Parametric Model of Greenhouse Gas Emissions for Fischer-Tropsch Fuels. Final Report

Transcription

1 Parametric Model of Greenhouse Gas Emissions for Fischer-Tropsch Fuels University of Dayton Research Institute Subrecipient Agreement No. FA Prime Cooperative Agreement No. FA Final Report Submitted to Dr. Philip Taylor University of Dayton Research Institute 300 College Park Dayton, Ohio Submitted by David T. Allen and Kirsten S. Rosselot Center for Energy and Environmental Resources The University of Texas at Austin Burnet Road, Bldg 133 (R7100) Austin, TX Jeremiah Miller, John R. Ingham and William Corbett URS Corporation 4141 Col. Glenn Highway Beavercreek, OH October 5,

2 Executive Summary The US Air Force is currently assessing the use of Fischer-Tropsch (FT) fuels, blended with conventional jet fuel, as aircraft fuel. However, a potential barrier to the use of FT fuels is the greenhouse gas emissions associated with their manufacture. Emerging guidelines for procurement of fuels place limits on the life cycle greenhouse gas emissions for fuels purchased by the US federal government. Using methodologies recommended by an Aviation Fuel Life Cycle Assessment Working Group, assembled by the US Air Force, this report describes estimates of life cycle greenhouse gas emissions for FT aviation fuels. FT aviation fuels can be made using a variety of feedstocks (e.g., coal, natural gas, biomass) and using a wide variety of processing configurations. The overall goals of the analyses described in this work were to assess (i) which FT processing parameters, feedstock choices, and greenhouse gas emission estimation assumptions would have the greatest impact on overall cradle-to-gate (total pre-combustion) greenhouse gas emissions, and (ii) whether the variability introduced by processing choices would be significant. To address these two issues, a model capable of estimating cradle-to-gate greenhouse gas emissions for FT fuels under a variety of user defined operating conditions was created. The model was run hundreds of times to generate an ensemble of FT processing scenarios that are described in this report. The model is available to other users for estimating greenhouse gas emissions for additional scenarios not considered in this work, and a user s guide to the operation of the model is included in this report. Overall, the greatest variability in greenhouse gas footprints for the ensemble of FT processing scenarios was found for the following process conditions: sequestration vs. emission of carbon dioxide emissions from syngas production and from the FT reactor the choice of catalyst in the FT reactor the use of biomass as a feedstock rather than coal the carbon number distribution of liquid hydrocarbons from the FT reactor Life cycle assessment methodological choices (i.e., allocation methods) also had a significant impact on the estimated greenhouse gas footprint. These parameters were often coupled, with the sensitivity of process conditions depending on the allocation choices. For example, in addition to the above list, the cradle-to-gate greenhouse gas footprint was sensitive to the fraction of unreacted syngas that is recycled to the FT reactor for some allocation choices. Collectively, these results suggest that potential fuel suppliers will need to document both the details of process configurations and their emission estimation methodologies to obtain precise greenhouse gas emission estimates. The results also suggest that improvements in processing technology (e.g., FT catalysts) could lead to significant reductions in greenhouse gas emissions associated with FT aviation fuel production. 2

3 Contents Executive Summary... 2 Chapter 1: Introduction... 5 Chapter 2: Syngas Manufacture Raw Material Acquisition Gasification Water-Gas Shift Syngas Cleanup Syngas Generation in the Model Chapter 3: Fischer-Tropsch Processing Process Description Reactor Feed Description Reactor Description Product Separation Autothermal Reformer Gas Cooling Carbon Dioxide Removal Recycle Stream Hydrogen Recycle Fuel Gas The Structure of the FT Reactor Model Chapter 4: Wax Upgrading Hydrocracker Model Liquid fuel products Estimating greenhouse gas emissions Chapter 5: Model Operation FT Reactor Model Operation Hydrocracker Model Operation Chapter 6: Ensemble Modeling of Variability Results Process Selections That Influence the Estimated Cradle-to-Gate Greenhouse Gas Footprint of SPK Life Cycle Assessment Selections That Influence the Estimated Cradle-to-Gate Greenhouse Gas Footprint of SPK Appendix A. The base case for electricity generation

4 List of Acronyms and Abbreviations ASU ATR bbl bpd CFR CO 2 E DoD FT GWP HHV IAWG LHV MDEA MMBtu SCF SPK UT VLE WGS Air separation unit Autothermal reformer barrel barrels per day combined feed ratio carbon dioxide equivalent U.S. Department of Defense Fischer-Tropsch Global warming potential Higher heating value Interagency Working Group Lower heating value methyl diethanol amine million Btus standard cubic feet Synthesized paraffinic kerosene University of Texas Vapor-liquid equilibrium Water-gas shift 4

5 Chapter 1: Introduction Figure 1-1 shows a simplified flowsheet of a process that produces liquid fuels from coal and biomass using gasification and Fischer-Tropsh (FT) synthesis. Coal and biomass are gasified to produce syngas (mainly carbon monoxide and hydrogen), which is then cleaned and modified through a water gas shift to achieve a desired CO/hydrogen ratio. The modified syngas is sent through a FT reactor, where hydrocarbon chains are formed. The output of the FT-reactor is sent to a hydrocracker to shorten the hydrocarbon chains that are longer than desired. Fuel gas is produced in the hydrocracker and in the FT reactor. Carbon dioxide is generated during the water-gas shift reaction and removed during gas cleanup, and more carbon dioxide is generated in the FT reactor. particulates H 2 O raw coal/biomass gasification syngas water gas shift fuel gas H 2 O 2 CO 2 fuel gas H 2O adjusted syngas Cl CO 2 H 2 S liquid fuels clean hydrocracker wax FT reactor syngas gas cleanup Figure 1-1. Simplified flowsheet of liquid fuel production from coal and biomass The US Air Force is assessing the use of Fischer-Tropsch (FT) fuels, such as the hydrocarbons formed in the process shown in Figure 1-1, in a variety of aircraft. However, a potential barrier to the use of FT fuels is the greenhouse gas emissions associated with their manufacture. The Energy Independence and Security Act of 2007 prohibits the federal government from purchasing alternative fuels for transportation that have greater greenhouse gas footprints than fuels produced from conventional petroleum sources. Specifically, Section 526 of the Energy Independence and Security Act of 2007 provides that: No Federal agency shall enter into a contract for procurement of an alternative or synthetic fuel, including a fuel produced from nonconventional petroleum sources, for any mobility-related use, other than for research or testing, unless the contract specifies that the lifecycle greenhouse gas emissions associated with the production and combustion of the fuel supplied under the contract must, on an ongoing basis, be less than or equal to such emissions from the equivalent conventional fuel produced from conventional petroleum sources. 5

6 Previous work done for the University of Dayton Research Institute by the University of Texas and URS (Allen et al, 2010), assessed the life cycle greenhouse gas emissions of FT fuels, made from three types of processes: 1) Steam methane reforming followed by FT wax production (cobalt catalyst) and upgrading (US average natural gas as a feed) 2) Coal (Kittanning #6 coal) gasification followed by FT wax production (cobalt catalyst) and upgrading 3) Coal/biomass (mixtures of switchgrass and Kittanning #6 coal) gasification followed by FT wax production (cobalt catalyst) and upgrading These greenhouse gas emission estimates were compared to a petroleum baseline fuel. The results, for greenhouse gas emissions that can be attributed directly to the fuel life cycle, are summarized in Table 1-1. The results are reported as the carbon dioxide equivalents per megajoule of the lower heating value (MJ LHV) of aviation fuel. Greenhouse gas emissions included releases of carbon dioxide, methane and nitrous oxide, but not particulate matter. Impacts of indirect land use were not included; details are available in the final report (Allen et al., 2010). Table 1-1. Global warming potentials (GWPs), expressed as equivalent carbon dioxide emissions per megajoule of lower heating value (g CO 2 E/MJ LHV), for Fischer-Tropsch fuels Life cycle stage Pre-combustion Combustion Total Petroleum baseline 14.3±4 73.2± Natural Gas feedstock 12±10 70±1 82 Coal feedstock 117±10 70± % coal 107±15 67± % switchgrass 85% coal 101±15 64± % switchgrass 75% coal 25% switchgrass 85±20 59±1 144 The results indicated that the greenhouse gas footprints of natural gas-derived FT fuels are less than the petroleum baseline, while coal-derived FT fuels have greenhouse gas footprints that are 110% larger than the petroleum baseline. Mixing biomass with coal had a moderate impact on the greenhouse gas footprint directly attributed to fuel production, decreasing the footprint by about 1% for each 1% of biomass added to the feed. The greenhouse gas emission estimates developed in previous work were based on a small number of processing conditions, however, a wide variety of process configurations are possible for FT fuel synthesis, and these different processing configurations could lead to significantly different greenhouse gas emission estimates. For example, whether or not the carbon dioxide generated during the water-gas shift reactions and in the FT reactor is sequestered has a significant effect on the net greenhouse gas emissions of FT fuels during the cradle-to-gate part of their life-cycle. Other processing parameters, such as choices of FT catalysts and internal recycle ratios, may also be important. 6

7 The goal of the work described in this report is to expand the modeling performed by Allen et al. (2010) by developing parametric models of greenhouse gas emissions of FT fuel processing. The parameters to be considered include a variety of process variables (e.g., fraction of the unreacted syngas recycled to the FT reactor; molecular weight distribution of the waxes produced in the FT reactor; fraction of heavy fuels recycled to the hydrocracker) that would be expected to vary among facilities producing FT fuels. In developing the parametric models, the FT fuel synthesis was divided into three main processing steps: syngas manufacture, FT synthesis, and FT wax upgrading. Chapter 2 of the report describes syngas manufacture and the associated greenhouse gas emissions for coal and biomass feedstocks. Chapter 3 describes the FT process to produce wax and describes the FT reactor model used in this work, and Chapter 4 describes the wax upgrading process for producing aviation fuel (synthesized paraffinic kerosene, or SPK), and describes the upgrading section model. Chapter 5 provides a tutorial on the use of the combined models, and Chapter 6 describes a series of parametric analyses performed with the model. References Allen, D. T.; Murphy, C.; Rosselot, K.S.; Watson, S.; Miller, J.; Ingham, J. and Corbett, W. Characterizing the Greenhouse Gas Footprints of Aviation Fuels from Fischer-Tropsch Processing Final Report to University of Dayton Research Institute from the University of Texas, Agreement No. RSC09006; Account No ; Prime Agreement No. F ; February,

8 Chapter 2: Syngas Manufacture This chapter describes the acquisition of raw materials and the processes for producing syngas used to generate Fischer-Tropsch (FT) fuels. The processes for producing syngas described here are largely drawn from the work previously reported by Allen et al. (2010), while the raw material acquisition values are based on the results of an Interagency Working Group (IAWG) study that studied the greenhouse gas emissions from production of FT aviation fuels (Allen et al, 2011). 2.1 Raw Material Acquisition Raw material acquisition includes all activities required to produce, process and transport raw materials to the FT production facility. The raw materials modeled in this work are coal and switchgrass. 2.1a Coal Acquisition In previous work by the UT and URS team, analyses were performed using Kittanning #6 coal. Previous work by the Interagency Working Group (IAWG) was based on Illinois #6 coal (Allen et al., 2011). The compositions of Kittanning #6 coal and Illinois #6 coal are similar, as shown in Table 2-1. This work assumes the use of Illinois #6 coal as a raw material so that the most recent work of the IAWG can be used in the analyses, and so that the analyses generated in this work are compatible with IAWG data synthesis activities. Table 2-1. Coal composition (weight %) Composition (weight %) Component Illinois #6 Kittanning #6 Moisture 11.12% 2.40% Carbon 63.75% 71.00% Hydrogen 4.50% 5.20% Nitrogen 1.25% 1.30% Chlorine 0.29% 0.40% Sulfur 2.51% 3.00% Ash 9.70% 9.40% Oxygen 6.88% 7.30% Total % % The IAWG analysis reported a total greenhouse gas footprint for coal delivered to the FT production facility of 91 g CO 2 E/kg coal, 78 g CO 2 E/kg coal from mining and 14 g CO 2 E/kg coal from transportation. 2.1b Switchgrass Acquisition In previous work by the UT and URS team, analyses were performed using a switchgrass described by Tarka (2009), which is the same as the composition of the switchgrass used by Allen et al (2011). The composition of the switchgrass is shown Table

