• No results found

IMPLEMENTATION OF OXYFUEL COMBUSTION IN A WASTE INCINERATION CHP PLANT : A Techno-Economic Assessment

N/A
N/A
Protected

Academic year: 2021

Share "IMPLEMENTATION OF OXYFUEL COMBUSTION IN A WASTE INCINERATION CHP PLANT : A Techno-Economic Assessment"

Copied!
110
0
0

Loading.... (view fulltext now)

Full text

(1)

IMPLEMENTATION OF OXYFUEL

COMBUSTION IN A WASTE INCINERATION

CHP PLANT

A Techno-Economic Assessment

By

MOSTAFA SALEH

ANTON HEDÉN SANDBERG

MÄLARDALEN UNIVERSITY

DEPARTMENT OF BUSINESS, SOCIETY AND ENGINEERING Master Thesis in Industrial Engineering and Management

Costumer: Mälarenergi AB Supervisor: Hailong Li

(2)

ABSTRACT

Global energy demand is predicted to rise in the coming decades, necessitating a shift to re-newable energy sources to mitigate greenhouse gas emissions. However, due to the inability to supply renewable energy around the clock, it is estimated that only by adding an important technology, carbon capture and storage (CCS), it could be possible to reduce 80% of the 1990s greenhouse gas emissions. CCS aims to reduce anthropogenic carbon emissions by capturing CO2 from flue gases, transporting, and then permanently storing or reutilizing industrially. The CCS approach includes three technologies: post-combustion capture, pre-combustion cap-ture, and oxyfuel combustion, with the latter being the emphasis of this thesis. Based on the case study of Mälarenergi’s Refused-derived waste-fired CHP plant, this thesis investigates the viability of converting existing non-fossil fueled CHP plants to oxyfuel combustion. A thorough technical investigation based on analyzing the impact of oxyfuel combustion on system perfor-mance was conducted through system modeling using a process simulator, Aspen plus. The model in this thesis considers the development of an air separation unit (ASU), a CHP plant, and a cryogenic CO2 purification unit (CPU). All of which are validated through calibration and

comparison with real-world data and similar work. To investigate the influence of employing oxyfuel combustion on the generation of both heat and electricity, two different scenarios were comprised, including recirculating flue gas before and after flue gas condensation. In addition, an analysis of the oxygen purity was conducted to assess the most optimal parameters with the least impact on system performance. Moreover, a detailed economic assessment comprising the costs of integrating oxyfuel combustion was also conducted. The findings of this thesis show that integrating waste incineration CHP plants with oxyfuel combustion for CO2 capture,

entails promising features under the condition of 97% oxygen purity and a flue gas recircula-tion system taking place after flue gas condensarecircula-tion. This is owing to (i) modest imposed en-ergy penalty of approximately 8.7%, (ii) high CO2 recovery ratio, around 92.4%, (iii) total

in-vestment cost of approximately 554 M$ during a 20-year lifetime, and (iv) cost of captured CO2

of around 76 $/ton. Aside from system modeling, this thesis presents an overview of the cur-rent state-of-the-art technology on the diffecur-rent separation and capture mechanisms. It is im-portant to highlight that the goal of this thesis is not to provide a comprehensive review but rather to present an overall picture of the maturity of the different mechanisms. The findings point to the cryogenic separation mechanism as the most mature technology for both oxygen production and capturing of CO2 during oxyfuel combustion.

Keywords: CCS, Oxyfuel Combustion, Carbon Capture, Gas Separation Techniques, Waste

(3)

PREFACE

This project is a dissertation submitted in fulfillment of the requirements for the degree of Master of Science in Industrial Engineering and Management at Mälardalen University in Swe-den, written by Mostafa Saleh and Anton Hedén Sandberg during the spring semester 2021. The project was carried out in cooperation with the energy company Mälarenergi AB, in the city of Västerås. This work is also linked to the research project ‘AI assisterad koldiox-idinfångning i biomassabaserade kraftvärmeverk’ (Projektnr: 51592-1). The support from Energimyndigheten is acknowledged.

We would like to express our deep sense of thanks and gratitude to the following people whose contribution have been incredibly important in completing this dissertation.

Professor Hailong Li, for the support, guidance, and encouragement received throughout this project. Thank you for your outstanding supervisory work. Working with you has been a tre-mendous honor and privilege. Marianne Allmyr and Per Tunberg, our supervisors at Mälarenergi, thank you for believing in us and allowing us to participate in Mälarenergi’s fu-ture objectives. Also, thank you for your assistance and provision of all necessary data.

A dept of gratitude is also owed to the following people who have contributed with inspiration and rewarding discussions. Yukun Hu, Associate Professor at University College London, Awais Salman, PhD at Mälardalen University, Filip Johnsson, Professor at Chalmers Univer-sity, Paul Fennell,Professor of Clean Energy at Imperial College London, and our examiner Anders Avelin, Professor and Head of Department at Mälardalen University.

Last but not least, we would like to thank our families and friends, without your love and sup-port none of this would indeed be possible.

Västerås in June 2021

(4)

SAMMANFATTNING

Global uppvärmning är ett påtagligt fenomen som mellan åren 1880–2012 resulterade i en ökning med 0,85 °C av medeltemperaturen på jorden, till stor del på grund av förbränning av fossila bränslen. Under de kommande decennierna förväntas det globala energibehovet växa, vilket kräver en övergång till förnybara energikällor för att bromsa utsläppen av växthusgaser. Eftersom vissa förnybara energikällor inte går att nyttja dygnet runt estimeras det att carbon capture and storage (CCS) är en nödvändig teknik för att begränsa utsläppen av växthusgaser till en önskvärd nivå. CCS är en metod för att fånga och lagra koldioxid från förbränningspro-cesser, såsom kraftvärmeverk. Metoden är förekommande i tre olika former: före förbränning (pre-combustion), efter förbränning (post-combustion) och syrgasförbränning (oxyfuel com-bustion). I detta arbete undersöks det sistnämnda som är baserat på att förbränna bränsle i en miljö som består av syre och recirkulerade rökgaser. För att tillföra syre används vanligtvis en luftavskiljningsenhet (ASU) och efter förbränningen tas det hand om koldioxiden i rökgasen via en kompressions- och reningsenhet (CPU).

Detta arbete syftade till att undersöka olika fångningstekniker och innebörden av syrgasför-bränning av avfallsbränsle (oxy-waste försyrgasför-bränning) för kraftvärmeverk. För detta genomfördes en fallstudie på Mälarenergis Panna 6, för att utvärdera den tekno-ekonomiska genomförbar-heten av att eftermontera syrgasförbränning på anläggningen. Detta arbete avgränsades till att endast behandla fångningen av koldioxid och därmed exkluderades både transport och lagring. Tillvägagångssättet baserades på en litterär kunskapsinsamling om olika fångningstekniker, tekniska egenskaper och ekonomiska förutsättningar för syrgasförbränning. Steady-state si-muleringar för olika scenarios, såsom olika belastningar och typer av rökgasrecirkulering (torr och våt), utfördes i Aspen Plus och databehandling skedde i MS Excel. Tre modeller användes för simuleringar, föreställande: luftavskiljningsenhet (ASU), kraftvärmeverk (CHP plant), kompressions- och reningsenhet (CPU). Validering av modellerna skedde genom jämförelse, för kraftvärmeverket, med verkliga data erhållen från Mälarenergi, och, för ASU och CPU, med tidigare forskning. I den tekniska delen diskuterades viktiga parametrar som energibortfall, syrets renhet, rökgaskomposition och recirkulering av rökgas. I den ekonomiska delen utvär-derades kostnaden för den infångade koldioxiden.

