University of South Carolina Scholar Commons

Theses and Dissertations

2018 Dynamic Simulation of a Solar Powered Hybrid sulfur Process for Production Satwick Boddu University of South Carolina

Follow this and additional works at: https://scholarcommons.sc.edu/etd Part of the Chemical Engineering Commons

Recommended Citation Boddu, S.(2018). Dynamic Simulation of a Solar Powered Hybrid sulfur Process for . (Master's thesis). Retrieved from https://scholarcommons.sc.edu/etd/4820

This Open Access Thesis is brought to you by Scholar Commons. It has been accepted for inclusion in Theses and Dissertations by an authorized administrator of Scholar Commons. For more information, please contact [email protected]. Dynamic Simulation of a Solar Powered Hybrid sulfur Process for Hydrogen Production by Satwick Boddu Bachelor of Technology, Indian Institute of Technology - Guwahati, 2014

Submitted in Partial Fulfillment of the Requirements For the Degree of Master of Science in Chemical Engineering College of Engineering and Computing University of South Carolina 2018

Accepted by: Edward P. Gatzke, Director of Thesis Stanford G. Thomas, Reader John W. Weidner, Reader Cheryl L. Addy, Vice Provost and Dean of the Graduate School

Abstract

The Hybrid Sulfur process is a thermo-electrochemical cycle used to produce hydrogen from water. The process requires a high temperature energy source for H 2SO 4 decomposition with temperature reaching 800°C. This step is followed by SO 2 - depolarized water . Using solar energy as the high temperature energy source allows for efficient environmentally friendly production of hydrogen. This method is an alternative to traditional photovoltaic electrolysis for hydrogen production. Making the process economically competitive is a major challenge. Operating the process with changes in the availability of solar energy also increases process complexity. The dependence of the process on solar energy requires analysis of the electrolysis and decomposition sections separately.

The Hybrid Sulphur process was modelled in ASPEN Plus for a target production rate of 500 gram moles of H 2 per second. The process simulation includes

H2SO 4 decomposition and O 2 separation of the SO 2/O 2 product from the H 2SO 4 decomposition. Given the transient nature of solar energy utilized for the decomposition reaction, analysis of the dynamics of the separation section is of primary importance. A dynamic simulation was developed with control schemes to stabilize the process. This simulation was analyzed for step changes in feed flowrate corresponding to the target hydrogen production rate of 500 gram moles per second. With the proposed controller

ii

configuration, the separation process exhibits time constants ranging from approximately

40 min for a step change in the overall production rate from 100% to 50%. The settling time for the same production rate change is approximately 60 min. The separation system can accommodate the system operating a 0% capacity by maintaining column flow with dilution water. At zero feed the process is functional but it just the recycles the water from the electrolyzer section through the system making it entirely redundant and uneconomical. To avoid shutdown of the separation section at low production rates, this work proposes to include holdup storage tanks for the product streams from the decomposition section. This will allow the distillation columns to run continuously, but the separation system must accommodate variable feed rates. Dynamic variation in the separation section caused by changes in the solar-powered decomposition reactor may thus be mitigated by use of gas and liquid holdup tanks. The results of these simulation prove vital in analysis of the viability of the future for large scale hydrogen production through high temperature Hybrid Sulphur process.

iii

Table of Contents

Abstract ...... ii List of Figures ...... v List of Tables ...... vi Chapter 1. Introduction ...... 1 Chapter 2. Chemistry of Hydrogen Production ...... 4 2.1 Hybrid Sulphur Cycle ...... 4 2.2 The Bayonet Decomposition Reactor ...... 7 Chapter 3. Simulation ...... 10 3.1 Steady state simulation ...... 10 3.2. Dynamic simulation ...... 15 Chapter 4. Results ...... 19 4.1 Steady state results ...... 19 4.2 Energy ...... 19 4.3 Dynamic simulation ...... 20 4.4. Analysis of the Dynamic response ...... 32 4.5. Limitations of the Simulation ...... 34 Chapter 5. Conclusion ...... 36 Chapter 6. Future Work ...... 37 References ...... 38 Appendix A: Aspen plus results ...... 40

iv

List of Figures

Figure 2. 1. Schematic diagram of an SO 2 – depolarized electrolyzer ...... 6

Figure 2. 2. Schematic diagram of Bayonet Decomposition Reactor ...... 9

Figure 3. 1. Schematic diagram of Hybrid Sulphur process (high temperature section) . 12

Figure 3. 2. Hybrid Sulphur process dynamic Simulation ...... 13

Figure 3. 3. Gas Separation section ...... 14

Figure 3. 4. Control schemes in the Aspen Dynamics simulation ...... 17

Figure 3. 5. Control schemes in the Aspen Dynamics simulation ...... 18

Figure 4. 1. Temperature profile of the decomposition reactor (step change: +25%) .... 21

Figure 4. 2. Sump Liquid level profile in the O 2 distillation column (step change: +25%) 22

Figure 4. 3. Pressure profile of the O 2 distillation column (step change: +25%)...... 23

Figure 4. 4. Liquid level profile of the SO 2 distillation column (step change: +25%)...... 24

Figure 4. 5. Pressure profile of the SO 2 distillation column (step change: +25%)...... 25

Figure 4. 6. Temperature profile of the decomposition reactor (step change: -50%) ..... 26

Figure 4. 7. Sump Liquid level profile in the O 2 distillation column (step change: -50%). 27

Figure 4. 8. Pressure profile of the O 2 Distillation column (step change: -50%)...... 28

Figure 4. 9. Sump Liquid level of the SO 2 distillation column (step change: -50%)...... 29

Figure 4. 10. Pressure profile of SO 2 distillation column (step change: -50%)...... 30

Figure 4. 11. Final product stream at zero feed flow rate ...... 31

v

List of Tables

Table 4. 1. Utilities for the steady state SO 2 production rate of 1161 Kmol/hr ...... 19

Table 4. 2. Tank sizes for liquid and gas storages for excess decomposition products ... 33

Table 4. 3. Results of control loop response for vital blocks ...... 34

Table A. 1 Components specified in Aspen Plus ...... 40

Table A. 2 Global reactions ...... 40

Table A. 3 High Temperature reactions ...... 41

Table A. 4 Decomposition output stream ...... 42

Table A. 5 Aspen results for the Output stream ...... 44

Table A. 6 Results of Energy analysis on Aspen Plus ...... 45

Table A. 7 The Gaseous stream entering the gas separation section ...... 46

Table A. 8 The liquid stream entering the separation section ...... 48

Table A. 9 Outlet stream after oxygen separation ...... 50

Table A. 10 Design specifications of the Oxygen distillation column ...... 52

Table A. 11 Design Specifications of the Sulphur Dioxide distillation column ...... 52

Table A. 12 Temperature control on reactors ...... 53

Table A. 13 Sump level control for oxygen separation ...... 53

Table A. 14 Pressure control for oxygen separation ...... 53

Table A. 15 Sump level control for Sulphur dioxide separation ...... 54

vi

Table A. 16 Pressure control for Sulphur dioxide separation ...... 54

vii

Chapter 1. Introduction

The need for renewable and sustainable energy is dictated by the fact of depleting fossil fuels and an intention to lower the greenhouse gas emissions to fight climate change.

Hydrogen is seen as a means to accomplish this goal in the long term. Hydrogen is an outstanding storage medium for energy, which is a key component in the . Electricity can be used to produce Hydrogen by splitting water by electrolysis.

Hydrogen is then stored and can be used to generate electricity using efficient fuel cells.

