Citation
Effect of Ambient Air Contaminants of the Performance of a Proton Exchange Membrane Fuel Cell

Material Information

Title:
Effect of Ambient Air Contaminants of the Performance of a Proton Exchange Membrane Fuel Cell
Creator:
O'SULLIVAN, GERARD MICHAEL ( Author, Primary )
Copyright Date:
2008

Subjects

Subjects / Keywords:
Carbon monoxide ( jstor )
Contaminants ( jstor )
Cylinders ( jstor )
Dioxides ( jstor )
Electric potential ( jstor )
Fuel cells ( jstor )
Gases ( jstor )
Hydrogen ( jstor )
Nitrogen ( jstor )
Oxygen ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Gerard Michael O'Sullivan. Permission granted to University of Florida to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
5/1/2005
Resource Identifier:
71279530 ( OCLC )

Downloads

This item has the following downloads:

osullivan_g ( .pdf )

osullivan_g_Page_20.txt

osullivan_g_Page_73.txt

osullivan_g_Page_53.txt

osullivan_g_Page_23.txt

osullivan_g_Page_61.txt

osullivan_g_Page_56.txt

osullivan_g_Page_70.txt

osullivan_g_Page_34.txt

osullivan_g_Page_62.txt

osullivan_g_Page_15.txt

osullivan_g_Page_51.txt

osullivan_g_Page_14.txt

osullivan_g_Page_19.txt

osullivan_g_Page_28.txt

osullivan_g_Page_13.txt

osullivan_g_Page_21.txt

osullivan_g_Page_10.txt

osullivan_g_Page_50.txt

osullivan_g_Page_32.txt

osullivan_g_Page_65.txt

osullivan_g_Page_33.txt

osullivan_g_Page_59.txt

osullivan_g_Page_27.txt

osullivan_g_Page_69.txt

osullivan_g_Page_52.txt

osullivan_g_Page_49.txt

osullivan_g_Page_25.txt

osullivan_g_Page_06.txt

osullivan_g_Page_64.txt

osullivan_g_Page_45.txt

osullivan_g_Page_37.txt

osullivan_g_Page_02.txt

osullivan_g_Page_36.txt

osullivan_g_Page_46.txt

osullivan_g_Page_79.txt

osullivan_g_Page_63.txt

osullivan_g_Page_22.txt

osullivan_g_Page_39.txt

osullivan_g_Page_29.txt

osullivan_g_Page_48.txt

osullivan_g_Page_76.txt

osullivan_g_Page_24.txt

osullivan_g_Page_11.txt

osullivan_g_Page_42.txt

osullivan_g_Page_40.txt

osullivan_g_Page_66.txt

osullivan_g_Page_68.txt

osullivan_g_Page_18.txt

osullivan_g_Page_67.txt

osullivan_g_Page_54.txt

osullivan_g_Page_47.txt

osullivan_g_pdf.txt

osullivan_g_Page_35.txt

osullivan_g_Page_55.txt

osullivan_g_Page_30.txt

osullivan_g_Page_38.txt

osullivan_g_Page_71.txt

osullivan_g_Page_05.txt

osullivan_g_Page_74.txt

osullivan_g_Page_80.txt

osullivan_g_Page_08.txt

osullivan_g_Page_31.txt

osullivan_g_Page_58.txt

osullivan_g_Page_03.txt

osullivan_g_Page_77.txt

osullivan_g_Page_43.txt

osullivan_g_Page_60.txt

osullivan_g_Page_17.txt

osullivan_g_Page_78.txt

osullivan_g_Page_16.txt

osullivan_g_Page_09.txt

osullivan_g_Page_44.txt

osullivan_g_Page_57.txt

osullivan_g_Page_75.txt

osullivan_g_Page_01.txt

osullivan_g_Page_12.txt

osullivan_g_Page_41.txt

osullivan_g_Page_72.txt

osullivan_g_Page_07.txt

osullivan_g_Page_04.txt

osullivan_g_Page_26.txt


Full Text












EFFECT OF AMBIENT AIR CONTAMINANTS ON THE PERFORMANCE OF A
PROTON EXCHANGE MEMBRANE FUEL CELL















By

GERARD MICHAEL O'SULLIVAN


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA


2005

































Copyright 2005

by

Gerard O'Sullivan

































This thesis is dedicated to my parents.















ACKNOWLEDGMENTS

I would first like to thank Dr. Herbert Ingley for giving me the opportunity to come

to this great university and work on a project that I held interest in. I would also like to

thank Dr. Jim Fletcher for coming all the way down from University of North Florida to

assist me whenever I needed help. Finally I want to thank Dr. William Lear and Dr. Yogi

Goswami for taking time to sit on my committee.

I am only here today because of the great sacrifices my parents made to raise their

children in the United States. They came to New York from Ireland months before I was

born and have always held me, my brother, and my sister first in their lives. I am forever

grateful for what they did and continue to do so I could have opportunities like this.

Florida was a great learning experience for me in general, it was my first time away

from my home and I would like to send a special thank you to all my friends in NY and

the friends I made here in Gainesville in the past year. There were some moments here in

Florida that I would not have made it without a little help from my friends. The biggest

lesson I have learned in my time here is all one needs in life to be happy is the love of

family and friends. I love them all.

And finally I must thank God. I could never begin to count the number of blessings

I have in my life and it is all God's doing.
















TABLE OF CONTENTS

page

A C K N O W L E D G M E N T S ................................................................................................. iv

LIST OF TABLES .............................................. ........ viii

LIST OF FIGURES ......... ......................... ...... ........ ............ ix

N O M E N C L A T U R E .......................................................................................................... x i

A B S T R A C T .......................................... ..................................................x iii

CHAPTER

1 IN TRODU CTION ................................................. ...... .................

2 BACKGROUND ................................... .. .......... ................. .3

Fuel C ell B asics ............................................................... ... .... ......... 4
E lectrochem ical R action .......................................................... ...............5..
C athode characteristics............................................ ........................... 6
Platinum catalyst ..................... .. ....................... .. ...... .................
Polarization Curves .................................................................... .7
H hydrogen and R reform ing Processes ........................................ .....................9
Fuel C ell C ontam nation .................................................. ............................... 10
Types of Contam inants.................. ......... .............. ...... ............... 11
Previous Research on Contamination of the Anode................. ..................12
M ethods of M itigating the Contamination .......................................................13
Challenges for Fuel Cell V vehicles ................................................... .. ... .......... 15
A ir Contam nation ........................................................ .............. .. 15
Carbon M monoxide ............................................. .. ...... ................. 16
N nitrogen Dioxide .................. ............................ .... ... ................. 16

3 EXPERIM EN TAL SY STEM ......................................................... ................ 18

Ballard 1.2 kW N exa Pow er M odule................................................ ....... ........ 18
Nexa Subsystems..................................... ... ....... .. ......20
H ydrogen system ........................................... ........ .... .......... .. 20
Oxidant air system ......................................... .. ........................20
C ooling sy stem ................................................................ ............... 2 1


v









S safety sy stem .............................................................. .. 2 1
Air/Contaminant Flow Measurement and Mixing...................................................21
M easu rem en t .................................................... ................ 2 3
P itot T ube .................................................................. 23
Differential Pressure Transducer..... .......... ........................................24
M ass Flow Controller.............. ............................ 25
G a se s .........................................................................2 6
L o ad B an k ..............................................................2 6
S a fety ................................................................2 8
G as Chrom autograph ........ ............................... ................. ........ ..... 29

4 M E T H O D O L O G Y ...................... .. .. ......... .. .. ........................ .......................... 3 1

Baseline Performance Testing ................................. ..................................... 31
Polarization Curve Generation ......... ....................... .................... 31
C on stant L oad T testing .............................................................. .....................32
C ontam nation T esting............ ... ...................................................... .. .... ........ 32
Polarization Curve G generation ........................................ ........................ 34
C constant L oad T testing ............................................................... ............... 35
Contam nation Recovery ...................................... ........... .... ........... .... 35
Temperature Effects.............. .. ................ ...................36

5 RESULTS AND DISCU SSION ...................................................... ................ 37

B aseline P olarization D ata............................................................... .....................37
B aseline C constant L oad D ata.............................. .......................... ............... 38
Contam nation R esults.................. ................. .............. ...... .............. .. 40
Carbon M monoxide ................. ................ ..... .............. ............... .. 40
C o n stan t lo ad ............................................................... .............. 4 0
Polarization curves .................................... .................... ..............46
Rapid injections of carbon monoxide ................ .................................47
Nitrogen Dioxide .................. ....... .......... ......... 50
Constant load ......... ... ....... .. ... .. ................ ................... 51
Rapid injections of nitrogen dioxide ..................... ................... ........ 53
Temperature Effects.............. .. ................ ...................54

6 CONCLUSION AND RECOMENDATIONS........................... .....................56

APPENDIX

A LABVIEW COMPUTER PROGRAM ........................................... ............... 59

B STAR T-UP PR O CED URE S ............................................... ............ ............... 61

N exa Startup Procedure ........................................................................... 61
N exa Shutdow n Procedure ............... ..... ....................................................... 61
SRI Instruments 8600 Series Gas Chromatograph ............................................... 62









Startup P procedure ............................. .................... .. .. ....... .... ............62
Test Procedure ................................... ........................... .... ....... 63
Shutdow n Procedures .................................................. .............................. 64

L IST O F R E FE R EN C E S ........................................................................... .......... .......... 65

B IO G R A PH IC A L SK E TCH ...................................................................... ..................66
















LIST OF TABLES


Table page

2.1 Typical levels of carbon monoxide according to various air quality reports..........16

2.2 Typicallevels of nitrogen dioxide according to various air quality reports. ............17

4.1 Concentration of contaminant mixture gases in cylinders. .....................................33

4.2 Contaminant flow rates for three air flow rates (30, 60, 90), for pure
contaminant gases (top) and for contaminant gas mixtures (bottom). ...................34

4.3 Selected concentration for testing based on emissions reports collected in
literate re rev iew ................................................................................... 3 5

5.1 Stack voltage and stack power allowable ranges under normal operation
conditions. ...........................................................................40

5.2 Percentage changes in voltage and power during rapid CO injections at a
constant load of 20 am ps ......................... ...................... ................. .. ......... ...... 49

5.3 Percentage changes in voltage and power during rapid CO injections at constant
load of 30 am ps. .................................................... ................. 50
















LIST OF FIGURES


Figure page

2.1 A typical polarization graph illustrating regions of control by various types of
overpotential (B lom en et. al, 1993) ................................... ............................. ......... 8

2.2 Cell voltage after changing the fuel gas at 400 mA cm-2; anode catalyst;
Pto.5Ruo.5; pure H2 and H2/100ppm CO; T = 80C (Divisek et. al, 1998). ..............13

3.1 Ballard 1.2kW Nexa unit and various components............... ..................19

3.2 Air/Contaminant mixing manifold, Ballard Nexa fuel cell, load bank,
gas cylinders, NexaMon OEM software. ..................................... ............... 22

3.3 Inline Delta tube with swagelok fittings connected to static and dynamic
pressure reading devices ............................................................................ ... .... 23

3.4 Linear output at 5VDC supply of the SDP-1000-L. The fine lines indicate
the maximum tolerances including a temperature variation from 0-50C. ..............24

3.5 Sensirion SDP- 1000 differential pressure transducer. ...........................................25

3.6 Fathom mass controller connected to Swagelok tubing .........................................26

3.7 Back view of load bank with several resistors in parallel, constructed at UNF.
Potentiometer was placed at center (not pictured). ...............................................27

3.8 Front view of load bank with several switches to activate resistors and
potentiom eter for slight adjustm ents ......................................................................28

3-9 Various safety measures; fire extinguisher, smoke/carbon monoxide detector
and gas cylinder supports. ............................................................. .....................29

3.10 SRI Instruments 8600 gas chromatograph. ................................... ............... 30

5.2 Confidence interval of polarization curves based on six tests. .............................38

5.3 Stack voltage and stack power at constant load of 20 amps. ..................................39

5.4 Stack voltage and stack power at constant load of 40 amps. ..................................39

5.5 Carbon monoxide test (10 ppm) at constant load of 40 amps..............................41









5.6 Carbon monoxide test (30 ppm) at constant load of 20 amps..............................42

5.7 Carbon monoxide test (100 ppm) at constant load of 20 amps.............................44

5.8 Carbon monoxide test (100 ppm) at constant load of 30 amps.............................45

5.9 Polarization curves for 60 and 100 ppm carbon monoxide.............................. 46

5.10 Rapid injections of CO (10, 30, 60, 80 ppm) at constant load of 20 amps. ............48

5.11 Rapid injection of CO (100,120,140 160 ppm) at constant load of 20 amps...........48

5.12 Rapid injection of CO (100,120,140 160 ppm) at constant load of 30 amps...........50

5-13 Nitrogen dioxide test (100 ppb) at constant load of 20 amps..............................51

5.14 Nitrogen dioxide (400 ppb) at a constant load of 20 amps. ..............................52

5.15 Nitrogen dioxide (Ippm) at constant load of 30 amps ....................................... 53

5.16 Rapid injections of nitrogen dioxide (800, 1000, 1200 and 1400 ppm) at
constant load of 20 am ps. .......................................... .............. ............. ..54

5.17 Temperature effects on fuel cell performance. Polarization curves and power
curves shown for temperatures (50F, 60F, 70F and 80F). ...............................55

6.1 A sample experimental system containing a singb PEM fuel cell and various
components to control all operating parameters (Moore et. al) ............................58
















NOMENCLATURE


AP Differential Pressure

Au Gold

Cf Flow Coefficient

CNCI Cyanogens Chloride

CO Carbon Monoxide

CVC Cell Voltage Checker

D Pipe Diameter [mm]

DP Differential Pressure

EPA Environmental Protection Agency

FC3 Fuel Cell Contamination Control Group

FCV Fuel Cell Vehicle

FS Full-Scale

GC Gas Chromatograph

H2 Hydrogen

HCN Hydrogen Cyanide

ICE Internal Combustion Engine









LFL Lower Flammability Limit

mA milli- amps

MEA Membrane Electrode Assembly

NO2 Nitrogen Dioxide

02 Oxygen

Pd Palladium

PEM Proton Exchange Membrane

PPB Parts per Billion

PPM Parts per Million

Pt Platinum

Rh Rhodium

SLPM Standard Liters per Minute

SO2 Sulphur Dioxide

TCD Thermal Conductivity Detector

VDC Voltage (Direct Current)

Ws Weight Density of Air [kg/m3]















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

EFFECT OF AMBIENT AIR CONTAMINANTS ON THE PERFORMANCE OF A
PROTON EXCHANGE MEMBRANE FUEL CELL
By

Gerard Michael O'Sullivan

May 2005

Chair: Herbert A. Ingley, III
Major Department: Mechanical and Aerospace Engineering

Proton exchange membrane fuel cells are considered the most promising candidate

for replacing the power source in vehicles because of their high power density and high

energy efficiencies at low temperatures. There are many barriers to entry for a fuel cell

vehicle but one commonly overlooked yet essential issue is air contamination. Research

in fuel cell contamination to date has dealt solely with its effect on the anode of the fuel

cell, due to contaminants that exist in reformed fuels. The first fuel cell powered vehicles

will be expected to perform alongside internal combustion engines and diesel engines

which emit harmful contaminants such as carbon monoxide, nitrogen oxides, and sulphur

dioxides. Therefore the adverse effects which these contaminants can have on the

performance of a fuel cell should be understood so that steps to mitigating the negative

effects can be taken.

The goal of this study is to determine the concentrations at which carbon monoxide

and nitrogen dioxide begin to adversely affect the performance on a Ballard Nexa fuel









cell. Several tests were designed to simulate the types of environments in which a fuel

cell powered vehicle would operate. The performance drops and recoveries were

calculated from the collected data. The effect of air temperature was studied.

The results showed significant losses in performance with CO levels above

100ppm. Transient response tests demonstrated the fuel cell's ability to recover

completely despite brief injections of very high concentrations of carbon monoxide. The

carbon monoxide mechanism for poisoning seems to be a function of both cell potential

and CO concentration. No adverse effects were discovered for NO2 concentrations up to

1.4 ppm for the time interval tested.














CHAPTER 1
INTRODUCTION

The use of proton exchange membrane (PEM) fuel cells as replacements for

conventional sources of power has received growing attention. Recent developments in

fuel cell technology have made PEM fuel cells a strong contender to replace the internal

combustion engine (ICE) in vehicles. These fuel cell vehicles (FCV) can reduce harmful

emissions and decrease our dependency on foreign oil. The transition to fuel cell

powered vehicles will take a very long time due to fueling infrastructure and consumer

adaptation of a new technology. These first FCVs will be expected to operate alongside

traditional ICE engines. The harmful emissions that these vehicles deposit into the air

may severely affect the performance of the fuel cell vehicles.

The platinum catalyst on the electrodes of a PEM fuel cell responsible for breaking

down the hydrogen and oxygen is very susceptible to contamination. Even small

concentrations of a particular contaminant can have an adverse effect on the overall

performance of a PEM fuel cell. Much of the research to date has dealt with this adverse

effect on the anode (fuel side of the fuel cell) because of the contaminants that exist in

reformed fuels. Very little information is available about the contaminants effect on the

cathode (air side of the fuel cell). Efforts must be turned to address the type of

environments these PEM fuel cell vehicles may operate in, since it is likely that fuel cell

vehicles will draw the oxidant it needs to operate from the surrounding ambient air.

The focus of this research is to study the effects of selected contaminants on the

cathode and determines the concentrations at which these detrimental effects begin to






2


take place. Various experiments were designed to simulate the conditions a FCV may

encounter. Transient response tests were used to assess the fuel cell's ability to recover

after severe contamination. The ability of a fuel cell to recover to its original

performance is important for understanding the mechanism of the contamination. The

experimental data was then analyzed to determine a particular contaminant's effect on the

fuel cell's performance and to observe and recovery after contamination had ceased.














CHAPTER 2
BACKGROUND

Proton exchange membrane fuel cells have become the most promising alternative

power generating device for the vehicles of the future. The burning of fuel in an internal

combustion engine is an inefficient process for extracting energy from a fuel. The

conversion of fuel into electricity in a fuel cell is a one step electrochemical process.

Since electricity is generated directly and does not involve any intermediate mechanical

or thermal process, fuel cells are more efficient than any other technologies.

With fuel cell technology still in its infancy there is much research being done on

various parameters that affect fuel cell performance. One particular concern is catalyst

contamination. The catalysts in a proton exchange membrane (PEM) fuel cell perform

the most important function; they break down oxygen and hydrogen molecules. These

catalysts are very sensitive to contamination and degrade in the presence of contaminants.

The contaminants adsorb onto the surface of the catalysts and prevent the oxidizer and

fuel from breaking down and hence inhibit the fuel cell's ability to function. Some of the

contaminants that affect PEM fuel cells are carbon monoxide, nitrogen oxides, sulfur

oxides, benzene, propane and various other chemicals that can be found in the ambient

air.

Research conducted on fuel cell contaminants has focused primarily on

contaminants entering on the anode side of the fuel cell. This is due to the fact that many

fuel cells use a reformer to convert fuels such as natural gas, methanol and other organic

fuels into hydrogen. This reforming process produces high concentrations of carbon









monoxide and other chemicals that can have a detrimental effect on the fuel cell.

Therefore much of the research has focused primarily on dealing with this adverse effect.

There has been very little information published about the effects of contaminants

on the cathode. Air is critical to the performance of the fuel cell in two ways. First, the

air provides the oxygen necessary to complete the electrochemical process of converting

the hydrogen into electricity and water. Second, the air carries water, a by-product of the

fuel cell reaction, out of the fuel cell. Otherwise the water would flood the fuel cell and

prevent the electrochemical process.

Fuel Cell Basics

Fuel cells are electrochemical energy conversion devices. A single fuel cell

consists of two electrodes. The fuel electrode or anode oxidizes the hydrogen fuel and

the air electrode or cathode reduces the oxygen. Fuel cells are characterized by the type

of electrolyte used between these two electrodes. Some examples of the types of fuel

cells are Polymer Electrolyte Membrane (PEM), Phosphoric Acid, Direct Methanol,

Alkaline, Molten Carbonate, and Solid Oxide.

PEM fuel cells use a solid polymer electrolyte placed between two porous

electrodes. The catalyst is deposited onto these electrodes and is typically platinum.

PEM fuel cells are the most promising power source for fuel cell powered vehicles. They

offer a very high power density and high energy conversion efficiencies while operating

at relatively low temperatures. PEM cells are also lightweight and compact making them

suitable for automotive applications.









Electrochemical Reaction

A fuel cell is an electrochemical device which converts the free-energy of an

electrochemical reaction into electrical energy. The simplest form of the overall fuel

cell's reaction is shown below:

H2 + 1/2 02 = H20 2-1

This reaction is a result of two separate reactions taking place on the electrodes.

Platinum is used as a catalyst on the electrode to increase the rate of the reaction kinetics.

However platinum is expensive, a limited resource and very susceptible to poisoning

from contaminants. The electrochemical reactions that take place on platinum

electrocatalysts are shown below:

Anode H2 2H+ + 2e- 2-2

Cathode 02 +4H+ + 4e- 2H20 2-3

The oxygen reduction reaction, shown in formula 2-3, which takes place on the

cathode, is at least three orders of magnitude slower than the anode reaction. Since there

is already a challenge in increasing the electrocatalytic activity of the cathode reaction it

is very important that the air coming into the fuel cell is relatively clean and free from

harmful contaminants which could further reduce the rate of reaction.