Resultaten av detta arbete visar att det är möjligt att fånga upp till 92,4% av den koldioxid som produceras i samband med förbränning av avfallsbränsle i en miljö med 22 mol% respektive 26 mol% O2. Detta med en teknik för luftseparering samt rökgaskomprimering och rening

ba-serad på kryogen destillering. Drygt 90% av utsläppen av NOx från lufteldad förbränning är

möjliga att undvikas samtidigt som utsläppet av SOx ökar upp emot 30%. För att hålla liknande

förbränningstemperatur som vid traditionell förbränning, måste en del av rökgasen returneras till pannan, antingen före rökgaskondenseringen, som våt, eller efter, som torr. Den typ av rökgasrecirkulering som används, torr eller våt, påverkar flera aspekter i systemet. Bland annat innebär torr rökgasrecirkulering att 10% större andel av rökgasen förs tillbaka till pannan jäm-fört med våt recirkulering. Jämjäm-fört med traditionell avfallsförbränning bidrar syrgasförbrän-ning, med torr rökgasrecirkulering, till en betydligt mindre volym på rökgasen, som i sin tur kan innebära att en minskad storlek på utrustningen för att behandla rökgasen behövs

(5)

använ-Oxy-waste förbränning orsakar på grund av luftavskiljning och koldioxidrening, en sanktion på systemets energiproduktion som varierar i storlek beroende på rökgasrecirkulering och sy-rets renhet. För 97 mol% O2 motsvarar detta bortfall 8,7% av systemets nettoeffekt under torr

rökgasrecirkulering och 15% under våt rökgasrecirkulering. En implementering av koldioxid-avskiljning på kraftvärmeverket skulle kräva ett stort markutrymme på cirka 6300 m2, relativt

höga investeringskostnader för ASU och CPU, totalt kring 93–96 M$, men framförallt skulle det innebära en kraftig ökning av driftskostnaderna. Kostnaden för fångad koldioxid ligger på 75,75 $/ton för torr rökgasrecirkulering och 78,84 $/ton för våt rökgasrecirkulering. På grund av den avsevärt högre energisanktionen under våt rökgasrecirkulering är det fördelaktigt att låta torr rökgas återvända till pannan, ur både det tekniska och det ekonomiska perspektivet.

Nyckelord: Avfallsförbränning, Koldioxidavskiljning, Koldioxidfångning,

(6)

CONTENT

1 INTRODUCTION ... 1 1.1 Background ... 2 1.2 Problem Definition ... 6 1.3 Purpose ... 6 1.4 Research Questions ... 6

1.5 Delimitation and Scope... 7

2 METHOD ... 8

2.1 Methodology Approach ... 8

2.2 Data Collection ... 9

2.3 Model Development and Simulation ... 10

2.4 Model Validation... 11 3 CAPTURE TECHNIQUES ... 12 3.1 Absorption ... 12 3.1.1 Physical Absorption ... 12 3.1.2 Chemical Absorption ... 13 3.2 Membrane-Based Separation ... 15 3.3 Adsorption ... 17 3.3.1 Physisorption ... 18 3.3.2 Chemisorption ... 19

3.3.3 Pressure Swing Adsorption ... 19

3.4 Cryogenic Distillation ... 20

3.5 Calcium Looping ... 22

3.6 Chemical Looping ... 22

3.7 Summarizing Findings on Separation and Capture Technologies ... 24

(7)

4.3 Oxyfuel Retrofits ... 30

4.3.1 Space Requirement ... 31

4.3.2 Flue Gas Recycling ... 31

4.3.3 Air Separation Unit (ASU) ... 32

4.3.4 Compression and Purification Unit (CPU) ... 33

4.4 Cost Assessment ... 34

5 CASE STUDY AT MÄLARENERGI ... 38

5.1 Mälarenergi AB ... 38

5.2 Description of Reference Plant ... 38

5.3 Description of The Integrated Oxyfuel System ... 40

5.3.1 Air Separation Unit (ASU) ... 40

5.3.2 CO2 Conditioning Process – Compression and Purification Unit (CPU) ... 42

5.4 Modeling & Simulation... 43

5.4.1 Air Separation Unit (ASU) ... 43

5.4.2 Boiler Island ... 45

5.4.3 Steam Generation (Rankine Cycle) ... 49

5.4.4 Compression and Purification Unit (CPU) ... 50

5.5 Techno-Economic Evaluation ... 52 5.6 Model Validation... 56 6 RESULTS... 57 6.1 Model Validation... 57 6.1.1 CHP... 57 6.1.2 ASU ... 58 6.1.3 CPU... 58

6.2 Dry and Wet Recycling ... 58

6.2.1 Recycling Ratio and Oxygen Concentration ... 59

6.2.2 Comparison of Flue Gases ... 60

6.2.3 Technical Performance ... 62

(8)

6.3.2 Technical Impact of Employing Different Oxygen Concentrations ... 65

6.4 Economic Performance ... 67

6.4.1 Dry Flue Gas Recycling ... 67

6.4.2 Wet Flue Gas Recycling ... 69

7 DISCUSSION ... 71

7.1 Validation of The Model ... 71

7.2 Additional Space Requirement ... 72

7.3 Incorporation of a Flue Gas Recirculation System ... 72

7.4 Incorporation of ASU and CPU ... 74

7.5 Economic Viability ... 75

8 CONCLUSIONS ... 77

9 SUGGESTIONS FOR FUTURE WORK ... 78

REFERENCES ... 79

APPENDIX 1: DRY FLUE GAS RECYCLING APPENDIX 2: WET FLUE GAS RECYCLING

(9)

LIST OF FIGURES

Figure 1: Illustration scheme of the dominant carbon capture technologies ...3

Figure 2: Number of publications on oxyfuel combustion ... 4

Figure 3: Brief illustration of the methodology ... 9

Figure 4: Illustration of chemical absorption ... 14

Figure 5: A simple schematic representation of the chemical-looping combustion process... 23

Figure 6: Illustration of oxyfuel retrofits... 30

Figure 7: Illustration of unit operation of cryogenic air separation unit ... 33

Figure 8: CO2 avoided, and CO2 captured ...35

Figure 9: Illustration of the process scheme of Plant 6 ... 39

Figure 10: Illustration of the ASU flow diagram ... 41

Figure 11: Illustration of the CPU flow diagram ... 42

Figure 12: ASU –Aspen flowsheet of the Compression and Purification Island ... 44

Figure 13: ASU -Aspen flowsheet of the Refrigeration & Distillation island ... 45

Figure 14: Aspen flowsheet of the oxy-waste boiler island (dry-FGR) ... 45

Figure 15: Aspen flowsheet of the oxy-waste boiler island (wet-FGR)... 48

Figure 16: Aspen flowsheet of the steam generation cycle ... 49

Figure 17: CPU – Aspen flowsheet of the flue gas compression island. ... 51

Figure 18: CPU – Aspen flowsheet of the cryogenic distillation island ... 51

Figure 19: Validation of the CHP model ... 57

Figure 20: Impact of dry-FGR on the combustion temperature ... 59

Figure 21: Impact of wet-FGR on the combustion temperature ... 59

Figure 22: Dry-FGR oxy-waste plant under different capacity loads. ... 63

Figure 23: Wet-FGR oxy-waste plant under different capacity loads. ... 64

Figure 24: The energy consumption of ASU and CPU during different oxygen purities... 65

Figure 25: The CO2 emissions, the CO2 avoided with oxyfuel, and the CO2 captured, for dry-FGR ... 68

Figure 26: The CO2 emissions, the CO2 avoided with oxyfuel, and the CO2 captured, for wet-FGR. ... 70

(10)

LIST OF TABLES

Table 1: Existing studies of oxyfuel combustion. ... 5

Table 2: Pros and cons – Different CO2 capture techniques. ... 24

Table 3: Economic performance from previous studies ... 37

Table 4: Main operating parameters of the CHP plant ... 39

Table 5: Ultimate analysis of refuse-derived fuel used in the modeling ... 40

Table 6: Aspen Plus unit operation block description of ASU ... 43

Table 7: Aspen Plus unit operation block description of oxy-waste boiler island. ... 46

Table 8: Aspen Plus unit operation block description of CPU... 50

Table 9: Reference costs for ASU and CPU ... 54

Table 10: Assumptions for cost assessment ... 55

Table 11: Validation of the ASU model ... 58

Table 12: Validation of the CPU model ... 58

Table 13: Comparison of flue gases during dry- and wet flue gas recycling. ... 60

Table 14: Flue gases from reference plant. ... 61

Table 15: Technical performance of oxy-waste combustion with dry- and wet FGR. ... 62

Table 16: Effect of different oxygen purities on flue gas composition. ... 65

Table 17: The impact of different oxygen purities on the oxy-waste system ... 66

Table 18: Economic performance dry-FGR ...67

(11)

NOMENCLATURE

Symbol Description Unit

LHV Lower Heat Value of fuel MJ/kg

p Pressure bar T Temperature K V Volume m3 η Efficiency %

ABBREVIATIONS

Abbreviation Description

AIC Annualized Investment Cost ASU Air Separation Unit

BECCS Bio-energy with Carbon Capture and Storage BioCCS Bio-energy with Carbon Capture and Storage CCS Carbon Capture and Storage

CFB Circulating Fluidized Bed CHP Combined Heat and Power CHP Combined Heat and Power CLC Chemical Looping Combustion CO2-eq CO2-equivalent

COA Cost of CO2 avoided

COC Cost of CO2 captured

COE Cost of Electricity

CPU CO2 Compression and Purification Unit

DeNOx NOx reduction system

FGD Flue Gas Desulphurization FGR Flue Gas Recycling

FGR Flue Gas Recycling

GHG Green House Gases

(12)

Abbreviation Description

IPPC Intergovernmental Panel on Climate Change LCOE Levelized Cost of Electricity

LHV Lower Heating Value

MEA Monoethanolamine

MS Microsoft

MSW Municipal Solid Waste

NGCC Natural Gas Combined Cycles

NOx Nitrogen Oxides

PSA Pressure Swing Absorption

RDF Refuse-Derived Fuel – sorted and dried combustible waste.

