Modelling of Pulverized Coal Power Plants in Carbon Capture and Storage (CCS) Networks

Transcription

1 Modelling of Pulverized Coal Power Plants in Carbon Capture and Storage (CCS) Networks Hugo Manuel Dias Rodrigues Dissertation for the Master s Degree Chemical Engineering Chair: Adviser: Adviser PSE: Member: Member PSE: Juri Prof. João Manuel Nunes Alvarinhas Fareleira Prof. Henrique Aníbal Santos de Matos Eng. José Alfredo Ramos Plasencia Vítor Manuel Geraldes Fernandes Javier Rodriguez October 2012

2

3 Modelling of Pulverized Coal Power Plants in Carbon Capture and Storage (CCS) Networks Hugo Manuel Dias Rodrigues Dissertação para obter o grau de Mestre em Engenharia Química Presidente: Orientador: Orientador PSE: Vogal: Vogal PSE: Júri Prof. João Manuel Nunes Alvarinhas Fareleira Prof. Henrique Aníbal Santos de Matos Eng. José Alfredo Ramos Plasencia Vítor Manuel Geraldes Fernandes Javier Rodriguez Outubro 2012

4

5 Education is the most powerful weapon which you can use to change the world. Nelson Mandela

6

7 Acknowledgments I would like to express my deepest gratitude to my supervisors Prof. Dr. Henrique Matos and Alfredo Ramos, for all the help they provided and the guidance during the past seven months. Also would like to thank all the gccs team especially Dr.Adekola Lawal for all the support and availability to analyze and discuss ideas. To all the people that made my life easier in London since all PSE to my cousin Joao Cunha which welcome me in his home for the first days. My flat mates and coworkers: Élton Dias, Mário Calado and Ricardo Fernandes for all the time we spend together and the adventures discovering London, its attractions and social life. To my friends and my girlfriend which helped not only in this step but throughout my degree and my life the deepest recognition. At last, but not less important, I would like to thank my family for giving me all the conditions and support all this years, to accomplish my objectives. To all of you a big thank you! iii

8

9 Abstract Carbon dioxide emissions to the atmosphere are getting the worldwide attention. Carbon Capture and Storage technologies promise to take an important role in climate change mitigation. However, nowadays no software can do a full chain analysis of the Carbon Capture and Storage for power plant systems. For that purpose Process System Enterprise with the support of Energy Technologies Institute (ETI) and other associates is developing a new tool-kit, which should have the models for entire chain addressing different types of power plants, injection and storage. In the scope of this work pseudo steady state component models of a supercritical power plant were developed as well as the composite model that fully represents a supercritical power plant. Two modes, design and operational were developed for the composite model to be able to simulate part load in addition with turbine following control. A daily cycle simulation was analysed and sensitivity studies were made on the boiler efficiency and on the reheat vapour temperature. The mathematical modelling was implemented in the commercial software gproms. This is a highly complex system where all the interactions and all the recirculation of information that appears in the steam cycle were studied and successfully captured by the model. The sensitivity analysis shows that an increase in the reheat vapour temperature improves the power production and the gross efficiency. Pointing out that an improvement of the equipment manufacture material will lead to more efficient power plants. Keywords Supercritical Pulverized Coal Power Plant, Turbine following control, Modelling, Simulation, gproms v

10

11 Resumo As emissões de dióxido de carbono estão cada vez mais a chamar à atenção do mundo e as tecnologias de captura de CO 2 prometem ter um papel importante na redução dos impactos causados por estas. No entanto, hoje em dia não existe um software capaz de simular toda a cadeia da captura do dióxido de carbono para as centrais termoelétricas. Por esta razão a empresa Process System Enterprise com o apoio do Energy Technologies Institute (ETI) e outras empresas está a desenvolver um novo produto que vai dispor de uma biblioteca com todos os modelos, abrangendo diferentes tipos de centrais até à injeção e armazenamento. No tema desta dissertação desenvolveu-se modelos de pseudo estado estacionário para uma central a carvão, mas também o modelo compósito que representa a central termoeléctrica supercrítica de carvão. Dois modos foram desenvolvidos: dimensionamento e operação para o modelo compósito, permitindo simular diferentes produções de energia juntamente com o controlo turbine following. Foi simulada uma produção diária e feita uma análise de sensibilidade à eficiência da caldeira e à temperatura do vapor reaquecido. O software usado neste trabalho foi o gproms. Este sistema é muito complexo onde todas as interações e recirculações de informação foram estudadas e fielmente capturadas pelo modelo compósito. A análise de sensiilidade do vapor reaquecido mostra que um aumento nesta provoca um aumento da produção eléctrica assim como da eficiência da central. Este fato mostra que um aumento da resistência térmica dos materiais levará no futuro a centrais com uma maior eficência. Palavras Chave Central Supercrítica de Carvão Pulverizado, Turbine following, Modelação, Simulação, gproms vii

12

13 Contents 1 Introduction Motivation State of the Art Original Contributions Dissertation Outline Background Review Carbon Capture and Storage Technologies Post Combustion Capture Pre-Combustion Capture Oxyfuel Compression and transportation Injection and storage Power Plant Steam Generation Steam Cycle A Carnot Cycle B Rankine Cycle C Reheat-Regenerative Cycle Integration with CCS Equipment Boiler Turbine Feedwater heater/deaerator A Deaerator B Feedwater Heater Condenser Flue Gas Desulphurization unit A Limestone forced oxidation (LSFO) Electrostatic Precipitator Gas/Gas Heater ix

14 2.3.8 Governor valve Stack Generator Control Boiler/turbine control Feed water control Steam temperature control Deaerator Control Condenser Control Feedwater heaters Control Materials and Methods gproms software Model development workflow gccs model library gccs structure Mathematical Modelling of PCPP Components Feedwater Heater Inlets Outlets Variables Nomenclature Equations Degree of freedom gproms interface Boiler Steam Condenser Inlets Outlets Variables Nomenclature Equations Degree of freedom Electrostatic Precipitator Inlets Outlets Variables Nomenclature Equations Degree of freedom Flue Gas Desulphurisation Inlets Outlets x

15 4.4.3 Variables Nomenclature Equations Degree of freedom Gas/Gas Heater Inlets Outlets Variables Nomenclature Equations Degree of freedom Control Valve Inlets Outlets Variables Nomenclature Equations Degree of freedom Supercritical Pulverized Coal Power Plant Modelling Design Mode Results and Discussion A Stream conditions B Key Performance Indicators C Equipment Parameters Operational Mode Control Mode Control loops A Condenser pressure B Deaerator drum level C Turbine following control loop D Others controls Step change in power plant load Daily Cycle Sensitivity analysis Boiler efficiency Reheat temperature Conclusions and Future Work Future Work Bibliography 81 Appendix A gccs Model Library A-1 xi

16 Appendix B Main Operating Conditions and Results B-1 xii

17 List of Figures 1.1 Carbon dioxide annual emissions from fossil fuels world wide, IEA values Global CO 2 emissions and Green House Gases emission reductions Technical options for CO 2 capture from coal-power plants Block diagram illustrating Power Plant with Post-Combustion CO 2 Capture Block diagram illustrating Power Plant with Pre-Combustion CO 2 Capture Block diagram illustrating Power Plant with Oxyfuel CO 2 Capture Scheme of an Advanced Supercritical Power plant Carnot thermodynamic cycle Rankine thermodynamic cycle Reheat regenerative thermodynamic cycle Representation of a power plant steam cycle Boiler schematic representation Schematic representation of a feedwater heater Schematic representation of a LSFO system Model development workflow gccs feedwater heater example of a specification gccs environment for a fwh simulation gccs boiler steam condenser model configuration gccs electrostatic precipitator model configuration Specific power relationship with the efficiency of an ESP unit gccs FGD model configuration Valve stem position for different flow characteristics gccs control valve model configuration Supercritical steam cycle flow diagram gccs diagram of a Supercritical Pulverized Coal Power Plant Flow diagram of the HP pressure feedwater heaters side Condenser pressure control loop Deaerator drum level control loop Turbine following control loop xiii

18 5.7 Load change and throttle pressure deviation for the different types of load control Load change and throttle pressure deviation results for 4% step change in load Daily cycle of the Portuguese National Grid from 3 of September Simulation results from the daily cycle schedule Boiler pressure and governor valve stem position response during the daily cycle schedule Feedwater and coal deviation from full capacity load for the daily cycle schedule Responses of the manipulated and controlled variable for a change of load from 90% to 100%, for the level control Responses of the manipulated and controlled variable during the daily load, for the condenser pressure control Variation of the feedwater heater and deaerator temperature and pressure during the day A.1 gccs boiler model configuration A-2 A.2 gccs turbine model configuration A-3 A.3 gccs deaerator model configuration A-4 A.4 gccs governor model configuration A-4 A.5 gccs drum model configuration A-5 A.6 gccs pump model configuration A-6 A.7 gccs blower model configuration A-7 A.8 gccs generator model configuration A-7 A.9 gccs recycle model configuration A-8 A.10 gccs recycle model configuration A-8 A.11 gccs controller model configuration A-9 xiv

19 List of Tables 2.1 Classification of pulverized coal power plants according to UNIPEDE.[1] Nomenclature of the variables used in the feedwater heater model Nomenclature of the variables used in the condenser model Nomenclature of the variables used in the electrostatic precipitator model Nomenclature of the variables used in the FGD model Nomenclature of the variables used in the GGH model Nomenclature of the variables used in the Control Valve model Main assignments in the flowsheet Key performance indicators (KPI) deviation Equipment parameters for the feedwater heaters and the condenser Equipment parameters for the turbines Specification trade-offs for operational mode Controllers parameters and stabilization time for every controller Table summarizing the controllers parameters Power plant Key performance indicators for different loads Boiler efficiency sensitivity study Reheat temperature sensibility study B.1 Main operating conditions deviation from reference [2].(for stream identification please see fig 5.1) B-2 B.2 Key performance indicators (KPI) and flue gas composition from reference [2] B-2 B.3 Main operating conditions from design results.(for stream identification please see fig 5.1) B-3 B.4 Key performance indicators (KPI) and flue gas composition from design results..... B-3 B.5 Equipment parameters for the feedwater heaters and the condenser B-4 B.6 Equipment parameters for the turbines B-4 B.7 Main operating conditions from operational results.(for stream identification please see fig 5.1) B-4 B.8 Key performance indicators (KPI) and flue gas composition from operational results... B-5 B.9 Main operating conditions for 95% load.(for stream identification please see fig 5.1)... B-5 xv

20 B.10 Key performance indicators (KPI) and flue gas composition for 95% load B-6 B.11 Main operating conditions for 90% load.(for stream identification please see fig 5.1)... B-6 B.12 Key performance indicators (KPI) and flue gas composition for 90% load B-7 B.13 Main operating conditions results for a reheat temperature increase of 1.2%.(for stream identification please see fig 5.1) B-7 B.14 Key performance indicators (KPI) and flue gas composition for a reheat temperature increase of 1.2% B-8 B.15 Main operating conditions results for a reheat temperature increase of 7.4%.(for stream identification please see fig 5.1) B-8 B.16 Key performance indicators (KPI) and flue gas composition for a reheat temperature increase of 7.4% B-9 B.17 Main operating conditions results for a reheat temperature increase of 14.5%.(for stream identification please see fig 5.1) B-9 B.18 Key performance indicators (KPI) and flue gas composition for a reheat temperature increase of 14.5% F-10 xvi

21 xvii

22 Abbreviations xviii Abbreviation AGR ASU CCGT CCS CFBD CLC CW DCA Deff DFGD DOF EDF EOR E.ON ESP ETI EU FGD FTR FWH gccs GHG GGH gproms HP IEA IGCC ITM IP LHV KPI LP LSFO MEA MSD OSTG PC PCC PFBC PSE RH SCR SH SNCR TTD UK USC WFGD Description Acid Gas Removal Air Separation Unit Combined Cycle Gas Turbine Carbon Capture and Storage Circulating Fluidized Bed Combustion Chemical Looping Combustion Cooling Water Drain Cooler Approach Drain effectiveness Dry Flue-Gas Desulfurization Degree of Freedom Electricite de France Enhanced Oil Recovery E.ON Energy Limited Electrostatic Precipitator Energy Technologies Institute European Union Flue-Gas Desulfurization Feedwater Temperature Rise Feedwater Heater CCS system modelling toolkit, based on gproms Green House Gases Gas Gas Heater gproms ModelBuilder High Pressure International Energy Agency Integrated Gasification Combined Cycle Ion Transport Membrane Intermediate Pressure Lower Heating Value Key Performance Indicators Low Pressure Limestone Forced Oxidized Monoethanolamine Model Specifications Document Once-Through Steam Generation Pulverized Coal Pos Combustion Capture Pressurized Fluidized Bed Combustion Process Systems Enterprise Ltd. Reheat Selective Catalytic Reduction Superheat Selective Non Catalytic Reduction Terminal Temperature Difference United Kingdom Ultra Supercritical Wet Flue-Gas Desulfurization

23 List of Symbols Latin Symbols Variable Description Units A Heat transfer area m 2 C Components Cv Valve flow coefficient kg Pa s 1 F Mass flowrate kg s 1 h Specific enthalpy J kg 1 H Head J kg 1 H Henry coefficient Pa m 3 mol 1 L Leakage M Molecular weight g mol 1 N Number of p Pressure Pa P Power W Q Heat duty W r CaCO3/SO 2 Molar ratio of limestone/so2 removed R f Rangeability factor T Temperature K U Overall heat transfer coefficient W m 2 K 1 V sp Valve stem position w Mass fraction W Work duty W wt Solids content x Mole fraction Greek Symbols Variable Description Units γ Isentropic index (blower) γ Valve flow exponent (valve) γ Mass concentration mg Nm 3 Γ Mole percentage % Differential f H Standard heat of formation J mol 1 lm T Logarithmic mean differential temperature K r H Θ Standard heat of reaction J mol 1 η Efficiency % κ Rate tph µ Purity Continued on next page xix

24 Table 2 Continued Variable Description Units ρ Mass density kg/m 3 τ Time constant s - ν Stoichiometric coefficient of the component Subscripts Variable A air ash ashout B cap cold condenser eq f F W gyp hot in lime max mindif f out removed s sat shell spec vapour var w Description Reactants Air Ash Outlet stream with only ash Products capture Cold Condenser equations Fraction Feedwater Gypsum Hot Inlet stream Limestone Maximum Minimum difference Outlet stream Removed Isentropic Saturation Shell Specified Vapour Variables Water Superscripts Variable act Aux cold dry F W hot normal Steam Description Actual Auxiliaries Cold Dry basis Feedwater Hot Normal conditions Steam xx

25 1 Introduction Contents 1.1 Motivation State of the Art Original Contributions Dissertation Outline

26 Since 1997 with Kyoto protocol, global concern with climate changes from carbon dioxide emissions have been increasing. In 2008, UK government took the next step committing by law to reduce by 2050, 80% of carbon dioxide emissions from 1990 level.[3] [4] Over the last decades, carbon dioxide emissions has been increasing at a very high rate, with new countries like China and India starting to growth economically so their energy demands, therefore, his carbon dioxide emissions. Besides, this increase in energy demand, a major issue is the use of coal which amongst the fossil fuels, releases the most CO 2 on combustion around 960 gco 2 e/kwh in comparison to natural gas with 443 gco 2 e/kwh, for example.[5] Figure 1.1: Carbon dioxide annual emissions from fossil fuels world wide, IEA values.[6] Analysing figure 1.1, in 2009, carbon dioxide emissions dropped in relation to the 2008, the cause of that reduction was the economic crisis. However, despite the slow global economic recovery, 2010 saw the largest single year increase in global human CO 2 emissions from energy, growing a whopping 1.6 billion metric ton from 2009, to 30.6 billion metric ton. Last year, a new high record emission was achieved, 31.6 billion metric ton. Reducing significantly the carbon dioxide emissions level requires several measures: increasing energy efficiency, diversify energy sources with renewable energies and implementation of carbon capture and storage chains. Figure 1.2, exemplifies exactly that where we can see that CCS technology will represent a decrease of 19% in CO 2 emissions in 2050 to meet the target. This contribution is more than from renewable energies and more than triple the contribution from nuclear. Nowadays, carbon capture and storage is a very expensive technology. Two major factors will determine viability of such a technology, the cost of technology itself (that tends to decrease as the technology gets more mature) and the cost of carbon dioxide emissions. As the largest contribution to CO 2 emissions is from the burning of fossil fuel, particularly in producing electricity, three main processes are being developed to capture CO 2 from power plants that use coal or gas. These are: Post Combustion Capture (PCC) Pre-Combustion Capture Oxy Combustion 2

27 Figure 1.2: Global CO 2 emissions and Green House Gases emission reductions [7] In the world there are 74 large scale integrated CCS projects, from this 14 are already in operation or under construction. Europe accounts with 20 of those projects, being United Kingdom the most active country in this area. This confirms the effort being done by the UK in reducing Green House Gases (GHG), and creating conditions to accomplish the targets, either by integrated large scale CCS project or by modelling tool-kit that allows a simple understanding of the system, make technoeconomical decisions, optimization studies and control tuning.[8] 1.1 Motivation Mitigating global warming is the major challenge of the next century, global conscience about the problem has been rising along with CO 2 emissions. Measures need to be taken to accomplish a significant reduction in carbon dioxide emissions and Carbon Capture Storage (CCS) technology promises to be an important technology for climate change mitigation. Coal power plants emitted in 2011 almost 14 billion metric tons of carbon dioxide representing the most significant share in global emission with 45% of the total carbon dioxide emissions. Applying CCS system to this kind of power plant will reduce significantly CO 2 emissions. [9] [10] In September of 2011, UK government by the Energy Technology Institute (ETI) delivered Process Systems Enterprise (PSE) and others stakeholders like EDF, E.ON, Rolls - Royce, CO 2 DeepStore and e4tech a project to build a carbon capture and storage modelling tool-kit. The completion time for this project is scheduled to spring of The tool-kit allows the users to study the differences and the effects on the system at part load, start up and shutdown, being able to identify the key issues. This project will help to support future design of integrated CCS in power plants, analysing economical effects on the entire cluster, help the owners and developers of power plants to understand the all concept from power generation to storage and its trade-offs, technology suppliers that want to understand new possibilities around the system and policy makers keen to understand the issues around the overall CCS system. 3