With their heavy dependence on climatic conditions, renewable energies like solar and wind must address energy storage and transportation problems. Hydrogen is a possible solution to these problems, paving the way towards a promising future for renewable energies. Hydrogen is vastly abundant in nature, however not in its pure state. Due to high reactivity, it is found bound to other elements to form compounds (water, hydrocarbons). Limited availability of Hydrogen in its basic diatomic form (H 2) imposes a need to find an effective production method.

Hydrogen can be produced from natural gas using high-temperature steam. This process, called steam methane reforming, accounts for about 95% of the hydrogen used today in the United States [1]. However this method involves consumption of fossil fuels and emission of greenhouse gases. process are attractive for hydrogen production due to the abundance of water in the nature and limited emission

1

of greenhouse gases when consumed. Direct electrolysis of water requires electricity generation. The overall efficiency is of this process likely to be only about 20-24% based on the Lower heating value (LHV) of hydrogen produced [2]. Processes which promise efficiencies of over 40% have been identified which use external heat generated by solar or nuclear sources to power thermochemical water splitting process [2,3,4]. More than

100 processes have been subjected to various analyses and research studies to determine the potential candidates to produce H 2 since 1960’s. Thermochemical cycles using sulfur often require a high temperature H 2SO 4 decomposition section. These processes are among the most attractive options due to their high thermal efficiencies compared to other competitive processes [5,6,7].

Among the thermochemical cycles examined by the NHI, the Hybrid

Sulphur process and Sulfur-Iodine (SI) process were recognized as effective for nuclear hydrogen production on large scale [8]. The Sulphur-Iodine cycle is a three step process with recycling of Iodine and sulphur dioxide. Under International Nuclear Energy Research

Initiative (INERI), the French CEA, General Atomics and Sandia National Laboratories (SNL) are jointly developing the sulfur-iodine process [9,10,11]. The Hybrid sulphur process involves a high temperature decomposition step which is followed by electrolysis. The

Hybrid Sulphur cycle has been identified as one of the most advanced processes for solar hydrogen production, with research activities funded under the DOE-EERE Solar

Thermochemical Hydrogen (STCH) program. SRNL is involved in the STCH project to determine the feasibility of the Hybrid Sulphur process and developing the components

2

of Hybrid Sulphur process: Bayonet decomposition reactor and SO 2 – depolarized electrolyzer (SDE) [2].

This work focuses on developing a separation section for the sulphuric acid decomposition products and analyzing the dynamic response for several process fluctuations (eg. Energy) in decomposition.

3

Chapter 2. Chemistry of Hydrogen Production

2.1 Hybrid Sulphur Cycle :

The Hybrid Sulphur cycle was first proposed by Westinghouse Electric Corporation in

1970s [12,13]. It came to be known as the Westinghouse process after undergoing years of development. The process involves oxidation and reduction reaction of sulphur. The entire process involves two major reaction steps: 1] High temperature decomposition of

H2SO 4 and 2] SO 2 depolarized electrolysis of water.

The decomposition of H 2SO 4 is endothermic, effectively proceeding with an external heat supply at temperatures higher than 800 OC. This decomposes sulphuric acid to produce sulphur trioxide (SO 3) and steam (H 2O):

H2SO 4 --> SO 3 + H 2O

Sulphur trioxide is dissociated to SO 2 and O 2 by further catalytic heating.

SO 3 -----> SO 2 + 1/2O 2

O2 is removed from the decomposition products as a by-product by distillation. The remaining mixture of H 2O, SO 2 and unreacted H 2SO 4 is fed to the anode of the sulphur dioxide electrolyzer (SDE). The SO 2 in the mixture is electrochemically oxidized at the anode to form H 2SO 4, protons and electrons.

4

+ SO 2 + 2H 2O -----> H 2SO 4 +2H + 2e-

The H 2SO 4 is recycled back to the decomposition reactor. The protons are transported across the membrane to the cathode where it recombines with electrons to form Hydrogen (H 2) [14].

+ - 2H + 2e ----> H 2

Sulphur – Dioxide Electrolysis (SDE) distinguishes the Hybrid Sulphur process from other the Sulphur cycles. The process has a standard cell potential -0.158 V at 25 OC.

This is 87% less than that of water electrolysis i.e. -1.229V [15]. This leads to far less consumption of electricity, making the process more efficient and a primary factor in low cost hydrogen production [3,4]. In practice, water electrolyzers have higher cell potentials of -1.7 to -2.0 V for most of the commercially viable current densities due to ohmic losses and electrode over-potentials. Similarly SDEs operate at a cell potential higher than -

0.243V at practical current densities due to the SO2 dissolved [4]. SRNL research was able to attain a potential of -0.6V at temperature higher than 100 OC and pressure around 10 bar for a current density of 500 mA/cm2 [4,16]. Thus, SDE appears to be a significant improvement over traditional water electrolysis.

5

Figure 2. 1. Schematic diagram of an SO 2 – depolarized electrolyzer

6

2.2 The Bayonet Decomposition Reactor:

A bayonet reactor is preferred to carry out the high temperature sulphuric acid decomposition. During the NHI, SNL identified the bayonet reactor as the fundamental design to vaporize and decompose the H 2SO 4 inside a single component [17]. The reactor is modelled to be a plug-flow reactor with the feed and heating fluid flowing though

0 concentric paths. The concentrated H 2SO 4 feed mixture enters the reactor at 120 C. It is

0 vaporized and superheated to around 1075 C where it decomposes into SO 3. Sulphur trioxide is catalytically decomposed into SO 2 and O 2. This gaseous mixture is cooled to 480

0 0 C with SO 3 re-associating with water to form H 2SO 4 and further cooled down to 250 C.

Due to the high-temperature operating conditions of the reactor it is a challenge to find an appropriate material that can resist these extreme conditions without corrosion or deterioration. Additionally, the material should have good heat transfer characteristics. Silicon Carbide (SiC) is proposed as a solution to these problems. SiC can be used as the tube material for both the external end and internal open closed tubes

[18]. As depicted in the Figure 2, the reactor consists of one closed ended SiC tube and an open ended SiC tube both co-axially aligned to form two concentric flow paths. A baffle tube may be included to enhance heat transfer. Concentrated liquid H 2SO 4 is fed at the open end to the annulus, where it is vaporized before passing through an annular catalyst bed. The decomposition reaction takes place in the catalyst bed, using heat provided by the external heat source. The products are SO 2, O 2, and H 2O in vapor form. These products return through the central tube and exchange heat with the feed through recuperation.

7

Cooled and partially condensed product exits out the open end [19, 20]. A major advantage of this reactor is that high temperature internal heat recovery is realized in a single heat exchanger device. Additionally, the reactor seals are only required at the low- temperature base of the reactor.

8

High Temperature Heat

Outlet: SO 2,

O2, H 2O

Figure 2. 2. Schematic diagram of Bayonet Decomposition Reactor

9

Chapter 3. Simulation

A model of high temperature section of Hybrid Sulphur process with the gas separation section and decomposition reactor is simulated on ASPEN plus. Given the complexity of designing a bayonet reactor, a generic stoichiometric reactor was setup with fixed conversion rates. The output stream from the reactors is passed into the separations section where distillation is employed. The steady state and dynamic responses of the gas separations are primarily analyzed for variations in the feed flow rate.

