acidification than bituminous coal combustion, because they have lower sulphur contents. Eutrophication impact is dependent on the type of boiler used. Cyclone boilers and wall-fired wet bottom PC boilers have the highest eutrophication values, due to their higher NOx emission factors compared to other types of boilers. 3.2. Chemical absorption carbon dioxide capture LCI model development and LCIA results In the chemical absorption CO2 capture LCI model developed, the inputs and outputs are calculated using engineering models. The model developed assumes a default efficiency value of 95% for the amine-based CO2 capture systems, however, the user can also

Fig. 6. LCI results of coal combustion in a PC wall-fired dry bottom boiler per 1 MWh electricity generated (where Hg0 is elemental mercury, Hg2+ is oxidised mercury and Hgp is particle-bound mercury).

296

A. Korre et al. / International Journal of Greenhouse Gas Control 4 (2010) 289–300

Fig. 7. LCIA results of coal combustion with alternative coal types and boilers per 1 MWh electricity generated (1,4-DB is 1,4-dichlorobenzene).

specify a desired level of CO2 capture efficiency. The schematic representation of the LCI model is shown in Fig. 8. In order to account for the technological differences that exist between alternative chemical absorption CO2 capture processes, the LCI model developed was set up for eight different solvents (Table 5).

In chemical absorption CO2 capture, solvent regeneration requires considerable energy. Previous research with a pilot CO2 capture plant have shown that, for a typical MEA solvent, sensible heat to bring the solvent to its boiling point, heat of desorption of CO2, and the evaporation heat of dilution water account for 49, 27

Fig. 8. A schematic representation of chemical absorption CO2 capture processes LCI model developed.

A. Korre et al. / International Journal of Greenhouse Gas Control 4 (2010) 289–300 Table 5 Solvent alternatives used in the chemical absorption CO2 capture LCI model developed. Solvent types

Solvent alternatives

Primary amine Secondary amine Tertiary amine Hindered amine Promoted potassium carbonate

MEA, DGA DEA, DIPA MDEA, TEA KS-1 K+/PZ

and 24% of the total heat duty, respectively (Dugas et al., 2006). The heat required for solvent regeneration is mainly determined by the type of solvents used and the solvent concentration, ranging from 2200 to 6000 kJ/kg CO2 captured. The heat required for sorbent regeneration is normally extracted from the steam turbine in the form of low pressure (LP) steam in coal-fired power plants and combined cycle gas plants using a reboiler (a heat exchanger). In this research, the heat requirements of alternative solvent regeneration processes were taken from the literature (Go¨ttlicher, 2004; Mimura et al., 2000). The heat-to-electricity ratio represents the equivalent loss of power generation capacity due to the extraction of steam for sorbent regeneration. Literature provides a wide range of heat-to-electricity ratios for LP steam, ranging from 0.14 to 0.25 (Go¨ttlicher, 2004; Rao et al., 2004; Feron, 2005), for different types of turbines and plant configurations. In the model developed, 0.14 was used for heat-to-electricity ratio for steam turbine power plants with currently available power generation technologies and a ratio of 0.19 was used for the future advanced power plants (Go¨ttlicher, 2004; Feron, 2005). Power consumption in the aqueous absorption/stripping unit includes the flue gas blower power, the direct contact cooler (DCC) circulation pump power (if used), and absorber wash water pump power, which are all inversely proportional to the CO2 concentration in the flue gas, and range from 0.2 to 1.3 MJ/kg CO2 removed (Bolland and Mathieu, 1998; Go¨ttlicher, 2004). In the model developed, it is assumed that oxidation and carbamate polymer-

297

Table 6 Reclaimer waste composition and properties (after Nsakala et al., 2001). Substance/property

Content/value

Amine NH3 NaCl Na2SO4 Na2CO3 Insolubles Total nitrogen Total organic carbon pH Specific gravity

