Considerations in Designing a Nu cl e a r P o w er Plan t with a Hydrogen and Biofuels F acilit y
Lauren Ayers, Matthew Chapa, Lauren Chilton, Robert Drenkhahn, Brendan Ensor, Jessica Hammond, Kathryn Harris, Anonymous Student, Sarah Laderman,
Ruaridh Macdonald, Benjamin Nield, Uuganbayar Otgonbaatar,
Alex Salazar, Derek Sutherland, Aditi Verma, Rebecca Krentz-Wee, and Elizabeth Wei
Decem b er 14, 2 011
Con ten ts
I I Bac kground 10
1 Core 11
1.1 Main Goals of the Co r e Group 11
1.2 Design P arameters 11
1.2.1 Biofuels Co ordi nation 11
1.2.2 Viabilit y to get Licensed and Built in the Up coming Decades 12
1.3 Design Opti ons and Ev aluation 13
1.3.1 Liquid Metal F ast Reactors 13
1.3.2 Molten Salt Reactors 14
1.3.3 Gas Co oled Reactors 14
1.3.4 Sup ercritical Co olan t Reactors 15
1.3.5 T able of Design Comparison 15
2.1 Goals of the Pro cess Heat Design Group 17
2.2 Design Challenges 17
2.3 P ossible Heat Exc hanger Designs 17
2.3.1 Straigh t Shell-and-T ub e 18
2.3.2 Mo dified Shell-and-T ub e 18
2.3.3 Plate 19
2.3.5 Ceramic 19
3 Hydrogen 21
3.1 Goals of Hydrogen 21
3.2 Design Opti ons and Ev aluation 21
3.2.1 Steam-Me t hane Reforming 22
3.2.3 W estinghouse Sulfur Pr o cess 23
3.2.4 Hydrogen from Urine 24
3.2.5 Hydrogen from Bacteria 25
3.2.6 High T emp erature S te am Electrolysis 25
4 Biofuels 27
4.1 Main Goals of Biofuels 27
4.2 Design P arameters 27
4.3 Design Opti ons and Ev aluation 27
4.3.1 P ossible Sources of Biomass 27
4.3.2 Electrofuels to Hydrogen Pro cess 29
4.3.3 Algae T ranses terifi c ation Pro cess 30
4.3.4 F ermen tation to Ethanol Pro cess 30
|
6.1 Pro cess Ov erview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . |
||
|
6.1.1 Core Ov erview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . |
||
|
6.1.2 Secondary Ov erview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . |
||
|
6.1.3 T able of Imp or tan t P arameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . |
||
|
6.2 Primary System Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . |
||
|
6.2.1 F uel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . |
||
|
6.2.2 Criticalit y Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . |
||
|
6.2.3 Sh utd o wn Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . |
||
|
6.2.4 Thermal Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . |
||
|
6.2.5 Depletion Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . |
||
|
6.2.6 Core Reactivit y F eedbac k P arameters . . . . . . . . . . . . . . . . . . . . . . . . . . . |
||
|
6.2.7 Natural Circulation and Flo w Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . |
||
|
6.2.8 Safet y Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . |
||
|
6.3 Secondary System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . |
||
|
6.3.1 Heat Exc hanger . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . |
||
|
6.3.2 Condensers and Compressors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . |
||
|
6.3.3 Acc i den t Scenario Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . |
||
|
6.4 Economics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . |
49 |
7.2 Heat Exc hangers 52
7.2.1 Choice of Material and W orking Fl uid 52
7.3 PCHE De sign 54
7.4.1 PCHE1 55
7.4.2 PCHE2 57
7.4.3 PCHE Conclusions 59
7.6 Heat Exc hanger at Biofuels Plan t 60
7.7 Heat Sin k 61
7.7.1 Purp ose, Lo cation, an d Comp onen ts 61
7.7.2 Heat Rate and Des ign 61
7.7.3 En vir onme n tal Concerns 61
7.8 Pro cess Heat System Costs 62
7.8.1 PCHE 62
7.8.2 Circulator 62
7.8.3 Piping 62
7.9 Heat Storage Details 63
7.9.1 Lithium Chloride 63
7.9.3 System Design 64
7.9.4 PCM Sizing and Design 67
7.9.5 Heat Storage Summary 70
7.10 Circulator 70
7.11 Heat Sin k 71
7.11.1 Purp ose, Lo cation, an d Comp onen ts 71
7.11.2 Heat Rate and Des ign 71
7.11.3 En vir onme n tal Concerns 71
7.12 Piping 71
7.13 Pro cess Heat System Costs 73
7.13.1 PCHE 73
7.13.2 Circulator 73
7.13.3 Piping 73
|
8.1 In tro duction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . |
||
|
8.2 High T emp erature Steam Electrolysis (HTSE) . . . . . . . . . . . . . . . . . . . . . . . . . . . |
||
|
8.2.1 HTSE Pro duction Plan t . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . |
||
|
8.2.2 Materials and Comp onen ts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . |
||
|
8.2.3 Material Corrosion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . |
80 |
|
|
8.3 Other De sign Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . |
||
|
8.3.1 Biofuels sh utdo wn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . |
||
|
8.3.2 Core sh utdo wn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . |
||
|
8.3.3 Mec hanical failures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . |
||
|
8.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . |
9 Biofuels 82
9.2.1 Densification 84
9.2.2 T ransp ortation to Site 85
9.3 Gasification 85
9.4 Acid Gas Remo v al 87
9.5 Fisc her-T ropsc h Reactor 89
9.6 F ractional Distillation and Refining 94
9.7 Biofuels Results Summary 97
IV Conclusions 98
10.1 Short-T erm F uture W ork 98
10.1.2 Biofuels Plan t 98
10.2 Long-T erm F uture W ork 99
10.2.1 Core 99
10.2.3 Biofuels Plan t 99
11 Economics Of Design 100
References 101
VI App endix A: Core P arameter QFD 109
VI I App endix B: Criticalit y Mo del 111
VI I I A pp endix C: In v estigation of PCHE thermal h ydraulics 120
IX App endix D: Implemen tation of switc hgrass as feedsto c k for a indus trial biofuels pro cess in a n uclear complex 126
X App endix E: Impact of gasifier design on FT pro duct selectivit y 130
List of Figures
1 Com bined Multiple Shell-P ass Shell-and-T ub e He at Exc hanger (CMSP-STHX) with con tin u
ous helical baffles [158] 18
2 A PCHE heat exc hanger made of Allo y 617 with straigh t c hannels [160]. The semi-circular
fluid c hannels ha v e a diameter for 2m m. 20
3 Steam methane reforming blo c k diagram [48] 22
4 Electrolysis pro cess blo c k diagram 23
5 The W estinghouse Sulfur P ro cess for h ydrogen pro duction [4]. 24
6 Sc hematic represen tation of the direct urea-to-h ydrogen pro cess [36]. 24
7 Sc hematic system arrangemen t of UT-3 pro cess [31]. 26
8 p ossible pro duction paths for commercial pro ducts 28
9 Sim ulated p oten tial for switc hgrass crop with one harv est p er y ear[147] 29
10 Basic outline of the pro cess tur ning biomass in to bio diesel fuel 31
11 V olumetric energy densit y of alternativ e commercial fuels burned for energy[30] 31
12 Radial view of th e core sho wing the la y out of the fuel assem blies, con trol r o ds, reflector, and shield. 35
13 T able of imp ortan t parameters f or the reactor, still to b e done are depletion calculations an d kinematics 36
14 K-effectiv e v ersus ro d p osition for final core mo del 37
15 Thermal Conductivities for differen t mate ri als in a fuel pin 38
16 T emp eratu re s at differen t lo cations in the f uel pin with v arying axial heigh t. This is sho wn
with the axially v arying lin e ar heat rate (in blue). 39
17 Sk etc h of mo del for LB E flo w calculations 41
18 Mass flux through do wn c hannel giv en an outlet temp erature of 650 C C for v arying inlet tem p eratures and v alues of D. Green represen ts D = 1 m, blue represen ts D = 2 m, and red represen ts D = 3 m. 42
19 V ariation in Reynolds n um b er for flo w through do wn c h annel giv en an outlet temp erature of 650 C C for v arying inlet temp eratures and v alues of D. Green represen ts D = 1 m, blue
represen ts D = 2 m, and red repr e sen ts D = 3 m. 43
20 The axis going up the screen is mass flu x ( k g / s − m 2 ), the axis on the lo w er left side of th e screen is inlet tem p erature ( C C), and the axis on the lo w er righ t side is outlet tem p erature ( C C). 44
21 Ov erview of the secondary system of the reactor 45
22 The no dal v alues of te mp erature, pressure, and mass flo w for the secondary system 46
23 Prin ted Circuit Heat Exc hanger (PCHE) Design 47
24 Geometry options for the in-reactor shell-and-tub e heat exc hanger. 47
25 An o v erview of the pro ce ss heat system 51
26 Axial temp erature profile of the hot and cold fluid and hot and cold c hannel w alls 56
27 Heat flux profil e for PCHE1 56
28 Minim um m ass flux and flo w qualit y in PCHE2 57
29 Axial temp erature profile of PCHE2 58
30 Heat flux profil e of PCHE2 59
31 The basic design of the storage heat exc hanger 64
32 A view of the cross-s ectional flo w area of the storage heat exc hanger. 65
33 The c harging la y out for the storage heat exc hanger. 66
34 The storage lo op during energy disc harge. 67
35 The mass flo w rate of the LBE as a function of time when the storage device disc harges. 70
36 Helium Pip e La y out. Adapted from [87] 72
37 HTSE Hydrogen Pr o duction Plan t 75
38 P o w er Requiremen t for W ater Electrolysis [121] 78
39 SOEC sc hematic 79
40 Sc hematic of prop osed biofuels pro duction plan t 83
41 Sites for switc hgrass gro w th in the U.S. 84
42 Sc hematic of silv agas gasification pro cess 86
43 T ypical acid gas remo v al pro cess after gasification of biomass without a com bustion c ham b er 87
44 Amine Acid Gas Remo v al Pro cess 88
45 LO-CA T acid gas remo v al pro cess 89
46 Slurry phase b ubble reactor sc hematic [100] 90
47 The effec t of feed r atio r = H 2 /CO on selectivit y at T = 300 Q C [100] 91
48 ASF distribution for c hain gro wth [100] 92
49 Effect of catalyst concen tration on con v ersion ratio 93
50 Effect of sup e r ficial v elo cit y , catalyst concen tration on the n um b er of co olan t tub es [122] 93
51 Heat transfer co e fficien t as a function of catalyst concen tration 94
52 Distiller Sc h e matic 95
53 Betc hel Lo w T emp erature Fisc her-T ropsc h Refinery Design 96
54 Core house of qualit y 109
55 PCHE no dalization [54] 120
56 PCHE v olume and heat transfer co efficien t as a function of c hannel diameter 122
57 Reynolds n u m b e r and Pressure drop as a function of S-CO 2 mass flo w rate 124
58 Heat transfer co e fficien ts and PCHE v olume as a function of S -C O 2 mass flo w rate 125
59 Map of biomass resources in T exas. Harris Coun t y is noticeable in dark green to the lo w er righ t. (National Renew able Energy Lab oratory) 127
60 Silv a gas and FICBC for pro duct selectivit y 130
61 Blo c k diagram of the UT-3 plan t. 134
62 Prop osed Calcium P ellet Design [131] 135
63 The m i xe d conducting mem bran e pro cess. O 2 is s epar ate d from the steam mixture and diffused across as ions with the aid of a pump. 136
64 Hydrogen storage options with corresp onding energy , op erating temp erature, and wt.% [123] 138
List of T ables
1 P o w er densit y comparison for differen t reactor t yp es 12
2 Reactor Design Comparison[88, 139, 62, 49, 120, 44, 98, 126, 136] 16
3 Principal F eatures of Heat Exc hangers (adapted from [133, 103]) 17
4 Inputs and Ou tputs Comparison for Biofuel Pro duction Pro cesse s 28
5 Comparison of Biomasses for Biofuel Pro duction 29
6 Summary of turb ine costs [54] 49
7 F ractional costs of the differen t sup ercritical CO 2 cycle designs [54] 50
8 P oten tial W orking Flui d Prop e r ties [18] 52
9 Summary of PCHE results 53
10 P oten tial Materials for PCHE 54
11 Impact of design parameters on PCHE v olume and pressure drop 54
12 Steam/h ydrogen and o xygen streams from the h ydr oge n plan t 60
13 PCHE capital cost 62
14 Relev an t ph ysical pr op erties of LiCl 63
15 The comp osition of Allo y 20. Adapted from [11]. 64
16 A summary of the heat storage results 70
17 Losses th rough helium transp ort pip es 73
18 PCHE capital cost 73
19 SOEC Outlet Mass Flo ws 76
20 Summary of option s for solid ele ctrol yte materials 80
21 Mass distributions of w o o d and switc h gras s [39] 86
22 Comp osition of Syngas Output from Silv agas Gasifier [118] 87
23 Comp osition of Syngas Output after Acid G as Remo v al 89
24 Pro ducts of Fisc her-T ropsc h Pro cess and Relativ e Boiling P oin ts 94
25 Straigh t c hannel and zigzag c hannel PCHEs 121
26 Hydrogen Pro duction Rates and Thermal P o w er Requiremen ts 133
27 Comparison of CVD (TEOS) and Zr Silica ceramic mem branes 135
P art I
In tro duction
The p olitical and so cial climate in the United States is making a transition to one of en vironmen talism. Comm unities are demanding green e n e rgy pr o duced in their o wn coun try in order to curb global w arming and dep endency on f oreign energy suppliers. Because of these trends, the o v erall design problem p osed to the 22.033 F all 2011 class is one of a green energy pro duction facilit y . This facilit y w as constructed to con tain a n uclear reactor, a h ydrogen pro duction plan t, and a biofuels pro duction plan t. This design w as c hosen so the reactor can pro vide heat and electricit y to supply b oth h ydrogen and biofuel pro duction plan ts their necessary input requiremen ts. S ince n uclear p o w er can also pro vide m uc h more energy , a cle an form of electricit y can then b e sold to the grid. The h ydrogen pro duction plan t’s main goal is to pro vide enough h ydrogen to p o w er a biofuels pro duction cycle. In order to mak e this plan t more economically feasible and in line with the gree n energy goal, no extra h ydrogen will b e sold. Instead, their sole pur p ose will b e to supply the biofuels facilit y with as m uc h h ydrogen i s required. In turn, the biofuels pro duction p lan t will pro duce b oun tiful bio diesel and b iogas ol ine for sale to the public.
Due to th e complexit y of this plan t, the design w ork w as sub divided in to four groups: (1) Core, (2) Pro cess Heat, (3) Hydrogen, and (4) Biofuels. The core gr oup w as resp onsible for creating a reactor design (including primary and secondary systems) that could p o w er h ydrogen and biofuel pro duction plan ts with b oth electricit y and heat in add ition to selling an y additional electricit y to the grid. The pro cess heat group w as resp onsible for transferring the heat pro vid e d b y th e reac tor to the h ydrogen and biofuels facilities (in addition to wherev er else heat w as ne eded). The h ydrogen group w as resp onsible for receiving the core heat from p ro cess heat and creating e nou gh h ydrogen to p o w er the biofuels pro cess, if not more. The biofuels group w as resp onsible for taking that h ydrogen and pr o cess heat and creating bio diesel an d biogasoline for sale to the ge n e r al public. There w ere t w o in tegrators to the com bine the design w ork from all four groups and presen t a cohesiv e green energy f ac il it y .
This design is quite significan t b ecause the w orld is starting to turn to w ards green pr o duction and in order to meet p eople’s gro wing energy needs, large, clean electricit y and gasolin e generating plan ts will need to b e built. F acilities that pro duce b oth carb on-emission-free elec tr ic i t y and fuel will hop efully b ecome a rising trend to help com bat global w arming an d other increasingly alarming en vironmen tal concerns.
P art I I
Bac kground
1 Core
1.1 Main Goals of the Core Group
The o v e r all goal of the pro ject w as to pro v e that n uclear plan ts can w ork in tandem with h yd roge n an d biofuels pro duction and furthermore that the p oten tial reactor t yp es are not just limited to high temp erature gas reactors but can also w ork at lo w er temp eratures, suc h as those pro duced b y liquid salt and metal co oled reactors. With those o v erarc hing goals in mind, the pri m ar y goals of the core group w ere to design a reactor with:
• sufficien tly hi gh outlet temp erature to b e useful for pro cess heat applications
• capabilit y to pro duce sufficien t elec tr ic i t y to supply o v erall plan t needs or at least 100 M W e
• tec hnology that is a viable alternativ e to the curren t com mercial reactor fleet
• a go o d c hance at b eing appro v e d and built within th e next few d e cades
The high outlet temp erature is necessary for h ydrogen and biofuel pro duction and is the most limiting factor in c ho osing reactors. The goal of b eing capable of pro ducing at leas t 100 MW e w as necessary only to insure prop er siz i ng of the reactor. A reactor not capable of pro ducing this amoun t of electricit y w ould not b e able to supp ly the amoun t of pro c ess heat needed. A large reactor design w as c hose n , b ecause these are the designs most lik ely to b e ap pro v ed for constru c ti on in the up coming decades. Sho wing the profi tabilit y of using a n uclear plan t t hat is already b e i ng considered for construction for biofuel pro duction adds extra incen tiv e to construct suc h designs. A viable reactor has to b e b oth tec hni c all y feasible and stand a reasonable c han c e at b eing licensed and built. Extremely exotic core designs w ere th us excluded b ecause of the unlik elin e ss that suc h des i gns w ould b e built or license d in the near future. In the end, the design c hosen, a le ad - b is m uth eutectic (LBE) co oled fast reactor with a sup ercritical CO 2 (S-CO 2 ) secondary cycle, pro vides a reactor that has a viable c hance at b eing appro v ed and constructed in the coming decades and also pro vides a high temp erature alternativ e to curren t ligh t w ater tec hnology . The o v erall design, not just the reactor plan t, pro v es that reactors of this t yp e can w ork profitably with biofuels pro duction. F urthermore, other reactor t yp es with similar or higher ou tle t temp eratures can b e e x trap olated to b e profitable as w ell.
1.2 Design P arameters
When considering whic h design to pursue, man y factors w ere i m p ortan t. In or der to help sort through the differen t parameters the group used the Qualit y F unctional Deplo ymen t metho d (QF D) and in particular a house of qualit y . An explanation of this metho d and the house of qualit y constructed can b e see n in App endix X.
The parameters outlined in the follo wing sections w ere the primary differen tiating factors b et w ee n differ en t designs and w ere what guided the c hoice of reactors. They are also in the house of qualit y constructed for this dec i s ion . Other factors that ar e not explicitly listed h e re w ere either not of significan t concern or w ere not factors that di ff eren tiated one reactor from another.
1.2.1 Biofuels Co ordination
Certain re actor t yp es w ould b e n ot b e compatible with the h ydrogen and biofuel pro duction pro cesses describ ed later in the r e p ort. Thi s is b ecause if the outlet temp erature of the w orking co olan t w as to o lo w, it w ould b ecom e impractical to heat it up f or use. Ho w ev er, certain designs that ha v e garnered a lot of atten tion as p oten tial candidates to b e used in the com i ng decades to replace exis t ing reactors are liquid metal fast reactors, including so dium and lead co oled reactors. These reactors op erate at m uc h higher temp eratures than t ypical ligh t w ater reactors ( L WR) and, while falling short of th e optimal temp erature, still pro duce high enough temp eratures to b e compatible with the biofuel and h ydrogen pro cesses. F urthermore, it w ould b e b e n e fi c i al if the n uclear plan t to ok up as little ph ysical space as p ossible so as to in tegrate w ell with the other plan ts on site. These t w o design parameters are describ ed in more detail b elo w.
|
P o w er Densit y (MW/m 3 ) |
|||
|
System |
Core a v erage |
F uel Av erage |
F uel Maxim um |
|
HTGR |
8.4 |
44 |
125 |
|
PTGR |
4.0 |
54 |
104 |
|
CANDU |
12 |
110 |
190 |
|
BWR |
56 |
56 |
180 |
|
PWR |
95-105 |
95-105 |
190-210 |
|
LMFBR |
280 |
280 |
420 |
T able 1: P o w er densit y comparison for differen t reactor t yp es
1.2.1.1 Reactor Outlet T emp erature
The ma jor design par am eter considered w as reactor outlet temp erature. The tem p eratures that the reactor m ust supply to the h ydrogen and biofu e l plan ts quite substan tial ( > 700 C C) and the curren t reactor fleet (except for gas reactors) cannot ac hiev e those temp e r atures . High temp erature gas re actors pro vide optimal temp eratures, but as will b e seen later on, fall short in man y other categories and th us w ere not the de facto reactor of c hoice. A temp e ratu re close to the requiremen ts could suffice with extra heating coming from other means (either electrical or b y burning excess h ydrogen or biofuels). The final design c hosen, w as LBE co oled fast reac tor with a S-CO 2 secondary lo op. Th is design has an outlet temp erature of 650 C C.
1.2.1.2 F o otprin t of the Reactor
A design parameter that the group considered imp ort an t w as size of the o v erall reactor plan t. It w as felt that a smaller reactor plan t w ould mak e construction c heap er and w ould allo w more flexibilit y when c ho osing a lo cation for the plan t. With biofuels and h ydrogen pl an ts that w ould also b e presen t, ha ving a smaller plan t w ould b e useful so as to limit o v erall size of the facilit y . A small reactor plan t also allo ws for more flexibilit y in ho w th e reactor can b e used indep enden t of the h ydrogen and biofuels pro duction plan ts. It can b e used in lo cations wh e re space is limited suc h as ships, d e n s ely pac k ed nations, etc. The LBE co oled reactor h ad a higher p o w er densit y and w as ph ysically smaller com p ared to other plan ts [88] as sho wn in T able 1.
This w as aided b y a liquid metal co olan t (whic h allo ws for greater heat r e mo v al p er v olume and th us more p o w er pro duction p er v olume) and b y a sec on dary S-CO 2 cycle [54]. The secondary S-CO 2 using a Bra yton cycle offers a sm all e r p lan t than the t ypical Rankine steam cycle primarily b ecause of the decrease in size of the turbines [54].
1.2.2 Viabilit y to get Licensed and Built in the Up coming D ecades
As men tioned previously , the viab ilit y of the reactor is a critical goal to the pro ject. Liquid me tal fast reactors, b oth so dium and lead co oled, ha v e b een built and op erated in the past, for example at D unrea y in Scotland for so dium and on So viet submarines f or lead, and th us ha v e pro v en trac k re cord s when it comes to tec hnical viabilit y . Whi le some of those designs, notably the so dium co oled Sup erphenix, ha v e faced problems, exp erience with these reactors and decades of design optimization ha v e led to more stable and promising designs. Still, the question arose as to ho w difficult the licensing pro ces s w ould b e for designs that the Nuclear Regulatory Commission (NR C) is unfamiliar with. None of the reac tor s under consideration had m uc h in w a y of a rec en t predecessor that the NR C had dealt with. Therefore, the primary though t pro cess w as to examine eac h design from the viewp oin t of a regulator and see whic h reactor w ould ha v e th e easiest time getting through licensing. A design suc h as the molten salt reactor (MSR) w ould b e v ery difficult to get licensed (ev en though reactors ha v e b een built of this t yp e in the Uni te d States, alb eit a whil e ago [70]) b ecause of the uneasiness of molten fuel paired with a to xic co olan t. Similarly , a reactor suc h as the So dium F ast Reactor (SFR) w ould b e faced with a problem in regards to prev e n ting its co olan t coming in to con tact with an y w ater or air b ecause the co olan t then reacts violen tly . Lead reactors face problems in k eeping the co olan t from solidifying and also face to xicit y issues. The secondary lo op is one facet of the design that is c hosen where it adds to the difficult y in the plan t lice nsin g b ecause there is no exp erience with S-CO 2 in a n uclear r e actor in the US. Nev ertheless, CO 2 is v ery inert and, if released, w ould cause little issue.
1.2.2.1 Safet y
Some of the designs p ossess op e rat ing c haracteristics and materials that offe r significan t p erformance adv an tages at the cos t of safet y . F or the c h os en LBE-co oled design, one safet y concern w as that the bism uth coul d b e neutronically activ ated. This w as deemed to b e a minor concern for t w o reasons. First, prop er shielding could minimize w ork er dose due to bism uth activ ation . Second, the co olan t is made of lead and bism uth, b oth high-Z elemen ts, whic h r e sults in an e x c ellen t built-in gamma shield.
With an y fast reactor, a p ositiv e v oid co efficien t of reactivit y is a concern and th e design prop osed in this study is no differen t. One safet y adv an tage is that LBE tak es a considerable amoun t of heat to b oil and the mo deration pro vided b y the lead-bism uth is miniscule. Therefore, the lo w p ossibilit y of co olan t v oiding w as deemed a minor concern. F urthermore, a LBE co oled core can b e des i gne d for natural con v ection circulation co oling whic h is an excellen t passiv e safet y feature that s h o u ld aid in loss of off-site p o w er acciden ts that ha v e recen tly (b ecause of the F ukushima acciden ts) b een fo cused on. The final design utilized this safet y adv an tage.
1.2.2.2 Sim plicit y of Design
Minimizing the amoun t of structures, systems, and comp onen ts (SSCs) is b eneficial in reducing construction and capital costs. A simple design allo ws for simple op eration, less parts to main tain, and less opp ortunit y to o v erlo ok an issue. Therefore reactors designs w ere also v etted based on ho w complex the design w ould b e. F or example, ha ving an in termediary lo op or an online repro cess i ng plan t added to the complexit y of the reactor. The range of designs considered w en t from simple single fluid des i g n s suc h as Su p ercritical W ater Reactors (SCWR) to SFRs con taining three in terfacing co olan t systems and MSRs con taining a on - l ine repro cessing plan t. The final LBE co oled design utilized a primary co olan t system of LBE and a secondary co olan t syste m of S-CO 2 , similar to curren t pressurized w ater r e actors (PWRs) in the sense that b oth use a primary and s econdary co olan t.
1.2.2.3 Materi al Concerns
In an y reactor, material stabilit y and c or ros i on resistance are imp ortan t considerations. In the high op erating temp eratures that ar e b eing emplo y ed in this plan t design, the abilit y of the materials to b e able to withstand suc h an en vironmen t b ecame a s i gnifican t conce r n. In fact, all high temp erature designs considered faced material corrosion concerns. The final LBE co oled design also faced corrosion issues and w as in the end limited b y creep lifetime. Curren t optimization of t his design includes searc hing for a material capable of handling th e high temp erature en vironmen t for longer p erio ds of time. Curr e n tly , there is a large push in researc h to find materials capable of handling these en vironmen ts whic h b o des w ell for high temp e r a t ure reactors.
1.3 Design Options and Ev aluation
1.3.1 Liquid Metal F ast Reactors
Liquid metal reactors, including the SFR and the Lead F ast Reactor (LFR) are among the mos t viable high temp erature designs. Most curren t SFR and LFR designs ha v e outlet temp eratures of ab out 550 C C [150] though these designs h a v e the capabilit y to go to higher temp eratures. F rom the l ite r ature, it is seems that an outlet temp erature of 550 C C w as c hosen to matc h reac tors to readily a v ailable turbines with w ell kno wn p erformance histories [150]. Giv en that, this sugges ted higher temp eratures ma y b e p ossible, further researc h w as done on the fundamen tal material concerns with SFR or LFR op eration and ho w these migh t con tribute to a hard maxim um temp erature. It w as found that b oth so dium and lead cores and c o olan t could op erate at higher temp eratures, but curren t cladding is the limiting factor. Cladding b egins to creep and ev en tually fails at temp eratures around 650 C C [150]. Because of th is limit, the a v erage temp erature of the cladding is k ept b elo w 550 C C. Materials researc h to push this b ou ndary se ems most dev elop ed for LFR materials, with the US Na vy in par tic u lar b elieving that outlet temp eratures could approac h 800 C C within the next 30 y ears [57].