3.1 Steady state simulation :

The Hybrid Sulphur simulation model built with ASPEN plus does not include the electrolysis process and is limited to decomposition section of the process. The decomposition section of the process is solar driven, which leads to variations in energy supply for the decomposition. This unpredictability in the high temperature section recommends a storage for SO 2 produced so that it can be used to run the electrolysis uninterrupted. Being dynamically different the electrolysis section and high temperature decomposition section are designed and analyzed separately. Therefore this simulation is limited only to the decomposition section. The model

10

consists of mainly two sections: 1) High temperature decomposition and 2) separation of gases.

The feed stream to the reactor contains H 2SO 4 at a concentration of 50 wt% at 120 0C and 1 bar. This feed is decomposed to Sulphur dioxide and oxygen by the decomposition section. A separator at 145 0C concentrates the sulphuric acid and is then passed through the decomposition section which is designed with two reactors and a heater to resemble the original bayonet reactor. The first reactor decomposes the

0 0 sulphuric acid at 950 C at 1 bar to SO 3. The second reactor at 950 C further decomposes

SO 3 to SO 2 and O 2. The material stream from the decomposition reactors are subjected to further cooling and passed into flash drums to separate liquid and gaseous components.

The gas and liquid streams from the high temperature decomposition section are passed into a separation section. The separation of gases was accomplished primarily through two distillation columns. The feed to the first distillation column

(number of stages = 15, diameter = 2.21 m, height = 9.144 m) is composed of the gas and liquid streams from the decomposition section. This column separates the oxygen, leaving

Sulphur dioxide along with water. Sulphur dioxide is completely separated from the water in the second distillation column (number of stages = 15, reflux ratio = 4, diameter = 4.92 m, height = 7.24 m). After separation, the SO 2 may be stored or passed into the electrolyzer section.

11

12

Figure 3. 1. Schematic diagram of Hybrid Sulphur process (high temperature section)

13

Gas Separation Section High temperature decomposition Section

Figure 3. 2 Hybrid Sulphur process dynamic Simulation

14

Figure 3. 3 Gas Separation section

3.2. Dynamic simulation:

The steady state simulation in Aspen plus is converted to Aspen Dynamics. The model must have controllers added to maintain stable operation. This allows the process to be analyzed under various production rates. Controllers are established over various blocks in the simulation and tuned to enable the system to vary the production rate of sulphur dioxide upto 50%. The prominent control schemes used for the blocks are:

1) Heaters with specified vapor fraction: Ratio control of input and output volumetric

flow with heat duty as the manipulated variable.

2) Heaters with specified outlet temperature: Feedback reverse acting control with

heat duty as manipulated variable.

3) Tanks: Feedback control with outlet flow as manipulated variable.

4) Reactors: Feedback control of reactor temperature with heat duty being the

manipulated variable.

5) Flash separators: Feedback control of pressure and liquid level of the blocks by

controlling the flow rate of gas and liquid streams respectively.

6) Distillation Columns: Sump level control by bottom stream flowrate and pressure

control by overhead stream flowrate.

15

The controllers on the distillation columns and other blocks are tuned to proper parameters for efficient transition between different production rates. The settling time for distillation columns post the step change were upto 2-3 hours for the above mentioned transitioned. The controllers needed large Gain to make the process rapid and smooth. Such large gains however raise serious stability issues and complexities in practical applications.

16

Property controlled : Liquid level Controlled : Liquid level and pressure

Set-point : 6.19 m Set-point : 2.83 m and 11.4 bar

Figure 3. 4. Control schemes in the Aspen Dynamics simulation

17

Property Controlled: outlet Temperature

O Set-point: 1350 C

Figure 3. 5. Control schemes in the Aspen Dynamics simulation

18

Chapter 4. Results

The simulation is run on ASPEN Plus which analyzes the system under steady state. The simulation is then converted to ASPEN dynamics and results were recorded for the changes in feed flowrates .

4.1 Steady state results :

O A feed stream of H 2SO 4 (50% wt) enters the system at a rate of 11,606 Kmol/hr at 120 C.

After carrying out both the high temperature decomposition and gas separations the simulation model managed to generate 1161 Kmol/hr of Sulphur dioxide which is stored/passed into the electrolysis section. Oxygen gas of 894 Kmol/hr is produced as a by-product.

4.2 Energy :

Energy analysis of the steady state simulation is in the table below .

Table 4. 1. Utilities for the steady state SO 2 production rate of 1161 Kmol/hr

Property Energy in Gcal/hr

Heating Utilities 393.6

Cooling Utilities 268.7

19

4.3 Dynamic simulation :

The dynamic simulation is subjected to changes in the input feed flowrate. The controllers placed on each block is tuned enough to make the transitions quick and effective. The simulation is analyzed for an increase to 14606 Kmol/hr and decrease to 5803 Kmol/hr from steady state for the input feed flowrate. The dynamic simulation does scale down much more than 50% but the ASPEN dynamics is ineffective in recognizing the empty trays. The step down in production rates is limited to 50% to ensure no empty trays. The dynamics of the decomposition reactor are limited to the temperature. A better design of the reactor would have enabled us to analyze the transience of the decomposition reaction.

Dynamic response for the feed flow rate changes of some blocks are depicted below. There was no oscillatory or inverse responses observed.

4.3.1 For an increase of 25% in feed flowrate :

The amount of SO 2 now produced is 1149 kmol/hr with traces of oxygen and 118 kmol/hr of water in the output stream. The feed flowrate is increased to 14606 kmol/hr (125%) at t=10.5 hr. The system settles back to the steady state in about 2 hrs.

20

Figure 4. 1. Temperature profile of the decomposition reactor (step change: +25%)

The feed flowrate is increased to 14606 kmol/hr (125%) at t=10.5 hr. The settling time for the process is 2.5 hr for a setpoint of 1350 oC with energy as the controller output . The response is non-oscillatory. No inverse response.

21

Figure 4. 2 Sump Liquid level profile in the O 2 distillation column (step change: +25%)

The settling time for the process is 3 hr for a setpoint of 2.837 m with bottom stream flowrate as the controller output. The dynamic response is non-oscillatory. No inverse response.

22

Figure 4. 3 Pressure profile of the O 2 distillation column (step change: +25%)

The settling time for the process is 1.5 hr for a setpoint of 11.4 bar with top gas stream flowrate as the controller output. The dynamic response is non-oscillatory. No inverse response .

23

Figure 4. 4. Liquid level profile of the SO 2 distillation column (step change: +25%)

The settling time for the process is 2 hr for a setpoint of 6.15 m with bottom stream flowrate as the controller output. The dynamic response is non-oscillatory. No inverse response.

24

Figure 4. 5. Pressure profile of the SO 2 distillation column (step change: +25%)

The settling time for the process is 2.25 hr for a setpoint of 1.2 bar with top gas stream flowrate as the controller output. The dynamic response is non-oscillatory.

Process has Inverse response .

25

4.3.2 Decrease of 50% in feed flowrate :

The feed flowrate is decreased to 5803 kmol/hr (50%) at t=2 hr. Amount of SO 2 now produced is 560 kmol/hr along with traces of O 2 and 622 kmol/hr of water in the output stream .

Figure 4. 6. Temperature profile of the decomposition reactor (step change: -50%)

The settling time for the process is 2.5 hr for a setpoint of 1350 oC with energy as the controller output. The dynamic response is non-oscillatory. Process have no Inverse response.

26

Figure 4. 7. Sump Liquid level profile in the O 2 distillation column (step change: -50%)

The settling time for the process is 3 hr for a setpoint of 2.837 m with bottom stream flowrate as the controller output. The dynamic response is non-oscillatory.