9.5 wt.% 0.02 wt.% 0.6 wt.% 6.6 wt.% 1.7 wt.% 1.3 wt.% 5.6 wt.% 15.6 wt.% 10.7 1.14

isation account for 40 and 60% of the total solvent loss, respectively (Rochelle, 2001). The total MEA loss due to degradation is conservatively estimated at around 1.5 kg MEA/tonne CO2 captured (Chapel et al., 1999; White et al., 2003). A CO2 capture system produces three waste streams: reclaimer waste, spent carbon from the activated carbon filters, and filter elements from the slip-stream solvent filters at the carbon bed. The actual amount of amine waste generated in the process depends on flue gas composition and plant operating conditions. For reclaimer waste, the model uses a typical value of 3.2 kg/tonne of CO2 captured (IEA GHG, 2006). The composition of the reclaimer waste is taken from the literature, as presented in Table 6 (Nsakala et al., 2001). The amount of spent carbon from activated carbon is assumed to equal the activated carbon consumption. As the filter elements mainly contain flue gas particulates, it is assumed that the amount of filter elements is equal to the flue gas particulates removed by the amine unit. The wastes from the capture system can be disposed of by incineration, and then treated as solid waste. The solid waste incineration is not included within the LCA system boundaries, as the amount of solid wastes associated with the functional unit used (1 MWh electricity generated) is small and may be neglected according to the LCA 1% emissions cut-off rule implemented in this study.

Fig. 9. LCI results of a MEA CO2 capture system per 1 MWh electricity generated (where Hg0 is elemental mercury, Hg2+ is oxidised mercury and Hgp is particle-bound mercury).

298

A. Korre et al. / International Journal of Greenhouse Gas Control 4 (2010) 289–300

In order to compare the life cycle performance of solvents, the chemical absorption LCI model was applied using three solvents, namely the MEA, KS-1 and K+/PZ. Fig. 9 presents the LCI results obtained for a MEA CO2 capture system with reference to 1 MWh electricity produced, using the LCI model described above. To complete the LCI data for the chemical absorption capture process, the LCI emissions corresponding to electricity consumption are added as upstream LCI data. These are calculated using the inhouse LCA model for power generation without capture. The solvent production LCI emissions are neglected according to the LCA 1% cut-off rule as the consumption of solvent is small. The LCIA results evaluated for the chemical absorption CO2 capture process using these three solvents have shown that CO2 capture process can significantly reduce the life cycle GWP, and KS-1 process can reduce more life cycle greenhouse gases because it consumes less energy and hence causes less greenhouse gas emissions related to energy production. Since CO2 capture processes can remove particulate matter and acid gases from the flue gas, these processes result in reduced values of life cycle human toxicity and photo-oxidant formation impacts. The removal of acid gases also reduces acidification and eutrophication and results in negative values for acidification and eutrophication for MEA and K+/PZ processes, when the capture process is considered

alone. The KS-1 process generates more NH3, which contributes to increased acidification and eutrophication impacts. Depletion of abiotic resources and ecotoxicity impacts are increased for all capture processes and are related to energy consumption, which in turn results in increased fossil fuel consumption and higher ecotoxicity related emissions. The LCI model presented here does not consider the loss of amine solvent which occurs from vaporisation and mechanical sources and is directly emitted to air since the volume emitted is considered to be relatively small in LCA terms (below the LCA 1% cut-off). However, it is worthy to note that amine emissions to air and further degradation of its compounds have adverse impacts on vegetation and human health. 3.3. Life Cycle Impact Assessment results for power generation with post-combustion carbon dioxide capture The LCIA results reported in Fig. 10 were calculated for coal power generation systems utilising alternative post-combustion CO2 capture technologies using the LCI models developed. The results show that, compared to power generation without CO2 capture, the post-combustion CO2 capture using different solvents (MEA, KS-1 and K+/PZ) can achieve 80% reduction in greenhouse

Fig. 10. LCIA results for alternative power generation systems with and without CO2 capture (per 1 MW electricity generated) for PC wall-fired, dry bottom boiler burning US Appalachian coal.