The hydrogen breaks down into protons and electrons. The hydrogen ions mitigate

through the solid polymer to the cathode where they are re-combined with the electrons

and oxygen to form water. The electrons pass through an external circuit and produce

electricity.









Cathode characteristics

It is often understood that the ideal of maximum efficiency of an electrochemical

energy converter depends upon the electrochemical thermodynamics whereas the real

efficiency depends on the electrode kinetics (Bockris, 1969).

Electrode kinetics is the fundamental theory that describes the direct

electrochemical process of converting chemical energy into electricity and is a relatively

new area of research. An electrode impregnated with a platinum catalyst is called an

electrocatalyst. Noble metals such as Pt, Pd and Rh and their alloys have been found to

be the catalysts of choice for oxygen reduction. However, even the best of these

catalysts, Pt, is at least 106 times less active for oxygen reduction than for H2 reduction.

This leads to high overpotentials and is the major catalytic limitation to fuel cell

efficiency (Hoogers, 2003).

The voltage drop caused by the oxygen reduction reaction is the major source of the

irreversible voltage drop in the fuel cell. This voltage drop is influenced by several

physical and operating parameters including the cell current density, the active catalyst

surface area, conductivities and thickness of the catalyst layer and the concentrations and

diffusion coefficient of oxygen (Jeng et al. 2004).

There are three sources of overpotential that combine to reduce the overall voltage

of the fuel cell and reduce its performance.

* The activation overpotential is the loss in voltage due to the lack of
electrocatalystis.

* The ohmic overpotential is due to thermodynamic losses due to electrolyte
resistance to ion transport.









* The concentration overpotential is caused by gas transport limitation through the
gas diffusion layer of the cathode to the active catalyst sites.
These overpotentials are predominately due to the slowness of the oxygen

reduction reaction.

The reduction of oxygen is governed by a number of possible reactions but the four

electron reduction is the most attractive reaction to catalyze due to the high potential

gives the highest possible cell voltage for the fuel cell. Below is the four electron oxygen

reduction reaction.

02 + 4H+ + 4e 2H20 (Eo = 1.229 V) 2-4

Platinum catalyst

Depositing platinum onto the surface of the electrode increases the rate of the

electrochemical reaction. Although platinum has been proven as the most effective

catalyst for oxygen reduction it is very sensitive to contamination. Small concentrations

(< 10ppm) of carbon monoxide cause degradation in the fuel cell's performance. This

adverse effect is probably due to the adsorption of the CO molecules onto to active

catalyst sites to the exclusion of oxygen molecules. Platinum is also an expensive and

scarce metal and therefore there is much effort in finding a way to reduce the amount of

platinum deposited on the electrode, some other metals that may be alloyed with the

platinum are Pd, Rh, Ni, and Au.

Polarization Curves

The performance of a fuel cell is often graphically represented using a polarization

curve. It is a curve relating the current density and the cell voltage when using a single

fuel cell. When a fuel cell stack is used the polarization curve relates the stack current to

the stack voltage. The maximum thermodynamic reversible cell voltage is 1.23 V. The

curve again shows the vital role that electrode kinetics plays in the performance of the











fuel cell. Low electrocatalytic activity has the greatest effect on the drop in efficiency.


As stated before the oxygen reduction reaction in the cathode electrode is much slower


than the anode reaction. At low current densities the entire loss in the fuel cell potential


from the reversible value is due to activation overpotential at the oxygen electrode.


(Blomen & Mugerwa). A typical polarization curve is shown in Figure 2-1.



1.5
THERMMoYNuAMI 1.0
REVERSIBL CELL NTHNSIC
POTENTIAL (El MAXIMUM
Th THE IDEAL CELL POT TtAL-CURIENT RELATlON EfI
CELL POTIETAL LOSSES DUE TO
\ ACTIVATtION OVRWPOTENTIL&
S IACK OF nECTROCATALYtS) UNEAR DROP IN CEL
POTENTIAL MINLVY DUE TO
J OHMIC LOSStS IN SOLUTION
S-ETeEN E RCTRODES
1-

I-

Oj

U tOSSES CAUSE OF [
eCBIAsU Of CELL
OTNTIMAL TO ZERO



0 o
0 0.5 1
CELL CURRENT, amperes



Figure 2-1. A typical polarization graph illustrating regions of control by various types of
overpotential (Blomen et. al, 1993).

This curve can be generated for any fuel cell and will always have the same general


characteristics due to the overpotential losses. However, other parameters that can affect


the performance will change the slope of the linear portion of the curve. There are many


parameters that influence the performance of a PEM fuel cell and they include operating


temperature, pressure and humidification of the hydrogen and air streams.









Theslope of the polarization curve will decrease as the operation temperature

increases. This is due to the increase in the exchange current density and proton

conductivity. However humidification of the membrane becomes a problem at operating

temperatures over the boiling point of water.

The performance of the fuel cell increases with the operating pressure of the fuel

cell. This can be seen by a positive shift of the polarization curve with increasing

pressure. As the operating pressure increase, the partial pressures of the reactant gases

increases leading to better performance.

When the gas stream humidification temperature is lower than the operating

temperature, the fuel cell's performance will decrease and can be shown with a negative

shift of the polarization curves. A parametric study on PEM fuel cell performance and

the corresponding polarization curves can be found in Wang L., et.al.

All voltage losses can be divided into two groups. The first group involves local

losses due to the transport and kinetic processes in and across the membrane electrode

assembly (MEA). The second group involves global losses caused by the along-the-

channel nonuniformity of feed gases and water concentration (Kulikovsky, 2002).

Hydrogen and Reforming Processes

Hydrogen fuel is the essential part of a fuel cell system but also one of the more

difficult problems to overcome when speaking about the possibility of fuel cell powered

vehicles. Hydrogen is the most abundant element on the planet. However the process of

extracting this pure hydrogen from hydrocarbons or water is a high energy process.

Some examples of how hydrogen is extracted from hydrocarbons, called reforming, are

steam reforming, partial oxidation, and coal gasification. Electrolysis is a process of

separating water into pure hydrogen and oxygen gases which would be ideal because of









the lack of any harmful contaminants. However, electrolysis is a very energy intensive

process.

Although pure hydrogen fuel is the ultimate fuel, it is likely that the first

commercialized fuel cell powered vehicles (FCV) will have to use a hydrocarbon fuel

comparable to fuels used in the internal combustion engine (ICE). This is because there

will need to be a transition period from today's fueling infrastructure to a widespread

infrastructure for the production and supply of pure hydrogen fuel. The first fuel cell

vehicles will be expected to perform alongside ICE engines and FCVs emitting

contaminants from reforming processes. Thus the problem of air contamination is as an

important problem to address as the hydrogen fuel problem.

Fuel Cell Contamination

As large scale commercialization of PEM fuel cells draws closer, efforts must turn

to issues relating to the environment in which these fuel cells will operate. Fuel cell

development to date has taken place in a controlled environment of a laboratory, free

from real-world contaminants. Unless the oxidant is supplied to the cells from a

contained source (e.g., bottled air), impurities present in the immediate atmosphere may

adversely affect their performance.

There has been much research done to date that deals with the negative effect

contaminants have on fuel cell performance, however a large majority of that research

was addressed to contaminant effects on the anode of the fuel cell. Because most of the

hydrogen fuel used to power fuel cells is reformed from a hydrocarbon there are

impurities, particularly CO, that need to be removed from the reformate before it enters

the fuel cell.









PEM fuel cells currently use noble metal catalysts supported on high surface area

carbons as active cathode catalyst layers for oxygen reduction (Larminie & Dicks 2002).

As the oxygen reduction process occurs via a surface decomposition process, a large

catalyst surface area increases the number of reaction sites, thereby increasing catalyst

utilization. Any impurities that can be drawn in from the environment that may block

these active catalyst sites will decrease the rate of oxidation and thus decrease overall

performance.

It is therefore critical to understand which contaminants affect the fuel cell's

performance and at what concentrations these effects become problematic. It is

important to also observe if the damage is reversible or permanent. Concentration and

exposure time are two important factors in the possible degradation of the catalysts.

Once the contaminants effects and the methods of contamination are understood then the

necessary steps can be taken to mitigate the adverse effect.

Types of Contaminants

There are many contaminants that exist in the ambient air that can have a

detrimental effect on the catalyst of the fuel cell, however the concentrations at which

they are hazardous are found only in more urban environments. Some of these are CO,

NO2, SO2, Benzene, and 1, 3 Butadiene. These contaminants are the type of pollutants

that are emitted in automobile exhaust. Carbon monoxide is a gaseous byproduct of

incomplete combustion in internal combustion engines. Nitrogen oxides are released into

the environment as a result of fossil fuel combustion. Nitric oxide is common in auto

exhaust but rapidly oxidizes to form nitrogen dioxide. Sulfur dioxide is generally emitted

from diesel burning engines. Volatile organic compounds such as benzene and 1, 3

butadiene are found in both gasoline and diesel engine exhaust.









When considering the use of fuel cells in a battlefield environment there are several

chemical warfare agents that could affect performance such as, sarin, sulphur mustard,

cyanogens chloride (CNC) and hydrogen cyanide (HCN). These particular warfare

agents have been known to seriously compromise the performance of the fuel cell.

Previous Research on Contamination of the Anode

Much of the research involving contaminants and PEM fuel cells has been focused

on the carbon monoxide problem when using reformed hydrocarbon fuels. Pure

hydrogen as a fuel source has many limitations for use in automobiles. This fuel must be

stored either in compressed gas form, cryogenically stored in liquid form or adsorbed

onto a metal hydride. Each of theses storage mechanisms have major disadvantages,

therefore the literature suggests hydrogen will be supplied by reforming a hydrocarbon

fuel such as methanol. The reformation of methanol results in a gas mixture of about 74

percent hydrogen, 25 percent carbon dioxide and 1-2 percent CO (Divisek et al., 1998).

Carbon monoxide levels can be reduced to as low as 5 ppm using separate gas treatment

processes.

Most PEM fuel cells use a platinum catalyst because of its effectiveness in

hydrogen oxidation at low operating temperatures. But even a very small concentration

of CO (<10 ppm) in the hydrogen fuel stream can substantially reduce the performance of

the fuel cell. CO chemisorbs on the platinum sites to the exclusion of hydrogen. This is

possible because the CO is more strongly bonded to platinum than hydrogen, as indicated

by a greater potential required for the oxidation of CO than hydrogen, and the sticking

probability of CO on platinum is 15 times stronger than that of hydrogen on platinum

(Divisek et al., 1998). So even very small concentration of CO in the fuel can result in










complete coverage of the catalyst's active sites for oxidation and therefore the

performance will drastically decrease.

A study conducted by Divisek et al. found significant drops in performance when

CO was present in the hydrogen fuel stream. They found that fuel cell performance

depends strongly on CO concentration and the catalyst used. They also found that the

recovery of a fuel cell to initial cell voltage is shorter by a factor of two as compared to

poisoning time. The graph showing recovery is shown below.




800
H/CO -0 H,
700 A 6 6



W<40t0-.H` O ........ = L-


1D 00 ---------- ------------- ------ -- -- -- -
300

200W
100
0 :1--+-II
0 20 45 70 95 120 145 170 195 220
time / min

Figure 2-2. Cell voltage after changing the fuel gas at 400 mA cm-2; anode catalyst;
Pto.5Ruo.5; pure H2 and H2/100ppm CO; T = 80C (Divisek et. al, 1998).

Methods of Mitigating the Contamination

There are five methods available to mitigate the effect of contaminants poisoning

on PEM fuel cell catalysts. They are the use of a platinum alloy catalyst, the injection of

oxygen in the fuel stream, the injection of hydrogen peroxide (H202), increasing the fuel

cell operating temperature, and effective water management.









Platinum is used because it offers high oxidation activity at the low operating

temperatures associated with PEM fuel. However, they are susceptible to poisoning in

the presence of very low concentrations of an impurity. Platinum alloys such as Pt-Ru,

Pt-Rh, and Pt-Au can increase the tolerance to contaminants whilst keeping a high

exchange current density. This is because of the lower oxidation potential of the alloy

metal compared to Pt.

The injection of oxygen into the fuel stream is done in two ways. The first method

directly injects between 2-5 percent oxygen into the anode, which can increase the

tolerance of CO up to 500ppm. The second is the injection of H202 in the anode

humidifier. The H202 decomposes to hydrogen and oxygen and the oxygen acts to

oxidize the CO and improve the tolerance of the catalyst (Baschuk & Li, 2001).

However this method of mitigation is not useful for the cathode.

A higher operating temperature has been shown to increase tolerance to

contaminants. However higher temperatures in PEM fuel cells is not feasible because the

membrane must be humidified and a higher temperature would accelerate the evaporation

of the liquid water and hence dehydrate the membrane.

A partnership between Donaldson Co and the Los Alamos National Laboratory

called the Fuel Cell Contamination Control (FC3) has been established to study the effect

of ambient contaminants on the performance, life and durability of PEM fuel cells.

Donaldson hopes to use its filter expertise to develop air filtration systems that can

prolong a fuel cell engine's life. The research being done by this group has not been

disclosed and a full report is expected in the spring of 2005. A specialized filter is

expected to be the solution to air contamination problem.









Challenges for Fuel Cell Vehicles

The single most important challenge for the acceptance of fuel cell vehicles (FCV)

as a replacement of internal combustion engines (ICE) is the need for the FCV

technology to match or exceed the existing ICE technology. This is not a simple

challenge because of the many limitations a new technology like FCV presents. Some of

those limitations are fuel infrastructure, on-board storage of fuel, on-board fuel

reforming, start-up time, drivability (vehicles transient response to change in demand),

catalyst contamination and most important, cost.

The simplest and best fuel cell systems for powering vehicles are those in which

the fuel is converted directly in the fuel cell without the need for any pre-processing.

There are four contenders, namely, the direct hydrogen PEMFC, the direct hydrogen

alkaline fuel cell, the direct methanol PEMFC, and the direct methanol or hydrogen fuel

cell with a liquid acid electrolyte (McNicol et al., 2001). However it is likely that the

first commercialized FCVs will use a liquid fuel that is compatible with the existing

fueling infrastructure that exists today. Consequently these first FCV's will also be

producing contaminants into the air and pose the same problem with air contamination as

the ICE engines.

Air Contamination

Fuel cell powered vehicles will need to draw in ambient air from their surroundings

to supply the oxidant needed. Storing the oxidant onboard the vehicle is not an expected

solution because of the additional weight an oxygen storage unit would present. The

various contaminants that exist in ambient air have been shown to have an adverse effect

on the fuel cell, particularly in polluted areas such as urban roadways, high traffic volume

areas, industrial areas and even airport environments.









Carbon Monoxide

In the United States the exhaust from automobiles and trucks account for up to 60%

of the CO released into the air. In major urban areas, motor vehicles are responsible for

95% of CO emissions. CO disperses quickly in air, so moderate and high levels of the

gas are usually detected in areas with significant motor vehicle traffic or within enclosed

spaces where CO may accumulate (EPA, 1997). Several air quality reports from the

United States and the United Kingdom were reviewed. CO levels were much lower in

rural areas and non-rush hour traffic, in the range of 1-32 ppm. Levels in major urban

areas such as Los Angeles and New York City where pollutants would accumulate in the

"valleys" between buildings where rush hour traffic would emit exhaust for hours could

be as high as 67 ppm CO in extreme situations.

Table 2-1. Typicallevels of carbon monoxide according to various air quality reports.
CO levels Mean Level Maximum Level

City 12 ppm 30 ppm

Rural 5 ppm 15 ppm


Nitrogen Dioxide

In the United States auto exhaust accounts for about 30-40% of nitrogen dioxide

emissions. Nitrogen dioxide is produced by oxidation of atmospheric nitrogen and nitric

oxide. Although nitrogen dioxide emissions have been on a decline due to the

introduction of catalytic converters installed in vehicles it still presents a contamination

problem for FCVs. Below are some of the levels of NO2 found in various literature.









Table 2-2. Typical levels of nitrogen dioxide according to various air quality reports.
NO2 levels Mean Level Maximum Level

City 200 ppb 400 ppb

Rural 80 ppb 200 ppb


These two contaminants were chosen based on a review of contaminants testing on

fuel cells. It is known that any potential negative effect these two contaminants may have

on a fuel cell is reversible for CO and NO2. Several other contaminants such as benzene,

1, 3 butadiene, and sulfur dioxide were considered but were found to have irreversible

damage to the fuel cell. So in the interest of not permanently degrading the fuel cell

purchased for this study, CO and NO2 were chosen as the two contaminants used to study

air contamination effects.














CHAPTER 3
EXPERIMENTAL SYSTEM

A 1.2 kW proton exchange membrane fuel cell manufactured by Ballard Power

Systems was purchased for this experiment. The air inlet of the fuel cell was modified so

contaminant gases could be inserted into the air stream at desired concentrations using

flow measuring devices and mass flow controllers. A LabView program was written to

control the flow rate of contaminant gases based on the required air flow rate of the fuel

cell. Finally the contaminated air mixtures were analyzed using a gas chromatograph to

confirm that the proper mixing of air and contaminants was being achieved.

Ballard 1.2 kW Nexa Power Module

Ballard Power Systems were partly responsible for the renewed interest in fuel cells

for road transportation during the 1990s and began development of PEM fuel cells. The

Nexa fuel cell is one of the first commercially available fuel cells for practical and

educational purposes. The Nexa unit is a small, low maintenance, and fully automated

fuel cell system designed to be integrated into products for portable and back-up power

applications, as well as an educational tool for universities and companies.

At full load the Nexa produces 1200 watts of unregulated DC power at a nominal

voltage of 26 volts, at idle the output voltage is 46 volts. The system operation is fully

automated through various control systems. A control board attached to the outside of

the Nexa receives various signals from several sensors within the fuel cell stack.

Communication into and out of the control board are through a RS-485 serial link to a

computer. The NexaMon OEM software allows the user to read and log several data







19


points such as stack temperature, stack voltage, stack current, hydrogen pressure,

hydrogen concentration, hydrogen consumption, oxygen concentration, air temperature

and purge cell voltage. A figure illustrating the Nexa unit is shown in Figure 3-1.


al Hydogen :1 P'ocecs
Ia Hyarogan Inki A.r Inlet
Priniur ial \
Vent fbrt k


Senior


A Cc'Jeng
Air .lu' el
\


Hydm pen
- 4--a"e
!~e~cuate


Figure 3-1. Ballard 1.2kW Nexa unit and various components.1



1 Courtesy of Ballard Nexa Power Module Operating Manual.









Nexa Subsystems

Hydrogen system

The Nexa unit operates on pure hydrogen from a contained source. The hydrogen

system monitors and regulates the supply of hydrogen through a pressure transducer,

pressure relief valve, solenoid valve, pressure regulator, and a hydrogen leak detector.

All these components ensure there is an adequate fuel supply while also maintaining safe

operating conditions for indoor use. Nitrogen and water migrate from the cathode to the

anode through the membrane and accumulate in the anode. This is monitored by

checking the cell voltage in a few key cells. Once the average voltage of these few cells

drops to a certain level they a purged. The purged hydrogen is discharged into the

cooling air system where it is diluted to a level far below the lower flammability limit. A

hydrogen leak detector ensures that hydrogen level is well below the critical limit. As

with any hydrogen system safety is of absolute importance.

Oxidant air system

The oxidant required for the fuel cell reaction is provided from ambient air drawn in

through an air compressor in the fuel cell stack. An intake filter removes particulates

found in air but is not able to protect the fuel cell stack from contaminant gases such as

CO and NO2. The load on the fuel cell governs the H2/02 reaction. An increased load

will increase the current and thus requires more oxidant to be introduced. As the load

increases, the compressor speed is adjusted upward accordingly.

The oxidant system also humidifies the incoming air using the water produced by

the fuel cell reaction. The by-product water produced is run through a humidity

exchanger where the incoming air receives both the water and heat from the reaction.

This is necessary to keep the membranes moist for good ion exchange.









Cooling system

The fuel cell stack is air cooled using a cooling fan that blows air over the entire stack.

As mentioned above, purged hydrogen is discharged into the cooling system where it is

diluted far below the lower flammability limit (LFL). This safety control system

automatically shuts down the fuel cell if the H2 concentration reaches 25% of the LFL.

Safety system

The Nexa has several safety systems that prevent equipment damage and allow for safe

operation indoors. Voltages, current, temperatures etc. are all monitored to ensure they

stay within a specified range. A hydrogen leak detector prevents the hydrogen

concentration from reaching the LFL. An oxygen concentration sensor ensures the

adequate amount of 02 is being supplied. A cell voltage checker (CVC) monitors the

voltage of certain cells; a drop in voltage in cells will shut down the fuel cell in order to

protect the stack. The solenoid valve prevents hydrogen from leaving the contained

source when the fuel cell is not in operation and finally the pressure relief valve ensures

no damage to the cells through over pressurizing the fuel cell stack. All these

components combine to provide monitoring, warning and alarms, and shut down

mechanisms to ensure safe operation at all times.

Air/Contaminant Flow Measurement and Mixing

A mixing manifold was constructed to be able to inject contaminants into the

incoming air stream of the fuel cell to get the desired concentration of contaminant going

into the fuel cell. A box was constructed around the air inlet and completely sealed. A

one inch PVC pipe was tapped into the front side of the box. The one inch PVC pipe

provides a place to measure the air flow rate of air being drawn into the fuel cell from the

air compressor as well as a place to inject the appropriate amount of contaminant gases








into the air stream. Using compressed cylinders of air was considered for more accurate

testing but due to the high volume of air the Nexa requires the amount of air cylinders

was simply not feasible. Several options were available for measuring the air flow

entering the fuel cell. A pitot tube/pressure transducer system was chosen as most

educational and economical way to accurately measure the air flow. The entire

experimental set-up can be found in Figure 3-3.