Process has no Inverse response.

27

Figure 4. 8. Pressure profile of the O 2 Distillation column (step change: -50%)

The settling time for the process is 2 hr for a setpoint of 11.4 bar with top gas stream flowrate as the controller output. The dynamic response is non-oscillatory.

Process has no Inverse response .

28

Figure 4. 9. Sump Liquid level of the SO 2 distillation column (step change: -50%)

The settling time for the process is 3 hr for a setpoint of 6.15 m with bottom stream flowrate as the controller output. The dynamic response is non-oscillatory.

Process has no Inverse response .

.

29

Figure 4. 10. Pressure profile of SO 2 distillation column (step change: -50%)

The settling time for the process is 5 hr for a setpoint of 1.2 bar with top gas stream flowrate as the controller output. The dynamic response is oscillatory. Process has inverse response.

30

Figure 4. 11. Final product stream at zero feed flow rate

The SO 2 production rate is very low (11.7 kmol/hr); the stream has a high composition of water at 0.98 mole-fraction.

31

4.4. Analysis of the Dynamic response

The important components of the simulation like distillation columns are stable for major step changes in the feed. The controller action is swift and had a settling time of around

2-3 hours for the vital blocks of the simulation (Figure 4.7, Figure 4.9). The average time constant of the process is estimated to be 40 min (table 4.3). The controllers are set to large gains in the order of 30 to attain a faster transient response. The simulation cannot handle feed flowrates more than 14606 kmol/hr, which is 25% higher than the steady state value. The production rates at 25% above the steady state value are rarely achieved due to the nature of solar energy. The 25% scale up is adequate. The decomposition process is subjected to rapid changes in energy input. The step changes in the feed above

8000 kmol/hr resulted in severe errors in the thermodynamic property package ENRTL –

RK.

Though the process does not scale down to zero directly for a single step change, the simulation does not break down at zero feed flow rate. The recycled water stream from the electrolyzer fuels the system in the case of zero feed flow and the system is just processing the water (Figure 4.11). This makes running the process at zero flow rate completely redundant. Shutdown of the separation system at lower production rates is an efficient option. A high level energy/economic optimization is needed to decide the flowrates below which it is inefficient to run the separation system. Storage tanks can be used to collect excess liquid and gaseous streams out of the decomposition section while the decomposition section is processing at higher production rates (100-125%). This gas

32

and liquid storage tanks can be used to mitigate the effect of variations of production rates and maintain the separation process at high flowrates to avail efficiency. The process breaks down for step changes larger than 8000 Kmol/hr, the storage tanks can be a solution to the problem. Tank sizes required are estimated for an average feed flowrate of 14806 kmol/hr which is 25% above the steady state value (table 4.2). The dynamics of the step changes showed no inverse or oscillatory responses except for the overhead pressure control for the SO 2 distillation column. Having a simple transient response for control process makes the simulation more feasible in reality.

Table 4. 2. Tank sizes for liquid and gas storages for excess decomposition products

Process Time Gas Tank size Liquid Tank size Capital Cost (hr) (m3; gallons) (m 3; gallons) ($)

4 3286; 868069 299; 78987 470,200

8 6572; 1736139 600; 158503 738,100

12 9859; 2604472 898; 237227 960,400

24 19717; 5208680 1797; 474717 1,507,200

The tanks are designed for gas flowrate = 821 m 3/hr, liquid flowrate = 75 m 3/hr at

Temperature = 40 oC, Pressure = 12 bar.

33

Table 4. 3. Results of control loop response for vital blocks

Value Time Constant Settling time (t s)

Reactor temperature (Figure 4.6) 30 min 2 hrs

Liquid level for O 2 distillation 45 min 3 hrs

(Figure 4.7)

Liquid level for SO 2 distillation 45 min 3 hrs

(Figure 4.9)

The above results of the process are observed for step changes in production rates from

100% to 50%.

4.5. Limitations of the Simulation

The simulation does not scale up for a step larger than 8000 kmol/hr in feed flow rate.

This limits the possibility of a quick shutdown of the process when necessary. Since the simulation is not equipped with utilities and energy streams, the distillation columns are not equipped with temperature control. The temperature control of distillation columns is vital for the top stream to have a high product composition (pure product). The percent of water in the sulphur dioxide stream increases whenever we step down the production rates. The controllers are tuned to pretty high gain values to achieve a faster transient response. Though the high gain values are adequate for quick transition, they lead to problems regarding the stability of the control loops. The simulation is stable at low production rates close to zero but the sensitivity of Aspen dynamics software to low

34

material content in the distillation trays is questionable. The separation section requires a shutdown at low production rates, the cutoff value for feed flowrate is unknown by the study. Further energy/economic optimization is required to determine the minimum feed flowrate required to run the separation process.

35

Chapter 5. Conclusion

Large scale hydrogen production is vital to a future of sustainable energy. The hybrid

Sulphur process is one which is recognized as an alternative for hydrogen production. The Aspen simulation of high temperature section is analyzed for the steady state and dynamic response of the separation section. The steady state simulation generates 1161 Kmol/hr of SO 2 and 894 Kmol/hr of O 2 as a by-product. The process is mainly solar powered which subjects the process to significant variations in energy supply in a day. The simulation is now equipped with control loops tuned to efficient parameters and analyzed dynamically. Maximum step change of 8000 kmol/hr in the feed flowrate is executed before simulation breakdown. Most of the blocks of the gas separations were able to re-stabilize back to their steady states within 2 -3 hours on average for a step change in production rates. The process had a time constant of 20 min.

At low flowrates, shutdown of the separation section and collecting the decomposition product streams into storage tanks is an efficient option. A rapid and stable dynamic gas separation section for the hybrid Sulphur process is simulated. Further work in this area can make hybrid sulphur process a feasible option for large scale hydrogen production.

36

Chapter 6. Future Work

The Hybrid Sulphur process requires further design to fully realize it as a practical option to produce hydrogen at an industrial scale. The high fidelity model of the reactor is desired. Building customized models of reactors with use of a software like Aspen custom modeler can lead to great precision in the design of the reactors. Heaters need to be replaced by heat exchangers and further improve the energy management by heat integration. Optimization of the process is required. The dynamic simulation which is now equipped with basic control schemes like feed-back control with high gain parameters to ensure a quick transition between production rates. This could lead to stability issues for the control loops and can lead to difficulties in practical execution. So, advanced control methods like model predictive control are needed to be employed to ensure better functioning of the separations section. The response of the separation system is on the order of 2-3 hours, indicating that transient response may be quite difficult to execute in reality without some thermal holdup / storage. The separations section of the simulation does not include removal of unreacted sulphuric acid. Additional process for sulphuric acid removal before electrolysis is also necessary.