28 The main objective of this work is the development of (pseudo) steady-state models of relevant unit operations in pulverized coal power plants as feedwater heaters, condenser, gas gas heater, electrostatic precipitators and flue gas desulphurization unit, as well as modelling the pulverized power plant as a component model using the existing gccs power plant library and implement turbine following control and other relevant power plant control for part load operation. A literature review was conducted to see what have already been done in this area. 1.2 State of the Art A literature review was made to see what has been done. There is a little information regarding supercritical steam cycles, however several attempts have been made for static and dynamic simulation but most deal with combined cycles or sub critical power plants. Supercritical power plant studies were made by Sergio Espatolero [11], where an optimization of the boiler cold-end was analyzed as well as its integration with the supercritical steam cycle. From the same author and Romeo [2], a supercritical power plant was designed to aim the optimal integration with the capture plant, especially the energy requirements of CO 2 amine scrubbing which require specific steam drawn offs from the turbine cycle. Both studies took place in the ASPEN software. Falah, [12] presents a static and dynamic simulation model of a supercritical once-through heat recovery steam generator (SC HRSG) and its application to investigate the load changes and startup processes for next generation high efficiency combined cycles. The work was developed in the commercial simulation software named Advanced Process Simulation Software (APROS). Miroslav Variny, [13] presents a part load operation study for a combined cycle to improve the efficiency provisioning the auxiliaries services. Performance studies were made by Chia-Chin Chuang [14], for a combined cycle power plant with variable condenser pressure and variable load. Some reports can be found that simulate sub critical and supercritical power plants, such as DOE [15] and ALSTOM [16], in commercial softwares such as ASPEN. This type of reports have case studies for real power plants considering introduction of capture plant or new arrangements to improve the efficiency of the power plant. Changliang Liu [17], presents and updated, from 2011, overview of the modelling and simulation of thermal power plants. The thermal process control is review as well as the thermal process modelling which the author categorizes in three main areas: simplified boiler-turbine models for CCS research, dynamic models of subsystem and thermal performance calculation and optimization model. He identifies as one of the main challenges in the future the improvement of the accuracy of the models with actual power plant data. Modelling and simulation is the base of optimal operation and control and plays an important role in energy saving in thermal power plants.[17] There still more work to do on improvements of the models and on the analysis of the supercritical power plant steam cycle. Some more studies such as the ones made by Espatolero [11] would be welcome, however accuracy of the models need to be 4

29 analyzed to perform part load operations studies. Part load operation, control and the integration with the capture plant are the main areas that can be further developed. 1.3 Original Contributions The main contribution in this work is the development in gproms of steady state models adequate for power plant simulation. The implementation of those models together with the existing gccs library to model a supercritical coal power plant. Moreover, the simulation of part load operation plant adding the adequate control system. 1.4 Dissertation Outline This dissertation is organized as follows: Chapter 2 presents a background review focusing on carbon capture and storage technologies such as post combustion; in power plant concepts and the thermodynamics behind the steam cycle; the power plant equipments and their concepts and uses in the system; and to finalize a control review was done to see what are the most typical types of control used and what is controlled. In the chapter 3 a explanation on the materials and methodology that were used during this work. The chapter 4 presents a description of the mathematical modeling developed, the concept and assumptions behind the models needed to be implemented for the gccs power plant library. In the chapter 5, the flowsheet construction is explained and the various modes possible to simulate: design and operational. Is also presented the controls strategies implemented especially the turbine following and a daily cycle simulation, along with the sensitivity analysis on the boiler efficiency and on the reheat temperature in the boiler. To finalize, chapter 6 presents the conclusions of the thesis and suggestions for future work. 5

30 6

31 2 Background Review Contents 2.1 Carbon Capture and Storage Technologies Power Plant Equipment Control

32 2.1 Carbon Capture and Storage Technologies The main technologies being developed to capture CO 2 from power plants that use coal or gas are: Post combustion (PCC), Pre combustion and Oxy Combustion. These systems differentiate from each other in the technology used and where CO 2 in removed. Pre-combustion capture is applicable to integrated gasification combined cycle (IGCC) power plants, while post- and oxy-combustion capture could be applied to conventional pulverized coal (PC)- fired power plants. Figure 2.1: Technical options for CO 2 capture from coal-power plants.[7] Analysing the figure 2.1, is possible to see that post combustion and oxyfuel combustion have similar concepts, being the main difference the use of oxygen instead of air in oxyfuel technology. This concept doesn t need capture plant after the power station which represents an advantage. On the other hand post combustion can be applied to newly designed fossil fuel power plants, or retrofitted to existing plants. Air is used along with coal in the boiler leading to relatively small CO 2 concentration being used different capture methods: absorbent, adsorbent, and membranes. Some other capture methods are being study with far less emphasis: converting CO 2 to mineral and employing biofixation. Pre-combustion technology use a totally different concept, first the coal is gasified with air or oxygen under high pressure, then it goes to water gas-shift reactor that converts CO to CO 2 while producing additional H 2, thus increasing the CO 2 and H 2 concentrations. CO 2 removal is the second step using Acid Gas Removal (AGR), then a hydrogen rich syngas used as a fuel in a combustion turbine combined cycle to generate electricity Post Combustion Capture Post combustion capture refers to the separation of CO 2 from flue gas, is mainly used in typical coal fired power plant but can also be applied to integrated gasification combined cycle (IGCC) and 8

33 natural gas combined cycle (NGCC).[3] Figure 2.2: Block diagram illustrating Power Plant with Post-Combustion CO 2 Capture.[3] In a coal-fired power plant, coal is combusted with air and released heat that vaporizes water from the steam cycle to produce energy. The combustion result composition has N 2, CO 2, H 2 O, O 2, SO x, NO x, and particulate matter. Firstly nitrous oxides are removed in selective catalytic reduction (SCR), then particulate matter is removed from the gas in an electrostatic precipitator (ESP) followed by flue-gas desulphurization unit to remove sulphur compounds. [8] Conventional pulverized coal power stations release the flue gas to the stack after pollutants control, but now with CO 2 emissions control flue gas will go into another process to reduce the amount of CO 2 emitted. The type of fuel has also influence in the application of this technology to natural gas because typical CO 2 coal composition is 10-15%, but for natural gas is 4-5% and for the same amount of power output generated coal power plant generate twice the CO 2 amount than natural gas. This means that although natural gas has less CO 2 to remove, apply carbon capture to coal power plants is less energy consuming for the same amount of CO 2 capture because the composition is higher. CO 2 capture process can be done by absorption, adsorption, membranes, mineralization and biofixation. Absorption is a chemical process where CO 2 is dissolve into a liquid solution. Actually almost every PCC projects use absorption based methods, being monoethanolamine (MEA) aqueous solution the most common example. In absorption based method flue gas containing CO 2 is contacted with solvent in gas liquid contactors (absorber), where mass transfer occurs with CO 2 transfers from gas phase to liquid phase. Then CO 2 rich solution is pumped to the regenerator where is heated and CO 2 is release since the solubility decreases with temperature increase. The lean solution is circulated back to the absorber, and the carbon dioxide collected is dried. [8] Chemical absorption process requires the extraction of a relatively large volume of low-pressure steam from the power plant s steam cycle, to regenerate the solvent, which decreases the gross electrical generation of the plant. Adsorption is another possible choice, but still in an earlier demonstration stage. Physically CO 2 is adsorbed in a surface of a solid sorbent material, which can be zeolites, zeolitic imidazolate and metal organic frameworks. Typically Van der Waals forces prevail for physisorption but chemisorption can happen as well with 9

34 stronger covalent bonding. These kinds of systems are usually applied in packed bed or fluidized bed systems. Adsorption could be a valid alternative to absorption analyzing heat capacities because solid heat capacities are lower than liquids, which would reduce the energy penalties. However other effects must be taken in account like heats of reaction, working capacity, and others. Membranes technology could also be applied to extract CO 2 from the flue gas stream. If the membrane material is more permeable to CO 2 than the other compounds, carbon dioxide will selectively permeate. Partial pressure is required to create a gradient high enough to allow CO 2 transport across the membrane, that can be obtain pressuring the flue gas side, or applying vacuum to the other side.[8] Potentially these units can allow more flexible operation as well as providing less energy consuming capture plant Pre-Combustion Capture Pre combustion as the name implies is the method that requires CO 2 capture before the combustion section, outlined in figure 2.3. Figure 2.3: Block diagram illustrating Power Plant with Pre-Combustion CO 2 Capture.[3] In the gasifier, fuel is converted into gaseous components by applying heat under pressure in the presence of steam and limited O 2. Syngas is the gasified product, rich CO and H 2 mixture. In the next step, syngas goes to the water gas shift reactor where in the presence of water, CO is converted to CO 2 and more hydrogen is produce, by the following reaction: H 2 O + CO CO 2 + H 2 (2.1) After this, carbon dioxide is separated from the hydrogen that will be burnt in the combustion turbine combined cycle to generate electricity. The capture process is accomplished under pressure by an acid gas removal process of absorption in a solvent followed by regenerative stripping of the rich solvent to release the CO 2. Acid gas removal systems can be chemical or physical absorbents. State of art processes are the physical ones, with the glycol-based Selexol process and the methanol-based Rectisol process. [3] Physical sorbents dissolve acid gases under pressure and released when the pressure is decreased or the temperature increased. In a AGR unit there is two stages, the first stage absorber/stripper to remove the sulphur and in the second stage the CO 2 is remove. [8] 10

35 In comparison, chemical absorbents react with the acid gases and require heat to reverse the reactions release the acid gases, this heat requires withdrawn from the steam cycle large amounts of steam-heat for solvent regeneration. However, initial capital costs are lower than physical processes. Pre combustion is usually associated to IGCC power plants but can also be applied to Natural gas combined cycles and Natural gas reforming and partial oxidation. Capture addition to IGCC reduces efficiency by 7-8 %, although this is lower than post combustion capture still very high and some major development in capture technology is required like better integration and IGCC component equipments (e.g. air separation unit, shift, gas turbines for hydrogen) improvements to decrease efficiency losses. [8] Oxyfuel Oxy-combustion is an alternative to post-combustion CO 2 capture for new and existing conventional PC-fired power plants that offers the potential for high percentage of CO 2 capture. The nitrogen in air that is use in the combustion of Pulverized coal power plants dilute flue gas CO 2 content, being later required capture plants. Oxyfuel concept is exactly taking this nitrogen inert out, burnt the fuel only with oxygen, this makes the flue gas CO 2 content extremely high about 90% in dry basis.[3] This technology depending on the regulations can only require minor purification (dry CO 2 ) or none thereby less costs than purification in PCC. Figure 2.4: Block diagram illustrating Power Plant with Oxyfuel CO 2 Capture.[3] A simplified process schematic of oxy-combustion CO 2 capture is shown in figure 2.4, flue gas recycle is need about 80% flue gas stream to keep the boiler combustion and heat transfer characteristics of combustion with air, having CO 2 as inert instead of N 2. At the moment, there still no significant commercial scale, however oxyfuel relies on normal conventional equipment that is already available at the scale necessary for power plant applications, and key process principles like power generation is obtained by optimized integration of the steam cycle. A major issue is the high capital and operational cost of the air separation unit that doesn t compensate impurities reduction in the flue gas. Improvements in ASU technology or the development of more cost-effective oxygen production are required. Two leading technologies to produce less energy consuming oxygen are: 11

36 Chemical Looping Combustion (CLC) Ion Transport Membrane (ITM) Chemical looping combustion idea is to separate oxygen from nitrogen by a reversible reaction suitable by solids. Then oxygen molecules are transported to a different vessel where oxygen is release and combustion occurs.[18] Ion transport membrane uses selected ceramic materials that at high temperature (1000 C) and moderate pressures allow oxygen to migrate through the solid. [8] Compression and transportation After capture the carbon dioxide it s important to reduce its volume to a more cost effective transport and storage. Carbon dioxide can be compressed to liquid or supercritical phase, critical point is at 31.1 C and 72.9 atm. Either compression or a combination of refrigeration/pumping is done to convert the CO 2 gas to a supercritical fluid.[3] The transport can be done by pipeline in supercritical phase, ship or road tanker if the CO 2 is in liquid phase at moderate pressures and low temperature. The method most frequently used is the pipeline transport strategy to take the CO 2 from the power plant to a selected location for permanent, safe underground storage or beneficial reuse.[19] Injection and storage Geologic storage involves the injection of CO 2 into underground formations that have the ability to secure CO 2 for long periods. Storage locations for carbon dioxide can be: [20] Deep saline formations Depleted or partially depleted oil fields Depleted or partially depleted natural gas fields Coal seams In partially depleted oil and natural gas fields, CO 2 can be used in enhance oil recovery (EOR) a technique already well established in the oil industry. [3] 2.2 Power Plant Coal is the most abundant fossil fuel in the world, being a relatively inexpensive energy source however it produces relatively high levels of pollution. In 2009, almost 40 percent of the total electrical power produce in the world was from coal source. [21] Of the number of coal-fired electricity plants, pulverized coal (PC)-fired power plant is the most used in electricity generation plants around the globe. The other options for coal electrical power generation are: Circulating Fluidized Bed Combustion (CFBD) Pressurized Fluidized Bed Combustion (PFBC) 12

37 Integrated Gasification Combined Cycle (IGCC) However pulverized coal (PC) -fired power plant has the highest reliability and commercial readiness for high electricity production capacity. In 2004, PC-fired power plant represented 99% of coal total electrical production.[1] Different PC-fired power plants types can be encountered as improvements appear along the years. Table 2.1, compare the different types. Table 2.1: Classification of pulverized coal power plants according to UNIPEDE.[1] Category Unit Subcritical Supercritical Advanced Ultra Supercritical Supercritical (USC) Live steam pressure MPa 16.5 > >30 Live steam temperature C >600 Reheat steam temperature C No reheat >600 Single Reheat C No Yes Yes No Double Reheat - No No No Yes Generating efficency % >46 A schematic figure for a pulverized coal power plant can be seen in figure 2.5. In this figure, the entire system interaction is capture, from steam generation and steam cycle integration systems to the power distribution grid. Figure 2.5: Scheme of an Advanced Supercritical Power plant.[22] Steam Generation A scheme of a pulverized power plant can be found in figure 2.5. The key equipment for steam generation is the boiler, where the coal is burned with air, to heat up and vaporize the water to generate electricity in the steam cycle. 13

38 Inside the boiler there is a control of nitrogen oxides (NO x ) like nitrogen dioxide and nitrogen dioxide that are produce in the combustion, this control is done in a Selective Catalyst Reduction (SCR) reducing NO x concentration to the legislation imposed limits. Uses ammonia to react with NO x in the presence of a catalyst and produce water and nitrogen. An air heater can be found in the boiler scheme as well, in the air heater the air is integrated with the hot flue gas to take advantage of the heat still available. Particulate removal of ash or particulate matter is done in a fabric filter or Electrostatic Precipitator (ESP). Ash removal is followed by the sulphur removal that is done in the Flue Gas Desulphurization unit (FGD), this uses limestone to react with the sulphur content of the flue gas and gets a sub product which is gypsum. After these purifications steps to control pollutants emissions, flue gas goes to the stack where is dispersed into the atmosphere Steam Cycle Energy production is always about transforming one type of energy into another, for example gasoline is first burnt transforming chemical energy into mechanical energy to move the cars. First law of thermodynamics implies the conservation of energy heat in a thermodynamic process. When one energy form is converted into another, the total amount of energy remains constant.[23] Power plant steam cycles try to take advantage of this principle to produce energy as much efficiently as possible, using thermodynamics. A thermodynamic cycle is a series of thermodynamic processes at the end of which the system returns to its initial state.[24] A Carnot Cycle Carnot cycle was the first thermodynamics cycle, represented in figure 2.6, four reversible steps which two are adiabatic (2 to 3 and 4 to 1) and the other two isothermal (1 to 2 and 3 to 4). This is an ideal cycle being impossible to apply in power plant design, efficiency of Carnot cycle is the maximum thermal efficiency that a power plant cycle could ever achieve. It s impossible to achieve because of technical difficulties in doing adiabatic heat exchanges, Figure 2.6: Carnot thermodynamic cycle.[25] letting turbines receive saturate steam because originate an outlet with low quality in vapour which cause turbines to have corrosion problems, having liquid with some vapour also cause cavitation problems in the pumps.[26] 14

39 2.2.2.B Rankine Cycle Rankine cycle, or vapour power cycle because working fluid changes phase from liquid to vapour within the system, gives a better description of a power plant. A schematic of the Rankine cycle in the figure 2.7 helps to understand the system thermodynamics. Figure 2.7: Rankine thermodynamic cycle.[25] Process 1 to 2: Pump pressurizes working fluid, having an increase of temperature in this isentropic step, pump requires external power input. Process 2 to 3: Boiler vaporizes high pressure water to the turbine, using coal or natural gas as heat source. Process 3 to 4: Ideally, isentropic turbine produces power from the expansion of the working fluid, reducing pressure and temperature. Process 4 to 1: Outlet stream from the turbines enters a condenser, where the vapor is cooled to saturation liquid. In a real Rankine cycle compression and expansion are not isentropic which will reduce the efficiency of the system. One of the major issues is that with a high boiler pressure or a low condenser pressure formation of liquid droplets appears in the low pressure side of the turbine, causing corrosion C Reheat-Regenerative Cycle The thermodynamic cycle used nowadays in power generation stations is the Reheat-Regenerative Cycle, as the name implies it s a combination of two thermodynamic cycles. Reheat cycles use two turbines instead of one, and the steam from the first turbine is reheated in the reheater of the boiler before going to the next one. This system has the following advantages: Increases the dryness of the steam, reducing the corrosion problem in the turbines. Increases efficiency. Boiler size decreases because of an increase of work done per kg of steam. The disadvantages of reheat systems are the higher maintenance and plant equipment cost because of the reheater and its long connections, and condenser capacity is increased because of the 15

40 dryness increase in the turbines. In regenerative cycles steam is extracted from the turbine at certain points during its expansion and the steam is used for heating the feed water, more extractions usually mean an increase in efficiency. This process reduces the energy requirement to heat the high pressure water in the boiler and the condenser duty since less steam is needed to condensate, often this system have the problems already describe in the turbines thus usually power plant design combine reheat cycles with regenerative cycles, obtaining higher efficiency than in any of the others cycles. Figure 2.8: Reheat regenerative thermodynamic cycle.[25] In order to understand the entire concept of a pulverized coal power plant is important to comprehend the termodynamic behind it. A typical flowsheet is much more complex than the scheme presented in figure 2.8 to illustrate reheat Regenerative cycle. In figure 2.9, a typical flowsheet is presented. Complexity is exponentially increased with the number of integration draw offs being taken from the turbines. Figure 2.9: Representation of a power plant steam cycle.[27] The main equipments for the steam cycle are: condenser, deaerator, feedwater heater, turbines, governor valve, boiler and pumps. Deaerator is an open feedwater heater where everything is mixed to bled dissolved gases, like 16