AT


-r/ S FE!


Figure 3-2. Air/Contaminant mixing manifold, Ballard Nexa fuel cell, load bank, gas
cylinders, NexaMon OEM software.


ir/


''''~~II
::









Measurement

The following section describes all the components selected for the measurement of

the air flow and delivery of the contaminant gases and why they were chosen as the best

option for the given application.

Pitot Tube

A model 300 series inline Delta Tube was purchased from Mid-West instruments.

It utilizes two averaging flow elements of equal area to sense stagnation and the static

differential pressure providing minimum permanent pressure loss. The flow elements are

placed at the center of a threaded 8" section of pipe and can be attached to the one inch

PVC pipe on either end. The one inch PVC pipe was chosen to give us a high enough

flow velocity through the pitot tube to get a reliable reading. At the maximum air flow

rate demanded by the fuel cell (90 SLPM) the pitot tube will measure a differential

pressure (dp) of 0.1 in of H20. Therefore the selection of a suitable pressure transducer

to read this extreme low end dp value was critical. A picture of the averaging pitot tube

and swagelok fittings is shown in figure 3-3.

















Figure 3-3. Inline Delta tube with swagelok fittings connected to static and dynamic
pressure reading devices.









Differential Pressure Transducer

A Sensirion SDP 1000 was donated by Sensirion to the University of Florida; it is a

low range differential pressure transducer with a measurement range from 0-2 in of H20.

This pressure transducer was best suited for this application because of its high resolution

at the low end of its measurement range. At less than 30% full scale (FS) the resolution

has a range of 0.0002-0.0008 in of H20. The transducer requires a 5V power supply and

has a linear 0.25-4V output, as shown in figure 3-2.


4.00
SDP1I000-L / SDP2000 -L


3.00



2.00
3
ca-






0 20 40 60 80 100

differential pressure [% full scale]

Figure 3-4. Linear output at 5VDC supply of the SDP-1000-L. The fine lines indicate
the maximum tolerances including a temperature variation from 0-50C.2


2 Courtesy of Sensirion SDP-1000 Datasheet.




































Figure 3-5. Sensirion SDP- 1000 differential pressure transducer.



Mass Flow Controller

In order to control the flow rate of contaminant gases from the cylinder into the air

stream, a Fathom GR series stainless steel mass flow controller was selected and

purchased. The range of the controller is 0-1.2 SLPM. The controller uses a capillary

type thermal technology to directly measure mass flow. The controller was powered by

115VDC and used 0-5VDC linear input and output voltages to monitor and control the

unit from a remote source. Swagelok tubing and fitting were used to transport gases from

the gas cylinders to the controller and eventually to the pvc inlet air pipe to the fuel cell.



































Figure 3-6. Fathom mass controller connected to Swagelok tubing.

Gases

A large cylinder of ultra high purity hydrogen gas was purchased from Praxair. A

2000 psi tank of 30 ppm nitrogen dioxide mixed with air was purchased from Spectra

Gases, as well as a 2000 psi tank of 3000 ppm carbon monoxide mixture with nitrogen.

A specialty stainless steel regulator was purchased for the NO2 gas cylinders. The CO

and H2 cylinders used a standard brass regulator. The concentrations of CO and NO2 in

the cylinders were selected in order to be able to use the same mass flow controller for

each contaminant gas.

Load Bank

A load bank was constructed at the University of North Florida and used to

dissipate the 1200 watts of power. Several resistors that connected in parallel could be









switched on/off were fixed to a large board, a potentiometer was also used to be able to

fine tune the current being drawn out of the cell. This load bank allowed for incremental

steps in current which allowed polarization curves to be created. Listed below are all the

resistors that were used:

* 5 250 watt, 5 ohm resistors
* 11 100 watt, 25 ohm resistors
* 4 100 watt, 5 and 10 ohm resistors placed in parallel yielding 15 ohms
* 1 600 volt, 2.5 amps, 50 ohm potentiometer


The front and back panel of the load bank are shown in figures 3-7 and 3-8.


Figure 3-7. Back view of load bank with several resistors in parallel, constructed at UNF.
Potentiometer was placed at center (not pictured).



































Figure 3-8. Front view of load bank with several switches to activate resistors and
potentiometer for slight adjustments.

Safety

The Nexa unit uses a supply of hydrogen fuel to operate. It is necessary to be

aware of and understand the safety requirements related to hydrogen and compressed

gases. Although the Nexa unit has several safety features that prevent unsafe operation

additional measures must be taken in order to ensure safe conditions in the laboratory.

Hydrogen is an extremely flammable gas; therefore no sources of ignition
were placed near the unit. A large boiler was in place within a distance
from the cylinders for another project but was disconnected from its power
source.

All gas cylinders were secured to the laboratory wall using wall supports
and regulators were all checked for leaks using a liquid leak detector.









* Oxygen depletion is an issue because the Nexa converts oxygen into water.
The laboratory area was kept well ventilated to ensure there was adequate
oxygen.

* Electrical connections checked before every test run to prevent
electrocution or fires.


* Carbon monoxide detectors, smoke alarms and a fire extinguisher were
placed near the Nexa unit.


Figure 3-9. Various safety measures; fire extinguisher, smoke/carbon monoxide detector
and gas cylinder supports.

Gas Chromatograph

A SRI Instruments 8600 series gas chromatograph was used to analyze the

contaminated air mixtures to ensure the mixing manifold was delivering the proper

amounts of contaminants. The gas chromatograph (GC) uses a mol sieve and a column to

physically separate the gaseous specie over time by heating up the column oven. The

specie are then directed into a thermal conductivity detector (TCD). The TCD registers a









potential difference across a Wheatstone bridge. This data is collected and displayed as

peaks over time in PeakSimple software.

The GC first needed to be calibrated with known concentrations of CO before

testing the sample taken from the mixing manifold. Two test gas cylinders with known

CO concentrations were used (20 and 100 ppm). The SRI 8600 is shown in figure 3-8.


Figure 3-10. SRI Instruments 8600 gas chromatograph.














CHAPTER 4
METHODOLOGY

In order to investigate the effects CO and NO2 gases on fuel cell performance

several tests were run using the Ballard Nexa unit. Several preliminary tests were

performed using clean air to obtain baseline performance data. A confidence interval was

calculated using the various clean runs to be able to distinguish between a significant

change and normal fluctuations in performance. Polarization curves were created at

various concentrations of both gases. Constant load test were performed on the selected

concentrations of the contaminant gases at three different loads. Recovery information

was also obtained from the constant load tests by switching from clean air to

contaminated air and vice versa. Data collected was then compared to published data.

Baseline Performance Testing

Polarization Curve Generation

In order to establish the basic performance of the Nexa fuel cell, six test runs were

completed. The Nexa was supplied with ultra high purity hydrogen from a cylinder and

air was drawn from the surroundings, which were held at relatively constant temperature

and humidity. The Nexa was allowed to warm up for a short time so that the initial stack

temperature was the same for each test run. This also ensured the Nexa had reached a

quasi-steady state. Load increments of 5 amps were chosen to give a full spectrum of

data points to create a polarization curve. Starting at an initial load of 5 amps the load

was increased every 5 minutes to ensure the fuel cell had reached a steady state at each

load step. At 45 amps the fuel cell had surpassed its maximum power output of 1200









watts and was producing 1300 watts. Voltage (V), current (amps), power (watts), air

flow (SLPM), air temperature (C), and stack temperature (C) were all logged every 10

seconds at each of the nine load steps. The voltage and current data collected from each

of the six test runs was used to construct a confidence interval. This confidence interval

was used to ensure any changes in voltages due to contamination were significant.

Constant Load Testing

Contamination effects are easier seen under a constant load. At constant load the

voltage is also constant, and any change in voltage can be attributed to the contamination

effects. It is then necessary to also get baseline performance data on how the fuel cell

performs under constant load over time. Three loads were selected to cover the full range

of the Nexa fuel cell stack. The loads selected were 20, 30 and 40 amps. These currents

best reflect the linear portion of the polarization curve, the part of the curve that is likely

to be affected by the contamination. Six runs, each 45 minutes, for each of the three

loads were completed to again form a confidence interval by which we can compare the

constant load contamination testing.

Contamination Testing

Selected concentrations of contaminants are introduced into the air inlet stream of

the fuel cell by using a mixing manifold constructed from a pitot tube, pressure

transducer, pvc pipe, mass flow controller and a LabView program written to control the

process. The flow rate of the air in the pvc pipe leading to the fuel cell is calculated from

the differential pressure reading from the pitot tube using the formula given below:


FlowRatear (SLPM) = 0.1789 Cf D2 A 4-1
W,









Cf= Flow Coefficient = 0.559

D = Pipe I.D. mm = 26.6446

W, = Weight density of air in kg/m3 at 600F and 14.696 psi = 1.22377

AP = Differential pressure mm



The desired concentration of contaminant gas in parts per million (PPM) is input

into the LabView program. Based on the desired concentration input and the air flow rate

calculated the necessary flow rate of contaminant gas is calculated using the formula

given below:

PPM
FlowRate Cona min antGas PPMDd FlowRateAzr 4-2
P CyhnderMzxture




Table 4-1. Concentration of contaminant mixture gases in cylinders.
Carbon Monoxide mixed in N2 Nitrogen Dioxide mixed in air

3000 ppm 30 ppm



Swagelok tubing carried the contaminant gas from the compressed cylinder to the

mass flow controller and into the pvc pipe leading into the air inlet of the fuel cell. The

concentration of the contaminated air is verified using a gas chromatograph. Flow rates

of contaminant gases for different flow rates of air are given in Table 4.2.










Table 4-2. Contaminant flow rates for three air flow rates (30, 60, 90), for pure
contaminant gases (top) and for contaminant gas mixtures (bottom).
Pure Gases Flow Rate (LPM)
Contaminants (PPM) 30 60 90
CO 0 0.0003 0 001- 0.0009
30 0 0009 j 0 0013 0.0027
NO2 0.1 3_0E-06 6 0E-63 9.0E-06
0.4 1.2E-05 2.4E-05 3.6E-05

Gas Mixtures Flow Rate (LPM)
Contaminants (PPM) 30 60 90
3000 CO* I10 0.1 0.2 0.3
PPM mixture 30 0.3 0.6 0.9
30 NO2** 0.1 0.1 0.2 0.3
PPM mixture 0.4 0.4 0.8 1.2
mixed with N2
** mixed with Air

The LabView program was written to control the process of injecting contaminant

gases into the inlet air stream. The program reads the air flow rate into the fuel cell and

then calculates the necessary flow rate of contaminant gases to obtain the desired

concentration input by the user. The front panel of the program graphically displays the

desired concentration input (PPM), the differential pressure produced in the pitot tube (in

of H20), the air flow rate (CCPM), the desired contaminant flow rate (CCPM) and the

actual contaminant flow rate (CCPM), which is feedback sent to the program from the

mass flow controller. The flow can also be read from the mass flow controller itself to

ensure the proper flow rate. A diagram of the front panel and the block diagram can be

found in the appendix A.

Polarization Curve Generation

For each contaminant a polarization curve is generated using the same procedure

used for the baseline performance tests. Two different concentrations for each

contaminant have been selected based on the emissions reports collected in the literature

review. The selected concentrations are shown in Table 4-2.









Table 4-3. Selected concentration for testing based on emissions reports collected in
literature review.
Contaminant Gas First Concentration tested Second Concentration tested

CO 10 ppm 30 ppm

NO2 100 ppb 400 ppb


Constant Load Testing

Constant load tests were also performed; at a constant load the voltage drop due to

the contamination effects will easily be seen. The same procedure used in establishing

the baseline performance will be used. The specific concentration of contaminant will be

tested for 45 minutes at each load (20, 30, 40 amps). If adverse effects are seen at a

certain concentration and load another test will be performed using that same

concentration and load. The Nexa will operate on clean air for 30 minutes followed by a

30 minute contamination then back to clean air for 30 minutes to observe recovery.

The effect each contaminant has on the fuel cell stack will be determined from the

slope of the voltage curve. Higher concentrations may be used if there is no change in

voltage with the concentrations selected

Contamination Recovery

Studying how the fuel cell stack recovers from harmful effects is as important as

studying the effects themselves. Therefore after the constant load test the contaminant

source will be turned off and the Nexa will be allowed to run on fresh air to see if the

voltage returns to the initial voltage. The literature suggests that CO and NO2 both have

reversible effects after contamination. Several contaminants were not selected to be used

in this report because of their known irreversible effects and it is not out intent to

permanently degrade the Nexa fuel cell purchased for this experiment.






36




Temperature Effects

To illustrate the different parameters that affect the polarization curve and hence

the performance of the fuel cell, the effects of temperature will be studied. It is known

from the literature that increased stack temperature increases the performance of the fuel

cell. Polarization curves will be calculated at four different temperatures in increments of

10F starting at 600F. It is expected that as the temperature increases the stack voltage

will increase.

















CHAPTER 5
RESULTS AND DISCUSSED

Baseline Polarization Data

Several polarization curves were generated to obtain baseline performance data of


the Nexa fuel cell under normal operating conditions. All tests were conducted at 70F


and 55% relative humidity. Figure 5-1 displays the polarization curves generated


Nexa Fuel Cell Baseline Performance


20 O
0 5 10 15 20 25 30 35 40
Stack Current (amps)

Figure 5-1. Six polarization curves generated under normal operating conditions, 700F,
atmospheric pressure and 55% relative humidity.



Based on the six test runs a confidence interval was constructed and is shown in


figure 5-2.


45




40




35



30
25 3




25


45 50












Confidence Interval for Polarization Curve

45

40

35

30

g
S25

0 20

15

10

5

0
0 5 10 15 20 25 30 35 40 45 50
Stack Current (amps)



Figure 5-2. Confidence interval of polarization curves based on six tests.

Baseline Constant Load Data

Although polarization curves give excellent information about the different


parameters that affect fuel cell performance it is easier to see the contaminants effect on


performance by testing the contaminants under a constant load. The three loads selected


were 20, 30 and 40 amps. These three loads best reflect the linear portion of the


polarization, the part of the curve that is most affected by changing operating conditions.


Figures 5-3 and 5-4 show the stack voltage and the stack power for constant loads of 20


and 40 amps, the data for 30 amps is similar to these two graphs and hence not shown.


Data was collected every 10 seconds for 45 minutes.
















Clean 20 amps Run 1


. r nY (h. nir AIrn Mnn .f r --- I a68UU


34


32


S30
E

o 28


26


> 24


22


20


18


0 10 20 30 40 50
Time (minutes)

-*-Stack Voltage Load -m-Stack Power




Figure 5-3. Stack voltage and stack power at constant load of 20 amps.



Clean 40 amp Run 3


45




41

a 39


37


33




3 31


29


27


25
0 10 20 30 40 50
Time (minutes)

--- Stack Voltage Load --w- Power]




Figure 5-4. Stack voltage and stack power at constant load of 40 amps.


750




700




650



600




550


1400



1300



1200








1000


Polff-rWN 1. M


e
-


-









To ensure that any change in stack voltage or power is significant when introducing

contaminants into the air stream an uncertainty interval was constructed for stack voltage

and stack power for each load and is shown in table 5-1.

Table 5-1. Stack voltage and stack power allowable ranges under normal operation
conditions.

Constant Load Stack Voltage (V) Stack Power (watts)

20 amps 34.5 + 0.155 690 4.3

30 amps 33.4 0.142 990 6.1

40 amps 30.5 + 0.138 1225 9.8


Contamination Results

Carbon Monoxide

The concentrations of carbon monoxide selected were chosen based on several air

quality reports gathered from the literature review. A mean value of 10 ppm represented

the type of concentration that would be found in typical urban areas. A maximum value

of 30 ppm represents the higher concentration that would exist in a high traffic volume in

a busy city area with little ventilation, (so called 'street canyons'). Both concentrations

were tested at the three selected constant loads. Polarization curves were also generated

using load increments of 5 amps up to the maximum 45 amps.

Constant load

The Nexa fuel cell was allowed to operate on clean air until it had reached a quasi

steady state. The selected concentration of carbon monoxide was then introduced at a

constant load and allowed to operate for 45 minutes. If any voltage drop was detected at

the end of the 45 minutes the fuel cell was allowed to continue running on clean air to

obtain recovery information.







41


The concentrations of 10 and 30 ppm were each tested at three loads (20, 30 and 40


amps). Neither concentration showed any significant change in either stack voltage or


stack power throughout all the tests when compared to the clean air tests. Figure 5-5


shows the 10ppm test at a constant load of 40 amps.


Carbon Monoxide (10 ppm) 40 amps

43 1400

41 1350

39
1300

" 1250

S1200
33 a.
o 1150
,r 31

1100
29

27 1050

25 1000
0 5 10 15 20 25 30 35 40 45 50
Time (minutes)
-*-Stack Voltage Load ----Power


Figure 5-5. Carbon monoxide test (10 ppm) at constant load of 40 amps.

The voltage held relatively constant at 30.5 volts while the power remained at


around 1225 watts. The fluctuations seen in figure 5-5 and in all the data collected are


due to the purge cell cycles. The Nexa fuel cell automatically performs these purge


cycles every 40-60 seconds. The voltage of certain individual cells within the stack are


monitored, once the voltage drops to a designated critical value the entire stack is purged


of the hydrogen fuel to release any contaminants on the anode side. This purge cycle


briefly increases the stack voltage.








42



The 10 ppm tests at the constant load of 20 and 30 amps produced similar results as


shown in figure 5-5 with the voltages and power remaining within the uncertainty


interval defined.


Similarly, the tests performed with 30 ppm did not produce any significant changes


in the fuel cell performance as can be seen in figure 5-6 at a constant load of 20 amps.


30 ppm Carbon Monoxide at Constant Load of 20amps

38 900

36
850
34
S800
E 32
3 750
2630

28 ----700
a)E
26--------------------------------------------------------------
a 26 -
S- 650
S24
S- 600
22
550
20

18 500
0 5 10 15 20 25 30 35 40
Stack Current (amps)
-*- Stack Voltage Load --- Power


Figure 5-6. Carbon monoxide test (30 ppm) at constant load of 20 amps.


Tests at 30 and 40 amps produced a similar graph to figure 5-6 and hence are not


shown because of the insignificant change in performance.


The data collected from the 10 and 30 ppm tests showed that the Nexa can tolerate


the typical levels of carbon monoxide that would be encountered in under normal traffic


conditions. Therefore, further review of the literature was conducted to find if there


existed conditions where the concentrations of carbon monoxide would be higher. A









study conducted at the Lincoln Tunnel in New York City found that levels as high as 100

ppm could be found during extremely heavy traffic1. Also, the concentration of carbon

monoxide in diesel emissions can be as high as 500 ppm; a FCV within a short distance

from direct diesel exhaust may encounter levels near 100 ppm after considering the

exhaust would be diluted somewhat in the air.

The Nexa was subjected to 100 ppm carbon monoxide and the data revealed a drop

in performance during the 45 minute test. To better understand the effect of the

contaminant a different test methodology was used, the Nexa was allowed to run for 30

minutes on clean air followed by 30 minutes of 100 ppm CO and then allowed to operate

on clean air for 30 minutes to show any recovery. Figure 5-7 shows this test performed at

a constant 20 amps.

The introduction of 100 ppm carbon monoxide produced a 1.67% drop in stack

voltage and a 1.61% drop in overall stack power. It is interesting to see that the carbon

monoxide seemed to affect the fuel cell immediately but it did not continue to degrade

the voltage over time, it remained in a quasi steady state. When looking at the

contamination portion of the graph in figure 5-7 the higher values of stack voltage are

due to the purge cycles which were explained previously, this rapid purge produces a

brief increase in voltage and power but then fell rapidly until the Nexa is purged again.


1 A study conducted by the CDC, 1993.








44



Clean air 30 minutes, 100ppm CO for 30 minutes, Clean air 30 minutes at a constant Load of 20 amps


40

39

38

37

36

35

" 34

33

32

31

30


Begin Flow of CO stopped,
Contamination allowed to recover
(100 ppm) using clean air


0 10 20 30 40 50 60 70 80 90
Time (minutes)

-- Stack Voltage -U- Power




Figure 5-7. Carbon monoxide test (100 ppm) at constant load of 20 amps.


Despite the reduction in stack voltage and power the Nexa did recover. The stack


voltage recovered to 98.97% of its original voltage while the power recovered to 98.96%


of its original power output before contamination.


This same concentration was performed at a higher load of 30 amps however the


test could not be run at 40 amps because of the limitation of the mass flow controller.


The range of the mass flow controller was selected to give excellent accuracy at the 10


and 30 ppm concentrations and was therefore out of range for the air flow demanded by


the fuel cell at 40 amps. Figure 5-8 displays the results from the 30 amp test.


0


-550



-500




450












Clean air for 30 minutes, 100 ppm Co for 30 minutes, Clean air for 30 minutes at a constant load of 30 amps

40 1050

39
1000


37
950
36
Begin
35 Contamination Flow of CO stopped, 900
S(100 ppm) allowed to recover
34 using clean air
3 850


32
800
31

30 750
0 10 20 30 40 50 60 70 80 90
Time (minutes)
-*-Stack Voltage -- Power


Figure 5-8. Carbon monoxide test (100 ppm) at constant load of 30 amps.


