Simulation and Optimization by Ethanol Autothermal Reformer for Fuel Cell Application by SyafiQ

7/22/2019 Simulation and Optimization by Ethanol Autothermal Reformer for Fuel Cell Application by SyafiQ

1/87

SIMULATION AND OPTIMIZATION OF ETHANOL AUTOTHERMALREFORMER FOR FUEL CELL APPLICATIONS

MUHAMAD SYAFIQ BIN ADAM

UNIVERSITI TEKNOLOGI MALAYSIA


2/87

PSZ 19:16 (Pind. 1/97)

UNIVERSITI TEKNOLOGI MALAYSIA

BORANG PENGESAHAN STATUS TESIS

JUDUL: SIMULATION AND OPTIMIZATION OF ETHANOL

AUTOTHERMAL REFORMER FOR FUEL CELL APPLICATION

SESI PENGAJIAN: 2006/2007

Saya: MUHAMAD SYAFIQ BIN ADAM

(HURUF BESAR)

mengaku membenarkan tesis (PSM/Sarjana/Doktor Falsafah)* ini disimpan di Perpustakaan Universiti

Teknologi Malaysia dengan syarat-syarat kegunaan seperti berikut :

1. Tesis adalah hakmilik Universiti Teknologi Malaysia2. Perpustakaan Universiti Teknologi Malaysia dibenarkan membuat salinan untuk tujuan pengajiansahaja.

3. Perpustakaan dibenarkan membuat salinan tesis ini sebagai bahan pertukaran antara institusipengajian tinggi.

4. **Sila tandakan (3 )SULIT (Mengandungi maklumat yang berdarjah keselamatan atau

kepentingan Malaysia seperti yang termaktub di dalam AKTA

RAHSIA RASMI 1972)TERHAD (Mengandungi maklumat TERHAD yang telah ditentukan

oleh organisasi/badan di mana penyelidikan dijalankan)

TIDAK TERHADDisahkan oleh

WW

P O___

(TANDATANGAN PENULIS) (TANDATANGAN PENYELIA)

Alamat Tetap: No. 32, Jln SG 10/11, Engr. Mohd. Kamaruddin bin Abd Hamid Taman Seri Gombak, Nama Penyelia68100 Batu Caves, Selangor.

Tarikh : 15th November 2006 Tarikh : 15th November 2006

3

CATATAN: * Potong yang tidak berkenaan

* * Jika Tesis ini SULIT atau TERHAD, sila lampirkan surat daripada pihak berkuasa/organisasi berkenaan dengan menyatakan sekali sebab dan tempoh tesis ini perlu

dikelaskan sebagai SULIT atau TERHAD.

Tesis dimaksudkan sebagai tesis bagi Ijazah Doktor Falsafah dan Sarjana secarapenyelidikan, atau disertai bagi pengajian secara kerja kursus dan penyelidikan atau

Laporan Projek Sarjana Muda (PSM).


3/87

I hereby declare that I have read this report and in my opinion this report is

sufficient in terms of scope and quality for the award of the degree of Bachelor of

Engineering (Chemical).

Signature : .

Name of Supervisor : Engr. Mohd. Kamaruddin Abd. Hamid

Date : 15th November 2006


4/87

SIMULATION AND OPTIMIZATION OF ETHANOL AUTOTHERMAL

REFORMER FOR FUEL CELL APPLICATIONS

MUHAMAD SYAFIQ BIN ADAM

A report submitted in partial fulfilment of the

requirements for the award of the degree of

Bachelor of Engineering (Chemical)

Faculty of Chemical Engineering and Natural Resources Engineering

Universiti Teknologi Malaysia

NOVEMBER 2006


5/87

ii

I declare that this thesis entitled Simulation and Optimization of Ethanol

Autothermal Reformer for Fuel Cell Applications is the result of my own research

except as cited in the references. The thesis has not been accepted for any degree andis not concurrently submitted in candidature of any other degree.

Signature : ..

Name : Muhamad Syafiq bin Adam

Date : 15th November 2006


6/87

iii

To my beloved mother and father


7/87

iv

ACKNOWLEDGEMENTS

Alhamdullilah. Finally my first thesis was finished. Thanks to God because

all of His merciless and the knowledge that was given, I did my work successfully.

An honour and respect to the Prophet Muhamad SAW. Peace be upon him.

I would like to thank to my supervisor; En. Mohd Kamaruddin. A word thank

you cant describe all of his guidance and encouragement that showed to me during

the progresses of this thesis. All the knowledge that I learned from him will come in

handy at the future.

I also wanted to thank my family for the moral support. Especially to my

parents, for effort and support those drive me to this level. This thesis was dedicated

to them.

Finally, to all of my friends that contributed to this thesis. Thank you very

much to all of them that help me either direct or indirect.


8/87

v

ABSTRACT

Fuel cell application from hydrogen was one of alternative energy that

being studied and widely accepted in industry. This case study focused on

optimization of hydrogen production for fuel cell applications. In this case study,

ethanol was chosen as a raw material and with autothermal reforming as a process ofproduce hydrogen. Using a commercial dynamic flow sheeting software, HYSYS

3.2, the process of hydrogen production was successfully simulated. In this research,

fuel processor consists of an autothermal reactor, three water gas shift reactors and a

preferential oxidation reactor was successfully developed. The purpose of this case

study is to identify the effect of various operating parameters such as air-to-fuel

(A/F) ratio and steam-to-fuel (S/F) ratio to get the optimum hydrogen production

while made carbon monoxide lower than 10 ppm. From the results, an optimum A/F

and S/F ratio are 5.5 and 1.5, respectively to produce 34 % of hydrogen and 10.055

ppm of CO. Under these optimum conditions, 83.6% of fuel processor efficiency was

achieved.


9/87

vi

ABSTRAK

Penggunaan sel bahan api daripada hidrogen merupakan salah satu tenaga

yang masih dikaji dan diterima dalam kebanyakan industri. Kajian ini memfokuskan

tentang pengeluaran hidrogen untuk penggunaan sel bahan api secara dinamik.

Dalam kajian ini, etanol dipilih sebagai bahan mentah dan pembentukan autoterma(auto thermal reforming) merupakan proses untuk menghasilkan hidrogen. Dengan

menggunakan perisian HYSYS 3.2, proses pengeluaran hidrogen ini berjaya

dilakukan secara simulasi. Dalam kajian ini, pemproses minyak mengandungi reaktor

autoterma,, tiga reaktor anjakan air gas dan reaktor pilihan pengoksidaan telah

berjaya dihasilkan. Kajian ini bertujuan untuk mengenalpasti kesan pengandelaian

parameter yang berlainan seperti ratio udara-ke-minyak (A/F) dan ratio stim-ke-

minyak (S/F) untuk mendapatkan pengeluaran hydrogen yang optimum sementara

CO dihasilkan rendah dari 10 ppm. Daripada keputusan ujikaji, nilai ratio A/F dan

S/F yang optima adalah 5.5 dan 1.5 masing-masing. Dengan ratio tersebut,34%

hydrogen dan 10.055 ppm CO dapat dihasilkan. Dibawah keadaan pengoptimaan ini,

sebanyak 83.6 % kecekapan pemproses minyak didapati.