41 oxygen and carbon dioxide. A plant could be design only with open feedwater heaters but then a pump would be required after each deaerator. With this configuration energy costs are reduced, having just two pumps that overcome the pressure drops in the system. A plant configuration with only feedwater heaters could also be designed having deaereation in the condenser. However, in past power plants where this was tried the deaeration wasn t as complete and control of boiler feed pump was more difficult. Most of modern power plants have one deaerator and the rest with feedwater heaters. For nomenclature purpose, the condensate steam from the condenser that goes through the closed low pressure feedwater heaters to the deaerator is called condensate. On the other hand, the outlet stream from the deaerator is called feedwater Integration with CCS Power plant system can be integrated with the capture plant and with the CO 2 compression section. For the capture plant usually some low pressure steam draw offs are taken from the power plant to be used in the reboiler of the amine scrubbing, the thermal energy is needed for amine regeneration, typically the reboiler temperature should not exceed 120 C for MEA solvents. Optimizing this integration point is crucial to avoid high energy penalties in the power plant.[2] The compression section requires energy for compression and cooling water for intercooling stages, therefore the feed water can be used in this stages to minimize cooling water requirements. 2.3 Equipment Boiler A steam generator is a closed vessel which is used to produce steam by application of heat from the combustion of fuel to the water. A steam generator can be also called boiler. There are main types of boiler, but the ones used in power plant exceed by far size and capacity of others. It s a unique piece of equipment design specifically to operate at high pressures, with large amounts of water and with a type of fuel. [23] Boilers can be classified as water tube or fire tube boilers, but all modern boilers are water tube. In water tube boilers, water and steam circulates through the tubes and hot flue gas flow over these tubes. This arrangement offer greater versatility, with more boiler capacity and pressure, and most efficient use of the furnace, super-heater and reheater. [28] Two different types of boiling system can be found in boilers, those that have a steam drum or those that don t have (termed once-through steam generators, OSTG). The most used is the steam drum type where the drum serves as a separation point from water and steam. This is the simplest model to control. The water heats up in the tubes before getting to the drum where steam is generated. Then water is separated from the steam and the remaining water is then returned to the tube to be heated. For the OSTG system the water is evaporated somewhere in the tube section. [28] 17

42 The boilers capacity and pressure are very wide, varying from 0.13 kg/s to 1260 kg/s and from 1 atm to pressures above the water critical pressure (217.7 atm). Figure 2.10: Boiler schematic representation.[29] A boiler configuration can be seen in the figure 2.10, representing a typical boiler used in pulverized coal power plant. The boiler has two different sides, the steam generating and the heat recovery system, the main components are: Furnace and convection pass Steam superheaters (primary and secondary) Steam reheater Economizer Steam drum Attemperator and steam temperature control system Air heater In steam generating side, the coal is first pulverized, however this step does not happen in the boiler. The idea behind pulverization is that if the coal were made fine enough, it would burn as easily and efficiently as a gas. To burn successfully coal, a large quantity of fine particles should be encountered along with some minimum coarser particles. The coal will ground and dried the primary air and then pneumatically taken to burners inside the boiler. This air is heated in an air preheater and then separated in primary and secondary air supply. The secondary air, about 70 to 80% of the total hot air, is directly taken to individual burners. The primary air with the remaining 20/30% will go the coal pulverisers. The coal and the air is rapidly mixed and quickly burned in the furnace, releasing heat and producing the flue gas. The air quantity is very important to insure total combustion, allowing all the carbon monoxide to convert to carbon dioxide. Also the amount of excess air is very important because if exceeds the theoretic excess air point the efficiency starts to drop and nitrogen oxides start to form as more 18

43 oxygen and nitrogen are available. The hot outlet flue gas will then rise and exchange heat from convection to the superheater steam, reheated steam, pass through the economizer to heat the feedwater and pass in the air heater at the end. The superheater, reheater and economizer are usually located in the flue gas horizontal and vertical down flow sections of the boiler. The material used in the furnace and convection pass walls is carbon steel or low alloys, which maintain metal temperature within limits. In the modern boiler, the superheaters and reheaters are needed to increase plant thermal efficiency by applying the reheat principle. Therefore in the heat recovery system the feedwater will pass first in economizer before entering the steam drum where it vaporizes to pressures that can be above supercritical. The function of the steam drum is not only to vaporize the water, but also to provide a storage reservoir that allows short terms imbalances between feedwater supply and steam production. [28] After the steam drum, the vapour will pass in two superheaters to insure supercritical feed to the high pressure turbine. The high pressure exhaust steam will also return to be reheated in a reheater before going to the intermediate pressure or low pressure turbine. The main difference between superheaters and reheaters is steam pressure, because superheaters pressure is much higher. Physically they are a single phase heat exchangers with steam flowing inside the tubes and the flue gas in the outlet of the tubes usually in counter current. The steam temperature control is very important to keep temperature in their set point. Complex combination can be used to do this temperature control, but the simpler method is attemperation, the addition of water or low pressure steam to high temperature steam to lower the temperature. This attemperators are usually located between superheaters or in the outlet of the second superheater to better temperature control. The importance of temperature control is to prevent thermal expansion from dangerously reducing turbine clearances and to avoid erosion from excessive moisture in the last stages of the turbine. In the section 5.3 the control systems will more detailed Turbine A steam turbine is a mechanical device that converts thermal energy in pressurised steam into useful mechanical work. The steam turbine derives much of its better thermodynamic efficiency because of the use of multiple stages in the expansion of the steam. This results in a closer approach to the ideal reversible process.[23] The steam from the boiler is expanded in a passage or nozzle where due to fall in pressure of steam, thermal energy of steam is converted into kinetic energy of steam, resulting in the emission of a high velocity jet of steam which impinges with the moving blades of the turbine. Steam turbines can be classified by type of operation, direction of flow, means of heat supply, means of heat rejection, number of cylinders, arrangement of cylinder based on general flow of steam, number of shaft and rotational speed. However, only some classifications will be further developed. [30] 19

44 Steam turbines have two principles of operation, impulse or impulse-reaction turbine. The impulse blading principle is that the steam is directed at the blades and the impact of the steam on the blades drives them round, the pressure at the outlet sides of the blades is equal to that at inlet side. In the impulse reaction turbine, the pressure drop is divided between the fixed and moving blades. The reaction blading principle depends on the blade diverting the steam flow and gaining kinetic energy by the reaction. Since this turbine uses both impulse and reaction principle is called impulsereaction turbine.[23] The direction of the flow in a turbine can also be classified as: axial flow turbine, radial flow turbine and tangential flow turbine. A consideration on the heat supply can be done as well, turbines can be: single pressure turbine, mixed or dual pressure turbine or reheated turbine. The last classification to be mentioned is the heat rejection type: pass-out or extraction turbine, regenerative turbine, condensing turbine, non-condensing turbine and back pressure or topping turbine. Turbines are used for a large range of power requirements. For a power calculation is a good approximation to admit adiabatic system: W s = F H (2.2) Isentropic efficiency can give a good idea of the turbine performance, assuming once again that the heat lost to the surroundings is near to be zero Feedwater heater/deaerator η s = h in h out h in h s,out (2.3) Feedwater heater (FWH) and deaerator are heat transfer equipments very important in the steam cycle efficiency, allowing an increase in temperature of the feedwater thus increasing the efficiency cycle and reducing the amount of heat duty lost in the condenser since less steam is needed. The number of feedwater heaters used in a steam cycle depends mainly on the turbines sizes, inlet and outlet steam conditions, the overall plant cycle and most important economic considerations. An economic analysis is the best way to determine the optimum number of FWHs, each new equipment introduces more capital costs with equipment, piping, valves, controls and instrumentation and space requirements. Maintenance costs will also increase with more pumping costs to overcome pressures drops and equipment maintenance. On the positive side, with more feedwater heaters higher is the efficiency of the power plant, being needed less fuel consumption which will economically improve the power plant performance. Feedwater heaters can be classified as open or closed heat exchangers depending if mixing occurs or not. 20

45 2.3.3.A Deaerator Open feedwater heater or deaerator mix the extract steam with the condensate to be heated. The objective of this equipment is to deaerate the condensate, releasing dissolved gases like oxygen, nitrogen, ammonia and carbon dioxide that appear as a result of chemical reactions and leaks.[24] The presence of a deaerator in the power plant is crucial since the presence of non-condensable gases is the major cause of corrosion in feedwater piping, boiler and condensate handling equipment. Corrosion is mainly cause by three factors: feedwater temperature, ph and oxygen content. Dissolved oxygen in the water will cause rapid localized corrosion in boiler tubes. Carbon dioxide will lower water ph levels and produce carbonic acid that will corrode the entire boiler system and piping. High temperatures are a corrosion catalyst since at high temperature conditions corrosion is more severe, increasing the effect of oxygen and ph corrosion.[31] Remove this non condensable gases could be done by chemical additions however mechanical removal is thermal and economically more efficient. Chemicals are only used when additional oxygen or other non-condensable gases removal is required to control corrosion in the power plant.[32] Mechanical deaeration is based on two principles: Henry s Law and gas solubility variation with temperature Henry s Law asserts that gas solubility in a solution decreases as the gas partial pressure above the solution decreases. Gas solubility in a solution decreases as the temperature of the solution increases and approaches saturation temperature. Both of these natural processes are exploited to achieve low gas concentration in the feedwater, therefore in design of a deaerator four conceptual ideas should be fulfilled: Feedwater temperature should achieve the saturation temperature to decrease gas solubility near to zero. Agitation should be promoted: firstly by spraying feedwater in thin films increases the surface area of the liquid in contact with the steam; secondly the water is cascaded over a bank of slotted trays, further reducing the surface tension of the water. Steam will be sprayed in the feedwater to sweep out any gases still dissolved. Steam layer should be above the water level to accomplish Henry s Law and decrease partial pressure of non condensable gases. Two types of steam deaerators can be found in a power plant: Spray type and Tray type. Deaerator equipment allows a smooth transition between high-pressure and low-pressure closed heater design and also provides proper suction conditions for the boiler feed pump in addition to crucial task of bled dissolve gases. Disadvantages of the deaerator are being large and heavy and require the boiler feed pump to move the liquid forward B Feedwater Heater Closed feedwater heater is usually referred as feedwater heater (FWH). Most of this heat exchanger equipments configuration is a standard shell and tube but can be also header type. The inlet 21

46 steam enters to the shell side and the feedwater passes in the tube side.[24] A feedwater heater is designed to preheat boiler feedwater by means of condensing steam extracted (or bled ) from a steam turbine. No mixing occurs between feedwater and the steam condensate, passing through different sides, being the heat transfer by convection and condensation.[33] The nomenclature of a feedwater heater is steam inlet to the extracted steam from the turbine, drain outlet to the condensate steam, feedwater inlet and feedwater outlet to the water that goes in the tubes. A drain inlet can also be an inlet in the shell side, being this usually an outlet drain from the feedwater heater downstream. Three different zones can be seen in a feedwater heater: Desuperheating zone Condensation zone Subcooling zone Condensation zone is present in every FWH, this is where the steam is condensed. The other two zones are optional depending on the function required. And can have multiple entries with a drain inlet and steam inlet, exemplified in figure Figure 2.11: Schematic representation of a feedwater heater.[24] Desuperheating zone is design to thermally assure dry wall conditions with a minimum pressure drop loss, this is the final zone of the feed water in the tubes. Dry wall conditions prevent flashing and provide maximum heat recovery. [34] Sub cooling zone is totally isolated and could be either internal to the FWH or external. Usually is an internal zone, isolated by a plate, and the drain outlet is cooled to a temperature lower than the saturation temperature. Feedwater heaters can be classified by operational pressure as: low pressures heaters, intermediate pressure and high pressure heaters. Low pressure usually extracts steam from the low pressure turbine and is located between the condensate pump and either the boiler feed pump or, if present, an intermediate pressure (booster) pump. In intermediate pressure heaters the steam is extracted from the intermediate pressure and are located between the booster pump and the boiler feed pump. 22

47 High pressure usually have a tube side pressure of at least 100 bar, extracts steam from the high pressure turbine and is located downstream from the boiler feed pump. Different configurations can be also found in industry like, horizontal, vertical: channel up and vertical: channel down. The most used is the horizontal version that as a better level control although it occupies more space. There are some parameters usually used to individual heater performance like: Feedwater Temperature Rise (FWTR) Terminal Temperature Difference (TTD) Drain Cooler Approach (DCA) Feedwater temperature rise is the difference between the feedwater outlet temperature and the feedwater inlet temperature. Terminal temperature difference is the saturation temperature of the extraction steam minus the feedwater outlet temperature. It s the most commonly used parameter used to control a heater performance, a decrease in this difference indicates an improvement in the heat transfer. Drain cooler approach the temperature difference between the drain cooler outlet and the feedwater inlet. An increasing DCA temperature difference indicates the level is decreasing.[35] Condenser Condenser is one of the essential equipments in modern power plants, increasing considerably the efficiency of the system. It s a heat transfer equipment that condenses all the steam from the turbine. For condensation to occur, the heat vaporization must be removed from the steam and exchanged with a cooling fluid. The cooling fluid can be water or more rarely air. Since the pressure of the last turbine is below atmospheric pressure, so is the pressure in the condenser, the optimum condenser pressure should be determined along with steam pressure outlet pressure. Physical limits impose limitations to the lower pressure possible, like circulating water flow parameters, inlet temperature and cooling water temperature rise. Typical operating pressure is around 0.05 bar.[24] There are two types of water cooled condensers: Surface and jet condensers. Surface condensers don t have direct contact between cooling water and condensate, usually are tube and shell heat exchangers, with water in tubes and condensing steam in the shell. This type of condenser arrangement has several advantages: the condensate can be used as boiler feed water, cooling water can be of poor quality since it doesn t contact with the steam and high vacuum can be obtained (around 0.03 bar) allowing an increase of efficiency. Since condensers operational pressure is below atmospheric air leaks into the system which needs to be continuously removed from the system to maintain low pressure. In small condensers the use of jet air ejectors is enough. For large condensers mechanical evacuators are used, which basically compress the air from low pressure to atmospheric pressure, this kind of equipment is specially design is each case.[23] 23

48 Some disadvantages of this kind of condenser arrangement are: capital cost is higher, maintenance and running cost are high and requires more space in the plant. A power plant requirement for a proper operating is to don t allow any sub cooling degree in the condenser, which reduces the efficiency of the system. To prevent this sub cooling, the outlet temperature of cooling water should be regulated to the saturation temperature of the steam. On the other hand, jet condensers allow contact between cooling water and the steam. The water needs to be of higher quality, and it s mixed with sprays. Jet condensers are not the most used in power plant designs. Air cooled condensers are not so used in power plant industry as well because of his disadvantages: higher condenser operating pressure which cause less cycle efficiency, higher capital and operational costs, larger space requirements and noise level increase. However, it has some advantages as well, minimizes water make-up requirements and eliminates cooling water blow down disposal problems, cooling tower freeze up, tower vapour plume and circulating water pollution restrictions. These advantages clearly points out that this equipment is best suitable when water is scarce and water legislation restrictions are tight. Two types of air cooled condensers can be considered: Jet condenser with dry cooling tower Direct air cooled condenser For the first type, part of the steam condensate is cooled in a dry cooling tower, being then returned to the condenser where it is sprayed into the steam flow, causing the steam to condensate. This condenser needs circulating water pumps and piping to work. In the direct air cooled condenser the steam is piped from the turbine to the steam coils where it condensates and is collected in a tank. This type of condenser needs large steam duct and produces a better vacuum then the jet condenser Flue Gas Desulphurization (FGD) unit Flue gas desulphurization unit is very important in modern power plants since sulphur emission legislation is very tight. Annually worldwide around 160 million tons are emitted to the atmosphere, nearly half of which are from industrial sources. Sulphur dioxide in combination with liquid water easily forms sulphuric acid the main constituent of acid rain.[28] Sulphur dioxide control in power plants is done with one of two strategies: use of low sulphur coal or implementation of scrubbers. Basically a chemical reaction occurs in the scrubbers, SO 2 from the flue gas stream reacts with a reagent. The efficiency of this kind of units can be high, around 95%. Commercially available desulphurization units can be wet, semi-dry and completely dry processes. Wet flue gas desulphurization unit is the most used technology worldwide with more than 85% of installed capacity, however dry processes are also used for low sulphur applications. Wet scrubbers most frequently selected for sulphur dioxide removal used limestone or lime as reagent, but can also use magnesium enhanced lime, ammonia and sodium carbonate. Limestone 24

49 Dissolving gaseous SO 2 SO 2 (g) SO 2 (aq) (2.4) and lime systems are non-regenerative therefore all the reagent is consumed in the chemical reaction A Limestone forced oxidation (LSFO) Limestone forced oxidation is the wet process most used in the industry, produces gypsum as by product that can be sold to cement industry, manufacture of wallboard, used as fertilizer or sent to a landfill. A scheme of the process and the absorber can be seen in the figure Figure 2.12: Schematic representation of a LSFO system.[28] Flue gas enters in a mid section in the absorber and goes upwards while limestone sprayed downwards in counter current. The bottom of the absorber is an integral reaction or recirculation tank, fresh reagent is added to the tank to replenish the alkalinity required to remove SO 2, the addition of air helps to oxide the gypsum. The products are pumped out to the slurry dewatering. The chemical reactions that describes the chemistry in the absorber are: Hydrolysis of SO 2 Dissolution of limestone Acid-base neutralization CO 2 stripping SO 2 (aq) + H 2 O HSO 3 + H+ (2.5) CaCO 3 (s) + H + Ca ++ + HCO 3 (2.6) HCO 3 + H+ CO 2 (aq) + H 2 O (2.7) CO 2 (aq) CO 2 (g) (2.8) In the reaction tank: Dissolution of limestone CaCO 3 (s) + H + Ca ++ + HCO 3 (2.9) 25