The stack voltage was reduced by 2.58% while the stack power was reduced by


3.30%. Although here the contamination seemed to steadily decrease the voltage as time


went on. Again the Nexa seemed to recover after contamination, the stack voltage


recovered to 98.69% or its original voltage and the power recovered 98.03% of its


original power output.


The drop in performance was more severe at the higher potential. The mechanism


for adsorption of CO onto platinum appears to be a function of both potential and CO


concentration. These concentrations of carbon monoxide (100 ppm) represent extreme


cases of pollution a fuel cell vehicle may encounter in an urban area with heavy traffic


and little ventilation. A step to mitigating this adverse effect would be to use the popular


alloy catalyst Pt-Ru, which can be effective up to 100ppm CO.











Polarization curves

Polarization curves were generated for 10, 30, 60 and 100 ppm CO. Beginning at 5


amps the load was increased increments of 5 amps every 5 minutes. As previously noted


10 and 30 ppm CO did not have any significant change on the fuel cell performance, the


polarization curves generated for these two concentrations fell within the confidence


interval shown in figure 5-2. Figure 5-9 shows polarization curves for 60 and 100 ppm.


The 60 ppm curve falls right outside the interval and therefore has and adverse effect on


the performance. Data points for 40 and 45 amps in the 100 ppm CO test could not be


taken due to the mass flow controller issues discussed.



Polarization Curve for 60 and 100 ppm


45
60 ppm Carbon
Monoxide


40
Clean Air with confidence
interval based on 6 test
runs.

>35

0


30


100 ppm Carbon
Monoxide
25




20
0 5 10 15 20 25 30 35 40 45 50
Stack Current (amps)

-polarization curve -A-60ppm -+ 100 ppm


Figure 5-9. Polarization curves for 60 and 100 ppm carbon monoxide.









Rapid injections of carbon monoxide

Upon realization that typical levels of carbon monoxide detected on urban area

streets had no effect on the performance of the fuel cell, attention was drawn to more

extreme cases where carbon monoxide might be at much higher levels. For example, if a

fuel cell vehicle was in traffic behind a diesel truck emitting its exhaust directly in the

vicinity of the air intake for the PEM fuel cell in the vehicle, this would yield a much

higher concentration.

A test was designed to try to recreate this type of situation in which a brief rapid

increase in CO concentration might be introduced into the fuel cell and look at the

recovery. The test was done for two constant loads; 20 and 30 amps. The fuel cell was

allowed to operate on clean air until a quasi-steady state was reached. When this state

was reached the fuel cell was allowed to operate for 60 seconds before introducing 100

ppm for 60 seconds. The CO flow was then stopped and allowed to recover for 60

seconds before a higher concentration was introduced. This procedure was continued for

concentrations of 120, 140 and 160 ppm. This test was also completed for 10, 30, 60 and

80 ppm to find the concentration at which contamination effects begin to take place. No

effects were seen until the concentration reached 100 ppm CO. Figure 5-10 shows this

test at a load of 20 amps.

Despite some minor fluctuations in the stack power and voltage, the data fell within

the range specified in table 5-1 and is therefore not significant up to and including 90

ppm. Significant drops in performance occurred once 100 ppm CO was injected. These

results are looked at closer and can be found in figure 5-11. This figure shows the test

completed for concentration 100-160 ppm at a constant load of 20 amps.








48




Constant Load of 20 amps with injections of CO (10,30,60,80 ppm)


900



850



800



750



700



650


0 1 2 3 4 5 6 7 8 9 10
Time (minutes)

-- Stack Voltage -- Pwer



Figure 5-10. Rapid injections of CO (10, 30, 60, 80 ppm) at constant load of 20 amps.



Constant Load (20 amps) with injections of Carbon Monoxide (100,120,140,160ppm)


38



37



Injection of 120ppm Injection of 160ppm
36 CO for one minute CO for one minute


- / I
- 35



34
Injection of 100ppm
CO for one minute
Injection of 140ppm
CO for one minute
33



32
0 1 2 3 4 5 6 7 8 9 10
Time (minutes)



Figure 5-11. Rapid injection of CO (100,120,140 160 ppm) at constant load of 20 amps.


Clean 10 ppm Clean 30ppm Clean 60ppm Clean 80 ppm Clean
- _,









The rapid injections of contaminants showed that effect of CO is immediate, on the

other hand the time for the fuel cell to recover to it original voltage was also very rapid.

In some cases the voltage actually improved after recovery. This minor rise in stack

voltage may due to the fact that under contaminated conditions the fuel cell stack may be

heating up, which would result in better performance of the stack. Raising the

temperature of the stack is another option for mitigating the negative effects of

contaminants.

Table 5-2. Percentage changes in voltage and power during rapid CO injections at a
constant load of 20 amps.
Concentrations (ppm) Percentage drop (Voltage/Power) Percentage Recovery(Voltage/Power)

100 0.82% / 1.21% 99.89% / 100.1%

120 1.34% / 1.33% 99.84% / 99.61%

140 2.11% / 2.54% 100.34% / 100.34%

160 3.02% / 3.18% 100% / 100.1%



The same test was conducted at a higher load of 30 amps and resulted in a

significantly higher drop in performance and is shown in Figure 5-12. The 30 amps test

was similar to the 20 amps test but the percent changes in performance were much higher

as shown in table 5-3.

These results again show that higher the stack potential results in a larger drop in

performance. When considering a vehicle powered by a PEM fuel cell this adverse effect

become a critical issue. Acceleration of a FCV requires a steep increase in power of the

stack. The contamination of the fuel cell would greatly inhibit the fuel cell from

producing the power necessary, and hence reduce the overall drivability of the vehicle.












Constant load (30 amps) CO injections (100,120,140,160ppm)


38


36


34


32

0
- 30


28


26


24


0 1 2 3 4 5
Time (minutes)


6 7 8 9 10


Figure 5-12. Rapid injection of CO (100,120,140 160 ppm) at constant load of 30 amps.

Table 5-3. Percentage changes in voltage and power during rapid CO injections at
constant load of 30 amps.
Concentrations (ppm) Percentage drop (Voltage/Power) Percentage Recovery(Voltage/Power)

100 1.08% / 1.15% 102.01% / 1(.99 %

120 4.22% / 4.61% 100.49% / 100.58%

140 8.04% / 8.42% 100.64% / 100.19%

160 12.39% / 12.52% 100.56% / 100.99%



Nitrogen Dioxide

The concentrations selected to represent the average and maximum levels found in


and urban area were 100 ppb and 400 ppb respectively. These concentrations were tested


separately at three constant loads (20, 30 and 40 amps). Polarization curves and the rapid


injection tests were also completed as in the carbon monoxide tests.


Injection of 120ppm Injection of 160ppm
CO for one minute CO for one minute





-y-4 ~r\ Wr


XU~


Injection of 140ppm
CO for one minute


Injection of 100ppm
CO for one minute








51



Constant load


The constant load tests conducted for nitrogen dioxide were the same conducted for


carbon monoxide. The Nexa was allowed to operate on clean air to reach a quasi-steady


state before the selected concentration of nitrogen dioxide was introduced. The mass


flow controller was designed for carbon monoxide flow, so a conversion factor was


needed to use NO2. The flow controller uses a thermal sensor technology which allows


the use of conversion factors from the calibrated gases to other gases. The conversion


factor, also called a K factor, was 0.74 percent of the flow rate reading.


It was found that neither the 100 ppb nor the 400 ppb had any significant change in


performance as can be seen in figures 5-13 and 5-14.



Nitrogen Dioxide (100 ppb) at constant load of 20 amps


34-- -

32

30

PR


6--

4--
76 1-----------------1---------------------


t4 1 ------------------------------------ T


20


0 5 10 15 20 25 30
Time (minutes)

-o- Stack Voltage Load -a- Power


35 40 45


Figure 5-13. Nitrogen dioxide test (100 ppb) at constant load of 20 amps.


o
.o

r0


o

V5


2

2








52




Nitrogen Dioxide (400ppb) at constant load of 20 amps


36


34


32


E
S30

0
-o
28


o 26

0
,24
o

22


20


18


0 5 10 15 20 25 30
Time (minutes)

-*-Stack Voltage Load -- Power


35 40 45


Figure 5-14. Nitrogen dioxide (400 ppb) at a constant load of 20 amps.




Higher concentrations were tested at constant loads to find any drop in


performance. The highest concentration was limited by the range of the mass flow


controller. The maximum concentration that could be sustained by the controller was


1000 ppb or 1 ppm. Again this concentration did not yield any adverse effects as shown


in figure 5-15.


MUU




750




700




650




600




550




500


0
m




a

U)
s
J
0,







53



Nitrogen Dioxide 1ppm at Constant Load of 30 amps

36 1200

35 1150

34 1100


3 33 1050 --
0 V5



0 31 950

30 900


29 850


28 800
0 10 20 30 40 50 60 70 80 90 100
Time (minutes)
-*-Stack Voltage Load -m-Power


Figure 5-15. Nitrogen dioxide (lppm) at constant load of 30 amps.

Rapid injections of nitrogen dioxide

The rapid injection tests performed with carbon monoxide yielded the most


interesting data and hence was again performed with nitrogen dioxide. The limits of the


mass flow controller prevented testing concentrations above 1 ppm N02 for the constant


load test because it required the flow controller to sustain a high flow rate for a


substantial amount of time. However the controller was able to handle high flow rates for


a short period of time (-1 minute) so higher concentrations could be reached in the rapid


injections tests. These tests were run at 20 amps because of the lower air flow rate


required by the air flow. The results are shown in figure 5-16.











Constant Load of 20 amps with N02 injections (800,1000,1200,1400 ppb)

38



37



36
Inject 800 ppb for Inject 1 2 ppm for
one minute one minute




34


Inject 1 ppm for Inject 1 4 ppm for
33 one minute one minute


32
0 1 2 3 4 5 6 7 8 9 10
Time (minutes)


Figure 5-16. Rapid injections of nitrogen dioxide (800, 1000, 1200 and 1400 ppm) at
constant load of 20 amps.

The results of the rapid injection tests did not yield any significant drop in


performance and it was concluded that the Nexa fuel cell was able to tolerate nitrogen


dioxide with no adverse effects. This agrees with what was found in the literature


(Moore et. al).


Temperature Effects

Temperature is one of the many factors that influence the performance of a fuel


cell. It was gathered from literature that an increase in temperature increase the overall


performance of the fuel cell. With increased temperature the exchange current density


increases which reduces the activation losses. However there is a limit to how high the


temperature can be due to humidification issues. If the membrane material in the catalyst


is not fully hydrated this could cause a decrease in active surface area in the catalyst. It












can be seen that at low current the effects of temperature are much more significant. This


may be due to the fact that at higher temperature the membrane may not be fully


hydrated. But with increases in current the rate of water production also increases


proportionally which in turn keeps the membrane moist and the active sites on the


catalyst open and hence improves the fuel cell performance. The effects of temperature


are shown in figure 5-17.



Temperature Effects on Polarization Curves


1600


A' 1400


-1200


1000


800


600


400


___ 200


0
5 10 15 20 25 30 35 40 45 50
Stack Current (amps)
S--T=50F ---T=60F -- T=70F -A-T=80F
-ix-T=50F -- T=60F --T=70F -A-T=80


Figure 5-17. Temperature effects on fuel cell performance. Polarization curves and
power curves shown for temperatures (500F, 600F, 700F and 800F).


45


40


35


30

ca
" 25
0
0
-r 20
U)
15


10















CHAPTER 6
CONCLUSION AND RECOMMENDATIONS

Based on experimental data it appears that the typical levels of carbon monoxide

and nitrogen dioxide found in urban areas do not significantly affect the performance of a

PEM fuel cell over the time interval tested. However in the experiment, the fuel cell

performance began to drop at CO levels of 100ppm and higher. Although very high,

these levels of CO could be found under extreme conditions in a heavy traffic volume

area.

The carbon monoxide poisoning mechanism seems to be a function of both cell

potential and CO concentrations. The drops in voltage and stack power were immediate

after introduction of the CO. It should be noted that for the time interval tested the

adverse effects seemed stay in a quasi-steat state. Future work should also consider the

effect of time on contamination. The Nexa fuel cell had a very high hydrogen

consumption rate and thus longer timed tests were not feasible.

The transient response tests revealed important information about the PEM fuel cell

ability to recover rapidly after the cessation of the contaminant. The fuel cell was able to

recover completely despite the brief injections of very high concentrations of the

contaminant.

The experiments involving nitrogen dioxide did not appear to have any significant

effect on the performance up to concentrations of 1.4 ppm in the time interval tested.

One reason the NO2 may not have affected the performance of the fuel cell at the

concentrations tested was the fact that nitrogen dioxide is extremely soluble in water.









The incoming air stream passes through a humidity exchanger, this process of adding

moisture to the air stream may have scrubbed out the NO2. The NO2 actually reacts with

water to produce nitric acid and NO:

3 N02(g) + H20(l) 2 HNO3(aq) + NO(g) (6-1)

However, if this reaction was taking place, there should have been some change in

performance because of the effect of the nitric oxide now present in the air. In future

work, the contaminants should be introduced into the air stream after the air has been

humidified. Another solution would be to not humidify the air at all, since the water is

being produced on the cathode already, it is more important to humidify the fuel stream.

This solubility effect was not likely an issue for the CO because it is almost insoluble in

water.

This study was primarily concerned with short-term exposure tests, longer tests

may reveal additional information about the contamination mechanism.

Although the Nexa fuel cell stack was a great learning tool for someone unfamiliar

with the technology, it is not ideal for experimentation. The Nexa has several subsystems

that protects itself from any unusual operating conditions. In future work in studying the

effect of air contaminants on the cathode, efforts should be turned to designing and

constructing individual single PEM fuel cells. Constructing several different membrane

electrode assemblies (MEAs) would allow a researcher to get direct information about

contaminant effects. There were many variables in the Nexa testing that were unknown

because the Nexa is a complex system.

With single PEM fuel cells, more severe contaminants with possible irreversible

effects could be tested. Also with single cells, the effect of operating parameters such as











pressures, temperatures and humidity temperatures on the negative effects of


contamination could be studied. The ability to control all these parameters would give a


clearer understanding of the contamination mechanism. A sample experimental system is


shown in figure 6-1.



constant temperature inon haust to
nd humidity generator



Data
Reservoir Flowmeters Sytm

\ Detector (GC)

Temperature and
humidity sensor


AdsarberCarbon vfumeo
Pump concentrationfilter cupbheckrd
Pressure sensor Mixerthe
Exhaust to
Pump |fume
cupboard

Challenge gas Carbon
generators platinum ct filter

Adsorber for volumetric
concentration check


Figure 6-1. A sample experimental system containing a single PEM fuel cell and various
components to control all operating parameters (Moore et. al)

Researching and testing different catalysts used in the electrodes is also


recommended for future work. Platinum alloy catalysts can significantly enhance the


tolerance level to harmful contaminants while also reducing the costs of the very


expensive pure platinum catalysts.



















APPENDIX A
LABVIEW COMPUTER PROGRAM


The following is the block diagram and front panel from the LabView program


written to control the flow rate of contaminant gases based on the flow rate of air going


into the fuel cell.



Fie Edit Operate Tools Browse Window Help

A


i' 7 trt iE&IE c r 4 N tt o2


If Pitot Pressure I|
Pressure in Hessu20e










_oncentration (rm)
y CMass Fow Controler -




.00033333333


r FC Feedback (cc m









rn__ .


VI
> :















Preq.'.iir~ (in H'?Fil


Airflow ((cpn)


Gas hlow ((:L:pi I
Cc'ncenltralion I pnI I



M IFC feedback I,:,C:_nil

l'r V-

f-


+ I k I ,r. F I ....


.1~T 231111I


171 -I


+I I al-I II


. I I .1


! -T,- sart PELwbVIEW IS :ont.nant conro l i nite PitIf-21 P MP~














APPENDIX B
START-UP PROCEDURES

Listed here are the startup/shutdown procedures for the Nexa Fuel Cell and the

startup/testing/shutdown procedure for the Gas Chromatograph.

Nexa Startup Procedure

1. Check whether sufficient hydrogen is present in the cylinder and replace

cylinder if necessary

2. Check the connections between the Nexa and the hydrogen cylinder

3. Check the load connections

4. Check the hydrogen pressure in the cylinder (70kPa-1720kPa)

5. Turn ON the hydrogen supply

6. Run the NexaMon software and click ON the main toggle ON/OFF button

7. Switch the 24V power supply ON

8. Check for the STANDBY status of the software

9. Enter log file name and location and start data logging

10. Turn ON the main 5V ON/OFF switch

11. Check whether the status has changed to STARTING

12. Check the status again after 30 sec. It should display RUNNING

Nexa Shutdown Procedure

1. Switch OFF the main 5V ON/OFF switch

2. Check is the status of the Nexa has changed to NORMAL SHUTDOWN

3. Check if the current level has dropped to zero









4. Switch OFF the 24V power supply

5. Switch OFF the main toggle ON/OFF button on the software

6. Turn OFF the hydrogen supply



SRI Instruments 8600 Series Gas Chromatograph

Startup Procedure

The chromatograph should be turned on 2 hours before any testing is done. This

allows the voltage baseline to stabilize.

-Set He pressure regulator to 70 psi

Turn on the computer and start the chromatograph program. (PeakSimple)

-Make sure the TCD current detector toggle switch, located on the right side of

the chromatograph, is in the off position. Turn on the main power. The

switch is located on the left side of the chromatograph.

Using the display on the front of the chromatograph, monitor the oven and

TCD temperatures. Wait until they have reached their setpoints before

continuing. This should be ~ 47C and 107C respectively. The setpoint

and actual values can be displayed by turning either of the two knobs next to

the LCD to the appropriate selection. The toggle switch below the display

makes the LCD correspond to either the upper or lower knob setting.

Also using the display, check that the Column 1 head pressure is 7 or 8 psi

and the Carrier 1 pressure is -49 psi. If either of these values are different,

there could be a blockage of the column or a carrier gas filter problem. Refer

to the manual for further help.









-Check for carrier gas flow at "TCD Ref Gas Exit" located on the left side of

the chromatograph. Simply immerse the end of the tube in water and look for

bubbles

Check for sample gas flow at the sample gas exit, located underneath the top

cover on the right hand side of the column box. Do this same manner as for

the reference exit.

-Move the TCD current toggle switch to the up position. This is labeled as

"high".

Test Procedure

This test procedure should be followed each time a sample is analyzed. The

chromatograph is very sensitive, so variations in procedure may cause bad results.

-Make sure the Attenuator Switch is set tol. the switch is located on the right

side of the chromatograph.

-Use the Zero Adjust Knob to zero the output voltage. The knob is located just

above the Attenuator Switch. The actual output voltage can be seen on the

software display and is labeled as "Stand by:". The voltage does not have to

be exactly zero to start, and it will drift during testing.

-A sampleof gas can now be taken. First the needle should be flushed of air.

This can be accomplished by withdrawing some sample gas and then ejecting

it. If this is done several times, any air that was in the needle should be

displaced.

-Withdraw a 1CC sample of gas and start a chromatograph run by selecting

RUN under the Chromatograph menu in PeakSimple.









-At 0.9 minutes begin injecting the sample into the septum. Finish injecting at

1 min. This slow injection method insures that no sample leaks back through

the septum. Do not rotate the needle while it is in the septum. This would

cause it to cut out a plug. Once the needle is withdrawn, check to make sure it

is not plugged.

-Once all the peaks are shown, stop the chromatograph and analyze the results

by selecting Results under the Analyze menu in the software. In order to see

concentration results, it may be necessary to load the calibration files.

Shutdown Procedures

-Turn off current by moving the TCD Current toggle switch to middle position

-Turn off the main power to the chromatograph

Turn off computer

Close carrier gas regulation valve.















LIST OF REFERENCES

Baschuk J.J., Li X., Carbon monoxide poisoning of proton exchange membrane fuel
cells, Int. Journal of Energy Research, 2001, 25: 695-713.

Betts, Daniel, Modeling and Analysis of Fuel Cell Engines for Transportation
Application, Masters Thesis, Univ. of Florida, Gainesville, 2000

Blomen L.J.M.J., Mugerwa M., Fuel Cell Systems, Plenum Press, New York, 1993.

Bockris, J. O'M, Fuel Cells: Their Electrochemistry, McGraw-Hill, New York, 1969.

Divisek J, Oetkin H-F, Peinecke V., Schmidt V.M., Stimming U., Components for
PEM fuel cell systems using hydrogen and CO containing fuels, Electrochimica Atca,
1998, 43(24): 3811-3815.

Environmental Protection Agency (EPA), Office of Air and Radiation, 1997 National
Air Quality: Status and Trends, Brochure, New York, December 1999.

Hoogers Gregor, Fuel Cell Technology Handbook, CRC Press, Boca Raton, 2003.

Jeng K.T., Kuo C.P., Lee S.F., Modeling the catalyst layer of a PEM fuel cell cathode
using a dimensionless approach, Journal of Power Sources, 2004, 128:145-151.

Kulikovsky A.A., The voltage-current curve of a polymer electrolyte fuel cell:
"exact" and fitting equations, Electrochemistry Communications, 2002, 4:845-852.

Larminie J., Dicks A., Fuel Cell Systems Explained, John Wiley & Sons, Ltd.,
Chichester, England, 2002.

McNicol B.D., Rand D.A.J, Williams K.R., Fuel cells for transportation purposes -
yes or no?, Journal of Power Sources, 2001, 100, 47-59

Wang L., Husar A., Zhou T., Liu H., A parametric study of PEM fuel cell
performances, Int. Journal of Hydrogen Energy, 2003, 28:1263-1272.