9 Table 2-2. Switchgrass composition (weight %) Component Composition (weight %) Moisture 11.12% Carbon 63.75% Hydrogen 4.50% Nitrogen 1.25% Chlorine 0.29% Sulfur 2.51% Ash 9.70% Oxygen 6.88% Total % The life-cycle values include cultivation (including the use of fertilizers), harvesting, and transportation of switchgrass to the FT production facility. Consequential and land use impacts are also included, along with the establishment of the switchgrass plots from pasture or crop lands, the construction of the farming and transportation equipment, and storage of the switchgrass until it is shipped to the FT production facility. The total IAWG greenhouse gas footprint for switchgrass delivered to the FT production facility is -1,500,000 g CO 2 E/ton switchgrass, with 93,000 g CO 2 E/ton switchgrass from switchgrass production, -1,700,000 g CO 2 E/ton switchgrass from carbon dioxide uptake during growth, 110,000 g CO 2 E/ton switchgrass from indirect and direct land use, and 38,000 g CO 2 E/ton switchgrass from transportation. 2.2 Gasification The gasifier data used in this work is drawn from Allen et al. (2010) and, to a lesser extent, from the IAWG report (Allen et al, 2011). The flow diagram for the gasification section of the FT facility is shown in Figure 2-1. The coal delivered to the facility is ground to a size less than 200 mesh, then fed into the gasifier. The switchgrass is processed through a debaler that removes the twine from the bales and begins to break down the bales. The switchgrass is then fed into an initial knife mill to reduce the particle size, followed by a second collision mill to further reduce the particle size to less than 1 mm. The switchgrass must be reduced to this size to ensure reliable feeding to the gasifier. The switchgrass and coal are fed through separate handling systems so that should the switchgrass handling system become plugged the gasifier can continue operation on coal only. The feedstocks are reacted with oxygen from a cryogenic air separation unit (described in Section 2.2a) in a Shell gasifier. The Shell gasifier was selected because it has proven ability to gasify a feedstock containing up to 30% biomass. It is a high temperature (~2500 o F), high pressure (450 psia) entrained flow gasifier. Under these conditions, a high quality syngas is produced that is low in tars, methane and other hydrocarbons. The composition of the syngas produced by the gasifier is affected by the composition of the feedstocks. As discussed before, the composition of coal and switchgrass from the IAWG report are used for gasification analysis. The composition of the syngas at varying concentrations of coal and biomass is shown in Table

10 Figure 2-1. Gasification flow diagram Table 2-3. Syngas composition based on feed composition Mole fraction in syngas Compound 100% Coal 84% Coal 69% Coal H 2 O CO O N CH CO COS H H 2 S HCl Total a Air Separation Unit Oxygen supplied to the gasifier and autothermal reformer (discussed in Section 3.5) is produced in a standard cryogenic air separation unit (ASU). The oxygen stream is 95% oxygen by volume with the remainder being mostly nitrogen. The ASU also produces a nitrogen stream that is 98.7% pure and can be used in other areas of the facility. More detailed information concerning the ASU is available in Allen et al. (2010). 2.3 Water-Gas Shift Iron catalyst FT reactors typically require a syngas feed with a molar H 2 :CO ratio of 1.1:1, while those with a cobalt catalyst typically require a syngas feed with a molar H 2 :CO ratio of 2.1:1. 10

11 The syngas exiting the gasifier has a molar H 2 :CO ratio between 0.49 for the 100% coal case and 0.57 for the 69% coal case. This ratio is adjusted by performing a water-gas shift (WGS) reaction on a portion of the syngas from the gasifier, taking into account that in addition to the syngas from the gasifier, an unreacted syngas stream is recycled to the FT reactor, an autothermal reformer syngas stream enters the reactor, and a hydrogen recycle stream is returned to the reactor. The overall molar H 2 :CO ratio entering the FT reactor is maintained at its desired level by varying the amount of syngas fed into the WGS reactor. The WGS reaction is described by: CO + H 2 O CO 2 + H 2 Any COS present in the syngas stream that undergoes the WGS reaction is hydrolyzed into H 2 S that can be removed during syngas cleanup. The hydrolysis reaction is described by: COS + H 2 O CO 2 + H 2 S Any syngas that does not undergo the WGS reaction is sent to a COS reactor to hydrolyze the COS so that it can be removed during syngas cleanup. The flow diagram for the WGS section is shown in Figure 2-2. COS Reactor Stream # Stream Name 18 Gasifier Syngas WGS Reactor Impurity Removal COS Syngas 20 WGS Syngas Syngas from Reactor Heat Removal 21 COS Cleaned Syngas WGS Adjusted Syngase Steam In 23 Ratio Adjusted Syngas 24 Acid Gas Removal 25 Ratio Adjusted Syngas 2.4 Syngas Cleanup Figure 2-2. WGS section flow diagram FT catalysts are susceptible to catalyst poisoning, and sulfur and chlorine must be removed from the syngas before it is sent to the FT reactor. The COS found in the gasifier syngas has already been converted to H 2 S that is easier to remove from the syngas. Ammonia, sulfur, and chlorine are removed from the syngas using the Rectisol process. This process also removes carbon dioxide from the syngas; for this analysis it is assumed that 96% of the carbon dioxide is removed. The carbon dioxide that is removed was produced in the gasifier, the WGS reactor and the COS reactor. The carbon dioxide stream from the Rectisol process can be sequestered or used for enhanced oil recovery. The sulfur level in syngas exiting the Rectisol process remains high enough to potentially poison the FT catalyst, and the sulfur levels are further reduced using zinc oxide polishing. 11

12 2.5 Syngas Generation in the Model Syngas calculations in the greenhouse gas emission model and their locations within the model s workbook are described in the final section of Chapter 3. References Allen, D. T.; Murphy, C.; Rosselot, K.S.; Watson, S.; Miller, J.; Ingham, J. and Corbett, W. Characterizing the Greenhouse Gas Footprints of Aviation Fuels from Fischer-Tropsch Processing Final Report to University of Dayton Research Institute from the University of Texas, Agreement No. RSC09006; Account No ; Prime Agreement No. F ; February, Allen, D.T., Allport, C., Atkins, K., Choi, D.G., Cooper, J.S., Dilmore, R.M., Draucker, L.C., Eickmann, K.E., Gillen, J.C., Gillette, W., Griffin, W.M., Harrison, W.E.,III, Hileman, J.I., Ingham, J.R., Kimler, F.A.,III, Levy, A., Miller, J., Murphy, C.F., O Donnell, M.J., Pamplin, D., Rosselot, K., Schivley, G., Skone, T.J., Strank, S.M., Stratton, R.W., Taylor, P.H., Thomas, V.M., Wang, M.Q., Zidow T. Life Cycle Greenhouse Gas Analysis of Advanced Jet Propulsion Fuels: Fischer-Tropsch Based SPK-1 Case Study. Final Report from the Aviation Fuel Life Cycle Assessment Working Group to the U.S. Air Force, AFRL-RZ-WP-TR-2010-XXXX, Draft February 4, Tarka, Thomas J. Affordable, Low-Carbon Diesel Fuel from Domestic Coal and Biomass, prepared for the Department of Energy, National Energy Technology Laboratory, 2009 available at: 12

13 3.1 Process Description Chapter 3: Fischer-Tropsch Processing The Fischer-Tropsch (FT) process converts syngas (carbon monoxide and hydrogen) into a wide range of hydrocarbons that are subsequently separated into a liquid products stream (FT wax) and a variety of lighter hydrocarbon byproducts and combustible gases. A process flow diagram of a typical FT processing section is shown in Figure 3-1. The FT reactor, described in Section 3.3, is usually a slurry bed reactor. Syngas (stream 1) produced in the syngas manufacture and cleanup section is combined with recycle streams internal to the FT processing section and fed to the FT reactor. The FT products exiting the reactor (stream 5) are separated into products and recycle streams as described in Section 3.4. FT wax (stream 9) is sent to the wax upgrading section and processed into liquid fuels as discussed in Chapter 4. Light hydrocarbons are partially combusted in an autothermal reformer, as described in Section 3.5, to supplement the syngas from the gasifier. Unreacted syngas is processed to remove water, carbon dioxide and hydrogen. A portion of the unreacted syngas (stream 2) is recycled to the reactor and the remainder of the syngas is assumed to be used as fuel gas to produce energy and/or heat for the FT production facility, or to produce electricity that is exported off-site. 13

14 Stream # Stream Name 1 Ratio Adjusted Syngas 17 2 Unreacted Syngas Recycle 3 ATR Syngas 16 H2 Removal 4 Combined Syngas 5 FT Reactor Output 6 Light HC's ATR Steam 14 8 ATR Oxygen CO2 Removal 9 FT Wax 7 and Unreacted Syngas 11 Condensed Water Unreacted Syngas ATR 13 Recovered CO2 Heat Removal Gas Cooling 14 Unreacted Syngas Unreacted Syngas 16 H2 Recycle Heat Removal 17 Fuel Gas 1 4 Catalyst 5 10 FT Reactors FT Products Separation Separator Heat Removal 9 Catalyst Make-up Catalyst Blow-down Figure 3-1: Fischer-Tropsch processing section flow diagram 14

15 3.2 Reactor Feed Description There are four reactor feed streams. The main feed to the reactor is the clean syngas, produced as described in Chapter 2. There are also three recycle streams internal to the FT processing section. The light hydrocarbons are removed from the FT products during the separation step and sent to an autothermal reformer where they are partially combusted to produce an additional syngas stream (Section 3.5). A portion of the unreacted syngas that exits the reactor is recycled back to the reactor as described in Section 3.8. The final reactor input stream is a hydrogen recycle stream that is described in Section Reactor Description The FT reaction is typically described as: (2n+1) H 2 + n CO C n H (2n+2) + n H 2 O Additional reactions produce carbon dioxide, oxygenated hydrocarbons, and unsaturated products. The production of these byproducts is highly dependent on catalyst selection and operating conditions. The FT reaction is exothermic and requires the removal of heat to maintain the desired reaction temperature. With a slurry-bed FT reactor, heat is removed from the reactor by heat-exchanger tubes immersed in the slurry reactor. The slurry reactor has a high catalyst surface area and a reasonably uniform temperature throughout. A single pass conversion of carbon monoxide to useful hydrocarbon products of 87% and higher have been reported in literature. The FT reactor model developed in previous work (Allen et al., 2010) held the FT wax profile constant. To increase the utility of the model, the ability to vary the composition of the FT wax has been included in the version of the model developed in this work. The distribution of FT products by carbon number is typically described by: W n = n(1- ) 2 n-1 (Equation 3-1) where n is the carbon number, W n is the weight fraction of the products that have a carbon number of n, and is the chain growth probability. The carbon number distribution of the product is controlled by the value, which is highly dependent on the process conditions. Typical values for an iron catalyst are approximately 0.70 to An value of 0.70 would result in the product distribution shown in Figure

16 Mass fraction Carbon number Figure 3-2: Carbon number distribution for value of 0.70 FT reactors with iron-based catalysts produce a mixture of heavy waxes and diesel to naphtha range products. Typical values for a cobalt catalyst are approximately in the 0.80 to 0.90 range. An value of 0.85 would result in the product distribution shown in Figure 3-3. Mass fraction Carbon number Figure 3.3: Carbon number distribution for value of 0.85 The model uses the weight fraction of waxes described by Equation 3-1 to generate the FT product suite. The user can also input a wax distribution and have the model calculate an value that best represents the input distribution. 16