The feasibilit y of LFRs op erating at these temp eratures is supp orted b y th e significan tly higher b oiling p oin t of lead (1749 C C as opp osed to 883 C C f or so dium) whic h w ould protect against v oidin g and th u s ensure
sufficien t co oling of the fuel in the ev en t of acciden t scenarios and protecting against other dangerou s tran sien ts. Ha ving a small, safe, and economically comp etitiv e reactor with a suitable outl e t temp erature w ould b e most app ealing to an y p oten tial clien ts. The LFR meets these requiremen ts w ell and aligns with these QFD criteria, see App endix X. The p ossible e xception to this ma y b e the LFR’s economic comp etitiv eness, giv en the high costs of building an y new reactor design, although burning actin ides or ha ving a p ositiv e breeding ratio ma y c hange this. These t w o pro cesses w ould allo w the reactor to b e fueled using relativ ely lo w uranium enric hmen t (with as lo w as 11% enric hmen t b eing rep orted for U-Zr fueled SFRs [97]) th us allo wing fuel to b e bred in th e reactor during op erati on and reducing costs. This ma y ev en tually lead to the p ossibilit y of fast reactors, including LFRs, b eing deplo y ed with a once-through fuel cycle and ha ving economics near that of curren t once-through L WR reactors [97]. LFR reactors can ha v e small fo otprin ts due to the densit y and thermal conductivit y of the le ad co olan t, whic h will b e useful in this pro ject giv en that the h ydrogen and biofuel plan ts ma y b e v ery large, reducing the impact of the facilit y as a whole as w ell as making the reactor m or e secure. LB E , with a melting p oin t of 123.5 C C, w as c hosen as the final co olan t c hoice b e cause it offe r s a lo w er melting p oin t than lead and generally do es not expand or con tract m uc h with co oling or heating [3] . Ho w ev er, it can pro du c e a dangerous radioisotop e of p olonium [3], and th us additional op erational handling pro cedures w ould need to b e follo w ed to ensure the safet y of the w ork ers.
1.3.2 Molten Salt Reactors
{ }
The next des ign in v estigated w as the MSR. This reactor design is capable of the requisite temp erature and has additional features that are quite attractiv e for the plan t needs. One of these features is the abilit y to scale the reactor to an y size, allo wing flexibilit y during future design iterations [126]. Using b oth a graphite mo derator and a molten-salt co olan t, this reactor also do es not require pressurization th us eliminatin g equipmen t costs and pro vidin g added safet y . In the case of an acciden t where the core o v erheats, a safet y plug w ould drain the molten core in to a sub - criti c al geometry , ending threat of a criticalit y inciden t b efore an acciden t escalates. As a fuel, the MSR w ould need to use medium-enric hed uranium, but w as also compatible with a thorium fuel cycle. This w oul d reduce fuel c osts substan tially b ecause of the great abundan c e of thorium as compared to urani um[74]. Some disadv an tages of the MSR include the presence of hazardous materials in the reac t o r v essel and salt. A salt b eing considered for extensiv e use is FLiBe, whic h con tains b eryllium and p oses a risk to w ork e rs on the premises. A second concern is the pro duction of h ydrogen flu oride (HF) in the core, whic h is lethal if it comes in con tact with h uman tissue T o xicSubstances2003 . Additional ly the h ydrogen b onding to the fluoride in the HF is tritium, a dangerous radioactiv e isotop e. Dep endi ng on the fuel c hosen, it could add proliferation concerns due to the repro cessing of the fuel on-site. A thorium fuel cycle w ould eliminate the p oten tial pro duction of plutonium and is adv an tageous in that resp ect. Though the MSR had some p oten tial for the pro ject, it w as not as practical or effec tiv e as s ome of the other designs considered.
1.3.3 Gas Co oled Reactors
The Adv anced Gas Reactor (A GR), High T emp e rat ure Gas Reactor (HTGR), P ebble Bed Mo dular Reactor (PBMR), and V ery High T emp erature Reactor (VHTR) gas-co oled reactors op erate with a th e rmal neutron sp ectrum and are mo derated with graphite. The HTGR, PBMR, and VHTR all use helium as their primary co olan t and the A GR design u s es CO 2 . The designs do differ in their op erating temp eratures, ho w ev er, all of these designs op erate at significan tly higher te mp eratures than tr aditional PWR or b oiling w ater reactors (BWRs). The PBMR sp ec ifi e s that the fuel is con tained with in small, tennis-ball-sized spheres, whic h is one of the fuel options for the HTGR or the VHTR. Besides these “p ebbles,” the oth e r fuel option for these reactor t yp es are small cylindrical c ompacts, whic h can then b e compiled to form more a prismatic reactor core [62]. F or either the p ebble or the compact fuel design, the fuel starts as tin y particles of uranium di o xide (UO 2 ) or uranium carb ide (UC) and is then coated with la y ers of carb on and silicon carbide (SiC). These la y ers for m the tri-isotropic (TRISO) fuel particles that can b e assem bled in to either the p ebbl e or compact form for the fuel.
The TRISO fuel particles pro vide these gas r e actors with sev eral passiv e safet y features. In the ev en t that there w as a loss of co olan t acciden t, the refractory m aterials of the coating (the carb on and the silicon carbide) w ould b e able to con tain the radioactiv e material. Additionally , in a loss of flo w acciden t (LO F A) or a l o ss of co olan t acciden t (LOCA), the large amoun t of graphite in the core w ould b e able to absorb a significan t amoun t of heat and a v oid melting of the reactor fuel [62, 88]. The high temp eratures of the VHTR
in particular allo w for high efficiency and has b een noted for its com p atibilit y for h ydrogen pro d uction, the end goal of the pr o cess heat for our o v erall design. While the gas reactors had ideal temp eratures for our design, other concerns, notably the size and economics/viabilit y of one b eing built prev en ted these reactors from b eing c hosen.
1.3.4 Sup ercritical C o olan t Reactors
{ }
The SCWR op erates at high pressure and temp erature (25 MP a and 550 C C) an d use w ater in sup e r c ri tic al state (neither a liquid nor a gas, but with pr op erties of b oth) as the co olan t [49]. Suc h reactors could op erate on either a fast or thermal neutron sp ectrum. A SCWR resem bles a BWR in des ign but differs in the conditions it op erates under. Sup ercritical w at e r has excellen t heat transfer pr op erties compared to normal w ater and th us le ss co olan t flo w is needed for the core than the BWR. There also exists a lar ge temp erature c hange across the core (greater than 200 C C). Due to the high temp eratures and use of the Bra yton cycle, this plan t has excellen t th e r m o dynamic efficiencies on the order of ˜45% [49]. The design for a t ypical SCWR is simple, m uc h lik e BWRs, with one lo op (although a SCWR of equal p o w er w ould b e smaller than a BWR). SCWRs ha v e no risk of b oilin g and can ac hiev e high electric p o w er lev els (1700 MW e). F urthermore, there is a significan t amoun t of exp erience with sup ercritical w ater in fossil fuel plan ts Alstom , but not in a neutron irradiation en vironmen t. Thes e conditions along with the high te mp eratures also giv e rise to significan t materials concerns and there is some difficult y ac hieving a negativ e v oid co efficien t, particularl y with fast designs.
Sup ercritical D 2 O (S-D 2 O) Reactors are under consideration in Canada [44] at the mom en t. They can use thorium as a fuel in a fast r e actor, otherwise, there is limited difference b et w een these and SCWR. Sup ercritical CO 2 reactors use S-CO 2 as the pr imary co olan t. S-CO 2 has a n um b er of adv an tages o v er sup ercritical w ater [120] including:
• It requires a lo w er pressure (7.4 MP a vs. 22.1 MP a) to k eep critical, and th us op erates at a lo w er pressure (20 MP a vs. 25 MP a)
• It i s b etter suited for fast reactors b ecause of th e absence of the strong h ydrogen mo derator that w ater has
• It has the adv an tage of ha ving a turbine rated at 250 MW e that has a diameter of 1.2 m and a length of 0.55 m (sup ercritical w ater uses turbines sized similarly to curren t L WR tec hnology)
F or all these reasons, S-CO 2 is attractiv e for a consolidated reactor plan t. Muc h researc h has b een done on its use as a secondary co olan t in an indirect cycle with lead or molten salt and the decision w as made to use this as our secondary co olan t for a n um b er of reasons including:
• Small fo otprin t
• Excellen t h e at transfer prop erties
• Excellen t th e r m o dynamic efficiency
• Lo w er required pressures than sup ercritical w ater
1.3.5 T able of Design Comparison
|
SFR |
LFR |
VHTR |
MSR |
SCWR |
|
|
Neutron Sp ectrum |
F ast |
F ast |
Thermal |
F ast or thermal |
F ast or thermal |
|
Outlet T emp. ( ' C) Co olan t Mo derator |
530 to 550 Na None |
near-term: 550-650 |
> 1000 Helium (CO 2 ) Graphite |
650 FLiBe Graphite |
550 |
|
long-term: 750-800 |
|||||
|
Pb or LBE None |
S - H 2 O CO 2 , D 2 O |
||||
|
Same as |
|||||
|
co ola n t |
|||||
|
Relativ e P o w er |
High |
High |
Medium to High |
High |
High |
|
P o w er Densit y |
High |
High |
Lo w |
N/A |
Lo w to Med . |
|
F easibilit y |
Med. |
Med. |
Lo w to Med. |
Lo w to Med. |
Med. |
|
On-line Refueling |
No |
No |
Y es |
Y es |
No |
|
F uel Enric hmen t |
MEU to HEU |
MEU |
LEU to MEU |
LEU |
LEU or nat U |
|
Pressure |
˜ 1 atm |
˜ 1 atm |
dep en ds |
˜1 atm |
20-25 MP a |
|
Price |
Ab o v e Avg. |
Ab o v e Avg. |
Ab o v e Avg. |
High |
Belo w a vg. |
|
Ph ysical Size |
Belo w Avg. |
Small |
V ery Large |
Large molten fuel |
Av erage |
|
Materials Concerns |
neutron activ ation |
need to melt c o olan t |
high te mp concerns |
sup ercritical fluid |
|
|
Na reactiv e w/air & w ater |
P o creation with LBE |
molten salt can b e corrosiv e |
FLiBe dange rs (Be/tritium/HF) |
||
|
Other Notes |
Actinide managemen t |
Can’t v oid co olan t |
Co olan t remains single pha se |
on-line repro cessing |
similar to curren t BWR designs |
|
LBE has go o d thermal prop erties |
|||||
|
Using He reduces corrosion |
|||||
|
Pb/LBE are not reactiv e w/w ater |
T able 2: Reac tor Design Comparison[88, 139, 62, 49, 120, 44, 98, 126, 136]
2 Pro cess Heat
2.1 Goals of the Pro cess Heat Design Group
A lead-bism uth co oled reactor w i th a secondary lo op of S-CO 2 has b een c hosen as the heart source for this facilit y that will couple the reactor with a h ydrogen and biofuels plan t. The heat pro vided b y the reac tor will b e transp orted to the h ydrogen an d biofuels facilities with minim um temp erature and pressure losses. The pro cess heat system will consist of high temp erature heat exc hangers (HX) in the s econdary lo op , heat exc hangers at the biofuels and h ydrogen plan ts, a heat storage system and piping connecting all these comp onen ts. V arious heat sink options are also under consideration. The goal of the pro cess heat sub group w as to determine an optimal la y out design from the v arious t yp es of tec h nology a v ailable to fulfill the three main tasks of exc hanging, transp orting, and storing heat. Once this has b een complete d , the ob jectiv e is to size and mo del these comp onen ts based on the op erating conditions and heat requiremen ts of th e reactor, biofuels, and h ydrogen plan ts.
2.2 Design Challenges
Ov erarc hing pro cess heat issues include w orking with large temp erature gradien ts, minimizing heat and pressure losses, and c ho osing robu s t comp onen ts. A ma jor design c hallenge is the fact that the core is outputting a temp erature on the order of 650 Q C whic h is significan tly lo w er than the input temp erature needed to p o w er some of the h ydrogen pro duction pr o cesses . In addition, it w as determined that a heat storage device w ould b e implemen te d in order to h e at the lead-bism uth co olan t to a p oin t just ab o v e its melting temp erature during the ev en t of a sh utdo wn.
It is imp erativ e that the materials c hosen for fabricating heat exc hangers for the p ro cess heat system are able to withstand op erating c on ditions of high temp eratures (up to 900 Q C) and pressures (up to 5 MP a). Susceptibilit y of candidate materials to stress corrosion crac king under constan t load as w ell as slo w-strain rate conditions, fracture toughness, and crac k gro wth b eha vior ha v e b ee n studied extensiv ely and literature indicates that All o y 230 and Allo y 617 are suitabl e for fabricating high temp erature heat exc hangers [72]. The op erating conditions for the heat exc hanger at the h ydrogen pl a n t will b e more sev ere due to tem p eratures higher than the core outlet temp erature. Stud ies indicate that Allo y C-22 and Allo y C-276, b ecause of their high tensile strength and ductilit y un til fracture, ar e sui table heat exc hanger materials for op eration in or near acidic en vironmen ts [127].
2.3 P ossible Heat Exc hanger Designs
The applicabilit y of heat exc hanger (HX) designs to this system w as ev aluated based on the f e asib ilit y of the heat exc han ge r tec hnology as w ell op erating temp eratures and pressures. E ff ectiv eness, size, heat transfer area p er unit v olume, w orking fluid options, heat los ses and pressure drops for the v arious designs w ere also primary considerations. Ho w ev er, sev eral of these considerations ha v e conflicting implications for HX de sign . F or example, compact heat exc hangers (suc h as the PCHE discussed b elo w) ha v e high eff ectiv enesses and heat transfer co effi cien ts but incur larger pressure losses.
The principal f e atu res of fiv e HX designs - shell and straigh t tub e, shell and helical tub e, plate, and prin ted circuit and ceramic heat exc hangers- are listed in T able 3.
|
HX T yp e |
Compactness ( m 2 / m 3 ) |
T emp. Range ( Q C) |
Max P (MP a) |
Multi- stream |
Multi pass |
Cleaning Metho d |
|
Straigh t Shell-and-T ub e |
˜100 |
˜+900 |
˜30 |
No |
Y es |
Mec h, Chem |
|
Helical Shell-and-T ub e |
˜200 |
˜600 |
2.5 |
No |
No |
Mec h |
|
Plate |
˜200 |
-35 to ˜+900 |
˜60 |
Y es |
Y es |
Mec h, Chem |
|
Prin ted Circuit |
2000 to 5000 |
-200 to ˜+900 |
˜60 |
Y es |
Y es |
Chemical |
|
Ceramic Heat Exc hangers |
N/A |
1200 |
N/A |
y es |
y es |
Mec h,Chem |
T able 3: Principal F eatures of Heat Exc han ge r s (adapted from [133, 103])
Courtesy of Elsevier, Inc., http://www.sciencedirect.com . Used with permission.
Figure 1: Com b ined Multiple Shell-P ass Shell-and-T ub e Heat Exc hanger (CMSP-STHX) with con tin uous helical baffles [158]
2.3.1 Straigh t Shell-and-T ub e
Shell-and-tub e HXs find extensiv e application in n uclear plan ts and pro c ess heat systems. T h e se HXs can b e designed to b e robust and suitable for s p ecial op e r ating conditions suc h as a radioactiv e en vironmen t. They can b e fabricated using Hastello y , Incolo y or graphi te an d p olymers [102]. The design can b e adapted to include fins if one of the w orking fluids is a gas as the heat exc hanger allo ws liqui d/liquid, gas/liquid, and t w o phase systems. These heat exc hangers are v ery large due to lo w heat transfer area p er unit v olume (˜100 m 2 / m 3 ) but allo w high op erating temp eratures (up to 900 Q C) and pressures (up to 30 MP a) [133]).
2.3.2 Mo dified Shell-and-T ub e
Helically baffled Straigh t shell-and-tub e heat exc hangers can b e mo dified b y in tro ducing m ultiple passes, or path w a ys for the liquid as w ell as b y adding baffles. Baffles are flo w-directing panels within the tub e that increase the efficiency of heat transfer. T ypical segmen tal b affl es are p erp endicular to the tub e and ca n result in dead zones with lo w er lo cal heat transfer. Helically baffled shell-and-tub e heat exc hangers, due to their geometry , are able to induce tu rbulen t flo w and therefore increase heat transfer r ate s.This turbulence com bined with the high shear stress mak es the helical shell-and-tub e exc hanger less lik ely to exp erience disruptiv e fouling. A helically baffled shell and tub e HX is sho wn in sho wn in Figure 1.
Another adv an tage of the helical baffle is that it reduces vibration s within the heat exc hanger (b ecause the fluid is crossing the tub e bund le at an angle instead of v ertically) an d therefore increases th e h e at exc hanger’s ph ysical stabilit y . Helical heat exc hangers can op erate at large p re ssures and are exp ected t o withstand a maxim um pressure of 70 MP a [95]. A recen tly published pap er [158] n umerically assesses the b enefits of a com bined m ultiple shell-pass shell-and-tub e heat exc hanger with con tin uous helical baffles (CMSP-STHX) as sho wn in Figure 1. In their reference system (and in the ma jorit y of systems to date) the inner shell pass uses standard segmen tal baffles while the outer shell pass uses con tin uous helical b affl es. This is d ue to the difficulties in man ufacturing helical baffles in the smaller cen tral shell. Despite this, it w a s found that with the same mass flo w rate and heat transfer as a con v en tional shell-and-tub e heat exc hanger with segme n tal baffles (SG-STHX), the helical v ersion exp erienced a p re ssure drop 13% lo w er than that of the standard heat exc hanger. If the pressure drop w as held constan t, the helical HX w as capable of a 6.6% higher mass flo w
rate and a 5.6% higher heat transfer rate than the standard heat exc hanger [158].
Helical tub e The shell and helical tub e HX is a v ariation on the shell and straigh t tub e de sign and consists of tub es spirally w ound and fitted in a shell. Spiral tub e geometry pro vides a h igher heat transfer area p er unit v olume (200 m 2 / m 3 compared to 100 m 2 / m 3 for straigh t shell-and-tub e t yp e HXs). This d e sign has b een pro v en b y its use in the High T em p erature Engineering T est Reac tor (HTTR) [64]. H elical t yp e HXs are w ell suited to gas/liquid systems. A disadv an tage of this design is the difficult y in main tenance of the helical coils [133, 129].These heat exc hangers find extensiv e application in n uclear plan ts and also as pro cess heat systems and can b e designed to b e v ery robust and suitab le for sp ecial op erating conditions suc h as a radioactiv e en vironmen t. They can b e fabricated using Hastello y , Incolo y , graphite and p olymers [102]. The design can b e adapted to include fins if one of the w orking fluids is gaseous and the heat exc hanger allo ws li quid/liquid, gaseous/liquid as w ell as t w o phase systems. These heat exc hangers ar e v ery large due to lo w heat transfer area p er unit v olume (˜100 m 2 /m 3 ) but allo w high op erating temp eratures (up to 900 Q C) and pressures (up to 30 MP a).
2.3.3 Plate
In a plate t yp e heat exc hanger, the heat transfer o ccurs through p lanar surf ac es whic h allo ws coun ter, cross, and parallel flo w configurations [135] and can b e fabricated from Hastello y and nic k el allo ys. T h e se heat exc hangers, ho w ev er, allo w b oth m ulti-pass and m ulti-stream capabilities and greatest ease of cleaning and main tenance as compared to the other designs review ed for this pro ject. There are sev eral v ariation s on plate t yp e designs, ho w e v er, the Ba v ex plate HX pro vides the highest op erating temp eratures (up to 900 Q C) [133, 129].
2.3.4 Prin ted C i rcuit
PCHEs (an example of whic h is seen in Figure 2) can op erate under high temp erature (˜900 Q C) and high pres sure (˜60MP a) conditions. They are t ypically used in p etro c hemical, refining, and upstream h ydropro cess in g industries. PCHEs c an incorp orate m ultiple pro cess streams in to a single unit and ha v e lo w mass/dut y ratios of ˜0.2 t/MW [16]. They are suitable for corrosiv e en vironmen ts and ha v e an effectiv enes s of up to 98%. In a PCHE, the fluid flo w c hannels, whic h are of the order of sev eral millimeters, are c hemically etc hed and the flo w can b e parallel, cross, coun ter flo w, or a com bination of all three. Also, the absence of gask et and braze material lo w ers the probabilit y of leak age [99]. Ho w ev e r , there is p oten tial for thermal stresses in the axial direction when there ar e sharp temp erature v ariations. This design also suffers from lo w capacit y fac t o r s due to the nee d for offline insp ection and repairs [111]. F urthermore, small flo w c hannels coul d result in fouling problems whic h w ould require offline repairs using c hemical metho ds [133]. Ho w ev e r, redundan t mo dules ma y b e installed to impro v e capacit y factors of the p ro cess heat system d uring main tenance and r e p airs. PCHEs ha v e not b een used previously for n uclear ap plications, but are u nder reviews as p oten tial HXs for the Next Generation Nuclear Plan t [124].
2.3.5 Ceramic
A m ulti-stream heat exc hanger capable of op erating at high temp eratures in the pres ence of b oth reducing and o xidizing fluid streams is required at the Biofuels plan t. This HX will b e required to dra w heat from w aste h ydrogen and o xyge n fluid streams to pro duce steam. Ceramic heat exc hangers fabricated from reaction b onded silicon carbide (RBSiC) or siliconized silicon carbide (SiSiC) are b est suited to this application. Both ceramic materials ha v e lo w er thermal conductivities than Ni-Cr allo ys use d for fabr ic ati ng HXs discussed earlier in this section. Ho w ev er, unlik e metal allo ys b oth Rb SiC and SiSiC ha v e demonstrated resistance to o xidativ e and reductiv e en vironmen ts at temp eratures up to 1200 Q C [103]. Multi-stream cross-flo w heat exc hangers can b e fabricated using either ceramic material.
Source: Li, Xiqing., et al. "Alloy 617 for the High Temperature Diffusion-Bonded Compact Heat Exchangers." Published in ICAPP 2008 , Anaheim, CA, June 8-12, 2008. © American Nuclear Society and the authors. All rights reserved. This content is excluded from our Creative Commons license. For more information, see http://ocw.mit.edu/fairuse .
Figure 2: A PCHE heat exc han ge r made of Allo y 617 with straigh t c hannels [160]. The semi-circular fluid c hannels ha v e a diameter for 2mm.
3 Hydrogen
3.1 Goals of Hydrogen
The h ydrogen pro duction plan t w as implemen ted in this design to partly satisfy the o v erall goal of designin g a n u c l e ar system that can pro duce at least 100 MW e and pro duces h ydrogen and biofuels. Before the c hoice of the h ydrogen pro duction metho d could b e made, the purp ose of the h ydrogen p ro duction plan t had to b e determined. The plan t w ould either pro vide h ydrogen for use in a large scale h ydrogen econom y or only w ould pro vide the necessary h yd roge n input for the biofuel pro duction plan t. Hydrogen is curren tly used in the p e trol e u m and c hemical industries; ho w ev er, a large-scale h ydrogen infrastructure in v olv es h ydrogen b ecoming the main source of energy for transp ortation in an effort to off set oil consumption and curb greenhouse gas emissions. Ho w ev er, researc h has sho wn that the c hemical prop erties and costs of building a h ydrogen infrastructure curren tly mak e a h ydrogen econom y unfeasible [37, 66]. Con trarily , biofuels can b e distributed using presen tly a v ailable distribution infrastructures. Also, biofuels can b e used while blended with tr aditional gasoline i n sligh tly mo d ified in ternal com bustion e n gines , or in new e ngi nes that can run solely on biofuels [29]. In an effort to immediately impact oil consumption and reduce gr e enhou s e gas emissions in an economical fashion, the purp ose of the h ydrogen plan t w as deemed as supplying only the amoun t of h ydrogen required for the biofuel pro duction facilit y .
With the purp ose of the h ydrogen pro duction plan t determined, the main des i gn considerati o n s used to compare differen t h yd rogen pro duction metho ds are as follo ws. The biofuel pro duction plan t requir e s
7.9. kg/s (682,560 kg/da y) of h ydrogen to ac hiev e the desired biofuel pro duction rate, and th us 7.9. kg/s w as the h ydrogen pro duction quota imp osed on this design. In an effort to minimize greenhouse gas emissions in this n uclear reactor system o v erall, a h ydrogen pro duction pro cess that pro duces no net gr e enhou s e gases w as desired. Al s o, the ob jectiv e to minimize p o w er consumption w as used to guide design c hoices, esp ecially to minimize steady-state electrical p o w er consumption since it is less energy efficien t than using solely thermal p o w er while considering the thermo dynamic efficiency of con v erting heat in to electricit y of ˜45%. The lead-bism uth n uclear reactor exit temp eratures are ˜600C, and th us h ydrogen pro duction m etho ds that are able to op erate closest to this tem p erature are preferable. This design consideration is motiv ated b y the previously stated ob jectiv e to minimize steady-state electrical p o w er consumpti on. If maxim um temp erature requiremen ts are higher than the a v ailable thermal temp eratures of appro ximately 600 C, then add itional electrical p o w er will b e required to raise the temp erature of reactan ts so c hemical pro cesses pro ceed prop erly . Also, maxim um temp erature requiremen ts coupled with the t yp e of r e actan ts in v olv ed i n eac h h ydrogen pro duction metho d w ere considered, motiv ated b y the tendency for greater material degradation concerns in corrosiv e, high-temp erature en vironmen ts. If these material concerns w ere not sufficien tly mitigated, they w ould jeopar dize the reliabilit y and longevit y of the h ydrogen pro duction plan t. The requiremen tto replace comp onen ts frequen tly w ould negativ ely effect b oth the economics of the h ydrogen plan t and also jeopardize consisten t biofuel pro duction capabilities. Lastly , th e commercial viabilit y of eac h h ydrogen pro duction approac h w as considered, suc h that scalabilit y to h ydrogen pro duction quotas w as conceiv able and that the scaled p o w er requiremen ts from p revious studies w ere fa v orable relativ e to the total thermal and electrical output of the lead-bism uth n uclear reactor used in this design. All of these design ob jectiv es and considerations w ere used to guide the design c hoices of c h o osing a h ydrogen pro duction metho d to reac h a 7.9. kg/s h ydrogen pro duction rate required b y the biofuel pro duction plan t.
3.2 Design Options and Ev aluation
A total of eigh t options w e re explored, b oth thermo-c hemical and electro c hemical, to iden tify the op ti mal pr o cess for h ydrogen pro duction. F our ma j or h ydrogen pro duction metho ds w ere in v estigate d : w ater electrolysis, high-temp erature steam electrolysis, thermo c hemical w ater splitting, and bacterial h ydrogen pro duction. Other h ydrogen pro duction metho ds using natu ral gas es w ere quic kly rejected due to the de sign ob jec tiv e to use a h ydrogen pro d uction metho d that partly fulfills the o v erall design goal of minimzing greenhouse gas emissions from this n uclear system design. Material concerns dominate the high te mp erature steam electrolysis and thermo c hemical w ater splitting due to relativ ely high temp eratures (500-900 C C) and corrosiv e reactan ts and pro du c ts. Ho w ev er, the w ater electrolysis and bacterial h ydrogen pro duction pro cess are dominated b y commercial viabil it y concerns.