BIOGRAPHICAL SKETCH

Gerard Michael O'Sullivan was born August 29, 1980, in Queens, New York City.

He is the first son of Michael and Margaret O'Sullivan and older brother to Sean and

Patricia O'Sullivan. Gerard attended Archbishop Molloy High School run by the Marist

brothers and spent most summers at the Mid-Hudson Valley Camp working with

mentally handicapped and deaf children.

Gerard attended a summer research program sponsored by the National Science

Foundation at Clemson University, which encouraged him to apply to graduate school

and continue his education. After graduating from the New York Institute of Technology

in May 2003, Gerard enrolled at the University of Florida. He graduated in May 2005

with a Master of Science in mechanical engineering, specializing in thermal science and

fluid dynamics.

Gerard plans to travel for a few months before returning to New York City.




Full Text

PAGE 1

btnnfbnrffb fbtf t ffbnrn rnrfrf nfrnfnfnnb fn fbrff brfn nfrnfn

PAGE 2

!"# %&r() *&+

PAGE 3

" ,"%, ,,'%' -&#%'#.%+#,/

PAGE 4

* 01br n(' ,# 4%#"&+%$%#+!)%! +!.%"%#(+ #% #" ,%&#(+ *%, #+'%-#"&#"%)' +#%%,#n('&),) 4% #"&/ .)%#-" % +!&))"%&2+.+ *%, ##) '&# &,, ,#%"%+%*%n+%%'%'"%) +&))n&+##"&1 )) &.%&! 2&. &4 +!# .% #. ##%%/ n.%'&%-&(,%"%%&#&3 -%, %+#,&'%& ,%"% -" )'%+ #"%+ #%#&#%,"%&.%%24.%)&+'.+#",$%3%2&, +&+'"&*%)2&%)'.%."%&+'. ,#% ,#+#"% *%,.%*% &#%3(2"&#"%'&+'-+# +(% ('"&*%#(+ # %, 4%#" ,/ ) '&&,&%&#%&+ +!%8% %+-%%!%+%&)7 #&,3 ,# .%2& 3...%+'n( 4%%+'&&)"&#)) %+',+' #"% %+',n.&'%"%%& +%,* ))%+# "%&,#&"%%2%%%..%+#,"%% ) '&#"&#n("&*%&'%# #"(#) ##)%"%)..3 %+',"%$ !!%,# )%,,&*%%&+%' .%"%%,))%%',) 3%"&,"%% 3&. )+' %+',n% "%.))/ +' +&))n(,#"&('+%*%$%! +#(+##"%+(.$%$)%,, +!, n"&*%+. 3%+' #,)),+!/

PAGE 5

* tr !% 01br ///////////////////////////////////////////////////////////////////////////////////////////////// nrtr //////////////////////////////// /////////////////////////////////////////////////////////////////////////// nrnfr //////////////////////////////////////////////////////////////////////////////////////////////// /////////// 8 bf //////////////////////////////////////////////////////////////////////////////////////////////// ////////// 8 trf //////////////////////////////// ///////////////////////////////////////////////////////////////////////////////////// 8 f 9 nfn ////////////////////////////////////////////////////////////////////////////////////////////////////////9 t0f //////////////////////////////////////////////////////////////////////////////////////////////// //////////: (%)%))&, -, //////////////////////////////// //////////////////////////////////////////////////////////////// ////////////; )%-#"%. -&)%&-# + ////////////////////////////////////////////////////////////////////////////////////// &#"'%"&&-#% ,# -, /////////////////////////////////////////////////////////////////////////////////// < )&# +(.&#&)# //////////////////////////////////////////////////////////////////////////////////////////// = & >&# *%, /////////////////////////////////////////////////////////////////////////////////////////////// = !%+&+'f%3. %,,%, //////////////////////////////////////////////////////////////// /////? (%)%))+#&. +&# ///////////////////////////////////////////////////////////////////////////////////////////// ,+#&. +&+#, //////////////////////////////////////////////////////////////////////////////////////// 99 %* %,%&-+#&. +&# "%+'% //////////////////////////////// //////////// b%#"', # !&# +!#"%+#&. +&# ////////////////////////////////////////////////////////// 9: "&))%+!%,(%)%))%" -)%, ////////////////////////////////////////////////////////////////////////////// +#&. +&# //////////////////////////////// //////////////////////////////////////////////////////////////// ////// &'% //////////////////////////////////////////////////////////////////////////////////////////////// 9< #!% '% //////////////////////////////// //////////////////////////////////////////////////////////////// /9< : fnbrb /////////////////////////////////////////////////////////////////////////////////////9@ t&))&41%8&2%'(% //////////////////////////////////////////////////////////////// //////////9@ %8&($,#%., //////////////////////////////// //////////////////////////////////////////////////////////////// / !%+,#%. ////////////////////////////////////////////////////////////////////////////////////////// 8 '&+# ,#%. //////////////////////////////////////////////////////////////////////////////////////// +!,#%. /////////////////////////////////////////////////////////////////////////////////////////////

PAGE 6

* r&3%##%. //////////////////////////////////////////////////////////////////////////////////////////////// A+#&. +&+#)%&,%.%+#+'b 8 +! ////////////////////////////////////////////////////// b%&,%.%+# //////////////////////////////// /////////////////////////////////////////////////////////////////////// #($% //////////////////////////////// //////////////////////////////////////////////////////////////// ///// 33%%+# &)%,,%&+,'(-% //////////////////////////////////////////////////////////////// // b&,,)+#)% /////////////////////////////////////////////////////////////////////////////////// &,%, //////////////////////////////// //////////////////////////////////////////////////////////////// //////////// &+4 //////////////////////////////// //////////////////////////////////////////////////////////////// //// r&3%# //////////////////////////////// ////////////////////////////////////////////////////////////////////////////////////////// &,.&#& //////////////////////////////////////////////////////////////////////////////////////////////////// ; b /////////////////////////////////////////////////////////////////////////////////////////////////////:9 t&,%) +%%3.&+-%%,# + ////////////////////////////////////////////////////////////////////////////////////:9 & >&# *%%+%&# //////////////////////////////////////////////////////////////// ////////////:9 +,#&+#&%,# +! ///////////////////////////////////////////////////////////////////////////////////////// +#&. +&# %,# +! //////////////////////////////////////////////////////////////////////////////////////////////// & >&# *%%+%&# //////////////////////////////////////////////////////////////// ////////////:; +,#&+#&%,# +! ///////////////////////////////////////////////////////////////////////////////////////// +#&. +&# %-*% ///////////////////////////////////////////////////////////////////////////////////// %.&#%33%-#, ////////////////////////////////////////////////////////////////////////////////////////////////////:< frrnrrrn /////////////////////////////////////////////////////////////////////////////////:= t&,%) +%)& >&# &#& //////////////////////////////////////////////////////////////////////////////////////////:= t&,%) +%+,#&+#&#& //////////////////////////////////////////////////////////////////////////////////////:@ +#&. +&# %,(#, //////////////////////////////////////////////////////////////////////////////////////////////// &'% //////////////////////////////////////////////////////////////////////////////////////////////// +,#&+#&' //////////////////////////////////////////////////////////////////////////////////////////////// & >&# -*%, /////////////////////////////////////////////////////////////////////////////////////// ;< f&' +5%-# +,&$.'% /////////////////////////////////////////////////////////// ;= #!% '% //////////////////////////////// //////////////////////////////////////////////////////////////// / +,#&+#&' //////////////////////////////////////////////////////////////////////////////////////////////// f&' +5%-# +,+ #!%+' '% //////////////////////////////////////////////////////////// %.&#%33%-#, //////////////////////////////////////////////////////////////////////////////////////////////////// < rnfbnr /////////////////////////////////////////////////////////// n tn1bfffb ////////////////////////////////////////////////////////////////////// t rf B ffr //////////////////////////////////////////////////////////////////////////////////////<9 %8&#&#( %% //////////////////////////////////////////////////////////////////////////////////////////////<9 %8&"(#'%% /////////////////////////////////////////////////////////////////////////////////////////<9 rfn+,#(.%+#,% %,&,.&#& /////////////////////////////////////////////////////

PAGE 7

* r#&#-%% //////////////////////////////////////////////////////////////////////////////////////////////// %,#%'% //////////////////////////////// //////////////////////////////////////////////////////////////// /////<: r"(#'%%, ////////////////////////////////////////////////////////////////////////////////////////// <; nrfr /////////////////////////////////////////////////////////////////////////////////////////////////// tnfn0 /////////////////////////////////////////////////////////////////////////////////////////////<<

PAGE 8

* nrtr &$)% !% /9 -&))%*%),&.'%--' +!#*& (&) #%#,/ //////////9< -&))%*%),+ #%+' '%--' +!#& &) #%#,/ ////////////9= ;/9 +-%+#&# +#&. +&+#. 8#%!&,%, +'%,/ //////////////////////////////////////:: ;/ +#&. +&+#)&#%,%% )&#%,D:7% -+#&. +&+#!&,%,#&+'+#&. +&+#!&, 8#%,##E/ ///////////////////// :; ;/: r%)%-#%'--%+#&# %,# +!$&,%' ,, +,%#,)%-#%' ) #%&#%%* %2/ //////////////////////////////// //////////////////////////////////////////////////////////////// /////// /9 r#&-#&!%+',#&-%))&$)%&+!%,(+'%.&)&# -+' # +,/ //////////////////////////////// //////////////////////////////////////////////////////////////////////////////// %-%+#&!%"&+!%,)#&!%&2% &+5%-# +,&# -+,#&+#.,/ ///////////////////////////////////////////////////////////////////////////////////////// ;? /: %-%+#&!%"&+!%,)#&!%&2% &+5%-# +,& #,#&+# )./ //////////////////////////////// //////////////////////////////////////////////////////////////// ///////

PAGE 9

8 nrnfr % !% /9 -&))& >&# &))(,#&# %! +,3+#*& #, *%%+# &)t)%+%#/&)7??:E/ //////////////////////////////////////////////////////////////// ///////////@ %))#&!%&3#%"&+! +!#"%(%)&,#.F+'%&#&)#F #/f(/F%+'AF ,%4%#/&)7??@E/ /////////////// 9: :/9 t&))& '/1%8&(+ #+'*& +#,/ /////////////////////////////////////////////////9? :/ A+#&. +&+# 8 +!.&+ 3'7t&))&%8&3(%)%)))'$&+47 !&, +'%,%8&bb#2&%/ //////////////////////////////////////////////////////////////// / :/: n+) +%%)#&#($% #",2&!%) ## +!,++%-#%'##&# -&+''+&. %,,%%&' +!'%* -%,/ ///////////////////////////////////////////////////////////////////////////////////////// :/; +%&#(##"% B B "% + % +%, +' -&#% #"%&8 .(.%&+-%,+-)(' +!&#%.&#%*& &# B / ////////////// :/ r%+, B 33%%+# &)%,,%#&+,'(-%/ //////////////////////////////// //////////// :/< &#"..&,,+#))%-++%-#%#2&!%)($ +!/ //////////////////////////////// ////////// :/= t&-4* %2'$&+4 #",%*%&)%, ,#, &&))%)-+,#(-#%&# %+# %#%&,&-%'&#% +#%D+#-#%'E/ ////////////////////////////////////////////////// :/@ +#* %23'$&+4 #",%*%&),2 #-"%,#&-# *&#%%, ,#,+' %+# %#%) !"#&'5(,#.%+#,/ //////////////////////////////////////////////////////////////// /////// :B & &3%#%&,%,F %8# +!( ,"%7,.%A-&+ '%'%#%-# &+'!&, +'%#,/ //////////////////////////////////////////////////////////////////////////////////////// :/9 rfn+,#(.%+#,,&#&"/ ////////////////////////////////////////////////////////////// +3 '%+-%+#%*&)& >&# *%,$&,% 8#%,#,/ ////////////////////////////////:@ /: r#&-#&!%+',#&-%#+,#&+#)3./ ///////////////////////////////////:? /; r#&-#&!%+',#&-%#+,#&+#)3./ ///////////////////////////////////:? & '%%,#9.E#+,#&+#)3./ /////////////////////////////////;9

PAGE 10

8 /< & '%%,#:.E#+,#&+#)3./ ///////////////////////////////// /= & '%%,#9.E#+,#&+#)./ ///////////////////////////////;; /@ & '%%,#9.E#+,#&+#)./ /////////////////////////////// /? & >&# -*%,.&'%/ /////////////////////////////////////;< f&' +5%-# +,.E#+,#&+#)3./ /////////////;@ /99 f&' +5%-# 9.E#-+,#&+# .,/ //////////;@ f&' +5%-# 9.E#-+,#&+#.,/ ////////// 9: #!%+' 8 '%%,## +,#&+#)3./ ////////////////////////////////// /9; #!%+' 8 '%$E#-+,#&+#./ ///////////////////////////////////// #!%+' 8 '%E#+,#& +#)./ //////////////////////////////// //////////// f&' +5%-# +,+ #!%+' '%D.E# -+,#&+#.,/ ///////////////////////////////////////////////////////////////////////////////////////// 9= %.&#%33%-#,(%)%))3.&+-%& >&# *%,2% -*%,"+%.&#%,D7E/ //////////////////////////////// //
PAGE 11

8 bf H 33%%+# &)%,,% ( 3)33 %+# n +!%+,") '% &'% %))#&!%"%-4% &.%#%..J 33%%+# &)%,,% +* +.%+#&)#%-# !%+: (%)%))+#&. +&# +# (%)%))%" -)% r () B r-&)% &,.&#& !%+ !%+ '% n n+#%+&).$(,# +! +%

PAGE 12

8 %)&..&$ ) # # . )) B &. b b%.&+%)%-#'%,,%.$) #!% '% 8!%+ &))&' (. b 8-"&+!%%.&+% t &#, )) b &#, )) # )&# +(. f" f"' (. rb r#&+'& #%,b +(#% rr( 8 '% "%.&)+'(-# #%#%-# #&!% %-#%+#E 1,1% !"#%+, # 4!A.:J

PAGE 13

8 $,#&-#"%, ,%,%+#%'##"%&'(&#%"%+ *%, #) '& &# &)(3 )).%+#"% f%C( %.%+#,"%%%%&,#%%+-% btnnfbnrffb fb tf t %& -"&%)r() *&+ b& "& K %$%#+!)%7 nnn b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

PAGE 14

8 -%))%*%&)%,#,%%'%, !+%#, .(&#%"%#,+* +.%+#, +" -"&(%) -%))%%'*%" -)%(&#%"%3.&+-%%-*% %,%% -&)-()&#%'."%)%-#%''&#&"%33%-# #%.&#%&,#(' %'/ "%%,(#,"%', !+ 3 -&+#,%,3.&+-% #%*%),$*% ./ &+, %+#%,+,%%,#,'%.+,#&#%#"%(%)%)),$ ) ##%-*% -%#%),#% %3+5%-# +,3*%!"-+-%+#&# ,&'%"% -& '%%-"&+ ,.3,+ +!,%%.,(+-# "%))#%+# &) & +-%+#&# '*%,%%33%-#,%%' ,-*%%'-+-%+#&# +, .#"% .%+#%*&)#%,#%'/

PAGE 15

9 f9 nfn "%%8-"&+!%%.&+%bE(%)%)),,%&-%.%+#, -+*%+# &)-%,%"&,%-% *%'! +!&##%+# %-%+#'%*%)%+#, 3(%)%))#%-"+)*%&'%b(%)%)),&##%+'%#%)&%#"%+#%+&) -$(,# +! +%nE+*%" -)%,"%,%(%)%))%" -)%,E&%'(-%"&.3( %. ,, +,+''%-%&,%'%+'%+-% )"%#&+, # 3(%)%)) %%'*%" -)%, ))#&4%*%)+!# .%'(%(%) +! +3&,#(#%&+'-+,(.% &'&&# 3+%2%-"+)"%,% ,#, ))$%%%-#%#&#%&)+!, '% #&' # &)+! +%,"%"&.3(). ,, +,"&#"%,%*%" -)%,'% #+#"%& .&%*%%)33%-#"%3.&+-%"%(%)%))*%" -)%, "%&# +(.&#&)#+#"%)%-#'%,b(%)%))%, $)%%&4 +! '+#"%!%+&+'+ ,*%(,-%# $)%+#&. +&# *%+,.&)) -+-%+#&# +,# -(&-+#&. +&+#&+"&*%+&'*%,%33%-#+#"%&)) 3 .&+-%b(%)%))(-"%%,%&-##%"&,'%&)# #"#" ,&'*%,% %33%-#"%+'%3(%), '%"%(%)%))E%-&(,%"%+#&. +&+#," ,# + %3.%'(%),% ##)% +3.&# ,*& )&$)%$(##"%-+#&. +&+#,3 3%-#"% -&#"'%D& '%"%(%)%))E33#,(,#%+%#''%,,"% %+* +.%+#,"%,%b(%)%))%" -)%,&&#%+7, +-%# 4%)"&#(%)%)) *%" -)%, ))&2"% '&+# #+%%',&#%3."%(+' +!&.$ %+# "%(," ,%,%&-" ,##"%33%-#,3%)%-#%-+#&. +&+#,"% -&#"'%+''%#%. +%,"%+-%+#&# +,#" -"#"%,%'%# .%+#&)33%-#,$%! #

PAGE 16

#&4%&-%& % .%+#,%%'%, !+%'# .(&#%"%+' # ,.& %+-(+#%&+, %+#%,%%,#,2%%%'#,,%,,#"%(%)%)),$ ) ##%-% &3#%%*%%-+#&. +&# "%$ ) #(%)-%))#%-% #, +&) 3.&+-%,.#&+#(+'%,#&+' +!#"%%-"&+ ,.#"% +#&. +&# "% %8% .%+#&)'&#&&,#"%+&+&)%##%. +%# -(&-+#&. +&+#,33%-#"% 3(%)%)),%3.&+-%+'#%*%%-%3#%-#&. +&# &'-%&,%

PAGE 17

: f t0f 8-"&+!%%.&+%(%)%)),"&*%$%-.%"%.# +!&)#%+&# *% %!%+%&# +!'%* -%3"%*%" -)%,"%(#%"%+ (%)+&+ +#%+&) -$(,# +! +%, +%33 %+#%,,8#&-# +!%+%! .(%)"% -+*%, (%)+#)%-# #+(%)%)),#%%)%-#"%. -&)-%,, r +-%)%-# # ,!%+%&#%'' %-#)+''%,#+*)*%++#%.%' &#%%-"&+ -&) "%.&)%,,7(%)-%)),%%33 %+##"&+& "%%-"+)! %, 1 #"(%)%))%-"+)# )) #, +3&+-"%%,(-%,%&-"$% +!'+% *& &.%#%,"!%-#(%)%))3.&+-%+%# -(&+-%+ ,&#&)# -+#&. +&# "%&#&)#, +&8-"&+!%%.& +%bE(%)%))3. #"%#.#&+#(+-# +F#"%%&+&!%+.%-(%,"%,% -&#&)#,&%*%%+, # *%+#&. +&# +''%&'% +"%%,%+-%+#&. +&+#, "%+#&. +&+#,',$"%(3&-%"%&#&) ,#,+'%*%+##"% >%+' 3(%).%&4 +!'+&+'"%+-% +" $ #"%(%)%)),$ ) ##(+-# .%"% -+#&. +&+#,"&#&33%-#b(%)%)),%&.'%7+ #%'%,7,(3 '%,7$%+>%+%+%+'*& "%"% -&),#"&#&+$%+#"%.$ %+# & f%,%&-"-+'(-#%(%)%))#&. +&+#,"&,(,% .& ) -+#&. +&+#,+#% +#"%+'%, '%"%(%)%))" ,,'(%#"%&-#"&#& 3(%)%)),%%3.%+*%#(%),("&,+&#&)!&,.%#"&"%!&+ 3(%),+#!%" ,%3. +!%,,(-%," !"-+-%+#&# ,&

PAGE 18

; .'%"%"%. -&),"&#&+"&*%'%# .%+#&)33%-#"%(%)%)) "%%3%(-"%%,%&-""&,(,% .& )&) +! #"#" ,'*%,%%33%-#/ "%%"&,$%%+*% ##)%+3.&# ,"%&$(#"%33%-#,+#&. +&+#, "%&#'% # -&)#"%3.&+-%"%(%)%)) +#2& ,#7#"% & '%,#"%+%-%,,& %#%"%)%-#"%. -&)%,,+*%# +! #"%!%+ +#)%-# #+'&#%%-#"% & %,&#%7& B '(-#"% 3(%)%))%&-# #"%(%)%))/#"%2 ,%#"%&#%(')"%(%)%))&+' %*%+##"%%)% -#"%. -&)%,,/ btnt (%)%)),%)%-#"%. -&)+%!+*%, +'%* -% +!)%(%)%)) -+, ,#,2 )%-#%,"%(%))%-#'%+'%' >%,"%!%+(%)+ #"% )%-#'%&#'%%'(-%,"% (%)%)),%"&&-#% >%"% )%-#)%%'$%#2%%+#"%,%2)%-#'%,/.%8&.%,"%,(%) -%)),%).%)%-#%%.$&+%bE7", -%-#%#"&) )4&) +%b#%&+&#%7&+'r) 8 '%/ b 3(%)%)),%& %)%-#)%)&-%'$%#2%%+2 %)%-#'%,"%&#&)# ,'% #%+#"%,%)%-#%,+',-&))&# +(. b(%)%)),%"%# +%-%(%)-%))%%'*%" -)%,"% 3% %%'%+, #&+'" !"%+%-+*%, %33 %+%,2" )%&# +! &#%)&# *%))%.&#%,b-%)),%&), !"#2% !"#+'--#&4 +!#"%. ,( #&$)%&(# *% -&# ,/