17 3.4 Product Separation The FT products exiting the reactor must be separated and sent to the appropriate areas of the FT production facility. The model handles the separation process by assigning specific molecules or fractions of specific molecules to specific product streams. This does not change based on FT products stream composition or flow rates. The FT products are split into three product streams. Unreacted syngas containing water, carbon dioxide, nitrogen, carbon monoxide, hydrogen, and C1 and C2 hydrocarbons, is separated and send for further processing (Sections ). The light hydrocarbons, containing C3 through C5 hydrocarbons and small amounts of C6 through C8 hydrocarbons, are separated from the FT wax and sent to the autothermal reformer for production of additional syngas. The FT wax, containing the C6+ hydrocarbons, is sent to the upgrading section for further processing, as described in Chapter Autothermal Reformer Light hydrocarbons (C3 to C5 with small amounts of C6 to C8) are fed to an autothermal reformer (ATR). The ATR combines the light hydrocarbons from the FT reactor with an excess of steam and oxygen from an air separation unit to produce additional syngas. The oxygen used in the ATR is provided by the ASU described in Section 2.2a. Autothermal reforming is preferable to steam reforming in this case because the process can be adjusted to give the desired molar H 2 :CO ratio by varying the input of steam and oxygen. The autothermal reforming reaction takes place in a pressure vessel where the oxygen, steam and hydrocarbons are partially combusted (autothermal referring to the partial combustion of the hydrocarbons that provides the required energy for reforming) then fed through a catalyst bed. The ATR modeled for this report is based on the analysis performed by Piña et al (2006). Piña et al studied the autothermal reforming of methane. Methane has a higher H 2 :C ratio (2.0:1) than the C3 to C6 hydrocarbons (1.24:1). The composition of the syngas used in this report has been adjusted to reflect the lower molar H 2 :C ratio. Figure 3-4 shows a simplified schematic of a typical ATR. Figure 3-4: Autothermal reformer (Piña et al, 2006) 17

18 The ATR can be operated at a variety of steam to carbon (S:C or S/C) input ratios. When the steam to carbon ratio is increased, the molar H 2 :CO ratio also increases and the CO:CO 2 ratio decreases. Figures 3-5 and 3-6 show the relationship between the steam to carbon ratio and the molar H 2 :CO and CO:CO 2 ratios used in this work CO:CO2 Ratio S/C Ratio Figure 3-5: CO:CO 2 ratio as a function of steam to carbon ratio H2:CO Ratio S:C Ratio Figure 3-6: Molar H 2 :CO ratio as a function of steam to carbon ratio For FT reactors using a cobalt catalyst, a molar H 2 :CO ratio of 2.1:1 in the combined feed streams to the FT reactor is required. For this report, the conditions in the ATR have been set to produce a syngas that has a 2.1:1 molar H 2 :CO ratio. The amount of gasifier syngas that is fed to 18

19 the WGS reactor is varied to make the overall molar H 2 :CO ratio fed to the FT reactor (including the unreacted syngas, the ATR feed, and the hydrogen recycle) meet reactor requirements. 3.6 Gas Cooling The unreacted syngas is separated from the FT products during products separation. The unreacted syngas contains water, carbon dioxide, nitrogen, carbon monoxide, hydrogen, and C1 and C2 hydrocarbons. The gas cooling results in a syngas that has a temperature less than the boiling point of water. This removes a majority of the water, which originated in the water-gas shift reactor, the FT reactor, or the ATR, from the unreacted syngas before it proceeds to further processing. 3.7 Carbon Dioxide Removal Following gas cooling, carbon dioxide is removed from the unreacted syngas. Carbon dioxide is removed using an methyldiethanol amine (MDEA) unit. For the purposes of this report it was assumed that the MDEA unit was run to capture 96% of the carbon dioxide in the unreacted syngas stream. More detailed information on the MDEA unit is provided in Allen et al. (2010). 3.8 Recycle Stream After the carbon dioxide is removed from the unreacted syngas stream, the main components of the stream are hydrogen and carbon monoxide. A portion of this stream can be recycled to the reactor to augment the syngas produced by gasification. The model allows recycle rates ranging from 0% to 75%. The previous work (Allen et al., 2010) held the unreacted syngas recycle rate constant at 50%. 3.9 Hydrogen Recycle The remaining unreacted syngas undergoes membrane separation to remove the hydrogen before the remainder of the unreacted syngas is combusted for power generation. The membrane separation removes 95% of the hydrogen from the unreacted syngas Fuel Gas The portion of the unreacted syngas that is not recycled is combusted in process heaters or to generate electricity. This stream is composed of nitrogen, carbon monoxide, and C1 and C2 hydrocarbon molecules, with small amounts of water, carbon dioxide, and hydrogen remaining. The carbon dioxide emissions from the combustion of the fuel gas contributes to the overall greenhouse gas footprint of the FT processing section The Structure of the FT Reactor Model This section describes the function of the worksheets in the FT reactor model and the flow of data between the sheets. The FT reactor model is a workbook titled FT Section Model v3.xlsx. Each worksheet in this workbook that models a piece of process equipment includes both a mass 19

20 and atomic balance. The model also contains an overall mass and atomic balance which can be found on the Overall Outputs worksheet. The overall mass and atomic balances account for the FT reactor as well as the WGS reactor and syngas cleanup. The Specifications worksheet is where the user provides variables that the model uses to calculate the results. The main user-defined variables are the H 2 :CO ratio, mass fraction of coal gasified, percent of oxygenated hydrocarbons in the products, percent of olefinic hydrocarbons in the products, reactor single pass carbon monoxide conversion, percent of converted carbon monoxide converted to carbon dioxide and percent unreacted syngas recycled. The user also chooses the composition profile of the wax by inputting either an FT reactor α value in cell B14 or a wax composition in cells F19:F268 of the Specifications worksheet (if wax composition values are provided, the model relies on them, and anything entered in cell B14 is ignored by the model). When the model runs, it first determines if the user has defined a wax. If the user has defined a wax, that wax composition is used as described later in this section; however, if the user has not defined a wax the model uses the α value in cell B14 to generate the wax composition. The user may select the mass fraction of coal gasified from a dropdown list. If the user selects one of the mass fractions populated in the list, the model also populates the syngas composition. If the user selects a user defined mass fraction of coal gasified, then the user must also define the composition of the syngas. If user defined is selected for the mass fraction of coal gasified instead of one of the values in the drop-down, the model will not calculate the greenhouse gas footprint for the acquisition and processing of raw materials. This is because the model in its present form is not capable of performing these calculations. The percent hydrogen removal, percent carbon dioxide removal and required wax production can also be selected by the user in the Specifications worksheet. The allowable wax delta and the allowable H 2 :CO ratio delta are model control inputs and help the model define when convergence has been reached. These two variables should not be changed unless the model will not converge. If any cells on the Specifications worksheet are highlighted in red after the model runs, the model did not converge and (as described in Chapter 5.1) the calculate button must be pressed. If the calculate button has been clicked several times and the model still will not converge, then the allowable wax delta or the allowable H 2 :CO ratio delta may need to be increased. The Results worksheet compiles all the results from all the worksheets into one location. The results are organized in a manner which allows them to be easily accessed and included in reports generated using this model. If any cells on this page other than the cell at the top left are highlighted in red, the model has not converged. The Flow Diagram worksheet contains the flow diagrams used for this model. The Gasification worksheet pulls the mass fraction of coal gasified from the specifications sheet. This is then compared to the three base case scenarios (1.0, 0.84 and 0.69). If the mass 20

21 fraction of coal gasified matches one of the base case scenarios then the sheet calculates the flow rates for coal, biomass and oxygen from the ASU based on the required flow rate of carbon monoxide which is defined on the Overall Inputs worksheet. The Gasification worksheet also includes syngas compositions for the three base cases. The GHG Footprint Data worksheet contains the greenhouse gas footprint data for the production and transportation of the raw materials, land use for switchgrass production, onsite processing of the raw materials, onsite electric and steam footprints, and the fuel gas combustion. The onsite processing of raw materials and the onsite electric and steam footprints are not included in the overall greenhouse gas footprint numbers since it is assumed in this model that the a baseline amount of power production from fuel gas combustion meets the onsite power demands for a particular base case, described in Appendix A. If additional fuel gas is generated, it results in excess electricity generated for export offsite. Likewise, if less fuel gas is generated than in the base case, electricity is purchased from offsite. These columns are highlighted in yellow to alert the user that they are not included in the overall greenhouse gas footprint, but are cataloged for completeness. All greenhouse gas footprint data is then normalized per kg of FT wax produced. The greenhouse gas emissions for methane and nitrous oxide are converted to carbon dioxide equivalents using the IPCC 2007 conversion factors. The Overall Inputs worksheet contains all the inputs for the FT reactor section of the model. There are three worksheets that provide values for the Overall Inputs worksheet: WGS Steam, ATR Steam, and ATR Oxygen. In the Overall Inputs worksheet, the composition of the syngas stream leaving the gasifier is taken from cells A17:B29 on the Specifications worksheet. Flow rates are calculated iteratively. Required wax production is also taken from the Specifications worksheet. The user-input value is compared to the calculated wax production from the FT Products Separation worksheet, and if it is too low, the flow rate of the gasifier syngas is increased. If the wax production is too high, the flow rate of the syngas is decreased. Once this is done, Excel recalculates the entire workbook. The resulting wax flow rate is again compared to the specified wax flow rate. Excel performs iterative calculations until the calculated wax flow rate is equal to the user specified flow rate (within the limits set by the allowable wax delta on the Specifications worksheet). The Overall Outputs worksheet combines information about all the output flow streams of the model in one place. These streams include the FT wax, streams from the acid gas removal process (this is two separate streams which can be found at the bottom of the WGS worksheet), water removal, carbon dioxide removal and fuel gas. The Overall Outputs worksheet also contains overall mass and atomic balances for the FT reactor model. For the mass balance calculations, the inputs are drawn from the Overall Inputs worksheet. The atomic balances may be off by numbers far less than 1. This is not an error with the model, but an artifact of the manner in which Excel performs iterative calculations and determines when convergence has been reached. The WGS worksheet takes the gasifier syngas stream from the Overall Inputs worksheet and divides it into two streams. One stream is sent to the WGS reactor to produce hydrogen as described in Section 2.3. The remainder of the syngas is sent to the carbonyl sulfide reactor. The percentage of gasifier syngas sent to the WGS reactor is controlled by the calculated overall 21