10/87

vii

LIST OF CONTENTS

CHAPTER TITLE PAGE

Tittle Page i

Declaration ii

Dedication iii

Acknowledgements iv

Abstract v

Abstrak vi

List of Contents vii

List of Figures xi

List of Tables xiii

List of Symbols xiv

I INTRODUCTION

1.1 Background Research 1

1.2 Problems Statement 2

1.3 Research Objective 2

1.4 Scopes of study 3

1.5 Thesis Organizations 4

II LITERATURE REVIEW

2.1 Introduction 6

2.2 Hydrogen Production for Fuel Cell Application in

General 6

2.2.1 Natural Gas 7


11/87

viii

2.2.1.1 Methane 7

2.2.1.2 Ethane 8

2.2.1.3 Propane 9

2.2.1.4 Butane 9

2.2.2 Alcohol 9

2.2.2.1 Methanol 10

2.2.2.2 Ethanol 10

2.2.2.3 Propanol 11

2.2.3 Petroleum Fractional 12

2.2.3.1 Kerosene 12

2.2.3.2 Gasoline 122.2.3.3 Diesel 13

2.3 Hydrogen Production for Fuel Cell from Ethanol 132.3.1 Steam Reforming 142.3.2 Partial Oxidation 15

2.4 Steam Reforming of Ethanol for Hydrogen

Production 16

2.5 Optimization simulation of Hydrogen Production 17

2.6 Summary 17

III METHODOLOGY

3.1 Research Tools 18

3.1.1 Aspen HYSYS 18

3.2 Research Activities 19

3.2.1 Data Collection 193.2.2 Base Case Stoichiometry 19

3.2.3 Base Case Validation 21

3.2.4 Auto-thermal Reactor Optimization 21

3.2.5 Heat Integration 21

3.2.6 Carbon Monoxide Clean Up 22

3.2.6.1 Water Gas Shift 22

3.2.6.2 Preferential Oxidation 22

3.2.7 Plant Wide Optimization 23


12/87

ix

3.2.7.1 ATR Optimization 23

3.2.7.2 Water Gas Shift Optimization 23

3.2.7.3 Preferential Oxidation Optimization 24

3.2.8 Temperature and Component Profile 24

3.2.9 Fuel Processor Efficiency 24

3.3 Summary 25

IV SIMULATION AND OPTIMIZATION OF HYDROGEN

PRODUCTION PLANT FROM ETHANOL FOR FUEL CELL

APPLICATION

4.1 Process Description of Hydrogen Production from

Ethanol 26

4.2 Modelling and Simulation of Hydrogen ProductionFrom Ethanol for Fuel Cell 27

4.2.1 Thermodynamic Properties 31

4.2.2 Physical Properties 32

4.2.3 Integration Algorithm 33

4.2.4 Mathematical Modelling of the Reactor

Operating 33

4.2.4.1 Linear and Non-Linear System 33

4.2.4.2 Material Balance 34

4.2.4.3 Component Balance 35

4.2.4.4 Energy Balance 36

4.2.5 Degree of Freedom Analysis 38

4.2.6 Analysis of Optimization Response 384.3 Summary 39

V RESULTS AND DISCUSSION

5.1 Results for Base Case Study 40

5.2 Results for Validation 43

5.3 Results for Heat Integration 44

5.4 Results for Carbon Monoxide Clean Up 465.4.1 Water Gas Shift 46


13/87

x

5.4.2 Preferential Oxidation 47

5.5 Plant Wide Optimization 48

5.5.1 ATR Optimization 49

5.5.2 Water Gas Shift Optimization 50

5.5.3 Preferential Oxidation Optimization 53

5.6 Temperature Profile of fuel Processor System 55

5.7 Component Profile of the Fuel Processor System 56

5.8 Fuel Processor Efficiency 57

5.9 Summary 57

VI CONCLUSION AND RECOMMENDATIONS

6.1 Summary 58

6.2 Conclusion 59

6.3 Recommendation 59

REFERENCES 61

APPENDIX

APPENDIX A Final result of simulation HYSYS 3.2 66


14/87

xi

LIST OF FIGURES

FIGURE NO. TITLE PAGE

3.1 Algorithm for methodology. 25

4.1 The operation conditions for the major unit

operation 27

4.2 The whole plant system by Aspen HYSYS 3.2 29

4.3 HYSYS simulation environment 30

4.4 Reactor operating 35

4.5 Block diagram of the simulation of hydrogenplant using Aspen HYSYS 3.2 39

5.1 Process flow diagram of the base case 41

5.2 The heater attachment on the ATR reactor 45

5.3 The heaters at the feed streams were exchangewith the heat exchanger 46

5.4 The WGS reactor 47

5.5 The PROX reactor 48


15/87

xii

5.6 Temperature of ATR vapour for varies air feedmolar flow 49

5.7 Molar flow of CO and H2 effluent for variesair feed molar flow 50

5.8 Molar flow of CO and H2 effluent for varieswater feed molar flow 51

5.9 Temperature to ATR outlet for varies water feedmolar flow 52

5.10 CO Molar flow in PROX effluent for varies airfeed molar flow 54

5.11 Temperature profile for the whole unit operation 55

5.12 H2 and CO profile for the whole unit operation 56


16/87

xiii

LIST OF TABLES

TABLE NO. TITLE PAGE

4.1 Physical property of the component 32

5.1 Molar flow of ATR effluent for base case 43

5.2 Validation for simulation effluent comparewith calculated effluent 44

5.3 Effluent molar flow after water gas shiftreaction for each reactor 47

5.4 Effluent molar flow after preferentialoxidation reaction 48

5.5 Molar flow of the effluent before optimizationfor ATR,HTS, MTS and LTS. 52

5.6 Molar flow of the effluent after optimizationfor ATR,HTS, MTS and LTS. 53

5.7 Molar flow of the effluent before and afteroptimization for PROX 54


17/87

xiv

LIST OF SYMBOLS

A Heat transfer area

a Parameter, cubic equation of state

b Parameter, cubic equation of state

C ConcentrationF Volumetric flow rate

g Local acceleration of gravity

H Molar or specific enthalpy

h Step size

k Kinetic energy

m Mass flow rate

MW Molecular weight

Nm Number of independent variables

Nom Number of manipulated variables with no steady state effect

Noy Number of variables that need to be controlled fromNm

Nss Number of variables needed to be specified

P Absolute pressure

Po Reference pressure

Pci Critical pressure, species iPri Reduced pressure, species i

Q Heat

Qr Heat generated by reaction

R Universal gas constant

r Rate of reaction

t Time

u Internal energy

V Volume


18/87

xv

Y Process Variable

Greek letters

Function, cubic equation of state

Error

Viscosity

Density

Potential energy

Acentric factor

Abbreviations

ATR Auto thermal reforming

ca. at approximate

CO Carbon Monoxide

CO2 Carbon Dioxide

et al. et alias: and others

etc. et cetera

H2 Hydrogen

HTS High Temperature Shift

LTS Lower Temperature Shift

PROX Preferential Oxidation

MTS Medium Temperature Shift

WGS Water Gas Shift


19/87

CHAPTER

INTRODUCTION

1.1 Background Research

Hydrogen was expected to become an important energy carrier for

sustainable energy consumption with a significantly reduced impact on the

environment. Hydrogens benefit and disadvantages differ from the fossil fuels

common place in advanced energy utilizing society. It is because characteristics of

hydrogen that cheap, easy to obtain, high efficiency, virtually silent operation and

less pollutant emissions. (Fuel cell store website, 2006)

From that perspective, researcher over the world tries to make use the

hydrogen as an alternative energy by converting into fuel cell. Hydrogen as fuel cell

technology currently needed in large quantities, and is projected to be the fuel of

choice for a number of advanced technologies that are being pursued. Fuel cell will

supply the energy that a global society requires to support the growing number ofpeople that demanding on fuel cell technology using hydrogen. (Fuel cell store

website, 2006)

For that purpose, some fossil fuels which have high hydrogen to oxygen ratio

were the best candidates to produce hydrogen. The more hydrogen present and the

fewer extraneous compounds was the idea to get it. One of the methods which

commonly being used was the steam reforming. Other established methods includepartial oxidation of residual oil, coal gasification, water electrolysis and etc. The new


20/87

2

technologies such as high-temperature electrolysis of steam, thermal cracking of

natural gas, thermo chemical water splitting, solar photovoltaic water electrolysis,

and plasma decomposition of water is still investigated its efficiency. These

technologies can be classified as thermal, thermo chemical, electrochemical,

photochemical, and plasma chemical methods. (Fuel cell store website, 2006)

Seven common fuels are the postulated hydrogen sources studied in this work

alcohol, natural gas, gasoline, diesel fuel, aviation jet fuel, and hydrogen itself.

Among the bio-fuel candidates for carriers of hydrogen, ethanol is of particular

interest because its low toxicity, low production costs, the fact that is a relative

clean fuel in terms of composition, relatively high hydrogen content and availabilityand ease of handling. Hydrogen can be obtained directly from ethanol by two main

processes; partial oxidation and steam reforming. (Fuel cell store website, 2006)

1.2 Problem Statement

In reality, chemical plants are never truly at steady state. Feed and

environmental disturbances, heat exchanger fouling, and catalytic degradation

continuously upset the conditions of a smooth running process. Optimization

simulation can help researcher to make better design, optimize, and operate process

or refining plant. In this research, ethanol is the main focus to study the steady state

behaviour. Furthermore, the optimization is the main case study that will make more

yield selectivity hydrogen. The important of this study is to identify design

parameters and also to estimate fuel processor efficiency.

1.3 Research Objectives

The main objective of this research is to simulate and optimize the hydrogen

production plant for fuel cell application using ethanol via autothermal reformer.


21/87

3

1.4 Scope of Study

To achieve above objective, several scopes has be drawn:

i. Base case simulation development

By using Aspen HYSYS 3.2, hydrogen production simulation plant was

being developed with data from Akande et al. (2005)

ii. Base case simulation validation

From base case simulation that being developed with Aspen HYSYS 3.2, it

was validated using theoretically data from total reactions stoichiometrycoefficient.

iii. ATR optimization

ATR was optimizing by optimized the air feed molar that enter the ATR

while monitoring the production of hydrogen and carbon monoxide (CO) in a

certain range of temperature.

iv. Heat integration

This system is used to increase the efficiency of the plant by using heat

exchanger to cool down the ATR vapour out with the hot stream from the

feed.

v. Carbon monoxide clean up

Carbon monoxide that produced by the total reaction in ATR need to be

reduced their concentration by introducing water gas shift reaction and

preferential oxidation reactions.

a. Water gas shift

Equilibrium reactors were placed to the plant to convert CO into carbon

dioxide (CO2). Three reactors were needed for conversion with water

gas shift (WGS) reaction as the main reaction.