A. Korre et al. / International Journal of Greenhouse Gas Control 4 (2010) 289–300

gas emissions, without a significant increase in other environmental burdens throughout the whole power generation life cycle. Power generation with CO2 capture systems yield higher values for depletion of abiotic resources and ecotoxicity since these plants require more coal and limestone for the same amount of energy produced and generate more bottom ash, fly ash and solid wastes. The disposal of ash and solid wastes causes ecotoxicity impacts. With respect to human toxicity, power generation with CO2 capture systems have lower values, as CO2 capture process can further reduce fly ash in the flue gas after the flue gas desulphurisation (FGD) process and this reduces the emissions of trace metals in the fly ash. However, captured fly ash is discharged in water or soil, and this shifts human toxicity impacts to ecotoxicity impacts. The CO2 capture process can also further reduce acid gases such as SOx, NOx, HCl and HF in flue gases after the FDG process, and hence power generation with CO2 capture systems do not increase the impacts of acidification, eutrophication, and photo-oxidant formation considerably throughout the life cycle. MEA capture system has the highest values in these three impact categories since the MEA capture processes consume more coal and the upstream processes of coal mining contribute to these three impacts. KS-1 capture has more benefits in most impact categories than MEA and K+/PZ, but KS-1 has a slightly higher value for acidification and eutrophication than K+/PZ, as KS-1 capture process generates more NH3. 4. Conclusions This paper described the development of a dynamic LCA framework for the ‘‘cradle-to-grave’’ assessment of alternative CCS technologies in fossil fuel power generation. The application of the LCA model developed is demonstrated using the post-combustion capture route with chemical absorption technology. The LCI models developed quantify flows of materials, natural resources, energy, intermediate products or emissions at component unit process level, based on fundamental physical/chemical principles or empirical relationships which, to a greater extent, account for the technological, spatial and temporal characteristics of the power generation systems in consideration. This approach not only addresses the limitations of conventional LCI models that use linear input/output coefficients, but also facilitates the screening of technological options in order to improve the life cycle environmental performance of a power generation system with CCS. The development of the LCI models at component unit process level, and the use of fundamental physical/chemical principles in the calculations, have improved the ability of the LCI models to handle the complexity of fossil fuel power generation systems and reduced the LCA model uncertainty. The models referred to in the literature address LCA needs of the existing power generation plants, however, they do not offer solutions for novel systems that are not commercially operated. The LCI methodology developed in this research provides an innovative and robust approach for conducting LCA for novel systems by configuring virtual systems at unit process level. The LCI models developed included the models of conventional coal combustion, the use of alternative solvents for chemical absorption CO2 capture, plant level air pollution control units (including particulate matter emissions control, NOx emissions control and SOx emissions control), and solid waste disposal. At plant emissions level, the post-combustion chemical absorption (MEA) CO2 capture unit captures 95% of the CO2 and has lower emission rates of PM-10, SO2, SO3, NO2, HCl, HF, Hg vapor and trace metals than that of a conventional power plant. Trace metal emissions to soils occur mainly at surface impoundments, and trace metal emissions to soils from landfills are not significant.

299

The LCIA results have shown that lignite and sub-bituminous coal combustion yield a relatively higher value in global warming potential (GWP) due their low heating value. Comparison of the three alternative boiler types used in coal combustion has shown that eutrophication impact is dependent on the type of boiler used. Cyclone boilers and PC wall-fired wet bottom boilers cause the highest eutrophication impacts due to their higher NOx emissions. Human toxicity and ecotoxicity are mainly caused by trace metal emissions, which depend on the type of coal burnt, and subbituminous coal combustion has the highest value in these impact categories. Sub-bituminous coal combustion and lignite combustion have the lowest values in photo-oxidant formation and acidification than bituminous coal combustion, because they have lower sulphur contents. These differences support the LCA modelling approach developed in this research at unit process level covering a wide range of emissions besides CO2 and major fossil fuel combustion emissions. The LCIA results of chemical absorption capture have shown that the KS-1 process reduces a relatively larger volume of life cycle greenhouse gases. By removing particulate matter and acid gases from the flue gases, the CO2 capture processes also reduces the life cycle human toxicity and photo-oxidant formation impacts. Depletion of abiotic resources and ecotoxicity impacts are increased for all capture processes. The LCIA results for pulverised coal power generation with post-combustion CO2 capture have shown that, compared to systems without capture, the CO2 capture process can achieve 80% reduction in greenhouse gas emissions, without a significant increase in other environmental burdens. The reduction of CO2 emissions through the implementation of post-combustion capture systems results in an increase in abiotic resource use. Therefore, one may question whether it is worth capturing and storing CO2 from power generation, which leads to further depletion of fossil fuel resources. This question can be answered by the comparative evaluation of the impacts of climate change against the impacts of abiotic resource depletion. The IPCC Fourth Assessment Report (2007) concludes that the adverse impacts of climate change on natural and human systems (such as altered frequencies and intensities of extreme weather, sea level rise, etc.) will be severe, irreversible and uncertain with a multicentury time scale. Furthermore, it is possible that dealing with the consequences of unabated climate change impacts would also lead to the depletion of abiotic resources in the long-term. Nevertheless, it is believed that the new and advanced energy technologies that are currently being developed will address the potential for fossil fuel shortages in the future. In this context, it is reasonable to take action now and stabilise the climate systems by reducing CO2 emissions, even if this process uses slightly more fossil fuel resources, which are expected to be somewhat less important in the medium and long-term future. References Akai, M., Nomura, N., Waku, H., Inoue, M., 1997. Life-cycle analysis of a fossil-fuel power plant with CO2 recovery and a sequestering system. Energy 22, 249–255. Bolland, O., Mathieu, P., 1998. Comparison of two CO2 removal options in combined cycle power plants. Energy Convers. Manage. 39 (16–18), 1653–1663. Brady, K., 2000. LCA and Global Climate Change—Solving the Measurement Puzzle. Available at: http://www.fivewinds.com/uploadedfiles_shared/LCA_N_Climate_ Change.pdf. Chapel, D.G., Mariz, C.L., Ernest, J., 1999. Recovery of CO2 from flue gases: commercial trends. In: Proceedings of the Canadian Society of Chemical Engineers Annual Meeting, Saskatoon, Saskatchewan, Canada. Coal in Sustainable Society (CISS), 2003. Case Study B17: electricity from CO2 Recovery Type IGCC. Available at: http://www.ciss.com.au. Corti, A., Lombardi, L., 2004. Biomass integrated gasification combines cycle with reduced CO2 emissions: performance analysis and life cycle assessment (LCA). Energy 29, 2109–2124. Doctor, R.D., Molburg, J.C., Brockmeier, N.F., Lynn, M., Victor, G., Massood, R., Gary, J.S., 2001. Life-cycle analysis of a shell gasification-based multi-product system with CO2 recovery. In: Proc. of the 1st Nat. Conf. on Carbon Sequestration, Washington, D.C., USA.