22 H 2 :CO ratio and the required H 2 :CO ratio. The H 2 :CO ratio is calculated on the Combined Syngas worksheet after all reactor feed streams have been combined. If the user-input ratio is higher than actual ratio, the percentage of gasifier syngas sent to the WGS is increased. If the user-input ratio is lower than the calculated ratio, the percentage of gasifier syngas sent to the WGS is decreased. Excel performs iterative calculations until the calculated H 2 :CO ratio is equal to the user input H 2 :CO ratio (within the limits set by the allowable H 2 :CO ratio delta). The WGS reaction is then modeled. The model only simulates the forward WGS reaction and not the reverse reaction. All syngas entering the WGS reactor is assumed to undergo the WGS reaction. The remainder of the syngas is sent to the carbonyl sulfide reactor where the carbonyl sulfide is hydrolyzed. The streams are then combined and impurities (carbon dioxide, hydrogen chloride, and hydrogen sulfide) are removed from the syngas stream. The Combined Syngas Input worksheet combines all streams entering the FT reactor. The cleaned syngas from the WGS worksheet, the ATR syngas from the ATR worksheet, the unreacted syngas recycle from the Syngas Recycle Split worksheet and the hydrogen recycle from the H2 Removal worksheet together comprise the feed for the FT reactor. The Input Wax worksheet is only used if the user has defined a user defined wax composition on the Specifications worksheet. This worksheet takes the user defined wax from the Specifications worksheet and determines the maximum mass fraction value. The wax is then truncated so that only carbon numbers equal to or higher than the maximum mass fraction value are used to determine the α value of the user defined wax. The wax is truncated because the lighter ends of the wax where the mass fractions are increasing may have had portions removed by the separation processes and therefore would not be representative of the actual FT products produced by the reactor. The truncated wax then has the mass fractions normalized so the total mass fraction for that portion of the wax is one. The Alpha Calculation worksheet is also only used if the user has defined a user defined wax composition on the Specifications worksheet. This worksheet contains the wax composition for every α value from 0.01 to 1 (increasing by intervals of 0.01). The mass fractions for the waxes from each α value are truncated and normalized to match the truncated and normalized user-defined wax developed on the Input Wax worksheet. The absolute value of the difference between the truncated user defined wax and the truncated alpha value wax is determined for every carbon number for each α value. The difference for each carbon number is then totaled and becomes the delta between that α value and the user-defined wax. The Alpha Value worksheet is another worksheet that is only used if a user defined wax composition was selected on the Specifications worksheet. The Alpha Value worksheet draws the delta values for each α value from the Alpha Calculation worksheet and determines which α value has the minimum associated delta, or the best fit α value. The best fit α value is then used to generate the FT products that are used for the rest of the model. In the Reactor Calculations worksheet, if the user has defined a user defined wax, then the α value from the Alpha Value worksheet is used to generate the FT products. The calculated α value is used because most wax compositions do not contain any information on the light ends of the product which have a large impact on the overall greenhouse gas footprint. If the user has 22

23 chosen to enter an α value instead of a user-defined wax, then the chosen α value is used to generate the FT products. The mass fractions of the FT products are then normalized so the total mass fraction is one. The average molecular weights from the Molecular Weights worksheet are used to determine the mole fraction of the FT products. The number of moles of carbon monoxide entering the FT reactor is taken from the FT Reactor worksheet and the number of moles of carbon monoxide converted to FT products is determined. The molar flow of each carbon number in the FT products is calculated based on the mole fractions of the FT products and the moles of carbon monoxide converted in the FT reactor. The number of moles of hydrogen, oxygen and carbon in the FT products are then determined so that overall mass and atomic balances can be performed on the FT Reactor worksheet. The percentages of oxygenates and olefins are included in these calculations. The numbers for the molar percentages of oxygenates and olefins for each carbon number are drawn from the Molecular Weights worksheet. The FT Reactor worksheet calculates the composition and flow rate of the entire product suite from the FT reactor. The syngas input stream is drawn from the Combined Syngas Input worksheet. The FT products and their molar flow rates are drawn from the Reactor Calculations worksheet. The molar flow of carbon monoxide is determined by multiplying the input molar flow of carbon monoxide by one less the FT reactor single pass carbon monoxide conversion given in the Specifications worksheet. The molar flow of carbon dioxide is determined by multiplying the input molar flow of carbon monoxide by the FT reactor single pass carbon monoxide conversion and the percentage of converted carbon monoxide that is converted to carbon dioxide. Nitrogen, methane, and ethane pass through the reactor unchanged. The molar flow of water is controlled to balance the oxygen mole balance for the reactor. The flow rate of hydrogen is controlled so that an atomic balance for hydrogen is obtained. The lower heating value (LHV) for the FT products (C1+) is calculated based on the molar flow and the LHV calculations on the Heating Values worksheet. The FT Products Separation worksheet separates the products into three streams. The unreacted syngas stream contains all non-hydrocarbons and the C1 and C2 hydrocarbons. This stream is sent to the H2O Removal worksheet for further processing. The light hydrocarbon stream contains all C3 to C5 hydrocarbons and a portion of the C6 to C8 hydrocarbons. This stream is sent to the ATR worksheet for syngas production. The FT wax stream contains any C6 to C8 hydrocarbons that are not contained in the light hydrocarbon stream and all C9+ molecules. This stream is sent to the upgrading section, which is described in Chapter 4. The ATR worksheet converts the light hydrocarbon stream into a syngas stream that can be recycled to the Combined Syngas Input worksheet. The light hydrocarbon stream is drawn from the FT Products Separation worksheet. This model locks the output H 2 :CO ratio from the ATR at 2.1. Based on this number, the steam to carbon ratio (S/C) in the ATR is 1.17, the oxygen to carbon ratio (O 2 /C) is and the carbon monoxide to carbon dioxide ratio (CO/CO 2 ) is There is excess steam in the ATR, and the excess steam passes through the ATR without reacting. This process completely converts all hydrocarbons that enter the ATR. In the H2O Removal worksheet, the unreacted syngas stream from the FT Products Separation worksheet is cooled to below the boiling point of water. The unreacted syngas 23

24 leaving the H2O Removal worksheet is saturated with water vapor, but a majority of the water has been removed from the unreacted syngas stream. The unreacted syngas is then sent to the CO2 Removal worksheet. The amount of carbon dioxide removed is based on the user-input value in the Specifications worksheet. The remaining unreacted syngas is sent to the Syngas Recycle Split worksheet. The Syngas Recycle Split worksheet draws the unreacted syngas from the CO2 Removal worksheet. The unreacted syngas stream is split into two streams: a recycle stream that is sent back to the Combined Syngas Input worksheet and an unreacted syngas stream that is sent to the H2 Removal worksheet for further processing. The percentage of the incoming syngas that is recycled to the reactor is selected by the user on the Specifications worksheet. The H2 Removal worksheet removes the hydrogen from the unreacted syngas stream. The percent of hydrogen removal is selected by the user on the Specifications worksheet. The hydrogen that is removed is recycled to the Combined Syngas Input worksheet. The Molecular Weights worksheet converts the mass percentages of paraffinic, olefinic and oxygenated products into mole fractions for C3+ hydrocarbons based on the user defined inputs from the Specifications worksheet. The Heating Values worksheet calculates the HHV and LHV using the group contribution method and also calculates the LHV using the Hileman method (Hileman et al, 2010). The Hileman method values are the values used for LHV in this model. References Allen, D. T.; Murphy, C.; Rosselot, K.S.; Watson, S.; Miller, J.; Ingham, J. and Corbett, W. Characterizing the Greenhouse Gas Footprints of Aviation Fuels from Fischer-Tropsch Processing Final Report to University of Dayton Research Institute from the University of Texas, Agreement No. RSC09006; Account No ; Prime Agreement No. F ; February, Hileman, J, R Stratton, P Donohoo. Energy content and alternative jet fuel viability. Journal of Propulsion and Power vol.26 no.6 ( ) doi: / Piña, Juliana; Bucalá, Verónica; and Borio, Daniel Oscar (2006) "Optimization of Steam Reformers: Heat Flux Distribution and Carbon Formation," International Journal of Chemical Reactor Engineering: Vol. 1: A25. 24

25 Chapter 4: Wax Upgrading The liquid hydrocarbon output (wax) generated during Fischer-Tropsch (FT) processing generally contains hydrocarbons that are too large to be used as jet fuel and that are made up almost exclusively of saturated, straight chain compounds. The wax must be upgraded to yield synthesized paraffinic kerosene (SPK) that can be blended with jet fuel from conventional refineries and used in aircraft. The basic process configuration for the upgrading section, shown conceptually in Figure 4-1, mixes the wax with an excess of hydrogen and feeds it to a hydrocracker. During hydrocracking, the feed is contacted with a catalyst at elevated temperature (typically >300 C) and pressure (typically >50 atm), creating lower molecular weight and more extensively branched compounds. Some of the hydrogen fed to the hydrocracker is consumed; the excess is recycled to the hydrocracker. The products of hydrocracking, which include fuel gas, naphtha, SPK and diesel, are separated using distillation. Some of the liquid output from the FT reactor may already be in the SPK distillation range and may be routed around the hydrocracker and sent straight to separation. This stream is labeled preseparated wax in Figure 4-1. The wax that goes directly to liquid fuel products without going through the hydrocracker is also called straight-run product. Hydrocrackers are generally operated in recycle mode so that material that is too heavy to be included in the hydrocracker product cuts (labeled heavy ends in Figure 4-1) is recycled back to the hydrocracker for further processing. If production of diesel is not desired, the diesel distillation cut is included in the heavy ends to be recycled. In a no-recycle configuration, a residual stream is generated that includes hydrocracker output too heavy to be included in the liquid product streams. makeup hydrogen water Compressor fuel gas C1-C4 naphtha wax from FT column Wax Hydrocracking High Pressure Separation Separation diesel SPK (to blending) recycle hydrogen heavy ends light ends from wax 4.1 Hydrocracker Model Figure 4-1. Wax upgrading unit 25

26 The model developed for this work predicts the output generated by hydrocracking a specified wax (see Chapter 3). The user may either provide a carbon number distribution for the wax being fed to the hydrocracker, as well as the amount of oxygenates and olefinic compounds present in the wax, if any, or accept the wax output of the FT model described in Chapter 3 as wax input to the hydrocracker model. (Note that cradle-to-gate greenhouse gas emissions of the SPK produced in the upgrading section are not estimated by the model unless the wax output of the FT model is used as the wax input to the hydrocracker model.) The user must also specify the fraction of bonds broken during hydrocracking, whether or not recycle of heavy ends is desired, and what degree and type of isomerization is expected in the reactor. Detailed information about running the model is contained in Chapter 5. The hydrocracker for upgrading FT wax to SPK and other liquid fuels can play a significant role in determining the overall greenhouse gas emissions of the process for making liquid fuels from natural gas, coal, and biomass because of its role in determining product slate and, to a lesser extent, because of the hydrogen consumption in the reactor. The model estimates the effect of wax feed characteristics on hydrocracker output and the impact of hydrocracker operation choices on both product slate and hydrogen consumption. During hydrocracking, bonds are saturated and compounds are hydrogenated so that nearly all the products of the hydrocracker are saturated alkanes (in some hydrocrackers, naphthenic compounds -- ringed alkanes -- are also produced). Also, molecules are broken (cracked) into smaller molecules, and straight-chain molecules can be isomerized to form branched molecules. The model developed in this work relies on well-known rules for the likelihood of a bond breaking during hydrocracking (Shah et al., 1988). In a hydrocracker, the two outermost carbonto-carbon bonds at both ends of a molecule are unlikely to break. If the next innermost bond is broken, propane is created. This bond has half the likelihood of all the other bonds in the molecule of breaking, and the likelihood of the other bonds breaking is all the same. For example, consider a molecule of n-nonane, which has nine carbons and eight carbon-to-carbon bonds: C - C - C C C C C C C As modeled, bonds 1, 2, 7, and 8 are unable to break. Bonds 3 and 6 have half the likelihood of breaking as bonds 4 and 5, and bonds 4 and 5 have an equal likelihood of breaking. The hydrocracker model developed in previous work (Agreement Number RSC09006; Account No , Prime Agreement No. F ) assumed that hydrocracking occurs in a two-phase (catalyst and a liquid reactant/product mixture) reactor. Most industrial, and even laboratory scale, reactors exhibit three-phase behavior (catalyst, low molecular weight material in a vapor phase, and a liquid phase containing higher molecular weight material). The distinctive feature of a three-phase reactor is that the material in the vapor phase no longer has access to the catalyst, which is assumed to be coated by liquid. This reduces the creation of very light ends in the hydrocracker, since low carbon number material enters the vapor phase and escapes the reactor without further reaction. 26