Figure 3: Steam methane re f orming blo c k diagram [48]
3.2.1 Steam-Methane Reforming
Steam reforming of natural gas (as sho wn in Figure 3) i s one of the most widely used in industry to da y for c hemical man ufacturing and p etroleum refining. Steam reforming first con v erts me th ane in to h ydrogen and carb on mono xide b y a reaction with steam o v er a nic k el catalyst. In a second step kno wn as a w ater gas shift reaction, the carb on mono xide from the first reaction is reacted with steam to form h ydrogen and carb on dio xide [125]. The c hem ical reactions can b e seen in Equations 1 and 2.
C H 4 + H 2 O → C O + 3 H 2 (1)
C O + H 2 O → C O 2 + H 2 (2)
These reactions o cc u r at 750 C C for Equation 1, whic h is a little high but within an acc eptab le range, and 350 C C and b elo w for Equ ation 2, whic h is w ell within th e acc eptabl e range. Steam reforming is ab out 70% efficien t, whic h is an adv an tage [89]. Ho w e v er, despite the efficiency , input temp erature, and pro v en feasibilit y of the te c hnology , steam methane failed the most critical parameter, whic h w as to b e carb on emission-free, b ecause it pro duces C O 2 . F or this reason, SMR w as one of the first pro duction options to b e discarded.
3.2.2 W ater Electrolysis
Electrolysis pro duces h ydrogen b y passing elec tr ic it y through t w o electro des in w ate r . This causes a disso ciation of H 2 O as H + and O - tra v el to the catho de and ano de resp ectiv e ly to form H 2 and O 2 . Tw o distinct categories o f electrolysis units are presen tly used for industrial pro duction of h ydrogen: alk aline electrolyzers and solid p olymer electrolyte (SPE) elec trol yz ers. The alk ali ne electrolyzers feature an aqueous solution of p otassium h yd ro xide used for its high conductivit y and resulting in a faster d is so ciation of w ater. F or the SPE, the electrolyte is a solid ion conducting m em brane that allo ws the H + ion to transfer from the ano de side of the m em brane to the catho de side, where it forms h yd roge n . Figure 4 sho ws a st yli z ed depiction of the electrolysis pro c ess.
Figure 4: Electrolysis pro cess blo c k diagr am
Both v ariations of w ater electrolysis f e atu re lo w op erating t e mp eratures at atmospheric press ur e as w ell as a reas onab le amoun t of energy re q uired to c harge the an o de and catho de. Alk aline electrolyzers require 100-150 C C while SPE electrolyzers require 80-100 C C. Both metho ds of w ater electrolysis are do cumen ted to require 1280-2133 MW electrical p o w er requiremen t to obtain the desired h ydrogen pro duction rate of 682,560 kg/da y . Though this electrical p o w er requiremen t is not inconceiv able, a h ydrogen p ro duction pro cess that requires a lo w er electrical p o w er requi re men t w ould b e preferred. The efficiency of the ES pro cess alone is 75%, ho w ev er, coupling this pro cess t o the 45% e fficiency of n uclear energy results in an o v erall energy efficiency of 25-45%[6]. F urthermore, the h ydrogen pro duction rate of w ater ES is far lo w er than what is required b y the biofuel p ro cess. The most p o w erfu l electrolyzer can pro duce just 1,048 kg/da y of h ydrogen, th us requiring 651 electrolyzers running in parallel to reac h the pro duction quota of 682,560 kg/da y . Though this is not an imp ossible task, the prosp ect of scaling up w ater electrolysis to ac hiev e the pro du c tion rate is not desirable if another solution alr e ady capable of p ro ducing t he necessary amoun t of h ydrogen is a v ailable while using less electricit y than this pro cess.
3.2.3 W estinghouse Sulfur Pro cess
The “h ybrid” W estinghouse Sulfur Pro cess (WSP) electrolyzer fe at ures four distinct steps of whic h one is thermo c hemical and one is electro c hemical. Oxygen is generated from H 2 SO 4 in a thermo-c h e mical reaction requiring 800 C C. The resulting SO 2 , H 2 O, and O 2 pass through the sulfuric acid v ap ori z er, and on to the o xygen reco v ery step where O 2 is remo v ed. SO 2 and H 2 O con tin ue on to the electro c hemical h ydrogen generation, where a bias of -1.5 V pro duces H 2 SO 4 and H 2 . T h e h ydrogen is remo v ed from the system, H 2 SO 4 passes bac k through the sulfu ric acid v ap orizer, and the pro cess b egins again at the thermo c hemical o xygen generation step [4]. This is sho wn as a basic blo c k diagram p ro cess in Figure 5.
Courtesy of Edward J. Lahoda. Used with permission.
Figure 5: The W estinghouse Sulfur Pro cess for h ydrogen pro duction [4] .
Though the pro cess lo oks elegan t, it w as one of the first designs d ism issed primarily due to the sulfuric acid that w ould presen t and w ould in tro duce significan t material corrosion concerns to the plan t. In addition to the op eratin g temp erature of 800 C required b y this pro cess in a highly acidic en vironmen t, th e coincidence of these t w o p roblems resulted in the rejection of the WSP as as a metho d of h ydrogen pro duction.
3.2.4 Hydrogen from Urine
Urea from natural h uman w aste con tains h ydrogen. Most imp ortan tly , these h ydrogen atoms are only w eakly b onded to the res t of the molecule and can b e easily rem o v ed through the use of an inexp ensiv e nic k el catalyst [36]. The basic pro cess is sho wn in Fi gure 6.
Bryan K. Boggs, Rebecca L. King, and Gerardine G. Botte. Chem. Commun . , 2009, 4859-4861. Reproduced b y permission of The Royal Society of Chemistry.
Figure 6: Sc hematic represen tation of the direct urea-to-h ydrogen pro cess [ 3 6] .
Ho w e v er, the amoun t of urine needed is proh ibitiv e. It w ould tak e man y gallons p er hou r to ac hiev e the r ate needed so transp orting and storing the urine w ould b e come problematic. F urthermore, o v er time urea h ydrolyzes in to ammonia whic h is not as easy to use. Th us, the pro cess w ould also ha v e to w ork quic kly . While the cost for the actual pro cess, the storage and transp ort considerations for this h ydrogen pro duction pro cess b egin to b ecome significan t factors in o v erall costs at large v olumes. Long-term storage lac ks feasibilit y b ecause of the h ydrolyzation problem and so the id e a w as de cid e d not to b e adv an tageous.
3.2.5 Hydrogen from Bacteria
There are four main w a ys to pro duce h ydrogen through biological pro cesse s: biophotolysis of w ater using al gae and c y anobacteria, ph oto- d e comp osition of organic comp ounds b y photosyn thetic bacteria, fermen tativ e h ydrogen pro duction from organic comp oun ds , and h yb rid sys t e ms using photosyn thetic and fermen tativ e bacteria [50].
Dark fermen tation h ydrogen pro duction w as determined to b e the most commercial viable bacterial metho d. Dark fermen tation is essen tially the s ame pro cess as algal biomass pro duction in large op en p onds follo w ed b y a dark fermen tativ e stage. This metho d w ould yield close to 12 mol of H 2 for eac h molecule of glucose metab olized [69]. Ho w ev er, the high cost for the v ery large v olume of ra w material requi red coupled with the risk of system failure from biological con tamination led to the rejection of this metho d [69, 85].
3.2.6 High T emp erature Steam Electrolysis
One of the more promising tec hnologies is High T emp erature Steam E lec trol ys i s (HTSE). Similar to the electrolysis of liquid w ater, HTSE uses a catho de and an ano de to split w ater molecules in to h ydrogen and o xygen. HTSE, though, has a h uge adv an tage o v e r liquid w ater electrolysis in efficiency . The lo w er en thalp y of formation of steam compared to w ater means that less electric p o w er is needed to break apart the molecules. Unlik e s ome oth e r pro cesses, the w aste heat output b y the core can b e directly used to ai d in heating the steam. The efficiency of the pro cess gro ws with temp erature b ecoming v ery efficien t for high temp erature reactors.
HTSE u s es v ery simple c hemistry and do es not create an y dangerous b ypro ducts other than the h ydrogen itself. The lac k of CO 2 output at an y stage of the pro c ess means that i n addition to not p olluting, it is also carb on neutral. There are materials concerns with ste am at high temp erature b ecoming highly corrosiv e, but these can b e met with further researc h in to ceramic materials.
While b eing v ery efficien t and clean, HTSE has a p roblem when it comes to pro duction v olume. It has y et to b e sho wn to b e viable at the pro duction rates re q uired, and an y increase in capacit y w ould require scaling up existing mo dels b y orders of magnitude. F or that reason, it w as not as attractiv e an option at first, but w as reconsidered l ate r and w ork w as done to pro v e its viabilit y for this facilit y .
3.2.7 Br-Ca-F e UT-3
The UT -3 pro cess in v olv es s olid -ga s reactions to pro duce h ydrogen [156] using the reactions in Equations 3-6 at the temp eratures indicated in paren theses [82].
C aB r 2 + H 2 O → C aO + 2 H B r (760 C C ) (3)
1
C aO + B r 2 → C aB r 2 + 2 O 2 (571 C C ) (4)
F e 3 O 4 + 8 H B r → 3 F eB r 2 + 4 H 2 O + B r 2 (220 C C ) (5)
3 F eB r 2 + 4 H 2 O → F e 3 O 4 + 6 H B r + H 2 (560 C C ) (6)
High temp erature steam cycles through t w o calcium an d t w o iron reactors and is split in to h ydrogen and o xygen, as sho w in Figure 7.
The UT-3 pro cess has b een w ell demonstrated, and has b een cited as b oth an economically and tec hni cally viable approac h for commercial h ydrogen pro duction [31]. The UT-3 pro cess can b e scaled somewhat confiden tly for our h ydrogen pro duction needs, though more sophisticated sim ulations of this h ydrogen pro duction pro ces s w ould b e required to confirm the analytical scalings t o desired h ydrogen pro duction quotas are v alid.
Courtesy of Elsevier, Inc., http://www.sciencedirect.com . Used with permission.
Figure 7: Sc hematic system arrangemen t of UT-3 pro cess [31].
4 Biofuels
4.1 Main Goals of Biofuels
The biofuel plan t’s goal w as to pro duce the greatest amoun t of fuel p ossible while utilizing the a v ailable resources of h ydrogen from the h ydrogen pro duction plan t and electricit y and pro c ess heat from the n uclear p o w er pl a n t.
4.2 Design P arameters
Biofuels had t w o m a jor decisions to mak e in designing a biofuels plan t: what kind of renew able biomass feedsto c k to use and with what pro duction pro cess w ould the feedsto c k b e con v erted. A literature searc h w as made with these t w o o v erall questions in mind and the options w ere ev aluated.
Concerning the c hoice of biomass, parameters of greatest imp ortance w ere energy d e n s it y , a v ailabilit y , costs, and comp etition with fo o d sources. The energy densit y of a crop refers to ho w m uc h e n e r gy can b e deriv ed f rom a giv en mass, and can b e appr o ximated b y the lo w er heating v alue (LHV) of a crop in MJ/kg. F or t w o crops treated with the same pr o cess, a grea t e r LHV roughly translates to more or higher qualit y fuel pro duction. Av ailabilit y of a bi om ass is also imp ortan t b ecause there m ust b e enough plan t material a v ailable to pro duce biofuels on a comme r c ial scale y ear-round. The design searc h allo w ed for the fact that some crops ma y not b e widely gro wn at the curren t time, but had the p oten tial to b e gro wn on a large scale if their demand increased. Although not the most imp ortan t factor, pro duction and transp ortation costs also w ere examined to a v oid exp ensiv e c hoices. F inally , crops w ere also ev aluated based on their usefulness as a fo o d crop b ecause of the negativ e public p erception asso ciated with using fo o d sources to create liquid fuels [96].
In c ho osing a pro duction pro cess, the tec hnical feasibilit y , pro c ess efficiency , temp erature of reaction, abilit y to utilize h ydrogen resources, main tenance requiremen ts, en vironmen tal impact, and final pro duct of the biofuels plan t w e r e all examined. The need for h ydrogen in a design w as esp ecially i m p ortan t b ecause the h ydrogen pro duction facilit y , whic h w ould b e coupled to the biofuels facilit y , w as more economically feasible as a supplier to the biofuels plan t than as a separate seller of h ydrogen. The temp erature required w as also restricted b y the pro ce ss heat that w ould b e a v ailable from th e n uclear core, whic h w ould reac h a maxim um temp erature of ab out 650 Q C. F easibilit y and efficiency w ere maximized while main tenance an d en vironmen tal impact w ere minimized as m uc h as p ossible. The t yp e of fu e l whic h w ould b e made also had to b e c hosen based on the quan tit y , demand, and asso ciated rev en u e whic h could b e generated [39].
4.3 Design Options and Ev aluation
Fiv e v arieties of biomas s and fiv e s yn thetic fuel designs sto o d out in the literature searc h and the ones most suited for the coupled plan t des ign w ere selected. As seen in Figure 8, there is more than one path to arriv e at the desired pro ducts. Biomass sour c es considered w ere switc hgrass, s or gh um, e n e r gy cane, sugar cane, and corn. The syn thetic fuel pro duction pro ces ses researc hed w ere either bio c hemical (microb e electrolysis; algae transesterification; ferme n tation to e th anol) or thermo c hemical (syngas con v ersion to ethanol or Fisc her- T ropsc h fuels). A s u m mary comparison of the v arious design options, whic h are explained in greater detail in the follo wing sections, is sho wn in T able 4. After consideration of the design parameters, syn gas con v e r s ion and the Fisc her-T ropsc h pro cess w ere c hosen to create biogasoline and bio diesel using switc hgrass feedsto c k, steam, ele ctri c it y , and h ydrogen.
4.3.1 P ossible Sources of Biomass
Biofuel pro duction required a biomass source to first b e se lected. A c ompar is on of fast-gro wing biomasses in T able 5, ho w ev er, sho ws that man y p oten tial biomas ses are also fo o d sources . F r om the non-fo o d sources, switc hgrass w as selected as the biomass b ecause of its v ery high energy densit y , reliabilit y , and the p oten ti al for scaling up to industrial lev els of gro wth in North America [147], as illustrated in Figure 9. Switc hgrass is ric h in ligno cellulose, whic h r e leases a large amoun t of energy when con v erted in to syngas (a mixture of carb on mono xide an d h ydrogen) via gasification [90]. It is desirable to gro w the switc hgrass on site b ecause transp ortation of biomass is extremely u nec onomical. Although the curren t cost p er ton is greater than
Figure 8: p os sibl e pro du c tion paths for comme rcial pro du c ts
|
Pro cess |
Reaction T emp. ( Q C) |
H 2 Input |
Steam Input |
F uel Pr o duced |
|
Algae T ransesterification |
25-60 |
no |
no |
Oil, Diesel, or Ethanol |
|
Microb e Electrolysis to Hydrogen |
25-100 |
no |
no |
Hydrogen |
|
F ermen tation to Ethanol |
190 |
y es |
y es |
Ethanol |
|
Thermo c hemical to Ethanol |
350 |
y es |
y es |
Ethanol |
|
Thermo c hemical to FT F uels |
350 |
y es |
y es |
Gasoline and Diesel Blends |
T able 4: Inputs and Outputs Comparison for Biofuel Pro duction Pro cesses
that for the other non-fo o d source of energy cane, this v alue is deriv ed from the amoun t of switc hgrass curren tly a v ailable, n ot from the amoun t that w ould b ecome a v ailable if switc hgrass b ecame a widely gro wn commercial crop [96]. Switc hgrass also has high feasibilit y b ecause it is easy to gro w and has b een used in man y recen t studies on large-scale biofuel p ro duction [68, 92].
Map courtesy of Pacific Northwest National Laboratory, operated by Battelle for the U.S. Department of Energy.
Figure 9: Sim ulated p oten tial for switc hgrass crop with one harv est p er y ear[147]
|
Crop |
Curren t Cost p er T on |
F easibilit y |
F o o d Source? |
Energy Densit y (MJ/kg) |
Dry T onnes p er Acre/Y ear |
|
Switc hgrass |
$ 60 |
high |
no |
17 |
11.5 |
|
Sorgh um |
$ 40 |
medium |
y es |
16.9 |
20 |
|
Energy Cane |
$ 34 |
medium |
no |
12.9 |
30 |
|
Sugar Cane |
$ 34 |
medium |
y es |
12.9 |
17 |
|
Corn |
$ 40-50 |
high |
y es |
13.4 |
3.4 |
T able 5: C ompar is on of Biomasses for Biofuel Pro duction
4.3.2 Electrofuels to Hydrogen Pro cess
Before the p ossibilit y of selling pure h ydrogen w as ruled out, an electrofuels pro cess w as briefly considered as an alte r nativ e w a y to pro duce h ydrogen in tandem with the first h ydrogen pro duction plan t. The elec tr ofuels pro cess is a biological means of pro ducing h yd rogen b y applying v oltage to a microbial ele ctrol ys is cell (MEC) con taining carb on-fixing microb es. The MEC allo ws microb es to surpass the p oten tial needed to pro du c e o xygen an d h ydrogen gas and to serv e as another me an s for the bioseques tr ation of CO 2 . Bes id e s the
redundancy of building t w o h ydrogen pro duction plan ts op erating on differen t systems in the sam e lo cation, the electrofuels pro cess w as also ruled out b ecause mass pro duction of h ydrogen fuel w as found to b e neither economically feasible nor safe at the presen t time.
4.3.3 Algae T ransesterification Pro cess
A second bio c hemical pro cess considered w as transesterification, where algae are used in c hemical r e action s to form new alcohols and esters. The b enefits of the transesterification pro cess are its high energy densit y and th e lo w temp erature requiremen t. Algae oil yield p er acre is higher th an other inpu ts and the pro cesses in v olv ed require only temp eratures of up to 60 Q C [45]. Algae can also capture carb on emissions from indu s tr y for reuse. Although transesterification has b een high ly dev elop ed in the United States, including a program from 1978 to 1996 to fund the dev elopmen t of renew able fuel from algae [134], it is still a difficult tec hnology to utilize. Signi fican t researc h is required to c ho ose the optimal strain of algae with a m axi m ized lipid con ten t in addition to the fact that algae harv esting is difficult. Capital costs for the initial building are also high b ec au s e the organisms m u s t b e protected from con tamination. Th us, ev en the most optimistic predictions put algae pro duction costs a t o v er $ 1.40/liter [45]. Additionally , algae transesterification only utilizes electricit y , but not the h yd roge n or high-temp erature pro cess heat resources that are a v ailable in this study .
4.3.4 F ermen tation to Ethanol Pro cess
F ermen tation of biomass to ethanol is another bio c hemical route for making biofuels. In this pro cess, biological feedsto c ks are first h ydrolyzed (br ok en do wn) in to basic sugars using a com bination of acids and enzymes, then allo w ed to fermen t in to ethanol. F ermen tation is orc hestrated b y incubating the sugars in a tank with carb on dio xide at sligh tly elev ate d temp eratur e s and pressures. The resulting ethanol is concen trated and purified via distillation. As with algae transes terifi c ation , ho w ev er, significan t rese arc h is still required to engineer the appropriate y east or bacteria whic h can efficien tly con v ert the h e micellulose of biomass in to ethanol. This pro cess w ould require electricit y and some pro ce ss h e at from the reactor, but again there is not space for a significan t h ydrogen input an ywhere in the fermen tation pro cedure [39].
4.3.5 Fisc her T ropsc h Pro cess
Another pro cess considered w as the thermo c hemical Fisc h e r-T ropsc h (FT) pro cess to either ethanol or bioga soline and bio diesel. In this pro c ess, biomass is com busted under lo w-o xygen conditions to create a pro ducer gas. Th is syn thetic pro ducer gas , or syngas, is then used in the FT pro cess to pro du c e a distribution of pro ducts that include bio diesel and bio jet fuels [90]. Finally , the arra y of pro ducts undergo a refini ng pro ces s to separate them in to commercially ready alternativ e fuels. A represen tativ e FT pro cess sc hematic is sho wn in Figure 10. While temp eratures of up to 350 Q C are needed at v arious steps in the pro ce ss, this is still comfortably lo w er than the temp eratures required b y the h ydrogen plan t an d within the range that the n uclear core pro cess heat will pro vide.
FT fuels are already b eing pro duced b y companies suc h as Ren tec h and Choren [39], so they are feasible. The w aste emissions of the pr o cess is lo w o v erall and consists of mos tl y c harred ash f rom the gasifier, trace acid gases suc h as H 2 S, and some CO 2 emissions [90]. Ho w ev er, as men tioned in the tec hn ical memorandum presen ted b y Jec h ura to the National Bio energy Cen ter, the CO 2 w aste comes primarily from the o xygen molecules of w ater, whic h are discarded as the h ydrogen molecules are tak en to refine F-T fuels [39].
Another adv an tage of this des i gn is that FT reactions pro duce a v ariet y of pro ducts that can b e refined in to gasoline, diesel, and jet fuel blends, whic h ha v e m uc h wider applications than pure ethanol. As the arra y in Figure 11 demonstrates, bio diesel fuel has the highest v olumetric energy densit y of the p ossible pro ducts [25], costs less p er gallon than fossil fuel diesel, helps reduce the US dep endence on foreign fossil fuels, and reduces the United States’ carb on fo otprin t [30]. F ur thermore , if implemen ted on a large scale, the estimated pro duction cos ts for high e n e r gy densit y FT fuel could b e as lo w as $ 1/gallon [39], whic h mak es it economically comp etitiv e.
F or all of these reasons, the FT pro cess h as b een selected as the optimal route for biofuels pro duction. In conclusion, a n uclear p o w er plan t will b e used to man ufacture Fisc her-T ropsc h fuels in a gasification-based pro cess using s witc hgrass, steam and electricit y from the n uclear c or e , and on-site pro duced h ydrogen. This
Figure 10: Basic outline of the pro cess turning biomass in to bio diesel fuel
Figure 11: V olumetric energy densit y of alternativ e comm ercial fuels burned for energy[30]
n uclear-p o w er-c ou pled plan t, in con trast to most existing biofuels plan ts, will ha v e a direct h ydrogen source and th us eliminate b oth the hea vy w ater consumption and the ma jorit y of CO 2 emissions, while the energy inputs from th e p o w er plan t further minimize other costs asso ciated the FT pro cess.
P art I I I
Results
5 Ov erall Plan t D e s ign
The en tire plan t la y out is sho wn on the follo wing page. In order to mak e the design w ork more manageable, the la y out w as divided in to four con trol v olume s: (1) Core, (2) Pro cess Heat, (3) Hydrogen, and (4) Biofuels.
34
• M assachuse 8 s 22 033 NUCLEAR SYSTEMS DESIGN PROJECT 2011
Institute of
Technology
O
O O'
o
8 iotueTs Di#ti i ia \ ion
‹ ,
6 Core
6.1 Pro cess Ov erview
6.1.1 Core Ov ervi ew
The fin al design c all s for a 3575 MWt lead-bism uth eutectic (LBE) co oled fast reactor with a secondary sup ercritical CO 2 (S-CO 2 ) system. The core utilizes urani um mononitrid e (UN) as its fuel and a fer ritic/martensitic stee l (T91) [67] with a 100 micron corrosion resistan t la y er as its c lad ding material [138]. The core is comp osed of 12 rings of hexagonal shap ed assem blies as sho wn in Figure 12. The inner te n rings house 253 fuel as sem blies of 10 0 fuel pins eac h and 18 b oron carbide (B 4 C) con trol ro ds. The 11 th ring is comp osed of a magnesium o xide (MgO) reflector and the 12 th ring is comp osed of B 4 C shield assem blies. The p o ol t yp e d e sign w as c hosen to p ermit as m uc h natural circulation of th e LBE as p ossible, supplemen ted b y pumps for full p o w er op eration. The primary heat exc hangers (HXs) are lo cated at the top of t he reactor v essel and u s e shell-and-tub e tec hnology . The outlet temp erature for the co olan t is 650 C C and the inlet temp erature is 484 C C.
Figure 12: Radial view of the core sho wing the la y out of the fuel assem blies, con trol ro ds, reflector, and shield.
6.1.2 Secondary Ov erview
The secondary lo op consists of a S-CO 2 Bra yton cycle, whic h extracts heat from the primary LBE lo op b y means of three heat exc hangers. A small amoun t of this h e at is transferred to the pro cess heat group with a separate HX. The S-CO 2 driv es a series of t w o 250 MW turbines p er lo op (of whic h there are three for a total of six turbines) with an in termediate compressor to ensure sup e rcri tic al it y is main tained. The S-CO 2 is then fed through a reheater and a condenser where it reac hes its minimal temp erature of 100 C C. It is then recompressed, reheated, and pump ed bac k to the p rimary heat exc hanger to b egin the c ycle anew. The total cycle effi ciency is calculated at 41.7% and pr o duces at least 1 GW of electric p o w er.
6.1.3 T able of Imp ortan t P aram e t e r s
Figure 13: T able of imp ortan t parameters for the reactor, still to b e done are depletion calculations and kinematics
6.2 Primary System Design
6.2.1 F uel
The fuel material c hosen w as uranium mononitride (UN). UN has some significan t adv an tages o v er uranium dio xide (UO 2 ), the t ypical fuel for commercial n uclear reactors. It has a similarly high melting p oin t as UO 2 , but has a substan tial increase in thermal condu c tiv it y [143]. The increase in conductivit y is necessary b ecause of the higher temp eratures the co olan t is op erating at and b ecause of the high p o w er densit y of the core (when compared to curren t comm ercial pl a n ts). Conserv ativ ely assuming a one meter activ e fuel heigh t (curren t pr o jections put the critical v alue near t w o meters), the maxim um cen terline fuel temp erature f o r UN is 1,862 C C as compared to 6,600 C C for UO 2 .
When usin g UN as fuel, a dra wbac k is the requiremen t that the isotop e nitrogen-15 is used in s t e ad of the m uc h more abundan t nitrogen-14. The natural abundance of N-15 is 0.366%. The r e qu irem en t to e nr ic h the nitrogen adds significan t fuel costs. Ho w ev er, there is less need to enric h the uranium when comparing to UO 2 , b e cause N-15 is a b etter mo derator and there is an increase in uranium in the fuel du e to increased densit y and c hange in s toic hiometry . T h e r e f ore , there is an adv an tageous offse t in the cost and proliferation risk when using the UN due to the decrease in uranium enr ic hmen t.
6.2.2 Criticalit y Calculations
T o calculate the criticalit y of our core, a mo del w as constructed to run in Mon te Carlo N-P article (M C NP ) co de. The co de for the final v ersion (with the ro ds fully withdra wn) can b e seen in App endix A. The final results sho wing k-effectiv e v ersus ro d p osition can b e seen in Figure 14.
Figure 14: K-effectiv e v ersus ro d p osition for final core mo del
Some notable p oi n ts on this figure are the t w o zones. In the lo w er 2/3 of the core the enric hmen t is
increased b y 2.5%. The upp er zone is sligh tly less enric hed to lo w er the w orth. Ha ving a lo w er w orth at the top of the core is use f ul b ecause reactors do not w an t to op erate with th e ro ds fully withdra wn for safet y reasons, hence the top part of th e core i s t ypically shado w ed and the fuel is not b eing used. Another p oin t is that the calculated critical ro d p osition is sligh tly less than 50% withdra w al. As describ ed in follo wing sections, thermal analysis w as done with a 1 m activ e fuel heigh t whic h then added additional conserv atism in to the calculations.
6.2.3 Sh utdo wn Margin
A requiremen t of all n uclear reactors in the United States is that with the most reactiv e con trol bl ade, or paired con trol blades, fully withdra wn with the reactor in a cold , xenon free situation , the core can b e sh utdo wn b y at least 1% [154]. As can b e seen in Figure 14, the reactor with the blades fully inserted is sh utdo wn b y nearly 15 %. Therefore, with one, or ev en t w o, blades fully withdra wn the reactor can meet the sh utdo wn requiremen t with plen t y of margin remaining as all k-effectiv e v alues w ere calculated for a cold, xenon-free condition.