SOx Sulfur Oxides

UNFCCC United Nations Framework Convention on Climate Change

DEFINITIONS

Definition Description

Mol% The moles of a specific component as a proportion of the total moles in a mixture.

Oxy-waste Integrated waste incineration oxyfuel combustion ton Metric ton: 1000 kg

Vol% The volume of a specific component as a proportion of the total volume of the mixture.

(13)

1

INTRODUCTION

This chapter aims to introduce the topic of this thesis. The chapter starts by overviewing the background of the studied area, followed by an elucidation of the problem that is to be investi-gated. Following that are sections that highlight the thesis’ objective and research questions. The chapter concludes by presenting the scope and delimitation of this thesis.

The world is undergoing extensive climate change, mostly recognized as global warming. Global warming intends the increase in global average surface temperature, which has in-creased with 0.85°C during 1880-2012 due to humankind’s greenhouse gas emissions, with carbon dioxide, nitrous oxide, and methane being some of the most frequent gases (Intergov-ernmental Panel on Climate Change [IPCC], 2014). The greenhouse gas with the most significant impact on climate change is CO2 (Songolzadeh et al., 2014), accounting for almost two-thirds

of the enhanced greenhouse effect (Blunden & Arndt, 2020). Most of the CO2 emissions are

directly caused by fossil fuel combustion, which in turn stands for about 84.3% of the global primary energy use and about 30.8% of Sweden’s (BP, 2020). To counteract further climate change, actions have been taken in recent years. For instance, during the 21st Conference of the Parties of the United Nations Framework Convention on Climate Change (UNFCCC), a common goal to decrease greenhouse gases (GHG) to maintain a global average temperature below 2°C above pre-industrial levels was established. This goal was adopted as part of the Paris Agreement and aimed to limit global temperature rise to 1.5 °C (UNFCCC, 2015). For this to happen, decarbonization of the energy sector and net-zero carbon dioxide emissions must be reached by 2050 (Rogelj et al., 2018). Therefore, many countries and states worldwide have set goals to reach net-zero CO2 emissions by the middle of the century (Global CCS Institute

[GCCSI], 2020; Schreyer et al., 2020). Sweden has set a target of reaching net-zero emissions by 2045, which would entail an 85% reduction in 1990 emissions (Klimatpolitiska vägvalsu-tredningen, 2020).

Furthermore, the global energy demand is expected to grow annually over the next decade (BP, 2020; World Energy Council, 2013), necessitating a complete transition to renewable energy sources to supply the growing demand while reducing GHG emissions (Rogelj et al., 2018). Renewable energy sources such as solar and wind energy are expected to maintain rapid growth in the next few years and contribute to achieving net-zero carbon dioxide emissions (International Energy Agency [IEA], 2020a). However, due to the inability to supply all renew-able energy around the clock, the European Climate Foundation (2010) declared that only by adding an important technology, carbon capture and storage (CCS), it could be possible to re-duce 80% of the 1990s greenhouse gas emissions in 2050.

(14)

1.1

Background

The theory of CCS goes back to 1977, when it first was suggested as a technique to prevent the then-emerging climate change (Marchetti, 1977). Since then, much research has been done on the subject. In 2020 there were 65 facilities with carbon capture and storage in the world, cor-responding to an increase of 33% from 2019. Of these facilities, 26 were completely opera-tional, with a total CCS capacity of 40 Mt CO2 yearly, while the others were in various stages of

development (GCCSI, 2020). In comparison, this would cover almost all of Sweden’s CO2

emis-sions in 2019, approximately 46.3 Mt CO2 (BP, 2020).

A variant of CCS is BECCS, BioEnergy with Carbon Capture and Storage, in which CO2 is

extracted from biomass rather than fossil fuels, as is the case with traditional CCS. As some of the municipal solid waste (MSW) is biogenic carbon, waste-to-energy combined with carbon capture is considered BECSS (Wienchol et al., 2020). The concept of BECCS is; biomass ab-sorbs CO2 from the air through photosynthesis while it grows. The biomass is thenceforth used

as fuel to generate energy in a power plant, in which the CO2 is captured through carbon

cap-ture. When the CO2 is captured in the power plant and stored, negative emission is achieved

(IEA, 2020b). Fuss and Johnsson (2021) point out three main reasons that BECCS has consid-erable potential in Sweden. The first reason is the existing forest industry and the associated carbon stock. The second is the nation’s clear ambitions to become a pioneer in taking climate responsibility and obtain negative emissions. The third reason is that BECCS has been pro-moted, and related targets have been proposed by a public inquiry, published in Klimat-politiska vägvalsutredningen (2020). One of the targets is to capture and store 1.8 million tons CO2-eq by 2030 annually. However, today, there are no economic benefits for companies that

achieve negative emissions. That is something the public inquiry has highlighted, and the pro-posal is to introduce government aids for those who contribute to negative emissions.

In short, the goal of CCS is to reduce anthropogenic carbon emissions by first capturing CO2

from flue gases, separating, transporting, and then permanently storing or reutilize industri-ally (Leung et al., 2014). CCS includes a series of different carbon dioxide separation methods, including absorption, physical and chemical absorption, membrane, and cryogenic process. These separation methods suffer from being energy-intensive due to substantial energy penal-ties for regeneration and additional compression work (Ma et al., 2016). Therefore, different carbon capture technologies with various advantages compete to become a low-cost solution for the carbon dioxide capture process. Among these technologies, three main approaches to capture CO2 can be distinguished: Post-combustion capture, Pre-combustion capture, and

Ox-yfuel combustion (Huang et al., 2018). Figure 1 presents the fundamentals of these technology approaches.

(15)

Figure 1: Illustration scheme of the dominant carbon capture technologies.

At present, post-combustion capture is considered the most mature technology for carbon cap-ture with the least impact on the existing system. However, it is quite far from becoming eco-nomically feasible since extracting CO2 from exhaust gases after combustion requires a

signif-icant amount of energy and an 8-12% efficiency penalty (Pettinau et al., 2017). In general, the exhaust gases in a post-combustion system are predominantly composed of a mixture of nitro-gen, carbon dioxide, water vapor, and oxygenated compounds. Such mixture requires treat-ment to remove particulate matter, nitrogen, and sulfur oxides before being transferred to a separation unit, in which a liquid solvent, usually an aqueous amine solution, chemically ab-sorbs the carbon dioxide (Basile et al., 2011). According to Breeze (2015), up to 90% of the carbon dioxide is chemically bounded and removed before the flue gases are released into the atmosphere.

Pre-combustion capture involves the capture and removal of carbon dioxide prior to combus-tion. The method is most widely used in Integrated Gasification Combined Cycles (IGCC) and Natural Gas Combined Cycles (NGCC), and other plants that are dedicated to producing syn-thesis gas (Jansen et al., 2015). The basic approach of pre-combustion capture employs a low-cost physical solvent (for example, rectisol or selexol) to dissolve CO2 under high pressure and

(16)

re-In contrast to post-combustion, the energy requirement for CO2 separation and compression

during pre-combustion involves an efficiency penalty of less than 9% (Pettinau et al., 2017). Nevertheless, pre-combustion capture is still unreliable, expensive, and needs further devel-opment (Jansen et al., 2015).