27 The modified hydrocracking model described here accounts for an unreactive vapor phase. It does this by connecting a series of small reactors in series. Vapor phase compounds from each subreactor are prevented from reacting in subsequent reactors. The subreactors are pictured conceptually in Figure 4-2. wax feed hydrocracker output products a) The no-recycle case. The hydrocracker is treated as a series of modeled sub-reactors with some fraction of bonds broken in each reactor. As the feed moves down the chain, more and more small molecules are created that enter the vapor phase and are no longer available for cracking in subsequent sub-reactors. The sub-reactor that the product output exits from is determined by the user input fraction of total bonds broken and must generally be interpolated between the output of two of the reactors in the chain. wax feed hydrocracker output recycle products b) The recycle case. As with the no-recycle case, the hydrocracker is treated as a series of sub-reactors with the same fraction of bonds broken in each reactor. As the feed moves down the chain, more and more small molecules are created that enter the vapor phase and are no longer available for cracking in subsequent sub-reactors. The subreactor that the product output exits from is determined by the user input fraction of total bonds broken and must generally be interpolated between the output of two of the reactors in the chain. The recycle stream is routed back and fed to the first reactor. The calculations must be iterated until the composition of the feed to the first reactor (recycle stream and wax stream combined) converges. Figure 4-2. Modeled hydrocracker subreactors The vapor phase is represented by species with user-selected carbon numbers less than a threshold value. In an actual hydrocracker, this threshold carbon number depends on the temperature and pressure in the hydrocracker, but predicting the behavior of the lighter compounds in the mixture is complex due to the isomers present in the mix. A single threshold carbon number based on reactor temperature and pressure is unlikely to be realistic. Because of isomerization, there is a distribution of threshold carbon numbers, each with different isomerization patterns, that would be found in the vapor phase in the reactor. A distribution 27

28 around the threshold carbon number, rather than a sharp carbon number cutpoint, is employed by the model. In each sub-reactor, a small fraction of the total bonds that are available for breaking is broken. Based on the probability rules as applied to the model, this excludes all of the two outermost bonds in each molecule and half of the third innermost bond in each molecule. The number of bonds broken by carbon number depends on both the number of available bonds at that carbon number and the amount of material at that carbon number that is present. Because of the probability rules described earlier, the largest molecules are consumed as the reactions proceed. 4.1a Probability calculations in the model The bond breaks for each subreactor are calculated in the calculate the breaks worksheet of the model. This worksheet is a matrix with the number of moles entering the subreactor for each carbon number contained in cells B10 to B257 and in cells E5 to IR5. The number of moles is based on 100 moles being fed to subreactor 1 and grows with each subreactor. The theoretical number of breakable bonds per molecule (row 2) and the vapor-liquid equilibrium effect on bond breakage (row 3) combine to give a value for the breakable bonds at each carbon number (row 4). Cell B5 calculates the fraction of total bonds that are broken in this subreactor, and cell B3 calculates the fraction of breakable bonds broken in the subreactor. The average number of breakable bonds per molecule, based on values calculated in row 1, is calculated in cell B2, and cell B1 has the terminal carbon number found in the wax (this terminal carbon number is the largest carbon number with at least mol % of the wax). Cells C8 to H8 have the carbon number range across which the vapor-liquid equilibrium takes effect (advanced users may wish to alter this range). The main purpose of this worksheet is to calculate the values in cells E10 to IR257. For each carbon number input, the same fraction of breakable bonds are broken and new smaller species are created, following the probability rules for bond breakage described earlier. Because the fraction of bonds broken in each subreactor is deliberately small, there is generally some unbroken material at each carbon number, except at the high end of the carbon numbers. The amount of unbroken material in each subreactor is assumed not to isomerize and must be tracked. This is done in rows 261 to 511 of the calculate the breaks worskheet. Column IS contains the sum of all the products of breaking at each carbon number this becomes the input for the next subreactor in the series. The model includes macros that copy this column into column B and into a column on the find break-fraction combo worksheet. When the macro is run, the output for subreactor 1 from column IS in the calculate the breaks worksheet is copied into column C of the find break-fraction combo worksheet, then the output of subreactor 2 is copied into column D, and so on. The input to subreactor 1 is copied into column B of the find break-fraction combo worksheet. The user-selected fraction of bonds broken occurs somewhere in between two of the subreactors, and once the macro has finished running and the results of all 29 subreactors are copied into the find break-fraction combo worksheet, the final results are interpolated between the two subreactors where the user-selected value of fraction of bonds broken lies. This occurs in cells A512 to AG763 of the find break-fraction combo worksheet. The totals for this section, less 28

29 unbroken material, are found in column AB. These results are used in the next part of the model, which assigns the hydrocracker output to product cuts. Cells A262 to AC511 of the find break-fraction combo worksheet continue to track the unbroken material, which does not isomerize. The totals for this section, found in column AC, are also used in the next part of the model, which assigns hydrocracker output to product cuts. In order to relate fraction of bonds actually broken to total bonds broken, the moles of bonds in each subreactor for each carbon number are calculated in cells AL11 to BO259 of the find break-fraction combo worksheet. 4.1b Modeled carbon number characterization compared to experimental results The carbon number distribution of the output predicted by the model has been compared to experimental results from a number of researchers. In order to generate these comparisons, the modeled fraction of bonds broken and the threshold carbon number was selected to best fit the experimental data. Figure 4-3 shows a comparison between modeled results and experimental results for hydrocracking a stream containing 88% C16 and 12% C17 (Sie et al, 1991). Figure 4-4 gives a comparison between modeled results and experimental results for hydrocracking a SMDS FT wax under high and medium severity conditions (Sie et al, 1991). Figure 4-5 gives a comparison between modeled and experimental results for hydrocracking the wax produced in an iron slurry FT reactor (Leckel et al, 2005) mole percent modeled results (VLE C# 15, of bonds broken) experimental data (Sie et al, 1991 Figure 10) carbon number Figure 4-3: Modeled and experimental carbon number distributions from hydrocracking a stream that is 88% C16 and 12% C17 29

30 10 wax feed mass percent carbon number modeled values medium severity case (no recycle; of bonds broken, VLE C#=18) modeled values high severity case (no recycle, of bonds broken, VLE C#=18) experimental medium severity products (SMDS medium) (Sie et al, 1991, Figure 12) experimental high severity products (SMDS heavy) (Sie et al, 1991, Figure 12) Figure 4-4. Modeled and experimental carbon number distributions from hydrocracking an SMDS Fischer-Tropsch wax mass percent carbon number wax feed (Leckel, 2005, Figure 3) modeled values, 7 MPa case (no recycle, VLE C# 29, of bonds broken) experimental values, 7 MPa case (Leckel, 2005, Figure 3) Figure 4-5. Modeled and experimental carbon number distributions from hydrocracking an iron slurry Fischer-Tropsch wax 4.2 Liquid fuel products Cracking is not the only reaction that takes place in a hydrocracker. Isomerization also occurs. Isomerization does not consume hydrogen, but it does have an effect on product slate, as boiling point decreases with increased branching. Isomerization also influences other important fuel 30

31 characteristics, such as density, heat of combustion, and combustion indices such as octane number. 4.2a Estimating volumes of liquid fuels produced Distillation cut points and other product properties are generally dependent on isomers present, not just on the carbon number of the product. The model relies on formulas that are dependent on carbon number for developing a set of representative isomers at each carbon number. Users must specify the extent to which isomerization is occurring (the fraction of carbons that are branched in the most commonly branched isomers that are produced) and whether creation of naphthenes is allowed. Shah et al (1988, Appendix C) data for low carbon number hydrocracker output were used to get an understanding of the upper bound of potential branching and creation of naphthenic compounds. (The extent and types of isomerization that occur in a reactor depend not just on reactor conditions but also on the type of catalyst used in the reactor.) As with the two-phase model created for the earlier work, compounds that aren t cracked in the reactor are now isolated before the isomerization step because it is assumed that if they do not contact the catalyst, they do not isomerize. Boiling points, which determine distillation cuts, are based on an adaptation of the Stein and Brown (1994) estimation method. The equations are (T b in K): T b (degrees K) = Σ( n i g i ) T b (corr) = T b T b (T b ) 2 [T b <= 700 K] T b (corr) = T b T b [T b > 700 K] boiling point coefficients: -CH CH 2 - straight >CH- straight CH 2 - ring >CH- ring b Liquid fuel calculations in the model The worksheet assign to product cuts is where the modeled result is isomerized, boiling point values are assigned, and material is placed into product cuts based on boiling point. There are five regimes in this sheet, each with carbon numbers 3 to 33: the regime at which the most common degree of branching is observed (columns D to AH), a regime where half as much branching as the most common degree of branching is observed (columns AI to BM), a regime where 1.5X as much branching as the most common degree of branching is observed (columns BN to CR), a regime for ringed compounds (columns CS to DW) and a regime for material that is not branched because it has not contacted the catalyst in the hydrocracker (columns DX to FB). All isomers of C34+ are combined in column FC. The group contribution value for boiling point is given in row 7, and the product cut that the material falls into are coded in row 8. Row 9 helps calculate the product cut values of row 8. The codes for the distillation cuts are given in Table 4-1. The SPK distillation cut is broken into seven sub-cuts that correspond to the 2008 military specifications for SPK. These specifications have changed, but the framework has been left in place in case future specifications resemble the 2008 specifications again. Advanced users 31

32 can alter the military specification temperatures by changing cells A6 to A11 in the model control worksheet. Group contribution equations and values for calculating the boiling point can be found in column B of the group cont values & equns worksheet. Table 4-1. Codes for product distillation cuts in the hydrocracker model Code Cut z fuel gas (C3 to C4) a naphtha (C5 to user-selected temperature #1) b, c, d, e SPK (>user-selected temperature #1 to 205 C) f, g, h SPK (>205 C to user-selected temperature #2) i diesel (>user-selected temperature #2 to user-selected temperature #3 j residual/recycle (>user-selected temperature #3) Cells D10 to FC257 of the assign to product cuts worksheet contain the number of moles of material at each boiling point and product cut by carbon number. Columns FD to FO sum the moles of product cuts for all isomers by carbon number, and columns FP to GA convert the moles to mass. The weight percent of the products is given in columns GB to GM. Total feed for subreactor 1 is calculated in column GT. When the model is operated in recycle mode, the amount of material to be recycled is not known until all the bonds have been broken. This amount is mixed with input wax and fed to subreactor 1, which changes the amount of material to be recycled. After calculating the recycle stream repeatedly, the profile of the material to be recycled stops changing with each modeled run. The model treats recycled material as if it were uncracked. However, this should not alter the overall results because that material isomerizes before it exits the hydrocracker as product. 4.2c Military specifications for SPK distillation cuts The temperature cutoffs for SPK-fraction distillation cuts are specified by the user, and can be adjusted so that the maximum amount of SPK product that meets military specifications is produced. The military specification (DoD, 2010) for the final boiling point of SPK is a maximum of 300 C, and there is no initial boiling point. Instead, military specifications require that at least 10% of SPK be recovered at 205 C and that the temperature at which 90% of the SPK product boils is at least 22 C higher than the temperature at which 10% of the SPK product boils. There are military specifications for SPK other than distillation cut fractions. These include maximum allowed aromatics, sulfur, freezing point, viscosity, naphthalenes, thermal stability, particulate matter, and filtration time. There are also allowed ranges for flash point temperature range, API, and electrical conductivity, and minimum required values for heat of combustion and water separation. Other than distillation cuts, the only properties that were estimated in this study were density and heat of combustion, as described below. 32