6.2.4 Thermal A nalysis
A k ey test of an y reactor is t o ensur e th a t at f ull p o w er the fuel is not molten. F urthermore, c hec ks are usually done for ligh t w ater reactors to ensure that there is sufficien t margin to b oiling and that the w ater in a fuel c hannel do es not dry out. An adv an tage of u s in g LBE is that the b oiling p oin t is 1,670 C C whic h is o v er t w o and half times the op erating te mp erature. Suc h a high b oiling p oin t remo v es the risk of b oiling and dry-out in this reactor design. A thermal analysis w as conducted w i th an axially v arying linear heat rate that follo ws a sin usoidal shap e. V alues used for differen t thermal cond uctivities of the materials can b e seen in Figure 15.
Figure 15: Thermal Conductivities for di ff eren t materials in a fuel pin
F or all thermal analyses, a p er pin mass flo w rate of 5.676 kg/s w as used. Thes e results c an b e seen in Figure 16.
Figure 16: T emp eratures at differen t lo cations in the fuel pin with v arying axial heigh t. This is sho wn with the axially v arying linear h e at rate (in blue).
As can b e see n in Figure 16, the fuel op erates nearly 1,000 C C b elo w the m eltin g p oin t of UN. F urthermore, the max cladding temp erature remains b e lo w 700 C C at all p oin ts. The LBE also remains substan tially b elo w its b oiling p oin t. Figure 16 also represen ts a conserv ativ e analysis b ecause it w as done with the activ e f uel heigh t set to only 1 m. Had the activ e fuel heigh t b een raised (at b e gin ning of life (BOL) this is 1.9 m) the op erating temp eratures b ecome ev en more fa v orable.
6.2.5 Depletion Analysis
In order to estimate the lifetime p erformance of the core, it w as initially though t that calculations w ould b e made using the ERANOS fast reactor co de, dev elop ed at CEA, F rance . This co de is suitable b ecause it can pro vide rapid computation of burn up, reactor kinetics, and other imp ortan t v alues b y solving the diffusion or transp ort equations for a 2D or 3D mo del of the core an d has b een w ell b enc hmark ed against Mon te Carlo co des, including M C NP . Unfortunately , the co de w as unable to b e ru n prop erly in the time a v ailable for design. Instead, the literature w as used to tak e a conserv ativ e estimate of the fuel in v en tory and burn up o v er the lifetime of the core. A 2400 MWt LFR had b een designed in detail, including burn up analysis, b y a group at MIT with a BOL fuel comp osition of ˜71/14/15% U/TR U/Zr a v eraged o v er three zones[150]. This core w as found to remain critical for 1,800 da ys, with K eff = 1.02 at BOL and the con trol
ro ds fully withdra wn. Giv en t hat our LFR design has a K eff closer to 1.04 at BOL with con trol ro ds fully withdra wn and a greater in v en tory of fertile material, it seems lik ely that this c or e could sustain criticalit y for at least 1,800 da ys, if not longer. The real limitation s to core lifetime are due to cladding creep. The cladding temp erature will exceed 600 Q C whic h will cause th e onset of creep, limiting the lifetime of the fuel to around 460 da ys. This is less th a n a t ypical L WR refueling cycle and do es not fully exploit the fuel. F uture w ork will ha v e to fo cus on reducing these temp eratures or c hanging to higher temp eratu re c l a d dings, suc h as o xide disp ersed allo ys as they ha v e limited creep though fail at the same maxim um temp erature. This will allo w the core to mak e full use of the fuel and impro v e the economics of a once through cycle.
It w as decided not to use a depleted or natural uranium blank et as a reflector around the core b ec au s e this w ould lik ely lead to a concen trated build up of plutonium and proliferation concerns. Instead it w as decided to use a magnesium o xide reflector and breed new fuel using the fertile material inside the fuel pins. W ork at MIT has s h o wn that this material signi fic an tly impro v es br e edin g whic h allo ws an increase of the lifetime of the fuel b y up to 50%, though burn up of the cladding and other structural mate r ials is still a limitation [97]. The reactor designed b y the MIT group [ 15 0] analyzed the core in v en tory after 1,800 da ys and an a v erage disc harge burn up of 77 MWd/kf . They estimated that there w o u ld not b e a significan t increase in the Pu in v en tory (only 1.7% across the en tir e core) though there w as up to a 3.1% i ncre ase in the outer zones. It will b e i m p ortan t in the future to c hec k what p ercen tages of this Pu are Pu-239 and Pu-241 to ensur e they do not exceed 90% of the total p lutonium and presen t a significan t pr oliferation risk. T ak en together, these studies suggest there will not b e a dangerous build up of Pu in our reactor core design though this r e mains to b e v erified using ERANOS.
One promising result w as the burning of minor actinides in the fuel at BOL. The amoun t of min o r actinides d e creased b y 24.1% from 1255 MH kg to 953 MH kg after 1,800 da ys. This suggests th e p ossibilit y of using minor actinides as f uel in future iterations of our reactor design whi c h w ould giv e the reactor an economic b o ost as w ell as allo wing it to b ecome part of the United States’ n uclear w aste solution.
6.2.6 Core Reactivi t y F eedbac k P arameters
When calculating core kinetics and safet y factors, estimations a gai n h ad to b e made from the literature due to an inabilit y to calculate these directly using the ERANOS co d e . Again, the LFR analyzed b y the MIT group w as u s ed due to its similar size to our design[150]. The Doppler co efficien t of an LFR is exp ected to b e negativ e due to the har d sp ectrum and w as found to b e -0.111 +/- 0.030 ¢ /K. The co olan t temp erature co efficien t w as found to b e p ositiv e but agai n quite small, 0.131 +/- 0.052 ¢ /K. Lead exhibits a small insertion of p ositiv e reactivit y with decreased densit y accoun ting largely for this p ositiv e co effic ien t[150]. The scattering c r os s section of lead also increases with higher temp eratures, th us increasing mo deration and leak age. Ho w ev er, when similar calculations w ere made o v er an SFR core, the use of an MgO reflector reduced this temp e r ature co efficien t due to the reduction in fuel enric hmen ts th at w ere p ossible[97]. These calculations will ha v e to b e v erified for this design using a co de lik e ERANOS .
6.2.7 Natural C i rculation and Flo w Analysis
One of the b enefits of an LBE co oled r e actor op erating at relativ ely high temp eratures is that it is p ossible for the c o olan t in the core to b e en tirely driv en b y natural circulation. It w as unclear whether this w ould b e p ossible with this design due to the large increase in the core size, but first order calculations w ere made to estimate the con tribution of natural circulation in order to reduce the load on the circulation pumps. Initially , it w as only imp ortan t th a t to c hec k the feasibilit y of using natural con v ection to driv e the co olan t and it w as assume d that the there w ould b e no significan t temp e r ature loss b et w een the heat exc hanger and the core and that the reactor pressure v essel (RPV) w as an adiabatic c ham b er. Figure 17 sho ws ho w this simple mo del w ould then lo ok. Note that there is a third LBE/CO 2 heat exc hanger going in to the page.
Figure 17: Sk etc h of mo del for LBE flo w calculations
The core c hannel and the do wn c hannels b et w een the steel frame w alls w ere mo deled as pip es with diameters e qu al to their width. The width b et w een the core c hannel and the edge of the RPV is lab eled on the left as D in Figure 17 and is one of the v ariables that ma y b e c hanged in order to further aid n atural con v ec ti on. Giv en that the p itc h to diamete r ratio is quite large at 1.6, the ro d bundles w ere appro ximated as an n ular flo w for the p urp oses of calculating the friction factor [141, 86]. The pressures in the core c hannel and one of the do wn c hannels w ere then balanced a n d a function for the mass flux foun d using Equation 7.
≈
2 D 2 ( ρ hot − ρ col d )
G
23 . 95 η
(7)
Figure 18 sho ws a plot of the mass flux through the core for a giv en temp erature rise across the core, an outlet temp erature of 650 C C, and a giv en width D. The horizon tal line sho ws the m ass flux of 2 5, 371 k g / s − m 2 required to co ol the core at full p o w er.
Figure 18: Mass flux through do wn c hannel giv en an outlet temp erature of 650 C C for v arying inlet temp er atures and v alues of D. Green represen ts D = 1 m, blue represen ts D = 2 m, and red represen ts D = 3 m.
The reactor will op erate with an inlet temp erature of 480 C C and with a temp erature rise of ab out 170 C C so from Figure 18, it is pro v en that natural circulation c an driv e a significan t amoun t, if not all, of the c o olan t flo w in the core. Dep end ing on the c hosen width, D, the remaining flo w will b e pr o vided b y p umps. Figure 19 sho ws ho w the Reynolds n um b er c hanges f or the flo w in the do wn c hannel with v arying inlet temp eratures and widths. In the cases where D < 2 m, the flo w remains laminar, but ab o v e that, it transitions to turbulen t flo w.
Figure 19: V ariation in R eyn olds n um b er for flo w throu gh do wn c hannel giv en an outlet temp e r ature of 650 C C for v arying inlet temp eratures and v alues of D. Green represe n ts D = 1 m, blue represen ts D = 2 m, and red represen ts D = 3 m.
This will increase the pressure loss in the system and reduce natural con v ection. In the case where for D = 3 m, the mass flux is closer to 35,000 k g / s − m 2 . This is still sufficien t to mee t the needs of the core, but is less than half of what a laminar r e gime could pro vide.
As discussed previously , in order to a v oid corrosion of the w alls the v elo cit y of the LBE m ust remain at or b elo w 2.5 m/s, so th is is the LBE v elo cit y b eing used to co ol the core. As suc h , D m ust b e optimized to only just meet the required mass flux. Giv en the simplicit y of this mo del, it w as assumed that som e losses are neglected so in order to comp ensate for this, a mass flux of 28,500 k g / s − m 2 w as targeted. This is ac hiev ed b y setting D = 1.9 m (whe n appro ximating to the nearest dec imeter). Figure 20 sho ws the mass flux plot for this setup with c hanging inlet and outlet temp eratures. Analysis will ha v e to b e conducted in order to decide whether it is w orth allo wing the reactor to remain at higher temp eratures during sh utdo wn in order to main tain this natural circulation or if pumps can driv e the en tire flo w.
Figure 20: The axis going up the scree n is mass flux ( k g / s − m 2 ), the axis on the lo w er left side of the screen is inlet temp e r ature ( C C), and the axis on the lo w er righ t side is outlet temp erature ( C C).
This plot can also b e used to predict the b eha vior of the co ol a n t dur ing acciden t situati ons . F ortunately , due to the high b oiling p oin t of LBE, it is un lik ely that v oids will form and the densit y of LBE drop significan tly . Ho w ev er, at significan tly high temp eratures, the flo w ma y transition to a turbulen t regime th us reducing the mass flux. In general, the natural circulation should suffice as long as a temp erature rise across the core of ab out 150 C C is main tained. Underco oling is therefore a large concern and a secondary heat dump in the RPV is going to b e necessary . This could also allo w an op erator to sustain natu ral circulation at lo w er core p o w ers during sh utdo wn b y pro viding heat to the core from the PCM storage and then mo ving it to this secondary heat dump. A final p oin t of consideration during acciden t analysis will b e deciding ho w long a high flo w of LBE caused b y high temp eratures in an acciden t c ou ld b e sustained without damaging the pip es due to erosion.
6.2.8 Safet y Systems
Another adv an tage of using a LBE-co oled fast reactor is its man y inheren t passiv e safet y features. The difference in inlet and outlet temp eratures of the reactor allo ws for a significan t natural con v ection mass flo w rate of 90,000 kg/s . Natural circulation should b e able to pro vide sufficien t sh utdo wn flo w pro vided the deca y heat remains hi gh; for long sh utdo wn,s heating will b e pro vided b y pro cess heat storage to k eep the LBE ab o v e its melting temp erature. This will also help main tain the delta T across t he core needed for natu ral circulation. LBE also pro vides a significan t heat sink with large margin to b oiling, essen tially
eliminating the issue of co olan t b oiling. Reactor V essel Auxiliary Co oling Systems ( R V A Cs) w i ll b e able to pro vide ad ditional heat remo v al, in case of loss of a heat sink in addition to a large in v en tory of pri m ary co olan t. Because of the large margin to melting, pressurization of the core is not needed and it can run at atmospheric pressure. Scenarios w ere analyzed to see what w ould happ en in the case of a transien t p o w er raise. Up to a 75% increase in p o w er could b e tak en b efore the fuel reac hed its melting p oin t. As temp erature is raised, the thermal condu c tiv it y of the LBE also go es up, whic h aids in heat remo v al. Another analysis w as used to determine at full p o w er if the bulk co olan t temp erature w ould cause fuel melting. T emp eratures of u p to 1,400 C C could b e withsto o d b y the fuel b efore melting. Ho w ev er, at suc h temp eratures the LBE gap w ould b oil and the clad ding w ould fail. Nev ertheless, this analysis sho ws the resiliency of the fuel m aterial . It is recommended that to main tain the strength of the T91 clad, the temp erature should ideally remain b elo w 700 C C. A t normal op erating tem p eratures, the clad al w a ys remains b elo w 700 C C and will still remain b elo w that v alue ev en with a 10% increase in p o w er. Bey ond this p oin t, the creep effects on the T91 b e gin to accelerate the clad and it loses m uc h of its lifetime. As it is, T91 can only b e op erated at 650 C C for a limited amoun t of time (less than t w o y ears) and in order to enhance safet y and pro vide b e tter longevit y , a new clad will need to b e c hosen.
6.3 Secondary System
A basic o v erview of the core’s secondary system can b e seen in Figure 21 with no de v alu e s sho wn in Figure 22.
Figure 21: Ov erview of the secondary system of the reactor
Figure 22: The no dal v alues of temp erature, pressure, and mass flo w for the s econdar y system
6.3.1 Heat Exc ha n ger
In selecting the heat exc hangers (HXs) for the reactor, sev eral ge ometries w ere considered and compared using data from the S T AR-LM reactor conceiv ed b y Argonne National Lab oratory . The design of the ST AR-LM reactor design also consists of a LFR with a S - CO 2 secondary cycle. As a result, the relativ e trends asso ciated with the ST AR-LM secondary system w ere assumed to b e analogous to the trend s that w ould b e obse r v ed in the pro duction of this LFR design. The initial ob jectiv e w as to use a Prin ted Circuit Heat Exc hanger (PCHE) design as the in-reactor heat exc hanger. The PCHEs are pro duced b y the British compan y , HEA TRIC, whic h is a division of Meggitt (UK) Ltd.[38]. These heat exc hangers consist of flat plates with semi-circular flo w c hannels that are stac k ed and adhered together throu gh diffusion b onding. A sc hematic of this t yp e of heat exc han ge r is sho wn b e l o w in Figure 23.
© Heatric. All rights reserved. This content is excluded from our Creative Commons license. For more information, see http://ocw.mit.edu/fairuse .
Figure 23: Prin ted Circuit Heat Exc hanger (PCHE) Design
The PCHE design allo ws for efficien t heat transfer b et w een w or king fluids (results in 44.1% cycle efficiency if used as the in-reactor heat exc hanger). Additionally , PCHEs are extremely compact with a heigh t of 2.76 m, whic h is a fraction of the heigh ts of other HX t yp es. This decrease in size could lead to c heap er and more delib erately compact system designs that w ould use a fraction of the materials in comparison to other design options.
The largest obstacle with using the PCHE design for the in-reactor heat exc hanger results from the reactor co olan t. While a PCHE could b e made to facilitate the heat exc hange from the LBE to the S-CO 2 , the flo w c hannels for the LBE w ould need to ha v e a significan tly larger diameter than the diame ter of the flo w c hannels for the S-CO 2 due to th e greater viscosit y and frictional factor for the LBE. Consequen tly , a simpler design for the in-reactor heat exc hanger w as c hosen.
F rom the p ossible HX designs, a basic shell-and-tub e heat e x c hanger w as c hosen b ecause of its simplicit y and ease of man ufacturin g compared some of the other designs, including the straigh t ann uli tub e HX. In order to maximize the heat transfer and flo w of LBE through the shell-and-tub e heat exc hanger, t w o geometries of the tub es within the heat exc hanger w ere considered: a rectangular, in-line geometry and a triangular, staggered geometry (b oth sho wn b elo w in Figure 24).
Figure 24: Geometry options for the in-reactor shell-and-tub e heat exc hanger.
Due to the viscosit y of the LBE, the staggered ge ometry of the tub es allo w e d for a gr e ater amoun t of flo w through the sub-c hannels and th us resulted in b etter heat transfer b et w een the LB E and the S-CO 2 so
this design w as ultimately c hosen [140].
F or the recup erator (a regenerativ e HX) in the secondary cycle that preheats the S-CO 2 b efore it returns to the in-reactor heat exc hanger, the PCHE heat exc hanger design w as c hose n . This design is more adv an tageous than a shell-and-tub e HX at this p oin t in the system b ecause the heat exc hange b et w een S-CO 2 to S-CO 2 do es n ot ha v e the same viscosit y and friction concerns that exist b et w een the LBE and S-CO 2 . The aforemen tioned b enefits of the compact PCHE design made it the b est c hoice for the recup erator comp onen t of the secondary system.
6.3.2 Condensers and Compressors
The compressors used increase the p res sur e b y a factor of 2.2 and use ab out 42 MW of p o w er to op erate. Compressors are placed b oth in b et w een the turbines (to ensure th e pressure k eeps the S-CO 2 sup ercritical) as w e ll as after the condenser to help bring the pressure bac k to its maxim u m v alue in or der to complete the cycle. The c ond e n s er uses an en vironmen tal w ater source (i.e. riv er, o cean) to decrease the temp erature of the S-CO 2 b efore it is recompressed and reheated to b e fed in to the primary heat exc h anger. Using a minim um temp erature of 100 C C guaran tees that the sup ercritical temp erature (˜31 C C) is met with sufficien t margin. The mass flo w of w ater required to co ol to this temp erature w as determined to b e 13,715 kg/s, w ell within reason for existing condensers already in use.
6.3.3 Acciden t Scenario Analysis
The main t w o acciden t scenarios that are essen tial to explore in this design are a loss-of-heat-s i nk (LOHS) ev en t without a scram in the core and a loss- of - generator - load (LOL) ev en t without a scram.
LOHS ev en t without a scram
F ollo wing this ev en t, the temp eratures throughout the system will b egin to rise b ecause the reactor is still pro ducing heat without it b eing prop erly remo v ed. As a result, the co olan t densit y will b egin to d e crease, whic h coupled with the D oppler effect in the fuel, will indu c e a decrease in reactivit y in the core. As time progresses follo wing thi s ev en t, heat will con tin ue to b e generated, but the rate at whic h this heat generation will o ccur will b egin to d e crease with time. Since the heat will not b e prop erly remo v ed b y the secondary cycle, th e Reactor V essel Auxiliary Co oling System ( R V A CS) will b e resp onsib le for the ma jorit y of the heat remo v al. During this time, the system temp eratures will con tin ue to increase due to the deca y heat pro duced b y the core, whic h will result in a greater amoun t of negativ e reactivit y b eing inserted i n the reactor. Ov er time, the rate of heat remo v al b y the R V A CS will exceed the heat generation rate from d e ca y heat, and the temp erature of the system will b egin to decrease, causing an increase in p ositiv e reactivit y . Consequen tly , the system will exp erience oscillations in the p o w er lev el as w ell as in the temp eratu re o v er time, but the actual p o w er lev el will remain lo w compared to the lev el at fu ll p o w er [107].
LOL ev en t without a scram
A LOL acciden t is c on s id e r e d to b e one of the most sev ere acciden t s that ma y o c cur in the sec on dary S-CO 2 Bra yton cycle, although the system’s resp onse d o es not p ose significan t safet y concerns o v erall. In this acciden t s cenario, the ass u m p tion is that the generator is separated from the grid. Initially , it is estimated that since the turbine energy that w ould normally b e s en t to the grid cannot b e transferred, the resp onse of the secondary cycle will b e to try to minimize the amoun t of departure from the normal function. Ho w ev er, the rotational sp eeds of the turbine will increase drastically . Im mediately follo wing th e LOL ev en t, the S-CO 2 Bra yton cycle con trol system will op en the turbine b ypass v alv e, whic h m ust o ccur at an extremely rapid rate (from fully closed to fully op en in less than 0. 5 seconds). Once this v alv e op ens, the flo w rate of th e S-CO 2 will d e crease. F ollo wing this ev en t, the tem p erature of the LBE c o olan t in the reactor will increase, in tro ducing a negativ e reactivit y insertion in to the core, whic h reduces the p o w er lev e l in the core [107].
6.4 Economics
Another imp ort a n t asp ect of assessing the feasibilit y of this LFR design with a secondary Bra yton cycle lo op is d e termini ng whether the reactor could b e economically comp etitiv e with other p o w er plan ts. In particular , it is of in terest to examine the cost comparison b et w een this system and a t ypical ligh t-w ater reactor ( L WR). In a pap er written b y V. Dostal, M. J. Driscoll, and P . Hejzlar, it is sho wn that estimating the cost of comp onen ts in the secondary cycle of this reactor is particularly difficult b ecause there are limited examples of these comp onen ts actually in use, and therefore, the cost estimate is pr im ar ily founded on e xp ert opinions [54]. Dostal, et al. [54] u tilize d the data asso ciated with the helium Bra yton cycle in order to pro vide an appro ximate cos t comparison, wh ic h will also b e done in this ec on om i c analysis. Ultimately , Dostal et al. assert that, “The sup ercritical CO 2 cycle is more efficien t than the steam cycle and its op erating and main tenance costs are not exp ected to exceed those of the steam cycle. Therefore, if the capital cos t of the sup ercritical CO 2 cycle is lo w er than that of the steam cycle the electricit y generation cost will b e lo w er as w ell” [54]. In a comparison b et w een the cost of the turb o-mac hinery for a helium Bra yton cycle and the c ost of the turb o-mac hinery in a S-CO 2 , it w as f ound that the cos t of the S-CO 2 cycle w ould b e less than that of the helium cycle. An o v erview of the cost comparison b et w een the t w o can b e found in T able 6.
|
T urb omac hinery |
Conserv ativ e |
Best Estimate |
Cons. |
Best Est. |
Cons. |
Best Est. |
|
T emp erature ( Q C) |
550 |
650 |
700 |
550 |
650 |
700 |
|
Efficiency (%) |
41.0 |
45.3 |
47.0 |
43.1 |
47.1 |
48.9 |
|
P o w er (MW e) |
738 |
815 |
846 |
776 |
848 |
880 |
|
T emp erature Ratio |
0.925 |
0.932 |
0.940 |
0.925 |
0.932 |
0.940 |
|
P o w er Rati o |
0.01 |
0.960 |
0.983 |
0.931 |
0.985 |
1.008 |
|
Pressure Ratio |
0.730 |
0.730 |
0.730 |
0.730 |
0.730 |
0.730 |
|
He T u r b i ne (K $ ) |
78,000 |
78,000 |
78,000 |
78,000 |
78,000 |
78,000 |
|
CO 2 T urbine (K $ ) |
47,455 |
50,945 |
52,614 |
49,035 |
52,272 |
53.952 |
T able 6: Summary of turbine costs [54]
The reheater in the secondary lo op’s (a HEA TRIC PCHE) cost can b e estimated using the w eigh t of the heat exc hanger. It is estimated that for stain le ss steel PCHE u nits, the cost is $ 30/kg, while the cost for titanium units is $ 120/kg [54]. In order to determine the fraction of metal in the PCHE, Equation 8 can b e used where f m is th e fraction of metal p er cubic meter, d is th e diameter of the semicircular c hannel, P is the c hann e l pitc h, and t is the thic kness of the heat exc hanger plate [54].
f m = 1 − 8 P t
(8)
In order to obtain the v olu m e of metal used in the PCHE p e r cubic meter of its v olume, Equation 8 m ust b e used with the parameters of the unit to obtain the fraction of m etal p er cubic meter. This result can then b e m ultiplied b y the total v olume of the unit to find the total mass of metal. The total c ost of the u nit can then b e calculated using the appropriate $ /kg n um b er.
T o summarize the estimations b et w een a steam cycle (c ou pled with a high-temp erature gas reactor HTGR), a helium cycle, and a CO 2 cycle, a comparison w as made b et w een the total capit a l cost of the cycle as w ell as the capital cost p er kW e as so ciated with the cycle. This comparison can b e found in T able 7.
This summary of fractional costs of v arious S-CO 2 cycles demonstrates that this design ma y b e econom ically comp etitiv e with other secondary cycles. As a res u lt of the compact size and efficiency of the S-CO 2 Bra yton cycle, the p oten tial cost reduction of the secondary cycle ma y help to offset the o v erall capital cost p er kW e of a system coupled w i th the lead fast reactor (LFR) design [148].
|
T urb o-mac hinery |
T emp e r ature ( C C) |
Costs |
vs. Steam Cycle |
vs. He li um Cycle |
|
Conserv ati v e Conserv ati v e Conserv ati v e Best Estimate Best Estimate Best Estimate |
550 650 700 550 650 700 |
Capital Cost/kW e |
0.865 |
1.0553 |
|
T otal Capital Cost |
0.922 |
0.896 |
||
|
Capital Cost/kW e |
0.784 |
0.956 |
||
|
T otal Capital Cost |
0.922 |
0.897 |
||
|
Capital Cost/kW e |
0.755 |
0.922 |
||
|
T otal Capital Cost |
0.922 |
0.897 |
||
|
Capital Cost/kW e |
0.822 |
1.004 |
||
|
T otal Capital Cost |
0.921 |
0.896 |
||
|
Capital Cost/kW e |
0.753 |
0.919 |
||
|
T otal Capital Cost |
0.922 |
0.897 |
||
|
Capital Cost/kW e |
0.726 |
0.886 |
||
|
T otal Capital Cost |
0.922 |
0.897 |
T able 7: F ractional costs of the differen t sup ercritical CO 2 cycle designs [54]
Figure 25: An o v erview of the pro cess heat system
7 Pro cess Heat
7.1 Pro cess Ov erview
—
The pro cess heat system is sho wn in Figure 25. Helium is the primary w orking fluid and is sho w in in blue in the figure. S CO 2 and w ater are sho wn in green and red resp e ctiv ely . 149.85 kg/s of He circulate through the system.
—
The Pro cess Heat system dra ws heat from the high temp erature S − CO 2 to heat helium up to 606.55 Q C. This pr o cess tak es place in the 9 PCHE1 h e at exc hangers connected in parallel to the S CO 2 lo op. 132 kg/s of the He flo w then passes through heat storage and 17.85 kg/s is re-routed around it.
The helium going through the heat storage device exp eriences a pressure drop. A circulator lo cated so on after the exit from heat storage is used to raise the pressure of the helium bac k to 5MP a. Main tain ing the lo w densit y w orking fluid at a high pressure impro v es its heat tran s f e r prop erties. The He, no w at 590 Q C passes through the 12 PCHE2 heat exc hangers connected in parallel at the h ydrogen plan t. The PCHE2 heat exc hangers are used to pro duce steam at 1.977 M P a and 559.49 Q C for high temp eratu re steam electrolysis for the pro duction h ydrogen.
Co oler He, no w 212.36 Q C, returns to th e PCHE1 heat exc hangers where it is heated bac k up to 605.55
Q C.
An emergency heat sink is connected parallel to the PCHE1 heat exc hangers to rapid ly remo v e heat from
the pro cess heat system in the ev en t of a reactor scram.
Not on the primary He lo op (sho wn in red) is a ceramic heat exc hanger that will b e lo cated at the h ydrogen plan t. This heat exc hanger will b e u s ed to dra w heat from the high temp erature h ydrogen and o xygen streams to pro duce steam at 182 Q C and 0.1MP a for the Biofuels plan t.