The principle behind oxyfuel combustion is to aid the capture of CO2 from flue gases by

elimi-nating nitrogen prior to combustion. As illustrated in Figure 1, the process comprises flue gas recirculation to provide a nitrogen-lean and carbon dioxide-rich combustion atmosphere. Thus, resulting in a flue gas mainly composed of carbon dioxide and water vapor, from which the latter can be easily removed through condensation (Miller, 2016). Although oxyfuel com-bustion is not yet commercially applicable, it is considered the most promising of the technol-ogies considered. Because it offers easy CO2 capture and a low-efficiency penalty, slightly

above 7 % (Escudero et al., 2016). Many studies have been conducted to assess the viability of oxyfuel combustion, and the number of publications has dramatically increased over the last decade, which can be envisaged in Figure 2.

Figure 2: Number of publications on oxyfuel combustion. Source: Web of Science as cited in Wienchol et al. (2020).

(17)

The results indicate promising progress in terms of cost-effectiveness and technological feasi-bility. Table 1 summarizes some of the promising findings from previous research on oxyfuel combustion integrated with fossil-fired power plants.

Table 1: Existing studies of oxyfuel combustion.

Studied Area Findings Reference

High-efficiency low LCOE combined cycles for sour gas oxy-combustion with CO2 capture

The results showed a better technical and eco-nomic performance of the sour gas-based oxy-combustion combined cycles.

Chakroun & Ghoniem (2015)

Techno-economic analysis for a large-scale Carbon Capture and Storage

Oxyfuel combustion offers better techno-eco-nomic performance than other commercial and mature carbon capture options.

Cormos (2016)

Feasibility assessment of CO2 capture retrofitted to

an existing cement plant

The results indicate that oxyfuel CO2 capture

tech-nology may be a better choice in terms of costs compared to post-combustion capture.

Gerbelová et al. (2017)

Cost of CO2 capture and

storage in fossil- and nat-ural gas-fired power plants

The LCOEs and mitigation costs of oxyfuel com-bustion in plants using low-rank coals are compa-rable to those of supercritical pulverized coal power plants with post-combustion.

Rubin et al. (2015)

Economic and technical overview of the future of CCS technology

The main barriers to implementing CCS are

eco-nomic rather than technical. Wennersten et al. (2015)

Key combustion funda-mentals in pulverized fuels (PF)-firing

High-concentration CO2 and H2O only marginally

boost overall radiative heat transfer in PF oxyfuel furnaces. However, the high concentrations greatly affect gas phase combustion.

Yin & Yan (2016)

Optimization of operating parameters of an oxyfuel combustion power gener-ation system

Increasing the purity of oxygen reduces net coal consumption while increasing net electrical effi-ciency.

Zhang et al. (2020)

Aside from Table 1, some experiments have been conducted to examine the feasibility of com-bining oxyfuel combustion with non-fossil-fired power plants. However, most of these experi-ments have been carried out in a laboratory-scale (Wienchol et al., 2020).

(18)

1.2

Problem Definition

According to Figure 2, there appears to be a niche for research into oxyfuel combustion as a technique of reducing carbon dioxide emissions. However, although non-fossil fuels such as waste are currently burned in stationary combustion plants, most of the current research is focused on the oxy-combustion of coal (Sher et al., 2018). In this regard, beyond laboratory-scale power plants, research on oxyfuel combustion of waste has received much less attention. Accordingly, there is a knowledge gap in evaluating the technical and economic performance of adapting oxyfuel combustion into waste incineration combined heat and power (CHP) plants. In this thesis, on behalf of Mälarenergi1, a techno-economic study of retrofitting oxyfuel

combustion into a 167 MW waste incineration CHP plant is carried out.

1.3

Purpose

This study aims to research the feasibility of integrating oxyfuel combustion into existing waste incineration CHP plants regarding economic and technological aspects. A case study on Mä-larenergi’s refuse-derived waste (RDF)-fired CHP plant, Plant 6, is included to acquire more in-depth knowledge. The purpose of this case study is to provide an overall picture of Mä-larenergi’s possibilities to apply oxyfuel combustion as a method of limiting carbon dioxide emissions. Furthermore, the present work aims to overview the current knowledge status of the various separation techniques to provide a base for decision-making for Mälarenergi.

1.4

Research Questions

Based on the knowledge gap regarding integrated oxyfuel carbon capture with RDF-fired CHP plants, the following research questions were developed in this study.

1. What are the primary considerations for retrofitting existing CHP plants with oxyfuel combustion?

2. From a technical standpoint, how does the integration of oxyfuel combustion affect the operation of the existing CHP plant?

3. Which separation technique is best suited to be applied in the air separation unit and the CO2 compression and purification process? In terms of efficiency, costs, and ability

to supply a large amount of high purity oxygen.

4. How feasible is the conversion process of existing CHP plants to oxyfuel combustion from an economic perspective?

(19)

1.5

Delimitation and Scope

The CCS chain includes separation, capture, transportation, and storage of carbon dioxide. This work is limited to study the separation and capture technologies, where oxyfuel combus-tion is the focus area. Thus, transportacombus-tion and end storage of CO2 is excluded. Moreover, due

to technical limitations in the modeling tool, Aspen Plus, only steady-state simulations are considered. Likewise, there is no possibility to specify the boiler type. Hence, general simula-tion results are obtained. Furthermore, for the sake of simplificasimula-tion and technical limitasimula-tions of the software of choice, the modeling is based on several assumptions, including the neglect of heat transfer in the furnace, the exclusion of heavy metal formation, and the exclusion of the flue gas cleaning process. For the economic part, the cost of heat is not addressed; instead, the cost of electricity is. Moreover, previous studies are used to support estimates of fixed and var-iable costs for the economic evaluation.

(20)

2

METHOD

A brief overview of the method used to conduct this study is presented in this chapter. The methodology procedure and the assembling of the theoretical framework and literature review of this study are described in the first section of the chapter. Following that, the gathering of necessary data and model development is to be presented. The chapter ends with a brief over-view of how the developed model will be validated.

2.1

Methodology Approach

At an early stage of this study, state-of-the-art of current separation and capture techniques were developed to enhance understanding of the investigated area. Accordingly, the literature review of this thesis comprises a theoretical framework with fundamental knowledge about the different separation and capture techniques and the characteristics of oxyfuel combustion. Be-sides, previous studies of integrating oxyfuel combustion and discussions with experienced personnel in the field of CCS and oxyfuel combustion were conducted to investigate the feasi-bility of adapting oxyfuel combustion. In this regard, the economic analysis has primarily been based on research and estimations of previous attempts of integrating oxyfuel combustion into existing power plants. Moreover, the economics includes the assessment of the costs of invest-ment, modifications, and the total amount of CO2 captured. Keywords such as CCS, Oxyfuel

combustion, Gas separation techniques, Oxygen-enriched combustion, Aspen Plus, and Cost assessment of CCS were used to gather relevant literature, using the search engines Discovery,

Primo, and Google Scholar.

Furthermore, using the process industry-designed software Aspen Plus V11, a case study about Mälarenergi’s waste incineration heat and power (CHP) plant, Plant 6, was analyzed. Aspen Plus is a steady-state flowsheet-based simulator in which various unit operation and process modules, including reactors, pressure changers, columns, heat exchangers, mixers, and sepa-rators, can be virtually connected via pipelines (materials and energy flows) to define a model in detail. Each unit can be defined through desired inputs, specific reaction mechanisms, de-sign, and efficiency. These specifications are then used to solve the process scheme module using a sequential modular (SM) approach, i.e., calculating the outlet flow characteristics of each module based on the inlet flow characteristics. Moreover, to ensure that the system oper-ates within the defined constraints, Aspen Plus provides manipulator functions based on Fortran code calculator. These can be used to define non-conventional fuels, assigning chemi-cal reactions, and establishing energy balances, for example.

(21)

The objectives of the simulation were to analyze the impact of integrating oxyfuel combustion and evaluate the best possible operating parameters with the least effort required. Accordingly, to assess the technical effects of implementing oxyfuel combustion, the simulation results, in-cluding flue gas comparison, power output, additional required work, and plant efficiency, were analyzed and compared to actual data from the CHP plant.