22/87

4

b. Preferential oxidation

To maximum reducing CO, preferential oxidation (PROX) reaction was

introduced.

vi. Plant wide optimization

It was develop to optimized all the reactors used in the plant developed using

Aspen HYSYS 3.2 and to reduced CO concentration to the specific

requirement.

a. ATR optimization

Its used to optimize the ATR temperature outlet for heat integration.

b. Water gas shift optimization

Its used to optimize water molar flow to the ATR and reduces the CO

concentration with WGS reaction.

c. Preferential oxidation optimization

It was formed to maintain the amount of air into PROX reactor that

reduced the CO concentration to the specification.

vii. Temperature and component profile

The profile of temperature and components for every unit operations involve

in this research was analyzed.

1.5 Thesis Organizations

This thesis involves the conclusion of the several tasks to achieve the

objective. Chapter Two is discuss about the literature survey that related in synthesis

of hydrogen for fuel cell applications. In this chapter, internal researched of

hydrogen production using ethanol by autothermal reforming was been concentrated.

This chapter is the major chapter because the development of the of hydrogenproduction are based on the literature survey that we had researched.


23/87

5

Chapter Three is about the methodology for the methods that we need in

scope. Fundamentally, there are five methods that we carried out. The next chapter;

Chapter Four, is optimization simulation of hydrogen production plant from ethanol

for fuel cell application. We are using Aspen HYSYS 3.2 as a simulator to simulate

the plant.

Chapter Five is the results and discussion based on the methodology that we

use and developed from chapter four. Finally, Chapter Six is the conclusion all what

we have done in this entire thesis.


24/87

6

CHAPTER II

LITERATURE REVIEW

2.1 Introduction

In this chapter, a general hydrogen production using natural gas, alcohol and

petroleum fractional of gas as an input will be reviewed. It is meant to provide a list

of hydrogen had been produced by a specific class of hydrocarbon such as natural

gas and alcohol. Methane, ethane, propane and butane are some example of natural

gas. (Fuel cell store website, 2006)

Alcohol can also be used to produce hydrogen with a different condition

either used a same method or not. The different between those methods was higher

selectivity of hydrogen. Methanol, ethanol and propanol were some of them.

Petroleum fractional such as kerosene, gasoline, and diesel too will produce

hydrogen by using same method like reforming. From this review, some significant

journal will be taken as references. (Fuel cell store website, 2006)

2.2 Hydrogen Production for Fuel Cell Application in General

Fuel cell requires hydrogen as its fuel source for generating power. Hydrogen

used in secondary power units is produced in a fuel processor by the catalytic

reforming of hydrocarbons. Diesel, jet fuel, gasoline, as well as natural gas, are


25/87

7

potential fuels that all have existing infrastructure of manufacture and distribution,

for hydrogen production for fuel cell applications. (Fuel cell store website, 2006)

2.2.1 Natural Gas

The lack of a hydrogen infrastructure and the unsolved hydrogen storage

problem has initiated the development of compact fuel reformers that are able to

produce a hydrogen-rich gas from fuels such as hydrocarbon. Methane, due to its

large abundance and high H: C ratio is an ideal source of hydrogen. Ethane, propane

and methane are the family of natural gas which they produce hydrogen-rich too.

(Liu et al., 2002)

2.2.1.1Methane

Fernandez et al. (2005) discussed and studied the hydrogen production by

sorption enhanced reaction process simulated by a dynamic one-dimensional pseudo-

homogenous model of a fixed-bed reactor, where a hydrotalcite-derived Ni catalyst

has been used as steam reforming catalysts.

Galvita and Sundmacher (2005) said that almost CO-free hydrogen gas, can

be produced by a novel steam reforming process of methane in a fixed bed reactor

which contains two different catalysts layers which go through a periodic

reduction/re-oxidation cycle.

The fluidized bed reactor was proposed by Lee et al. (2004) in order to

overcome the reactor plugging problem due to carbon deposition, which was resulted

in the shut-down of the fixed bed reactor system. Several kinds of activated carbons

were employed as the catalyst to examine the reaction activity.


26/87

8

Oxidized diamond is proposed by Nakagawa et al. (2004) as an effective

catalyst support material for decomposition of methane. Oxidized diamond-

supported Ni catalyst produced a high yield of hydrogen by the decomposition of

methane at 823 K.

Bingue et al. (2004) describes that transient filtration combustion waves

formed in a porous matrix of randomly arranged alumina pellets are studied

experimentally for rich and ultra-rich methane/air waves with oxygen enrichment

and depletion.

2.2.1.2Ethane

The catalytic decomposition of ethane was studied by Chin et al. (2005) over

a Ni/SiO2 catalyst at temperatures ranging between 450 and 650 C.

Wang et al. (2003) proved that formation rates of the more valuablehydrocarbons and hydrogen are remarkably enhanced by selective permeation of

hydrogen product in the membrane reactor. It was also found that formation rate of

methane as a side product is effectively suppressed by selective permeation of

hydrogen though the membrane tubes.

The key reactions forming the higher hydrocarbons involved addition of

radicals to unsaturated bonds (Shebaro et al., 1997). Recent model calculations forassociation reactions in hydrocarbon pyrolysis and flames have emphasized the role

of chemically activated association and isomerization in overcoming entropic

inhibitions, particularly for benzene formation.


27/87

9

2.2.1.3 Propane

Aartun et al. (2005) compared Rh-impregnated alumina foams and metallic

micro channel reactors upon production of hydrogen-rich syngas through shortcontact time catalytic partial oxidation (POX) and oxidative steam reforming (OSR)

of propane.

Resini et al. (2005) compared the both catalyst and suggest the palladium-

based catalyst, the steam reforming of propene is faster and more selective than

steam reforming of propane.

Silberova et al. (2005) investigated partial oxidation and oxidative steam

reforming of propane over 0.01 wt.% Rh/Al2O3 foam catalysts and concluded high

selectivity to hydrogen was obtained for both reactions.

2.2.1.4Butane

Avci et al. (2003) found the major difference between the two catalysts at 648

K, at which Pt-Ni/-Al2O3 showed superior performance in terms of selective

hydrogen production that resulted in lower carbon dioxide and methane formation.

2.2.2 Alcohol

Alcohols as fuel have been proven to be effective in the near complete

elimination of emissions of benzene, olefins, complex hydrocarbons and SO2. In

particular, methanol and ethanol are now seriously considered as a source for fuel-

cell-powered vehicles. While propanol too produce high selectivity of hydrogen with

various support of certain catalysts. (Wanat et al., 2005)


28/87

10

2.2.2.1Methanol

Basile et al. (2005) showed that the methanol reforming (MR) gives methanol

conversions higher than traditional reactors (TRs) at each temperature confirming thegood potential of the membrane reactor device for this interesting reaction system.

Liu et al. (2004) described that prepared catalysts showed high activity and

selectivity towards hydrogen formation and explained their catalytic performances

during oxidative methanol reforming for the production of hydrogen reaction

conditions.

Xu et al. (2004) found that the alkali-leached Ni3Al powders show a high

catalytic activity for the methanol decomposition and made rate of hydrogen

production increases rapidly with increasing reaction temperature.

2.2.2.2Ethanol

Both Vaidya and Rodrigues (2005) said that this production is simple and

cheap and hence steam reforming of ethanol to produce hydrogen for fuel cells is

attractive. The entire process of ethanol steam reforming coupled with selective CO2

removal by chemisorptions will enable production of high-purity H2 and hence is

very promising.

Aupretre et al. (2005) conclude that Rh is the most active metal in the steam

reforming reaction, especially in ethanol steam reforming (ESR)but the conditions

plead in favor of a support that is non-acidic and moderately basic.

A reaction mechanism is proposed by Mattos and Noronha (2005b) to explain

the catalytic tests. The effects of reaction conditions and catalyst reducibility on the


29/87

11

performance of the Pt/CeO2 catalyst in the partial oxidation of ethanol were

described.

H2 production and CO2/COx ratio obtained over Ni-based catalysts supportedon Al2O3 are compared by Fierro et al. (2005) with those obtained over NiCu/SiO2

and Rh/Al2O3 catalysts and suggest that its provided very good activity and

selectivity for ethanol partial oxidation reaction with high selectivity to H2.

A series of Pt catalysts supported on alumina modified by Ce and/or La were

discussed by Navarro et al. (2004) involving the production of hydrogen by oxidative

reforming of ethanol. When both ceria and lanthana were present on the support

substrate the platinumceria interaction was diminished, reducing the promoter effect

in the production of hydrogen by oxidative reforming of ethanol.

2.2.2.3Propanol

CeO2 resulted in the highest selectivity and fairly higher stability for the

steam reforming among the supported Rh catalysts. Mizuno et al. (2003) concluded

that Rh=CeO2 is actually superior to any other catalyst for the steam reforming of

IPA.

Wanat et al. (2005) have shown that different alcohols have very different

selectivity in catalytic partial oxidation at short contact times even at high

temperatures. Rapid adsorption of alcohols as alkoxy species leads to complete

dissociation to H2 and CO. 2-Propanol gave lower conversions and less H2 and CO

than the other alcohols, but produced the most chemicals.