F ollo wing sections desc r ib e op erating conditions and individual comp onen ts of the p ro cess heat system in detail. Also presen ted in conclusion is a summary of the pro cess heat sys t e m costs.
7.2 Heat Exc hangers
7.2.1 Choice of Material and W orking Fluid
In c ho osing a w orking fluid, k ey f ac tor s to consider are practicalit y , heat capacit y , and viscosit y . As T able 8 sho ws, helium has the most attractiv e w orking fluid prop erties for the system under consideration. PCHEs w ere c hosen as t w o of the four heat exc hangers (HXs) in the pro ce ss heat system . PCHEs ha v e a maxim um op erating pressure and tem p erature of 60 MP a and ˜900 C C [55] resp ectiv ely . High ac hiev able op erating temp eratures and pressures w ere decisiv e factors in the selection of PCHEs for dra wing heat from the S-CO 2 and heating w ater at the h ydrogen plan t.
|
Fluid { 5MP a; [200 C C, 700 C C] } |
Heat Capacit y [ J / kg − K ] |
Viscosit y [P a-s] |
Boiling T emp erature ( C C) |
|
Carb on Dio xide (CO 2 ) |
[1,079.5, 12,37.8] |
− 5 − 5 [2 . 337 × 10 , 4 . 064 × 10 ] |
263.94 |
|
W ater/Steam (H 2 O) |
[4,476.1, 2,351.5] |
− 4 − 5 [1 . 35 × 10 , 3 . 678 × 10 ] |
14.28 |
|
Helium (He) |
[5,188.9, 5,190.6] |
− 5 − 5 [2 . 740 × 10 , 4 . 533 × 10 ] |
-264 |
T able 8: P oten tial W orking Fluid Prop erties [18]
The heat exc hangers at the h ydrogen plan t that will mak e use of streams of sup erheated steam, h ydrogen, and o xygen are y et to b e designed and will b e sub ject of futu re w ork. T able 9 sho ws a summary of the PCHE op erating p arame ters. The follo wing sections discuss eac h PCHE in fu rther detail. The impact of v arious design parameters on PCHE v olume, pressure drop, op erating cost and capital cost is presen ted. Axial temp erature, heat flux profiles, and cost estimations are sho wn and f ouling related problems are discussed. A design strategy for the biofu e ls heat exc h anger is outlined and futu re w ork is describ ed.
|
unit |
PCHE1 |
PCHE2 |
|
|
Num b er of un its |
9 |
12 |
|
|
Heat rate/unit |
MW |
35 |
26 |
|
T otal heat rate |
MW |
315 |
312 |
|
Lo cation |
S-C O 2 lo op |
Hydrogen plan t |
|
|
PCHE t yp e |
zigzag c hannel |
straigh t c hannel |
|
|
Hot fluid |
S-C O 2 |
He |
|
|
Flo w confi guration |
coun terflo w |
coun terflo w |
|
|
Cold Fluid |
He |
H 2 O |
|
|
T hot in |
Q C |
630 |
600 |
|
T cold in |
Q C |
199.6 |
40.81 |
|
T hot out |
Q C |
314.73 |
212.36 |
|
T cold out |
Q C |
605.07 |
559.48 |
|
P hot in |
MP a |
20 |
5 |
|
P cold in |
MP a |
5 |
2 |
|
delta P hot |
kP a |
8.043 |
9.812 |
|
delta P cold |
kP a |
23.749 |
13.874 |
|
mass flo w hot |
kg/s |
90 |
12.68 |
|
mass flo w cold |
kg/s |
16.65 |
7.4371 |
|
v cold inlet |
m/s |
7.2625 |
0.0243 |
|
v hot outlet |
m/s |
1.0621 |
4.1990 |
|
Re hot |
6,487.67 |
1,646.55 |
|
|
Re cold |
1,177.35 |
492.34 |
|
|
h tc hot |
W/m 2 -K |
2398 |
850.01 |
|
h tc cold |
W/m 2 -K |
2,271.98 |
25,314.58 |
|
h tc total |
W/m 2 -K |
1,087.71 |
735 |
|
v olume |
m 3 |
0.9172 |
1.3 |
|
length |
m |
0.30699 |
1.09 |
|
d hot |
mm |
2 |
2 |
|
d cold |
mm |
2 |
2 |
|
pitc h hot c hannel |
mm |
2.65 |
2.65 |
|
pitc h cold c hannel |
mm |
2.65 |
2.65 |
|
n um b er of hot c hannels |
293,888 |
186,368 |
|
|
n um b er of cold c hannels |
293,888 |
93,184 |
|
|
plate thic kn e ss |
mm |
1.6 |
1.6 |
|
n um b er of hot plates |
1,312 |
832 |
|
|
n um b er of hot plates |
1,312 |
416 |
T able 9: Summary of PCHE results
Helium has the highest heat capacit y of the examined w orking fl uids, while its lo w vi s cosit y mak es pumping feasible b y minimizing frictional losse s and pump w ork. Its extremely lo w b o i ling temp erature of
-264.15 C C means there is no c hance of t w o phase flo w en tering the system, whic h could p oten tially damage
fragile equipmen t lik e compressor blades. Another adv an tage of He is the fact that it is a noble gas and therefore c hemically inert, whic h significan tly reduces the prosp ect of a con taminated or corrosiv e w orking fluid. Although the price and a v ailabilit y of helium is questionable due to recen t shortages [17] , this non tec hnical reason w as neglected for the pu rp oses of this study .
In order to c ho ose a HX material, one m ust consider the tensile strength, thermal conductivit y , and co efficien t of thermal e xp ansion. Additionally , it is desirable to kno w a material’s corrosion resistance, the ease with whic h it can b e man ufactur e d , and the an ticipated lif e ti m e. Allo y 617, a nic k el-c hromium-cobalt molyb den um allo y , and Allo y 230, a ni c k el-c hromium-tungsten-cobalt allo y , w ere in v estigated and the results of the researc h conducted are sho wn in T able 10.
|
Allo y 230[14] |
Allo y 617[15] |
|
|
thermal conductivit y , k at 650 C C |
21.4 W m − K |
23 W m − K |
|
tensile strength at 650 C C |
675 MP a |
627 MP a |
|
mean co efficien t of thermal e xp ansion |
15.2 µm for 25-800 C C mC |
15.1 µm for 20-760 C C mC |
T able 10: P oten tial M ate r ials for PCHE
T o b e rigorous, a calculation of maxim um stress w as p erformed using Equation 9 where p is the design pressure, r out is the plate thic kness, and r in is the c hannel radius [54].
max · ou t in −
p r 2 + r 2
2 r − r
2 2
τ = 1 (9)
out in
F or a PCHE with a plate thic kness of 1.6 mm, a c h annel diameter of 2 mm, and exp eriencing a pressure of 20 MP a, th e maxim um stress is 21.57 MP a. Ho w e v er, the sharp edges of the semi-circular PCHE c hannels require the calculation of a radius of curv ature for the c hannel edge at whi c h the design stress of Allo y 617 is not b e exceeded.
Equation 10can b e used to calculate the increas e in stress due to the presence of a crac k [144]. W e treat the sharp edge of the semi-circular c hannel as the crac k here, ρ is th e radius of curv ature of the crac k and a is the crac k length whic h is tak en here to b e the radius of the c hannel.
σ max
= σ (1 + 2 a ) (10)
ρ
Using Equation 10, w e find that for radii of curv ature less than 100 ] m, the design stress of Allo y 617 is exceeded. A radiius of curv ature 200 ] m pro duces a s tr e ss of 114MP a whic h is b elo w the design stres of Allo y 617. It is prop osed that PCHEs b e fabricated suc h that the edges of the semi-circular c hannels h a v e radii of cur v atures 200 ] m or greater.
Actual op erating pressures that the PCHE heat exc han ge r s will encoun ter throughou t the system range from atmos ph e r ic (˜0.1 MP a) to 20 MP a. System temp eratu re s range f rom ro om temp erature (˜20 C C) to 606.5 C C. By c ompari ng results from Equation 9 with T able 10, it is eviden t that heat exc hangers fabricated from either of these allo ys w ould op erate w ell b elo w the design stress at all p oin ts within the system. Allo y 617 w as ultimately c hosen due to the demonstrated feasibilit y of their use in man u facturing PCHE heat exc hangers [160].
7.3 PCHE Design
The impact of c hannel d iam eter, mass flo w rate, and c hannel configuration (straigh t or zigzag c hannels) on the PCHE v olume w as studied. T able 11 outlines the impact of design c hoi c es on the PCHE v olume and pressure drop across the HX. A discussion of design c hoices and their impact on the op erating parameters is presen ted in App endix of this rep ort. A summary of the findings along with the implications of de sign c hoices on costs is discussed in this section.
|
P arameter c hannel c on figuration |
Change in parameter |
V olume |
Pressure drop |
Capital cost |
Op erating cost |
|
straigh t |
t |
t |
t |
t |
|
|
zigzag |
t |
t |
t |
t |
|
|
hot fluid m ass flo w rate |
t |
t |
t |
t |
t |
|
c hannel d iame ter |
t |
t |
t |
t |
t |
|
n um b er of PCHE units |
t |
t |
t |
t |
t |
T able 11: Im p ac t of design paramete r s on PCHE v olume and pressure drop
The v olume of a PCHE is se n s it iv e to sev eral design parameters. A s tr aigh t c hannel configuration results in lo w er Reynolds n um b ers, lo w er Nusselt n um b ers, and p o orer heat transfer for b oth the hot and cold fluids.
All other design parameters remaining unc hanged, the PCHE v olu m e for a straigh t c hannel PCHE is larger than th at of a zigzag c hannel PCHE. Similarly , a higher mass flo w rate for the hot fluid and a smaller c hannel diameter impro v e heat transfer and result in a redu c tion in the PCHE v olume. F or the same total p o w er, ha ving a larger n um b er of PCHE uni ts increases the total heat exc hanger v ol ume , but reduces the pressure drop.
The capital cost of the PCHE is in v ersely prop ortional to the PCHE v olume. A lo w er PCHE v olume results in a lo w er co st of f abrication as w ell as lo w er materials costs [54]. Op e rati ng costs are prop ortional to the hot fluid mass flo w rate an d p res sur e drop across the PCHE unit. A larger mass flo w rate of the hot fluids reduces the amoun t of h ot fluid a v ailable for electricit y pro duction and the asso c i a t e d addition of pumps and compressors further increases the capital cost of the plan t .
In summary and as seen in T able 11, a higher mass fl o w r a t e , sm al le r c h a n nel diameter, and zigzag c hannel confi guration minimize the PCHE v olume, but adv ersely impact the pressure drop.
7.4 Pro cess Heat PCHEs
The follo wing se ctions presen t and describ e the heat flux and temp erature profiles of the pro cess heat PCHEs. Small fluid c hannel v elo cities w ere c hosen to minimize fouling and reduce pressure drops wherev er p ossible.
7.4.1 PCHE1
As indicated in T able 9, PCHE1 is lo cated at the S-CO 2 reactor lo op and consists of 9 units of 35 MW zigzag c hannel PCHEs with a coun terflo w configuration. Zigzag c hannels w ere c hosen for PCHE1 to increase flo w turbulence, impro v e heat transfer, and reduce the heat exc hanger v olume. Zigzag c hannels, ho w ev er, increase the pressure drops b y making the flo w more turbulen t. A zigzag c hannel design w as feasible for PCHE1 due to th e friction factors for S-CO 2 and helium b eing lo w er than that of w ater. As helium passes through PCHE1, it is heated from 199.6 Q C to 605.07 Q C.
Axial T emp erature Profiles
The temp erature profiles of S-CO 2 , He, the S-CO 2 c hannel w all, and the He c hannel w all are sho wn in Figure
26. The fluids in PCHE1 units do n ot undergo ph as e c hanges and the heat transfer mec hanism is single phase forced con v ection. The highest w all temp erature is 594.7 Q C. Allo y 617 has b een sele cted for heat exc hanger fab rication b e cause it has b een pro v en acc eptab le for temp eratures up to 982 C C [1].
Figure 26: Axial temp erature profile of the hot and cold fluid and hot and cold c hannel w alls
Axial Heat Flux Profile
The axial heat flux profile for PCHE1 is sho wn in Figure 27. The heat flux is highest at the S-CO 2 inlet and He outlet (where x = 0) and decreases with an increase in the temp eratures of the cold fluid (He). As a result of b oth fluids b eing in a single phase, there are no abrupt c hanges in the heat flux or heat transfer deterioration due to a transition from single to t w o-phase flo w or due to surpassing critical heat flux.
Figure 27: Heat flux profile for PCHE1
7.4.2 PCHE2
PCHE2, lo cated at the Hydrogen plan t, consists of 12 units of 26 MW, straigh t c hannel PCHEs with a coun terflo w configuration. Straigh t c hannels w ere c hosen for PCHE2 to reduce the pressure drop on the w ater side. The HX with He in PCHE2 is used to heat w ater from 40.81 Q C to 559.48 Q C.
Flo w Qualit y and Mass Flux
Flo w stratification in a horizon tal tub e results in a c ri tica l heat flux v alue of zero on the v ap or s id e [137]. Equation 11 [61] and Equation 12 [110] w ere used to calculate the minim um H 2 O mass flux needed to prev en t flo w stratifi c ation .
1 ) 2 (11)
where
x 0 . 65 + 1 . 11 X LT
( dP ) l
µ 1 − x ρ
X 2 = f r ic = ( f ) 0 . 25 ( ) 1 . 75 ( f ) (12)
f r ic
LT ( dP ) v µ g x ρ g
The minim um mass flu x as a function of flo w qualit y is sho wn in Figure 28. In order to prev en t flo w stratification, a c hannel mass flu x of 50 k g / m 2 − s w as c hosen for w ater.
Figure 28: Minim um mass flux and flo w qualit y in PCHE2
Axial T emp erature Profile
The axial temp erature profiles of He, H 2 O, the He c hann e l w all, and the H 2 O c hannel w alls are sho wn in Figure 29. PCHE2 also has a c ou n terflo w configur a t ion whic h means that the He and w ater flo w in opp osite directions. The PCHEs w ere mo deled using a no dal mo del implemen ted in F ortran [54].
Figure 29: Axial temp erature profile of PCHE2
The region mark ed as ’excluded’ on Figure 2 9 is b eliev ed to b e an artifact of the co de due to the extremely small temp erature c hanges that tak e place in this axial region. The region has b een excluded from the PCHE2 geometry , ho w ev er, this do es not affect the inlet or outlet temp eratures of either fluid. Besides the excluded region, the H 2 O temp erature profile is indicativ e of:
1. sup erheated v ap or with single phased forced con v ection at the He inlet and H 2 O outlet;
2. t w o phase flo w with n ucleate an d transition b oiling;
3. and sub-co oled liquid with single phased forced con v ection at the He outlet and H 2 O inlet.
The sharp gradien t of the He temp erature in the t w o phase region is indicativ e of the high heat tran s f e r co efficien ts for n ucleate b oiling. The large temp erature sw i ngs seen in this HX are l ik ely to result in a reduced heat exc hanger design life due to thermal stre sses. F uture w ork on this HX will in v olv e in v estigating heating H 2 O in PCHEs arranged in series, instead of in a single PCHE . Shell-and-tub e heat exc hangers will also b e in v estigated further as a p oten tial alternativ e design c hoice .
Axial Heat Flux Profile
The axial heat flux profile for PCHE2 is sho wn in Figure 30. The heat flux in the sub-co oled region (whic h b egins at 0.68 m and con tin ues to the left on the x-axis) is seen to decrease as the w ater approac hes n ucleate b oiling. This can b e explained using the temp erature profiles in Figure 29. In mo ving a w a y from the He outlet and H 2 O inlet, the temp erature gradien t across the hot and cold fluids decreases. This decrease in the temp erature grad ien t manif e sts itse lf as a lo w er heat flux. Ho w ev er, with the creation of bubbles and the onset of n ucleate b oiling, the heat transfer prop erties are greatly impro v ed and the heat flux increases and reac h e s a maxim u m near the e n d of the t w o phase region.
Figure 30: Heat flux profile of PCHE2
The heat fl ux for single phase forced con v ection in th e sup erh e ated v ap or decreases on approac hing the He inlet and H 2 O outlet. This can again b e attributed to a decreasing temp erature gradien t across the hot and cold c hannels.
In Figure 30, the largest heat flux, 252 kW/m 2 , is seen to o ccur near the end of the t w o phase region at a flo w qualit y b et w een 0.85 and 0.9. According to the Gro enev eld lo ok - u p tables, at a pressure of 2 MP a, mass flux of 50 k g / m 2 − s , and a flo w qu alit y b e t w een 0.85 and 0.9, the c r itical heat flux is b et w een 658 kW/m 2 and 373 kW/m 2 [65]. Although the CHF v alues in the Gro enev eld lo okup tables are for circular c han nels , they are w ell ab o v e the largest heat flux in the 26 MW PCHE. Therefore, the op erating conditions for PCHE2 pro vide a significan t margin to CHF.
F uture w ork should address d e termini ng correction factors for the se micircular PCHE c hannels in order to precisely compute v al ue s for CHF.
7.4.3 PCHE Conclusions
A zigzag c hannel configur ation w as c hosen for PCHE1 and a straigh t c hannel confi guration for PCHE2. Straigh t c hannels w ere c hosen for P C HE 2 in order to minimize pressure drops on the w ater side. The ideal configuration for PCHE2, ho w ev er, w ould b e a zigzag c hannels for He and straigh t c hann e l s for H 2 O. This design w as unable to b e completed in th e design timeframe giv en due to limitations of the computational mo del. F utur e w ork should explore this option and also th at of c ar rying out the heat exc hange b et w een He and H 2 O in m ultiple stages of PCHEs or in shell-and-tub e HXs. A Matlab mo del for sizing a coun ter flo w shell-and-tub e HX with He in the shell and H 2 O in tub es is b eing dev elop ed and will b e a v ailab le for future studies.
7.5 F ouling
F ouling adv ersely affects b oth the p erformance and the design li fe of a heat exc h anger. Dep osition of impurities in the flo w passages results i n a r e d uc t ion of the equ iv alen t diameter and reduction of the heated p erimeter. This increases the pressure drop across the heat exc hanger and lo w ers its heat rate [80].
F ouling is exacerbated b y high fluid v elo cities, rough surfaces and impurities in the fluid. PCHE1 is a S-CO 2 /He HX and PCHE2 is a He/H 2 O HX. Although s mall amoun ts of impurities will b e presen t in b oth He and S - CO 2 , the PCHE c hann e ls for these fluids are lik ely to exp erience less fou ling than those con taining w ater [78]. The addition of c hlorine to the feedw ater and the use of 200 µ m s trai ners can coun ter biofouling [109]. Addition of c hlorine ho w ev er, migh t create the problem of separating c hlorine from the steam b efore using the steam for high temp erature electrolysis for h ydrogen pro du c ti on.
T able 12: Steam/ h ydrogen and o xygen streams from t he h ydrogen plan t
|
Prop ert y |
Unit |
|
|
Inlet temp erature |
836 |
C |
|
Outlet temp erature |
71 |
C |
|
H 2 O mass flo w rate |
17.57 |
kg/s |
|
H 2 mass flo w rate |
7.9 |
kg/s |
|
O 2 mass flo w rate |
62.6 |
kg/s |
|
Heat rate from H 2 O/H 2 stream |
158.21 |
MW |
|
Heat rate from O 2 stream |
49.29 |
MW |
PCHE fouling studies indicate no c hange in HX effectiv eness, but a 55% higher press u re drop for op erating times of 500-660 hours [109]. Ho w ev er, an 18 mon th fuel cycle will require 12,960 PCHE op erating hou rs. The pressure drop related effe cts of fouling can therefore b e mitigated b y installing redundan t PCHE units for b oth PCHE1 and PCHE2, rerouting flo w, and carrying out c hemic al cleaning on the offline system during the reactor fuel cycle. The other option is to increase the pumping and compression inputs to c ou n ter the effects of larger pressure drops. Ho w ev er, c ho osing the latter option will require quan tif ying thermal effectiv eness deterioration o v er PCHE op erating times of the order of 12,960 hours, whi c h is not a simple task.
7.6 Heat Exc hanger at Biofuels Plan t
Tw o heat exc hangers ma y b e used to extract heat from the steam + H 2 and O 2 streams. Another option w ould b e to not utilize the heat from the fluid streams at t he h ydrogen plan t. This w ould requi re the use of condensers that du m p this heat to th e atmosphere or a b o dy of w ater. This w ould also increase the energy requiremen t of the Biofu e l s plan t. A rigorous cost-b enefit analysis of the differen t design options is needed here and ma y b e the sub ject of future w ork.
The follo wing s ection describ e s prosp ec ti v e materials and prop oses a metho dology for designing the HXs as w ell as p ossible limitations.
Prosp ectiv e Material s
Sev eral exp erimen tal studies fo cuse d on strength, o xidation resistance, and thermal fatigue ha v e b een con ducted to ev aluate the use of reaction b onded silicon carbine (RBSiC) and siliconised silicon carbide (SiC) as materials for HX fabrication. These exp erimen tal studies ev alu ate d the use of RBSiC and SiSiC b y fabri cating and testing a cross flo w HX for reco v ering w aste heat from com bustion p ro ducts of a gas fired furnace. While b oth materials demonstrated go o d o xidation resistance, exp osure to reducing en vironmen ts cause d the formation of pits on the SiC surface. F urthermore, b oth S iC materials ev aluated are capable of withstand ing exp osure to 1200 Q C. Ho w ev er, in comparison to Allo y 617, whic h will b e use d for fabricating PCHE1 and PCHE2, b oth c eramics ha v e lo w er thermal conductivi tie s. RBSiC demonstrated sup erior temp erature strength and o xidation prop erties[103].
Heat Rates Extractable of Fl uid Streams from the Hydrogen Plan t
T able 12 sho ws the prop erties of the fluid stream s at th e Hydrogen plan t. These streams ma y b e used to pr o duce steam at 182 Q C for use at the biofuels plan t. The thermal efficiency of high temp erature gas furnaces can b e increased b y reco v ering heat from the com bustion pro ducts b efore exhausting them to the atmosphere. As indicated earlier, prop osals for heat r e co v ery include the use of ce ramic heat exc hangers fabricated using RbS iC or SiSiC.
Cho osing a Heat Exc hanger Design
W e prop ose that b oth shell and tub e designs as w ell as cross flo w ceramic HX designs b e ev aluated for this purp ose. The former h a v e lo w er effectiv ensses but are c heap er to fabricate, the latter due to higher
effectiv enesse s require the use of few er materials but are also lik ely to ha v e higher fabrication costs. F urther, the narro w er c hannels of a cross-flo w heat exc han ge r are lik ely to create fouling probl e ms.
If it is found that the shell and tub e HX is the design of c hoice, it is prop osed that smaller shell and tub e units b e fabricated and connected in parallel with an adequate n um b er of redundan t units. This w ould allo w p erio dic main tenance while k e epin g the Pro cess Heat system online.
Prop osed Design Metho dology
W e prop ose dev eloping matlab co des for ev aluating b oth the cross -flo w and shell and tub e designs.
F or eac h design, it is prop osed that a hot and cold c hannel b e selected as the unit cell. The matlab mo del should split the c hann e l in to sev eral no des, use appropriate fluid prop erties at eac h no de and k eep trac k of fluid temp eratures and no dal heat transfer rates. This metho dology is similar to that emplo y ed b y the mo del use d in this w ork for designing Pro cess Heat PCHEs.
The heat extractable from either of the streams is sufficien t to pro duce the s team required for the biofu e ls plan t. Ho w ev er, use of the steam/h ydrogen stream adds complexit y to the design problem as the slip ratio for this stream can no longer b e appro ximated as b eing equal to 1 once the steam condenses.
The considerations outlined here should inform future w ork for designing this heat exc hanger.
7.7 Heat Sink
7.7.1 Purp ose, Lo cation, and Comp onen ts
In the ev en t of a core sh utdo wn, heat will still need to b e remo v ed from pl an t comp onen ts, therefore, helium will still b e pump ed for some time. The emergency heat sink forms a lo op with the PCHE at the core and exists in parallel to the lo op con taining heat s tor a ge and the heat exc hangers for the biofuels and h ydrogen facilities. V alv es directing helium out of the core heat exc hanger and in to the heat s tor age facilit y w ould redirect the flo w in to this heat sink where it w ould in teract with a plate t yp e heat exc hanger that has sea w ater from the Gulf of Mexico as its cold fluid. Suc h a heat exc hanger should b e deriv ativ e of the marine heat exc hangers of Sondex Inc [19], as w ell as made of titanium to prev e n t sea w ater corrosion. After losing heat to the sea w ater, the helium wil l b e pump ed bac k t o the core heat exc hanger.
7.7.2 Heat Rate and Design
Based on the deca y heat c alcul ation [88] sho wn in Equation 26, th e 3575 MWt n uclear reactor is exp e cted to pro duce a deca y heat of 120 MW after 30 seconds, with 40 MW of dec a y heat b eing pro duced one hour after sh utdo wn. In one hour after a sh utdo w n from full p o w er, the core has pro duced 179 GJ of deca y heat whic h a v erages to around 50 MW.
Q dot ( t ) = Q dot, 0 · 0 . 066 · t − 0 . 2 (13)
·
k g − K
Ob viously not all of this deca y heat is going thr ough the pro ces s heat system. A more realistic analysis tak es the maxim um heat from the core PCHE as Q dot, 0 = 35 M W 9 units = 315 M W . Av erage heat in the one h our after sh utdo wn i s 5 MW. F ollo wing the lead of the San O nofre Nuclear Generating station in southern California, the maxim um allo w able temp erature difference b et w e en the s ea w ater in le t and outlet temp eratures w as set at 19 C F (10 C C) [22]. Giv en w ater’s heat capacit y of appro ximately 4190 J in the temp erature range of in terest and the heat transfer rate seen in Equation 27, a sea w ater flo w rate of 120 kg/s, or 455 gallons/s is needed.
7.7.3 En vironmen tal Concerns
Q dot = m dot · c p · D T (14)
The En vironmen tal Protection Agency’s Clean W ater Act Section 316 Thermal Dischar ges states that the lo cation, design, construction , and capacit y of co oling w ater in tak e structures m ust reflect the b est tec hnology a v ailable f o r minimizing adv e r s e en vironmen tal impact [2]. The goal of limiting thermal p ollution can b e approac hed in m ultiple w a ys. Heat treatmen ts that reduce fouling should b e p erformed gradually to driv e
a w a y marine life - oth e r w i s e the lo cal fish w ould exp erience a thermal sho c k and some w ould die. Limiting the sea w ater temp erature increase, taking i n colder w ater from deep, offshore lo cations, and utilizing diffusers at the ou tle t to recom bine the sea w ater and reac h thermal equilibriu m quic k er are all steps that help preserv e the lo cal en vironmen t.
7.8 Pro cess Heat System Cos ts
7.8.1 PCHE
The cost of a PCHE is prop ortional to the metal fraction f m . The metal fraction for PCHE1 and PCHE2 w as calculated using Equation 28 and w as found to b e 62.95% [54].
f m = 1 − 8 P t
(15)
F ollo wing the calculation of f m , a metho dology sim i lar to Dostal’s [54] w as used to estimate the capital cost of the PCHEs. These capital costs are outlined in T able 18. The cost of Allo y 617 w as n ot a v ailable at the time of this calc u lation and a $ 30/kg v alu e of steel as quoted b y Heatric [54] w as used for this cost estimation. Allo y 617 costs ha v e b een requested from Hun tington Metals and these cost calculations will b e up dated when the Allo y 617 c osts b ecome a v ailable.
|
HX |
Material |
V olume (m 3 ) |
W eigh t (kg) |
Cost (K $ ) |
|
PCHE1 |
Allo y 617 |
8.2548 |
69,010.128 |
2,070.30384 |
|
PCHE2 |
Allo y 617 |
15.6 |
130416 |
3,912.48 |
T able 13: PCHE capital cost
7.8.2 Circulator
The circulator will ha v e to b e custom b uilt so the exact price is not kno wn, but curren tly man ufactur e d circulators that accomplish similar goals cost around $ 250,000.