The whole simulation process was divided into three parts: an air separation unit (ASU), the actual CHP plant, and a compression and purification unit (CPU). These are covered in greater detail in Chapter 5. The CHP plant was developed with inspiration from the CHP model pro-posed by Salman et al. (2017), and the ASU process was modeled following the propro-posed Linde Double Column model in Hu et al. (2010). The CPU process follows the proposed two-stage cryogenic model in Li et al. (2013). Figure 3 briefly illustrates the methodology of this study.

Figure 3: Brief illustration of the methodology.

2.2

Data Collection

Fuel properties and annual operating data of the reference CHP plant were received from Mälarenergi per Excel sheet. The data was in raw format and needed preprocessing. Hence, using MS Excel, the data was preprocessed, and inaccurate data was filtered and analyzed. In this context, inaccurate data refers to operational data collected during a maintenance

(22)

shut-2.3

Model Development and Simulation

The developed model in this study is based on two interconnected processes: oxy-waste combustion with flue gas recirculation, and CO2 purification unit. The first process involves

developing a CHP plant equal to that of Mälarenergi, Plant 6, and a cryogenic air separation unit (ASU). The second process entails modeling a compression and purification unit (CPU) developed to purify the CO2-enriched flue gas. To simulate the operations of the model,

equation of state (EOS) models2, such as the Peng-Robinson equation of state physical

property3 and the Steam-TA property method4, were used. The Peng Robinson property

method was used for calculations of thermodynamic properties in all of the aforementioned processes. The Peng Robinson approach was chosen due to being highly consistent between experimental and simulation performance and applicable in both gas and liquid phases. The empirical form of this equation is expressed in Eq. (2 − 1) (Peng & Robinson, 1976),

𝑝 = 𝑅𝑇

(𝑐 + 𝑉-) − 𝑏−

𝑎

(𝑉-+ 𝑐)(𝑉-+ 𝑐 + 𝑏) + 𝑏(𝑉-+ 𝑐 − 𝑏) (2 − 1) Where a, b, and c are component-specific parameters. For pure materials, the values of these parameters are stored in the Aspen Plus database, while for mixtures, they are determined using mixing rules for mixtures. The Steam-TA property method, on the other hand, was used to assess the water/steam properties of the model.

The developed CHP model comprises the development of a circulating fluidized bed (CFB), various heat exchangers, pressure exchangers, and other ancillary components. The prepro-cessed data was used to simulate the studied CHP plant, both with and without oxyfuel com-bustion, to obtain comparable data. However, since Aspen Plus only provides steady-state sim-ulations, the simulation was conducted at full capacity load. To replicate the CHP plant’s actual operation as closely as possible, a sensitivity analysis in Aspen Plus was conducted by varying the capacity load from 60 to 100%. Moreover, different flue gas recycling methods (wet and dry) were examined to study the behavior of the combustion characteristics during oxyfuel combustion. Besides, simulations with different oxygen purities: 80, 90, 95, and 97 mol% were performed to evaluate the best possible operating parameters with the least efficiency penalty.

2 Equation of states is thermodynamic equations used to relate state variables including pressure, volume, and

temperature (PVT) to describe the properties of fluids, mixture of fluids, and solids.

3 The Peng-Robinson equation of state is a built-in EOS in Aspen Plus, used to express fluid properties in terms of

(23)

2.4

Model Validation

As previously mentioned, the developed model of this study is divided into three main pro-cesses: CHP, ASU, and CPU. Each model is validated separately. The CHP model is validated through calibration with actual operating data received from Mälarenergi. The validation is based on a comparison of the model’s heat and power production output to the CHP plant’s actual heat and power production. This was accomplished using Aspen Plus and MS Excel, with the former being used to collect model output data and the latter to preprocess actual CHP plant data and compare the two data sets to validate the CHP model.

The ASU and CPU models, on the other hand, are validated by comparing the power consump-tion of producing different oxygen purities and purifying CO2 with published data in Hu et al.

(24)

3

CAPTURE TECHNIQUES

In this chapter, a theoretical review of the current knowledge status of the different separation and capture techniques is presented. The chapter aims to identify the most feasible technique for oxygen production and capture of carbon dioxide from flue gas.

3.1

Absorption

Absorption between two substances includes one absorbent, which absorbs another substance, the absorbate. Furthermore, the absorbent can also be called solvent (Wu et al., 2018). The absorption technology is divided into two categories, chemical, and physical absorption. These are mainly defined in how the absorption takes place, whether there is a chemical reaction between the absorbent and absorbate or not (Fang & Zhu, 2015).

3.1.1 Physical Absorption

Physical absorption implies physical solvents to absorb desired components from gases, such as carbon dioxide from syngas, without chemically reacting with them (Rackley, 2017). It is one of the leading commercial alternatives for pre-combustion carbon capture and natural gas purification (Nookuea, 2020; Tan et al., 2016), and Vega et al. (2018) claim that it is highly recommended to use this alternative in pre-combustion CO2 capture.

The physical absorption process takes place according to Henry’s Law, which makes it depend-ent on pressure and temperature. The partial pressure and the temperature of the desired gas component determine its solubility in the solvent, of which the solubility increases linearly with the partial pressure. Simply put, the desired gas component, often CO2, migrates from the gas

to the solvent if the solubility is greater in the solvent. To favor solubility, high partial pressure and low temperature are appropriate (Olajire, 2010; Strube & Manfrida, 2011). Physical ab-sorption is more suitable for CO2 capture in the pre-combustion stage rather than the

post-combustion stage, which can be explained by the lower partial pressure of CO2 in

post-com-bustion gases (Cormos et al., 2016), and higher concentration of CO2 in pre-combustion

(Nookuea, 2020; Tan et al., 2016). Additional to high partial pressure, physical absorption in low temperatures promotes the highest efficiency and economy (Kohl & Nielsen, 1997; Vega et al., 2018).

(25)

The process of physical absorption can be divided into two steps, the absorption step and the solvent regeneration after the separation (Scholes et al., 2013). The absorption itself happens in one step, making physical absorption unperturbed towards limitation due to fixed stoichi-ometry5, unlike chemical absorption (Nookuea, 2020). However, the absorption of CO2 is an

exothermic reaction that increases the solvent’s temperature, and thus, the temperature change has to be kept in control (Tan et al., 2016). The solvent regeneration, done by pressure reduction or by heating, decreases the solubility of the CO2, resulting in the CO2 being

de-sorbed. After the solvent regeneration, the lean solvent is recycled to absorb once again (Olajire, 2010; Scholes et al., 2013; Strube & Manfrida, 2011; Wang et al., 2011).

Speight (2007) points out that Rectisol and Selexol are two of the most common physical sol-vent processes for cleaning gases. One of the reasons is that the solsol-vents can be regenerated, by pressure reduction, without requiring too much energy (Strube & Manfrida, 2011). The low heat duty for regeneration of the solvent is also one of the advantages of physical absorption in general (Cormos et al., 2016). However, these processes have low absorption capacity and in-volve high costs (Nemitallah et al., 2017).

Rectisol is a process using chilled methanol to remove carbon dioxide and hydrogen sulfide (H2S) from syngas and other streams with moderate or less concentration of CO2 (Vega et al.,

2018). The process must take place at a low temperature, and thus, the cost of cooling the sol-vent is the most significant disadvantage of Rectisol (Speight, 2007).The other common phys-ical absorption option, Selexol, uses a mixture of dimethyl ethers of polyethylene glycol as a solvent, often to separate H2S and CO2 from syngas (Padurean et al., 2012). Selexol has

ad-vantages in low toxicity and low vapor pressure (Yu et al., 2012).

3.1.2 Chemical Absorption

Chemical absorption is another well-established technology for CO2 capture (Fang & Zhu,

2015; He, 2018; Yu et al., 2012) and the most utilized one for post-combustion at low scale plants (IPCC, 2005; Kothandaraman et al., 2009). However, the technology has over the years been deemed insufficiently developed for large-scale applications (Kothandaraman et al., 2009), but it is one of the most promising technologies that are likely to be implemented in the near future for larger-scale post-combustion CO2 capture (Rao et al., 2002; Rochelle, 2009).

Chemical adsorption is suitable if carbon capture is going to be implemented at an existing plant (Yu et al., 2012). Moreover, the technology is expensive and energy-intensive (He, 2018; Jiang et al., 2019).