37

References

1. Fuel cell technologies office, Energy efficiency and renewable energy, Department of Energy, April 2016. 2. Cladio Corgnale, William A. Summers, Solar hydrogen production by Hybrid Sulphur process, International Journal of Hydrogen Energy 36(2011) 11604-11619 3. Claudio Corgnale, Sirivatch Shimpalee, Maximilian B. Gorensek, Pongsarun Satjaritanun, John W.Weidner, William A.Summers, Numerical modeling of a bayonet heat exchanger-based reactor for sulfuric acid decomposition in thermochemical hydrogen production processes, International journal of Hydrogen energy, volume 42, issue 32, 10 August 2017, Pages 20463-20472, 4. Maximilian B. Gorensek, William A. Summers. Hybrid sulfur flowsheets using PEM electrolysis and a bayonet decomposition reactor. International journal of hydrogen energy 34 (2009) 4097 – 4114. 5. Ginosar DM, Glenn AW, Petkovic LM, Burch KC. Stability of supported platinum sulfuric acid decomposition catalysts for use in thermochemical water splitting cycles. Int. J Hydrogen Energy 2007; 32(4):482–8. Brown LC, Besenbruch GE, Lentsch RD, Schultz KR, Funk JF, Pickard PS, et al. High efficiency generation of hydrogen fuels using nuclear power. Final Technical Report from General Atomics Corp. to US DOE. GA-A24285; 2003. 6. Funk JE. Thermochemical hydrogen production: past and present. Int. J Hydrogen Energy 2001; 26 (3):185e90. 7. L.C. Brown, G.E. Besenbruch, R.D. Lentsch, K.R. Schultz, J.F. Funk, P.S. Pickard, et al., High efficiency generation of hydrogen fuels using nuclear power, Final Technical Report from General Atomics Corp. to US DOE. GA-A24285 (2003). 8. M.B. Gorensek, W.A. Summers, C.O. Bolthrunis, E.J. Lahoda, D.T. Allen, R. Greyvenstein, Hybrid sulfur process reference design and cost analysis, report, May 12, 2009; South Carolina. 9. Giovanni Cerri, Coriolano Salvini, Claudio Corgnale, Ambra Giovannelli, Daniel De Lorenzo Manzano, Alfredo Orden Martinez, Alain Le Duigou, Jean-Marc Borgard, Christine Mansilla, Sulfur–Iodine plant for large scale hydrogen production by nuclear power, International Journal of Hydrogen Energy, Volume 35, Issue 9, May 2010, Pages 4002-4014.

38

10. S. Goldstein, J.M. Borgard, X. Vitart, Upper bound and best estimate of the efficiency of the iodine sulphur cycle, Int J Hydrogen Energy, 30 (6) (2005), pp. 619- 626. 11. International Nuclear Energy Research Initiative, 2006 Annual Report. DOE/NE- 131.Washington, DC: United States Department of Energy, 2007. p. 113. 12. Lee E. Brecher and Christopher K. Wu, Electrolytic decomposition of water, Westinghouse Electric Corp., Patent 3,888,750, June 10, 1975. 13. L.E. Brecher, S. Spewock, C.J. Warde, The Westinghouse Sulfur Cycle for the thermochemical decomposition of water, Int J Hydrogen Energy, 2 (1) (1977), pp. 7-15. 14. M.B. Gorensek, Hybrid sulfur cycle flowsheets for hydrogen production using high temperature gas-cooled reactors, Int. J Hydrogen Energy, 36 (20)(2011), pp. 12725-12741. 15. M.B. Gorensek, J.A. Staser, T.G. Stanford, J.W. Weidner, A thermodynamic analysis of the SO2/H2SO4 system in SO2-depolarized electrolysis, Int. J Hydrogen Energy, 34 (15)(2009), pp. 6089-6095. 16. A.G. Niehoff, N.B. Botero, A. Acharya, D. Thomey, M. Roeb, C. Sattler, R. PitzPaal, Process modelling and heat management of the solar hybrid sulfur cycle, Int. J Hydrogen Energy, 40 (2015), pp. 4461-4473. 17. R. Moore, P. Pickard, E. Parma, M. Vernon, F. Gelbard; Integrated boiler, superheater, and decomposer for sulphuric acid decomposition; Sandia Corp. (2010), US Patent No. 7645437 B1. 18. Maximilian B. Gorensek, Thomas B. Edwards; Energy Efficiency Limits for a Recuperative Bayonet Sulfuric Acid Decomposition Reactor for Sulfur Cycle Thermochemical Hydrogen Production, Ind. Eng. Chem. Res. 2009, 48, 7232–7245. 19. Maximilian B. Gorensek, William A. Summers. Hybrid sulfur flowsheets using PEM electrolysis and a bayonet decomposition reactor. International journal of hydrogen energy 34 (2009) 4097 – 4114. 20. E.J. Parma, M.E. Vernon, F. Gelbard, R.C. Moore, H.B.J. Stone, P.S. Pickard, Modeling the sulfuric acid decomposition section for hydrogen production, Proceedings of 2007 int. topical meeting on safety and techn. of nucl hydrogen production, Control and Mgmt, Boston (June 24–28, 2007), pp. 154-160.

39

Appendix A: Aspen plus results

Table A. 1 Components specified in Aspen Plus

Component ID Component name H2O WATER H2SO4 SULFURIC-ACID H2 HYDROGEN O2 OXYGEN SO2 SULFUR-DIOXIDE SO3 SULFUR-TRIOXIDE H3O+ H3O+ HSO3- HSO3- HSO4- HSO4- SO3-- SO3-- SO4-- SO4--

Thermodynamic package: ENRTL-RK

Table A. 2 Global reactions

Reaction Type Stoichiometry 1 Equilibrium H2SO4 + H2O <--> H3O+ + HSO4- 2 Equilibrium H2O + HSO3- <--> H3O+ + SO3-- 3 Equilibrium 2 H2O + SO2 <--> H3O+ + HSO3- 4 Equilibrium H2O + HSO4- <--> H3O+ + SO4-- 5 Equilibrium SO3 + H2O <--> H2SO4

40

Table A. 3 High Temperature reactions

Reaction Type Stoichiometry 1 Equilibrium H2SO4 <--> H2O + SO3 2 Equilibrium H2SO4 <--> H2O + 0.5 O2 + SO2

Feed stream input specifications Pressure : 1 bar Temperature : 393 K Mass Flow rate : 98 kg/sec

Mass Fractions : H 2O – 0.501, H 2SO 4 – 0.499 Chemical reactions in the Reactors

H2SO 4 --> H 2O + SO 3

H3O+ + HSO 4- --> H 2SO 4+ H 2O

H2O + SO 3 --> H 2SO 4

H2SO 4 --> H 2O+ 0.5 O 2+ SO 2 Reactor type : RStoic (Stoichiometric) Temperature : 1350 oC Pressure : 1.3 – 1.5 bar Type of Reactor : Stoichiometric Total Heat duty : 332.690 Gcal/hr

41

Table A. 4 Decomposition output stream

Description Units DECOMP

Phase Vapor Temperature C 1350 Pressure bar 1.33 Molar Vapor Fraction 1 Molar Liquid Fraction 0 Molar Solid Fraction 0 Mass Vapor Fraction 1 Mass Liquid Fraction 0 Mass Solid Fraction 0 Molar Enthalpy kcal/mol -41.80483607 Mass Enthalpy kcal/kg -1549.781454 Molar Entropy cal/mol-K 9.661512004 Mass Entropy cal/gm-K 0.358169856 Molar Density kmol/cum 0.009855128 Mass Density kg/cum 0.265838782 Enthalpy Flow Gcal/hr -447.8709538 Average MW 26.97466535 Mole Flows kmol/hr 10713.37663 H2O kmol/hr 8000.821035 H2SO4 kmol/hr 1.21E-14 H2 kmol/hr 0 O2 kmol/hr 902.2407285 SO2 kmol/hr 1804.481457 SO3 kmol/hr 1.99E+00 H3O+ kmol/hr 1.973275666 HSO3- kmol/hr 0 HSO4- kmol/hr 1.773123859 SO3-- kmol/hr 0 SO4-- kmol/hr 0.100075904 Mole Fractions H2O 0.746806662 H2SO4 1.13E-18 H2 0 O2 0.08421628 SO2 0.16843256 SO3 0.000185463