50 Acid-base neutralization CO 2 stripping Sulfite oxidation HCO 3 + H+ CO 2 (aq) + H 2 O (2.10) CO 2 (aq) CO 2 (g) (2.11) O 2 (g) + 2HSO 3 2SO 4 + 2H+ (2.12) Precipitation of gypsum Ca ++ + SO 4 + 2H 2 O CaSO 4.2H 2 O (2.13) Parameters that control SO 2 removal capability can be chemical or physical. Chemical parameters are: inlet SO 2 concentration, stoichiometry, ph and chlorine concentration. The physical are: liquid to gas ratio, tray pressure drop and nozzle pressure. The alkalinity is the key to every effect on the sulphur removal Electrostatic Precipitator Particulate emissions need to be control in a power plant, and several technologies are available commercially like: electrostatic precipitators, fabric filters (baghouses), mechanical collectors and venture scrubbers. A fabric filter is comprised of multiple compartment enclosure with each compartment containing rows of fabric bags in the form of round, flat, or shaped tubes, or pleated cartridges. Fabric filters collect particles with sizes ranging from submicron to several hundred microns in diameter at efficiencies generally in excess of 99 or 99.9 percent.[36] However ESP has been the most commonly used in particulate matter control. Efficiencies of 99.9% can be obtained in medium and high ash coals. Basically a dry electrostatic precipitator electrically charges the ash particles in the flue gas to collect and remove them, is comprised of a series of parallel and vertical metallic plates. Each plate contains electrodes which are positively charged. When the particulate gas enters the electrostatic precipitator and is struck with a negative charge electrode, the positively charged plate act as a magnet and pulls the particulate gas to them.[28] Gas/Gas Heater This is heat exchanger equipment, it is important to heat up the outlet flue gas from the FGD to avoid plume formation, since flue gas is saturated with water vapour, condensation is inevitable. This condensation can be extremely acidic leading to the formation and accumulation of acidic deposits. Usually, the GGH cools down the temperature of the flue gas to the FGD in order to increase the temperature from the FGD. 26

51 2.3.8 Governor valve The pressure in the steam cycle is a very important variable in to control. The governor valve purpose in a power plant is to regulate the high pressure into the first HP turbine. When a power demand changes the governor valve will change this pressure drop to meet the final required power Stack Stack is a vertical structure used to convey gaseous products of combustion. The objective is to disperse the pollutants at higher altitude to ease down its influence on the surroundings, with tall structures the chemicals in the flue gas are neutralize in the air before they reach the ground Generator The generator converts the mechanical shaft energy it receives from the turbine into electrical energy. Has a stationary stator and spinning rotor with cooper as conductor. The rotor can spin with of values from 3600 rpm. The production of energy only happens when the power production is synchronized with grid power. The rotor is protected in chamber cooled with hydrogen because of its high heat transfer coefficient and low viscosity to prevent windage losses. 2.4 Control Control is a constant and important presence in industrial world. Instrumentation and control are essential to maintain a normal equipment operation, promoting safety and economic profit. In control terminology, an output variable or measure variable value is compared to the setup point. Depending on the error between those two variables the controller will adjust the manipulated value to some obtain the desired set point value. The proportional control is the simplest type of control where the manipulated variable is proportional to the error signal. The control is always directly or inversely proportional, depending upon the control configuration. This way, if a positive variation in the manipulated variable produces an answer with positive value it is a directly proportional. On the other hand, if a positive variation in a manipulated variable produces a negative variation in the set point it s an inversely proportional system.[28] In this kind of control an increase of the proportional term (or gain) will reduce the final offset but increase the time required to get to stationary state. An introduction of an additional control parameter can eliminate the implicit offset of the proportional control. The addition of Integral term reduces the offset completely but can make the system be less stable as well as take longer to get to stationary state without any offset. The integral term does a repetitive integration of the error signal along the time the deviation occurs. A further and last improvement can be made to the response and stability problems of the system with the addition of derivative control. Derivative control is used to reduce the magnitude of the 27

52 overshoot produced by the integral component and improve the combined controller-process stability. However, it has some problems with the noise in the error signal since the derivative term determines the derivative of the error signal Boiler/turbine control The control of the boiler-turbine system is essential for a proper response to power demands. Three main control options can be presented to a boiler/turbine system: Boiler following control Turbine following control Coordinated boiler turbine control Boiler following control system implies that the boiler response follows the turbine response. In this control mode a change in power demand will implicate a response to the throttle pressure, and a change in the boiler pressure produces a change of the firing rate. The turbine following control is the opposite since the turbine response follows the boiler response. A change in power demand set point will produce a change of the firing rate, therefore the boiler pressure will change and to main the throttle pressure constant, the turbine control valves change position. Coordinated boiler turbine control combines both previous control modes to exploits their advantages and minimize the disadvantages. In simple terms power demand changes and throttle pressure changes are responsibility of both boiler and turbine systems. Coordinated boiler turbine provides a faster response than turbine following systems, however is not as fast as boiler following systems since power error is limited to maintain a balance between boiler response and stability Feed water control The objective of this control is to make sure that the inlet feedwater is equal to the evaporation rate. As simple as this sounds, some difficulties appear within this task, drum level measurement is hard because of swelling or shrink due to steam evaporating and the interactions that changes may cause in other location in the boiler, therefore measure the level is not sufficient to say that feedwater supply needs to be regulated.[37] Three main control strategies can be implemented in this case: one-, two- or three-element feedwater control systems. The one-element feedwater control is basically done by measuring the level and changing the feedwater supply to maintain the level correct, however as said before phenomenon s related to steam vaporizing make this control system very inefficient. Two-element feedwater control uses another concept, measuring the outlet steam flowrate and matching the inlet feedwater to that value, with level measurement assuring correct drum level. This kind of control system is not very used in power stations. 28

53 Three-element control is a cascade control system that uses three variables: drum level, steam flow, feedwater flow. This type of control is the most used in power stations since it is the most accurate. Several combinations of this control can be made, however a frequently used is the following: the level of the steam drum is controlled and measured in the secondary loop of the cascade, the primary loop will be the difference between the feedwater flow and the steam flow measurement, however when this values differ the control will be faster and will trim the level variations. The manipulated variable is the feedwater valve. Controlling the system this way corrects the error produce by the feedwater valve Steam temperature control The outlet steam temperature from the boiler is very important it can be affected by the firing rate, excess air, feedwater temperature, changes in fuel, ash deposits in the heat transfer surfaces and the specific burner combination in service.[37] The control of the temperature can be done by the following methods: Attemperation - attemperator is an apparatus that is used for reducing the steam temperature by spray high purity water into an interconnecting steam pipe. Gas proportioning dampers - are used in control the steam temperatures by splitting the flue gas flow rate opening the damper, and rearranging the amount of heat being transfer to the superheaters and reheaters. Gas recirculation - recirculation of the flue gas after passing through the superheaters, reheaters, economizer and air heater. Excess air - increasing the excess air will also change the heat absorption pattern within the furnace. Burner selection - steam temperature can be regulated by selective burner operation. Movable burners - can be a solution as well changing the pattern of the combustion zone will affect the steam temperature. Differentially-fired divided furnaces - divided furnace sections thus heating up different sections, e.g. one heating the superheaters and the other generating steam and providing heat to the reheater Separately-fired superheaters - has the name implies the superheaters are independent of the boiler Deaerator Control The deaerator principles where already seen in the section A, thus it is clear that the two key parameters must be controlled for a performance. The key issues are: maintenance of the steam pressure at an optimum value and keep a considerable level of water in the equipment.[38] The pressure is maintained by the amount of steam being taken off the turbine draft. The level is controlled by manipulating the condensate inlet flow a typical control for this level can be the method 29

54 already described for the steam drum control system as three elements. For this an additional outlet flow measurement is required to trim level variations.[39] The temperature can be also a controlled variable using some electric immersion heaters which are capable of maintaining the required minimum temperature under no steam flow conditions Condenser Control In the condenser it s crucial to control the pressure, since as low as this pressure can be the higher is the thermal efficiency of the power. Thus the pressure in this unit is always measured and controlled by the cooling water inlet flowrate.[38] The level in the condenser is also important to assure, if the level drops a make-up of water will be introduce in the system, this make up is important to face off the water and steam losses during the process. If an increase in the level occurs that can be related to swelling effects already described or by a failure in the feedwater pumping system, therefore requiring immediate operator actions.[39] Feedwater heaters Control In a feedwater heater the level control is essential to good maintenance of the heat transfer efficiency. The maximum heat transfer takes place when the largest tube area is exposed to the steam without allowing steam blow-through. Condensate is allowed to drain from the shell through the normal drain. When the tubes become submerged in condensate, heat is transferred to the condensate rather than the tubes with the feedwater inside, resulting in poor heater efficiency. Therefore, the level control is implemented in feedwater heaters, the level is measured and the manipulated will be the outlet drain condensate valve, obviously changing the flowrate of the outlet drain. The steam draft flow is usually imagined as a control system where the temperature is controlled by the amount of steam taken from the turbine however FWH have a self-regulating feature. There are no control valves on the extraction steam supply lines. The steam flow adjusts itself by a thermal equilibrium process. When the feedwater temperature approaches the saturated steam temperature then condensation of the extraction steam diminishes and therefore the flow of extraction steam to the feedwater heater tends towards zero. 30

55 3 Materials and Methods Contents 3.1 gproms software Model development workflow gccs model library gccs structure

56 3.1 gproms software The software used in this work was the gproms from the Process Systems Enterprise the company that developed it and holds its rights. It is a platform for high-fidelity predictive modelling for the process industries. Advanced Process Modelling environment where it is possible to create first principle models or simply use the gproms model libraries. Models are built using fundamental chemistry, physics, chemical engineering, operating procedures and other relationships to fully capture with high fidelity the behaviour of the system being modeled. Different activities can be simulated such as: steady state and dynamic simulation, parameter estimation, model-based experiment design and optimisation. gproms is an equation based modelling system resulting on a numerical solution of all the equations in a model or a flowsheet at the same time. This type of numerical resolution has several advantages increasing the robustness and fastness in comparison with traditional sequential-modular simulation approaches. gproms offers a number of products and complete libraries for diverse applications, products examples are: gsolids, gcrystals, gsaft, gflare, gfuelcell and in the interest of this work the future product gccs. 3.2 Model development workflow gccs is a new product that is now being developed by PSE and as previously stated this work is integrated in it, being necessary to develop the models and basis from scratch, making them obviously physical and chemically correct as well as, in a robust and simple modelling writing to be integrated in large flowsheets. The evolution of the work is described in the figure 3.1 where a model development flowchart is presented. The first step is to study the technical specification required for each model, doing a detailed bibliographic review getting in touch with all the concepts, objectives, specific characteristics and physical constraints of the model. Analyze the equations of the system and its variables to know the degrees of freedom (DOF). The next step is to make the model specification document (MSD) where all the assumptions, equations, variables and degrees of freedom of the model are clearly explained. After the MSD is reviewed the implementation of the model in gproms starts, in this step all that was stated in the MSD is transcript to gproms language, created the user dialog box according to the expected assignments and the model report. Tests within physical constraints of the system are applied to the model and the results checked, usually model refinement is required changing equations, improving thermodynamics calculations or developing initialization procedures to increase robustness of the model. The final step of a model development is to check the results of predictive simulations against data processing from the partners and confirm the accuracy of the model. If the model does not pass the 32

57 predictive simulations refining is necessary to make it more accurate. After being review by the chief technologist the final model document can be produced. Figure 3.1: Model development workflow.[40] Once the models are tested and verified connecting different models to make a flowsheet is the next step and will complement the analysis made for the DOF of model in the early stages of the workflow. The degree of freedom of the flowsheet is now required and can eventually affect the dialog box options of simple models. The construction of a flowsheet is done connecting model by model, to avoid recirculation of information which causes some troubles in gproms. The solution found to solve this issue is explained in the chapter 5.1. The development of a power plant flowsheet requires several models that were divided by the members of the gccs power plant team. Connecting them together and initialise a complex flowsheet is a hard task, requiring the deep knowledge of all models to be able to calculate their DOF and the global value for the entire flowsheet. After being tested in the flowsheets some models can return to the early stages of development requiring a new arrangement of assignments. 3.3 gccs model library In the early stages of the work a gccs library was already been developed by the several members of the power plant team. The models can be categorized by their importance in representing the power plant configuration, those models are described as main equipments of the power plant, however to build a consistent flowsheet some other auxiliary simple models were introduced and for last the models introduced only for control purpose can be listed as well. 33

58 The main equipments available in the libray were: the boiler, the turbine, the deaerator and the governor valve. The auxiliary models were: the drum, the pump, the blower, the generator, the recycle breaker, the source coal, source air, source utility, sink utility, sink waste, the stack and the junction. The following missing models in the library were mathematically modeled: feedwater heater, condenser, electrostatic precipitator, flue gas desulphurization unit, gas gas heater and the control valve. Appendix A is the review and summary of all the characteristics and specification options available for all the models used. Typically all the models have a design and operational mode, design mode is used when the stream conditions are fixed and the parameters of the equipments such as areas and efficiencies are calculated. In operational mode those parameters are specified and the stream conditions are determined. 3.4 gccs structure gproms is a very powerful tool that solves difficult mathematical problems. It allows the user to write first principle models that can later be used as drag and drop to build large and complex flowsheets. The models communicate with the exterior through ports and connections, which depending on the type can pass different types of information. In the gccs there are several types of connections: UtilityFluid, ProcessFluid, Coal, Power and Control ports. After connecting all the models, the assignments must be done in the process. If the model have the specification dialog box the assignments are automatically updated in the process otherwise must be done manually. The advantages of the process is that with only one model several processes can be used simultaneously. The physical property package Multiflash was used to obtain the thermodynamics properties of the compounds. An example of a property call is : T hot,sat out = PhysProp.DewTemperature(p out, 1) (3.1) This equation written gives the saturation temperature of the water at the outlet pressure. the multiflash package is possible to choose the thermodynamics more adequate to the different connection types. For the UtilityFluid connection was used the steam tables 1995 version: IAPWS-95, as the name suggests this connection was used for all the steam cycle. The ProcessFluid connection that is used for the flue gas side used the Peng-Robinson 1978 equation of state. In 34

59 4 Mathematical Modelling of PCPP Components Contents 4.1 Feedwater Heater Boiler Steam Condenser Electrostatic Precipitator Flue Gas Desulphurisation Gas/Gas Heater Control Valve

60 In the power plant several models were required and the responsibility of modelling each piece of equipment was divided by the members of the gccs power plant team. The models developed in this work are described in this chapter: the feedwater heater, boiler steam condenser, electrostatic precipitator, flue gas desulphurization unit, gas gas heater and the control valve. The description of the models only include the assumptions used and the general equations, the manipulation of equations to make the models more robust and the initialisation procedures are not presented since it is classified information. 4.1 Feedwater Heater The FeedWaterHeater (FWH) model is used to pre-heat boiler feed water using condensing steam from the turbines and is usually used to determine that steam flowrate drawn off from the turbine. This is a lumped and steady-state model. There can be two or more inlet streams, these streams are first mixed and the properties of the mixture are estimated by mass and energy balances. The only exception is the inlet pressure which is assumed to be the minimum of all the inlet steam pressures. It is assumed that the inlet and outlet feedwater is in liquid phase and the outlet steam condensate can be liquid or vapour phase. For the design of the equipment logarithmic mean temperature difference is used and a constant overall heat transfer coefficient is assumed to estimate the heat transfer area required. The model includes temperature relationships such as terminal temperature difference and drain effectiveness, these type of temperatures relationships are more or less constant in the feedwater heaters relating the temperatures with the pressure of the system. They are very important in part load operations because the pressures of the system start to decrease but these relations are kept constant Inlets One UtilityFluid inlet port representing the inlet feed water. One UtilityFluid inlet port representing the inlet steam from the turbines. One array of UtilityFluid inlet port representing the inlet(s) drain(s) from another feedwater heater Outlets One UtilityFluid outlet port representing the outlet steam condensate drain from the feedwater heater. One UtilityFluid outlet port representing the heated feed water from the feedwater heater. 36

61 4.1.3 Variables Nomenclature Table 4.1: Nomenclature of the variables used in the feedwater heater model. Symbol Definition Units Array Size F in Mass flowrate of the hot side kg s 1 F FW Mass flowrate of the cold side kg s 1 T in Inlet temperature of the hot side K T out Outlet temperature of the hot side K Tin FW Inlet temperature of the cold side K Tout FW Outlet temperature of the cold side K T hot,sat out Dew temperature of the steam K TTD Terminal temperature difference K DCA Drain cooler approach K FWTR Feed water temperature rise K Deff Drains effectiveness K T in Inlet temperature difference K T out Outlet temperature difference K p in Inlet pressure of the hot side Pa p out Outlet pressure of the hot side Pa p FW in Inlet pressure of the cold side Pa p FW out Outlet pressure of the cold side Pa p shell Pressure drop on the hot side Pa p FW Pressure drop on the cold side Pa h in Inlet specific enthalpy of the hot side J kg 1 h out Outlet specific enthalpy of the hot side J kg 1 h FW in Inlet specific enthalpy of the cold side J kg 1 h FW out Outlet specific enthalpy of the cold side J kg 1 Q Heat duty W A Heat transfer area m 2 lm T Logarithmic mean differential temperature K U Overall heat transfer coefficient W m 2 K Equations The inlet port connections are used to obtain the inlet conditions to the hot stream in the FWH via a material and energy balance. The inlet pressure is assumed to be the lowest pressure among the Inlet ports. Mass balance for the mixing section: Energy balance for the mixing section: Nin aux F [Steam] + i=1 Nin aux F [Steam].h [Steam] + i=1 A p-h flash call is used to obtain the inlet temperature: F [aux] i = Fin (4.1) F [aux] i.h [aux] i = Fin.h in (4.2) T in = PhysProp.PHFlash(p in, h in, 1) (4.3) p in = min i (p [Steam], p [aux] i ) (4.4) 37

62 Energy balance for the heat exchange: Q = F in.(h in h out ) (4.5) Q = F FW.(h FW out h FW in ) (4.6) Design calculation: Q = U.A. lm T (4.7) A counter current flow is asssumed therefore the logarithmic mean temperature difference is defined by: T in = T hot in T cold out (4.8) T out = T hot out T cold in (4.9) Outlet pressure calculation: lm T = T in T out ln ( T in / T out ) (4.10) p out = p in p shell (4.11) The enthalpy estimation for the feedwater stream: p FW out = p FW in p FW (4.12) h FW in = PhysProp.LiquidEnthalpy(Tin FW, p FW in, 1) (4.13) h FW out = PhysProp.LiquidEnthalpy(T FW out, p FW out, 1) (4.14) Oulet drain enthalpy : h out = PhysProp.Enthalpy(T out, p out, 1) (4.15) Outlet saturation temperature: T hot,sat out = PhysProp.DewTemperature(p out, 1) (4.16) Temperature differences: TTD = T hot,sat out T FW out (4.17) DCA = T out T FW in (4.18) FWTR = T FW out T FW in (4.19) Drain effectiveness: Deff = T hot,sat out T hot,sat out T out T FW in (4.20) 38