30/87

12

2.2.3 Petroleum Fractional

The Polymer Electrolyte Membrane (PEM) fuel cell requires hydrogen as its

fuel source. In order to avoid storing high-pressure hydrogen, the fuel can begenerated in an onboard fuel processor. For transportation applications, the primary

focus is on reforming gasoline, because a production and distribution infrastructure

already exists. For auxiliary power units, the focus is on reforming both gasoline (for

automotive applications) and diesel (for trucks and heavy-duty vehicles). For

portable power generation, the focus has been on reforming natural gas and liquefied

petroleum gas. (Cheekatamarla and Lane, 2005)

2.2.3.1 Kerosene

The auto thermal reforming of desulphurised kerosene was examined with a

15 kW (based on the lower heating value of Jet fuel) test rig. Lenz andAicher (2005)

successfully performed experiment at steam to carbon ratios of S/C = 1.52.5 and air

to fuel ratios of= 0.240.32.

Suzuki et al. (2000) was discussed about long sustained run of hydrogen

production using HD-kerosene was successfully achieved on the CRI-101CE catalyst

(Ru/CeO2Al2O3). Highly dispersed Ru/Al2O3 catalyst can be obtained by using

ruthenium trichloride and aqueous ammonia in the catalyst preparation.

2.2.3.2Gasoline

A numerical model of a simple reforming system, based on a partial oxidation

process, has been developed by Minutillo (2005) and tested it using the experimental

data of a plasma-assisted reformer. The conversions of methane, propane, heptane,

toluene and gasoline to hydrogen have been investigated and a thermodynamic


31/87

13

analysis of the reforming system has been conducted by means of the AspenPlus

software.

Otsuka et al. (2002) proposed and investigated a new technology usinggasoline as a fuel for solid polymer electrolyte fuel cell through the decomposition of

gasoline range alkanes into hydrogen and carbon and figured the method can supply

high purity hydrogen without CO and CO2.

2.2.4 Diesel

In order to show efficient catalysts for hydrogen generation from diesel

autothermal reforming Cheekatamarla and Lane (2005) showed that bimetallic

catalysts exhibited superior performance to the commercial catalyst and the

monometallic counterparts which showed that the enhanced stability is due to a

strong metalmetal and metalsupport interaction in the catalyst.

The reforming process efficiency has been shown by Tsolakis and Megaritis

(2004) to improve considerably with water addition up to a certain level after which

the adverse effects of the exothermic water gas shift reaction become significant.

Methanol, natural gas, gasoline, diesel fuel, aviation jet fuel, ethanol, and

hydrogen are compared by Brown (2001) for their utility as hydrogen sources for

proton-exchange-membrane fuel cells used in automotive propulsion.

2.3 Hydrogen Production for Fuel Cell from Ethanol

Fuels containing hydrogen generally require a fuel reformer that extracts

the hydrogen from any hydrocarbon fuel; ethanol for example. Ethanol appears as an

attractive alternative to methanol since it is much less toxic, offers a high octane


32/87

14

number, a high heat of vaporization and a low photochemical reactivity. There was

several method of producing hydrogen using ethanol. Steam reforming was the

popular way to produce follow by partial oxidation.

2.3.1 Steam Reforming

Oxidative steam reforming of ethanol for hydrogen production in order to

feed a solid polymer fuel cell (SPFC) has been studied over several catalysts at on

board conditions (a molar ratio of H2O/EtOH and of O2/EtOH equal to 1.6 and 0.68

respectively) and a reforming temperature between 923 and 1073 K. Two Ni (11 and

20 wt.%)/Al2O3 catalysts and five bimetallic catalysts, all of them supported on

Al2O3, were tested by Fiero et al. (2005).

By using high temperatures, low pressures and high water-to-ethanol ratios in

the feed favour hydrogen production. Vaidya and Rodrigues (2005) Ni, Co, Ni/Cu

and noble metal (Pd, Pt, Rh)-supported catalysts to produce hydrogen by using steam

reforming. They said that this entire process of ethanol steam reforming coupled with

selective CO2 removal by chemisorptions will enable production of high-purity H2

and hence is very capable.

Akande et al. (2005) were estimated the effects of catalyst synthesis method

(i.e. precipitation (PT), co-precipitation (CP) and impregnation (IM)), Ni loading and

reduction temperature on the characteristics and performance of Ni/Al2O3 catalystsfor the reforming of crude ethanol for H2 production. The result showed the type of

species generated by the synthesis method, the PT catalysts were more reducible than

the CP and IM catalysts.

Comas et al. (2004) analysed ethanol steam reforming with and without the

presence of CaO as a CO2 sorbent. They founds Both processes show the same

behaviour with pressure and water to ethanol ratio, atmospheric pressure and water to


33/87

15

ethanol relations higher than three are favourable conditions for higher hydrogen

productions without carbon formation.

Sun et al. (2004) proved that the catalyst Ni/Al2O3 exhibits relative loweractivity for ethanol steam reforming and hydrogen selectivity. But they found that the

catalysts Ni/Y2O3 and Ni/La2O3 exhibit relative high activity for ethanol steam

reforming at 250 C with a conversion of ethanol of 81.9% and 80.7%, and a

selectivity of hydrogen of 43.1% and 49.5%, respectively. When temperature

reached 320 C, the conversion of ethanol increased to 93.1% and 99.5% and the

selectivity of hydrogen was 53.2% and 48.5%.

From the endurance tests Freni et al. (2003) founded out at low gas hourly

space velocity (10,000 h-1) for 630 h showed that Ni/MgO catalyst possesses

adequate characteristics to be proposed as an efficient catalytic system for the

production of hydrogen for MCFC by steam reforming of ethanol.

Liguras et al. (2002) found that, under certain reaction conditions, the 5%

Ru/Al2O3 catalyst is able to completely convert ethanol with selectivity toward

hydrogen above 95%. They found it from investigated of the active metallic phase

(Rh, Ru, Pt, Pd), the nature of the support (Al2O3, MgO, TiO2) and the metal loading

(05 wt.%).in the temperature range of 600850 C .

2.3.2

Partial Oxidation

The performance of Pt/Al2O3, Pt/ZrO2, Pt/CeO2 and Pt/Ce0.50Zr0.50O2

catalysts as the support were being studied upon on each individuals catalyst. Mattos

and Noronha (2005a) showed that the support plays an important role on the products

distribution of the partial oxidation of ethanol.


34/87

16

From the effect of reaction conditions and catalyst reducibility on the

performance of the Pt/CeO2 catalyst, Mattos and Noronha (2005b) found that at low

conversions, the ethanol dehydrogenation dominates, forming acetaldehyde, whereas

at high conversions the decomposition of ethanol is favoured, producing CH4, H2,

and CO.

2.4 Steam Reforming of Ethanol for Hydrogen Production

Akande et al. (2005) reported that the effect of catalyst synthesis method, Ni

loading and reduction temperature on the characteristics and performance of

Ni/Al2O3 catalysts were estimated. They investigated that which method will

produces highest selectivity hydrogen yield.

The feed for this process was crude ethanol. Based on this composition, the

general equation representing the reforming of crude ethanol can be represented as in

equation below.

C2:12H6:12O1:23 + 3:01H2O 2:12CO2 +6:07H2 (2.1)

Upon experiment of synthesis catalysts, three methods of synthesis: co-

precipitation, precipitation and impregnation were investigated. The reactor used to

obtain experimental data was BTRS model number 02250192-1 supplied by

Autoclave Engineers, Erie, PA, USA. Crude ethanol was delivered to the reactor

chamber by means of a HPLC pump regulated at the desired flow rates. Thereactions were carried out at atmospheric pressure and reaction temperature of 400

C. The product mixture during reaction was passed through a condenser and gas

liquid separator to separate the gaseous and liquid products for analysis.

As a result of the type of species generated by the synthesis method, the PT

catalysts were more reducible than the CP and IM catalysts. Catalysts prepared by

precipitation generally exhibited lower crystallite sizes of NiO species than thecorresponding catalysts prepared by co-precipitation. The catalysts prepared by


35/87

17

impregnation had the largest crystallite sizes except IM10 which had the smallest

crystallite size. In terms of H2 yield, CP15 gave the highest yield because the CP

catalysts gave the highest H2 selectivity as compared to corresponding catalysts

prepared by precipitation and impregnation.

2.5 Optimization Simulation of Hydrogen Production

Based on the literature reviews that have been done, there were few

researchers did on optimization simulation of hydrogen production using ethanol as a

raw material for fuel cell application. However, some of them did research

hydrocarbon on simulation. Jimnez (2006) using Aspen HYSYS to study the

viability of using a new catalyst to Methanol to a hydrogen rich product gas and

compare their production potential. Ozdogan et al. (2005) shows by using

hydrocarbon fuel as source in HYSYS 3.1 to compare two liquid hydrocarbon fuels.

They studied the effect of average molecular weights of hydrocarbons, on the fuel

cell processing efficiency.

2.6 Summary

Generally, there are many articles and journal on hydrogen production for

fuel cell application but when we are grouping that journal, we can conclude that,

there are three major groups that can synthesis hydrogen for fuel cells. Three of themare natural gas, alcohol and petroleum fraction. Additionally, there are many

processes that produce hydrogen such as steam reforming, autothermal reforming,

partial oxidation reforming, etc. Focus of this literature survey is to find a research

about ethanol as an input for hydrogen production by autothermal reforming. There

are a researchers had done the research about ethanol but a few had done research it

in simulation.