7.8.3 Piping
The cost p er meter of the en tire piping apparatus (from All o y 625 all the w a y out to the Gemcolite 2600) is appro x im ately $ 36,507. T h e r e f ore , the total cost for 40 m of piping (the total amoun t that will b e in the system) is around $ 1.46 M.
7.9 Heat Storage Details
Heat will b e stored in order to b e able to k eep the LBE molten f or an extra t w o w e eks after the core stops pro ducing enough deca y heat to do so on its o wn. A laten t heat thermal energy storage system w as c hosen instead of sensible heat, as it can rele ase heat at a constan t temp erature and can store more energy p er unit m ass of storage material [52]. Some widely used categories of phase c hange materials (PCMs) include paraffins, salt h ydrates, and me tals. The PCM selected m ust ha v e a melting p oin t in the desired op erating temp erature range of the system [27]. This w as the most imp ortan t criteria in pic king a PCM for the system, as the op erating temp erature is dep enden t on the heat pro vided b y the core and demanded b y the h ydrogen and biofuels p lan ts.
The lo cation of th e storage device is to b e directly after the heat exc hanger with the core secondary lo op where the temp erature of the system is app ro ximately 500-600 C C. Molten salts - name l y , c hlorides and carb onates - w ere therefore examined as p oten tial PCMs as the range of melting temp eratures for molten salts w as mos t app ropriate for this syste m [73] and other categories of PCMs melt at a lo w er tem p erature than is desirable. Lithium c hloride, LiCl, w as ev en tually c hose n as it is p ossible to push the op erating temp erature of the system just ab o v e the melting p oin t of LiCl, allo wing sensible heat storage past the melting p oin t to b e neglected.
7.9.1 Lithium Chlori de
LiCl w as c hosen as the PCM as its melting p oin t of 605 C C w as most appropri ate for the pro c ess heat system. This is also matc hed to the outlet temp erature of the core secondary lo op (606.5 C C). A list of the relev an t prop erties of LiCl can b e seen in T able 14. The heat capacities and laten t heat of fusion w ere calculated using the Shomate equation for LiCl [18].
|
Prop ert y |
V alue |
|
Melting P oin t |
605 C C [24] |
|
Δh C of F usion |
470 kJ/kg [18] |
|
c P (solid) |
1.132 k J / kg − K [18] |
|
c P (liquid) |
1.538 k J / kg − K [18] |
|
k (thermal conductivit y) |
0.534 W / m − K [112] |
|
Densit y |
2,068 kg/m 3 [18] |
T able 14: Re l e v an t ph ysical prop erties of LiCl
As can b e se en from T able 14, the heat capacit y of LiCl is significan tly less than the l a t e n t heat of fusion. Heating the salt from 603 C C to 604 C C therefore stores less than 0.1% of the energy that heating from 604 C C to 605 C C b ecause of the phase c hange at 605 C C.
The heat capacit y of the solid is imp ortan t b ecause, initially , the LiCl will need to b e heated up from ro om temp erature (˜20 C C) to 605 C C. This in v olv es storing heat as sensible heat, so it is b etter f or the heat capacit y of LiCl to b e smaller as it will in v olv e less time and energy to heat the LiCl up to its me lti ng p oin t. This will tak e around 660 kJ/kg of LiCl, whic h is significan t compared to the 470 kJ/kg stored in the phase c hange. Ho w ev er, it is only imp ortan t to the system design that there is negligible sensible heat storage after the melting p oin t. When the reactor is sh ut do wn and th e primary lo op needs t he stored energy to k eep the lead bism uth molten, the laten t heat, whic h is a hi ghe r qualit y heat than the sensible heat, is a v ailable immediately , as opp osed w aiting for the material to co ol b efore reac hing the melting p oin t. The sensible heat stored b et w een the melting p oin t and the op erating temp erature of 606.5 C C can b e ignored as it is less than 0.5% of the energy stored as laten t heat.
7.9.2 Allo y 20
Because the PCM melts, it cannot ha v e direct con tact with the h e l ium flo wing thr ough the system as this could result in flo w of the molten PCM and con tamination of the lo op. Therefore, the PCM m ust b e con tained in a cladd ing. It is imp ortan t to selec t a material that is c hemically compatible with the PCM. The u s e of a molten salt me an s corrosion is a p oten tial problem [151] b ecause molten salts are go o d conductors of
electricit y . Allo y 20, a nic k el-c hromium-molyb den um allo y , w as c hosen as the con tainmen t material in order to alleviate this problem as it is compatible with LiCl [119] d ue to its extreme resistance to c hloride ion corrosion [12]. The c hemical comp osition of the allo y can b e se en in T able 15.
|
Minim um (%) |
Maxim um (%) |
|
|
Nic k el |
32.5 |
35.0 |
|
Chromium |
19.0 |
21.0 |
|
Molyb den um |
2.0 |
3.0 |
|
Manganese |
none |
2.0 |
|
Copp er |
3.0 |
4.0 |
|
Silicon |
none |
1.0 |
|
Carb on |
none |
.06 |
|
Sulfur |
none |
.035 |
|
Phosphorus |
none |
.035 |
|
Niobium |
none |
1.0 |
|
Iron |
balance |
balance |
T able 15: T h e comp osition of Allo y 20. Adapted from [11].
7.9.3 System D esign
A custom heat exc hanger w as des i gned for the purp ose of storing heat in the LiCl. The design, whic h w as inspired b y plate t yp e heat exc hangers, can b e seen in Figure 31.
Figure 31: The basic design of the storage heat exc hanger
The LiCl will b e stored in slabs, stac k ed on top of eac h other with gaps for the helium to flo w th rough, and con tained b y Allo y 20. A more detailed view can b e seen in Figure 32.
Figure 32: A view of the cross-sectional fl o w area of the storage heat exc hanger.
The o v erall dimensions o f the tank will b e 20 m long, 18 m wide, and 11.41 m tall. There will b e 10 slabs of LiCl, with eac h b eing 20 m long, 18 m wide, and 1.13 m tall. Eac h slab will b e con tai ned b y Allo y 20, with a thic kness of 1 cm (dimension t w all in Figure 32). The gap b et w e en s lab s will b e 1 cm. There will b e a total of 11 gaps, allo wing the helium to flo w b oth o v er and under the top an d b ot tom slabs of LiCl.
Assumptions
The heat exc hanger w as designed suc h that eac h of the slabs of LiCl could b e treated equally . A n um b er of assumptions w ere made (refer to Figure 32 for dimensions):
1. Ab o v e the melting p oin t, heat is stored in the LiCl on ly as laten t heat; sensible heat is neglected. Only energy stored as laten t heat will b e used for energy storage.
2. No con v ection o ccurs within the LiCl; heat transfer within the LiCl o ccurs through conduction only .
3. t P C M << L P C M ; the end effects of eac h slab can b e ignored.
4. The helium temp erature is isothermal for an y giv en v alue of y .
5. During sh utd o wn, the LBE e n ters the shell-and-tub e heat exc hanger at 140 C C; heating it to 150 C C is sufficien t to k eep i t molten throughout the primary lo op.
6. The storage will b e used to k eep the LBE molten for up to t w o w eeks.
7. The a v erage mass flo w rate of the LBE during sh utdo wn is 1,900 kg/s.
8. There are no transmission losses b et w een the storage heat exc hanger an d th e shell-and-tub e heat exc hanger with the primary lo op.
Storage La y ou t : Charging
When the storage system is b eing c harged, the h ydr oge n and biofuels plan ts wil l b e v alv ed off from the main lo op so the helium cycle through the core heat exc hanger and the storage device only . Figur e 33 sho ws this la y out graphically . A preheater will use electricit y from the grid to heat the incoming helium to 705 C C, allo wing the c harging time to b e drastically reduced from on the order of y ears to on the order of da ys.
Figure 33: The c harging la y out for the storage heat exc hanger.
Storage La y out : Disc harging
When th e system is disc hargin g, the main lo op will b e v al v ed off and the v alv es to the primary lo op will op en as sho wn in Figure 34. Helium will flo w through the lo op b et w een the storage h e at exc hanger and a shell-and-tub e heat exc hanger with the core’s primary lo op.
Figure 34: The storage lo op during energy disc harge.
7.9.4 PCM Sizing and Design Mass and V olume of LiCl
If the LBE has an a v erage mass flo w rate of 1,900 kg/s in sh utdo wn, the shell-and-tub e heat exc hanger will need to transfer 2.85 MW of energy from th e helium to the LBE. With a mass flo w rate for helium of 132 kg/s, this results i n the helium co oling b y 25 C C. Under as sumpt ion 8, th e storage device th us needs to disc harge
×
× ×
×
2.85 MW of energy . Und e r assumption 6, this mea n s that the storage device m ust store 3 . 447 10 6 MJ of energy . If the energy that is to b e used is laten t heat only (assumption 1), using the prop erties in T able 14, the mass of LiCl needed can b e calculated to b e 7 . 334 10 6 kg, or 3,546.31 m 3 . T en slabs of LiCl with the dimensions 20m x 18m x 1.13m results in 4,068m 3 , for an actual total of 8 . 41 10 6 kg 3 . 9 . 954 10 6 MJ of energy .
Flo w Prop erties of Helium
m 2 K
The Reynolds n um b er for the helium flo wing through the heat exc hanger w as calculated to b e in the turbulen t flo w regime. The Gnielinski relation [149] w as u s ed to calculate the Nusselt n um b er. The Nusselt n um b er w as then used to calculate the con v ection co efficien t of 241.589 W .
Pressure Drop
In calculating the pressure drop acros s the storage heat exc hanger, the effects of gra vit y and acceleration w ere ignored. Th us, only the form losses and friction losses w ere tak en in to accoun t. The friction pressure drop w as calculated according to Equation 16 where f is the friction factor, L is the length of the heat exc hanger, D e is the equiv alen t diameter of the flo w area, G is the mass flux, and ρ is the densit y of helium
[40].
ˆ x G 2
e
D p f r ic = f D
dx (16)
2 ρ
T o find the friction factor, f, the roughness of th e slabs w as estimated using the v alue for roughness of stainless steel allo ys [43]. The roughness and Reynold’s n um b er w ere used to find the v alue of f on Mo o dy’s c hart [108]. The v alue of f w as found to b e 0.035. The pressure drop due to f riction w as calculated to b e 0.778 MP a.
The form losses can b e calculated according to Equation 17 where k f or m is the form loss co effic ien t and v 1 and v 2 are the v elo cities directly upstream and do wnstream of the lo cation of in terest, resp ectiv ely [40].
ˆ G 2 v 2 − v 1
D p f or m = k f or m 2 ρ dx + ρ 2
(17)
k g − K
The form acceleration losses, the pressure drop due to the v elo cit y , and densit y c hange w ere calculated for the inlet and the outlet. The densit y w as calculated using Equation 18 where R specif ic is the sp ecific gas constan t of helium, 2080 J .
p = ρR specif ic T (18)
The form acceleration losses w ere calculated to b e -0.001 MP a, whic h are negligible compared to the friction drop.
If the p ip es are constructed so that there are no abrupt c hanges in pip e shap e and area, the con tribution due to the form loss co efficien t should b e small, although not negligible, in comparison to the friction co efficien ts. An estimate of 1 MP a will b e used as the total pressure drop across the storage heat exc hanger. Due to this large pressure drop, a circulator will b e lo cated directly do wnstream of the storage device to raise the helium bac k up to its normal op eration al pressure of 5 MP a.
Charging Time
·
The preheater w i ll create a greater d iff erence in temp erature b et w een the melting p oin t of LiCl and the temp erature of the heating fluid. F or this geometry of flat slabs of LiCl, the distance b et w een the phase fron t and the heating s u rface can b e describ e d using Equation 19 where k is the thermal c on ductivit y of the LiCl in W/m-K and D h f is the laten t heat of fusion in J/m 3 [104].
s char g e
( t ) = 2 t k · ( T mel t − T sur f )
D h f
(19)
×
Setting s to equal 0.565 m (half the thic kness of the PCM as the phase fron t will b e propagating in from b oth sides), the time required to fully c harge the PCM is equal to 2 . 9 10 6 s, or 33 da ys and 12 hours. Ho w ev er, this complete c hargi ng time w on’t b e needed ev ery time the plan t is started up; the PCM will b e insulated suc h that it sta ys w arm for at least a mon th durin g the refueling p erio d, as this insulation is necessary for the purp ose of k eeping the LBE molten a mon th after sh utdo wn.
Disc harging
The heat flux densit y (p o w er/unit area) of the storage device can b e calculated using Equation 20 where q” is the total p o w er divided b y the total surface area, k is the thermal conductivit y of the LiCl in W/m-K and s is the distance b et w een the phase fron t and the surface of the LiCl [104]. This d e r iv es from the thermal resistance mo del of heat transfer.
q ”( t ) = k · ( T mel t − T sur f )
s
(20)
m c
The temp erature of the helium can b e solv ed as a function of x as s een in Equation 21 where β = π hD e
dot p
[23].
T 0 − T ( x ) = e − β x (21)
T 0 − T H e,in
T 0 is 605 C C as this is the temp erature of the PCM when melted. The temp erature difference of the helium will th us b e equal to the p o w er absorb ed d ivided b y the mass fl o w rate and heat capacit y . Using Equations 20-21, an appropriate helium inlet temp erature (and therefore outlet temp eratur e ) can b e solv ed for that will allo w for the correc t amoun t of p o w er to b e dra wn from the storage during disc harge.
Com bining Equ ations 20-21 and in tegrating o v er b oth the length of the PCM (20 m) and the time for disc harge (1,209,660 s) giv es Equation 22.
ˆ 1209660 s ˆ 20 m k LiC l ( T mel t − [ T mel t − T H e,in ] e − β x ) E stor e d
0 0
2 t
LiC l
melt H e,in
f h f
sl ab
LiC l
k [( T − T ) e − β x ] · dx · dt = n w
(22)
The total energy stored is di vided b y the w id th of the LiCl slabs b ecause the total p o w er densit y w as in tegrated only in one di m ension: th e length of the LiCl slab. By p lugging in the kno wn v alues, the inlet temp erature of the helium that allo ws the storage to b e disc har ge d o v er t w o w eeks can b e found. The inlet temp erature of the helium w as found to b e 419.73 C C; this inlet temp erature remo v es all th e laten t energy from the storage o v er the course of t w o w eeks.
As the phase fron t mo v es, for a constan t helium inlet temp erature, the p o w er dra wn from the storage will c hange. The outlet temp erature of helium and the p o w er exc h anged to the LBE will therefore also v ary with time. Th us, it will b e necessary to v ary the mass flo w rate of the LBE in order to k eep the c hange in temp erature of the LBE equal to 10 C C. In tegrating the p o w er exc hanged in the LBE-helium heat exc hanger o v er the t w o w eek disc harging p erio d m ust giv e the total energy used from storage d uring disc harge as sho wn in Equation 25.
ˆ
12 0 9660 s
m dot ( t ) · c p,LB E · D T · dt = E stor ed (23)
0 s
The mass flo w rate of the LBE at an y giv en time should b e equal to the p o w er b eing released b y the storage at that time divided b y th e heat capacit y of LBE and the c hange in te mp erature of the LB E (10 C C). The p o w er b eing released b y the storage at an y giv en time can b e found b y in tegrating only o v er x in Equation 22, and then m ultiplying b y the n um b er of slabs and the width of th e slabs:
ˆ 20 m k LiC l ( T mel t − [ T mel t − T H e , in ] e − β x )
0 2 t LiC l
melt
H e,in
q ( t ) = [
k [( T − T ) e − β x ] · dx ] · n sl abs · w LiC l (24)
f h f
As ev erything in this equation is kno wn, the mass flo w rate of the LBE as a function of time can b e ev aluated:
√
q ( t ) 9 . 986 · 10 6
m dot ( t ) = c
This function is plotted in Figure 35.
p,LB E
· D T
=
LB E t
(25)
Safet y
If one of the slabs w ere to burst, the helium lo op w ould b e con taminated with molten LiCl. While the PCM cladding w as c hosen to b e resistan t to corrosion with LiCl, the piping m at e ri al and the heat exc hangers in the rest of the main lo op w ere not. Con tamination of the rest of the lo op could also re sul t in LiCl solidifyin g in the pip es b ecause th e temp erature of the helium drops b elo w the freezing p oin t of the LiCl as it mo v es through the h e at exc hangers to the biofuels and h ydrogen facilities.
Therefore, it is imp ortan t that if a slab op ens and molten LiCl is exp osed to th e helium that the storage device is is ol ate d and the h e li um flo w is rerouted around the storage d e v ic e. The pressure thr oughout the in terior of the slabs and the flo w of the LiCl within the slab s h ould b e monitored; if a s lab w ere to burst, the pressure will drop and molten LiCl will flo w to w ards the leak, rev ealing that th e slab has b egun to leak.
Figure 35: The mass flo w rate of the LBE as a function of time when the storage device disc harges.
In suc h an ev en t, the storage device can b e v alv ed off and th e helium flo w directed to anot he r pip e flo wing to w ards the h ydrogen and biofuels plan ts (whic h has b een incorp orated in to the pro cess heat design as seen in Figure 25). This will also result in less of a pressure drop as the pressure drop of the storage heat exc hanger is large compared to that of the piping. Therefore, in the ev en t that the storage is v alv ed off, the circulator that is lo cated directly after storage should b e sh ut off.
7.9.5 Heat Storage Summary
|
Result |
V alue |
|
Mass of LiCl |
6 8.41 × 10 kg |
|
V olume of LiCl |
4068 m 3 |
|
T otal Stored Energy (Laten t) |
6 3.9539 × 10 MJ |
|
Pressure Drop (friction) |
0.778 MP a |
|
Pressure Drop (form acceleration) |
-0.001 MP a |
|
T otal Pressure Drop (Estimate) |
1.0 MP a |
|
Reynolds Num b er - Helium |
7,152.2 |
|
Charging Time |
33 da ys , 12 hours |
|
Disc harging T i m e |
2 w eeks |
|
Helium Inlet T emp erature (Char ging) |
705 C C |
|
Helium Inlet T emp erature (Steady State) |
606.5 C C |
|
Helium Inlet T emp erature (Disc harging) |
419.73 C C |
|
Mass Flo w Rate (Charge/Disc harge/Steady) |
132 kg/s |
T able 16: A summary of the heat storage r e sults
7.10 Circulator
The circulator should b e lo cated in the area with the greatest pressure drop in the system in order to recup erate the helium pressure. F or this design, a circulator w i ll b e placed at the 1 MP a drop immediately follo wing the heat s tor ag e system as sho wn in Fi gure 25. Here, the pressure needs to increase from 4 MP a bac k up to 5 MP a, therefore, the circulator will need to ha v e a compression ratio of 1.25. Since the mass flo w rate is high (149.85 kg/s) in addition to the pres sur e ratio, there are no curren tly m an ufactured circulators on the mark et to place in this system. A custom circulator [87] is a p ossibilit y for this plan t. Th is h ybrid circulator is exp ecte d to use arou nd 4 MW due to its large mass flo w rate and pressure ratio.
7.11 Heat Sink
7.11.1 Purp ose, Lo cation, and Comp onen ts
In the ev en t of a core sh utdo wn, heat will still need to b e remo v ed from pl an t comp onen ts, therefore, helium will still b e pump ed for some time. The emergency heat sink forms a lo op with the PCHE at the core and exists in parallel to the lo op con taining heat s tor a ge and the heat exc hangers for the biofuels and h ydrogen facilities. V alv es directing helium out of the core heat exc hanger and in to the heat s tor age facilit y w ould redirect the flo w in to this heat sink where it w ould in teract with a plate t yp e heat exc hanger that has sea w ater from the Gulf of Mexico as its cold fluid. Suc h a heat exc hanger should b e deriv ativ e of the marine heat exc hangers of Sondex Inc [19], as w ell as made of titanium to prev e n t sea w ater corrosion. After losing heat to the sea w ater, the helium wil l b e pump ed bac k t o the core heat exc hanger.
7.11.2 Heat Rate and Design
Based on the deca y heat c alcul ation [88] sho wn in Equation 26, th e 3575 MWt n uclear reactor is exp e cted to pro duce a deca y heat of 120 MW after 30 seconds, with 40 MW of dec a y heat b eing pro duced one hour after sh utdo wn. In one hour after a sh utdo w n from full p o w er, the core has pro duced 179 GJ of deca y heat whic h a v erages to around 50 MW.
Q dot ( t ) = Q dot, 0 · 0 . 066 · t − 0 . 2
(26)
·
k g − K
Ob viously not all of this deca y heat is going thr ough the pro ces s heat system. A more realistic analysis tak es the maxim um heat from the core PCHE as Q dot, 0 = 35 M W 9 units = 315 M W . Av erage heat in the one h our after sh utdo wn i s 5 MW. F ollo wing the lead of the San O nofre Nuclear Generating station in southern California, the maxim um allo w able temp erature difference b et w e en the s ea w ater in le t and outlet temp eratures w as set at 19 C F (10 C C) [22]. Giv en w ater’s heat capacit y of appro ximately 4190 J in the temp erature range of in terest and the heat transfer rate seen in Equation 27, a sea w ater flo w rate of 120 kg/s, or 455 gallons/s is needed.
7.11.3 En vironmen tal Concerns
Q dot = m dot · c p · D T (27)
The En vironmen tal Protection Agency’s Clean W ater Act Section 316 Thermal Dischar ges states that the lo cation, design, construction , and capacit y of co oling w ater in tak e structures m ust reflect the b est tec hnology a v ailable f o r minimizing adv e r s e en vironmen tal impact [2]. The goal of limiting thermal p ollution can b e approac hed in m ultiple w a ys. Heat treatmen ts that reduce fouling should b e p erformed gradually to driv e a w a y marine life - oth e r w i s e the lo cal fish w ould exp erience a thermal sho c k and some w ould die. Limiting the sea w ater temp erature increase, taking i n colder w ater from deep, offshore lo cations, and utilizing diffusers at the ou tle t to recom bine the sea w ater and reac h thermal equilibriu m quic k er are all steps that help preserv e the lo cal en vironmen t.
7.12 Piping
There are three lo cations that helium piping will need to b e installed:
1. from the core heat exc hanger to heat storage (5 m),
2. from heat storage to the biofu e ls heat exc h anger (5 m),
3. and from t he biofuels heat exc hanger bac k to the core heat exc hanger ( 30 m).
The temp erature of the w orking fluid (helium) as it is exiting the heat exc hanger with the secondary lo op wil l b e at 606 Q C. The a v erage high and lo w temp eratures for South e rn T exas are 27 Q C and 17 Q C, resp ectiv ely [41]. This high discrepancy in temp eratures requires t w o la y ers of insulation. Gase ou s helium flo w also requires a flo w liner to reduce friction. Due to th e high pressure of this system (5 MP a), there needs to b e a pressure b oundary pip e in place as w ell. The la y out for the pip e is sho wn in Figure 36.
Figure 36: Helium Pip e La y out. Adapted from [ 87]
|
Length of Pip e (m) |
T emp eratu re Drop ( Q C) |
Pressure Drop (P a) |
|
5 |
0.007 (negligible) |
56.87 |
|
30 |
0.041 (negligible) |
2047.28 |
T able 17: Loss es through helium transp ort pip es
The helium gas flo ws in the inner circle with a mass flo w rate of 149.85 kg/s. The next la y er is an Allo y 625 [71] flo w liner of a thic kness of 50 mm. The next la y er i s an aerogel la y er[106]1 mm thic k . F ollo wing the aerogel la y er is a 50 mm thic k Allo y 304 stainless stee l la y er [47]. The outer la y erthat in teracts with atmospheric conditions is a ceramic insulator from RSI called Gemcolite 2600 [75]of thic kness 1mm.
The applied ho op stress on this 50 mm thic k Allo y 625 is appro ximately 81 MP a [152]. The des i gn limit for Ha ynes man ufactur e d Allo y 625 is around 350 MP a [71] so this is w ell b elo w design limits and a safe thic kness to use.
Using accepted equations for temp erature and press u re losses, kn o wn prop erties ab out eac h of the mate rials, and calculated parameters in the pro cess heat s y s tem, the losses in T able 17 w ere found.
7.13 Pro cess Heat System Cos ts
7.13.1 PCHE
The cost of a PCHE is prop ortional to the metal fraction f m . The metal fraction for PCHE1 and PCHE2 w as calculated using Equation 28 and w as found to b e 62.95% [54].
f m = 1 − 8 P t
(28)
F ollo wing the calculation of f m , a metho dology sim i lar to Dostal’s [54] w as used to estimate the capital cost of the PCHEs. These capital costs are outlined in T able 18. The cost of Allo y 617 w as n ot a v ailable at the time of this calc u lation and a $ 30/kg v alu e of steel as quoted b y Heatric [54] w as used for this cost estimation. Allo y 617 costs ha v e b een requested from Hun tington Metals and these cost calculations will b e up dated when the Allo y 617 c osts b e6come a v ailable.
|
HX |
Material |
V olume (m 3 ) |
W eigh t (kg) |
Cost (K $ ) |
|
PCHE1 |
Allo y 617 |
8.2548 |
69,010.128 |
2,070.30384 |
|
PCHE2 |
Allo y 617 |
15.6 |
130416 |
3,912.48 |
T able 18: PCHE capital cost
7.13.2 Circulator
The circulator will ha v e to b e custom b uilt so the exact price is not kno wn, but curren tly man ufactur e d circulators that accomplish similar goals cost around $ 250,000.
7.13.3 Piping
The cost p er meter of the en tire piping apparatus (from All o y 625 all the w a y out to the Gemcolite 2600) is appro x im ately $ 36,507. T h e r e f ore , the total cost for 40 m of piping (the total amoun t that will b e in the system) is around $ 1.46 M.
8 Hydrogen
8.1 In tro duction
After ev aluating m ultiple metho ds of h ydrogen pro duction discussed earlier, the t w o b est options for this h ydrogen pro duction facilit y w ere determined to b e the t he UT-3 pro duction pro c ess and high temp erature steam electrolysis (HTSE). The UT-3 h ydrogen pro duction pro cess is attractiv e b ecause it has b een pro v en commercially for large-scale pro duction of h ydrogen [32] and had manageable material concerns com p are d with other thermo-c hemical w ater splitting metho ds suc h as the sulfur-io dine pro cess [157]. The other leading approac h, HTSE, is attractiv e due to the simplicit y of the pro cess relativ e to the UT-3 metho d: an electric p oten tial is main tained to separate steam in to its constituen ts, h ydrogen and o xygen. Also, HTSE do es n ot pro duce en vironmen tally harmful b ypro ducts and th us satisfies one of the primary ini tiativ es to minimize greenhouse gas emissions from this facilit y . Ho w ev er, the primary motiv ation for the decision to utilize the UT-3 pro ces s w as that HTSE had not b een demonstrated on the sc al e of h ydrogen pro duction required for the biof uel pro duction plan t, 7.9 kg s -1 , and scalings of steady-state electrical p o w er requiremen ts for a large scale HT S E h ydrogen pro duction plan t w ere not definitiv e. A description of the UT-3 h ydr o gen pro duction plan t design and the reasoning for the e v en tual ab andonmen t of this ap proac h in fa v or of the HTSE pro ces s is pro v ided in App endix E.