Unlike physical absorption, this type of absorption includes chemical reactions. These reac-tions occur between the absorbate and the liquid solvent (absorbent) and cause weak bonds between the two substances (Olajire, 2010; Vega et al., 2018).

(26)

For CO2 capture, the bonds are created due to a dynamic interface between gas and liquid,

where the CO2 dissolves into the liquid absorbent and encounters absorbent molecules to

cre-ate chemically distinct products. As CO2 and the absorbent keep bond with each other,

prod-ucts transfer into the liquid absorbent, and thus, CO2 molecules are separated from the flue gas

to the absorbent (Puxty & Maeder, 2016). After the absorption, CO2 is desorbed from the

ab-sorbent in a stripping column by hot counterflow steam. After that, the abab-sorbent is regener-ated (Olajire, 2010; Oyenekan & Rochelle, 2006; Puxty & Maeder, 2016). The heating of the absorbed CO2, which breaks the CO2 and the absorbent bonds, requires much energy and

makes the desorption process contributing to an energy penalty on the system (IEA, 2004; Puxty & Maeder, 2016; Rochelle, 2016). Below, an illustration of CO2 molecules being absorbed

by a liquid solvent is attached.

Figure 4: Illustration of chemical absorption. Adopted from Puxty & Maeder (2016).

In terms of solvent selection, amine solvents are considered the most widely used and mature type (Fang & Zhu, 2015; Tan et al., 2016; Rochelle, 2009). Furthermore, monoethanolamine (MEA) is the most used amine solvent (Puxty & Maeder, 2016; Strube & Manfrida, 2011; Wang et al., 2011). MEA got a high absorptive capacity and a rapid reactivity towards CO2 (Jiang et

al., 2019; Koronaki et al., 2015), which makes the ability to absorb low concentration CO2 one

of the advantages of the solvent (Peng et al., 2012). However, one disadvantage with MEA and other amine-based solvents is that degradation of the solvent occurs, especially in oxidative environments, which contributes to higher costs due to solvent refill. Auspiciously, MEA is the cheapest amine (Rochelle, 2009). To prevent degradation, the flue gases need to be cleaned from SO2 and NOx before the solvent can absorb the CO2. Otherwise, the two dioxides affect

the system performance (Oyenekan & Rochelle, 2006; Wang et al., 2011; Zhuang et al., 2011). Cleaning the flue gas of SO2 and NOx can be a difficult and expensive procedure (Zhuang et al.,

(27)

In addition to amine solvents, ammonia can also be used as an absorbent (He, 2018; Koronaki et al., 2015; Strube & Manfrida, 2011). Absorption with ammonia-based solvent does not entail the same heat from chemical reactions as amine-based absorbents do. Ammonia-based sol-vents can be regenerated at high pressure, it got a potentially higher CO2 capacity than

amine-based solvents such as MEA, and it can operate despite oxygen occurring in the gas (Fang & Zhu, 2015). Also, ammonia-based absorbents have a higher tolerance to SO2 in the flue gas and

can remove pollutant components in flue gases (Peng et al., 2012). However, some disad-vantages are the high volatility of ammonia, which contributes to a more complex carbon cap-ture process, the unstable products (Peng et al., 2012), and the heat requirement for solvent regeneration, albeit it is believed to be less than for amine-based absorption (Jilvero et al., 2012).

3.2

Membrane-Based Separation

Membrane separation is a physical separation process in which gas mixtures composed of two or more components are selectively separated into a retentate stream and a permeate stream through a semi-permeable barrier – membrane (Khalilpour et al., 2015). Gas separation by membranes is a pressure-driven process induced by the difference in pressure between down-stream and updown-stream sides (Chong et al., 2016). For separation, the process takes advantage of differences in diffusivity, solubility, absorption, and adsorption abilities of different gases on different membrane materials (Míguez et al., 2018). Membranes for gas separation can be classified as organic (polymeric), inorganic (metallic, carbon, and microporous ceramics), hy-brid, and facilitated transport (Basile et al., 2011), as well as porous and non-porous (Olajire, 2010). The selection between these membranes depends on the design and operating condi-tions of the capture system used (Míguez et al., 2018). The performance of the membrane relies on two key expressions: permeability and selectivity that are affected by the membrane mate-rial, structure, thickness, configuration, and the design of the capture system (Bernardo et al., 2009).

Permeability describes the ability of the membrane to allow the permeable gas to diffuse (due to pressure difference across the membrane) and can be measured in terms of gas permeation flux, membrane thickness, and area and pressure difference across the membrane. Selectivity is defined as the ratio of permeabilities of different gases through the same membrane. To achieve separation, a large difference between gas permeabilities is preferred, i.e., differences in molecular size, molecular weight, and affinity to membrane material (Ji & Zhao, 2017).

(28)

Membrane-based separation follows five different mechanism approaches: size sieving, sur-face diffusion, solution-diffusion, facilitated transport, and ion transport (Olajire, 2010). Size sieving or size exclusion discriminate according to the size of particles or molecules, using po-rous membranes. Therefore, the greater the difference in molecular/particle size, the more ap-propriate it is to use this size sieving mechanism (Nath, 2017). Surface diffusion is used when the membrane material has a higher affinity for a specific component than another, i.e., dis-criminating according to the chemical affinities between components and membrane materials (Ji & Zhao, 2017). Solution-diffusion is a typical mass transfer mechanism in polymeric mem-branes, where separation is achieved between different permeants due to differences in solu-bility and absorbasolu-bility (Ismail et al., 2005). The mechanism uses dense non-facilitated mem-branes, in which permeants first dissolve into the membrane and then diffuse through it (Kha-lilpour et al., 2015). Due to a combination of low solubility and/or low diffusivity, the solution-diffusion mechanism is often limited by a low permeate flow rate (Ji & Zhao, 2017). Facilitated transport membranes follow the same approach as solution-diffusion but without any limita-tion on the permeate flow rate (Ji & Zhao, 2017). This is because facilitated transport mem-branes also contain an active transport mechanism that increases the permeability and selec-tivity of the membrane material (Khalilpour et al., 2015). The ion transport mechanism is used to separate an O2/N2 mixture by converting oxygen molecules into oxygen ions at the feed

in-terface through a surface-exchange reaction (Ji & Zhao, 2017).

Attention to membrane-based separation has taken off in recent years, and many studies have been dedicated to carbon dioxide separation and sequestration (He, 2018). CO2-selective

mem-branes allow separation of CO2 from different gas streams, such as flue gas in post-combustion

systems, hydrogen in pre-combustion systems, and oxygen/nitrogen in oxyfuel combustion systems (Basile et al., 2011). In pre-combustion capture, shifted syngas is generated consisting of CO2 and H2. The shifted syngas is decarbonized to obtain a stream of pure H2 (Jansen et al.,

2015). Therefore, for this purpose, H2-selective membrane that favors H2 permeation and CO2

-selective membrane that preferentially permeates CO2 are used (Giordano et al., 2019). Due to

infinite selectivity, the ideal membrane type for H2/CO2 separation is metal. However, at

ele-vated temperatures, inorganic membranes, such as carbon, alumina, and silica, can be used to separate the H2/CO2 mixture (Basile et al., 2011).

In post-combustion capture, after the water vapor has been condensed, the exhaust gas mostly contains a mixture of CO2 and N2, with the former having a much lower concentration than the

latter (Breeze, 2015). Accordingly, CO2-selective membranes with high thermal and chemical

stability are required in post-combustion capture (Brunetti et al., 2010). Moreover, due to mi-nor differences in molecular size between CO2 and N2 (the diameter of the former is slightly

smaller than the latter), the separation of the mixture depends on both surface diffusion and solution diffusion (Ji & Zhao, 2017).

(29)

The standard design of an oxyfuel membrane gas separation technique is that air is drawn from the ambient into the membrane module, and the O2/N2 mixture is separated based on

diffu-sivity and solubility differences, with oxygen collected upstream due to the high diffudiffu-sivity and nitrogen collected downstream (Chong et al., 2016). In Burdyny and Struchtrup (2010), two membrane types are addressed for the separation of oxygen: polymeric and high-temperature ion transport membranes, where the former is capable of producing oxygen-enriched air of various concentrations and the latter producing purities of close to 100% oxygen concentra-tion. In the ion transport membranes, oxygen molecules are converted into oxygen ions on the surface of the membrane and are transported through the membrane by an applied electric voltage or oxygen partial pressure difference (Basile et al., 2011). The membranes used in this mechanism are O2-selective in principle and usually inorganic (fluorite-based and

perovskite-based) (Ji & Zhao, 2017).