36/87

18

CHAPTER

METHODOLOGY

3.1 Research Tools

This research was carried out using various computational tools. Aspen

HYSYS 3.2 simulator was used for process flow sheeting to provide data regional

analyses. Aspen HYSYS 3.2 simulator was also used to perform the new process

model control structure for H2 production using ethanol as a raw material for fuel cell

application.

3.1.1 Aspen HYSYS

HYSYS was a product of AEA Technology, which is now part of Aspentech

Engineering Suite (AES). HYSYS has been chooses as the process simulator for this

research because of two main advantages over the other software packages. It caninteractively interpret commands as they entered one at a time. Other requires

execution after new entries. HYSYS has the unique feature that information

propagates both in forward and reverse directions, performing back-calculation in a

non-sequential manner. The bi-directionality often makes iterative calculations

unnecessary and the solution is fast.


37/87

19

3.2 Research Activities

3.2.1 Data Collection

From theoretical analysis and the report that have been done by Akande et al.

(2005), variables, variables relationship, approximate correlations, dynamic

characteristic and etc., about hydrogen production from ethanol is collected. Other

journal that related to this case study was collected too. Fierro et al. (2005) elaborate

the reaction that might be occurring while ethanol steam reforming reaction was

reacted. Vaidya et al. (2005) show that some reaction that using by ethanol. Reaction

such as ethanol steam reforming, ethanol cracking, and the others were collected to

comparable.

3.2.2 Base Case Stoichiometry

Vaidya et al. (2005) showed that the reaction is strongly endothermic andproduces only H2 and CO2 if ethanol reacts in the most desirable way.

22223 263 COHOHOHCHCH ++ (H = 174kJmol-1) (3.1)

However, other undesirable products such as CO and CH4 are also usually

formed during reaction.

COHOHOHCHCH 24 2223 ++ (H = 256kJmol-1) (3.2)

OHCHHOHCHCH 24223 22 ++ (H = -157kJmol-1) (3.3)

Total oxidation of ethanol to H2 and acetaldehyde respectively, the main

reactions are being given by:

OHCHOCHOOHCHCH 23223 5.0 ++ (H = -175kJmol-1) (3.4)


38/87

20

Other reactions that can also occur are: ethanol dehydrogenation to

acetaldehyde, ethanol dehydration to ethylene, ethanol decomposition to CO2 and

CH4 or CO, CH4 and H2.

24223 HOHCOHCHCH + (H = 68kJmol-1) (3.5)

OHHCOHCHCH 24223 + (H = 45kJmol-1) (3.6)

4223 5.15.0 CHCOOHCHCH + (H = -74kJmol-1) (3.7)

2423 HCHCOOHCHCH ++ (H = 49kJmol-1) (3.8)

They suggested the occurrence of several reactions: acetaldehyde formed by

dehydrogenation of ethanol is decomposed to CH4 and CO or undergoes steam

reforming.

442 CHCOOHC + (H = -21kJmol-1) (3.9)

2242 32 HCOOHOHC ++ (H = 180kJmol-1) (3.10)

Water reforms the C1 products to hydrogen.

2224 42 HCOOHCH ++ (H = 160kJmol-1) (3.11)

224 3HCOOHCH ++ (H = 210kJmol-1) (3.12)

2242 422 HCOOHHC ++ (H = 210kJmol-1) (3.13)

2262 522 HCOOHHC ++ (H = 350kJmol-1) (3.14)

In addition, the following reactions occur when O2 is present:

OHCOOCH 2224 22 ++ (H = -800kJmol-1) (3.15)

224 25.0 HCOOCH ++ (H = -36kJmol-1) (3.16)

2224 2HCOOCH ++ (H = -320kJmol-1) (3.17)

225.0 COOCO + (H = -280kJmol-1) (3.18)

22 COOC + (H = -390kJmol-1) (3.19)

Other reaction:

24 2HCCH + (H = 75kJmol-1) (3.20)


39/87

21

62242 HCHHC + (H = -140kJmol-1) (3.21)

242 22 HCHC + (H = -52kJmol-1) (3.22)

3.2.3 Base Case Validation

Validation was done by comparing the mole fraction of the effluent by

calculation from total reaction and the mole fraction of the effluent of the ATR as

simulated in Aspen HYSYS 3.2 simulation.

3.2.4 Autothermal Reactor Optimization

Optimization for ATR was done by varies the feed air molar to get the best

flow rate of air when entered the ATR. Two case studies have been developed in this

optimization. The first one is about to monitor the molar flow rate of CO and

hydrogen at ATR vapour stream after varying the air molar flow rate. The second

case study is to monitor the temperature at the ATR vapour stream after varying the

air flow rate within the same range as the first case study. The optimize air molar

flow rate need to be above 700C when flow out at the ATR vapour stream. This is

for usage of heat energy in heating the feed stream.

3.2.5 Heat Integration

All hot and cold streams were systematically arranged to build system heat

integration. By apply a heat exchangers to a process, the heat from ATR vapour

stream was being cooled down by the feed stream; water, air, and ethanol. This can

really save a lot of energy and achieve target required.


40/87

22

3.2.6 Carbon Monoxide Clean Up

Carbon monoxide is a dangerous gas that should be aware and not profitable.

Several reactions may produce it as main product or by-product. So, the cleaningmethod is required need and converts it to other relevant component. Water gas shift

and preferential oxidation can reduce CO and are being used in this research entirely.

3.2.6.1Water Gas Shift

Water gas shift is the first stage to reduce the CO after reaction in ATR. CO

will be converted into hydrogen and carbon dioxide when mixed with steam. There

were three equilibrium reactors that being attached after stream that flow out from

ATR reactor. The first reactor is called high temperature shift (HTS), followed by

medium temperature shift (MTS) and end with low temperature shift (LTS). Stream

from ATR vapour will entered this entire three equilibrium reactor, and will react on

this reaction:

222 HCOOHCO ++ (H = -42kJmol-1) (3.23)

3.2.6.2Preferential Oxidation

The next stage CO cleans up was preferential oxidation reactions. The

conversion reactor was attached after WGS stage. It was performed in order to

reduce the CO concentration out of the LTS to the ppm levels required for the fuel

cell. The PROX reactor was modelled as a conversion reactor based on two reactions

to oxidize CO. the reactions were

225.0 COOCO + (H = -280kJmol-1) (3.24)

OHOH 222 5.0 + (H = -240kJmol-1) (3.25)


41/87

23

3.2.7 Plant Wide Optimization

Plant wide optimization was being done to optimize the usage of the reactors

for the whole plant. ATR optimization optimized the stream out ATR reactor, watergas shift optimization optimized the HTS, MTS and LTS reactors and preferential

oxidation optimization optimized the PROX reactor.

3.2.7.1 ATR optimization

ATR optimization was studied by monitoring the temperature at the ATR

stream out. The temperature of the stream must be above than 700 C. These was

important because it will affect the heat exchanger network if the required

temperature not in right conditions. The case study one was optimizing the air flow

and set the range of the molar flow rate that can be manipulated. The next case study

was to monitor the highest hydrogen that can be choosing in the range of air molar

flow rate in first case study.

3.2.7.2 Water Gas Shift Optimization

WGS optimization was conducted by varying the water feed molar flow rate

to get the best water feed molar flow rate to optimized the efficiency of the reactors

except for PROX reactor. For this optimization, case study three developed to

monitor the CO and hydrogen concentration in each reactor except PROX after

varying water feed molar flow rate. While case study four was developed to monitor

the temperature of HTS inlet after varying the water molar feed rate within the same

range as case study three. The optimized water molar flow rate was taken at the point

where the temperature for HTS inlet is above 100C.


42/87

24

3.2.7.3Preferential Oxidation Optimization

PROX optimization was conducted by varying molar flow rate of air in the

additional air stream that directed to PROX reactor. The purpose was to reduce theconcentration of CO in PROX effluent to approximately 10 ppm while making sure

that the effluent temperature is in range 60C to 100C. Case study five was

developed to monitor the CO concentration in PROX vapour stream after varying the

air molar flow rate in the new air stream in a certain range.

3.2.8 Temperature and Component Profile

By looking at the temperature and component profile, we investigated the

behaviour of every unit operations. This is needed to find out the different for each

reactor and their effect. Next, the conditions like the temperature and the component

on overall plant also were studied well.

3.2.9 Fuel Processor Efficiency

The system fuel processor efficiency can be calculated by :

(Lenz and Aicher , 2005)

CxHyOzCxHyOz

COCOHH

LHVn

LHVnLHVn += 22 (3.26)


43/87

25

3.3 Summary

This chapter basically show the methodology that need to accomplish. The

method are describe in detail from stoichiometry mathematical analysis calculation,base case development with HYSYS, validation, heat integration model, clean up

model, plant wide optimization, components and temperature analysis to fuel

processor efficiency. All of them are systematically do as a Figure 3.1.