Ho w e v er, since the UT-3 pro cess inheren tly requires substan tial surplus steam, or another high heat capacit y flu id, to carry r e actan ts and pro ducts throughout th e cycle, the amoun t of electrical p o w er required to heat the reacting and excess steam from temp eratu re a v ailable, 560 C, to 760 C w as ultimately impractical relativ e to the electrical output of the lead-bism uth reactor implemen ted in th is design; substan tial additional electrical p o w er w ould ha v e to b e imp orted to this UT-3 h ydrogen pro duction plan t. Therefore, HTSE w as reconsidered as a viable option. Though the temp erature of steam required for efficien t steam electrolysis is also up w ards of 800 C, the efficiency of separating steam at that temp erature along with no longer requiring excess steam as a w orking fluid of a system pro v es the elec tr ic al requiremen ts of this t yp e of h ydrogen pro duction plan t to b e more practical than a UT-3 plan t. In an effort to further reduce electrical p o w er requiremen ts in this HTSE plan t, a design using regenerativ e heating i s prop ose d to minimize the e lectrical p o w er requi re men ts required to raise steam f rom 560 C to 800 C. Ultimately , utilizing stac k ed s oli d o xide electrolyzer cells, an energy efficien t HTSE plan t has b een implemen ted in this design to supply the biofuel pro duction plan t with the necessary h ydrogen input rate of 7.9 kg s -1 .
8.2 High T emp eratur e Ste a m Ele ctr oly sis (HTSE)
8.2.1 HTSE Pro duction Plan t
Mass Flo ws and Thermal Energy Calculations A high temp erature steam electrolysis system has b een designed with recup erativ e heating in an effort to minimize the steady-state electrical requiremen ts for a h ydrogen p ro duction rate of 7.9 kg s − 1 required b y the biofuel pro duction plan t. The design, along with mass flo ws and p o w er requiremen ts is presen ted in Figure 37.
Figure 37: HTSE Hydrogen Pro duction Plan t
Steam is pro vided b y pro cess heat system at a temp erature of 559 C and pressure 0.105 MP a, with a mass flo w of 88.07 kg s − 1 . The temp erature of the steam from pro cess heat is limited b y the output temp erature of the lead-bi s m uth reactor; ho w ev er, a temp erature of 800 C is required for high temp erature electrolysis to b e an efficien t means of h ydrogen pro duction [59]. Th us, the mass flo w of steam coming from the pro cess heat system m ust b e heated to 800 C b efore en tering the w ater splitting solid o xid e electrolyzer cells (SOEC) to ac hiev e necessary steam con v ersion efficiencies.
Using simple energy considerations, the p o w e r required to raise this mass flo w of steam to 800 C is f ound using
Q ˙ = m ˙ ( h H 2 O , 800 C − h H 2 O , 559 C ) = 47 . 9 M W
Th us, initially b e f ore the recup erativ e heating system i s in op eration, the ste am will b e raised to 800 C through electric heating. Assuming a 40% efficiency of electric heaters, the initial electrical requiremen t for heating 88.07 kg s − 1 of steam from 559 C to 800 C w as found to b e 119.8 MW. This required electrical p o w er is the desired p o w er to off set through recup erativ e heating pro ce ss that will b e discussed.
Steam at 800 C en ters a steam splitter, whic h will partition the 88.07 kg s − 1 mass flo w in to a parallel structure of SOECs, whic h has b een depicted in the flo w c hart as three parallel SOEC units. Ho w ev er, the n um b er of solid electrolyzer units in par allel will b e m uc h larger, and the precise v alue will need to b e determined in future study . Assuming a steam con v ersion of 80% in the SOECs, whic h seems reasonable from other mathematical mo dels of the SOEC pro cess, [77] the mass flo ws of the exiting H 2 O ( g ) /H 2 ( g ) mixture and O 2 ( g ), whic h are passiv ely separated in the SOECs, are found stoic himetrically using the kno wn w ater electrolysis pro cess, is presen ted in T abl e 19.
|
Material |
|
|
H 2 O ( g ) |
17 . 57 |
|
H 2 ( g ) |
7 . 9 |
|
O 2 ( g ) |
62 . 6 |
|
T otal |
88 . 07 |
T able 19: SOEC Outlet Mass Flo ws
The temp eratures of the output steams are increased b y 36 C due to t he op eration of the SOECs at areal curren t densities of 7000 A m − 2 . This relativ ely large curren t densit y leads to a do cumen ted increase in temp erature during the electrolysis pro cess [77] whic h enab les the p ossibilit y of recup erativ e heating since the output stream temp eratures are higher than the required input steam temp erature of 800 C. Th us, during startup, this plan t will rely solely on electric heating to raise t he temp erature of steam from 559 C to 800 C, but once recup erativ e heating b egins, the electrical requiremen t can b e greatly offset using reclaimed thermal p o w er from the output streams of the plan t. The separate exit streams of H 2 O ( g ) /H 2 ( g ) and O 2 ( g ) are passed through t w o heat exc hangers with secondary lo ops of w ater and steam in order to co ol the exit pro ducts b efore sending the streams to the biofuel pro d uction plan t, th us reclaiming the energy th at w ould b e lost from th e system otherwise. Heat exc hanger system 3 reduces the temp erature of the H 2 O ( g ) /H 2 ( g ) from 836 C to 120 C to a v oid substan ti ally differen t flo w rates b et w een gaseo u s h ydrogen and liquid w ater, and 836 C to 71 C for O 2 ( g ) for and the resulting p o w er extracted b y heat exc hangers (2) and (3) are presen t in 37.
Th us, the amoun t of p o w er remo v ed from the exiting streams of pro ducts is 159 . 4 M W . Tw o secondary systems using a t w o phase liquid-to-steam heat remo v al cycle are implem en ted to claim this reco v erable p o w er. W ater from external sources en ters the heat exc hanger systems (2) and (3) at 35 C, and is heated through a series of heat exc han ge rs to 800 C. Though it w ould b e desirable to heat the exiting secondary steam flo w to 836 C, the practical exit temp erature is limited to 800 C b y finite heat exc hangers. Ev en though the total reco v erable p o w er is greater than the required thermal p o w er to raise the incoming stream of s team at 559 C to 800 C, due to temp erature constrain ts b et w een 800 C and 580 C for the secondary system in h e at exc h anger 1, only 19.78 MW out of the 43.77 MW required thermal p o w er to raise ste am from 559 C to 780 C thermal p o w er can b e transferred from the secondary streams to the input stream, raising the temp erature to 661 C. T h e remaining p o w er required to raise the temp erature of inpu t steam from 661
C to 800 C is supplied through constan t electrical p o w er, while once again assuming 40% efficiency , w ould b e 70.24 MW. Th us, o v erall, the recup erativ e heating cyc l e s reduce the electrical p o w er requiremen t f or this plan t b y 49.6 MW while still assuming a 40% electrical heater efficiency .
Through the recup erativ e heat pro cess, heat e x c hanger system 1 transfer 19.78 MW from the recup erativ e heating lo ops to the inlet stream of steam, raising the temp erature of the steam flo wing at 88.07 kg/s from 559 C to 661 C. Ho w ev er, there is a total of 159.4 MW a v ailable from the recup erativ e heating system, th us, there is a surplus of p o w er in the secondary streams exiting heat exc hanger system 1 at 580 C of 139.6 MW. The recup erativ e heating output steam, whic h corresp ond to mass flo w of 39.66 kg s − 1 at 580 C, are sen t to pro cess heat f or their uti liz at ion to p o w er the biofuel plan t. Once the excess energy is extracted from the secondary s team b y the pro cess heat system, the no w liquid w ater at 35 C is returned to the h ydrogen pro duction plan t and cycles, forming closed secondary recup erativ e heating lo ops. Th us, not only is recup erativ e heating offsetting our electrical re q uireme n t for the h ydrogen pro duction plan t b y 49.6 MW, it is also p o w ering th e biofuels pro duction plan t.
Lastly , the co oled H 2 O ( l ) /H 2 ( g ) that exits heat exc hanger 3 at 120 C is co oled further in the condenser to acceptable ro om temp erature c ond itions for lefto v er w ater rejection at the end of the cycle. The t w o-phase mixture is separated naturally and the 7.9 kg s − 1 of H 2 (g) is pipp ed to the biofuel pro duction plan t. The o xygen output is not needed for an y purp ose in the biofuel pro duction plan t, and it could either b e sold or v en ted appropri ate ly to the atmosphere. P ossible uses of thi s o xygen, along with the abandoned idea of h ydrogen storage that w as not longer deemed required for this design, is presen te d in App endix F.
SOEC Electrical P o w er Requiremen ts An estimation for the SOEC p o w er requiremen t can b e found using the follo win g form ula used in previous studies whic h relate th e molar h ydrogen pro duction rate to the total curren t in the en tire SOEC system [63],
N ˙ H = I
2 2 F
where F is the F arada y n um b er (96487 J V − 1 mol − 1 ). The total molar pro duction rate of h ydrogen is 3910.9 mol s − 1 , and th us corresp onds t o a curren t of I = 754.7 MA. The electrolyzer cells w ould seem to op erate at roughly 0.54 V compared the the appro ximately 1 V [26] required for w ater electrolysis using ratios of the electrical energy requiremen ts depicted in Figure 38. The dotted line depicting the electrical energy demand is a c haracterization of the magnitude of the electric p oten tial required to split w ater molecules in to h ydrogen and o xygen. A t the phase transition from liquid to gas, and as steam temp erature increases, the amoun t of electrical p o w e r required to split w ater molecules b ecomes increasingly sm al le r . Using steam at 800 C, an appro ximate electrical p o w er requiremen t of 402.68 MW is required. Due to the uncertain t y in the actual efficiency of these cells relativ e to the ideal p o w er requiremen ts depicted in Figure 38, further analysis will b e required to presen t a more definitiv e electrical p o w er requiremen t for t he SOECs. Nev erthe less, this calculation of electrical p o w er requiremen ts will b e used as a lo w er b ound of th e electrical p o w er requiremen ts for this h ydrogen pro duction plan t. Som e studies suggest an electrical p o w e r requiremen t of
3.1 kWh p er normal m 3 of h ydrogen pro duced in similar planar stac ks of SOECs, whic h w ould require an electrical p o w er requiremen t of 983.1 MW for w ater splitting, coupled with the steady-state p o w er require men ts to raise input steam f rom 661 C to 800C, w ould lead to a total electrical p o w er requiremen t for the h ydrogen pro duction plan t of 1053.3 MW [77]. Though this electrical p o w er requiremen t is relativ ely large compared to the theoretical v alues, this figure is still reasonable on the scale of e l e ctricit y pro duction of this reactor system. This electrical p o w er requiremen t will b e used a conserv ativ e upp er b ou nd on the electrical p o w er requiremen ts of this h ydrogen pro duction plan t that has b een demonstrated e xp erimen tal ly . Th us, HTSE app ears to b e fa v orable approac h to large-scale h ydrogen pro duction, and future w ork in v olving more sophisticated sim ulations of HTSE h ydrogen pro duction and subsequen t optimization will more definitiv ely determine th e electrical p o w er requiremen ts for this h ydrogen pro duction plan t. I n addition, it is imp ortan t to note the h ydr o gen pro duction plan t will b e p o w ering the biofuel pro duction plan t with the surplus heat remo v e d from ou tle t streams of residual steam, h ydrogen, and o xygen, and th us will offset some of the elec trical p o w er requiremen ts that w ould b e required in the biofuel pro du c ti on p lan t without the recup erativ e heating system.
Courtesy of Elsevier, Inc., http://www.sciencedirect.com . Used with permission.
Figure 38: P o w er Requiremen t for W ater E lec tr olys i s [121]
8.2.2 Materials and Comp onen ts
SOEC The fundamen tal principl e b ehin d high temp erature steam electrolysis (HTSE) is that as the tem p erature of the steam increases, the energy required to disso ciate the H 2 O molecule decreases. Lik e standard w ater electrolysis, HTSE can pro duce large quan tities of emissions-free h ydrogen while also eliminating the need for exp ensiv e c hemical catalysts [161]. T o split the H 2 O molecule, a Solid Oxide Electrolysis Cell (SOEC) is used to apply an electrical p oten tial s u bstan tial enough to d is so ciate O 2 and H 2 .
→
39 is a sc hematic of an SOEC cell. The c ell unit consists of three distinct sections: a p orous, conductiv e catho de, an o xygen-ion-conducting electrolyte, and a p orous, conductiv e ano de. As sho wn, H 2 O is inciden t on the catho de and undergo es the c hemical reaction: H 2 O + 2 e − H 2 + O 2 − . The H 2 remains at the catho de while the applied electro c hemical p oten tial dra ws the o xygen through th e electrolyte to w ards the ano de. In doing so, the e- are lib erated from the ions allo w i ng the formation of molecular O 2 at the an o de. This molecule tra v els through the p orous mate r ial and is collected in a stream at the outlet of the cell [7]. The rate of h ydrogen pro duction directly corresp onds to the temp erature of the gaseous H 2 O en tering the unit y as w ell as curren t applied to the catho de/ano de [161].
SOEC w e r e initially dev elop ed in a tubular formation, c hosen as the optimal configuration to a v oid sealing p roblems [77]. The SOEC tub es w ould b e rep eated to ac hiev e greater h ydrogen pro duction resultin g in 10-cell and 24-cell configurations [63, 121]. It w as found through exp erimen ting that th e tub e formations required a longer curren t path, increasing the O hmic resistance within the ce ll . The planar formation of SOECs b oth shortens the curren t path as w ell as p ermits a high pac king densit y , making it the more effi cien t c hoice [77]. SOEC units no w consists of “stac ks” of cells, connected b y conductiv e in terconnects and ha v e b een op erated successfully up to 60 la y ers [63].
Before exploring the materials used in the SOEC, it is imp ortan t to iden tify r e qu irem en ts necessary for efficien t H 2 and O 2 separation:
1. The electrolyte m ust b e
(a) dense
(b) c hemically stabl e
Courtesy of Elsevier, Inc., http://www.sciencedirect.com . Used with permission.
Figure 39: SOEC sc hematic [105]
(c) ha v e h igh ion ic conductivit y (and therefore, high curren t efficiency , resulting in a high con v ersion efficiency f or the cell)
(d) gas-tigh t, suc h that recom binati on of H 2 and O 2 will not o ccur
(e) thin, to mini m ize the effects of Ohmic resis tan c e
2. Electro des should ha v e suitable p orosit y to supp ort gas transp ortation.
3. Thermal expansion co efficien ts should b e s imilar for the electro des and the electrolyte to minimize mec hanical stress within the cell.
4. In terconnect material m ust b e c hemically stable in reducing/o xidizing en vironmen ts.
Electrolyte Material Yttria-stabilized zirconia (YSZ) is the most common electrolyte material used in SOEC cells. ZrO 2 b oasts high o xygen ionic conductivit y , go o d me c hanical strength, and a high melting p oin t (˜2937 K). A necessary dopan t (Y 2 O 3 ) is added to pro vide stabilization for the lattice structure, consequen tly lo w e r ing ionic c ond uctivit y . T abl e 20 summarizes the other options for electrolyte materials. Though not ha ving the highest ionic condu c tiv it y , YSZ w as deemed to b e the b est c hoice o v erall for mas s pro duction of electrolysis ce ll s on the b as i s of cost and c hemical c ompatib ilit y in reducing e n vironmen ts. [105].
Catho de Material T o ac hiev e th e c hem i c al and ph ysical stabilit y required of the catho de in a highly o xidizing/reducing en vironmen t, the t w o viab le options for SOEC catho des are noble metals (e.g. Pt) and non-precious metals (e.g. Ni and Co). While b oth options p erform similarly , the hi gh cost of nob le metals mak es non-precious metal electro des the more app ealing option. S p ecifically , Ni is selected due to its high electro c hemical reactivit y as w ell as its abilit y to induce h ydrogen reduction. The m aterial i s com bined with the YSZ to form a cemen t lik e material used as the catho de [7].
|
Name |
T yp e |
Conductivit y (S/cm) |
Optimal T emp erature (K) |
Commen ts |
|
YSZ |
Stabilized zirconia |
0.13 |
1273 |
Ov e rall b est c hoice |
|
ScSZ |
Stabilized zirconia |
0.18 |
1273 |
Due to in tricate pro duction pro cedure, the cost of ScSZ is exorbitan t |
|
LSGM |
Dop ed LaGaO 3 |
0.17 |
973 |
Requires reduced op erating temp erature; problematic reaction b et w een LSGM and Ni |
|
GDC |
Ceria-based o xides |
0.10 |
1073 |
Chemically unstable in a reducing en v iron men t |
|
SDC |
Ceria-based o xide |
0.08 |
1073 |
Chemically unstable in a reducing en v iron men t |
|
BaCeO 3 |
Proton-conducting electrolyte |
0.08 |
1073 |
Lo w conductivit y |
T able 20: Summary of options for solid electrolyte materials
Ano de Material Lik e the catho de, only t w o classes of materials successfully function for use as the SOEC ano de: noble metals (less attractiv e due to cost) and electronically conducting mixed o xides. Lan than um stron tium manganate (LSM) exhibits a similar thermal expansion co efficien t to the YSZ electro de and sustains a high p oten tial drop across the SOEC. T o date, no other materials are explored for application in the SOEC cell [105].
8.2.3 Material C orr osion
Ov er time in an y construct mate r ials b egin to degrade. In the HTSE h ydrogen pro duction plan t th e t w o main concerns are corrosion of the piping and degradation of the e lectrolyzer cells. Ut ilization of corrosion resistan t ceramics that can withstand 800 C temp eratures could mitigate these concerns and increase the longevit y of this plan t. F uture w ork will require in v estigating p ossible c eramics for this pu rp ose.
Metallic Piping Corrosion A t high temp eratures man y materials exhibit unfortunate pr o cesses w h ic h lead to structural failures. T h e h igh temp erature steam necessary for the electrolyzers to r un, will inevitably tak e its toll on an y piping and turbines used to transp ort it. Material expansion, c r e ep, and o x idation are all problems that m ust b e addressed in the design [114]. F or these reasons, there is considerable in terest in nic k el based sup erallo ys whic h are more resistan t to these concerns. In suc h materials, a protectiv e la y er of c hromium o xide forms prev en ting direct con tact b et w een the metal and the high temp erature o xidizing en vironmen t. It has b een sho wn that the Ni-based sup erallo y Ha ynes 230 has sup erior o xidation resis tan c e when compared to other sup erallo ys [42]. This mate r ial could b e used in the p iping and heat exc hangers of the plan t.
Electrolyzer Corrosion Despite extensiv e researc h in t o longer-liv ed fuel cells, there is still m uc h w ork to b e done. Problems suc h as con tact b et w een differen t materials, deterioration of the electro des, and blo c king of reaction sites are all ongoing concerns to the lifetime of the electrolyzer cells. F uture researc h is required to optimize electrolyzer cells to op erate as w ell as p ossible for extended p erio d s of time, b oth to ensure prop er pro duction rates during sc heduled op eration, but also to minimize replacemen t costs that will negativ ely e ffect the long-term economics of this reactor system. Curren t problems to mitigate are as follo ws. The c hanges in temp erature caused b y heating up the electrolyzer will stress the materials as they expand at differen t rates. Thermal cycles m ust b e reduced as m uc h as p ossible to minimize thermal stresses that will degrade the p erformance and structural in tegrit y of th e se cells. F urthermore, c h rom i um from the in terconnects can p oison the electro des reducing the electrical conductivit y , and th us researc h i n to ceramic coatings can b e used to slo w this pro cess. The delamination of the electro de-electrolyte in terface in the o xygen electro de is another ma jor problem that m ust b e addressed through additional materials researc h. Silicon p oisoning on the electro de coming from the seal of the cells themse lv es or the steam can also in terf e re
with conductivit y . F or this reason demineralized w ater should b e used for t he steam and differen t coatings should b e considered in future w ork [60].
Economics Using estimates from Idaho National Lab oratory , the total capital cost w ould b e on the order of 1.3 billion dollars. Giv en an exp ected lifetime of the electrolyzer cells of 3 y e ar s , the y early replace men t cost for a third of th e cells is estimated at around 40 million dollars ann ually [8].
8.3 Other Design Considerations
8.3.1 Biofuels sh utdo w n
If the biofu e l s plan t sh u t do wn unexp ectedly , there w ould b e a large amoun t of h ydrogen that w ould need to b e dealt with safely . First, a v alv e in the first pip e whic h brings ste am from the pro ce ss heat section in to the h ydrogen plan t w ould close, so that no more h ydrogen w ould b e pr o duced. Sec on d, there w ould b e an emergency system to cycle the h ydrogen gas and steam bac k through the plan t. The v oltage w ould b e turned off so that the steam w ould remain steam and not b e split in to more h ydrogen. Small amoun ts of h ydrogen gas w ould slo wly b e bled off from the main stream and flared in safe amoun ts.
8.3.2 Core sh utdo wn
If the reactor sh u t do wn unexp ectedly , the HTSE plan t w oul d lose electricit y and the temp erature of the steam w ould decrease. Without electricit y , there w ould b e no v oltage and electrolysis w ould not o ccur. The plan t w ould stop pro ducing h ydrogen, but there w ould b e no ma jor safet y concerns in the ev en t of an unexp ected core sh utdo wn.
8.3.3 Mec hanical failures
There will b e a redundan t compressor to ensure that if a main compressor fails, the h ydrogen pro duction facilit y can con tin ue with routine op eration s . Ho w ev er, since the compressor condenses steam in to w ater, failure of b oth compressors w ould n ot cause catastrophic failu re in an y case.
8.4 Conclusions
The HTSE h ydrogen pro duction metho d w as u ltimate ly c hosen for this d e sign to supply the required 7.9 kg s -1 of h ydrogen for t he biofuel pro duction p lan t. Steam at 559 C en ters the h ydrogen pro duction plan t from the pro cess heat system and m ust b e heated to 800 C to ac hi e v e acceptable steam electrolysis efficiencies. The inlet steam te mp erature is increased to 800 C using b oth a recup erativ e h e ati ng system an d electrical heating. The o xygen, h ydrogen, and unseparated steam pro ducts are co oled after increasing to 836 C in the electrolyzer cell system , pro v iding 159.4 MW of total reco v erable p o w er. Only 19.78 MW is able to b e transferred to the i nlet steam due to differen tial temp eratu re limitations, raising the input steam temp erature from 559 C to 661 C, while the r e main ing 139.6 MW is reclaimed b y the pro cess heat system to p o w er the biofuel pro duction plan t. The rec u p erativ e h e ati ng system that w ould reduce the steady-state electrical p o w er requireme n ts b y a significan t 49.6 MW an d also pro vides exce ss heat to p o w er the biofuel pro duction plan t. Th us, the recup erativ e heating system is adv an tageous for o v erall energy efficiency of this n uclear reactor system b y reco v ering the thermal p o w er a v ailable in the outpu t streams of this h ydrogen pro duction plan t. The steady-state electrical p o w er requiremen t to raise the inlet steam from 661 C to 800 C is 70.24 MW. Steady-state electrical p o w er requiremen ts b oth to h e at inlet steam from 661 C to 800 C and to main tain the ele ctric p oten tial in the electrolyzer cells are conserv ativ ely estimated at 1053.3 MW, and will lik ely b e able to b e reduced with more sophisticated sim ulation s of this h ydrogen pro duction plan t design and subsequen t optimization in future w ork. It is also imp ortan t to note that an y increase i n the a v ailable pro cess heat temp erature ab o v e 559 C will lead to a reduction in electrical p o w er requiremen ts of this design as w ell, and th us higher reactor output temp eratures w ould b e adv an tageous for more energy efficien t HTSE h ydrogen pro duction designs.
9 Biofuels
9.1 Pro cess Ov erview
V arious biofuels pro du c ti on designs w ere considered, with parameters of greatest imp ortance b eing the biomass cost, a v ail abilit y , and comp etition with fo o d sources; the carb on emissions, tec hnical feasibilit y , and capital cost; and the qualit y , quan tit y , and commercial viabilit y of th e fuel pro du c ed. A Fisc her-T ropsc h (FT) man ufacturing plan t and refinery using switc hgrass feedsto c k w as c hosen as the most fa v orable w a y to pro du c e syn thetic gasoline and d ies el. A thermo c hemical route to FT fuels w as selected whic h uses gasification and gas-clean-up to form a pro ducer gas, or syngas, that is fed in to th e FT reactor. This approac h util iz es the a v ailable h ydrogen and steam resources, mini m izes c ap ital cost and main tenance requiremen ts, and results in a distri bution of h ydro carb ons that can b e refined in to gasoline and diesel blends. If implemen ted on a large scale, the estim at e d pro duction c osts for high e n e r gy densit y FT fuel could b e as lo w as $ 1/gallon [39], whic h also mak es it economically comp etitiv e. FT fuels are already b eing pro duced from biomass b y companies suc h as Ren tec h and Choren [39], giving them a high feasibilit y , and the w aste emissions of the pro cess - c harred ash from the gasifier, trace acid gases suc h as H 2 S, and some CO 2 remo v al [90] - are lo w o v erall.
The prop osed biofuels pro duction plan t is sho wn in Figure 40. As seen in the figure, the main sections of the plan t include a biomass preparation area, a Silv agas @ gasifier and tar remo v al unit, a Rectisol acid gas remo v al system, a bubble c olu m n Fisc her-T ropsc h reactor, a fractional distillation unit, a Sasol r e fi ning unit, and biogasoline and bio diesel storage sites. The ma jorit y of the biofuel plan t energy wil l b e suppl ie d b y e xcess pro cess heat f rom the n uclear reactor. Hydrogen for the FT and refining pro cesses will b e supp lie d from the h ydrogen pro duction plan t, and minor electricit y costs will b e co v ered using electricit y from the n uclear r e actor.
A feedsto c k flo w rate w as c hosen based on the amoun t of a v ai lable switc hgrass, pro cess heat, and h ydrogen resources. An estimated one million dry metric tons of switc hgrass can b e gro wn for a p o w e r plan t in one y ear, whic h has led to designs in the literature that use 3,500 tons/da y (tp d) of biomass, so switc hgrass a v ailabilit y w as not the limiting factor in our pl a n t design. The curren t biofuels plan t design uses 2,903 tp d or 24.38 kg/s of switc hgrass.
F or p o w er, electricit y will b e dra wn from the n uclear p o w er plan t for basic facilit y needs, while 40 MW of heat energy will b e sen t from the n uclear reac tor to th e biof uels pl an t in the form of H 2 O. This energy will b e used primarily to w arm the air and steam inputs in to the Silv agas @ gasifier, to p o w er the FT reactor, and to heat FT liquids durin g distillation. The biofuels plan t wil l also supply 19 MW of heat bac k to the pro cess heat system b y co oling the syngas immediately prior to acid-gas remo v al.
The h yd rogen plan t will sup ply H 2 to the refining pro cedure, whic h needs 7.9 kg/s for naph tha h ydrotreat men t, distillate h ydrotr e atmen t, w ax h ydrotreatmen t, and C 6 /C 5 isomerization [53, 51, 93]. All refining will b e conducted on-site to meet the targets of the o v erall design of the complex. The biofuels plan t exp ects to op erate at 20 hours a da y and will giv e prior n otic e whenev er p ossible b efore sh utdo wns, since h ydrogen pro duction m ust s t o p or v en t their gas whenev er biofuels pro duction pauses. The biofuels pl an t will further ensure that prop er heat dumps are managed whenev er pro duction stops and will c o ordinate with reactor sh utdo wns to ascertain that the necessary p ro cess heat is alw a ys a v ailable.