63 4.1.5 Degree of freedom Analysing the number of variables and the number of equations the degrees of freedom of the model are: DOF = N var N eq = ( [Steam] + 3.N [aux] ) 20 = [Steam] + 3.N [aux] (4.21) The feedwater heater specification will include the following mandatory specifications, note that in some options no inlet steam flowrate is provided in the model but an additional specification replaces that one and the steam flowrate drawn off is determined: 1. Inlet streams: Feedwater inlet - Temperature, pressure and flowrate. Drain inlet(s) - Temperature, pressure and flowrate. Steam inlet - Temperature and pressure. 2. Overall heat transfer coefficient 3. Pressure drop on the cold side 4. Pressure drop on the hot side Additional Specifications: 1. With steam flow calculated Design Any one of: Drain outlet temperature, Drain cooler approach or Drain effectiveness Any one of: Outlet feedwater temperature, Terminal temperature difference or Feedwater temperature rise. Operational Heat transfer area Any one of: Terminal temperature difference or Feedwater temperature rise, Drain cooler approach or Drain effectiveness 2. With steam flow known Inlet stream: Steam flowrate. Design Any one of the outlet temperatures or temperatures approaches. Operational Heat transfer area gproms interface This section is only presented for the feedwater heater to exemplify and show how the gproms inputs should be done and how the results are presented. One of the feedwater heaters specification can be seen in the figure 4.1: 39

64 Figure 4.1: gccs feedwater heater example of a specification. Using this specification in a flowsheet and specifying the inlets. The outlets of the feedwater heater, the steam flowrate, the heat transfer area and the logarithmic mean temperature difference are calculated. An example of the gproms software environment is presented in the figure 4.2. Figure 4.2: gccs environment for a fwh simulation. 40

65 4.2 Boiler Steam Condenser The BoilerSteamCondenser model is used for condensation of low-pressure steam, using cooling water. This is a lumped and steady state model that is usually used to calculate the value of the cooling water flowrate. There can be two or more inlet streams, these streams are first mixed and the properties of the mixture are estimated by mass and energy balances. The only exception is the inlet pressure which is assumed to be the minimum of all the inlet steam pressures. Figure 4.3: gccs boiler steam condenser model configuration. Is assumed that the inlet and outlet cooling water is in liquid phase and the outlet steam condensate is saturated. For the design of the equipment logarithmic mean temperature difference is used and a constant overall heat transfer coefficient is assumed, to estimate the heat transfer area required for the system. This model have a control port that passes the pressure in the condenser to the controller, the model is presented in the figure Inlets One array of UtilityFluid inlet ports representing the inlet steam to condenser. One UtilityFluid inlet port representing the inlet cooling water to condenser Outlets One UtilityFluid outlet port representing the outlet steam condensate from the condenser. One UtilityFluid outlet port representing the cooling water return from the condenser. One ControlSignal port for pressure measurement Variables Nomenclature Table 4.2: Nomenclature of the variables used in the condenser model. Symbol Definition Units Array Size F hot Mass flowrate of the hot side kg s 1 F cold Mass flowrate of the cold side kg s 1 Tin hot Inlet temperature of the hot side K Tout hot Outlet temperature of the hot side K Tin cold Inlet temperature of the cold side K Tout cold Outlet temperature of the cold side K T hot,sat out Dew point temperature of the steam K T mindiff Minimum temperature difference K T in Inlet temperature difference K Continued on next page 41

66 Table 4.2 Continued Symbol Definition Units Array Size T out Outlet temperature difference K p hot in Inlet pressure of the hot side Pa p spec Pressure specified by the user Pa p condenser Pressure in the condenser Pa p cold in Inlet pressure of the cold side Pa p cold out Outlet pressure of the cold side Pa p cold Pressure drop on the cold side Pa h hot in Inlet specific enthalpy of the hot side J kg 1 h hot out Outlet specific enthalpy of the hot side J kg 1 h cold in Inlet specific enthalpy of the cold side J kg 1 h cold out Outlet specific enthalpy of the cold side J kg 1 Q Cooling load for the condenser W A Heat transfer area m 2 lm T Logarithmic mean differential temperature K U Overall heat transfer coefficient W m 2 K Equations The inlet port connections are used to obtain the inlet conditions of the hot stream in the condenser via a material and energy balance. The inlet pressure is assumed to be the lowest pressure among the Inlet ports. Mass balance for the mixing section ([Inlet] i is the ith Inlet port): Energy balance for the mixing section: Nin hot i=1 F [Inlet] i = Fhot (4.22) Nin hot i=1 F [Inlet] i h [Inlet] i A p-h flash call is used to obtain the inlet temperature: = Fhot h hot in (4.23) T hot in = PhysProp.PHFlash(p hot in, h hot in, 1) (4.24) Pressure in the condenser will be: p hot in = min i (p [Inlet] i ) (4.25) It s important to track down if p condenser is higher then p hot in The outlet pressure of the cooling water will be: p condenser = min(p hot in, p spec ) (4.26) which is physically impossible. p cold out = p cold in p cold (4.27) Energy balance for the heat exchange: Q = F hot (h hot in h hot out) (4.28) 42

67 Q = F cold (h cold out h cold in ) (4.29) Design calculation: Q = UA lm T (4.30) Assuming counter current flow, logarithmic mean temperature difference is defined: T in = T hot in T cold out (4.31) T out = T hot out T cold in (4.32) lm T = T in T out ln ( T in / T out ) (4.33) One of the assumptions of the model is that the outlet condensate is saturated therefore the outlet temperature of the condensate fluid will be: T hot,sat out T hot,sat out = T hot out (4.34) = PhysProp.DewTemperature(p hot out, 1) (4.35) Minimum temperature difference of the system is always between the outlet condensate temperature which is the saturation temperature and the outlet temperature of the cooling water: Specific enthalpy estimation for the cooling water: T mindiff = T sat hot T cold out (4.36) h cold in h cold out Specific enthalpy for the working fluid: = PhysProp.LiquidEnthalpy(Tin cold, p cold in, 1) (4.37) = PhysProp.LiquidEnthalpy(T cold out, p cold out, 1) (4.38) h hot out = PhysProp.LiquidEnthalpy(T hot out, p hot out, 1) (4.39) Degree of freedom Analysing the number of variables and the number of equations the degrees of freedom of the model are: DOF = N var N eq = ( Nin hot ) 18 = Nin hot (4.40) The condenser specification will include the following mandatory specifications: 1. Inlet streams: Steam inlet(s) - Temperature, pressure and flowrate. Cooling water inlet - Temperature and pressure. 2. Overall heat transfer coefficient 3. Cooling water side pressure drop Additional Specifications: 43

68 1. With cooling water flow calculated Design Condenser pressure Any one of: Outlet cooling water temperature or Minimum temperature difference Operational Heat transfer area Any one of: Condenser pressure, Outlet cooling water temperature or Minimum temperature difference 2. With cooling water flow known Inlet stream: Cooling water flowrate. Design Condenser pressure Operational Heat transfer area 4.3 Electrostatic Precipitator The Electrostatic Precipitator model removes the required particulate matter from the inlet stream, the outlet ash concentration or the unit efficiency can be specified. The model estimates the amount of power required to perform this operation with empirical equations. The pressure drop across the equipment is not estimated, must be an input from the user. This is a steady-state and lumped model. No Figure 4.4: gccs electrostatic precipitator model configuration. explicit modelling of the particulates themselves or the equipment is done and no temperature change is considered. The outlet flue gas is in gas phase nothing condenses and only ash is removed Inlets One ProcessFluid inlet port representing the inlet flue gas Outlets One ProcessFluid outlet port representing the outlet flue gas Variables Nomenclature 44

69 Table 4.3: Nomenclature of the variables used in the electrostatic precipitator model. Symbol Definition Units Array Size F in Inlet mass flowrate kg s 1 F out Outlet mass flowrate kg s 1 F ashout Outlet ash mass flowrate kg s 1 w in Inlet mass fraction C (Components) w out Outlet mass fraction C w ashout Auxiliary outlet ash mass fraction C x in Inlet mole fraction C x out Outlet mole fraction C x dry in,o 2 Mole fraction of the oxygen in the inlet dry flue gas % x dry out,o 2 Mole fraction of the oxygen in the outlet dry flue gas % γ in,ash Concentration of the ash in the inlet flue gas mg Nm 3 γ out,ash Concentration of the ash in the outlet flue gas mg Nm 3 p in Inlet pressure Pa p out Outlet pressure Pa p Pressure Drop in the ESP Pa h out Outlet specific enthalpy J kg 1 ρ in Density of the inlet stream kg m 3 ρ normal in Normal density of the inlet stream kg m 3 ρ normal out Normal density of the outlet stream kg m 3 κ ash,cap Ash capture rate tph η Efficiency of ash removal % P Electrical consumption of the unit W Equations Efficiency expression: Conversion between molar and mass fractions: η = γ in,ash γ out,ash γ in,ash.100 (4.41) M i w out,i = x out,i C i=1 (M ix out,i ) (4.42) M i w in,i = x in,i C i=1 (M ix in,i ) (4.43) Mass balance: F in w in,i = F out w out,i + F ashout w ashout,i (4.44) Total Composition Restriction: C x out,i = 1 (4.45) i=1 In the outlet ash stream, only ash is present so: w ashout,ash = 1 (4.46) w ashout,i = 0 (4.47) Outlet pressure: p = p in p out (4.48) 45

70 The concentration should be reported in terms of mg/nm 3, corrected to dry gas at 6% O 2 because this is how it is reported to the authorities and checked if the values are within the limits. γ in,ash = f(w in,ash, ρ normal in, x dry in,o 2 ) (4.49) γ out,ash = f(w out,ash, ρ normal out, x dry out,o 2 ) (4.50) (Concentration should be converted from kg.nm 3 to mg.nm 3 ) The inlet and outlet concentration correction for the oxygen in the dry flue gas: The outlet enthalpy is: x dry in,o 2 = x in,o (4.51) 1 x in,h2o x dry out,o 2 = x out,o (4.52) 1 x out,h2o ρ normal in = PhysProp.Density(T standard, p standard, w in ) (4.53) ρ normal out = PhysProp.Density(T standard, p standard, w out ) (4.54) h out = PhysProp.VapourEnthalpy.(T out, p out, w) (4.55) The power consumption of the ESP is quite variable, however there is a clear relationship between capture efficiency and the power, to achieve greater efficiencies more power will be required. The specific power relationship with the efficiency can be found in figure 4.5. A two branch equation was found fitting the relation in figure 4.1, using TableCurve 2D version The equation was rearranged to obtain the following power (W) expression: F in if η > 99.9 ρ in P = (4.56) η η η 2.F in otherwise ρ in Figure 4.5: Specific power relationship with the efficiency of an ESP unit.[41] 46

71 To estimate the power requirement is necessary to know the density of the inlet flue gas: ρ in = PhysProp.Density(T, p in, w in ) (4.57) The ash capture rate of the ESP unit is estimated by the following expression (conversion is required to have the ash capture rate in tonnes per hour): Degree of freedom The degrees of freedom of the model are: κ ash,cap = (γ in,ash γ out,ash ). F in ρ in (4.58) DOF = N var N eq = ( C ) (12 + 4C ) = 5 + C (4.59) The electrostatic precipitator specification includes the following mandatory specifications: 1. Inlet stream - Temperature, pressure, composition and flowrate. 2. Pressure drop across the equipment. Additional Specifications: 1. Outlet ash concentration or the removal efficiency. 4.4 Flue Gas Desulphurisation The flue gas desulphurisation (FGD) unit removes sulphur dioxide of the flue gas to a certain specification level. Either the sulphur content in the outlet flue gas stream is specified or the unit efficiency (of sulphur removal). The model estimates the power and material requirements based on the outlet sulphur dioxide content specification. The outlet temperature and composition is also estimated by the model. The wet flue gas desulphurisation system is modelled, specifically the limestone forced oxidation (LSFO). The following reaction will take place in the FGD unit: Figure 4.6: gccs FGD model configuration. CaCO O 2 + 2H 2 O + SO 2 CaSO 4.2H 2 O + CO 2 (4.60) This is a steady-state and lumped model. Desulphurisation operation is not detailed modelled, ratio and stoichiometric data used for estimation of the material requirements. The outlet water content in the flue gas is determined assuming saturation and in the enthalpy balance the model only takes into account the liquid and vapour components, the solids are ignored. Another assumption is that the outlet liquid content is only water and dissolved carbon dioxide. 47

72 4.4.1 Inlets One ProcessFluid inlet port representing the inlet flue gas Outlets One ProcessFluid outlet port representing the outlet flue gas Variables Nomenclature Table 4.4: Nomenclature of the variables used in the FGD model. Symbol Definition Units Array Size η Efficiency of SO2 removal % F in Inlet mass flowrate kg s 1 F out Outlet mass flowrate kg s 1 F w Inlet water in the system kg s 1 F SO2 removed Removed mass of sulphur dioxide kg s 1 F lime Required limestone mass flowrate kg s 1 F air Required air mass flowrate kg s 1 F gyp By-product gypsum mass flowrate kg s 1 w in Mass fraction of the inlet stream C w out Mass fraction of the outlet stream C x out Molar fraction of the outlet stream C w air Mass fraction of the air stream C w w Mass fraction of the water stream C x dry out,o 2 Mole fraction of the oxygen in the outlet dry flue gas % wout w Mass fraction of the outlet liquid flowrate x w out Mole fraction of the outlet liquid flowrate γ out,so2 Concentration of the SO 2 in the outlet flue gas mg Nm 3 T in Flue gas inlet temperature K T Inlet temperature of the reactants K T out Outlet temperature K p in Inlet pressure Pa p out Operational pressure Pa p Pressure drop Pa h in Inlet specific enthalpy to the FGD J kg 1 h out Outlet specific enthalpy of the FGD J kg 1 h gyp Limestone s water specific enthalpy J kg 1 h lime Gypsum s water specific enthalpy J kg 1 h w Make up water specific enthalpy J kg 1 h air Air specific enthalpy J kg 1 wt lime Limestone s solids content wt gyp Gypsum s solids content w gyp Outlet water content in solid gypsum ρ normal out Normal density of the outlet stream kg m 3 ρ in Density of the inlet stream kg m 3 ρ w out Density of the water kg m 3 r H Θ Standard heat of reaction J mol 1 r CaCO3/SO 2 Molar ratio of limestone/so2 removed H CO2,H 2O Henry coefficient for carbon dioxide in water Pa m 3 mol 1 µ lime Purity of the limestone P Power requirement W 48

73 4.4.4 Equations and Efficiency definition is: η = F inw in,so2 F out w out,so2 F in w in,so2 (4.61) F SO2 removed = F inw in,so2 F out w out,so2 (4.62) Conversion between molar and mass fractions: M i w out,i = x out,i C i=1 (M ix out,i ) (4.63) Total Composition Restriction: C x out,i = 1 (4.64) i=1 The pressure in the system is given by the following equation: p = p in p out (4.65) Mass balance for the nitrogen: w in,n2 F in + w air,n2.f air = w out,n2.f out (4.66) For carbon dioxide: w in,co2.f in + F SO2 removed.m CO 2 M SO2 x out,co2 = w out,co2.f out + (1 w w out).f gyp. 1 wt gyp wt gyp (4.67) = H CO 2,H 2O.(1 x w out) p out (4.68) The carbon dioxide is the most important component in the system this way its losses must be well taken in account, so the dissolution in water will be function of the operational temperature: [42] ( ) ρ w out H CO2,H 2O = exp (4.69) T out M H2O ρ w out = PhysProp.LiquidDensity(T out, p out, w w ) (4.70) For water, estimation of the vapor pressure [43] must be done: p vapour = (T out T ref ) (T out T ref ) (T out T ref ) 3 (4.71) Where T ref = K. x out,h2o = p vapour.x out w (4.72) p out The mass balance for the water is presented in the following equation: w in,h2o.f in + F w + F lime. 1 wt lime wt lime = w out,h2o.f out + w out w.f gyp. ( ) 1 wtgyp + w gyp + 2.F SO2 wt gyp M H2O removed M SO2 (4.73) 49

74 The conversion between mass and molar fraction for the liquid outlet, water and carbon dioxide, is: w out w.(m CO2.(1 x out w ) + M H2O.x out w ) = M H2O.x out w (4.74) Total mass balance: F in + F lime wt lime + F water + F air = F out + F gyp wt gyp (4.75) For all the components except sulphur dioxide, water, oxygen, carbon dioxide and nitrogen the following equation is applied: Estimation of material requirements: F in.w in,i = F out.w out,i (4.76) F lime.µ lime = r CaCO3/SO 2. M CaCO 3.F SO2 removed (4.77) M SO2 F Air = F SO2 removed w air,o2. M O 2 M SO2 (4.78) Typically the gypsum brings a certain percentage of water (w gypsum ) and the rest of limestone that didn t react. Enthalpy balance: ( ) 1 wtlime F in.h in + F lime.h lime + F air h air wt lime The enthalpy of reaction is calculated by: F gyp = f(w gypsum, F SO2 removed, r CaCO 3/SO 2, µ lime ) (4.79) ( = F out.h out + w gyp + 1 wt ) gyp F gyp.h gyp + r H Θ. F SO2 removed (4.80) wt gyp M SO2 r H Θ = Σν B f H Θ B Σν A f H Θ A = Jmol 1 (4.81) γ out,so2 Concentration correction for the oxygen in the dry flue gas : = f(w out,so2, ρ normal out, x dry out,o 2 ) (4.82) x dry out,o 2 = x out,o 2 1 x out,h2o (4.83) Enthalpy calculation: ρ normal out = PhysProp.Density(T standard, p standard, w out ) (4.84) h gyp = PhysProp.LiqudiEnthalpy(T out, p out, w w ) (4.85) h lime = PhysProp.LiquidEnthalpy(T, p out, w w ) (4.86) h water = PhysProp.LiquidEnthalpy(T, p out, w w ) (4.87) The lime and gypsum are only liquid enthalpy beacause of the assumption already stated before that the solids enthalpy are ignored, note that in liquid phase can only be found water and/or carbon dioxide. For h Air, h out and h in vapour enthalpy calls are used to determine the required enthalpy. 50