Base Case Development with HYSYS Validation

Input Output

Temperature and Component Analysis

Plant Wide Optimizations

Clean Up Model

Heat Integration Model

Stoichiometry Mathematical Analysis

Figure 3.1 : Algorithm for methodology.


44/87

26

CHAPTERV

SIMULATION AND OPTIMIZATION OF HYDROGEN PRODUCTION

PLANT FROM ETHANOL FOR FUEL CELL APPLICATION

4.1 Process Description of Hydrogen Production from Ethanol

The process simulation package Aspen HYSYS 3.2 has been used along with

conventional calculations in this study. Figure 4.1 presents the investigated operation

conditions for major fuel processing units (ATR, HTS, MTS, LTS, and PROX). The

selection of these operating conditions are based on theoretical studies aiming at

producing hydrogen rich and carbon monoxide poor mixtures in an efficient manner

at acceptable conversions.

It started from the feed stream; ethanol, air and steam at 1 atm enter the ATR

reactor. Then the outlet stream will enter WGS reactor. There were three reactor in

WGS section; HTS, MTS and LTS. Finally, the outlet entered the PROX reactor and

the product was ready to enter fuel cell.


45/87

27

To Fuel

CellATR

reactor

WGS

reactor

PROX

reactor

EthanolAir

Steam

100C 100C 70C

Air

Figure 4.1: The operation conditions for the major unit operation

4.2 Modelling and Simulation of Hydrogen Production from Ethanol forFuel Cell

The hydrogen production from ethanol for fuel cell was simulated using

HYSYS software as a figure 4.2 shows it. Typically, the simulation process takes the

following stages:

i. Preparation Stage

a) Selecting the thermodynamic model

b) Define chemical components

ii. Building Stage

a) Adding and define streams

b) Adding and define unit operations

i. Auto-thermal reforming reactor

ii. Water gas shift reactor

1. High temperature shift reactor

2. Medium temperature shift reactor

3. Low temperature shift reactor

iii. Preferential oxidation reactor

c) Connecting streams to unit operations

d) Add auxiliary unit

i. Heater


46/87

28

ii. Cooler

iii. Heat exchanger

iii. Execution

a) Starting integration

b) Optimization the whole plant


47/87

29

Figure4.2:Thewholepla

ntsystemb

yAspenHYSYS3.

2


48/87

30

HYSYS simulator is made up of four major parts to form a rigorous

modelling and simulation environment.

i) A component library consisting of pure component physical properties.

ii) Thermodynamic packages for transport and physical properties

prediction.

iii)Integrator for dynamic simulation and/or solver for steady-state

simulation.

iv)Mathematical modelling of unit operation.

For this study, each of above components is described in below.

HYSYS

SimulationEnvironment

PhysicalPropertyLibrary

Unitoperation

Model

Integrator/ Solver

Thermo-DynamicPackage

Figure 4.3: HYSYS Simulation Environment


49/87

31

4.2.1 Thermodynamic Properties

In order to define the process, the thermodynamic property packages used to

model steady-state of ethanol must be specified. The feed for the hydrogenproduction is considered to be relatively ideal mixture of ethanol and oxygen.

Ethanol is the primarily characterized as a C2H5OH. The Peng-Robinson Equation of

State (EOS) is used to model the thermodynamics of hydrogen production for both

steady-state and dynamics operations (HYSYS Reference, 2000):

(4.1)

)(bV + )( iiii

i

ii bVbV

T)(aRTP =

The terms;

ci

ciiir

iP

TRTTa

22);()(

= (4.2)

ci

cii

P

RTb = (4.3)

Where according to Peng Robinson (1976);

21= , 21+= , 45724.0= , 07779.0= , 30740.0=cZ .

Therefore,

22/12 )]1)(26992.054226.137464.0(1[);( riiiir TT ++=

For dynamics modelling of hydrogen production, the Peng-Robinson

Equation of state was found to simulate hydrogen production faster than the real

time. When performing the dynamics simulation, Aspen HYSYS permits a user

selected thermodynamics calculation procedure.


50/87

32

Additionally, the allowable maximum and minimum temperature and

maximum pressure over which dynamics are calculated and is user defined in Aspen

HYSYS. For the Aspen HYSYS model the default values were selected. Usually the

default minimum and maximum temperature value in flow sheet, respectively. The

maximum pressure was selected to be 1 atm above the highest pressure in the flow

sheet (HYSYS reference, 2000).

4.2.2 Physical Properties

Components that entered the ATR for the process hydrogen production was

ethanol, water and air. Additionally, the component such as carbon monoxide, carbon

dioxide, hydrogen, nitrogen, oxygen, acetaldehyde, ethylene, methane and carbon

need to define in HYSYS environment. All components are present in room

temperature. The pure component properties of the feed stock are listed in Table 4.1.

Table 4.1: Physical property of the component

Component Molecular formula MW(kg/kmol) (kg/m3) BP (C)

Ethanol C2H4OH 46.069 795.98 78.25

Oxygen O2 31.999 1137.68 -183.95

Water H2O 18.015 997.99 100.00

Nitrogen N2 28.014 806.37 -195.80CarbonMonoxide CO 28.010 799.39 -191.45

Carbon dioxide CO2 44.010 825.34 -78.55

Hydrogen H2 2.016 69.86 -252.60Acetaldehyde C2H4O 44.05 777.00 19.85

Methane CH4 16.04 299.39 -161.52

Ethylene C2H4 28.05 383.23 -103.75

Carbon C 12.01 1642.06 -

MW was the molecular weight, was density and BP was boiling point. The

densities were taken at 25C.


51/87

33

4.2.3 Integration Algorithm

A dynamic model is represented by a set of ordinary differential equations

(ODEs) in Aspen HYSYS. In order to solve the model, an implicit Euler method isused to integrate the ODEs. The fixed step size implicit Euler method explains here

is known as the rectangular integration. It can be described by extending a line slope

zero and length h (the step size) from tn to tn+1 on af(Y) versus time plot. The area

under the curve is approximately by a rectangle of length h and heightfn+1(Yn+1) in a

function of the following form (HYSYS Documentation, 2000).

(4.4)

To provide a balance between accuracy and speed, Aspen HYSYS employs a

unique integration strategy. The volume, energy and speed composition balances are

solved at different frequencies. Volume balances are defaulted to solve at every

integration step, whereas energy and composition balances are defaulted to solve at

every 2nd and 10th integration step, respectively. The integration time step can be

adjusted in Aspen HYSYS to increase the speed or stability of the system. The

default value of 0.5 second was selected.

4.2.4 Mathematical Modelling of the Reactor Operating

4.2.4.1 Linear and Non-Linear Systems

A linear first-order Ordinary Differential Equation (ODE) can be described as

follows:

(4.5)

)(f(Y

:, Y

dY

where =1

nt

nndtYfY

dtnt

+

+=+

1

)

)(KfYdY

=+ udt


52/87

34

In a non-linear equation, the process variable Y may appear as a power,

exponential, or is not independent of other process variables. Here are two examples:

(4.6))(3 KfY

dt

=+ udY

(4.7))(KfYYdt=+ 2 u

dY

The great majority of chemical engineering processes occurring in nature are

nonlinear. Nonlinearity may arise from equations describing equilibrium behaviour,

fluid flow behaviour, or reaction rates of chemical systems. While a linear system of

equations may be solved analytically using matrix algebra, the solution to a non-

linear set of equations usually requires the aid of a computer.

4.2.4.2 Material Balance

The conservation relationships are the basis of mathematical modelling in

HYSYS. The dynamic mass, component, and energy balances that are derived in the

following section are similar to the steady-state balances with the exception of the

accumulation term in the dynamic balance. It is the accumulation term which allows

the output variables from the system to vary with time. The conservation of mass is

maintained in the following general relation:

Rate of accumulation of mass = mass flow into system - mass flow out of system


53/87

35

Figure 4.4: Reactor operating

(4.8)oio FF = oi

dt

Vd )(

Where:

Fi= the flow rate of the feed entering the reactor tank

i = the density of the feed entering the reactor tank

Fo = the flow rate of the product exiting the reactor tank

o= the density of the product exiting the reactor tank

V = the volume of the fluid in the reactor tank

4.2.4.3Component Balance

Component balances can be written as follows:

Rate of accumulation of component j =

Flow of component j into system

- Flow of component j out of system

+ Rate of formation of component j by reaction


54/87

36

Flow into or out of the system can be convective (bulk flow) and/or

molecular (diffusion). While convective flow contributes to the majority of the flow

into and out of a system, diffusive flow maybe come significant if there is a high

interfacial area to volume ratio for a particular phase. For a multi-component feed for

a perfectly mixed tank, the balance for componentj would be as follows:

VRCFCFdt

Vd )Cjjoojii

jo+=

((4.9)

Where:

Cji = the concentration of j in the inlet stream

Cjo = the concentration of j in the outlet stream

Rj = the reaction of rate of the generation of component j

For a system with NC components, there are NC component balances. The

total mass balance and component balances are not independent; in general, you

would write the mass balance and NC-1 component balances.