33 4.2d Estimating density of liquid fuel products The model results can be used to estimate the density of the modeled liquid fuel hydrocracker products. Density of the product streams is calculated using a relationship based on their hydrogen mass fraction (Hileman et al, 2010). This relationship, contained in column G of the group cont values & equns worksheet, is density (g/ml) hydrogen mass fraction Hydrogen mass fraction for each product stream was calculated using the average molecular weight generated by the model. Military specifications require the density of SPK to be between and kg/l at 15 C (DoD, 2010). Modeled densities for SPK are in this range only in the 250 C to 300 C range, which, at least when the group contribution methods described here are used to predict density and boiling points, means that both the density specification and the distillation temperature specifications simultaneously is precluded. 4.2e Estimating heat of combustion of liquid fuel products The model results can be used to estimate the heat of combustion of the modeled hydrocracker product streams. The net heat of combustion of the product streams is calculated using a relationship based on their hydrogen mass fraction (Hileman et al, 2010). This relationship, found in column H of the group cont values & equns worksheet, is LHV (MJ/kg) hydrogen mass fraction As with density, hydrogen mass fraction for each product stream was calculated using the average molecular weight generated by the model. In cases where the gross heat of combustion was desired, it was calculated from the net heat of combustion using the relationship LHV HHV hydrogen mass fraction This relationship is given in column I of the group cont values & equns worksheet. Military specifications require the net heat of combustion of SPK to be 42.8 MJ/kg or greater (DoD, 2010). 4.2f Product property calculations in the model The worksheet assign to product cuts is also where the estimated product properties are calculated. The formulas and values used to create the estimates are in the group cont values and equns worksheet. By carbon number, the number of atoms of each species of hydrocracker output are given starting in column A of the assign to product cuts worksheet in row 268, the molecular weight in row 269, the mass fraction of hydrogen in row 270, and the mass of carbon and hydrogen in products resulting from 100 moles of feed to the hydrocracker in rows 271 and 272. Boiling point formulas are in rows 274 to 281. Rows 274 to 321 contain acentric factor, critical pressure, and critical temperature estimates that ultimately were not used in the modeling because they were not providing reasonable results for the modeled conditions. Mass percent by carbon number is calculated in rows 323 to 324, density in row 326, and lower heating value in 33

34 row 328. Beginning in column FD, estimated property values by product cut rather than carbon number are given. Row 269 gives molecular weight, row 272 gives hydrogen mass fraction, row 273 gives density, row 274 gives lower heating value, and row 275 gives higher heating value. Rows 332 to 968 calculate the temperatures at which 10% by mass and 90% by mass of the SPK is recovered (this is to test that the military specifications for distillation gradient is met). Estimated wax properties (column BX to CD) and the properties of the straight run cut (the preseparated wax) (columns AE to BV) are given in the wax input worksheet. 4.3 Estimating greenhouse gas emissions Greenhouse gas emissions assigned to the SPK stream from the upgrading process depend on utility demands for heat and energy for separators and compressors, as well as the amount of hydrogen consumed in the hydrocracker and the relative volumes of co-products produced. In this analysis, it was assumed that the greenhouse gas footprint of constructing the upgrading section were insignificant. 4.3a Stoichiometric hydrogen demand during hydrocracking A simple means of estimating hydrogen consumption in the hydrocracker is to calculate the stoichiometric hydrogen demand the hydrogen required to saturate and hydrogenate all the molecules and to take the place of carbon bonds when molecules are broken. Stoichiometric hydrogen demand can be estimated if the average molecular weight of the wax feed and the average molecular weight of the hydrocracker output are known, along with the concentration of olefinic bonds and oxygenated compounds. For a given feed, the smaller the molecular weight of the output, the higher the stoichiometric hydrogen demand. The fraction of oxygenated molecules and olefinic bonds in the wax are important to hydrogen consumption if they are present in significant quantities. Modeled values were used to calculate the stoichiometric hydrogen demand in the hydrocracker and from that, an estimate of the greenhouse gas emissions from hydrogen consumed during hydrocracking. This stoichiometric hydrogen consumption is equal to the hydrogen required to crack the molecules entering the hydrocracker, the hydrogen required to consume the oxygen in the oxygenated compounds entering the hydrocracker, and the hydrogen required to saturate the olefinic bonds entering the hydrocracker. The model is configured such that the user selects the amount of SPK produced. This determines the amount of wax required. In order to calculate the hydrogen demand, the volume of feed to the hydrocracker had to be calculated. For the recycle case, this is the wax feed plus the recycle stream. The wax feed rate was calculated by performing a carbon balance around the hydrocracker. The equation for the carbon balance is 34

35 carbon in wax input carbon in fuel gas carbon in naphtha carbon in SPK carbon in residual m W NW MWW m F m N m S m D m R NF NN NS ND NR MWF MWN MWS MWD MWR In this equation, N is the average number of carbon atoms per molecule, m is the mass flow rate, and MW is molecular weight. The subscripts W, F, N, S, D, and R stand for wax input, fuel gas, naphtha, SPK, diesel, and recycle, respectively. For a saturated alkane, N is equal to (MW - 2) N 14 The oxygen and olefin concentrations in the input to the hydrocracker are not large enough to change the molecular weight of the feed, so a value for the molecular weight that assumes saturated alkanes gives reasonable results. Substituting and rearranging gives m W MWW 2 Carbon input MWW 14 m F MWF 2 m N MWN 2 MWF 14 MWN 14 m S MWS 2 m D MWD 2 MWS 14 MWD 14 m R MWR 2 MWR 14 m W m F MWF 2 m N MWN 2 m S MWS 2 m D MWD 2 m R MWR 2 MW F 14 MW N 14 MW S 14 MW D 14 MW R 14 MW MWW 2 14 Per mol of hydrocracker feed (recycle plus wax), the moles of hydrogen molecules required for cracking is equal to N I 1 mol H 2 for cracking number of broken bonds N O 1 mol H2 1 mol H 2 mol input mol input mol input In this equation, N (the average number of carbons per molecule) is derived from the molecular weights of the hydrocracker input and output that were calculated by the model. The subscript O designates output (including the portion of the output that is recycled) and the subscript I designates input to the hydrocracker. Per mole of wax feed, the moles of hydrogen molecules required to consume the oxygen in oxygenated compounds and to saturate the bonds in the olefinic compounds is given by W 35

36 and mol H 2 mol H 2 to oxygenate 2 mol wax mol wax mol H 1 mol H 2 to saturate 2 mol wax mol wax f oxygenated f olefinic In these equations, f is the fraction of molecules in the wax feed that are olefinic or oxygenated. The recycle stream is assumed to be 100% saturated alkanes, so only the wax portion of the total feed to the hydrocracker has oxygenated compounds, not the recycle portion. That portion is mols of wax mol fraction of input that is wax mols of wax mols of recycle Molar consumption can be converted to mass consumption by applying the molecular weights. Greenhouse gas emission factors of kg methane/scf H 2, kg nitrous oxide/scf H 2, and kg carbon dioxide/scf H 2 ( kg CO 2 E/SCF H 2 ) were applied in order to estimate the greenhouse gas emissions for the modeled stoichiometric hydrogen demand. These values were taken from Table 4-28 of Skone and Gerdes (2008), and are intended to reflect average cradle-to-gate emissions from hydrogen production at refineries nationwide, assuming hydrogen is produced at a modern steam methane reforming plant using natural gas as the feed, with subsequent hydrogen purification by pressure swing adsorption. 4.3b Combustion of fuel gas for heat and electricity generation Fuel gas is produced in both the FT reactor and in the hydrocracker, with the fuel gas produced in the FT reactor of lesser quality. In every modeled scenario, it is assumed that all of the fuel gas produced is burned either in process heaters or to generate electricity. Carbon dioxide emissions from burning the fuel gas are estimated based on the carbon species present in the fuel gas and on the volume of fuel gas produced, with complete combustion assumed. A base case for all processes within the facility fenceline, including switchgrass and coal processing, gasification, fuel gas purification and water gas shift reactions, the FT process, and the upgrading section was developed that produced just enough fuel gas in the hydrocracker and the FT reactor to meet the energy needs of the entire process. In this base case, described in more detail in Appendix A, 69% of the feed to the gasifier is coal and the hydrocracker is operated in recycle mode with only naphtha and diesel produced. The FT reactor has an iron catalyst. The ratio of the combined higher heat content of the fuel gas produced in the FT reactor and in the upgrading section to the heat content of the wax in this base case is If run conditions result in a ratio higher than this, then a credit can be taken for excess electricity generation, and if run conditions result in a ratio lower than this, electricity must be purchased. In other words, if process configurations result in less total energy from combustion of fuel gas produced than is produced in the base case on a per unit of energy wax produced and upgraded, it is assumed that no other significant overall process demands for process heat or electricity will occur and additional electricity demands will be met by purchasing electricity. National average values of greenhouse gas emissions for electricity are applied: kg methane/kwh, kg nitrous oxide/kwh, and kg carbon dioxide/kwh (Skone and Gerdes, 2008). Because of the sensible heat made available during the processes, steam can be generated 36

37 without burning fuel gas and the fuel gas is used only to superheat the steam before generating electricity. Thus, the efficiency of electrical generation is high, at 83%. If process configurations result in more fuel gas being produced than is produced in the base case, it is assumed that the excess fuel gas will be burned to produce excess electricity and a displacement credit based on the national greenhouse gas emission estimate for that excess electricity is applied. As with the case for less fuel gas being produced than the base case, the excess fuel gas is needed only to superheat steam for electricity generation and the efficiency is high. 4.3c Displacement credits for naphtha produced The ASTM boiling range of light straight-run gasoline is F (Skone and Gerdes, 2008, Table 4-1). The naphtha produced in the hydrocracker is similar to straight run gasoline because it is paraffinic and because of its distillation temperature range. This type of naphtha is usually used as feedstock for production of olefins. A credit can be taken for the cradle-to-gate greenhouse gas emissions that would have been released due to production of this naphtha stream at a conventional refinery. In an analysis of the greenhouse gas emissions from conventional refineries (Skone and Gerdes, 2008), special naphthas and petrochemical feedstocks are grouped with the "light ends" streams at a refinery, along with still gas and LPG. The cradle-to-gate greenhouse gas emissions of all light ends are estimated as being equal on a per barrel basis, at kg methane per barrel, kg nitrous oxide per barrel, and 60.9 kg carbon dioxide per barrel (74 kg CO 2 E/bbl). For naphtha less than 401 F and special naphthas, the HHV is MMBtu/bbl (Skone and Gerdes, 2008, Table 4-18). The HHV of the modeled naphtha stream (which varies depending on the modeled scenario) determines the credit assigned to naphtha production. naphtha credit kg CO 2 eq 74 HHV of modeled naphtha bbl naphtha MMBtu HHV MJ bbl naphtha MMBtu 4.3d Displacement credit for diesel produced A nationally representative estimate of the cradle-to-gate greenhouse gas emissions from diesel production at conventional refineries was used to calculate the displacement credit for diesel produced in the hydrocracker. This value is kg methane per barrel, kg nitrous oxide per barrel, and 82.5 kg carbon dioxide per barrel (96.5 kg CO 2 E/bbl) (Table L-1 of Skone and Gerdes, 2008). 4.3e Greenhouse gas footprint calculations in the model The greenhouse gas footprint calculations are found in the results worksheet of the hydrocracker model. Rows 15 to 59 of this worksheet contain flow rates for wax, preseparated wax, fuel gas, naphtha, SPK, diesel, residual, hydrogen, recycle, coal, and switchgrass. Rows 63 to 92 of this worksheet have estimated properties of the streams, including density, molecular weight, average carbon number, heating values, and the ratio of wax needed to produce the fuels in the hydrocracker to the wax produced by the FT model. Rows 95 to 113 contain literature and assumed values for atomic mass, oxygenated and olefinic compounds found in wax, stream 37