75 Power expression was provided by one of the partners in this project (E-On), therefore is confidential and is only represented the relationship behind: P = f(f SO2 removed, F in, ρ in ) (4.88) ρ in = PhysProp.Density(T in, p in, w in ) (4.89) Air stream assigned: Water stream assigned: 0.79 for N 2 w air = 0.21 for O 2 0 other components 1 for H 2 O w w = 0 other components (4.90) (4.91) Degree of freedom The degrees of freedom of the model are: DOF = N var N eq = (5C + 35) (4C + 25) = C + 10 (4.92) The FGD specification will include the following obligatory specifications: 1. Inlet stream - Temperature, pressure, composition and flowrate. 2. Any of this: Operational pressure Pressure drop 3. Any of this: Efficiency Outlet SO 2 mass fration Outlet SO 2 mole fration Outlet SO 2 concentration 4. Limestone in slurry Additional Specifications: 1. In advanced mode (in standard mode this values are assigned as default) Molar ratio limestone/sulphur dioxide removed Limestone purity Slurry temperature Solids content in gypsum 51

76 4.5 Gas/Gas Heater The gas/gas heater model is a simple heat exchanger, exchanging heat from two gas streams. The user needs to assign the hot outlet temperature or the heat duty. The pressure drop and the heat transfer efficiency should be specified by the user as well. This a steady state and lumped model Inlets One ProcessFluid inlet port representing the inlet hot stream usually from the ESP unit. One ProcessFluid inlet port representing the inlet cold stream typically from the FGD unit Outlets One ProcessFluid outlet port representing the outlet hot stream that goes to the FGD. One ProcessFluid outlet port representing the outlet cold stream that goes for the stack Variables Nomenclature Table 4.5: Nomenclature of the variables used in the GGH model. Symbol Definition Units Array Size F hot Mass flowrate of the hot side kg s 1 F cold Mass flowrate of the cold side kg s 1 w hot Mass fraction of the hot side C w cold Mass fraction of the cold side C Tin hot Inlet temperature of the hot side K Tout hot Outlet temperature of the hot side K Tin cold Inlet temperature of the cold side K Tout cold Outlet temperature of the cold side K p hot in Inlet pressure of the hot side Pa p hot out Outlet pressure of the hot side Pa p cold in Inlet pressure of the cold side Pa p cold out Outlet pressure of the cold side Pa p Pressure drop on the hot side Pa h hot in Inlet specific enthalpy of the hot side J kg 1 h hot out Outlet specific enthalpy of the hot side J kg 1 h cold in Inlet specific enthalpy of the cold side J kg 1 h cold out Outlet specific enthalpy of the cold side J kg 1 η Heat transfer efficiency % Q Heat duty W Equations Energy balance for the heat exchange: Outlet pressure calculation: η = η = Q F hot.(h hot in.100 (4.93) hhot out) Q F cold.(h cold out h cold (4.94) in ).100 p hot out = p hot in p (4.95) 52

77 Vapour enthalpy calls are used to determine the: h hot in Degree of freedom p cold out = p cold in p (4.96),hcold in,hhot out and h cold Analysing the number of variables and the number of equations the degrees of freedom of the model are: out. DOF = N var N eq = (2C + 17) 8 = 2C + 9 (4.97) The Gas Gas Heater specification will include the following obligatory specifications: 1. Inlet streams: Inlet hot stream - Temperature, pressure, composition and flowrate. Inlet cold stream - Temperature, pressure, composition and flowrate. 2. Heat transfer efficiency 3. Pressure drop across both sides. Additional Specifications: 1. Hot stream outlet temperature or heat duty. 4.6 Control Valve This model describes a valve, determining the flow as a function of the pressure difference and the valve stem position. The model is characterized in terms of the valve flow coefficient (Cv, gpm/psi 0.5 ), the inherent flow characteristic (linear, equal-percentage, quick-opening), the rangeability factor and the leakage fraction. The flow-characteristic options for the valve can be linear, quick-opening or equal-percentage and are illustrated in Fig Figure 4.7: Valve stem position for different flow characteristics.[44] 53

78 The flow coefficient (Cv) describes the flow versus pressure relationship through a valve. definition, Cv is the number of gallons per minute of 60 o F water which will pass through a valve with fixed pressure drop of 1 psi. The rangeability factor of a valve (used only in the equal-percentage calculation) is the ratio of the valve flow coefficient at maximum valve stem position to the valve flow coefficient fraction when the valve stem position is at its minimum. In real engineering sense, this rangeability factor should always be greater than 1. By Figure 4.8: gccs control valve model configuration. The dynamics of the valve are modelled via a time delay equation on the stem-position of the valve. This model can have both liquid or gas inlet streams and no phase change occurs in the system. It is assumed isenthalpic expansion and irreversible flow. Since this is a control valve, the stem position setting is an input from the controller connected in the control port. The representation of the control valve model can be found in the figure Inlets One UtilityFluid inlet port representing the inlet water or vapour stream. One ControlSignal inlet port representing the stem position setting Outlets One UtilityFluid outlet port representing the inlet water or vapour stream Variables Nomenclature Table 4.6: Nomenclature of the variables used in the Control Valve model. Symbol Definition Units Array Size F Mass flowrate kg s 1 F max Maximum mass flowrate across the valve kg s 1 w Mass fraction in the valve - C T in Inlet temperature K T out Outlet temperature K p in Inlet pressure Pa p out Outlet pressure Pa h in Inlet specific enthalpy J kg 1 h out Outlet specific enthalpy J kg 1 p Pressure drop across the valve Pa Cv Valve flow coefficient kg Pa s 1 Cv f Fraction of valve flow coefficient V sp Valve stem position Vsp act Actual valve stem position L f Leakage fraction τ Time constant s γ Valve flow exponent R f Rangeability factor 54

79 4.6.4 Equations Due to the isenthalpy assumption in the valve the inlet and outlet enthapies are equal: h out = h in (4.98) h in = PhysProp.Enthalpy(T in, p in, 1) (4.99) Note that in the flow is implicit that the inlet flow is equal to the outlet since only one flow variable was defined, the same condition can be applied for the composition. The outlet temperature is determined by a p-h flash call: T out = PhysProp.PHFlash(p out, h out, 1) (4.100) Valve dynamics can be described by the following equation: Inherent flow characteristics can be one of the following options: act dvsp τ = V sp Vsp act (4.101) dt Linear EqualPercentage QuickOpening CASE: Cv f = R(V f Cv f = Cv f = V sp act + L f (4.102) 1 + L f act sp 1) + L f 1 + L f (4.103) act πvsp sin( 2 ) + L f (4.104) 1 + L f The flow across the valve will be a function of the maximum allowable flow for the valve, depending on the stem position opening and the inherent flow characteristics: F = Cv f.f max (4.105) The pressure drop in the system is defined as: F = Cv.Cv f. p 1 γ (4.106) p = p in p out (4.107) Degree of freedom Analysing the number of variables and the number of equations the degree of freedom of the model is DOF = N var N eq = (C + 17) 8 = C + 9 (4.108) The valve specification includes the following obligatory specifications: 1. Inlet streams : Temperature, pressure and composition. 55

80 2. Inlet control port : Stem position. 3. Flow coefficient 4. Maximum allowable flow across the valve Additional Specifications: 1. In advanced mode (in standard mode these options are assigned with default values): Time constant Leakage fraction Flow exponent Rangeability factor (only necessary in equal percentage inherent characteristics) Initial Conditions The user must provide initial conditions for each of the state variables. The following options are supported from the model specification dialog: 1. Steady state 2. Dynamic Specify actual valve stem position dv act sp dt = 0 (4.109) 56

81 5 Supercritical Pulverized Coal Power Plant Modelling Contents 5.1 Design Mode Operational Mode Control Mode Sensitivity analysis

82 The topic in this chapter is the development of a composite model representing a supercritical pulverized coal power plant (PCPP) with the models already explained in the chapter 4 and the models available on the gccs library. The idea behind the modelling of the PCPP is to be able to verify the accuracy of the models taken in account all the assumptions at the same time, to simulate with a high degree of accuracy part load operations with turbine following control implemented and for a future work to connect the power plant with the capture plant, compression, liquefaction and storage. This will allow the partners and clients to study sensibility issues and relationships in the entire carbon capture cluster such as the effects of CCS on power plant operations. In this chapter is presented a comparison between the results obtained by the gccs tool-kit and the results presented in the paper Designing a supercritical steam cycle to integrate the energy requirements of CO 2 amine scrubbing from Luis M. Romeo, Sergio Espatolero and Irene Bolea, the simulations in it were carried out using Aspen Plus software (Aspen 2003). [2] 5.1 Design Mode The design mode is one of the 3 steps to simulate part load operations, in this step the results presented in Romeo et al. [2]are matched by the gccs fixing the main operating conditions (feedwater flowrate, temperatures and pressures of the streams) to obtain the equipments design parameters (size, area, efficiency, etc).the flowsheet configuration is the same as the paper and is presented in the figure 5.1. Figure 5.1: Supercritical steam cycle flow diagram. Streams conditions can be found in the appendix B.[2] As it can be seen the system has the expected steam cycle having a regenerative reheat cycle with 8 feedwater heaters, four low pressure FWH and four high pressure feedwater heaters. A deaer- 58

83 ator unit, four pumps, one boiler, high, intermediate and low pressure turbines and a turbine pump ( turpump ) are also found in this power plant. Some notes should be made for this system: Turmpump relationship with the pump 4 FWH 8 with outlet drain in vapour phase No governor valve represented No flue gas side represented or information provided The turpump and the pump 4 (P4) present a relationship suggesting that the turbine provides the shaft power required by the pump. Another note is that the last feedwater heater (FWH 8) is not a real FWH in a sense that the drain from this unit is in vapour phase. The feedwater heaters outlet drain should only be saturate or sub cooled water however the gccs FWH model does not have that assumption, therefore it s possible to have outlet drain in vapour phase without any problem. From the figure 5.1 no governor valve is represented which in the gccs is included since it is crucial for control purpose in part load operations. Another thing is that the paper does not present any data on the flue gas side, except the flue gas flowrate and composition. The gccs flowsheet can be seen in the figure 5.2. A simple explanation of the gccs model library used in this work can be found in the appendix A. Figure 5.2: gccs diagram of a Supercritical Pulverized Coal Power Plant. Legend: A - Source coal; B - Source air; c - Boiler; D - Governor Valve; E - Turbine; F- Generator; G - Condenser; H - Drum; I - Pump; J - Feedwater heater; K - Deaerator; L - Electrostatic Precipitator; M - Blower; N- Gas/Gas Heater; O - Flue gas desulphurization unit; P - Stack; Q - Recycle breaker The development of the flowsheet was done step by step adding one model at a time, one issue that soon needed to be solved was the incapability for the gproms to solve the recirculation of information that naturally exists in the steam cycle. The solution found was the introduction of the 59

84 recycle breaker model which basically simulates a source and a sink and at the end of each simulation replaces automatically the source values by the solution found until the system converges. The degrees of freedom of the power plant was analysed and the necessary assignments were made, the most important are summarized in the table 5.1. All the models are assigned with the stream conditions to match the results and obtain the operational parameters for the next step. The stream conditions and main operating conditions from the reference [2] are presented in the appendix B. Table 5.1: Main assignments in the flowsheet. Model Boiler Turbine GovernorValve BoilerSteamCondenser FeedWaterHeater Deaerator Drum PumpUtility ElectrostaticPrecipitator Blower GGH FGD Coal Air Specification Superheat (SH) temperature and pressure Reheat (RH) temperature and pressure drop Inlet pressure Outlet temperature or vapour fraction Stem position Operating pressure Minimum temperature difference Outlet temperatures Operating pressure Residence time Volume occupation Discharge pressure Efficiency Discharge pressure Outlet hot temperature Outlet limit concentration Coal ultimate analysis Air composition and humidity One additional assignment, specified in the process was the outlet pressure of the last turbine set equal to the pressure in the condenser. This assignment is necessary to meet the degrees of freedom of the power plant and will influence the flux of information of the cycle. The last turbine pressure is known so all the pressures are known backwards therefore the outlet RH pressure is already known so the RH pressure in the boiler needs to be a pressure drop specification. Another important thing to understand is that the feedwater flowrate is specified on a special recycle breaker that only closes the temperature and pressure because obviously if no flowrate is lost during the cycle what is specified is equal to what returns. However, if the flowrate of water was not specified the expected power should be specified in the generator model. The introduction of the governor valve requires a change of the specifications because it introduces a pressure drop, reducing the pressure and the temperature at the high pressure turbine. To avoid this deviation a study was done to discover the right SH conditions in the boiler to obtain the expected conditions after the governor valve (the outlet SH conditions from Romeo et al. [2]). 60

85 The relation already stated between the turpump and the pump 4 is represented by a recycle break that sets the power produced in the turbine equal to the power required by the pump. It s important to notice that the last FWH (FWH 8) is not specifying two temperatures because its flow is already known, was determined by the FWH 5. Some values were adjusted to match the results such as the vapour fraction in the low pressure turbine 5 and in the turpump because having a fraction of liquid increases the turbine isentropic efficiency therefore the power production. Although the Romeo et al. [2]] doesn t mention the vapour fraction in the last turbines the outlet temperatures are at saturation conditions which suggestes that steam moisture can be found in those turbines. The values of humidity in the air, oxygen mass fraction in the outlet flue gas, fraction of carbon in ash and fraction of ash in the flue gas were adjust to get the best results possible in comparison with Romeo et al. [2] Results and Discussion A Stream conditions The design mode is based on the specification of the pressures and temperatures. The flowrates of steam drawn from the turbines could be used to verify the model. The only exceptions are the drawn off steam flowrates from the turbines but nothing significant because the biggest error has the value of 6%. This is due to the steam tables used that have minor differences and because the FWH 5 determines the flowrate of the drawn off 6, but if this value is already a little deviated then the temperature of 29 will be deviated as well (3% deviation), please see figure 5.2 and 5.3 for a better understanding of the system. Note that if the temperature 29 is wrong so the error in the flow will increase not only due to steam table differences. A representation of the high pressures feedwater heaters can be found in figure 5.3. The stream conditions and the deviation can be found in the appendix B. Figure 5.3: Flow diagram of the HP pressure feedwater heaters side B Key Performance Indicators Key performance indicators obtained were compared with the ones from the reference [2] and are presented in the table 5.2. The results show that the power obtained is 1.71% lower than the expected 61

86 and the coal and flue gas flowrate were matched adjusting the fraction of unburned carbon. Both coal and flue gas flowrate are calculated in the boiler model. Table 5.2: Key performance indicators (KPI) deviation. KPI % Flue gas compostion % Gross power -1.7 CO Net Power -0.7 N Specific CO 2 emissions 2.4 H 2 O 0.0 Gross efficiency -1.8 O Coal flowrate 0.1 SO Flue gas flowrate 0.0 Although there is a lack of information in the Romeo et al. [2] to made a better comparison the results are in an acceptable range. The biggest error is the specific CO 2 emissions with 2.4% because the power is lower than the expected and the carbon dioxide in higher. The excess oxygen and the air humidity were adjusted to match the oxygen composition and the hydrogen composition. The air has an humidity of 60%. Analyzing the gross efficiency may be concluded that since the same amount of coal is used, less power is generated which is probably due to limitations on the steam cycle. The turbines models used from Aspen used by the reference [2] probably have different assumptions which make the gccs turbines generate less power. The gross efficiency is defined as the power generated divided by the ideal power produced which is equal to the lower heating value of the coal times the coal flowrate C Equipment Parameters In design mode, the equipment parameters and sizes are calculated, as a simplification is just like sizing the equipments fixing the expected performance. To clarify, in design the temperatures and pressures of every stream are specified in order to obtain the equipment parameters and sizes which will be used in operational mode to allow part load operation. In part load operation when the areas and sizing are assigned the temperatures and pressures change according to the load. For the heat transfer equipments such as feedwater heaters and condenser the overall heat transfer coefficient is constant and specified with the value of 5000 W m 2 K 1 which is within the typical range of W m 2 K 1 for the FWH and W m 2 K 1 for the condensers. [28] In the following tables are presented the values obtained for the feedwater heaters, condenser and turbines. The results obtained for the feedwater heaters terminal temperature difference is close to the typical range that should be around 3 K for low pressure FWH and around 15 K for high pressure. [35] Parameters Table 5.3: Equipment parameters for the feedwater heaters and the condenser. Feedwater Heaters Condenser TTD (K) Area (m 2 )

87 Table 5.4: Equipment parameters for the turbines. Parameters HP 1 HP 2 IP 1 IP 2 LP 1 LP 2 LP 3 LP 4 LP 5 Turpump η (%) Stodola s constant In the table 5.4 are presented the equipment parameters for the turbines. The efficiency is more or less constant around 90% in all the turbines with exception for low pressure turbines (3 and 4) where the efficiency starts to decrease and in the LP5 where is really high due to two phase appearance. The vapour fraction in the LP5 turbine is 94% and 91% in the turpump this values are within the acceptable values not lower than 85-90% because lower vapour fractions can damage the turbine. [45] Stodola s constant is a parameter used to relate through an empirical equation the pressure ratios and the flowrate of vapour being expanded in the turbine. 5.2 Operational Mode The second step is to introduce equipment parameters determined in the design mode (5.3 and 5.4). The objective of this step is to ensure that changing the specifications (can be seen 5.5) in the same stream conditions are achieved. This will allow in a further step to go to part load operations with no fixed temperatures or pressures in the steam cycle but only fixed equipment performance parameters. A large number of specification trade-offs are implemented in this step, for some models the specification dialog already gives an indication switching from design to operational mode. Examples are the turbines, the feedwater heaters and the condenser. However the deaerator doesn t have that option and the specification is changed from operational pressure specification to pressure drop specification. The specification trade-offs are summarized in the table 5.5. Table 5.5: Specification trade-offs for operational mode. Model Design Mode Operational mode Condenser Pressure Pressure CW oulet temperature Area FWH Drain tempetarature TTD Feedwater oulet temperature Area Deaerator Operational pressure Pressure drop Drum Residence time Design volume Volume of liquid Level Turbine Inlet pressure Stodola s constant Outlet temperature or vapour fraction Efficiency These are the specification changes for the operational mode, all the operational data was determined by the models in the design step. At this stage the results for the stream conditions and key 63