42

H3O+ 0.000184188 HSO3- 0 HSO4- 0.000165506 SO3-- 0 SO4-- 9.34E-06 Mass Flows kg/hr 288989.7493 H2O kg/hr 144137.0312 H2SO4 kg/hr 1.19E-12 H2 kg/hr 0 O2 kg/hr 28870.62062 SO2 kg/hr 115603.7436 SO3 kg/hr 159.0823279 H3O+ kg/hr 37.53697181 HSO3- kg/hr 0 HSO4- kg/hr 172.1208388 SO3-- kg/hr 0 SO4-- kg/hr 9.613761648 Mass Fractions H2O 0.498761743 H2SO4 4.12E-18 H2 0 O2 0.099901885 SO2 0.400027143 SO3 0.000550477 H3O+ 0.00012989 HSO3- 0 HSO4- 0.000595595 SO3-- 0 SO4-- 3.33E-05 Volume Flow cum/hr 1087086.493

43

Table A. 5 Aspen results for the Output stream

Description Units Output stream

Phase Mixed Temperature C 29.7 Pressure bar 10 Molar Vapor Fraction 0.00645125 Molar Liquid Fraction 0.99354875 Molar Solid Fraction 0 Mass Vapor Fraction 0.004962149 Mass Liquid Fraction 9.95E-01 Mass Solid Fraction 0 Molar Enthalpy kcal/mol -75.41655593 Mass Enthalpy kcal/kg -1258.260746 Molar Entropy cal/mol-K -1.94E+01 Mass Entropy cal/gm-K -0.32407865 Molar Density kmol/cum 16.61787725 Mass Density kg/cum 996.0281074 Enthalpy Flow Gcal/hr -96.34013014 Average MW 59.93714435 Mole Flows kmol/hr 1277.440065 H2O kmol/hr 111.1810133 H2SO4 kmol/hr 0.00E+00 H2 kmol/hr 0 O2 kmol/hr 4.76E+00 SO2 kmol/hr 1161.473458 SO3 kmol/hr 0 H3O+ kmol/hr 0.012589233 HSO3- kmol/hr 0.012589233 HSO4- kmol/hr 0 SO3-- kmol/hr 4.13E-16 SO4-- kmol/hr 0.00E+00 Mole Fractions H2O 8.70E-02 H2SO4 0.00E+00 H2 0 O2 3.73E-03 SO2 0.909219532 SO3 0

44

H3O+ 9.86E-06 HSO3- 9.86E-06 HSO4- 0 SO3-- 3.24E-19 SO4-- 0 Mass Flows kg/hr 76566.10959 H2O kg/hr 2002.957084 H2SO4 kg/hr 0.00E+00 H2 kg/hr 0 O2 kg/hr 1.52E+02 SO2 kg/hr 74409.56479 SO3 kg/hr 0 H3O+ kg/hr 0.239480834 HSO3- kg/hr 1.020643024 HSO4- kg/hr 0 SO3-- kg/hr 3.31E-14 SO4-- kg/hr 0.00E+00

Table A. 6 Results of Energy analysis on Aspen Plus

Utilities Actual Target Available % of Savings Actual Total Utilities 673.8 374 299.8 44.49 [Gcal/hr] Heating Utilities 393.6 243.7 149.9 38.08 [Gcal/hr] Cooling Utilities 280.2 130.3 149.9 53.49 [Gcal/hr] Carbon Emissions 0 0 0 0 [kg/hr]

45

Table A. 7 The Gaseous stream entering the gas separation section

Property Units Gaseous Stream Phase Vapor Temperature C 39.85008337 Pressure bar 12 Molar Vapor Fraction 1 Molar Liquid Fraction 0 Molar Solid Fraction 0 Mass Vapor Fraction 1 Mass Liquid Fraction 0 Mass Solid Fraction 0 Molar Enthalpy kcal/mol -26.60771233 Mass Enthalpy kcal/kg -605.2980878 Molar Entropy cal/mol-K -2.368338931 Mass Entropy cal/gm-K -0.053877275 Molar Density kmol/cum 0.479446523 Mass Density kg/cum 21.07552529 Enthalpy Flow Gcal/hr -38.46899243 Average MW 43.95803137 Mole Flows kmol/hr 1445.783536 H2O kmol/hr 3.132724175 H2SO4 kmol/hr 0 H2 kmol/hr 0 O2 kmol/hr 902.0699375 SO2 kmol/hr 540.5808747 SO3 kmol/hr 0 H3O+ kmol/hr 0 HSO3- kmol/hr 0 HSO4- kmol/hr 0 SO3-- kmol/hr 0 SO4-- kmol/hr 0 Mole Fractions H2O 0.0021668 H2SO4 0 H2 0 O2 0.623931532

46

SO2 0.373901667 SO3 0 H3O+ 0 HSO3- 0 HSO4- 0 SO3-- 0 SO4-- 0 Mass Flows kg/hr 63553.79804 H2O kg/hr 56.43690317 H2SO4 kg/hr 0 H2 kg/hr 0 O2 kg/hr 28865.15552 SO2 kg/hr 34632.20562 SO3 kg/hr 0 H3O+ kg/hr 0 HSO3- kg/hr 0 HSO4- kg/hr 0 SO3-- kg/hr 0 SO4-- kg/hr 0 Mass Fractions H2O 0.000888018 H2SO4 0 H2 0 O2 0.454184587 SO2 0.544927395 SO3 0 H3O+ 0 HSO3- 0 HSO4- 0 SO3-- 0 SO4-- 0 Volume Flow cum/hr 3015.526169

47

Table A. 8 The liquid stream entering the separation section

Property Units Liquid Stream Phase Liquid Temperature C 39.96349707 Pressure bar 11.76 Molar Vapor Fraction 0 Molar Liquid Fraction 1 Molar Solid Fraction 0 Mass Vapor Fraction 0 Mass Liquid Fraction 1 Mass Solid Fraction 0 Molar Enthalpy kcal/mol -69.14762547 Mass Enthalpy kcal/kg -3057.264421 Molar Entropy cal/mol-K -35.73324893 Mass Entropy cal/gm-K -1.579895041 Molar Density kmol/cum 47.04293471 Mass Density kg/cum 1063.992767 Enthalpy Flow Gcal/hr -883.7194407 Average MW 22.61748281 Mole Flows kmol/hr 12780.18493 H2O kmol/hr 11470.39433 H2SO4 kmol/hr 2.32E-15 H2 kmol/hr 0 O2 kmol/hr 0.169763076 SO2 kmol/hr 1203.123232 SO3 kmol/hr 1.64E-28 H3O+ kmol/hr 53.248809 HSO3- kmol/hr 53.24823406 HSO4- kmol/hr 0.000536284 SO3-- kmol/hr 1.53E-05 SO4-- kmol/hr 4.03E-06 Mole Fractions H2O 0.897513956 H2SO4 1.81E-19 H2 0 O2 1.33E-05 SO2 0.094139736 SO3 1.29E-32 H3O+ 0.004166513