4.2.4.4Energy Balance

The Energy balance is as follows:

Rate of accumulation of total energy =

Flow of total energy into system

- Flow of total energy out of system

+ Heat added to system across its boundary

+ Heat generated by reaction

- Work done by system on surroundings


55/87

37

The flow of energy into or out of the system is by convection or conduction.

Heat added to the system across its boundary is by conduction or radiation. For a

CSTR with heat removal, the following general equation applies:

)()()(])[(

iiooroooooiiiii PFPFwQQkuFkuFdt

Vkud+++++++=++

(4.10)

Where:

u = Internal energy (energy per unit mass)

k = Kinetic energy (energy per unit mass)

= Potential energy (energy per unit mass)

V = the volume of the fluid

w = Shaft work done by system (energy per time)

Po = Vessel pressure

Pi = Pressure of feed stream

Q = Heat added across boundary

Qr= Heat generated by reaction: DHrxnrA

Several simplifying assumptions can usually be made: The potential energy

can almost always be ignored; the inlet and outlet elevations are roughly equal. The

inlet and outlet velocities are not high; therefore kinetic energy terms are negligible.

If there is no shaft work (no pump), w=0.

The general energy balance for a 2-phase system is as follows:

rvvlliiillvv QQHFhFhFhVHVdt

d+++=+ ][ (4.11)


56/87

38

4.2.5 Degree of Freedom Analysis

There are two types of degree of freedom. The first one is dynamic degrees of

freedom, Nm (m denotes manipulated). Nm is usually easily obtained by processinsight as the number of independent variables that can be manipulated by external

means. In general, this is the number of adjustable valves plus other adjustable

electrical and mechanical devices. Second is steady state degrees of freedom, Nss

which is the number of variables needed to be specified in order for a simulation to

converge. To obtain the number of steady state degrees of freedom we need to

subtract fromNom which is the number of manipulated variables with no steady state

effect andNoy which is the number of variables that need to be controlled fromNm

As a result equation 4.12 is obtained

)( oyommss NNNN += (4.12)

In any process simulation work, it is essential that the degrees of freedom

analysis be carried out to determine the number of variables to be specified.

4.2.6 Analysis of Optimization Response

The case study for certain section of plant was selected; an optimization

analysis will carry out to show the efficiency of the plant wide. Selected process

inputs were changed when the process had been optimized. Corresponding process

outputs were monitored to get the scope required.


57/87

39

4.3 Summary

Basically, this chapter is about the development of the simulation using

Aspen HYSYS 3.2. All the data that gathered from literature surveys are used. Forthe simulation of HYSYS, the equation of state that used is Peng-Robinson to

calculate the stream physical and transport properties. Mass and energy balances

have established for all cases. A block diagram about the simulation of hydrogen

plant using Aspen HYSYS 3.2 is shown in Figure 4.3.

Enter Aspen HYSYS 3.2

C2H4OH

Selecting thermodynamic model# Peng-Robinsion

Define chemical component

Addin & define stream

Adding & define unit operationATR reactor

O timization

O2

N2

CO2H2OH2CO

01 Ethanol

01 Air

03 water

Start inte ration

WGS reactor

Adding & define unit operationPROX reactor

Plant wide optimization

O timization

Figure 4.5: Block diagram of the simulation of hydrogen plant using Aspen HYSYS

3.2


58/87

40

CHAPTER V

RESULTS AND DISCUSSIONS

5.1 Results for Base Case Study

The base case of this study was developed by introducing all the raw

materials which were ethanol, air and water into a single autothermal reactor (ATR)

in vapour phase. The feed were entered the ATR in a different stream as shown in

Figure 5.1. Since ethanol and water are in liquid phase at room temperature; 25 oC,

these two materials need to be converted to gas phase first. This process was done

by heating the materials with heaters until 100 oC as the ethanol and water boiling

point are 78.4 oC and 100 oC, respectively. Air too was being heat up to 100 oC to

increase the rate of reaction. The reactor was set up to be operated at 1 atm. The

molar flow rate of the raw materials is being evaluated from the total reactions of all

reactions that occur in the reactor with basis of 100 kgmole/hr of ethanol.


59/87

41

Figure 5.1: Process Flow Diagram of the Base Case

Thermodynamic aspects of ethanol steam reforming have received a fair

amount of attention in the literature review by Vaidya et. al. (2006). The reaction is

strongly endothermic and produces only H2 and CO2 if ethanol reacts in the most

desirable way. The basic reaction scheme; (3.1) and (3.2), was as follows:

22223 263 COHOHOHCHCH ++

COHOHOHCHCH 24 2223 ++

In autothermal conditions, conversion of ethanol gives rise mostly to the

production of acetyldehyde which has been detected as the only product till complete

conversion of both ethanol and oxygen. However, after total oxygen conversion, H2

is also produced. Total oxidation of ethanol to H2 and acetaldehyde respectively, the

main reactions (3.4) are being given by:

OHCHOCHOOHCHCH 23223 5.0 ++


60/87

42

Other reactions that can also occur are: ethanol dehydrogenation to

acetaldehyde (3.5), ethanol dehydration to ethylene (3.6), ethanol decomposition to

CO2 and CH4 (3.7) or CO, CH4 and H2 (3.8).

24223 HOHCOHCHCH +

OHHCOHCHCH 24223 +

4223 5.15.0 CHCOOHCHCH +

2423 HCHCOOHCHCH ++

At low temperature in steam reforming conditions, acetaldehyde too reacts

and produces CO and H2 (3.10).

2242 32 HCOOHOHC ++

When O2 occur, methane will react and turn out total oxidation; (3.15), and

partial oxidation; (3.16) and (3.17).

OHCOOCH 2224 22 ++

224 25.0 HCOOCH ++

2224 2HCOOCH ++

Steam reforming of methane will give more production of hydrogen. The

reactions of the process; (3.22) and (3.19) are given by:

242 22 HCHC +

22 222 COOC +

All of the chemical reactions are assumed to occur adiabatically under

conversion conditions. All these 13 reactions are reacting in an autothermal reactor

(ATR) in vapour phase. Total reaction for all the reactions (5.1) are given as:

222223 238625.57 HCOCOOHOOHCHCH ++++ (5.1)


61/87

43

From the total reaction, the feed ratio that should be introduced into the

reactor is 7:5.5:2 for ethanol over oxygen over water. Taking basis 100 kgmole/hr of

reactant ethanol, the flow rate for oxygen and water is 78.5714 kgmole/hr and 28.57

kgmole/hr respectively. This will make the air flow rate is 374.1597 kgmole/hr. The

flow rate of the ATR effluent is given in Table 5.1:

Table 5.1: Molar Flow of ATR Effluent for Base Case

Master Component Molar Flow (kgmole/hr)

Ethanol 0

Oxygen 0

Nitrogen 295.5783Water 69.2490

Carbon monoxide 129.6610

Carbon dioxide 43.4019

Hydrogen 259.3220

Ethylene 0

Acetaldehyde 0

Methane 0Carbon 0

5.2 Result for Validation

Validation was done by comparing the mole fraction of the effluent bycalculation from total reaction and the mole fraction of the effluent of the ATR as

simulated in Aspen HYSIS 3.2 simulation.


62/87

44

Table 5.2: Validation for simulation effluent compare with calculated effluent

Master Component Calculated Simulated Error

Ethanol 0 0 0

Oxygen 0 0 0

Water 0 0.0869 -

Carbon monoxide 0.1387 0.1626 0.0239

Carbon dioxide 0.1040 0.0544 0.0496

Hydrogen 0.3987 0.3253 0.0734

Nitrogen 0.3586 0.3708 0.0122

Ethylene 0 0 0

Acetaldehyde 0 0 0

Methane 0 0 0

Carbon 0 0 0

From Table 5.2, errors for all components are very small, ranging from 1.2%-

7.3%. Since the errors are small, we can conclude that the simulation model

developed using Aspen HYSYS 3.2 is valid and can be used as a real plant for

further analysis.

5.3 Results for Heat Integration

The feed stream was basically in a room temperature condition; 25C.

The temperature required to enter the ATR reactor was 100C. In order to achieve

this target, three heaters were installed to the feed stream. Figure 5.2 show the

diagram of the heater being attached. The outlet temperature from ATR was above

700C, so we can apply the heat exchanger network.


63/87


64/87

46

Figure 5.3: The heaters at the feed streams were exchange with the heat exchanger.

5.4 Results for CO Clean Up

CO needs to be cleans up for the safety. The CO was produced in ATR. This

component was unprofitable and dangerous to environment, needs to be cleans up by

using water gas shift reaction and preferential oxidation reactions.

5.4.1 Water Gas Shift

The ATR effluent was passed through ATR cooler to cool down its

temperature to the desired HTS inlet temperature. The HTS was performed the water

gas shift reaction (3.23) in which CO was converted to meet the specification. Then

the outlet from HTS was being cool to enter MTS reactor. This process was repeated

until to LTS reactor. Figure 5.4 show the WGS reactor being attached after outlet

ATR reactor. The Table 5.3 shows the molar flow of the component out of all

reactors involved. From the water gas reaction, the composition of CO decreased


65/87

47

from 16.27% to 7.6%. Meanwhile the composition of hydrogen was increased from

32.53% to 41.20%.