Compared to traditional separation methods, membrane separation entails many advantages such as lower capital and processing cost, smaller unit sizes, simpler operation, better energy-efficiency, lower environmental impact, and ability to be applied in remote areas (Chong et al., 2016; Dai et al., 2016; Ismail et al., 2005; Ji & Zhao, 2017; Khalilpour et al., 2015). However, several issues are associated with membrane-based separation for CO2 capture, such as

high-temperature sealing, chemical and mechanical stability, and aging resistance (Brunetti et al., 2010; Ji & Zhao, 2017). Moreover, membranes cannot always achieve high degrees of separa-tion, making multiple stages or recycling necessary (Olajire, 2010). In post-combussepara-tion, for instance, the concentration of CO2 in flue gasses is low, which means that large amounts of

gasses will need to be processed. Besides, high temperatures of the flue gas will rapidly melt the membrane, meaning that the gasses must be cooled to below 100 °C before the membrane separation occurs (Brunetti et al., 2010). In the case of oxyfuel combustion, the membrane-based separation mechanism is still at its early stage of development and is unfavorable for commercialization due to high-temperature requirements and the resulting high cost of air separation (Chong et al., 2016; Ji & Zhao, 2017).

3.3

Adsorption

The definition of adsorption is the adhesion of atoms, molecules, or ions from a dissolved sur-face, gas, or liquid to a surface. Unlike absorption that treats the whole material in the capture process, adsorption only occurs at the material surface (Ben-Mansour et al., 2016; Speight, 2007). The separation technique is used in several areas, including removal of water vapor and acid compounds from gases, and gas purification in the industrial sector (Lecomte et al., 2010). Adsorption has been explored for CO2 separation in both pre-combustion and

post-combus-tion (Drage et al., 2009) and is also an alternative for separating oxygen from air (Lecomte et al., 2010). However, for post-combustion, adsorption is still under development to be commer-cial while absorption by MEA is already mature (Jiang et al., 2019).

(30)

The adsorption process is based on an adsorbing material that takes up an adsorbate selec-tively. The adsorption can be expressed as a function of the partial pressure, in which it acts under Henry’s law and is linear to the pressure. However, adsorbents have a limited adsorption capacity which with time will imply significantly higher demand in pressure to keep adsorbing, leading to a deviation from the proportional adsorption-pressure relation (Smit, 2014). The main issues and drawback with adsorption for carbon capture is the limited adsorption capacity, especially at low pressure, high temperature required for regeneration, and the sen-sitivity towards other components than CO2 in the flue gas, such as water vapor (Ding et al.,

2019; Yu et al., 2012).

The adsorption process mainly occurs in two different ways, physically and chemically. Physi-cal adsorption, physisorption, deal with weak van der Waals forces, and chemiPhysi-cal adsorption, chemisorption, deal with chemical bonding such as covalent bonding (Ben-Mansour et al., 2016; Songolzadeh et al., 2014; Yu et al., 2012; Webb, 2003).

3.3.1 Physisorption

Physisorption is the most common type of adsorption, whereas two possible sorbents for phys-ical adsorption are zeolites and activated carbons (Aaron & Tsouris, 2005; Lecomte et al., 2010; Smit, 2014; Yu et al., 2012). Physical adsorption is most adaptable in low temperatures and has an advantageous low heat of adsorption, which keeps the energy penalty low if the pressure decreases or temperature increases during desorption (Rackley, 2017; Webb, 2003). However, two disadvantages of the technique are low CO2 selectivity and low adsorption kinetics

(Nem-itallah et al., 2017).

Zeolites are one of the most promising adsorbents, and thus, a lot of research has been done on the materials to date (Samanta et al., 2012). It is promising due to the zeolite’s properties, such as the chemical composition and pore size (Sayari et al., 2011). However, zeolites have a hydrophilic behavior which implicates that the flue gas must be dried before CO2 adsorption.

Otherwise, impurities as moisture or water vapor could inhibit adsorption (Samanta et al., 2012; Sayari et al., 2011; Yu et al., 2012).

Activated carbon has some advantages compared to other adsorbents in the low cost, relatively high adsorption capacity (Songolzadeh et al., 2014), wide availability since it can be produced from several sources such as coal and biomass, and high level of thermal stability (Sayari et al., 2011; Yu et al., 2012). It is also tolerant to moisture and easy to regenerate (Songolzadeh et al., 2014). Activated carbons are most suitable for CO2 adsorption at high pressure due to the urge

of a high adsorption capacity (Sayari et al., 2011). However, activated carbon has a lower selec-tivity for carbon dioxide over nitrogen than zeolites, particularly at low CO2 partial pressure

(31)

3.3.2 Chemisorption

Chemisorption is achieved by chemical reactions between the adsorbate and the adsorbent, creating CO2-based molecule structures, after which it is regenerated (Ben-Mansour et al.,

2016). For chemical adsorption, amine-based adsorbents (Ben-Mansour et al., 2016) and metal oxides are a common choice of sorbent (Smit, 2014). This type of adsorption has the advantage compared to physical adsorption that for CO2 capture being feasible for CO2

sepa-ration from high volume flue gases with a low partial pressure of CO2 (Ben-Mansour et al.,

2016; Samanta et al., 2012). Jiang et al. (2019) indicate that chemical adsorbents such as amine-based have a higher capacity for sorption than physical adsorbents such as activated carbon and absorbents such as MEA. Also, some chemical sorbents can be used in high tem-peratures, in which Rackley (2017) indicates metal oxides work in temperatures up to 600-1000 °C. Chemisorption requires, due to chemical reactions, high activation energy (Králik, 2014). However, a disadvantage with chemical adsorption compared to physical adsorption is the time required for chemical reactions between adsorbent and adsorbate, which is longer than adsorption by, e.g., van der Waal forces. Nevertheless, chemical adsorption contributes to a higher amount of energy in the form of adsorption heat than physical adsorption (Deng et al., 2019; Králik, 2014; Wang et al., 2014). This separation technique also entails a decrease in sorption capacity over numerous cycles (Nemitallah et al., 2017).

3.3.3 Pressure Swing Adsorption

Pressure Swing Adsorption (PSA) is one of the leading technologies for gas separation by solid sorbents, using pressure change to adsorb and desorb species, such as CO2 and H2, from gases

(Tan et al., 2016). It is used for gas separation at most of the biogas plants in Sweden, at some natural gas plants worldwide (He, 2018), at hydrogen plants (Damen et al., 2006), and at dif-ferent power plants for both pre- and post-combustion CO2 capture (Raza et al., 2019; Strube

& Manfrida, 2011). It can also be used for oxygen separation from air (Wu et al., 2018). With PSA, gas separation can be made at low temperature and high pressure. The process involves subjection of the inlet gas to a pressure above atmospheric pressure and adsorption of the de-sired species. After the adsorption, species are released when pressure reduction down to at-mospheric pressure occurs, which enables desorption of the captured species into a separate stream (Kwon et al., 2011; Yu et al., 2012).

PSA is advantageous due to its excellent regeneration rate and well accomplishment at ambient temperatures; thus, a low energy demand (Aaron & Tsouris, 2005; Kwon et al., 2011; Wu et al., 2018). Be that as it may,impurities might be added to the gas at low pressures and adversely affect the adsorption performance of the desired gas species. PSA also has a disadvantage in low cost-effectiveness, for CO2 capture, in flue gases with low CO2 concentration. Furthermore,

feed gas losses can occur when pressure reduction takes place, and thus the rate of the inlet stream might be troubled so that the pressure becomes somewhat unstable (Kwon et al., 2011).

(32)

Beyond PSA, Temperature Swing Adsorption (TSA) is another commercially accessible tech-nology for gas separation and purification (Nemitallah et al., 2017; Tan et al., 2016). TSA differs from PSA in the fundamental by change of temperature instead of pressure, in which the pres-sure is kept approximately constant (Songolzadeh et al., 2014; Wu et al., 2018). The process involves a temperature increase of a desorber by some external source, such as hot air or steam, and a chilled adsorber by some cooling medium (Jiang et al., 2019; Yu et al., 2012). One draw-back with TSA is that it is a less rapid process than PSA due to the time required for heating and chilling is greater than the time for pressurization in PSA. However, when TSA is inte-grated into a plant, it could be possible to use waste heat for the heating processes (Rackley, 2017; Yu et al., 2012).