48

HSO3- 0.004166468 HSO4- 4.20E-08 SO3-- 1.20E-09 SO4-- 3.15E-10 Mass Flows kg/hr 289055.6128 H2O kg/hr 206642.3656 H2SO4 kg/hr 2.27E-13 H2 kg/hr 0 O2 kg/hr 5.432214713 SO2 kg/hr 77077.84924 SO3 kg/hr 1.32E-26 H3O+ kg/hr 1012.934522 HSO3- kg/hr 4316.977573 HSO4- kg/hr 0.052058249 SO3-- kg/hr 0.001224945 SO4-- kg/hr 0.000386914 Mass Fractions H2O 0.714887919 H2SO4 7.86E-19 H2 0 O2 1.88E-05 SO2 0.266654048 SO3 4.56E-32 H3O+ 0.003504289 HSO3- 0.014934765 HSO4- 1.80E-07 SO3-- 4.24E-09 SO4-- 1.34E-09 Volume Flow cum/hr 271.6706559

49

Table A. 9 Outlet stream after oxygen separation

Property Units Oxygen stream Phase Vapor Temperature C 26.88771605 Pressure bar 11.4 Molar Vapor Fraction 1 Molar Liquid Fraction 0 Molar Solid Fraction 0 Mass Vapor Fraction 1 Mass Liquid Fraction 0 Mass Solid Fraction 0 Molar Enthalpy kcal/mol -0.205062615 Mass Enthalpy kcal/kg -6.417987594 Molar Entropy cal/mol-K -4.808721623 Mass Entropy cal/gm-K -1.51E-01 Molar Density kmol/cum 0.460825347 Mass Density kg/cum 14.72393791 Enthalpy Flow Gcal/hr -0.184131891 Average MW 3.20E+01 Mole Flows kmol/hr 897.9300849 H2O kmol/hr 3.054659771 H2SO4 kmol/hr 2.84E-27 H2 kmol/hr 0.00E+00 O2 kmol/hr 8.95E+02 SO2 kmol/hr 8.73E-05 SO3 kmol/hr 3.51E-29 H3O+ kmol/hr 0.00E+00 HSO3- kmol/hr 0 HSO4- kmol/hr 0.00E+00 SO3-- kmol/hr 0 SO4-- kmol/hr 0.00E+00 Mole Fractions H2O 0.00340189 H2SO4 3.16E-30 H2 0.00E+00 O2 9.97E-01 SO2 9.72E-08 SO3 3.91E-32 H3O+ 0.00E+00 HSO3- 0

50

HSO4- 0 SO3-- 0 SO4-- 0.00E+00 Mass Flows kg/hr 28689.9731 H2O kg/hr 55.03055107 H2SO4 kg/hr 2.78E-25 H2 kg/hr 0 O2 kg/hr 28634.93696 SO2 kg/hr 0.005591474 SO3 kg/hr 2.81E-27 H3O+ kg/hr 0.00E+00 HSO3- kg/hr 0 HSO4- kg/hr 0.00E+00 SO3-- kg/hr 0 SO4-- kg/hr 0.00E+00 Mass Fractions H2O 0.001918111 H2SO4 9.70E-30 H2 0.00E+00 O2 9.98E-01 SO2 1.95E-07 SO3 9.79E-32 H3O+ 0 HSO3- 0 HSO4- 0 SO3-- 0 SO4-- 0 Volume Flow cum/hr 1948.525815

51

Table A. 10 Design specifications of the Oxygen distillation column

Property Value Units Number of Trayed/Packed stages 15 Total height 9.144 meter Total head loss 1.638571554 meter Total pressure drop 0.162438708 bar Number of sections 2 Number of diameters 2

Start End Diameter Section Internals Tray Section Limiting Stage Stage Height Type Type Pressure Stage Drop 1 9 1.6152 m 5.4864 m TRAY SIEVE 0.08684 9 bar 10 15 2.2138 m 3.6576 m TRAY BUBBLE- 0.07559 15 CAP bar

Table A. 11 Design Specifications of the Sulphur Dioxide distillation column

Property Value Units Number of Trayed/Packed stages 13 Total height 7.9248 meter Total head loss 1.22384896 meter Total pressure drop 0.114322 bar Number of sections 1 Number of diameters 1

Start End Diameter Section Internals Tray Section Limiting Stage Stage Height Type Type Pressure Stage Drop 2 14 4.9206 m 7.9248 m TRAY SIEVE 0.11432 2 bar

52

Control Loop Tuning

Table A. 12 Temperature control on reactors

Set point 1350 oC Gain 5 Integral time 20

Table A. 13 Sump level control for oxygen separation

Set point 2.8375 m Gain 30 Integral time 30

Table A. 14 Pressure control for oxygen separation

Set point 11.4 Gain 20 Integral time 12

53

Table A. 15 Sump level control for Sulphur dioxide separation

Set point 6.15 m Gain 50 Integral time 30

Table A. 16 Pressure control for Sulphur dioxide separation

Set point 1.2 bar Gain 20 Integral time 12

54

Aspen plus control panel for simulation run ->Processing input specifications ...

INFORMATION THERE ARE HENRY COMPONENTS DEFINED IN THIS CASE, THE BINARY DATABANK WILL BE SEARCHED AUTOMATICALLY FOR ANY AVAILABLE HENRY CONSTANTS.

THE PAIR PARAMETERS FOR ELECNRTL OPTION SET HAS BEEN RETRIEVED FROM DATABANK ENRTL-RK.

INFORMATION PURE COMPONENT PARAMETERS FOR SOME COMPONENTS ARE RETRIEVED FROM DATABANK ELECPURE. THESE PARAMETERS ARE USED WITH ELECNRTL METHOD. UNLESS YOU ENTER YOUR OWN PARAMETERS IN PROP-DATA PARAGRAPHS. PARAMETER MW RETRIEVED FOR COMPONENT H2SO4

PARAMETER PC RETRIEVED FOR COMPONENT H2SO4

PARAMETER TC RETRIEVED FOR COMPONENT H2SO4

PARAMETER ZC RETRIEVED FOR COMPONENT H2SO4

PARAMETER RKTZRA RETRIEVED FOR COMPONENT H2SO4

PARAMETER VC RETRIEVED FOR COMPONENT H2SO4

PARAMETER DGFORM RETRIEVED FOR COMPONENT H2SO4

PARAMETER DHFORM RETRIEVED FOR COMPONENT H2SO4

PARAMETER OMEGA RETRIEVED FOR COMPONENT H2SO4

PARAMETER PLXANT RETRIEVED FOR COMPONENT H2SO4

PARAMETER CPAQ0 RETRIEVED FOR COMPONENT SO4--

PARAMETER THRSWT(ELEMENT/3) RETRIEVED FOR COMPONENT H2SO4

PARAMETER CPAQ0 RETRIEVED FOR COMPONENT HSO4-

55

PARAMETER PLXANT RETRIEVED FOR COMPONENT H2O

PARAMETER THRSWT(ELEMENT/3) RETRIEVED FOR COMPONENT H2O

STRUCTURE FOR COMPONENT H2O HAS NOT BEEN DEFINED. PCES CANNOT USE GROUP-CONTRIBUTION METHODS TO ESTIMATE MISSING PROPERTIES USE THE STRUCTURES PARAGRAPH TO DEFINE STRUCTURES OF THIS COMPONENT.

STRUCTURE FOR COMPONENT H2SO4 HAS NOT BEEN DEFINED. PCES CANNOT USE GROUP-CONTRIBUTION METHODS TO ESTIMATE MISSING PROPERTIES USE THE STRUCTURES PARAGRAPH TO DEFINE STRUCTURES OF THIS COMPONENT.

STRUCTURE FOR COMPONENT H2 HAS NOT BEEN DEFINED. PCES CANNOT USE GROUP-CONTRIBUTION METHODS TO ESTIMATE MISSING PROPERTIES USE THE STRUCTURES PARAGRAPH TO DEFINE STRUCTURES OF THIS COMPONENT.