88 performance indicators are once more analyzed however as expected just by switching the specification the results obtained are the same. Therefore the results are not presented in this section and are presented the Appendix B. 5.3 Control Mode In this section some control loops will be introduced in the system to try to reproduce the system response of a real power plant. Although, the only model with full dynamics is the drum, the implementation of the controls loop will confirm and help to understand the system response to load changes. Obviously with load changes the pressures, the temperatures and the feedwater flowrate changes. Therefore the specification determined in the first step and tested in the operational mode are now used and kept constant over the course of load changes. To successfully model the control system, the controlled variable that was assigned in the operational mode is now given in the controller model as a set point. The controller will send the signal to the valves where the stem position is changed according to the desired response. The control loops implemented were: the pressure control in the condenser, the level control in the deaerator drum and the turbine following control loop. In the turbine following the power and the superheat pressure are controlled using the coal flowrate and changing the stem position in the governor valve. In this control mode, the power is specified in the power controller and not the feedwater flowrate in the recycle as it was specified in design mode. The controllers used were not optimized because that was not the objective of this thesis. A daily cycle will be studied in this section Control loops A Condenser pressure This control loop which is presented in the figure 5.4, is usually implemented in power plants where the pressure is controlled with the cooling water flowrate. If the cooling water flowrate increases the pressure will decrease because more heat will be transfer and more steam will condense. Figure 5.4: Condenser pressure control loop. 64

89 Naturally there is a limit to this heat exchange, being the amount of flowrate possible to pump and the outlet temperature of the cooling water which can t increase more than the saturation temperature of the steam at that pressure B Deaerator drum level The drum model wasn t initially a model required for the gccs library. However, it was necessary to add these drums (after the condenser and deaerator) to serve as buffers in load changes. The explanation is that the models are steady state, so when a controller is introduced for example in the condenser, it is just like specifying the flowrate of cooling water and then the pressure is calculate. If the load decreases so will the steam flowrate to the condenser, therefore the pressure will drop instantaneously and since the condenser model sets the temperature to the saturation this will decrease with the pressure. The problem with this happening is that will produce a perturbation in the feedwater heaters and eventually will crash due to temperature crossovers. With the introduction of this dynamic drum models, even though the pressure still decreases in the condenser that change will be diluted in the tank. One future work could be introducing dynamics in the condenser and the deaerator models to better capture the system responses. In spite of these problems in a real power plant the drums exist and are controlled as it can be seen in the figure 5.5. The condenser tank is typically controlled by the makeup water, losses due to leakage, steam venting or non recoverable steam usage need to be compensated. [24]. Because there isn t any water loss in the system this control loop is not represented in the flowsheet. Figure 5.5: Deaerator drum level control loop. Legend: A - Drum; B - Deaerator; C - Feedwater heater; D - pump; E - Recycle breaker; F - Governor valve; G - Controller The deaerator s drum serves as a surge tank for boiler feedwater and in order to control the level the inlet feedwater into the deaerator is manipulated. In terms of modelling that is exactly what is represented and the outlet feedwater from the tank is specified from the governor valve steam requirement. In the deaerator s tank the residence time is not specified as it was in operational mode, but is specified (as set point in the controller) the level which is the controlled variable. 65

90 5.3.1.C Turbine following control loop The last, but nevertheless the most important control loop is the turbine following as described in the section 2.4.1, it is a load demand adjusting system. Basically, when a load increase is required the first step would be to change the firing rate (flowrate of coal) which will induce an increase of pressure in the boiler. The pressure controller will then send a signal to the governor valve to decrease the pressure by opening the valve. The figure 5.6 presents the turbine following control as it is implemented in the gproms. Figure 5.6: Turbine following control loop. Legend: A - Source coal; B - Source air; C - Boiler; D - controller; E - Governor valve; F - Turbine; G - Generator; H - Recycle breaker. This type of control loop is more used when a frequency power change is required. For a typical grid energy production a boiler following mode is more suitable since uses stored energy in the boiler to provide immediate load response D Others controls The power plant have much more controls loop like the ratio controller with the air and the fuel and the level controllers in the feedwater heaters that were not represented. The feedwater heaters and the deaerator control were not represented due to the lack of dynamics in the system. An important thing to note is that the actual flow drawn off from the turbine is not controlled but is pressure driven Step change in power plant load The reference [28] presents a comparison between the load control loops (figure 5.7). This figure helps to understand the system dynamics with the different types of control. The turbine control has the behaviour already explained in the previous section and the boiler following control response has a decrease in the throttle pressure when an increase of power demand is required. This is due to the fact that when a power increase is required the governor valve is immediately adjusted which will make the pressure to decrease, only after this the coal will be adjusted to make the pressure go back to the set point. The coordinated or integrated control instate the best of both controls systems being the fastest and having less pressure deviation than the boiler following control. A positive step change was simulate and the results are presented in the following charts. Since no relative deviation was defined in the reference, a step of 4% was defined for the simulation from 90% to 94% of load change. 66

91 Figure 5.7: Load change and throttle pressure deviation for the different types of load control. The step change was done within the typical rate change allowed in a power plant. Physical restrictions in supercritical power plants usually set the limit rate of load change to 8%per minute. [46] In the simulation this will help to obtain less sharp and more smooth results, as it can be seen in the figure (5.7) the power response seems to be second order. However in the simulation the results suggests only first order response and that response is only due to the controller that adds one order to the system. To get a better response is only possible by adjusting the controllers or adding dynamics to the system. This is a limitation on this simulation in terms of control because even thought the results represent correctly the general behavior of the system, it currently does not accurately predict the dynamics of the system. Analysing the figure 5.8, it can be concluded that the pressure is well controlled since it is almost constant having no significant deviation, the biggest deviation achieves 1.2%. Although the reference step change is unknown, the stabilizing times are close to the expected and still show that the power controller reaches the set point without any problem. The stabilizing times are presented in the table 5.6, there was no intention to match the time frames but the relationship between them, in other words the pressure should always stabilize faster than the power reaching the set point. The stabilizing time is defined as the time when the variable becomes within 99.5 % of the set point objective. In figure 5.8, can be seen the different controllers tested and their parameters can be found in table 5.6. Table 5.6: Controllers parameters and stabilization time for every controller. Controller Gain Power Controller Pressure Controller Reset Stabilization Reset Stabilization Gain time (s) time (min) time (s) time (min) Reference

92 Figure 5.8: Load change and throttle pressure deviation results for 4% step change in load. The power controller has a direct action and the pressure controller an inverse action on their corresponding manipulated variables the firing rate and governor valve stem position, respectively. The controllers were tuned manually however the first guess for the parameters were based on the values obtained from reference [47]. The values from the reference found are not exactly as it was supposed to be, since they are used for coordinated control which is completely different. Nevertheless, especially the reset times can be valuable providing a typical value since those are more likely to keep constant. The gain of the controllers certainly should not be the same because the measured variables are different. Therefore the reference controller is the initial guess for the manual tunning, it has the literature parameters for the power controller and the reset time of the literature for the pressure controller. The gain of the pressure controller starts with 1, since the gain is unidimensional. Observing the gain of the controllers and the stabilizing time can be concluded that as bigger it is the gain the faster the response becomes. The controller 1 with the power gain of 0.15 is the quickest to get to the set point only taking 1.1 minutes to reach it and 1.7 minutes to stabilize the pressure which does not represent correctly the expected behavior. The controller that shows the closer behavior to expected is the controller 2, taking more time to reach the set point (2.4 minutes) than to stabilize the disturbance in the pressure (1.8 minutes). The set of parameters used in the controller 2 will be used to simulate the daily cycle in the following section. 68

93 However, in spite all this efforts to find the controller is important to know that in the limit if the controllers were to good (high gain and lower reset time) the response should be seen immediately, the entire dynamics is given only by the controllers and the governor valve Daily Cycle In a country power demand there are 3 types of energy load requirements: the base load, the intermediate load and the peak load. The base load is the minimum level of demand on an electrical supply system over 24 hours. Base load power sources are those plants which can generate dependable power to consistently meet demand, typically nuclear and coal power plants are used. Base load power plants produce continuous, reliable, efficient power at low cost and often are relatively inefficient at less than full output. Electricity demand fluctuates over the course of a day, the power demand is then met by intermediate and peak load power plants. Usually base load power plants produce between 30% and 40% of the power requirements and intermediate meet 30% to 60%. The rest is fulfilled by the peak load power plants.[48] Peak load generators, such as natural gas, have low fixed costs, low plant load factor and high marginal costs. Also coal can be used as intermediate and peak load power plants, but natural gas is much more flexible and faster than coal power plants. For purpose of this work, coal is considered base load and is analyzed the daily demand for the Portuguese Electrical Grid (provided by REN) for the 3 of September of 2012.[49] Figure 5.9: Daily cycle of the Portuguese National Grid from 3 of September Results Analyzing the figure 5.9, the schedule for the daily cycle of the power is decided. Clearly from 1 a.m to 8 a.m less power is required therefore the power plant could work at 90% load. From 8 a.m to 5 p.m the power plant is working at full capacity, then drops to 95% until 10 p.m when it goes back to the 90% load. the simulation is done with the previous control loop explained: condenser control, drum level control and turbine following control. The simulation of this production schedule is presented in the following figure

94 Figure 5.10: Simulation results from the daily cycle schedule. The controllers implemented were manually tuned, the condenser controller and the level control. The parameters used for the turbine following control are the ones used in the controller 2 of the step test in the previous section. The controllers parameters are summarized in the table 5.7. Table 5.7: Table summarizing the controllers parameters. Controller Gain Reset time (s) Power Boiler pressure Condenser pressure 2 3 Deaerator tank level 8 The level controller was define as a proportional controller since typically in industry level is controlled with proportional controllers, having a high gain ensures that the level stays close to the set point but at the same time can make the system unstable, therefore a trade-off between both must be determined. As previously stated, the change in set point is considered in terms of the physical limits typically allowed in real power plants, so in order to be inside those limits the same rate of load change was used 8% per minute. [46] Primarily, are presented the results to the variables directly affected by the load change, the coal, the governor valve stem position, feedwater flowrate and boiler pressure. When the coal flowrate is increased (power demand increased, more heat required) the pressure in the boiler increases (figure 5.8), then the governor valve will open decreasing the pressure drop and letting more vapour to flow across and going into the turbines which will decrease the pressure. This physical response can be seen in the following figures 5.11 and Observing the figure 5.11, is possible to see that the pressure is always well controlled being the biggest deviation with the value of -4.3% which represents about 13 bar decrease in bar of set point pressure, however it is possible to reduce this results decreasing the rate of load change which is already in the upper limit. This deviation occurs in the change from 100% to 95% ate 5p.m, and is interesting that a negative step causes more impact than a positive one, which is related to the fact that the controllers were tune for positive disturbances. 70

95 Figure 5.11: Boiler pressure and governor valve stem position response during the daily cycle schedule. Figure 5.12: Feedwater and coal deviation from full capacity load for the daily cycle schedule. This is a important restriction in load change because the pressure in the boiler should not change to much and depends on the materials and type of boiler, but is certainly of crucial importance in the power plant safety and performance. At 8 a.m, the biggest change in load of the day happens, changing the load from 90% to 100% which makes the governor valve to saturate since it was designed for full capacity. In terms of feedwater flowrate and coal requirements, the deviation tracks the evolution of load demand. The deviation of the feedwater flow is exactly the same as the load, but for the coal a production of 90% load only saves 9% of coal this will cause the gross efficiency of the power plant to drop to around 1%. The rest of the steam cycle is also affected by the load changes, for example the level controller in deaerator tank (figure 5.13). The deviation in the level set point is irrelevant and the main reason is that what doesn t get in the deaerator tank is accumulated in the condenser tank. When the load increases so does the feedwater that goes into the boiler, to adjust this increase in the outlet of the tank, the valve after the condenser drum changes the stem position to have more flow out. The outcome of this action is the decrease of the level in the condenser tank. Since this control loop is the only real dynamic loop, will be presented a closest look on a step from 90% to 100%. Analyzing the figure is possible to see a typical proportional response. The manipulated variable is a valve and the stem position is presented in the figure The valve goes to half open in the full 71

96 Figure 5.13: Responses of the manipulated and controlled variable for a change of load from 90% to 100%, for the level control. load since all the valves were sized for double the flow. The level also achieves the set point without any error because the first control arrangement was for 100% load therefore when it comes back it will get to the same point exactly. The response in figure 5.13 is first order. The proportional action is present, for example the steady state in 90% load is not at the step point but has an error and that error is due to the proportional control. The proportional control only sets the inlet flow equal to the outlet but all the accumulate error is never compensated has it is with the integral term of a PI controller. The level change in the condenser tank is not presented, however since the tanks are the only two dynamic models in the power plant, whatever one loses is accumulated in the other. The condenser pressure control is a proportional and integral controller with the values of the table 5.7. The response for the daily cycle is presented in the following charts, figure Figure 5.14: Responses of the manipulated and controlled variable during the daily load, for the condenser pressure control. As it can be seen, the pressure is not so well controlled during part load operations, however this results need to be carefully analyzed since this suddenly increase in pressure is mainly due to the lack of dynamics in the model. This exemplifies the need to add the tank model since this increase in pressure would immediately be seen in the feedwater heaters which would cause them to fail at some 72

97 point along the line. The cooling water flowrate is being manipulated for part load operations. Therefore, for a decrease in the load, a decrease in the inlet steam pressure of the condenser will occur and in order to maintain the pressure at its set point less cooling water flowrate will be required. Looking for the overall performance of the system when the load changes, the entire cycle would change as well - stream temperatures, pressures, flowrates. For example the temperature and pressure of the feedwater heater 6 and the deaerator are presented in the figure In part load operations, less power is required therefore less steam flow is required. Since the turbines have the same performance with less flow the pressures and temperatures will start to decrease and thats when the TTD of the FWH and the pressure drop of the deaerator are of extreme importance, because they follow the decrease in temperature and pressure of the turbine system to keep the steam cycle integration system working. Figure 5.15: Variation of the feedwater heater and deaerator temperature and pressure during the day. Another important thing to analyze is the vaporization of the steam in the last turbines and in all of the turbines, making sure that none had vapour fraction below the mechanical limit of 85%. The simulation results determined that the vapour fraction along part load operations varies but in none of the cases drops below 90% of steam. This assures that the mechanical limits of the equipments are not being exceeded in the simulation. he turpump power is being determined by the amount of the power that the pump 4 requires. In part load operations the same happens and obviously is less the power required by the pump because less feedwater is circulating. In order for the turpump to achieve the power required, the steam drawn off will decrease. The key performance indicators are presented for the different loads in the table 5.8 A direct relation is visible in this table since the interaction between all the key performance indicators seems to be suggesting some proportionality. Even though none of the variables decreases in the same degree has the load they follow the same direction. Exception to this is the specific carbon dioxide emissions because the systems is less efficient and the amount the coal is not saved in the same degree the atmospheric emissions will be more severe per megawatt produced. 73

98 KPI Table 5.8: Power plant Key performance indicators for different loads. Reference Case Power plant load 100% 95% 90% Gross power (MW) % % Gross efficiency (%) % -1.2 % Specific CO 2 emissions (gco 2 /kwh) % 1.2% Coal flowrate (kg/s) % -8.9% Flue gas flowrate (kg/s) % -8.9% The coal tendency is the same as the flue gas, the reason for this is simple, if for example, less 2 % of coal is burned in the boiler the flue gas flowrate will decrease exactly 2%. If the amount of flue gas decreased so the amount of carbon dioxide sent to the stack. However, if the power decreased more the specific carbon dioxide emissions will increase per megawatt. The gross efficiency decrease as well with the load which shows that the power plant is losing money comparing to 100% load. The relative difference between the power and the coal deviation gives the gross efficiency relative deviation. Which is due to the fact that the steam cycle in part load operation does not maintain the same performance in the heat integration. This results corroborate the previous stated that base load power plants typically lose efficiency in part load operations. 5.4 Sensitivity analysis To study the effects of some assumed performance parameters in the equipments, a sensitivity analysis was conducted for the boiler efficiency. This will provide ideas on how important is the boiler efficiency in terms of power generation. An important factor that can be analyzed is the effect of the reheat steam temperature in the power plant performance Boiler efficiency In this section was studied the effect of the boiler efficiency in the system. Efficiencies of 2% and 4% above and below of the assumed valued were simulated and the results for the key performance indicators are presented in the table 5.9: KPI Table 5.9: Boiler efficiency sensitivity study, relative deviation to the reference case is presented. Boiler efficiency cases Reference Case Study cases 94 % 90 % 92 % 96 % 98 % Boiler efficiency (%) -4.3 % -2.1 % 2.1 % 4.3 % Gross power (MW) % 0.0 % 0.0 % 0.0 % Gross efficiency (%) % -2.1 % 2.1 % 4.3 % Specific CO 2 emissions (gco 2 /kwh) % 2.2 % -2.1 % -4.1 % Coal flowrate (kg/s) % 2.2 % -2.1 % -4.1 % Flue gas flowrate (kg/s) % 2.2 % -2.1 % -4.1 % Analyzing the results, is easy to conclude that the main effect is not on the power generation 74

99 side since that all the temperatures and pressures in the steam cycle are kept constant. However changing the efficiency will determine how much coal is required, since that less heat is lost due to inefficiencies. The amount of coal used will determine the increase or decrease of the power plant efficiency. As it can be seen in table 5.9, the increase of the boiler efficiency is directly proportional to all key performance indicators. For example an increase of 2% in the boiler efficiency will lead to a decrease of 2% in the coal as expected, due to a enhancement in the heat transfer of the boiler, therefore less flue gas will be sent to the stack (2%). The specific carbon dioxide emissions will decrease 2% because the coal burned is reduced so less carbon dioxide will be emitted for the same power output. The opposite happens to the overall gross efficiency of the power plant which increases 2% due to a decrease of the coal required, in order to achieve the same power. The conclusions for this sensitivity are the expected which is higher the efficiency of the boiler the better, because improves the consumption of fuel and therefore increases the overall efficiency of the power plant and decreases the amount of carbon dioxide sent to atmosphere, reduce the inefficiencies in the heat transfer of the boiler can significantly improve the power plant performance. The enhancement is proportional to the improvement of the boiler efficiency Reheat temperature The reheat temperature is a condition of the power plant design, typically for supercritical power plants the temperature can be above 600 o C which is the case for this power plant. The effect of the reheat temperature in the overall performance of the power plant will be the objective in this section. It s expected an improvement of the performance of the power plant with the reheat temperature. This is because higher temperatures allow a better use of the steam s energy by the turbines. Table 5.10: Reheat temperature sensibility study, relative deviation to the reference case is presented. KPI Reheat temperature cases Reference Case Study cases o C o C o C o C Reheat temperature (%) 1.4 % 7.2 % 14.5 % Gross power (MW) % 3.4 % 7.0 % Gross efficiency (%) % 0.7 % 1.4 % Specific CO 2 emissions (gco 2 /kwh) % -0.7 % -1.4 % Coal flowrate (kg/s) % 2.7 % 5.5 % Flue gas flowrate (kg/s) % 2.7 % 5.5 % Table 5.10, presents the results for three different temperatures with an increase of 1.4%, 7.2% and 14.5%. As it is possible to observe the gross power in this case is affected by this change in the system. The key performance indicator that improved the most with reheat temperature, was the power production, for the 1.4% increase in the reheat (RH) vapour temperature the power increases 0.7%. In terms of numbers it represents an increase of 3 MW for an increase around 9 o C in the temperature, which clearly confirms the importance of the reheat temperature in the power generation. 75