Additional to the two recently named technologies, more technologies are available, such as Vacuum Swing Adsorption (VSA), Electric Swing Adsorption (ESA), and some hybrid pro-cesses as Pressure Temperature Swing Adsorption (PTSA) and Vacuum Temperature Swing Adsorption (VTSA) (Ben-Mansour et al., 2016; Kwon et al., 2011; Yu et al., 2012).

3.4

Cryogenic Distillation

Distillation is one of the most widely used thermal separation techniques (Ebrahimzadeh et al., 2016). The process is termed cryogenic distillation when liquifying gas mixtures to induce a phase change for separation (Wilcox, 2012). Gas separation by cryogenic distillation involves high pressure and extremely low temperature, at which gas mixtures are separated into their components due to differing boiling points and volatilities (Leung et al., 2014). The volatility of a component describes its molecules’ tendency to evaporate as a result of escaping the vapor-liquid interface (Rackley, 2017). For carbon capture, cryogenic distillation can be used in two ways, either to produce a nitrogen-lean atmosphere for oxyfuel combustion through cryogenic air separation or to capture CO2 from flue gas via a series of compression, refrigeration, and

separation steps (Xu et al., 2014).

The principle of cryogenic air separation can be divided into three major process islands: pression and purification, refrigeration, and distillation (Smith & Klosek, 2001). In the com-pression and purification island, ambient air is pretreated through filtration, comcom-pression, cooling, and dehydration to remove process containments, including water, CO2, and

hydro-carbons. The cooled high-pressure air is then passed on to the refrigeration island, in which it is further cooled and liquified by a countercurrent heat exchanger.

Lastly, to distill the liquified air into oxygen, nitrogen, and if needed, argon streams, distillation columns are used. These operate with various vapor-liquid contact devices, such as trays or packaging materials (Song et al., 2019), with cryogenic operating temperatures between -195.8 and -183.0 °C, representing the boiling temperatures of nitrogen and oxygen (Rackley, 2017). Argon streams are usually considered optional because the boiling point of argon is quite sim-ilar to that of oxygen, -185.87 versus -183.0 °C, making optimal purity both difficult and

(33)

ex-As previously mentioned, cryogenic distillation may be of interest for capturing CO2 from flue

gas. Cryogenic CO2 separations operate similarly to air separations in that the CO2-containing

gas mixture is first pretreated and compressed to extract moisture. The compressed gas is then passed through a series of heat exchangers to lower the temperature (Wilcox, 2012) before being transferred to a cryogenic column, where the temperature and pressure are further ma-nipulated, allowing the CO2 to liquefy while the other components remain in gaseous form

(Aaron & Tsouris, 2005). As a result, the component stream is mainly divided into two parts after the distillation column: top and bottom products, with the former containing vapor and the latter containing liquid (CO2). A portion of this CO2-rich stream is recycled back to the

distillation column to provide vaporization heat, while the remaining is separated and ex-tracted from the separator (Song et al., 2019).

Due to the benefits it entails, cryogenic distillation is a commonly employed technology in to-day’s industry. It yields high purity of captured CO2, up to 99% (Song et al., 2019). This high

CO2 purity has an advantage in that it can either be solid or liquid, making transportation

through pipelines or tankers for sequestration considerably easier (Aaron & Tsouris, 2005). In addition to being ready for transportation, the liquefied CO2 is produced at relatively low

pres-sure and requires no additional preconditioning, thus reducing the energy penalty (Song et al., 2019; Leung et al., 2014). Moreover, cryogenic technologies are based on a well-established process involving refrigeration and compression, both of which are proven and commonly used technologies (Yousef et al., 2018) and do not require the use of any additives or chemical rea-gents (Ghasem, 2020). Despite its extensive industrial use and significant benefits, cryogenic distillation requires a substantial amount of energy, which often accounts for more than half of the plant’s operating costs (Ebrahimzadeh et al., 2016). Another drawback of the cryogenic process is that it necessitates the removal of potential freezing compounds, such as water, NOx,

(34)

3.5

Calcium Looping

Calcium-looping (Ca-looping) is a post-combustion capture technology that uses high-temper-ature calcium sorbents, such as limestone to capture CO2 from flue gases. The Ca-looping

cap-ture process comprises the use of two interconnected fluidized bed reactors: carbonator and calciner, in which the Ca-sorbent is reversibly reacted between its carbonate form (CaCO3) and

its oxide form (CaO), to separate CO2 from flue gases (Fennell, 2015). In the carbonator, the

calcium oxide particles: CaO, reacts with CO2 to form CaCO3 and then transport it to the

cal-ciner, where the reaction is reversed, and a pure stream of CO2 and CaO reactant is produced.

The reactant is then returned to the carbonator to begin a new cycle, while the primarily CO2

-containing stream is further transferred to compression and storage (Coppola & Scala, 2020). The Ca-looping capture process offers a relatively small decrease in power plant efficiency (Pawlak-Kruczek & Baranowski, 2017). However, one major issue with Ca-looping is sorbent deterioration with cycles, which is a significant disadvantage of this approach. According to Fennell (2015), limestone from natural sources loses roughly 15 - 20% of its potential to absorb CO2 during each cycle from calcination to carbonation. However, because limestone is a

low-cost natural resource, high purge rates from the Ca-looping system can be maintained without compromising the process’s profitability.

3.6

Chemical Looping

Chemical looping combustion (CLC) is carbon capture technology that operates similarly to oxyfuel combustion in that it extracts oxygen from the air to create a nitrogen-lean combustion environment without the need for an air separation unit. In its place, it uses two interconnected fluidized bed reactors: an air reactor and a fuel reactor, with a metal oxide carrier (MexOy)

circulating between the two (Breeze, 2015), as illustrated in Figure 5. Several materials have been identified as suitable oxygen carriers, including iron oxide (Fe2O3), nickel oxide (NiO),

copper oxide (CuO), and di-manganese trioxide (Mn2O3) (Luo et al., 2018). The metal oxide

carrier is heated in the presence of air in the air reactor, where it preferentially captures oxygen and leaves a stream of nitrogen. The oxygen-rich carrier is then transferred to the fuel reactor, where it is fully or partially reduced due to the combustion reaction (Song et al., 2019). For example, if methane is considered as the fuel, chemical looping combustion proceeds accord-ing to the followaccord-ing reaction,

𝐶𝐻3(4)+ 4𝑀𝑒8𝑂:(;)→ 4𝑀𝑒8𝑂:=>(;)+ 𝐶𝑂?(4)+ 2𝐻?𝑂(4) (3 − 1)

Following this fuel oxidation step, the metal oxide carrier is looped back into the air reactor for re-oxidation, while the produced CO2 in the flue gas is easily extracted through water vapor

Figure

Figure 1: Illustration scheme of the dominant carbon capture technologies.
Figure 7: Illustration of unit operation of cryogenic air separation unit.
Figure 8: CO 2  avoided, and CO 2  captured.
Figure 11: Illustration of the CPU flow diagram. Source: Li et al. (2013).
+6

References

Related documents

Initially, the free carrier absorption measurements are done as explained in chapter 4 and the data for variation of absorption of probe light with respect to time is extracted. This

There are most often three positions in a large scale combustion process where SCR is placed; either directly after the combustion chamber (high-dust position), after a flue gas

A comprehensive set of biochar mineralisation experiments in soil was retrieved by Lehmann (2015) which includes results from various biochar samples (feedstock, production

The Uppsala heat and power system consists of a district heating network together with several production units for district heating and cooling, steam and electricity.. In

Problemet är idag att det inte finns någon klar och tydlig kunskap om vilka produkter eller material som innehåller höga halter av oönskade metaller och att det just är

According to the DAF reactors manufacturer, Purac, the design values for Margretelund WWTP’s DAF process was a surface load of 4 m/h with a percentage of total flow rate

Keywords: The Shining, Stanley Kubrick, Stephen King, adaptation studies, metafiction, postmodernism, The Uncanny, film theory, film adaptation... Previous Research

This study examines the relationship between per capita GDP and per capita emissions of the greenhouse gas carbon