STRUCTURE FOR COMPONENT O2 HAS NOT BEEN DEFINED. PCES CANNOT USE GROUP-CONTRIBUTION METHODS TO ESTIMATE MISSING PROPERTIES USE THE STRUCTURES PARAGRAPH TO DEFINE STRUCTURES OF THIS COMPONENT.

STRUCTURE FOR COMPONENT SO2 HAS NOT BEEN DEFINED. PCES CANNOT USE GROUP-CONTRIBUTION METHODS TO ESTIMATE MISSING PROPERTIES USE THE STRUCTURES PARAGRAPH TO DEFINE STRUCTURES OF THIS COMPONENT.

STRUCTURE FOR COMPONENT SO3 HAS NOT BEEN DEFINED. PCES CANNOT USE GROUP-CONTRIBUTION METHODS TO ESTIMATE MISSING PROPERTIES USE THE STRUCTURES PARAGRAPH TO DEFINE STRUCTURES OF THIS COMPONENT.

STRUCTURE FOR COMPONENT H3O+ HAS NOT BEEN DEFINED. PCES CANNOT USE GROUP-CONTRIBUTION METHODS TO ESTIMATE MISSING PROPERTIES USE THE STRUCTURES PARAGRAPH TO DEFINE STRUCTURES OF THIS COMPONENT.

STRUCTURE FOR COMPONENT HSO3- HAS NOT BEEN DEFINED. PCES CANNOT USE GROUP-CONTRIBUTION METHODS TO ESTIMATE MISSING

56

PROPERTIES USE THE STRUCTURES PARAGRAPH TO DEFINE STRUCTURES OF THIS COMPONENT.

STRUCTURE FOR COMPONENT HSO4- HAS NOT BEEN DEFINED. PCES CANNOT USE GROUP-CONTRIBUTION METHODS TO ESTIMATE MISSING PROPERTIES USE THE STRUCTURES PARAGRAPH TO DEFINE STRUCTURES OF THIS COMPONENT.

STRUCTURE FOR COMPONENT SO3-- HAS NOT BEEN DEFINED. PCES CANNOT USE GROUP-CONTRIBUTION METHODS TO ESTIMATE MISSING PROPERTIES USE THE STRUCTURES PARAGRAPH TO DEFINE STRUCTURES OF THIS COMPONENT.

STRUCTURE FOR COMPONENT SO4-- HAS NOT BEEN DEFINED. PCES CANNOT USE GROUP-CONTRIBUTION METHODS TO ESTIMATE MISSING PROPERTIES USE THE STRUCTURES PARAGRAPH TO DEFINE STRUCTURES OF THIS COMPONENT.

* WARNING IN PHYSICAL PROPERTY SYSTEM UNSYMMETRIC ELECTROLYTE NRTL MODEL GMENRTLQ HAS MISSING PARAMETERS: Dielectric constant (CPDIEC) MISSING FOR SO3 . CPDIEC OF WATER WILL BE ASSUMED.

* WARNING IN PHYSICAL PROPERTY SYSTEM NRTL BINARY PARAMETERS FOR ALL COMPONENT PAIRS ARE ZERO, YOUR RESULTS MAY NOT BE ACCURATE. PLEASE REVIEW AND PROVIDE BINARY PARAMETERS AS APPROPRIATE.

Flowsheet Analysis :

COMPUTATION ORDER FOR THE FLOWSHEET: T1 P1 SHXCONC VAL1 SFLCONC SHXD1 P2 BAYOH DECOMP BAYOC SHXRCVRB SCONDENS SFLDECO SHXD2 FLSEP1 CMPRSEP1 SHXSEP1 CMPRSEP2 SHXSEP2 FLSEP2 PSEP1 SHXSEP3 TSEP2 VAL3 KSEP SPSEP1 HXSEP5 VSEP1 FLSEP3 VAL4 KDESOR PSEP2 HXSEP4 MIXSEP1 HXPREP2 VAL2 HPBFW LPBFW MPBFW

->Calculations begin ...

Block: T1 Model: MIXER

57

Block: P1 Model: PUMP

Block: SHXCONC Model: HEATER

Block: VAL1 Model: VALVE

Block: SFLCONC Model: FLASH2

Block: SHXD1 Model: HEATER

Block: P2 Model: PUMP

Block: BAYOH Model: RSTOIC

Block: DECOMP Model: RSTOIC

Block: BAYOC Model: HEATER

Block: SHXRCVRB Model: HEATER

Block: SCONDENS Model: HEATER

Block: SFLDECO Model: FLASH2

Block: SHXD2 Model: HEATER

Block: FLSEP1 Model: FLASH2

Block: CMPRSEP1 Model: COMPR

Block: SHXSEP1 Model: HEATER

Block: CMPRSEP2 Model: COMPR

Block: SHXSEP2 Model: HEATER

Block: FLSEP2 Model: FLASH2

Block: PSEP1 Model: PUMP

58

Block: SHXSEP3 Model: HEATER

Block: TSEP2 Model: MIXER

Block: VAL3 Model: VALVE

Block: KSEP Model: RADFRAC

Convergence iterations: OL ML IL Err/Tol 1 1 10 7233.0 2 1 10 5376.2 3 1 10 4174.3 4 1 10 2083.1 5 1 10 221.78 6 1 10 924.18 7 1 10 293.28 8 1 10 445.25 9 1 10 416.97 10 1 10 91.714 11 1 10 600.58 12 1 6 15.008 13 1 10 91.213 14 1 2 3.0478 15 1 1 0.30009

Block: SPSEP1 Model: FSPLIT

Block: HXSEP5 Model: HEATER

Block: VSEP1 Model: VALVE

Block: FLSEP3 Model: FLASH2

Block: VAL4 Model: VALVE

Block: KDESOR Model: RADFRAC

Convergence iterations: OL ML IL Err/Tol 1 1 3 1467.4

59

2 1 2 209.92 3 1 1 42.254 4 1 1 5.5939 5 1 2 0.40024

Block: PSEP2 Model: PUMP

Block: HXSEP4 Model: HEATER

Block: MIXSEP1 Model: MIXER

Block: HXPREP2 Model: HEATER

Block: VAL2 Model: VALVE

Utility HPBFW Model: GENERAL

Utility LPBFW Model: GENERAL

Utility MPBFW Model: GENERAL

->Generating block results ...

Block: BAYOC Model: HEATER

Block: SHXRCVRB Model: HEATER

Block: SCONDENS Model: HEATER

Block: SHXCONC Model: HEATER

Block: P1 Model: PUMP

Block: P2 Model: PUMP

Block: SHXD1 Model: HEATER

Block: SHXD2 Model: HEATER

Block: SHXSEP1 Model: HEATER

60

Block: SHXSEP2 Model: HEATER

Block: SHXSEP3 Model: HEATER

Block: PSEP1 Model: PUMP

Block: HXSEP5 Model: HEATER

Block: PSEP2 Model: PUMP

Block: HXSEP4 Model: HEATER

Block: HXPREP2 Model: HEATER

->Simulation calculations completed ...

*** No Warnings were issued during Input Translation ***

*** Summary of Simulation Errors ***

Physical Property System Simulation Terminal Errors 0 0 0 Severe Errors 0 0 0 Errors 0 0 0 Warnings 2 0 0 ** ERROR CHEMISTRY (GLBAL) NOT CONVERGED; RMSERR= 0.2224 ; SUM OF DELX = 0.3945 .

61