PAGE 19

frnbrn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

PAGE 20

< brrntrnt n#,#%+(+'%,##"& #"% '%&).&8 .(.33 %+-+%)%-#"%. -&) %+%+*%#%'%+',"%)%-#-"%. -&)"%.+&. -,"%%&,"%%&) %33 %+-%+',"%)%-#'% +%# -,t ,7?
PAGE 21

= "%+-%+#&# %+# &) ,&(,%,#&+,#) #&# "% !&,' 33(, &"%&#"'%##"%-# *%&#&)# #%, "%,%%+# &),&%%'. +&#%)#"%)+%,,"% %'( -# %&-# +/ "%%'(-# + ,+%+(.$%, $)%%&-# +,#"% %)%-#%'(-# ,"%.###&-# *%%&-# #&#&)%(%"% !%+# &) *%,"%" !"%,#, $)%%))#&!%"%(%)%))%)2, "%)%-#!% %'(-# %&-# +/ L;%B E B; brnbrr % # +!&# +(."%3&-%"%)%-#'%+-%&,%,"%&#%#"% %)%-#"%. -&)%&-# )#&# +(."&,$%%*%+ ,"%#33%-# *% -&#&)#%'(-# #,*%%+, # *%+#&. +&# .&))+-%+#&# +, DNE3&+ '%&(,%'%&'&# + +#"%(%)-%)),3.&+-%" &'*%,%%33%-#,$&$)"%&', 3"% .)%-(%,#-# *% -&#&)# #%,"%8-)(, +.%-(%,)&# +(.,),&+%8+, *%+' ,-&-%%#&)+'#"%%3%"%%,(-"%33#+ +' +!&&%'(-%"%.(+# &# +(.'% #%#"%)%-#'%7%"%%#& ),"&#.&))' #"#"% &# +(.%'7f & nbrnt "%3.&+-%(%)%)),#%&" -&))%%,%+#%'(, +!&& >&# -*%#,&*%%)&# +!#"%(%+#'%+, #+'#"%%)))#&!%"%, +!& +!)% 3(%)%))"%+&(%)%))#&-4 ,%'#"%& >&# -*%%)&#%,#"%#&-4-(%+## #"%#&-#&!%"%&8 .(."%.. -%*%, $)%%))#&!% ,"% -*%!& ,","%* #&)%#"&#%)%-#'% +%# -,& +#"%3.&+-%"%

PAGE 22

@ 3(%)%))2)%-#&#&) --# #,"%%&#%,#33%-#"% +33 %+,#&#%'$%3%"%!%%'(-# %&-# +"%&#"'%%)%-#'% ,.(-",)% #"&+#"%&'%%&-# #%+#'%+, # %," %+# %),,+#"%(%)%))#%+# &) 3."%%*%, $)%*&)(%,'(%&-# *&# %+# &)#"%+%)%-#% Dt)%+O(!%2&E-&)& >&# *%,"+ % B 9/ % B 9/-&))& >&# &))(,#&# + %! +,3+#*& #, *%%+# &) t)%+%#/&)7??:E/ ,*%-&+$%!%+%&#%'(%)%))+'2 )))2&"&*%"%&.%!%+%&) -"&&-#% ,# -,'(%#"%%+# &),%,%*%"%&.%#%,"&#-&+&33%-# #" %3.&+-% ))"&+!%"%)"% +%&# 3"%*%"%%%& &.%#%,"&#+3)(%+-%"%3.&+-%b(%)%))+'#"%+-)('%&# +! #%.&#%7%,,%+'"(. 3 -&# "%!%+ &+' #%&.,

PAGE 23

? "%,)"%& >&# (*% ))%-%&,%&,"%&# %.&#% +-%&,%," ,,'(%##"%+-%&,% +#"%8-"&+!%%+#'%+, #&+' -+'(-# #%*%. 3 -&# "%%.&+%$%-.%,%.#&# +! #%.&#%, *%"%) +#&#%/ "%3.&+-%"%(%)%))+-%&,%, ##"%&# +!%,,%"%(%) -%))" ,&+$%%% # *%" 3#"%& >&# *% # +-%&, +! %,,%,"%&# %,,%+-%&,%7#"% # &)%,,%,"%%&-#&+#!&,%, +-%&,%,)%&' +!###%3.&+-% 1"%+#"%!&,#%&.. 3 -&# %.&#%,%#"&+#"%&# +! #%.&#%7"%(%)%)),3.&+-% ))%-%&,%&+'-&+$%"+ #"&+%!&# *% ," 3# "%& >&# *%,&.%# -#(b(%)%))%3.&+-%+' #"%%,+' & >&# -*%,&+$%+&+!/7%#/&)/ ))#&!%,%,&+$%' '% +#2/"% ,#+*)*%,&) ),%,'(%#"%&+, #+' +%# -%,,%,++'&-,#"%%.&+%)%-#'% &,,%.$)bE"%%-+*)*%,)$&)),%,-&(,%"%)+! B #"% B -"&++%)(+ 3. #%%'!&,%,+'&#%+-%+#&# D0( 4*,4 nt tt !%+(%),#"%,,%+# &)#(%)%)),#%.$(#),"%% 33 -(#$)%.,-%"%+&4 +!&$"%, $ ) #(%)%))%% *%" -)%,!%+ ,#"%#$(+'&+#)%.%+#"%)&+%#%*%"%-% ,, %8#&-# +!#" ,%!%+.&$+,&#% ,&" !"%+%!%,, r.%8&.%,!%+ ,%8#&-#%'.&$+,7-&))%%3. +!7&% ,#%&.%3. +!7&# &)'&# +'-)!&, 3 -&# )%-#) , %,, ,%&# +!&#% +#%!%&+!&,%," -2('$%'%&)$%-&(,%

PAGE 24

#"%&-.3(+#&. +&+#,%*%7%)%-#, ,*%+%+#%+, *% %,,/ )#%!%3(%),"%)# .&#%(%)7 #, 4%)"&# "% ,# -.%&) >%'(%)%))%%'*%" -)%,E ))&*%%&(%) -&$)%(%),%'+#"% +#%+&)$(,# +! +%nE" ,,$%-&(,%#"%% 2 ))+%%#$%&+, # .'&,(%) +! +3&,#(-#% #2 '%,%& +3&,#(-#(%"%'(-# +',%!%3(%)"% ,#(%)%)) *%" -)%, ))$%8%-#%'3.&)+!, '%+! +%,+,. ## +! -+#&. +&+#,.%3. %,,%,"(,"%%. #&. +&# ,,& .#&+#%.''%,,,#"%!%+(%))%./ rnbrn ,&!%-&)%.%&) >&# b(%)%)),&2,),%7%33#,.(,# #,,(%,%)&# +!#"%%+* +.%+#+" -"#"%,%(%)%)), ))&#%/ (%)%)) '%*%)%+##%"&,#&4%&-%+&+#))%'%+* +.%+#&&#%% 3.%&) B 2)'-#&. +&+#,+)%,,"%'&+#,) %'#"%%)),. -+#& +%',-%%/!/7$#)%'& E7 # %,%,%+# +#"% ..%' &#%# ."%%& &'*%,%)33%-#"% 3.&+-% "%%"&,$%%+.(-%,%&-"'+%##%"&#%&), #"#"%+%!&# *%33%-# -+#&. +&+#,"&*%(%)%))3.&+-%%*%)&!%.&5 #"&#%,%&2&,''%,,%'#+#&. +&+#33%-#, "%&'%"%(%)%))%-&(,%##"% !%+(%)%#2%(%)%)),,%3.%'.&#"%%% # %,&# -(&)7#"&#+%%#%.*%'."%%3.&#%$%3%#+#%, #"%(%)%))/

PAGE 25

99 b(%)%)),(%+ #)(,%%%#&)-&#&)#,#%!",3&-%%& -&$+,,&-# *%-&#"'%&#&)#)&,!%%'(-# +D&. + % -4, ,"%%'(-# %,,-,* &,3&-%'%# %,,7&&!% -&#&)#(3&-%%& +-%&,%,"%.$%%&-# #%,#"%%+-%&, +!-&#&)# (# ) >&# # %,"&#&$%&2+ 3."%+* +.%+#"&#& #"%,%-# *%-&#&)# #%, ))'%-%&,%"%&#% '&# +'#"(,'%-%&,%&)) 3.&+-%/ n#,#"%%3% # -&)#(+'%,#&+'2" -"-+#&. +&+#,33%-#"%(%)-%)), 3.&+-%+'&#"&#-+-%+#&# +,"%,%33%-#,$%-%%.&# -/#, .#&+#),%*%3"%'&.&!%,%*%, $)%.&+%+#+-%+#&# + % % .%%2.#&+#&-#,#"%, $)%%&'&# "%&#&)#, +-%"%+#&. +&+#,33%-#,+'#"%%#"',+#&. +&# %(+'%,#"%+#"% +%-%,,&#%&+$%#&4%+# # !&#%"%'*%,%33%-# trnrt "%% %&+#&. +&+#," ,##"%.$ %+# #"&#&+"&*% '%# .%+#&)33%-#"%&#&)#"%(%)%))"%*%"%+-%+#&# +,#" -" #"%%"&>&'(,%+.%$&%+* +.%+#,.%"%,%% %+ >%+%7&+'(#&' %+%"%,%+#&. +&+#,%"%)(#&+#, #"&#%. ##%' +&(#$ )%8"&(,#&'%,&!&,%'(-# +-%#%$(,# + +#%+&).$(,# +! +%, #!%'%,&%%)%&,% +# #"%+* .%+#,%,(#, )(%)$(,# # -'%,-..+&(# %8"&(,##&')' >%,#.+ #%+' 8 '%(3' '% ,!%+%&)). ##% 3.' %,%)+ +!%+! +%,&# )%!&+ -.(-"&,$%+>%+% $(#&' % +%%"!&, +%&+'' %,%)+! +%8"&(,#

PAGE 26

1"%+-+, '% +!#"%,%(%)%)),&$&##)%3 %)'%+* +.%+#"%%&%%*%&) -"%. -&)&3&%!%+#,#"&#)'&33%-#3.&+-%(-"&,7,& +7,()(,#&'7 -+!%+,") '%nE!%+-+ '%E"%,%# -()&&3&% &!%+#,"&*%$%%+++#% )-. ,%"%3.&+-%"%(%)%)) ttrnbrn b(-"%%,%&-" +** +!-#&. +&+#,b(%)%)),"&,% %3(,% "%&.'%%."%+(, %3.%&(%),% !%+&,&(%)-%"&,& #&# +,%+&(#.$ )%," ,(%)(,#$% ,#%'% #"%+-%,,%'!&,.7+ -&))#%' ) C( '.' ,$% +#%#&) '%&-"%,%,,#&!%%-"&+ ,.,"&*%&5' ,&'*&+#&!%, #"%%3%#"% #%&#%(!!%,#,!%+ ))$%, %%3. +!&& 3(%)(-"&,.%#"&"%%3.&# .%#"&%,()#, +&!&, 8#% $(#=; -%+#!%-%+#&' '%+'9 B -%+# ,%4%#)/7??@E & '%%*%),-&+$%%'(-%'#,), +!,%&#%!&,%&#.%+# %,,%, #b3(%)%)),%&# +(.&#&)#$%-&(,%#, %33%-# *%+%,, !% '&# #)2&# +!#%.&#%,(#*%&*%.&))+-%+#&# N.E+#"%!%+(%)#%&.&+,($,#&+# &))%'(-%"%3.&+-% #"%(%)%))/"%. ,$,"%&# +(. #%,"%8-)( !%" ,, $)%%-&(,%"%,%#+!)'&# +(.#"&%+7&,+' -&#% %&#%%+# &)%C( %'"%'&# "&!%+7&+'#"%# -4 +! $&$ ) #&# +(., .%,#!%" &#"&#!%&# +(. D ,%4%#)/7??@E*%+*%,.&))+-%+#&# +#"%(%)&%,()#

PAGE 27

9: -%#%*%&!%"%&#&)#,&-# *% #%,'&# &+'#"%%3%"% 3.&+-% ))&,# -&))-%&,% #+' (-#% ,%%#)3 !+ 3 -&+#3.&+-%"% &,%,%+##"%!%+(%),#%&."%"(%)%))%3.&+-% '%+',#+!)+-%+#&# +&+'#"%&#&)#%"%),"&##"% %-*%(%) -%))+ # &)-%))#&!% ,#%&&-#2,%'# ,+ +!# .%"%&" %-*%,"+$%)/ % B %))#&!%&3#%"&+! +!#"%(%)&,#.F+'%&#&)#F #/f(/F%+'A F ,%4%#/&)7??@E/ rtnrnnrrnbrn "%%% *%%#"',*& )&$)%#. # !&#%"%%33%-#+#&. +&+#,,+ +! b(%)%))-&#&)#,"%%#"%%&# +(.))&#&)##"%+5%-# %+ +#"%(%)#%&.7#"%+5%-# !% '%E7 +-%&, +!#"%3(%) -%))%&# +!#%.&#%7&+'%33%-# *%&#%&+&!%.%+#

PAGE 28

9; )&# +(.,%'$%-&(,%#3%," !" '&# -# ##"%&# +! #%.&#%,,, &#%2 # b(%)%*%7#"%&%(,-% $)%#,+ +! #"%%,%+-%*%+-%+#&# +,+ #)&# +(.))(-"&,# B f # B f"7&# B &+ +-%&,%#"%)%&+-%#+#&. +&+#," ),#%%+!&" !" %8-"&+!%(%+#%+, # ,,$%-&(,%"%%'&# %+# &)"%)) .%#&).&%'##/ "%+5%-# + +#"%(%),#%&.,+%+#2&"% ,#%# %-#) +5%-#,$%#2%% B -%+#+ +##"%+'%7" -"-& +-%&,%#"% #% &+-%."%%-,#"%+5%-# 3+#"%+'% "(. 3 %"%'%-%,!%+&+&+'#"%%+&-#,# >%#"%+' .*%"%#%&+-%"%-&#&)#Dt&,-"(4O %*%" ,%# # !&# ,%3("%&#'% !"%%&# +!#%.&#(%"&,%%+,"+#+-%&,%)%&+-% -+#&. +&+#,%*%" !"%#%.%&#(%, b(%)%)),,%&, $)%$%-&(,%#"% .%.&+%(,#$%"(. 3 %'&+'&" !"%#%.&#% ('&--%)%&#%"%*&&# "% C( '&#%+'"%+-%'%&#%"%%.&+%/ #+%," #2%%+&)',+'#"%)&.&# +&)&&# -&))%'#"%(%)%))#&. +&# +#:E"&,$%%%,#&$) ,"%##"%%33%-# .$ %+#+#&. +&+#,"%3.&+-% 3%&$ ) #b(%)%)), +&)',,% #, )#%8%# ,%*%) )#&# #%.,"&#& (%)%))+! +%, 3%"%%,%&-"$% +!'+%" ,,$%% ,-)%'&+'&()%#,%8%-#%' +#"% &) >%' )#% %8%-#%#$%"%)(# #&. +&# %./

PAGE 29

tnt "% +!)%.#.#&+#"&))%+!%"%--%&+-%(%)%))*%" -)%, E &,%)&-%.%+#+#%+&)$(,# +! +%,DnE,#"%+%%'"% #%-"+)#.&#-8-%%'#"%%8 ,# +!n%-"+)" ,,# .% -"&))%+!%$%-&(,%"%.& #&# +,+%2%-"+) 4%%,%+#,% #",% #&# +,%(%)+3&,#(-#(% B $&'#&!%(%) B $&'(%) %3. +!7,#&# B .%7 *&$ ) #*%" -)%,&+, %+#%,,%"&+!%+'%.&+'E -&#&)##&. +&# +'.#.#&+#7-# "% .%,#+'$%,#(%)%))# %.,3% +!*%" -)%,%,%+" #"%(%),-+*%#%'' %-#) +#"%(%)%)) #""%+%%'% B %,, "%%%+#%+'%,7+&.%)"%' %-#!%b7#"%' %-#!% &)4&) +%(%)%))#"%' %-#%#"& b7&+'#"%' %-#.%#"&!%+(%) -%)) #"& C( '&%)%-#)%Db-#&)/7%*%#,) 4%)"&#"% 3 ,#.%&) >%, ))%& C( '(%)#"&# ,&# $)% ##"%8 ,# +! 3(%) +! +3&,#(-#%#" ,#,&+,%C(%+#)#"%,% ,#,2 ))), '(+!-+#&. +&+#,+#"% &%"%&.%%. #"& +#&. +&# +&, #"%+! +%, n rnbrn (%)%))2%%'*%" -)%, ))+%%'&2+&.$ %+# ."% ( (+' +!, ##"%'&+#%%'%# +!#"%'&+##"%*%" -)%,+#+%8%-#% ,(# %-&(,%"%'' # &)% !"#%+,#&!%+ #(%,%+#"% *& #&. +&+#," ,#+&.$ %+# "&*%$%%+,"+#*% &'*%,%%33%-# "%(%)%))# -(&)))(#%'&%&,(-"&,$&'2&7" !"#&33 -)(.% &%&, +'(,# &)%&,+'%*%+& #+* +.%+#,

PAGE 30

9< n+#"%+ #%#&#%,#"%%8"&(,#.(#$ )%,+'#(-4,--(+# "%%)%&,% +##"% n+.&5$&+&%&,7.*%" -)%,%%, $)% ?P%. ,, +,' ,,%,C( -4) & 7,'%&#%&+'" !%*%),"% !&,%(&))#%-#% +&%&, #" !+ 3 -&+#.*%" -)%&3 3 -2 #" +%+-),% ,-%,"%%.&--(.(&#%7??=E%*%&)& C(&) #%#,"% + #%#&#%,+'#"%+ #%' +!'.%%%* %2%%*%),%%.(-% &)&%&,+'++ B (,&33 +#"%&+!% B .%* %),.&5$& &%&,(-"&,+!%)%,%2 #"%%)(#&+#,('&--(.(&#%+#"% Q*&))%R$%#2%%+$( )' +!,"%%(,&33 -('%. #8"&(,#,( $%," !"&,+%8#%.% #(&# ,/ &$)% B -&))%*%),&'%--' +!#& C(&) #%#,/ #t b"n$ nr . . %nrn n+#"%+ #%#&#%,&(#8"&(,#--(+#,$ B ;P+ #!%' '% %. ,, +, #!%+' 8 '%,(-%'&# #."% #%+&+'+ # '%)#"(!"+ #%+' '%. ,, +,"&*%$%%'%-) +%'(%#"% +#-# &#&) --+*%#%, +,#&))%' +*%" -)%,## ))%,%+#,-+#&. +&# %. ,%)2%%"%%*%),3+' +*& #%&#%

PAGE 31

9= &$)% B -&))%*%),+ #!%' '%--' +!#& C(&) #%#,/ %#'t b"n$ nr "%,%2 +#&. +&+#,%%",%+$&,%%* %2+#&. +&+#,%,# 3(%)%)),#,+#"&#%+# &)+%!&# *%33%-#"%,%#2+#&. +&+#,&*% (%)%)),%*%, $)%+%*%&)#"%-+#&. +&+#,(-"&,$%+>%+% 97&' %+%7&+',(3' '%%%+, '%%'$(#%%3*%%*%, $)% '&.&!%"%(%)%))+#"%+#%%,#.&+%+#)&' +!#"%(%)%)) -"&,%'" ,,#%%,%+&,"%#2+#&. +&+#,%## & #&. +&# 33%-#,/

PAGE 32

9@ f: fnbrb 18-"&+!%%.&+%(%)%))&+(3&-#%&))&% r#%.,&,-"&,%3" ,8% .%+#"% +)%#"%(%)%))2&,.' 3 %', -+#&. +&+#!&,%,('$%+,%#% +#"% #%&.# %, %'-+-%+#&# +, +! 3).%&, +!'%* -%,&+'.&,,)2+#)%,& %2&.2&, ##%# -+#)"%)&#%+#&. +&+#!&,%,$&,%"%%C( %'& )&#%"%(%) -%)) +&))"%+#&. +&#%'& 8#%,%%&+& )%, +!&!&,.&#&# -+3 ."&#"%. 8 +'-+#&. +&+#,&,$% +!&-" %*%'/ *+% t&))&%#%.,%%#)%, $)%"%%+%2%' +#%%,#+(%)%)), 3'#&+,#&# +!#"%??,+'$%!&+'%*%).%+#b(%)%)),/"% %8&(%)%)),%"% ,#.%&))*& )&$)%(%)-%)),&-# -&)+' %'(-&# +&)%,"%%8&(+ #,.&)).& +#%+&+-%7&+'()(#&#% 3(%)%)),#%.'%, !+% '#+#%&#%' +#'(-#,#&$)%&+'$&-4 B % & -&# ,7,%)),&+%'(-&# +&) *%, # %,+'+ %,/ #()#"%%8&'(-%,&##,%!(&#%%#. +&) #&!%#,7&#')%"%# #&!% ,#,"%#%.&# ,() &(#&#%'#(!"*& +#)#%.,+#&'##&-"%'##"%, '%3 #"%%8&%-% *%,*& !+&),%*%&)%+,, #" +"%(%)%))#&..(+ -&# +#+' "%+#&'%"&r B % &) +4# -%"%%8&bb#2&%)),"%%#%&'&+'),%*%&)'&#&