100 Another thing to note is that the coal does not increase in the same proportion as the power, which is a good thing meaning that the efficiency of the power plant will increase with the temperature improvement. For example, in the case two the temperature increases 7.2% and in order to achieve the required temperature, the coal increases 2.7%, therefore the flue gas will also increase in the same proportion of the coal flowrate. For this case the gross power increases more than the coal (3.4%), which will lead to higher power plant efficiency of 0.7%. Even thought the coal increases the performance of the power plant is improved producing more power without having to increase the coal in the same proportion, also the environmental performance improves decreasing 0.1% the amount of carbon dioxide sent to the atmosphere per Megawatt produced. The steam conditions in the steam cycle can be found in the appendix B. However with bigger values of reheat temperature, bigger are the deviation of the temperatures, pressures and flowrates in comparison to the reference case. Even thought the efficiency of the power plant increases because the steam cycle heat integration is improved with this increase. The feedwater entering the boiler has an increase in the temperature in the same range of value as the gross efficiency. So for example, for an increase of the RH vapour temperature of 7.2% the feedwater s temperature increases in 0.7% which will allow the system to save coal leading to an increase of the gross efficiency in 0.7%. The conclusion is that the power generated is highly improved by the increase of the reheat temperature however a limitation on equipment side makes impossible to increase that temperature has high as desired because of materials restrictions on how high the temperature can be. New materials are nowadays being discover and adapt to improve this restrictions. Modern chrome and nickel-based super alloys in the steam generator, steam turbine, and piping systems can withstand prolonged exposure to this high temperature steam.[27] Future materials will push further the efficiencies and incomings of supercritical power plant generations leading to temperatures around C helping the plants to achieve above 50% efficiencies.[50] 76

101 6 Conclusions and Future Work Contents 6.1 Future Work

102 The subject of this dissertation was the development of a supercritical pulverized coal power plant model and some of the relevant sub-models. The objective was the simulation of part load operation along with some control strategies. The control strategy implemented was turbine following control and with it a daily cycle of a power plant was simulated. The main original contributions of this work were the development from scratch in the gproms software of the sub models and the composite power plant model. Also the understanding of the system which allows the simulation of part load operations fixing the equipments design parameters. Nothing like this was ever implemented in gproms, therefore the first step was to prove the accuracy of the models and the design mode which prove that the modelling accuracy of the models used are at the same level as the ones used by Romeo et al.[2] with the ASPEN software. The power deviation is the only meaningful difference having less 1.7% of gross power because of the assumptions used for the gccs turbine models. However due to this error the other key performance indicators suffer some deviations exception done to the coal required and the flue gas flowrate which were matched. A supercritical pulverized coal power plant was simulated producing MW of gross power with an efficiency of 46.25% and a specific carbon dioxide emission of g/kwh of gross power. The coal consumption is kg/s producing kg/s of flue gas. The power plant has 350 kg/s of feedwater circulating the system. The understanding of the system allowed the introduction of a second simulation mode called operational mode. The operational mode made possible the change of specification based on the results of the design mode without losing the accuracy already achieved. Some of the important variables in the system were the terminal temperature difference and the deaerator pressure drop that allow part load operations and establish a relation between the temperatures and the pressures of the system. The control loops implemented showed the expected response which was proved by the step change done in the load. However due to the lack of dynamics in the models the introduction of dynamic drum models was crucial to part load operations. Since that the tanks were the only two dynamic models whatever losses in one was accumulated in the other. The tanks were used as buffers which was of extreme importance because when a change in the pressure of the condenser occurred the change would be felt in the FWH line which would cause them to fail since the temperature would be far away from the expected. With the tank, the pressure in the condenser can change but the tank acts as a buffer until the controller puts the pressure back to its set point. The daily cycle simulation presents a clear relation between the key performance indicators and the load. The decrease in power production of 5% does not mean that the coal consumed decreases 5%, actually the coal consumption decreases around 4.5% which makes the efficiency of the power plant worse in 0.6%. This is due to the fact that the steam cycle in part load operation does not maintain the heat recovery efficiency. The flue gas flowrate is directly proportional to the consumed coal and so is the carbon dioxide 78

103 emitted however the specific carbon dioxide emissions will increase 0.6% due to the greater decrease of power than the flue gas emission. The daily cycle helped to understand the dynamic in part load operation. To produce less 5% of power, less feedwater is required in the same proportion (5%). With less feedwater the outlet temperatures and pressure in the turbine will be reduce and all the steam cycle will be rearranged. The amount of reduction is different from the steam cycle position but is accentuated with the load decrease. The boiler efficiency was considered constant along part load operations however a sensitivity analysis was perform and the conclusions indicate that the improvement of 2% of the boiler efficiency impacts positively the power plant gross efficiency in 2%. This is explained with the fact that the heat transfer in the boiler is improved therefore the coal consumption reduces 2%. The sensitivity analysis was also conducted on a reheat temperature which showed that an increase in the reheat temperature of 7.2% would cause an increase of 3.4% in the produced power. With the reheat temperature increase the steam cycle integration is improved and the feedwater temperature entering the boiler is increased in 0.7% which enables a save in the coal consumption. Therefore the gross efficiency of the power plant is increase in 0.7%. The sensitivity analysis showed two ways of improving the power plant efficiency the first one the boiler efficiency can be improved with new approaches and materials to improve the heat transfer efficiency. The reheat temperature clearly affects the steam cycle completely and an optimization of the steam cycle with those conditions could increase up to 50% or more the efficiency of the power plant. However nowadays those temperatures are not possible to achieve but in a close future new materials will allow higher temperatures and pressures. The objective of this thesis was fully accomplished with the supercritical pulverized coal power plant being able to successfully simulate the part load operation. However some work can still be done to improve the accuracy of the simulations. In the following section are described some of the future ideas to complement the work done so far. 6.1 Future Work Power plant modelling is a widely studied subject however, there is a shortage of full chain CCS process modelling capability. Further work includes integrating the capture plant, and other parts of the chain with the power plant. In further steps the transmission and injection could also allow a study and optimization of the entire carbon capture chain. Another approach should be the improvement of some steam cycle models such as: condenser, deaerator and feedwater heater. The integration of dynamics is crucial to more accurate part load simulation, this would change the system degree of freedom making the flow to be pressure driven. A dynamic feedwater heater and deaerator could allow a study in the level controllers. Also the dynamics in the boiler and the coal milling could be explored to open a variety of control studies, having meaningful responses for the implemented controls such as turbine following. The optimization of power plant with cost estimation incorporated in the models could determine the number 79

104 of feedwater heaters and drawn offs required. Some of these ideas are already in place and this is just the first gccs library with the first steady state model that will be improved later. 80

105 Bibliography [1] P. Lako, Coal-fired power technologies: Coal-fired power options on the brink of climate policies, ECN, [2] L. M. Romeo, I. Bolea, and J. M. Escosa, Integration of power plant and amine scrubbing to reduce CO 2 capture costs, Applied Thermal Engineering, vol. 28, no. 8-9, pp , [3] DOE/NETL Carbon Dioxide Capture and Storage RD&D Roadmap, National Energy Technology Laboratory, December [4] C. Beales, UK emissions target, , last accessed in: 20 June [5] B. K. Sovacool, Valuing the greenhouse gas emissions from nuclear power: A critical survey, Energie policy, vol. 36, no. 8, pp , [6] J. Abraham, Very worried about escalating emissions? You should be, The conversation, 2011, Last accessed in: 14 June [7] The global status of CCS: 2011, Global CCS institute, 2011, Last accessed in: 20 June [8] Global CCS institute - CO 2 Capture technologies, Global CCS institute, 2011, globalccsinstitute.com, Last accessed in: 25 June [9] R. Crume, Monster increase in carbon dioxide emissions - But is anyone paying attention? Renewable energy magazine, 2011, Last accessed in: 30 June [10] Global carbon-dioxide emissions increase by 1.0 Gt in 2011 to record high, International Energy Agency, 2012, Last accessed in: 13 June [11] S. Espatolero, C. Cortés, and L. M. Romeo, Optimization of boiler cold-end and integration with the steam cycle in supercritical units, Applied Energy, vol. 87, no. 5, pp , [12] F. Alobaid, J. Ströhle, B. Epple, and H. G. Kim, Dynamic simulation of a supercritical oncethrough heat recovery steam generator during load changes and start-up procedures, Applied Energy, vol. 86, no. 7-8, pp ,

106 [13] M. Variny and O. Mierka, Improvement of part load efficiency of a combined cycle power plant provisioning ancillary services, Applied Energy, vol. 86, no. 6, pp , [14] C. C. Chuang and D. C. Sue, Performance effects of combined cycle power plant with variable condenser pressure and loading, Energy, vol. 30, no. 10, pp , [15] J. M. Klara, Bituminous Coal and Natural Gas to Electricity Final Report, Department of Energy (DOE), May [16] Engineering feasibility and economics of CO 2 capture on an existing coal-fired power plant, Allstrom, June [17] C. Liu, H. Wang, J. Ding, and C. Zhen, An Overview of Modelling and Simulation of Thermal Power Plant, In: Proceedings of the 2011 International Conference on Advanced Mechatronic Systems, Zhengzhou, China, August [18] Chemical Looping, Imperial College of London - Carbon capture and storage, 2012, www3.imperial.ac.uk/, Last accessed in: 13 June [19] S. Holloway, A. Karimjee, and M. Akai, Carbon Dioxide Transport, Injection and Geological Storage, IPCC Guidelines for National Greenhouse Gas Inventories, [20] Depleted Oil & GasFields for CO 2 Storage, IEA Greenhouse Gas R&D Programme, 2000, Last accessed in: 13 June [21] Electricity generation, OECD Factbook : Economic, Environmental and Social Statistics, 2012, Last accessed in: 13 June [22] J. Katzer, The Future of Coal-Based Power Generation With CCS, MIT Energy Initiative, http: //web.mit.edu/coal/, Last accessed in: 22 June [23] A. Raja, A. P. Srivastava, and M. Dwivedi, Power Plant Engineering, New Age International Publishers, [24] L. F. Drbal, P. G. Boston, and K. L. Westra, Power Plant Engineering: Black & Veatch. Kluwer Academic Publishers, [25] T. Ghosh and M. Prelas, Energy Resources and Systems: Fundamentals and Non-Renewable Resources. Springer Science and Business Media B.V., 2009, vol. 1. [26] E. G. de Azevedo, Termodinamica Aplicada, 2nd ed. Escolar Editora, [27] T. Riordan, New Generation Strategy Ultra-Supercritical Technology, American Electric Power, 2006, Last accessed in: 10 July [28] T. Babcock and W. Company, Steam/its generation and use, 41st ed. The Babcock and Wilcox Company,

107 [29] Wikipedia, Wikimedia Foundation, Inc., 2011, Last accessed in: 15 July [30] R. Beardmore, Thermodynamics-Steam Turbine, [31] The Dearator Principle, Last accessed in: 10 July [32] Basic Principle Of Deaerator and Deaeration. [33] Basics of closed feedwater heaters, Heat Exchange Insitute,Inc, Last accessed in: 10 July [34] Standards for Closed Feedwater Heaters, Heat Exchange Institute, Inc, [35] D. Hite, Heat Rate and Feedwater Heater : Level Control, Magnetrol International & Orion Instruments, [36] Air pollution Control Cost Manual, 6th ed. United States Environmental Protection Agency, [37] D. Lindsley, Power-Plant Control and Instrumentation: The Control of Boilers and Hrsg Systems, The institution of Electrical Engineers, [38] Fundamentals of Power Reactors, CNSC, January [39] Engineering Manual Industrial Boilers, Grundfos Industry, [40] A. Ramos, ETI CCS System Modelling Tool-kit Project Power Systems Update, Process System Enterprise, July [41] K. Parker, Electrical Operation Of Electrostatic Precipitators, ser. IEE Power and Energy. The Institution of Engineering and Technology, 2002, vol. 41. [42] A. Schumpe and G. Quicker, Gas Solubilities in Microbial Culture Media, Advances in Biochemical Engineering/Biotechnology, vol. 24, pp. 1 38, [43] R. Perry and D. Green, Perry s Chemical Engineers Handbook, 7th ed. McGraw-Hill, [44] B. W. Bequette, Process Dynamics: Modeling, Analysis and Simulation, 1st ed. Prentice Hall, [45] P. B. Meherwan, Handbook for cogeneration and combined cycle power plants. ASME Press, [46] Technology Status and Project Development Risks of Advanced Coal Power Generation Technologies in APEC Developing Economies, October 2008, Last accessed in: 20 July [47] R. D. Paranjape, Modeling and Control of Supercritical Coal Fired Boiler, May 1996, scribd.com, Last accessed in: 10 July

108 [48] M. Cordaro, Understanding Base Load Power: What it is and Why it Matters, New York AREA, October [49] REN, Last accessed in: 3 September 2012., [50] G. Stamatelopoulos, G. Scheffknecht, and ES.Sadlon, Supercritical boilers and power plants: experience and perspectives, In: Proceedings of power-gen Europe, Düsseldorf, Germany,

109 A gccs Model Library A-1

110 A.1 Main power plant models valve. The main power plant equipments are: the boiler, the turbine, the deaerator and the governor A.1.1 Boiler model The boiler model configuration is represented in figure A.1, with the ports connections. The boiler has an inlet and outlet port for the reheat steam, inlet port for coal and air, outlet port for the ash deposits from the coal and for the outlet flue gas, one inlet port for the feed water, one outlet for the outlet superheat steam and one control port for superheat pressure measurement. Figure A.1: gccs boiler model configuration. Specification options The following options will be available in the specification dialog. 1. Boiler properties tab Fraction of ash in flue gas Fraction of carbon in ash Boiler flue gas outlet pressure Boiler fractional efficiency Temperature of bottom ash Excess oxygen or outlet mass fraction of oxygen in the flue gas or molar fraction in dry flue gas 2. Steam properties tab Reheat outlet temperature Superheat outlet temperature Reheat outlet pressure or pressure drop in the RH side A-2

111 Superheat outlet pressure or pressure drop in the SH side This options are for the boiler operating conditions however this is a composite model and the air heater is include in this model it is possible to choose if the air heater exist and in positive case what are the specification for the unit. Also the tramp air (the air that enter the boiler) can be specified. 3. Air heater properties tab Heat exchanger effectiveness or the air outlet temperature or flue gas inlet temperature Flue gas pressure drop Air pressure drop Leakage fraction Heat transfer efficiency 4. Tramp air tab Inlet fraction Temperature outlet temperature Pressure A.1.2 Turbine model The turbine model configuration is represented in figure A.2, with the ports connections. The turbine has an inlet and outlet port for the steam and an inlet port for the power generated by the previous turbine and an outlet power port that sends the power generated in the turbine plus the power given by the inlet power port. Figure A.2: gccs turbine model configuration. Specification options The following options will be available in the specification dialog. 1. Design mode Inlet pressure Outlet pressure or outlet vapour fraction 2. Operational mode A-3

112 Stodola s constant Isentropic efficiency A.1.3 Deaerator model The deaerator model configuration is represented in figure A.3, with the ports connections. The deaerator has an array of inlets and one outlet port for the feed water, one inlet steam port, one outlet bleed steam port and one control port for pressure measurement. Figure A.3: gccs deaerator model configuration. Specification options The following options will be available in the specification dialog. Bled fraction Operational pressure or pressure drop A.1.4 Governor valve model The governor valve model configuration is represented in figure A.4, with the ports connections. The governor valve has one inlet and outlet port for the superheat and one inlet control port where the stem position is set. This model is very similar to the control valve the only exception is the fact that the maximum allowable flow is not applicable here. Figure A.4: gccs governor model configuration. Specification options The following options will be available in the specification dialog. 1. Specification mode A-4

113 In advanced mode (in standard mode these options are assigned with default values): Time constant Leakage fraction Flow exponent Rangeability factor (only necessary in equal percentage inherent characteristics) 2. Inherent characteristic Linear Equal Percentage Quick opening 3. Pressure drop Pressure drop known (specify stem position) Specify flow coefficient 4. Initial condition specification Steady state Dynamics: Specify actual valve stem position A.2 Auxiliary power plant models The auxiliary models used in the flowsheet were: the drum, the pump, the blower, the generator, the recycle breaker, the source coal, source air, source utility, sink utility, sink waste, the stack and the junction. A.2.1 Drum The drum model configuration is represented in figure A.5, with the ports connections. The drum has one inlet and outlet port for the water and one control port for measurement of liquid height. Figure A.5: gccs drum model configuration. A-5

114 Specification options The following options will be available in the specification dialog. 1. Design mode Residence time Volume liquid occupation 2. Operational mode Total volume Residence time Initial conditions: Liquid height Drum pressure A.2.2 Pump The drum model configuration is represented in figure A.6, with the ports connections. The drum has one inlet and outlet port for the water and one control port for measurement of liquid height. Figure A.6: gccs pump model configuration. Specification options The following options will be available in the specification dialog. 1. Design mode Residence time Volume liquid occupation 2. Operational mode Total volume Residence time Initial conditions: Liquid height Drum pressure A-6

115 A.2.3 Blower The drum model configuration is represented in figure A.7, with the ports connections. The blower has one inlet and outlet port for the flue gas and one power connection port. Figure A.7: gccs blower model configuratio. Specification options The following options will be available in the specification dialog. Isentropic efficiency Outlet pressure or pressure ratio or pressure difference A.2.4 Generator The drum model configuration is represented in figure A.8, with the ports connections. The generator has one control port for power measurement and one array of power port, with the objective of getting all the power generated. Figure A.8: gccs generator model configuration. Specification options The following options will be available in the specification dialog. 1. Total power demand Electrical power Efficiency 2. Specify efficiency Efficiency A.2.5 Recycle breaker The recycle breaker model configuration is represented in figure A.9, with the ports connections. The recycle has one inlet and outlet port and can have any connections type (utilityfluid, power, control, etc). A-7