300

A. Korre et al. / International Journal of Greenhouse Gas Control 4 (2010) 289–300

Dugas, R., Rochelle, G.T., Seibert, F., 2006. CO2 capture performance of a monoethanolamine pilot plant. In: The 8th International Conference on Greenhouse Gas Technology, Trondheim, Norway. Durucan, S., Korre, A., Munoz-Melendez, G., 2006. Mining life cycle modelling: a cradle-to-gate approach to environmental management in the minerals industry. J. Clean. Prod. 14, 1057–1070. Feron, P.H.M., 2005. Progress in post-combustion CO2 capture. In: European CO2 Capture and Storage Conference Towards Zero Emission Power Plants. Brussels, 13–15 April 2005 Available at: http://www.eu.int/comm/research/energy/pdf/ 14_1000_feron_en.pdf. Accessed at 12 July 2006. Fiaschi, D., Lombardi, L., Manfrida, G., 2000. Life cycle assessment (LCA) and exergetic life cycle assessment (ELCA) of an innovative energy cycle with zero CO2 emissions. In: The 5th Int. Conf. on Greenhouse Gas Control Technologies, Cairns. Goedkoop, M., Spriensma, R., 2001. The Eco-indicator 99 A Damage Oriented Method for Life Cycle Impact Assessment: Methodology Report, no. 1999/ 36A, 3rd edition. Pre Consultants b.v., Amersfoort, the Netherlands. Goedkoop, M., Huijbregts, M.A.J., Heijungs, R., De Schryver, A., Struijs, J., van Zelm, R., 2009.In: ReCiPe 2008—A Life Cycle Impact Assessment Method Which Comprises Harmonised Category Indicators at the Midpoint and the Endpoint Level. Go¨ttlicher, G., 2004. The Energetics of Carbon Dioxide Capture in Power Plants, Report, USDOE National Energy Technology Laboratory (NETL). Available at: http://www.netl.doe.gov/publications/carbonseq/refshelf.html. Guine´e, J.B., Gorre´e, M., Heijungs, R., Huppes, G., Kleijn, R., Koning, A., 2001. Life Cycle Assessment: An Operational Guide to the ISO Standards, Final Report, Centre of Environmental Science – Leiden University (CML). Intergovernmental Panel on Climate Change (IPCC), 2001. Climate Change 2001: The Scientific Basis. Cambridge University Press, Cambridge, UK. Intergovernmental Panel on Climate Change (IPCC), 2005. Carbon Capture and Storage. Cambridge University Press, Cambridge, UK. Intergovernmental Panel on Climate Change (IPCC), 2007. Climate Change 2007: Synthesis Report. Contribution of Working Groups I, II and III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Core Writing Team, Pachauri, R.K., Reisinger, A. (eds.)]. IPCC, Geneva, Switzerland, 104 pp. International Energy Agency (IEA GHG), 2006. Environmental Impact of Solvent Scrubbing of CO2, 2006/14. International Energy Agency (IEA), 1995. Industry Attitudes to Steam Cycle Clean Coal Technologies Survey of Current Status. Available at: http://www.iea.org/ dbtw-pd/textbase/nppdf/free/1990/ciab1995.pdf. International Energy Agency (IEA), 1999. Electric Power Technology: Opportunities and Challenges of Competition. Available at: http://www.iea.org/textbase/ nppdf/free/1990/elecpower99.pdf. International Energy Agency (IEA), 2003. Control and Minimisation of Coal-fired Power Plant Emissions. Available at: http://www.iea.org/dbtw-wpd/textbase/ papers/2003/CoalFiredFossilFuels.pdf.