PAGE 33

9? +#,(-&,#&-#%.&#(%7,#&-#&!%7,#&-4-(%+#7!%%,,% !%+-+-%+#&# !%++,(. +-+-%+#&# %.&#% &!%%))#&!% %))(,#&# +!#"%%8&(+ #,,"+ % B 9/ % B t&))&'/1%8&(+ #+'*& +#,/9 9(#%3t&))&'%8&2%b'()%%&# +!b&+(&)/

PAGE 34

%"b-!ttrt r "%%8&(+ #&#%, %!%+.+#& +%',-%"%!% ,#%..+ #,+'%!(&#%,"%(!%+#"&%,,(%#&+,'(-% %,,%%) %3*&)*%7,%+ '*&)*%7%,,%%!(&+'&!%%&4'%#%-# ))"%,%.+#,+,% #"%% ,&+&'%C(&#%3(%)2" )%),& +#& + +!,&3% &# +!-+' # +, % #!%+&+'&#% &#%"%&#"'%#"% &+'%!"#"%%.&+%+'&--(.(&#%+#"%+'%" ,,+ #% -"%-4 +!#"%%)))#&!%+ %2%%)),+-%"%*%&!%#&!%"%,%%2%)), %#& %*%)"%!%'/"%!%'!%+ ,' ,-"&!%' +#"% +!& #%."%%#, )(#%#%*%)&$%)"%%)&..&$ ) # # !%%&4'% #%-#+,%,"&#!%%*%),%))$%)"% # -&) #, 2 #"&!%+,#%.&3%#,3$,(#%.#&+-%/ #"nrr "%'&+#%C( %'"%(%)%))%&-# ,* '%'..$ %+# &2+ #+& %,,+#"%(%)%))#&-+ +#&4% )#%%.*%,# -(&#%, 3+& $(#,&$)%%-##"%(%)%)),#&-4+#&. +&+#!&,%,(-"&, "%+#"%(%)%))+,"%A%&-# + +-%&,%) 2 ))+%&,%"%%+#&+'#"(,%C( %,%'&+## +#-%,"% +-%&,%,#"%%,,%' ,'5(,#%&'&--' +!) "%'&+#,#%.),. 3 %,"%++!& +!#"%&#%(-% #"%(%)%))%&-# "% B '(-#&#%-%' ,#""(. # %8-"&+!%"%%"%+-. +!& %-% *%,#"#"%&#%+'"%&#"%%&-# ,,+%-%,,&%%"%%.&+%,,#%8-"&+!%/

PAGE 35

nr "%(%)%))#&-4 -%'(, +! +!&+#"&#), "%%+# %#&-4/ ,.%+# +%'&$*%!%!%+ ,' ,-"&!%' +##"% +!,#%."%%#, )(#%'&%)"%2%)&..&$ ) # #E" ,&3%#+##%. &(#&# -&))"(#,+#"% (%)%))3"%-+-%+#&# %&-"%,P"%/ -rr "% %8&"&,%*%&)&3%##%.,"&#%*%+#C( %+#'&.&!%+'&))&3% &# +,#&!%,7-%+#7#%.&#%,#-&%))+ #%'#+,%#"% ,#& # +&3 %&+!%!%%&4'%#%-#%*%+#,"%!% -+-%+#&# +.%&-" +!#"%/+-+-%+#&# %+,+,%,"% &'%C(&#%&.(+#,$% +!,) %%))#&!%-"%-4%E+ #,"% #&!%%#& +%)),F&)#&!%-%)),2 ))"(#2+#"%(%)%)) '%# %-#"%,#&-"%)%+ '*&)*%%*%+#,!%+%&* +!#"%-+#& +% ,-%"%+"%(%)%)),%&# +' +&))"%%,,%%) %3*&)*%+,(%, &!%"%%)),#"%,, > +!#"%(%)%))#&-))#"%,% -+#,$ +%* '%+ # +!7&+ +!&+'&)&.,7&+',"(# .%-"&+ ,.,+,%,&3%&# &#)) .%,/ n.rnr,btrn"n 8 +!. &+ 3'&,+,#(-#%#$)%+5%-#+#&. +&+#,+#"% ++!& #%&."%(%)%))!%#"%%, %'-+-%+#&# +#&. +&+#+! +#"%(%)-%))2&,+,#(-#%'&#"% +)%#&+'-%#%)%&)% +%+-" &,&' +##"%+# '%"%"%+* '%,)&-%#%&,%"%& )&#% $% &2+ +#"%(%)%))#"% & %,,,%))&,&-%#+5%-##"% &#%.(+#+#&. +&+#!&,%,

PAGE 36

+#"% #%&., +!-%,,%'+'%, &,+, '%%'%--&#% #%,# +!$(#(%"% !(.%& #"%%8&%C( %,"%.(+# ) +'%, 2&, .+#%&, $)%/%*%&)# +,%%*& )&$)%%&, +!#"%& ) %+#% +!#"%3(%)%))##($%A%,,%#&+,'(-%#%.&,",%+&,# %'(-&# +&)&+'%-&)&#--&#%)%&,%#"% )"%+# % %8% .%+#&)%# B &$% % B :/ % B A+#&. +&+# 8 +!. &+ 3'7t&))&%8&3(%)%)))'$&+47!&, +'%,%8&bb#2&%/

PAGE 37

btr "%) +!,%-# %,$%,))"%+#,%)%-#%3"%%&,%.%+# #"% )+''%) *%"%+#&. +&+#!&,%,+'"%%%,%+&, "%%,# "%! *%+&) -&# +/ nr .'%)% %,+) +%%)#&($%&,-"&,%' B 1%,#+,#(.%+#, n# ) >%,#2*%&! +!3))%.%+#,C(&)%&%+,%,#&!+&# +'#"%#&# 33%%+# &)%,,%* +!. + .(..&+%+#%,,%,"%3))%.%+#,% &-%'&#"%%+#%%&'%'@R,%-# %+'-&+$%##&-"%'##"%+-" #"%+'/"%+-&,",%#*%&" !"%+(!" 3)*%) #"%# #($%##%) &$)%%&' +!/#"%&8 .(. ) &#%%.&+'%"%(%)%))bE"%#($% )).%&,%' 33%%+# &) %,,%E"%%3%"%%)%-# ( #&$)%%,,%&+,'(-% #%&#" ,%8#%.% ))(%&, # -&)/ -#(%"%*%&! ##($% &+',2&!%)4 ## +!, ,,"+ + % B :/ % :B:n+) +%%)#&#($% #",2&!%) 4 ## +!,++%-#%'##&# -&+''+&. %,,%%&' +!'%* -%, /

PAGE 38

&nrnbtt t %+, r&,#%'$%+, #"%+ *%, #) '&F#, )&+!%' 33%%+# &)%,,%#&+,'(-% #&.%&,%.%+#&+!%. B ,%,,(%&+,'(-%2&,$%,#( #%'" -&# %-&(,%#, %,)(# &#"%)#,%&,%.%+#&+!%#%,,"&()-&)%DrE"%%,(# "&,&+!% B "%&+,'(-%%C( %,%+' "&, +%& B 7&,"+ + % B % B +%&#(##"% B B "% +% +%, +' -&#% #"%&8 .(.%&+-%,+-)(' +!&#%.&#%*& &# B / (#%3r%+, +r 9&#&,"%%#/

PAGE 39

% B r%+, B 33%%+# &)%,,%#&+,'(-%/ bttr n'%#+##"%)&#%+#&. +&+#!&,%,"% +'%+#"% ,#%&.7&&#".f% %,#& +)%,,,#%%)&,,)+#)%&,%)%-#%'&+' -"&,%"%&+!%"%+#)%, B b"%+#))%%,& ))& #"%.&) %-"+)#%-#)%&,%&,,)"%+#)%&,%%'$ +'(,% B +%&+(##&!%,.+ #+'-+#)"% (+ #%.%-%2&!%)($ +!&+' ## +!%%%'#&+,#!&,%, #"%!&,) +'%,#"%+#)%&+'%*%+#(&))"% +)%# #"%(%)%))/

PAGE 40

% B
PAGE 41

,2 #-"%A32%% 8%'#&!%'7&%+# %#%2&,),%'#$)%# 3 +%(+%"%%+#$% +&2"%%))" ,'$&+4 &))%'+-%.%+#&) ,#%-%+#" -"&))%& >&# -*%,%&#% ,#%'$%)%))#"% %, ,#,#"&#%%%'K B &##7.%, ,#, 99S&##%, ,#, S&##.%, ,#,&-% &) )%) %)' +., S#7..%+# %#% "%+#+'$&-+%)"%$&+4&%2+ + %, B =&+': B @/ % B= t&-4* %2'$&+4 #",%*%&)%, ,#, &&))%)-+,#(-#%&# %+# %#%& ,&-%'&#%+#%D+#-#%'E/

PAGE 42

% B +#* %23'$&+4 #",%*%&),2 #-"%,#&-# *&#%%, ,#,+' %+# %#%) !"#&'5(,#.%+#,/ -r "%%8&(+ #%,&(!%+(%)&#%# ,+%-%,,& &2&% &+'(+'%,#&+'#"%&3%#%C( %.%+#,%)&#%#!%+&+'-%,,% !&,%,)#""%%8&(+ #"&,%*%&)&3%#%&#%,#"&#%*%+#,&3%&# &'' # +&)%&,%,(,#$%&4%+ +'%%+,%&3%' # +, +#"%&&# !%+ ,+%8#%.%))&..&$)%!&,F"%%3%-%,!+ # 2%%&-%+%&"%(+ #&!% )%&,)&-% #" +&' ,#&+-% 3."% +'%,+"%5%-#$(#&,' ,-++%-#%'.#,% ,-%/ ))!&,) +'%,%%%%'##"%)&&#&)) +!&))#, &%!(&#,%%))"%-4%'%&4, +!& C( )%&'%#%-#/

PAGE 43

8!%+'%%# ,,(%$%-&(,%"%%8&+*%#,!%+ +#&#% "%&&#%&&,%%))%+# )&#%#+,%"% %2&,'%C(&#% +/ )%-# -&)++%-# ,"%-4%'$%3%*%#%,#(+#%*%+# %)%-#(# %,/ & '%'%#%-#,7,.%)&.,+'& %8# +!( ,"%%% &-%'+%&"%%8&(+ #/ % B & &3%#%&,%,F %8# +!( ,"%7,.%A-&'%'%#%-# &+'!&, +'%#,/ /btbr fn+,#(.%+#,,% %,!&,"&#&&,%'#+&)%#"% -+#&. +&#%'& 8#%,+,(%#"% 8 +!.&+ 3'&,'%) *% +!#"% &.(+#, -+#&. +&+#,"%!&,&#&E%,& %*%+'&(.+# -&))%&#%"%&,%% .%&# +!("%(."% ,%&%#"%+' %-#% +#"%.&)+'(-# ##%-#E"%%! ,#%,

PAGE 44

%+ # &) 33%%+-%-,,"%&#,#+% '!%" ,'&#& ,-)%-#%'&+'' ,&%&, &4,# .%%&4r .%#2&% "% ,#+%%'%'#%&) &#%2 #"++-+-%+#&# +,$%3% #%,# +!#"%,&.%&4%+."%. 8 +!.&+ 3) 2%,#!&, +'%, #"+ +-%+#&# +,%%(,%'D+'.E"%fn,"+ + % B @/ % B 9/fn+,#(.%+#,!&,&#&"/

PAGE 45

:9 f; b n'%#+*%,# !&#%#"%33%-#,!&,%,(%)%))3.&+-% ,%*%&)%,#,2%%(+(, +!#"%&))&%8&(+ #%*%&)%) +&%,#,%% 3.%, +!-)%&+& & +$&,%) +%3.&+-%'&#&+3 '%+-%+#%*&)&, -&)-()&#%, +!#"%& )%&(+,$)%,# +!( ,"$%#2%%& !+ 3 -&+# -"&+!%+'.&))(-#(&# +, 3.&+-%& >&# -*%,%%-%&#%'&# *& -%+#&# +,"!&,%,+,#&+#&'#%,#%% %3.%+#"%%)%-#% -+-%+#&# +,"%+#&. +&+#!&,%,##%%' 33%%+#&',%-*%+3.&# 2&,),& +%'"%+,#&+#'#%,#,,2 #-" +!.)%&+& -+#&. +&#%'& +'* -%*%,&&#&)%-#%2&,"%+-.% '#($ ,"%''&#&/ btntrn nbrnbrn '%#%,#&$) ,#"%$&, -3.&+-%"%%8&(%)%)), 8#%,#(+,%% -%#%"%%8&&, %2 #"(#&" #!%+ +'%+' & &,&2+."%(+' +!,7" -"%%"%)'&#%)&# *%),#&+#%.&#(% &+'"(. #"%%8&&,))%'#&.# .%"&#"% + # &)#&#%.&#%2&,"%&.%&-"#%,#" ,&),+,(%'"%%8&"&%&-" %& C(&, B ,#%&#&#%'+-%.%+#,.%%",%+#*%(),-#(. '&#&+#,%&#%)& >&# (*%#&# +!&#+ + # &)."%' 2&,+-%&,%'%*% +(#%,%+,%"%(%)-%))&%&-"%'&#%&' #&#%#&)',#%#."%(%)%))&',,,%'#,&8 .(.%

PAGE 46

2&##,+'&,'(+&##,#&!%E7-%+#&.E%2&##,E7& 3)rbE7& #%.&#%DE7&+',#&-#%.&#(%E%%)) )!!%'%*% ,%-+',#%&-"%+ +%',#%,"%#&!%+'-%+#'&#&-)%-#%'&"% 8#%,#(+,&,,%'#+,#(-#+3 '%+-%+#%*&)" ,+3 '%+-%+#%*&) 2&,%'#%+,%-"&+!%,)#&!%,'(%+#&. + &# %% !+ 3 -&+#/ trrtrn +#&. +&# 33%-#,&%&, %%%+(+'%+,#&+#)'/#+,#&+#'#"% #&!% ,),+,#&+#&+'&"&+!%#&!%&+$%## $(#%'#"%+#&. +&# %33%-#,#,"%++%-%,,& &),%#$&,%) +%%3.&+-%'&#&"%(%)%)) 3 .,'%+,#&+#) # .%%%',%%%)%-#%'#%"%()&+!% "%%8&(%)%)),#&-"%',%)%-#%'%%."%,%%+#, $%,#%3)%-##"% +%&# "%& >&# *%7#"%&#"%*%"&# 4%) #33%-#%"%#&. +&# r (+,7%&+(#%,7&-"%%% )',%%-.%#%'#&!& +.+3 '%+-% +#%*&)2" -"%&+-%"% -+,#&+ #'-+#&. +&# %,# +!/ rnbrntrn r%)%-#%'-+-%+#&# ,+#&. +&+#,%+#'(-%' +##"% +)%#,#%&. #"%(%)%)) +!&. 8 +!.&+ 3'-+,#(-#%'##($%%,,% #&+,'(-%*-7.&,,)+# )%&+'&& %2&. ##%+#+#"% %,,"%)&#%3"% +#"%)%&' +!#"%(%)%)),-&)-()&#%'. #"%' 33%%+# &)%,,%%&' +!."%##($% +!#"%.(&! *%+$%)K 9=@? / E DB9

PAGE 47

:: )33 %+# /..
PAGE 48

:; &$)% B +#&. +&+#)&#%,%%& )2&#%,7% -+#&. +&+#!&,%,#&+'+#&. +&+#!&, 8#%,##E/ "%& %2&.2&, ##%#+#"%%,, +5%-# +!-+#&. +&+# !&,%,+#"%+)%# #%&."%&.%&',"% )&#% +##"%3(%)%))&+' #"%+-&)-(&#%,"%+%-%,,&)&#%+#&. +&+#!&,%,#& #"%'%, % +-%+#&# + +(#"%%"%3+#+%)3"%&.&" -&)),&#"% '%, %-+-%+#&# (#bE7#"%' 33%%+# &)%,,(%'(-%' +#"%##($%D E7#"% )&#%bE7#"%'%, %'-+#&. +&+#)&#%DbE +'#"% &-#(&)+#&. +&+#)&#%DbE7" -" ,%%'$&-4,%+#"%&.3."% .&,,3)+#)%"%)&+&),%&3."%.&,,)+#)%#,%)3 %+,%"%)&#%' &!&.3"%+#+%)&+'#"%' &&.&+$% 3+#"%&+' nbrnbrn &-"-+#&. +&+#& >&# +-*%,%+%&#%'(, +!#"%&.%%% (,%'"%$&,%) +%3.&+-%%,#,233%%+#+-%+#&# +,&-+#&. +&+#"&*%$%%+, %)%-#%'$&,%"%. ,, +,%#,)%-#% +#"% #%&#(% %* %2"%%)%-#%'-+-%+#&# ,%"+ &$)% B

PAGE 49

&$)% B: r%)%-#% -%+#&# %,# +!$&,%. ,, +,%#,)%-#%' ) #%&#%%* %2/+#&. +&+#&, # +-%+## +#%,#%' r%-+'+-%+## +#%,#%' 9. :. 9$ ;$ trrtrn +,#&+# )&'#%,# ,%%&),3.%'F #&+,#&+#)'#"%#&!% #"%+#&. +&# 33%-#, ))&, )$%%%"%&.% -%%% +%,#&$) ," + #"%$&,%) +%3.&+-% ))$%%'"%3 -+-%+#&# +#&. +&+# ))$% #%,#%'. +(#%,#%&-D.,E3'*%,%33%-#,%,%%+&#& -%#& +-+-%+#&# +'&+#"%%,# )) %3.%(, +!#"&#&.% -+-%+#&# +&+')"%%8& ))%&#%+-)%&+& +(#%,)% +(#%+#&. +&# +#"%+$&-#)%&+& +(#%,%*%%-*% "%33%-#&-"-+#&. +&+#"&,#"%(%)%))#&-4 ))$%'%#%. +%'"% ,)"%#&!%-*% !"%+-%+#&# ,&%' 3"%% ,"&+!% #&!% #""%+-%+#&# +,%)%-#%' rnbrn r#(' +!""%(%)%)),#&-%-*%,"&.3(33%-#,,&,. #&+#, ,#(' +!#"%33%-#,"%.,%)*%,"%%3%3#%"%+,#&+#'#%,#"%+#&. +&+# ,-% ))$%+%3+'#"%%8& ))$%&))%'#(+%,"& #%% 3"% #&!%%#(+,"% + # &))#&!%"% #%&#%(!!%,#,#"&# ""&*% %*%, $)%33%-#,3#%+#&. +&# %*%&)-+#&. +&+#,%%%)%-#%#,% +#" ,%#$%-&(,%"% + %*%, $)%33%-#,+' # +#%+# .&+%+#)&'%#"%%8&(%)%))-"&,%'" ,8 .%+#

PAGE 50

:< brrt ))(,#&#%#"%' 33%%+#&.%#%,#"!%-#"%& >&# +-*%+'"%+-% #"%3.&+-%"%(%)%))7#"%%33%-#,%.&#% ))%#(' %#,2 3."% #%&#%"&#+-%&,%',#&-4#%. &#%+-%&,%,#"%3.&+-%"%(%) -%)))& >&# (*%,2 ))$%&)-()&#%&#' 33%%+#%.&#%,+ +-%.%+#, #&# +!&##,8%-#%#"&#,"%#%.&#% +-%&,%,"%,#&-#&!% 2 ))+-%&,%

PAGE 51

:= f frrnrrrn btnnbrnbrb r%*%&))& >&# (*%,%%!%+%&#%'#& +$&,%) +%3.&+-%'&#& #"%%8&(%)%))(+'%.&)&# +!-+' # +,))%,#,%%+'(-#%'&# &%)&# *%"(. # % B 9' ,&"%& >&# *%,!%+%&#%' btbbnbfrr trbb % B 9/r & >&# *%,%+%&#%'(+'%.&)&# +!-' # +,7= &#."% -%,,% +%)&# *%"(. # t&,%"% 8#%,#(+,+3 '%+-%+#%*&)2&,+,# (-#%'&+' ,"+ 3 % B

PAGE 52

:@ trfrrfr trb % B +3 '%+-%+#%*&)& >&# *%,$&,% 8#%,#,/ btntrrbrb )#& >&# (*%,! *%%8-%))%+#+3.&# $(#"% 33%%+# &.%#%,"!%-#(%)%))3. &+-%#,&, %#%%"%+#&. +&+#,33%-# 3.&+-%%,# +!#"%+#&. +&+#,(+'%&-+,#&+#"%%%',%)%-#% 2%%."%,%%%',$%,#%3)%-##"% +%&# "% & >&# +7"%#"%( *%#"&#,#33%-#%"&+! &# +!-+' # +, %, B & B ;,2"%#&-#&!%&+'#"%#&-2%+,#&+#)', &.7#"%'&#&.,, )&"%,%2&",+'"%+-%" &#&&,))%-#%'%* %%-+',. +(#%,/

PAGE 53

:? tbr !"r#b t b t $ffrb%& t t n n b'( frfr r fr % B :/r#&-#&!%+',#&-%#+,#&+#)3. tbr)!"r* n n t $ffrb%&b '( frfr r % B ;/r#&-#&!%+',#&-%#+,#&+#)3./