Figure 5.4: The WGS reactor

Table 5.3: Effluent molar flow after water gas shift reaction for each reactor

Master

Component

ATR HTS MTS LTS

Nitrogen 295.5390 295.5390 295.5390 295.5390

Water 69.2453 2.8729 0.1864 0.1596

Carbon

monoxide

129.6629 63.2925 60.6040 60.5772

Carbon dioxide 43.3924 109.7628 112.4513 112.4781

Hydrogen 259.3257 325.6961 328.3846 328.4114

5.4.2 Preferential Oxidation

Effluent from the LTS was cooled down first to the required PROX inlet

temperature. Preferential oxidation reactions, (3.24) and (3.25) took place in PROX

conversion reactor. CO was oxidized to CO2 and the H2 was oxidized to H2O,

simultaneously. Additional air was attached to the PROX reactor with zero molar

flow as shown in Figure 5.5. This extra air stream was needed in the optimization.


66/87

48

Figure 5.5: The preferential oxidation reactor

Table 5.4: Effluent molar flow after partial oxidation reaction

Master Component After LTS After PROX

Nitrogen 295.5390 295.5390

Water 0.1596 0.1596

Carbon monoxide 60.5772 60.5772

Carbon dioxide 112.4781 112.4781

Hydrogen 328.4114 328.4114

5.5 Plant Wide Optimization

Plant wide optimization was set to optimize the hydrogen production while

minimize the CO concentration with several constraints.. WGS optimization

optimized the water molar flow rate and increase the water gas shift reaction in HTS,

MTS and LTS while PROX optimization optimized the air molar flow in PROX air

feed stream and decrease the concentration of CO to ppm level required for the fuel

cell in the PROX reactor.


67/87

49

5.5.1 Result for ATR Optimization

Optimization for ATR was done by varying the air molar flow rate to get the

best flow rate of air to be introduced into the ATR. Two case studies were developedin order to do this optimization. The first case study was developed to monitor the

temperature at the ATR vapour stream after varying the air molar flow rate from 100

kgmole/hr to 1500 kgmole/hr. The second case study was developed to monitor the

molar flow rate of carbon monoxide and hydrogen after varying air molar flow rate

within the range that was chosen from first case study. The optimized air molar flow

rate was taken at temperature of the ATR vapour stream is above 700 oC. This is

because the heat from the stream can be used later for heat integration.

The results for case study one and case study two are presented in Figure 5.6

and Figure 5.7. From Figure 5.6, the temperature out of ATR is over 700 oC only

after the molar flow rate of air greater or equal 350 kgmole/hr. With that air molar

flow rate range, the hydrogen and CO molar flow rate was monitored. From figure

5.7, the flow rate of hydrogen produced by the ATR is decreasing when of air molar

flow rate greater than 350 kgmole/hr. Then it began constant after 550 kgmole/hr.

0

200

400

600

800

1000

1200

100 350 600 850 1100 1350

air - Molar Flow kgmole/h

ATRout-Temp

eratureC

.

temperature

Figure 5.6: Temperature of ATR vapour for varies air Feed molar flow


68/87

50

210

220

230

240

250

260

270

100 350 600 850 1100 1350


MasterCompMolarFlow(Hydrogen)kgmole/

.

100

105

110

115

120

125

130

135

MasterCompMolarFlow(CO)kgmole/h

.Hydrogen CO

Figure 5.7: Molar flow of CO and H2 effluent for varies air feed molar flow

The air molar flow rate was chosen at temperature 760C which is 370

kgmole/hr. This is suitable flow rate because at this rate hydrogen molar flow rate

begin to decrease. At that slope, hydrogen is 259.322 kgmole/hr.

5.5.2 Water Gas Shift Optimization

In WGS optimization, one case study was developed to optimized value of

feed water molar flow to reduce concentration of CO through water gas shift

reaction. Figure 5.8 shows the result of case study where the concentration of H2 and

CO after ATR was monitored. Another case study was developed to know how

temperature of the effluent will affect the water molar flow and the result was shown

in Figure 5.9.


69/87

51

Water molar flow rate was optimized from 30 to 300 kgmole/hr. As we can

see from Figure 5.8, the H2 show an increasing slope and the increasing is a bit

slower at 388 kgmole/hr. The optimum water molar flow rate was taken when H2 at

its higher molar flow rate. So, the value of water molar flow rate that was chosen was

150 kgmole/hr. At this point, H2 produced the greatest flow rate and CO reduced the

lowest flow rate. At that state, temperature was 250.1C as a Figure 5.9 show it.

320

330

340

350

360

370

380

390

30 70 110 150 190 230 270

water - Molar Flow kgmole/h

C

ompMolarFlow(H2)kgmole/h

.

CompMolarFlow(CO)kgmole/h

.

Hydrogen CO

Figure 5.8: Molar flow of CO and H2 effluent for varies water feed molar flow


70/87

52

50

100

150

200

250

300

350

400

450

500

30 70 110 150 190 230 270


ATRout-TemperatureC

.

temperature

Figure 5.9: Temperature to HTS for varies water feed molar flow

Table 5.5 compares the effluent produced by ATR, HTS, MTS and LTS

before and after WGS optimization being done. The increasing in water molar flow

rate did not affect the reactions in ATR, so the effluent of ATR did not change except

for steam. Other reactors show the same similarity, which were CO and steam being

reduced and H2 and CO2 were increased. From the ATR to LTS, CO was reduced

from 13.53% to 0.02% and H2 was increased from 27.05% to 40.56%.

Table 5.5: Molar flow of the effluent before optimization for ATR, HTS, MTS and

LTS.

Component ATR HTS MTS LTS

Nitrogen 295.5390 295.5390 295.5390 295.5390

Water 69.2453 2.8749 0.1864 0.1596

CO 129.6629 63.2925 60.6040 60.5772

CO2 43.3924 109.7628 112.4513 112.4781

Hydrogen 259.3257 325.6961 328.3846 328.4114


71/87

53

Table 5.6: Molar flow of the effluent after optimization for ATR,HTS, MTS and

LTS.

Component ATR HTS MTS LTS

Nitrogen 295.5390 295.5390 295.5390 295.5390

Water 190.6743 72.6090 61.4964 61.3008

CO 129.6629 11.5976 0.4850 0.2894

CO2 43.3924 161.4577 172.5703 172.7659

Hydrogen 259.3257 377.3910 388.5036 388.6992

5.5.3 Preferential oxidation optimization

Figure 5.10 show the result of the concentration of CO in ppm after varying

the air molar at PROX reactor. The concentration of CO in PROX was required

under 10 ppm. After being optimized, air molar flow was setup at 550 kgmole/hr,

where the 10.055 ppm and the temperature at 112.6 C. Table 5.7 compare the

effluent of PROX reactor after optimization was achieved. The concentration of H2

was decreased from 42.31 % to 34.02% because H2 was reacted with O2 in PROX to

produce H2O


72/87

54

0

40

80

120

160

200

100 250 400 550 700 850 1000 1150 1300 1450


COm

olarflow-ppm

.

CO

Figure 5.10: CO Molar flow in PROX effluent for varies air feed molar flow

Table 5.7: Molar flow of the effluent before and after optimization for PROX

PROXComponent

Before After

Nitrogen 295.5390 434.50

Water 61.3008 81.0109

CO 0.2894 0.0109

CO2 172.7659 199.9891

Hydrogen 388.6992 368.9891


73/87

55

5.6 Temperature Profile of Fuel Processor System

Figure 5.11 presents the temperature profile for the whole process starting

from the temperature of raw materials feed into the reactor until the temperature ofPROX vapour. The temperatures start up with 100 oC. The temperature rose up after

flow out from ATR (955.8 oC ). Then the temperature occurred at ATR will be used

to heating the raw materials by heat integrations process. The temperature slowly

cooled down to 100 oC during the heat integration process. At the first WGS reactor

(HTS), the temperature raise to 240.4 oC but then was set to cool at 100 oC before

enter the MTS reactor. After flow out from MTS reactor, the temperature rises just a

little and that same goes to LTS reactor. The temperature of effluents feed into theprox reactor were set to 70 oC. Finally the temperature of prox vapour is at 112.6 oC.

Figure 5.11 : Temperature profile for the whole unit operation

0

200

400

600

800

1000

feed

ATRi

ATRo HE

1HE

2HE

3

HTSC

o

HTSo

MTS

Co

MTS

o

LTSC

oLT

So

PROXc

o

PROX

Unit operations

Temperature,C


74/87

56

5.7 Component Profile of Fuel Processor System

Figure 5.12 shows the profile of CO and H2 component through the whole

plant. The main objective of this study was to maximize the production of H2 and inthe same time to reduce the concentration of CO as lower as possible. Therefore, it is

important to monitor concentrations of H2 and CO. The behaviour of the two

component profile was very different after ATR. This happened because CO was

being clean up in the plant where WGS reaction converted it into CO2 and H2 while

PROX reaction converted CO and H2 into CO2 and H2O with pres

Documents

Simulation and Optimization by Ethanol Autothermal Reformer for Fuel Cell Application by SyafiQ