International Energy Agency (IEA), 2004. Prospects for CO2 Capture and Storage. Available at: http://www.iea.org/textbase/nppdf/free/2004/prospects.pdf. ISO 14040 (E), 2006. Environmental Management – Life Cycle Assessment – Principles and Framework. International Organization for Standardization, Geneva. Koornneef, J., Keulen, T.V., Faaij, A., Turkenburg, W., 2008. Life cycle assessment of a pulverized coal power plant with post-combustion capture, transport and storage of CO2. Int. J. Greenhouse Gas Control 2 (4), 448–467. Lee, B.C., Lesley, L.S., 1992. Trace Elements—Emissions from Coal Combustion and Gasification. IEACR/49, IEA Coal Research, London. Lombardi, L., 2003. Life cycle assessment comparison of technical solutions for CO2 emission reduction in power generation. Energy Convers. Manage. 44, 93–108. Mimura, T., Matsumoto, K., Iijuma, M., Mitsuoka, S., 2000. Development and application of flue gas carbon dioxide recovery technology. In: 5th International Conference on Greenhouse Gas Control Technologies (GHGT-5). CSIRO Publishers, Cairns, Australia, ISBN 0 643 06672 1. Nsakala, N., Marion, J., Bozzuto, C., Liljedahl, G.N., Palke, s.M., Vogel, D., Gupta, J.C., Guha, M., Johnson, H., Plasynski, S., 2001. Engineering Feasibility and Economics of CO2 Capture on an Existing Coal-Fired Power Plant, Final Report, DOE National Energy Technology Laboratory, U.S. Pehnt, M., Henkel, J., 2009. Life cycle assessment of carbon dioxide capture and storage from lignite power plants. Int. J. Greenhouse Gas Control 3 (1), 49–66. Rao, A.B., Rubin, E.S., Berkenpas, M.B., 2004. An integrated modelling framework for carbon management technology (vol. 1), Final report, Centre for Energy and Environmental Studies, Carnegie Mellon University, Pittsburgh, PA. Available at: www.iecmonline.com/documentation/tech_04.pdf. Accessed 19 July 2005. Rochelle, G., 2001. CO2 capture by aqueous absorption/stripping opportunities for better technology. In: 2001 Conference Proceedings of Workshop on Carbon Sequestration Science. National Energy Technology Laboratory, Available at: http://www.netl.doe.gov/publications/proceedings/01/. . ./Rochell_Sep.pdf. Accessed 19 July 2005. Stro¨mberg, L., 2006. ENCAP Integrated Project, ENCAP CO2 Seminar in Billund, Denmark, 16 March 2006. United States Environmental Protection Agency (US EPA), 1998. AP 42, vol. I, 5th edition Available at: http://www.epa.gov. United States Environmental Protection Agency (US EPA), 2006. Life cycle assessment: principles and practice. Available at: http://www.epa.gov/ORD/NRMRL/ lcaccess/pdfs/lca101_allchapters.pdf. Accessed 11 February 2008. van Zelm, R., Huijbregts, M.A.J., van de Meent, D., 2009. USES-LCA 2.0—a global nested multi-media fate, exposure, and effects model. Int. J. Life Cycle Assess. 14, 282–284. White, C.M., Strazisar, B.R., Granite, E.J., Hoffman, J.S., Pennline, H.W., 2003. Separation and capture of CO2 from large stationary sources and sequestration in geological formations—coalbeds and deep saline aquifers. J. Air Waste Manage. Assoc. 53, 645–715.