PAGE 54

+,%"&#"&+!%+,#&-)#&!%%, !+ 3 -&+#"%+ +# (+! -+#&. +&+#,+#"% #%&.+(+-%#& +# +#%*&)&,+,#(-#%'#&-)#&!% &+',#&-2%&-"'&+' ,,"+ +#&$)% B 9/ &$)% B 9/r#&-#&!%&+',#&-2%))&$)%&+!%,'%.&)&# -+' # +,/ rnbrntrt "%+-%+#&# +,& '%%)%-#%2%%",%+$&,% +,%*%&)& C(&) #%#,!&#"%%3."% #%&#%%* %2.%&*&)(%.%%,%+#% #"%+-%+#&# "&#($%+#-&)$&+%&,&8 .(.*&)(% .%%,%+#,"%" !"%--%+#&# #"&#('%8 ,# & !"#&33 -(.% &$(, #%& #) ##)%*%+# )&# 7D,&))%T,#%%#&+,E"-+-%+#&# 2%%%,#%&#"%#%%%)%-#%'-+,#&+#)',& >&# -*%,%%&),+%&#% (, +!) +-%.%+#,.#"%&8 (.. trr "%%8&(%)%))&,))%'#%&#%)%&+& # ) #"&%&-"%&C(&, ,#%&#&#%/"%%)%-#%'-+-%+#&# &.'%&,"%+ +#'(-%'&# -+,#&+#'&+'&))%'#&#%3. +(#%,3 #&!%&,'%#%-#%'&# #"%"% +(#%,"%(%)%))&,))%'#+# +(%(++ )%&+& # $#& %-%+3.&# +,#&+# & r#&-#&!%E r#&-%2&##,E :;/U ;/: ::/;U
PAGE 55

;9 "%+-%+#&# +,%%&-#%,#%'&#"%%', &.E% #"%+-%+# &# "%'& !+ 3 -&+#"&+!% +% #"%#&-#&!% ,#&-%#&))"%%,#,"%+-%'#"%-)%&+& #%,#, % B ,","%.%,##+,#&+#./ t+r,rf#!)! n n $ffrb%& '( frfr r % B & '%%,#9.E#+,#&+# )3./ "%#&!%"%)%)&# *%)+,#&+###,2" )%"%2%%.& +%'&# &&##,"%3)(-#(&# +,,%%+ + % B +' &))#"%'&#&)%-#%'&% '(%"%!%%)))%,"%%8&(%)%))&(#&# -&))3 .,"%,%!% -)%,%*% B %-',"%#&!%%#& + +' '(&)-%)), #" #"%#&-4&% .+ #%+-%"%)#&!%&'%, !+&#%' # -&)*&)(%"%+# %#&-4 ,!% "%!%+(%)%)%&,%+#&. +&+#,"%+ % '%" ,!%)% %3)+-%&,%,"%,#&-#&!%

PAGE 56

"%#%,#,&##"%+,#&+#).-%', )&%,(#,&, ,"+ + % B #""%#&!%,%%.& + +! #" +#"%(+-%#& +# +#%*&)'%3 +% r )&)"%%,#,3.%' #.' '(-%& !+ 3 -&+#"&+!%, +#"%(%)%))3.&+-%,&+$%%%+ + !% B <&#-+,#&+#.,/ *!t+r,rftrr& !b t b t b trb%& t t n n b b '( rfr r % B
PAGE 57

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

PAGE 58

;; tbrf*!fr-#!!t.*!fr-tbrf*!frrr& t n b t n b $ffrb t t n n'( frfr nfr trfrfr #!! b't.bb' frbrf % B =/& '%%,#9.E#+,#&+#)./ %,#%#"%%'(-# +,#&-#&!%&%"%%8&' %-*% "%#&#&!%%-*%%'#?=P#, +&)#&!%" )%"%%%-*%%'#?


PAGE 59

tbrf*!fr-#!!t*!fr-tbrf*!frrrb*! t n b t n b $ffrbn b b '( frfr nfr trfrfr #!! b't.bb' frbrf % B & '%%,#9.E#+,#&+#)./ "%#&-)#&!%&,%'(-%" )%#"%#&-2%&,%'(-% :/:P)#"(!""%%"%+#&. +&# %%.% ##%&' )%-%&,%"%#&!%&, .% 2%+#!& +#"%%8&%%.%'#%-*%3#%+#&. +&# "%#&-4#&!% %-*%%'#?@/
PAGE 60

;< nbrnt & >&# -*%,%% !%+%&#%'t%! ++ +!&# &."%'&,+-%&,%' +-%.%+#,.,*% +(#%,,%* )#% .' +#"&*% !+ 3 -&+#-"&+!%"%(%)%))%3.&+-%7#"% & >&# -*%,!%+% &#%'"%,%2+-%+#&# +,%)) #" +#"%+3 '%+-% +#%*&)2+ + % B % B ?,",& >&# *%,& "%-*%&)), !"#, '%"%+#%*&)&+'#"%%3%"&,+'&'*%,%33%-# #"%3.&+-% &#&+#,.,+#"%.%,#(% #&4%+'(%#"%&,,)+#)% ,,(%, ,-(,,%'/ r!r!! trb rrf t! !"r# $% t!"r# $% "r&f'( f r#r)tf)f )* % B & >&# -*%,.&'%/

PAGE 61

;= n0rntb! %&) >&# #"&## -&))%*%),&'%'%#%-#%$&+&%& ,#%%#,"&'+33%-#"%3.&+-%"%(%)%))7&##%+# &,&2+#.% %8#%.%&,%,"%%&.'% !"#$%#(-"" !"%%*%),8&.%7 3 3(%)%))%" -)%&, +# &33 -$%" +'&' %,%)(-4%. ## +!#,8"&(,#' %-#) +#"% + #"% +#&4%"%b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

PAGE 62

;@ trr& !'f0fr1frt.#!-*!-/!-2! t n b tnb $ffrbt t n n b b '( frfr bttn ttn fttn rbttn % B 9/&' +5%-# +, .E#+,#&+#)3./ tr& !'f0fr1frt+r,r#!!-# !-#)!-#/! t n b tn $ffrb +,f(! "-(!f +,f(! "-f ,f(! "-(! +,f(t! "-f % B 99/&' +5%-# 9.E#-+,#&+#.,/

PAGE 63

;? "%& +5%-# +,+#&. +&+#,"%'#"!%-#,..%' &#%"% "%"&+'#"% .%"%(%)%)) %-*% # +&)#&!%&,&),& n+,%&,%,"%)#&!%-#(&)) .*%'&3#%%-*%" ,% +,#&#&!%&"%&-#"&#(+'%+#&. +&#%'-+' # ,"%(%)%))#&-4.&$% "%&# 2" -"(%, (#+$%##%3.&+-%"%#&-& +!#"% #%.&#%"%#&,+#"% # !&# +!#"%%!&# *%33%-#, -+#&. +&+#,/ &$)% B %-%+#&!%"&+!%,)#&!%&2% &+5%-# +,&# -+,#&+#. ,/+-%+#&# +,D.E %-%+#&!%'D)#&!%A2%E %-%+#&!% f %-*%D)#&!%A2%E 9 /@PA9/9P ??/@?PA9/9P 9 9/:;PA9/::P ??/@;PA??/<9P 9; /99PA/;P 9/:;PA9/:;P 9< :/PA:/9@P 9PA9/9P "%& .%%,#&,+'(-#%'&#" !"%.+'%,(#%' & !+ 3 -&+#)" !"%3.&+-%+' ,"+ % B "%.%,# 2&, )&"%.%,##"%-%+#"&+!%,3.&+-%%%(-"" !"% &,"+ #&$)% B :/ "%,%%,()#,!& +,2"&#" !"%"%#&-%+# &)%,()#,&&!% 3.&+-%1"%+-+, '% +!&*%" -)%%%b(%)%))" ,'*%,%%33%-# $%-% # -&),,(%--%)%&# %C( %,,#%%+-%&, %+%"% ,#&-"%+#&. +&# "%(%)%))('!%&#)+" $ #"%(%)%)) '(+!#"%%+%-%,,&+'"%+-%%'(-%"%&)) *&$ ) #"%*%" -)%/

PAGE 64

trrb*!t.fr1fr#!!-# !-#)!-#/! t b t b tnb $ffrb +,f(! "(!f +,f(! "-(!f +,f(! "-(!f +,f(t! "-(!f % B 9/&' +5%-# .E# +,#&+#.,/ &$)% B %-%+#&!%"&+!%,)#&!%&2% &+5%-# +,&# -+,#&+#.,/+-%+#&# +,D.E %-%+#&!%'D)#&!%A2%E %-%+#&!% f %-*%D)#&!%A2%E 9 9/@ PA 9/9 P 9/9 PA 99/?? P 9 ;/ PA ;/<9 P 9/;? PA 9/@ P 9; @/;PA@/; P 9/ < ;PA 9/9? P 9< 9/:?PA9/ P 9 /< PA9 /?? P %nrn "%+-%+#&# +,%)%-#%'#%%,%+#"%&*%&!%+'.&8 .(.%*%), &$&+&%&%% %,-# *%)"%,%-+-%+#&# +,%%%,#% ,%&#%)&#%%-+,#&+#',D.E& >&# *%,+'#"%& +5%-# #%,#,%%),-%#%'&,+#"%-&$.'%%,#,

PAGE 65

trr "%+,#&+# #%,#,+'(-#%'+ #%+' 8 '%%%#"%&.%+'(-#%' -& '%"%%8&&,))%'#%&#%)%&+& #%&-"&C(&, B ,#%& ,#&#%$%3%#"%%)%-#%-+-%+#&# + #!%+' '%&,+#'(-%"%&,, 3)+#)% 2&,'%, !+%'&.'%)7,+*%, 3&-#&, +%%'%'#%"%)+#)%%,"%.&)%+,%-"+)2" -"&)), #"%%+*%, &-#,"%&) &#%!&,%,#"%!&,%,"%+*%, 3&-#7&),-&) )%'&&-#7&,-%+#"%)&#%%&' +!/ n#&,"&#+% #"%#"%"%'& !+ 3 -&+#-"&+!% 3.&+-%,&+$%%%+ + %, B 9:& B 9;/ fr3ff#!!+rrb !b t b t $ffrb%& t t n n b frfr r % B 9:/ #!%+' '%%,#9#+,#&+# )./

PAGE 66

fr3ff)!!+rrb !b t b t $ffrb%& t t n n b'( frfr r % B 9;/ #!%+' '%$E#-+,#&+#)./ !"%+-%+#&# +,%%%,#%&#-+,#&+#', +'&+ 3.&+-%"%" !"%,#+-%+#&# &, #%"%&+!%"%&,,) -+#))% "%&8 .(.+-%+#&# "&#)'$%(,#& +%"%+#)%&, .!& +#" ,-+-%+#&# +#%)'&+'*%,%%33%-#,,2 + % B 9

PAGE 67

fr3ff#trr&*!b t t n b $ffr%&b b '' frfr r % B 9/ #!%+' '%E#+,#&+#./ n 0rntnrn "%& +5%-# %,#,3.%2 #"-&.'%%)'%'#"%# +#%%,# +!'&#&+'"%+-%&,!& 3.%2 #"+ #!%+' '%"% #,"% .&,,3)+#)%%*%+#%'#%,# +!-+-%+#&# +,$*% .3"%+,#&+# )'#%,#$%-&(,%#%C( %'#"%)-+#))%(,#& &" !")&#%3 ,($,#&+# &).(+# .%%*%"%-+#))%&,$)%+')% !")&#%, &# .%V. +(#%E" !"%-%+#&# +,-('$%%&-"%' +#"%& +5%-# +,#%,#,"%,%#%,#,%%(+&#.$%-&(,%"%% 3)&#% %C( %"% )"%%,(#,&%"+ + % B 9
PAGE 68

trr& !'f0. fr1fr2!!-#!!!-# !!-#)!!+ t n b n $ffrb +,fb#( !f +,f!( !f +,f*!( !f +,f*!( !f % B 9
PAGE 69

-&+$%%%#"&##%+##"%33%-#,%.&#(%%.(-.% !+ 3 -&+#" .&'(%#"%&-##"&## !"%#%.&#%"%%.&+%& %() &#%(# #" +-%&,%,+-(%+#"%&#%&#%-# ),+-%&,%, # +&))" -" #+%%"%%.&+%.,#+'#"%-# *% #%,"% -&#&)#+&+'"%+-% .*%,"%(%)%))%3.&+-%"%33%-#,%. &#% &%"+ + % B 9=/ $4rbffrt trb t b t'( ./0 ./t0 ./n0 ./b0 ./0 ./t0 ./n0 ./b % B 9=/%.&#%33%-#,(%)%))%3.&+-%& >&# *%,+' %*%,"+%.&#%,E/

PAGE 70

f< rnfb nr t&,%8% .%+#&)&#&#&,"&##"%-&))%*%),&'% &+'+ #!%+' '%+' $&&%&,#, !+ 3 -&+#)33%-#"%%3.&+-% b(%)%))"%# .%+#%*&)%,#%'/ %*%#"%8% .%+#7#"%(%)%)) 3.&+-%$%!&+##%*%),+'" !"%)#*%!"7 #"%,%)%*%),('$%(+'(+'%8#%.%+' # +,+&"%&&33 -(.% &%& "%&.'%,+ +!.%-"&+ .%%.,(+-# "-%)) %+# &)++-%+#&# +,"%#&!%&+',#&-%2%%..%' &#% &3#%+#-# "%#"('$%%#"&#"%# .%+#%*&)%,#%#"% &'*%,%%33%-#,%%.%',#&+&&, B ,#%&'#&#%(#%4,"('&),+, '%#"% %33%-# .%+#&. +&# "%%8&(%)%))"&&*%!% -+,(. &#%&+'#"(,+!%# .%'#%,#,%%%&, $)% "%&+, %+#%,%%,#,%*%&)% .#&+#+3.&# $(# "%b(%)%)) &$ ) #%-*%& ')&3#%"%-%,,&# "%+#&. +&+#"%(%)%))&,$)%# %-*%%#%), #%"% %3+5%-# +,*%!"-+-%+#&# ,"% -+#&. +&+# "%8% .%+#,+*)* +!+ #%' '% +#&*% !+ 3 -&+# %33%-#"%3.&+-%+-%+#&# +,.+#"% .%+#%*&)%,#% +%%&,"%.&"&*%33%-#%#"%3.&+-%#"%(%)%))&#"% -+-%+#&# +,%,#%'&,"%&-##"&# #!%+' '%,%8#%.%)($%+&#%

PAGE 71

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n3(#%4 +,#(' +!# "% %33%-# +#&. +&+#,"%&#"'%7%33#,"('$%(+%'#, !+ +!&+' -+,#(-# +!+' '(&) +!)%b(%)%)),+,#(-# +!,%*%&)' 33%%+#%.&+% %)%-#'%,,%.$) %,b,E('&))%,%&-"%#' %-#+3.&# $(# -+#&. +&+#33%-#,"%%%%& &$)%, +#"%%8&#%,# +!#"%%(+4+ $%-&(,%#"%%8&,.%8,#%. 1 #", +!)%b(%)%)),7.%%*%%#&. +&+#, #, $)% %*%, $)% %33%-#,('$%%,#%), #", +!)%% )),#"%33%-#&# &.%#%,(-"&,

PAGE 72

%,,%,7#%.&#%,&+'"(. #%.&#%,"%+%!&# *%%33%-#, -+#&. +&# ('$%,#(' %"%$ ) ##+#&))#"%,%&.%#%,('! *% -)%&%(+'%,#&+' "%+#&. +&# +.%-"&+ ,.&.%8% .%+#&)#%., ,"+ + % B 9/ % B 9/&.%8% .%+#&)#%.+#& + +!& +!)%b(%)%))+'*& (, -+#,+#))&# &&.%#%,b%#&)E f%,%&-" +!+'#%,# +!' 33%%+#&#& )#,,%' +#"%)%-#'%,,), %-.%+'%'(#%)&# +(.))&#&),#,&, !+ 3 -&+#)+"&+-%"% #%&+-%)%*%).3(+#&. +&+#," )%),%'(+!#"%#,"%*% %8%+, *%%&# +(.&#&)#,

PAGE 73

n tn1bff fb "%) +! ,"%)-4' &&.&+'+#+%)"%& %2&. 2 ##%+#+##"%)2&#%+#&. +&+#!&,%,$&,%+#"%)&#% +! +#"%(%)-%))/

PAGE 75

<9 n rf B ffr ,#%'"%%&%"%#&#,"(#'%%,"%%8&(%)%))+'#"% ,#&##%,# +!A,"(#'%%3"%&,.&#&"/ %"b-rr 9/ "%-4"%#"%(33 %+#!% ,%,%+##"% +'% &%&-% +'%3%-%,,& "%-4#"%++%-# ,%#2%%+#"%%8&+'#"%!%+'% :/ "%-4#"%&'-++%-# +, ;/ "%-4#"%!%%,,%#"%+'%& B &E "%!%+,
PAGE 76

;/ r2 #-"%% r2 #-"%& +#!!)%A$(##"%#2&% %/ B r%#%%,,%%!()&# B "%-%+',#&#"%-&#&&./D%&4r .%E B b&4%%#"%%+#'%#%-##!!)%2 #-"7)&#%+#"% !"# '% #"%&#&, +"%3 # +#"%& %"% ,2 #-" ,)&#%"%%3# '%"%&#&"/ B +!#"%' ,&"%3+#"%&#&.+ ##"%+&+' %.&#%,& #(+# )#"%&*%%&-"%#"% %# +#,$%3% -+# +( ,"($% ;=+'V%,-# *%)"%%#+# &+'&-#(&)&)(%,&+$%' ,&%'$+ +!% #"%"%2+%8# #"%"% &#%%)%-# "%!!)%,2 #-"$%)2"%' ,& .&4%,"%%,+'# #"%#"%)% %## +!/ B ), +!#"%' ,&"%-4#"&#"%(.+"%&%,,(%, &+'#"%& %%,,%, 3 #"%"%,%*&)(%,&%' 33%%+# #"%%-($%4&!%"%(.& %!&, )#%%.%3% #"%& +(&)#"%"%)

PAGE 77

<: B "%-4& %!&,)2#f%3&,8 #R&#%"%%3# '% #"%&#& ...%,%"%"%($% 2&#%+') $($$%, B "%-4&.%!&,)#"%&.%!&,8 #)&#%'(+'%+%&##" % -*%"% !"#&+', '%"%)(." ,&.%&++%, #"%%3%%+-%8 #/ B b*%"%%+#!!)%2 #-""%, # ,&$%)%'&, Q" !"R/ tr ,%,#-%%)'$%)%'%&-"# .%&&. )%,+&)%"% -&#&,*%%+, # *%7,& &# +, %%&&(,%$&'%,(#,/ B b&4%%#"%##%+( #-" ,,%##"%,2 #-" ,)&#%"% !"# '%"%&#&"/ B ,%"%%'5(,#%#"% )#&!%"%+,&#%'(,# &$*%"%##%+( #-"%-#(&)##&!%-&+$%%%"% ,#2&%' ,&+' ,&$%)%'&,r#&+'$KR"%#&!%,"&*% $%8&-#)%#&#&+' # )) 3# +!#%,# +!/ B &.%!&,&2$%&4% ,##"%+%%')%"('$%)(,"% ,&+$%&-,"% #"'&2 +!,%&.%!&,+'#"%+%5%-# +! #3" ,,'+%%*%&) .%,7& "&#&,+#"%+%%')%,"('$% ,&-%'/ B 1 #"'&2&.% !&,+',#&#&#&(+%)%-# +! f(+'%#"%&#&%+%&4r .%/

PAGE 78

<; B #. +(#%,$%! +5%-# +!#"%&.%+#"%%(. + ," +5%-# +!&# ,)+5%-# .%#+,%,"&#&.%%&4,$&-4#(!" #"% %#(.&#%"%+%%')%2" )%#, +"%%#(." ,( -&(,%#(#)(!+-%"%%%')% #"&2+7-"%-4#&4%%# ,(!!%'/ B +-%))#"%&4,%"+7,#"%&#&+'&+&)%"%%,()#, %)%# +!f%,(#,(+'%"%+&)>%%+( +#"%#2&%/'%#,%% -+-%+#&# %,(#,7 #.&+%-%,,&)'"%&) &# )%,/ -r,t B 3%+#* +!#"%%+##!!)%2 #-"#. '')% # B 33"%& %#"%"&#& B 3.% B )%& %!&,%!()&# )*%/

PAGE 79

nrfr t&,-"(4/6/ &.'%,+ +!8-"&+!%%.&+%(%) -%)),+#6+&)+%%,%&'5K B =9:/ t%##,&+ %)'%) +!&+&) ,(%)%))+! +%,&+,#&# -&# +7&,#% ,"%, ,+ 3) '&& +%,* ))% t).%/6/b/6/7b(!%2&(%)%))#%.,)%+(.%,,%247??:/ t ,7b(%)%)),K"% )%-#"%. ,#b-&2 B ))%247?
PAGE 80

<< tnfn0 %& -"&%)r() *&+&,(!(,#(%%+,%2 # %,"% ,# -"&%)+'b&!&%#r() *&+&+''%#"%%&+&+ &# &r() *&%&'&##%+'%'-"$ ,) !-(+ "%& ,# "%,+'+##(..%,#"% B (',&))%&.4 +! # .%+#&))+' -&'&+''%&3" )%+/ %&'&##%+'%'&(..%%,%&-&.%"%&# +&)%+-% (+'&# &#)%.,+ *%, #2" -"%+-&!%' .&'(&#%-) &+'-+# +(%" ,'(-&# 3#%&'(&# +!."%%2+,# #(#%%-"+) &%&'%)%'&#"%+ *%, #3) '&%&'(&#% & 2 #"&&,#%%+-%+.%-"&+ -&)+! +%% +! 7,&) > +! #"%.&)%+-%&+' 3)( ''+&. -,/ %&&+,&*%)%2.+#",$%3%%#+ +!#%2 #