38 temperatures, heating values for certain species, carbon dioxide equivalency factors, the base case electricity demand, and the efficiency of onsite electricity generation. Rows 116 to 120 have the excess/deficit fuel gas calculations, rows 123 to 152 have the greenhouse gas emission factors, and rows 156 to 183 have the greenhouse gas estimates. Rows 186 to 188 contain the carbon sequestration estimates and rows 191 to 203 give unit conversion values that were used in the calculations. References U.S. Department of Defense (DoD). Detail specification: turbine fuel, aviation, kerosene type, JP-8 (NATO F-34), NATO F-35, and JP (NATO F-37). MIL-DTL-83133G. 30 April Hileman, J, R Stratton, P Donohoo. Energy content and alternative jet fuel viability. Journal of Propulsion and Power vol.26 no.6 ( ) doi: / Leckel, D. Hydrocracking of iron-catalyzed Fischer-Tropsch waxes. Energy and Fuels, 2005, 19, Shah, PP, GC Sturtevant, JH Gregor, MJ Humbach, FG Padrta, KZ Steigleder. Fischer-Tropsch wax characterization and upgrading, final report. US Department of Energy. Available through NTIS DE June 6, Sie, ST, MMG Senden, HMH van Wechem. Conversion of natural gas to transportation fuels via the Shell Middle Distillate Synthesis process (SMDS). Catalysis Today, 8 (3), (1991). Skone, TJ, K Gerdes. Development of baseline data and analysis of life cycle greenhouse gas emissions of petroleum-based fuels. US Department of Energy, National Energy Technology Laboratory, Office of Systems, Analysis and Planning. November 26,

39 Chapter 5: Model Operation This chapter presents detailed instructions for using the greenhouse gas emission estimation model. The two sections of the chapter correspond to the model sections: Section 5.1 contains a tutorial for the Fischer-Tropsch (FT) reactor model and Section 5.2 contains a tutorial for the hydrocracker model. The hydrocracker model can be operated so that it is automatically linked to the FT reactor model so that cradle-to-gate greenhouse gas emissions for production of SPK can be estimated. Figures in this chapter reflect screen shots made in Excel 2007; other versions of Excel may have a different appearance. 5.1 FT Reactor Model Operation The FT reactor model is an Excel workbook called FT Section Model v3.xls. This workbook can be used to model the input and output streams of an FT reactor. The model takes a set of user defined variables and generates an FT wax component distribution, estimates the syngas requirements, and determines the associated greenhouse gas emissions from the point of raw material acquisition to the FT reactor for a given quantity of FT wax produced. The model varies the syngas input to the reactor and the amount of syngas sent to the water-gas shift (WGS) reactor in order to meet the constraints imposed by the user. The model achieves this by taking advantage of intentional circular references and iterative calculations in Excel. When the model is launched, a warning message about circular references in the workbook may be displayed. Since the circular reference is intentional, respond by clicking on the OK button if this message appears. Next, click on the ribbon (in Excel 2007) or the File tab (in Excel 2010) in the upper left hand side of the window (indicated by the red arrow in Figure 5-1), then click on the Options button at the lower right of the drop down menu (indicated by the green arrow in Figure 5-1). 39

40 Figure 5-1. First step for enabling iterative calculations Choose the Formulas tab on the left side of the pop-up box (indicated by the blue arrow in Figure 5-2). Ensure that the box for enable iterative calculations is checked, that maximum iterations is set to 32767, and that the maximum change is set to (indicated by the black arrow in Figure 5-2). Model settings that can be altered by the user are contained in the worksheet titled Specifications, which is shown in Figure 5-3. Since this is an iterative model, Excel undergoes a series of iterations until it converges on the desired results when changes are made to the model s settings. The process of iterating may take a few minutes, depending on the speed of the computer being used. While iterations are in progress, the values of various cells change rapidly and Excel may temporarily appear to freeze up until convergence is reached. Excel unfreezes when calculations are complete. 40

41 Figure 5-2. Second step for enabling iterative calculations The calculate button (indicated by the yellow arrow in Figure 5-3) must be clicked after each round of iterations is complete in order to verify that further iterations are not necessary; if the model has converged, clicking on the calculate button will not launch another round of iterations. The wax output level and molar H 2 :CO ratio calculated by the model are displayed at the top of the Specifications worksheet (in cells B1 and D1). If these cells are highlighted in red, it is a warning that convergence has not been reached. This model is capable of simulating the performance of a FT reactor using either a cobalt- or iron-based catalyst. The user-defined variables associated with each type of catalyst are noted below. 41

42 Figure 5-3. FT reactor model specifications worksheet The user-selected elements of the Specifications worksheet are: 1. The molar ratio of H 2 :CO in the overall combined feed to the FT reactor (cell B2). As discussed in Chapters 2 and 3, the model meets the user-selected ratio by adjusting the amount of syngas from the gasifier that is sent to the WGS reactor. Cobalt catalyst FT reactors require a higher molar H 2 :CO ratio than iron catalyst reactors because the WGS reaction is not promoted in cobalt catalyst FT reactors. The typical ratio for an FT reactor with a cobalt-based catalyst is 2.1:1; a reactor with an iron-based catalyst reactor is typically fed a stream with a molar ratio of 1.1:1. 42

43 2. The percentage by mass of gasifier feed that is coal (cell B3). Users may choose 100% coal by mass, 84% coal by mass, or 69% coal by mass from a drop-down box. The remainder of the feed is assumed to be switchgrass. If one of these three coal feed ratios is chosen, the associated values from the gasifier model populate the user defined syngas composition (in cells B19:B29 of the Specifications worksheet). It is possible to enter a different value for the fraction of gasifier feed that is coal, but in that case the user must also define the composition of the syngas produced by the gasifier. 3. The mass percent of oxygenated (cell B4) and olefinic (cell B5) hydrocarbons that are found in the FT wax. Cobalt catalyst FT reactors typically produce smaller amounts of oxygenates and olefins than iron catalyst FT reactors. Shah et al (1988) found that the wax from an iron catalyst FT reactor was 2.9% oxygenates and 7.4% olefins by mass. 4. The FT reactor single pass carbon monoxide conversion (cell B6). This is the total percentage of carbon monoxide entering the reactor that does not exit the reactor as carbon monoxide (regardless of what the carbon monoxide is converted to). A cobalt catalyst reactor may have a single pass conversion in the 70% range while the single pass conversion for an iron-catalyzed reactor system may be as high as 90%. 5. The percent of converted carbon monoxide that is converted to carbon dioxide (cell B7). This can be less than 2% for a cobalt catalyst reactor, while for an iron catalyst reactor it may be as high as 35% to 40%. 6. The percent hydrogen removal (cell B8) and percent carbon dioxide removal (cell B9). These values reflect the fact that at an actual plant, separation processes are not perfect. The hydrogen recovered in the reactor s gas purification section is recycled to the FT reactor, which in turn impacts the amount of syngas from the gasifier that is fed to the WGS reactor. For this reason, decreasing the amount of hydrogen removed in this step can increase the overall carbon dioxide production during water-gas shift. 7. Required wax production (cell B10). Six million kg/day of FT wax generally produces about 50,000 BPD of liquid fuel products from the upgrading section. Exactly how much liquid fuels are produced in the upgrading section depends on wax composition and on process choices in the upgrading section. The hydrocracker model calculates how much wax is needed to make a desired quantity of liquid fuels given the particular wax and upgrading section process selections, and when the hydrocracker and FT models are linked, the hydrocracker model adjusts values from the FT model so that they correspond to the wax production calculated by the hydrocracker model. 8. Percent of unreacted syngas that is recycled (cell B11). This value can range from 0% to 75%; the user selects values from a drop-down box that has 5% increments. Increasing the recycle ratio returns more carbon monoxide to the system and decreases the demand on raw material needed for gasification, which can lower the overall facility greenhouse gas footprint. Recycle ratios in excess of 75% are not realistic and create issues that do not allow the model to converge. 9. The allowable wax delta (cell B12) and allowable molar H 2 :CO ratio delta (cell B13.) These are both model variables that help the model define when it has converged. It is not recommended that model users change these values. 10. The FT reactor α (cell B14) and user defined wax composition (cells E19:E268). Only one of these sets of values should be filled out. If the user has defined the wax in cells E19:E268, the model will generate a comparable wax. If the user has not defined a wax composition 43

44 (i.e., cells E19:E268 have been left blank) the model generates a wax using the α value from cell B a FT Reactor Model Demonstration Case Study In this model demonstration, you will operate the FT reactor model to produce a wax similar to the wax reported by Sie et al (1991), using a cobalt catalyst and assuming that the feed to the gasifier is 16% switchgrass. You are also asked to assume that 50% of the unreacted syngas stream is recycled to the FT reactor and that the wax produced contains 2.9% oxygenates and 7.4% olefins by mass. Run the model to produce 5,995,488 kg/h of wax, with 95% hydrogen removal and 96% carbon dioxide removal. Do not change the default values for the wax and H 2 :CO molar ratio deltas. To run the model, make sure that Excel is configured to perform iterative calculations, as described earlier in this chapter, and open the Specifications worksheet. First, we are going to tell the FT reactor model to produce a wax similar to the wax reported in Sie et al (1991). This wax is one of the example waxes whose composition is provided in the wax input worksheet of the hydrocracker model. Open the hydrocracker model, which is in a workbook titled VLE hydrocracker model.xls, go to the wax input worksheet, and copy cells T10:T257 from that worksheet to cell E21 in the FT reactor model s specifications worksheet. The FT reactor model will freeze momentarily as it performs calculations. Since this is a cobalt catalyst, set the molar H 2 :CO ratio to 2.1, the carbon monoxide single pass conversion to 70%, and the percent of carbon monoxide converted to carbon dioxide to 2%. Select 0.84 for the mass fraction of coal gasified because you have been given that the feed is 16% biomass, and select 50% for the percent of unreacted syngas that is recycled. Enter the appropriate values for oxygenates, olefins, production of wax, and the removal rates of carbon dioxide and hydrogen from the first paragraph of this section. Make sure cell B14 is blank; the model must either be given an α value or a wax composition, but not both. Figure 5-4 is a screenshot that shows what the Specifications worksheet looks like when all of the information has been entered. 44

45 Figure 5-4. FT reactor model demonstration case study inputs You may have noticed that every time you change a value, the model goes through another round of iterations (advanced users may want to turn the iterating feature off before entering all the values in the specifications worksheet and then turning it back on so that the model runs). The 45

46 model has finished calculating if you click on the calculate button at the bottom left of the workbook window and nothing happens. Neither cell B1 nor D1 is highlighted in red when iterations are complete. Based on the wax composition that was entered by the user, the model has calculated an value of 0.89 (found in cell F3 of the Alpha Value worksheet). Table 5-1 shows some of the contents of the results tab. Table 5-1. FT reactor demonstration case study results Run Conditions User Specified Model Unit Actual Required H 2 :CO moles/mole Ratio to Reactor Allowable H 2 :CO 0.02 moles/mole Delta FT Wax Production kg/day Allowable Wax 200 kg/day Production Delta Reactor 494 o K Temperature Reactor Alpha 0.89 FT Single Pass 70.0% Conversion % of CO converted 2.0% to CO 2 % of Oxygenated 2.9% HC's in Products by Weight % of Olefinic HC's in 7.4% Products by Weight % H2 Removal 95.0% % CO2 Removal 96.0% % Syngas Recycle 50.0% Mass Flow Rates Inputs (FT Value Unit Source Production Section) Gasifier Syngas 34,533,308 kg/day calculated to meet daily demand for FT Wax WGS Steam 7,225,315 kg/day calculated to meet daily demand by WGS reactor based on Syngas Flow ATR Steam 1,124,154 kg/day calculated from report by Pina et al and adjusted to account for higher carbon # hydrocarbons ATR Oxygen 1,072,401 kg/day calculated based on needs of the ATR ATR Oxygen (O 2 only) 1,024,970 kg/day Outputs (FT Value Unit Source 46

47 Production Section) FT Wax 5,995,420 kg/day demand from hydroprocessing section Acid Gas Removal 1,399,221 kg/day removal of all HCl, H2S and % of CO2 defined in Run Conditions above Acid Gas Removal 859,965 kg/day (CO2 only) Condensed Water 10,389,897 kg/day water removed from unreacted syngas stream Recovered CO 2 21,455,053 kg/day CO 2 removed from unreacted syngas stream Fuel Gas 4,715,589 kg/day remainder of unreacted syngas after removal of water, CO 2, recycle stream and H2 Fuel Gas 44,806,898,035 kj/day calculated using LHV values listed on Molecular Energies worksheet Fuel Gas 46,453,054,635 kj/day calculated using HHV values listed on Molecular Energies worksheet Gasification Value Unit Source Section Coal Feedrate 17,855 Tonnes/day Calculated based on report " Characterizing the Greenhouse Gas Footprints of Aviation Fuels from Fischer Tropsch Processing" Allen et al, Biomass Feedrate 3,401 Tonnes/day Calculated based on report " Characterizing the Greenhouse Gas Footprints of Aviation Fuels from Fischer Tropsch Processing" Allen et al, Gasification Oxygen 15,222,089 kg/day Calculated based on report " Characterizing the Greenhouse Gas Footprints of Aviation Fuels from Fischer Tropsch Processing" Allen et al, Gasification Oxygen (O 2 only) 14,556,846 kg/day Calculated based on report " Characterizing the Greenhouse Gas Footprints of Aviation Fuels from Fischer Tropsch Processing" Allen et al, Gasifier Syngas H 2 :CO Ratio moles/mol Calculated from gasifier data Table 5-2 shows the carbon number distribution of the FT Liquids sent to the hydrocracking section through carbon number 40, while Figure 5-5 shows a comparison of the user-defined wax to the wax generated by the FT reactor model. 47

48 Mass Fraction Input Wax Calculated Wax Carbon Number Figure 5-5. Comparison of user-defined wax to model-generated wax for the demonstration case study 48

49 Table 5-2. FT wax composition for demonstration case study 5.2 Hydrocracker Model Operation Carbon Number Mass Percent The workbook VLE hydrocracker model.xls can be used to model hydrocracker output for a given wax or for a wax generated by the FT model. It provides estimates of physical properties of hydrocracker products, hydrogen consumption, and flow rates of all hydrocracker streams for a given flow rate of SPK as well as greenhouse gas emission estimates. If a wax generated by the FT model is used as a feedstock to the hydrocracker, the hydrocracker model provides cradle- 49

50 to-gate greenhouse gas emission estimates for SPK produced in an FT production facility, from coal mining and switchgrass cultivation to liquid fuels ready for transport. In this model, the hydrocracker is treated as a series of about 20 subreactors. This is done in order to protect light compounds that would be found in the vapor phase from further cracking. For example, in the first subreactor, there may be constituents that would be in the vapor phase under the given reactor conditions. Within that subreactor, some new vapor phase products are created. These newly created vapor phase products are not allowed to react in the next subreactor or in any subsequent subreactors. The modeling portion of this workbook has the following sections, each with its own worksheet: 1 calculate the breaks calculates the carbon number of products created in each subreactor for each carbon number input 2 find break-fraction combo determines where in the sequence of subreactors the user-selected fraction of bonds broken is located and calculates the hydrocracker product stream by carbon number 3 assign to product cuts assigns isomerization to the hydrocracker products, assigns them to distillation cuts, and calculates physical properties In addition to choosing the wax to be upgraded, the user selects many values. The upper and lower boiling point of the SPK stream can be altered by the user so that estimated properties of the SPK stream meet military specifications as closely as possible. The extent of cracking is controlled by the user so that particular product yields can be optimized. The user decides whether heavy fractions are recycled to the hydrocracker and whether those heavy fractions are in the heavier-than-spk or heavier-than-diesel range. Isomerization in the model is based on user-input values for extent of isomerization and extent of naphthenes created. The user also chooses the carbon number of compounds that are found mostly in the vapor phase in the hydrocracker; this value is related to the temperature and pressure in the hydrocracker. In particular, in the wax input worksheet, the user chooses: whether to use the wax created in the associated FT model or provide the weight percent by carbon number for another wax the weight percent of olefinic and oxygenated compounds in the wax In the model control worksheet, the user chooses values that determine the product yields from the given wax: initial and final boiling points of the SPK fraction the fraction of bonds broken whether recycle of heavy ends is desired and what temperature fraction is recycled, e.g., heavier than SPK or heavier than diesel the temperature cutoff for preseparated wax, if any (some waxes contain material that is already in the desired distillation fraction range) a vapor-liquid equilibrium (VLE) carbon number that represents the carbon number that is mostly contained in the vapor phase (the advanced user may also wish to alter the spread over which carbon numbers around that carbon number are affected by VLE in the calculate the breaks worksheet) 50

51 the desired flow rate of SPK in barrels per day (bpd) (all flow rates are based on this userchosen value) the fraction of branched carbon atoms found in the most commonly observed isomer class the level of ringiness in the liquid fuel products (high, medium, or none), which determines the percentage of compounds that are ringed (0.055, , or none, respectively) the distillation cut temperatures contained in cells A6-A11 of the model control worksheet (this is an option for advanced users and would be unusual) The model calculates the combined feed ratio (CFR) to characterize how much material is being recycled to the hydrocracker. CFR is equal to the total feed divided by the unrecycled feed. Thus, if 33% of the feed to the hydrocracker is recycled material, the CFR is 1.5. CFR is found in cell M11 of the model control worksheet. In recycle mode, the amount of material that is recycled to the hydrocracker depends on whether diesel or SPK is the heaviest product desired, but it also depends on the fraction of bonds broken. If more bonds are broken, less material is recycled. Thus, the user controls CFR by controlling the fraction of bonds broken (cell A25 in the model control worksheet). In no recycle mode, residual can be generated. This is material that is heavier than the heaviest desired product. When the model is operated in no-recycle mode, CFR is 1 and the fraction of bonds broken is adjusted by the user to control the amount of residual produced. The model s calculations rely on instructions contained in macros, so the macro feature of Excel must be enabled in order for the model to operate. In older versions of Excel, an enable macros dialogue box opens when a workbook that uses macros is launched. In later versions, a warning bar appears under the toolbar above the worksheet cells and the user can click on the bar to enable macros. Once the model is run, run conditions and model warnings, if any, can be viewed in the upper right corner of the "results" worksheet, which also provides all the flow rates and greenhouse gas emission estimates. The flow rate calculations will not work if no SPK is produced in the hydrocracker, and there is a limit to how high the fraction of bonds broken can be set. Compounds that don't break in the hydrocracker are assumed to not be isomerized as well. Generally, the smaller a compound is when it enters the hydrocracker, the less likely it is to be broken or cracked (in some cases, it is 100% unlikely). Distillation cuts and other properties were estimated based on group contribution methods and other formulas whose results depend on isomerization. When the recycle stream is combined with the wax stream, it is reborn as straight chain alkanes in the model. While this is not strictly correct, it is unlikely to have much impact on the results because these recycled compounds will be isomerized as they are cracked into product streams. To run the model, the user must enter the desired conditions in the yellow highlighted cells in the "model control" and "wax input" worksheets. Then, the user pushes the appropriate button on the model control page (recycle or no recycle) to run the model. After the model runs, the user may wish to adjust the final boiling point of the naphtha and SPK streams so that distillation specs are met with the highest possible volume of SPK created for a given volume of wax input. In recycle mode with SPK as the heaviest desired product, the model must be run again if the 51

52 final boiling point of the SPK fraction is changed. If the carbons and mass are not balancing (at top right of "results" worksheet), push the button to run the model again. The best way to illustrate use of the model and demonstrate use of the model s features is to use an example. 5.2a Modeling demonstration case study We wish to upgrade the output wax of the FT reactor model described in the first section of this chapter, which was operated to produce a wax similar to that described by Sie et al, The upgrading section will be operated under recycle conditions producing diesel with a final boiling point of 370 C as the heaviest product with a combined feed ratio (CFR) of 1.5. The wax will not be pre-separated; in other words, there is no straight-run product. Our goal is to produce 50,000 bpd of liquid product (diesel, SPK, and naphtha). We think that the most commonly branched group of isomers will have 0.11 of their carbon atoms branched and that no ringed compounds will form in the hydrocracker. Furthermore, we believe that at least half of the C10 compounds will be found in the vapor phase in the hydrocracker. Characterization of the Sie et al wax is given as an example wax in column T of the wax input worksheet. This characterization was entered into the FT model as described in the previous section. In the wax input worksheet of the hydrocracker model, we set the value of cell R1 to m so the modeled parallel to the Sie wax is entered as the wax input in cells N10 to N257, as shown in Figure 5-6. Figure 5-6. Wax inputs for demonstration case study 52

53 Next, we go to the model control worksheet. We enter d in cell A18 because the heaviest desired product is diesel. In cell A19, we put 370 because we want the final boiling point of the diesel to be 370 C. In cells A20 and A21 we enter reasonable values for the final boiling point of the SPK and naphtha cuts (265 C and 136 C); these values will have to be reassessed after the model runs. We input a reasonable value for the barrels per day of SPK produced (20,000) in cell A22; again, we will adjust this number after the model runs until the total liquid product is 50,000 bpd. We enter y in cell A24 because we want recycle operation, and 0.02 in the fraction of bonds broken cell (A25) because that is a good starting point when recycle is desired. We select 0.11, none, and 0 in cells A26 to A28 because that describes the isomerization we expect. We enter 10 in cell A29 because at least half the C10s are expected to be found in the vapor phase where they are protected from further cracking, and we don t want preseparation so we enter a 0 in cell A30. These values are shown in Figure 5-7. Figure 5-7. Model inputs for demonstration case study Then we push the recycle mode button and wait for the model to finish running. In recycle mode, the model has to calculate the output of all 20 reactors, find the recycle stream, add it to the wax stream, and recalculate all 20 reactors again. The model does this seven times when the recycle button is clicked, which is usually enough times to converge on the wax+recycle input to the reactor. If the sum of the values for the six mass and atomic balances in the results worksheet is more than 0.05%, a message saying to run the model again will be displayed. When the model finishes running, it stops on the results worksheet. In the top left corner of this worksheet (cells B2:B9), the run conditions are summarized. The top right corner (cells E2:E11) shows whether run integrity was compromised. If a message to run the model again is 53

54 displayed, go back to the model control worksheet and click to run the model in recycle mode again. CFR is calculated in cell M11 of the model control worksheet. As described at the beginning of this section, we want a CFR of 1.5, and our run returned a value of 1.3, so we need to go back to the model control worksheet and adjust the fraction of bonds broken downwards to meet that condition. We try a few rounds of adjustments, ending up with a value for fraction of bonds broken of Now the CFR is 1.5. To finish, we need to adjust the final boiling point of the naphtha and SPK streams so that the milspecs for distillation temperatures and net heating value are met. There is no range at which both the estimated distillation temperatures and estimated density values simultaneously meet military specifications: the estimated densities do not meet the density requirement unless the initial boiling point of the of the SPK stream is above 205 C, and in order to meet the distillation military specification, at least 10% of the stream must be recovered at 205 C. In order to most closely meet the density requirement, the highest initial boiling point for SPK that satisfies the military specifications is selected. For this run, that means setting the final boiling point of the naphtha stream at 194 C (cell A21 on the model results worksheet). The largest volume of SPK is created when the final boiling point of the SPK stream is set at 300 C (cell A20 on the model results worksheet, and this value also results in the highest possible density estimate. Total liquid output from the hydrocracker is calculated in cell M2 of the model control worksheet. We adjust the flow rate of SPK (cell A22 in the model control worksheet) from the hydrocracker to obtain the desired 50,000 bpd of total liquids. The flow rate of SPK from the hydrocracker must be set to 18,103 bpd in order to get a total liquids flow rate of 50,000 bpd. Figure 5-8 shows the final values in the model control worksheet for the demonstration case study. 54

55 Figure 5-8. Final model input screen for demonstration case study Let s take a detailed look at the flow rate results and some of the physical property estimates, which are summarized in the table below. 55