9.2 Switc hgrass
Switc hgrass ( p anicum vir gatum ) w as selected as the op timal bio energy crop b ecause of its v ery high energy densit y , abili t y to gro w in dry climates, repro ducibilit y on p o or land, and exclusion f rom use as a fo o d crop [96, 92, 147] . The places where switc hgrass can gro w are mapp ed in Figure 41. Among other alternativ e s for biomass feedsto c k, it has a more w ater-efficien t C 4 Carb on Fixation cycle and higher energy densit y of ligno cellulose. Up to one m i llion dry metric tonnes of switc hgrass could b e gro wn f or a p o w er plan t in one y ear [90], so the amoun t of biomass needed is limited b y other factors s u c h as the amoun t of heat a v ailable for gasification and h ydrogen a v ailabl e f or refining. It w as calculated that an input of 24.38 dry kg/s of feedsto c k at 20% moisture w ould b e required to fuel the curren t plan t design.
Figure 40: Sc hematic of prop osed biofuels pro duction plan t
Map courtesy of Pacific Northwest National Laboratory, operated by Battelle for the U.S. Department of Energy.
Figure 41: Sites for switc hgrass gro wth in the U.S.
9.2.1 Densification
Switc hgrass p ro duction and preparation will b e outsourced to farmers and their facilities to tak e ad v an tage of an extan t external net w ork, to remo v e the logistical burden of main taining on-site facilities, and to impro v e the transp ortation condition of the biomass it s elf.
First, unpro cessed switc hgrass bales ha v e a bulk densit y of around 150 kg/m 3 [128] and are not economical to transp ort. F ortunately , the microscopic structure of switc hgrass (SG) is p orous. When large amoun ts of pressure are applied to small surface areas of biomass, the particles b ecome compressed and fill in thes e spaces, and then f urther in teract with in termolecular b onding. The m oi s tu re serv es as a binding agen t for the lignin and cellulose, activ ating the b onding mec hanisms of the macromolecules at high pressure. F riction causes heating up to 90 Q C, whic h al lo ws the b onds to consolidate up on co oling [81]. The molecular b onding and the secondary s t ructure resulting from that b onding during the application of force main tains the p ellets in cohesiv e units suc h that loss of material due to disin tegration is negligible. A p elletizer at 137 MP a with a screen size of 3.2 mm can p ro duce p ellets at 12% moisture with a densit y of 1000 kg/m 3 [128], whic h is an order of magnitude higher than the ra w material. Immediately after harv esting, suc h a p elletizer will b e emplo y ed b efore switc hgrass is shipp ed to the plan t. Bales tak en d irec t ly from a switc hgrass farm will need to stored in a climate-con trolled, dry lo cation. Wh e n moisture con ten t is ac hiev ed, they will then b e hoisted b y mac hine from the holding facilit y in to a large hopp er, whic h will feed directly in to a grinder. This will allo w for smo other pressing, as the groun ded material will then pro ceed to the pneumatic p elletizer (see Fi gure 41). The p ellets will b e then b e arranged in to a transp ortable medium. In densified form, the bulk feedsto c k tak es up less v olume and allo ws f or truc ks to consisten tly haul more biomass p er trip. With few er trips, the CO 2 emissions during transp ortation are re d uce d , and th us the amoun t of greenhouse gases pro duced in our pro cess that are als o lo w ered. Accoun ting for 20% moisture, it w as calculated that 30.48 kg/s of p ellets w ere needed as an input, whic h accoun t for 24.38 kg/s (2903 tp d) of useable biomass.
9.2.2 T ransp ortation to Site
The capital costs of transp ortation are generally $ 0.028/ton/km for rail an d $ 0.137/ton/km for truc k s , with the latter ha ving a greater rate of c hange p er distance tra v eled but a lo w initial fixed cost [84]. The plan t should b e lo cated in a gro wing region suc h as Minnesota or T exas (See map in Figure 41) [101] so that shipping of feedsto c k b y truc k will b e under 200 km and meet optimal pri c in g. Although th e r e w ould b e a clearer adv an tage to railw a y for b oth long-distance transp ort and insertion of materials in to an con tin uous pro cess, it p oses negativ e implications for the plan t’s reaction time to reactor sh utdo wn p erio ds. Land v ehicles can b e dispatc hed most efficien tly in the ev en t of feedsto c k shortage and can b e readily halted if necessary . They also do not r e qu ire large c hanges to existing infrastructure [162].
The input pro c ess will r e ly on an unloading mec hanism and a temp orary storage system, whic h cannot b e fully automatic and wil l require some degree of man ual lab or. The unloading do c k will n e ed to c on s ist of sev eral ba ys and ha v e direct access to an area of lo w traffic congestion. P ellets from the te mp orary s tor a ge area will b e fed to a con v ey or b elt th at leads to a w eighing hopp e r , whic h will then feed directly in to the gasification c ham b er in fixed amoun ts for particular durations of the pro cess (see Figure 41).
Considering that gas ifi c ati on pro ce eds at a te mp erature of 682 Q C, there m ust b e considerable distance
b et w een the gasification c ham b er and the holding area, an d the con v ey ors and hoists will need to consist of firepro of material. In the ev en t of a reactor sh utdo wn, the holding area will need direct access to short- term storage. The engineers will need to sc hedule the reshipmen t of feedsto c k with the refueling of the core suc h that do wn time b et w een start-up of the reactor and that of th e b iofuels facilit y will b e v ery short. F urthermore, it is p oss i ble that shipmen ts ma y fluctuate with the p erennial gro w i ng cycle of switc hgrass , suc h that it ma y b e necessary to arrange for bac kup suppliers in cas e of em ergency shortage. All p ossible lo cations of suppliers m ust b e ev aluated for t ypical w eather conditions that ma y pr o v e troublesome to pro viding a consisten t source of feedsto c k. If this is a decisiv e factor, then the c ompl e x m ust tak e fo oting based on that lo calit y .
Before the biofuels facilit y can op erate, ground transp ortation m ust b e established b et w een the reactor complex and the energy crop source suc h that feedsto c k can b e deliv ered in a con trolled and consisten t manner. Pro cessing SG in to a dense form allo ws for a more economical transp ortation option and a v ery con trolled injection in to gasification pro cesses, whic h allo ws the o v erall pr o cess to b e more streamlined and manageable. Ho w ev er, as opp osed to the other sectors of the reactor complex, there are external v ariables that w eigh in to the a v ailabilit y of this biological resource, whic h will require considerable foresigh t and logistical co ordination during op eration.
9.3 Gasification
The switc hgrass will b e fed b y lo c khopp er in to Ren tec h’s paten ted S ilv agas dual fluidized b ed gasifier [58], where the biomass will b e con v erted to syngas. Silv agas w as selec ted b ecause of its comme r c ial a v ailabilit y , its op eration at atmos ph e r ic pressures, and the adv an tages of dual fluidized b ed gasification: gasification and com bustion o ccur in separate c ham b ers as sho wn in 42, whi c h prev en ts N 2 and CO 2 dilution of the syngas. In the com bustion c ham b er, preheated air com busts with switc hgrass c har and raises the air temp erature to 916 Q C. This heat is transferred to the gasification c ham b er via a fluidized sand b e d with v ery little air transfer, and the lefto v er “flue gas” is v en ted or collected. In the gasification c ham b er, switc hgrass gasifies in the presence of sup erheated steam and the hot sand b ed, forming a syngas of primarily CO, H 2 , and CO 2 . Unreacted switc h grass and co oled sand is sen t bac k to the com bustion c ham b er, wh ile useful syngas is sen t on to gas clean up an d the FT reac tor .
The gasification c h am b e r emplo ys switc hgrass input of 24.38 kg/s (dry) and sup erheated steam in put of
4.42 kg/s at 182 Q C, whic h react at the high temp e r a t ure of 682 Q C supplied b y the fluidized sand b ed. The main reactions whic h o ccur here are the
Boudouard reaction,
w ater-gas or steam reaction,
C + C O 2 ↔ 2 C O + 172 k J /mol (29)
C + H 2 O ↔ H 2 + C O + 131 k J /mol (30)
© ASME. All rights reserved. This content is excluded from our Creative Commons license. For more information, see http://ocw.mit.edu/fairuse .
|
Biomass T yp e |
Switc hgrass |
W o o d |
|
Carb on |
49.4% |
51.35% |
|
Hydrogen |
6.11% |
6.10% |
|
Oxygen |
44% |
42.29% |
|
Sulfur |
0.12% |
0.09% |
|
Nitrogen |
0.64% |
0.17% |
Figure 42: Sc hematic of silv agas gasification pro ce ss
T able 21: Mass distributions of w o o d and switc hgrass [39] methanation or h ydrogasification reaction,
C + 2 H 2 ↔ C H 4 − 74 . 8 k J /mol (31)
and gas-shift reaction,
C O + H 2 O ↔ C O 2 + H 2 − 41 . 2 k J /mol (32)
Other reactions are also presen t, but at slo w er rates and in smaller amoun ts. Since the total amoun t of minor comp ounds suc h as sulfur and nitrogen is less than 1% in switc hgr as s and steam , they do not pla y a significan t role in the gasification reactions. The listed c hem ical reactions o ccur at v arying rates and approac h differen t equilibrium lev els dep e n ding on the gasifier design. In the Silv agas pro cess, the carb on con v ersion rate is ab out 73% [34], and all the carb on comes from the sw i tc hgrass biomass. The comp osition of syngas formed from w o o d c hips in the Silv agas gasifier is kno wn, and is comparable to the exp ected syngas comp osition after switc hgrass gasification b ecause of the similarit y in com p osition of the t w o biomasses [39], as sho wn in T able 21. After gasification, syngas exits through the top of a cyclone to b e purified and then sen t to the FT reactor. The final syngas output will b e at a temp erature of 682 Q C and a rate of 33.2 kg/s [58].
Ab out 27% of the carb on con ten t do es not react [34]. This par tic u late biomass, called c har, is fi ltere d out b y the cyclone and s en t bac k to the com bustion c ham b er to b e burned. Heated air at 4 kg/s and 354 Q C, as calculated from the sp ecifications of the Silv agas paten t, en ters the com bustion c ham b er and c om busts with the switc hgrass c har to generate temp eratures of 916 Q C, whic h supplies the energy for the gasifier [58].
|
Comp ound |
% b y V olume |
Mass Flo w Rate (kg/s) |
|
CO |
47 |
10.97 |
|
H 2 |
18 |
0.303 |
|
CO 2 |
14.3 |
5.25 |
|
CH 4 |
14.9 |
1.99 |
|
C 2 H 4 |
4.7 |
1.10 |
|
C 2 H 6 |
1.1 |
0.276 |
|
other |
< 1 |
˜ 0 |
|
T otal |
100 |
19.89 |
T able 22: C omp osition of Syngas Output from Silv agas Gasifier [118]
© Nexant, Inc. All rights reserved. This content is excluded from our Creative Commons license. For more information, see http://ocw.mit.edu/fairuse .
Figure 43: T ypical acid gas remo v al pro cess after gas i fication of biomass without a com bustion c ham b er
This heat is transferred to the gasification c ham b er via a fluidized sand b ed with v ery little air transfer. Com bustion heats the flu idized sand b ed, after whic h the remaining ”flu e gas” exits the com bustor and is collected for separate disp osal.
The comp osition of syngas lea vi ng the Silv agas reac tor is compiled in T able 22 [118]. Other elemen t s suc h as nitrogen, sulfu r, and phosphorus are also presen t throughout the pro cess in small amoun ts, bu t th e ir v olumetric flo ws are not calculated b ecause they com p ose < 1% of the final syngas. Unfortunately , the ra w syngas can not y et b e sen t to the FT reactor b ecause carb on dio xide, and to a lesser exten t sulfur, nitrogen, and other trace elemen ts, w ould p oison the F T reaction.
9.4 Acid Gas Remo v al
Syngas cannot b e sen t directly to the FT reactor after Silv agas gasification b ecause it con tains v arious elemen ts, suc h as carb on dio xide and sulfur, whic h could p oison the FT p ro cess. In par tic u lar, CO 2 m ust b e remo v e d b ec au s e it in teracts with H 2 more quic kly than CO do es; the presence of CO 2 in teractions w ould lo w er the energy output of the FT reactor and pro duce larger, unfa v orable particles [115]. T race elemen ts suc h as sulfur and nitrogen, while to o scarce to b e n ote d in the literature, m ust also b e remo v ed to prev en t long-term buildup and damage. A sc hematic of a t ypical acid gas remo v al pro cess can b e seen in Figure 43. The ma jor steps are particulate remo v al, co oling and compression, and acid gas remo v al. Tw o steps from the sc hem ati c , tar refor m i ng and ZnO b e d sulfur remo v al, can b e b ypassed b ecause the Silv agas pro cess utilizes a com bustion to w er coupled to th e gasification, whic h minimizes tar creation. Once the acid gase s are remo v e d do wn to acceptable lev els, the syngas is sen t to the F-T reactor follo wing a heat addition and depressurization of the syngas to 15 bar at 245 Q C.
Syngas purification and conditioning b egins with th e separation of coarse particles in the Silv agas reactor. Initial purification is ac hiev e d using cyclones, whic h are large cen trifuges use d to separate solid particles from the syngas. These cyclones are used b efore the co oling of the syngas to prev en t condensation and dep osition [115]. The solid particles remo v ed are primarily tars, h ydro carb ons with C 10+ that ar e recycled bac k to the com bustor to further b e brok en do wn. The r e maini ng ra w syngas lea v es the cyclones at 682 Q C, and gas clean up con tin ues with syngas co oling to 107 Q C and remo v al of unreformed tar u s i ng a w ater scrubb er.
Courtesy of Plant Process Equipment, Inc. Used with permission.
Figure 44: Amine Acid Gas Remo v al Pro cess
A mass flo w rate of w ater is used from an outside source to lo w er the tem p erature of the syngas to 107 Q C o v er three stages. Syngas at 107 Q C and 1 bar is sen t to the Syngas V en tur i Scrubb er, C-200. This is a paten ted pro cess that uses mo ving w ater to remo v e ammonia, particulates, and halides from the syngas. The mass flo w rate is adjusted to ensure that the syngas is quenc hed to the desired temp erature of 107 Q C b efore it can b e sen t to the compressor [115]. An y remaining condensate in the syngas is remo v ed using a paten ted mac hine, the Syngas Compressor K O Drum, V-300. The co oled syngas is compressed to 30.7 bar with a four stage horizon tally split cen trifugal compressor with in terstage c o olers at 43 Q C, w h ic h increases the solubilit y of acid gases , b efore b eing sen t to the acid gas remo v al pro cess [115].
Lastly , t w o stages of acid gas clean up remo v es p ois on ous molecules, primarily H 2 S, NH 3 , and CO 2 , w h ic h can n e gativ ely affect the F-T reaction. The first stage, an amine remo v al plan t, filters the H 2 S conce n tration from ˜400 ppm to ˜10 ppm. Acceptable F-T conditions are 0.2 ppm, so a paten ted LO-CA T pro cess is next emplo y ed to further reduce H 2 S and CO 2 lev els.
The amine acid gas remo v al pro ce ss can b e seen in Figure 44 [13]. An amine is an organic comp ound that includes nitrogen that dissolv es with acid gases. The input is sour syngas, whic h refers to syngas th at con tains acid gases. The sour gas is first sen t through a cyclone separator as precautionary measure, in case the particulate remo v al pro cess did not remo v e all solid particles. Next, the syngas is sen t to an amine con tactor, diethanolamine (HN[CH 2 CH 2 OH] 2 ). This amine w as c hosen for the particularly lo w pr e ssure acid gas remo v al pro cess to minimize the o v erall net energy requiremen t and ac hiev e the desired CO 2 and H 2 S remo v al [115, 13]. In the amine con tactor, acid gas comp osed of amine, CO 2 , and H 2 S separates from syngas and is sen t to an amine regenerator, while the treated syngas is recyc l e d through an amine co ol e r and sen t bac k to the top of the amine con tactor. The amine regenerator separates amine from CO 2 and H 2 S, then reheats an d recycles it bac k to the exc hanger [13]. Carb on dio xide and sulfur are co oled in an amine co ole r and exp elled. CO 2 can b e sequestered through a n um b er of differen t pro cesses and H 2 S can either b e v en te d or sen t to a sulfu r reco v ery plan t to b e s old commercially .
Diethanolamine-treated syngas at 107 Q C and 30.7 bar is sen t to a paten te d LO-CA T pro ces s . This pro cedure, pictured in Figure 45, decreases the amoun t of H 2 S in the “sour” syngas from a concen tration of
˜10 ppm to less than 0.2 ppm [46].
Again, the input to b e cleaned is listed as a sour gas, ev en though it has just come from the amine remo v al plan t, and again, a cyclone is used as a b eginning filter. The syngas then en ters a c ham b er con taining o xygen flo w and an iron catalyst. Acid gases from syngas are remo v ed along with CO 2 and the pro ducts of H 2 O
Courtesy of Merichem Company. Used with permission.
Figure 45: LO-CA T acid gas r e mo v al p ro cess
|
Comp ound |
Mass Flo w (kg/s) |
|
CO |
10.97 |
|
H 2 |
0.303 |
|
CO 2 |
0 |
|
CH 4 |
1.99 |
|
C 2 H 4 |
1.10 |
|
C 2 H 6 |
0.276 |
|
H 2 S |
< .2 ppm |
|
T otal |
14.64 |
T able 23: Com p osition of Syngas Output after Acid Gas Remo v al and S Q (F e) [46]. Sulfur is remo v ed through the c h e mical reaction sho wn in Equation 33.
H 2 S + 1 / 2 O 2 → H 2 O + S Q (F e) (33)
Finally , the syngas fr om the LO-CA T pro cess needs to b e de p res sur iz ed do wn to 15 bar and heated to 245 Q C b efore it can en ter the FT reactor. A depressurizer will b e used to ac hiev e the correct pr e ssure conditions and the heat addition will come from the pro c ess heat group [115]. Clean syngas is sen t on to the FT r e actor at a mass flo w of 14.64 kg/s and with the comp ositions sho wn in T able 23; there is virtually no loss of gases apart from CO 2 and sulfur rem o v al.
9.5 Fisc her- T ropsc h Reactor
The Fisc her-T ropsc h (FT) reactor is the heart of the biofuels p ro duction plan t b ecause it con v erts syngas in to the long carb on c hains of gasoline and diesel f uels. The reactor is filled with inert oil in whic h F e catalyst particles are s u s p ended. Pro cessed syngas en ters from the b ottom of v essel through an inlet nozzle a n d is bubbled up v ertically in a c h urn turbulen t flo w regime to maximize the mass an d heat transfer [122]. As the syngas rises, the CO and H 2 gases form in to longer h ydro carb ons through the e xoth e r m ic reactions listed b elo w and the heat generated in these reactions is absorb ed b y v ertical co olan t tub es. A slurry phase bubble column FT reactor, pictured in Figure 46, has b een c hosen for its isothermal op erating condition and go o d heat transfer [100]. The reactor is 7 m in diameter and 30 m in heigh t and op erates at a temp erature of 240 Q C and pressure of 24.0 bar.
Courtesy of Elsevier, Inc., http://www.sciencedirect.com . Used with permission.
Figure 46: Slurry phase bubble reactor sc hematic [100]
© source unknown. All rights reserved. This content is excluded from our Creative Commons license. For more information, see http://ocw.mit.edu/fairuse .
Figure 47: The effect of feed ratio r = H 2 /CO on selectivit y at T = 300 Q C [100]
− −
Inside a FT r e actor, carb on mono xide and h ydrogen gas in teract in the reactions sho wn in Equati ons 34 and 35 to form long, straigh t h ydro carb on c hains called paraffins where ( C H 2 ) is a meth ylene group that p olymerizes in to larger molec u lar c hains [100] and Equation 35 is the comp eting w ater shift reaction whic h should b e minimized.
C O + 2 H 2 → − ( C H 2 ) − + H 2 O + 170 k J (34)
H 2 O + C O → C O 2 + H 2 (35)
The rate of these t w o reactions and the distri bution of their final paraffinic p ro ducts dep ends strongly on the c hoice of catalyst and the en tering H 2 /CO ratio of the syngas stream, whic h in turn go v ern the probabilit y of c hain gro wth, α , and to some exten t the reaction temp erature. The most commonly used FT catalysts are, in decreasing order of activit y , Ru, F e, Ni, Co, and Rh [142]. Althou gh Ru is the most activ e in pro ducing high C n um b er molecules, it is also v ery costly , so the second most activ e catalyst, F e ( α = 0.9) w as c hosen for our FT reactor [ 100]. Our design curren tly emplo ys a feed syngas ratio of around H 2 /CO = 2.0, whic h is dictated b y the biomass comp osition. The mass fraction of paraffins with carb on n um b er n and molecular form ula in final FT liquid can b e found using Equation 36 where α is the c hain gro wth probabilit y [122] that can b e calculated using Song et al’s mo del using temp erature and H 2 /CO ratio
[33] in Equ ation 37.
AS F
χ n = n (1 − α AS F ) 2 α n − 1 (36)
· − −
α = 0 . 2 3 + 0 . 63 [1 0 . 0039( T 533 K )] (37)
H 2 /C O + 1
Figure 47 and Figur e 48 sho w effects of H 2 /CO ratio and t he c hain gro wth pr obabilit y factor on pro duct selectivit y in an FT reactor.
Other significan t parameters in an FT reactor are the carb on con v ersion ratio, the sup erficial v elo cit y of syngas, and the n um b er of co olan t tub es required to k eep temp erature con trolled.
Carb on con v ersion ratio is the fraction of CO molecules in the f e ed syngas stream that is con v erted in to larger paraffins [100]. The con v ersion factor increases with increased catalyst concen trati on, as sho wn
Figure 48: ASF distribution for c hain gro wth [100]
in Figure 49, and is also influenced b y reactor temp erature and H 2 /CO ratio of feed syngas, among other factors. A carb on con v e r s i on ratio of ε smax = 0.4 is considered the limit for feasible commercial op eration of F-T plan t [122]. Our design has a ratio ε s = 0.35, whic h maximizes con v ersion while sta ying a reasonable margin b elo w the limit.
The total flo w rate of syngas through the reactor is c haracterized b y the sup erficial v elo cit y , whic h is
defined as the v olumetric flo w rate of syngas p er unit cross sectional area of reactor, where n ˙ denotes the
molar flo w rate and P and A represen t the p re ssure and the cross sectional area of the reactor resp ectiv ely , is sho wn in Equation 38.
V ˙
U s = A =
n ˙ R T P A
= 0 . 05 m/s (38)
.Giv en our reactor size of 7 m diameter b y 30 m heigh t, along with our syngas fl o w rate of 14.64 kg/s, from Equation 38, the sup erficial v elo cit y is found to b e U s = 0.12 m/s.
Curren tly , the FT reactor temp erature has b ee n c hosen to op erate at 245 Q C based on oth e r published
mo dels [100]. This temp erature is con trolled b y the co olan t flo w rate, whic h dep ends on the n um b er of v ertical co olan t tub es needed to absorb heat and to main tain c h urn turbulen t flo w regime. As illustrated in Figur e 50, the n um b er of tub es needed increases with faster flo w rate and larger catalyst concen tration. The n um b er of co olan t tub es that corresp onds to the sup erficial v elo cit y of this design is 6,000 and pitc h is
0.15 m.
Heat transfer co efficien t is estim ated to b e 1,450 W/m 2 from Figure 51. This giv es 41 MW for the heat exc hanged with the co olan t a s sho wn in Equation 39.
Q ˙ = 6000 · 30 m · π · 0 . 05 m · 1450 W / m 2 = 41 . 0 M W (39)
F or co olan t flo w, w e prop ose to use a one lo op system at pressure 33.5 bar whic h corres p onds to the saturation pressure at 245 Q C. The co olan t en t e rs the FT reactor as a condensed saturated liquid at 245 Q C and lea v e s the FT reactor as a saturated v ap or. The mass flo w rate in this primary heat exc hange lo op is
23.5 kg/s. Ho w e v er, if the heat w ere to b e dump ed in to th e en vironmen t, w e m ust follo w the EP A’s limit on
Courtesy of Elsevier, Inc., http://www.sciencedirect.com . Used with permission.
Figure 49: Effect of catalyst concen tration on con v ersion ratio
Courtesy of Elsevier, Inc., http://www.sciencedirect.com . Used with permission.
Figure 50: Effect of sup erfi c i a l v elo cit y , catalyst concen tration on the n um b er of co olan t tub es [122]
Courtesy of Elsevier, Inc., http://www.sciencedirect.com . Used with permission.
Figure 51: Heat transfer co efficien t as a fun c ti on of catalyst c on c en tration
|
F raction |
Carb on Num b er |
Boiling P oin t ( Q C) |
Mass Flo w (kg/s) |
|
Liquid P etrol eum Gas |
1-5 |
< 40 |
2.02 |
|
Ligh t Naph tha |
5-8 |
30-90 |
2.79 |
|
Hea vy Naph tha |
8-12 |
90-200 |
2.30 |
|
Distillate |
12-20 |
200-300 |
2.65 |
|
Hea vy W ax |
20+ |
300-350 |
1.46 |
T able 24: Pro ducts of Fisc her-T ropsc h Pro cess and Relativ e Boiling P oin ts
the p ermitted tem p erature c hange of co olan t w ater; that is, the co olan t w ater temp erature can b e no more than 20 Q F o v er th e in tak e w ater temp erature [5]. Giv en that follo wing suc h guidelines w o u ld require a w ater mass flo w r ate of 888 kg/s, the heat ge n e r ate d fr om the FT reaction ma y ha v e to b e v en ted or sen t bac k to the heat pro ces sin g system in order to b e feasible.
In conclusion, our iron-catalyzed FT p ro cess will pro duce 11.22 kg/s of v arying l e ngt h carb on c hains, as brok en do wn in T able 24. A fair amoun t of n aph tha and distillate, precursors of gasoline and diesel, is pro duced, but a significan t amoun t of hea vy w ax is also pro duced whic h will need to b e h ydro crac k ed in to more u s eable pro ducts. All the FT liquids pro duced in the reactor are sen t on to distillation, where the ligh ter carb on c hains are separated out, and then to the refinery t o increase qualit y of the pro ducts using h ydrogen gas.
9.6 F ractional Distillation and Refining
Outputs from the FT reactor are still primarily in the form of straigh t carb on c hains called paraffins. Therefore, to mak e FT liquids ready for commercial use , they m ust first b e distilled and refin e d to impr o v e the fuel qualit y . The first step in refining crude FT liquids is distillat ion b y b oiling at 350 Q C. The v arious h ydro carb on comp onen ts of crud e oil are called fractions, whic h are separated from one anoth e r b y a pro cess of fractional distillation. F ractional distillation op erates on the principle that differen t substances b oil at differen t temp eratures. As the gases rise up the di s til lation to w e r , they co ol and settle out in to distilled fractions with the hea viest comp ounds, whic h ha v e the highest b oilin g p oin t, settling out fi rs t . T able 24
Figure 52: Distiller Sc hematic
sho ws the differen t p ro ducts of lo w temp erature Fisc her-T r opsc h Syn thesis and the b oiling p oin ts.
Crude oil is stored in tanks with the capacit y of 20 million gallon s b efore distillation. In s id e the fractional distillation c ol umn, horizon tal bubble pl ate s lo cated at differen t heigh ts collect the fractions, whic h co ol and condense at the prop er b oiling p oin t. The crud e oil is initially v ap orized in the ab s ence of air via a furnace at 350 Q C causing most of the oil to ev ap orate. As the v ap or mo v es up th e column, eac h fraction condenses at a differen t temp erature and liquid fr ac tion s are collected in the tra ys. The h ydr o carb on con ten t with a b oiling p oin t higher than 350 Q C is funneled in to a v acuum distillation unit whic h re-distills at a higher pressure. Heat exc hangers are used throughout this pro cess to recycle heat. The hot n aph tha whic h has a lo w b oiling p oin t is co oled while the crude oil is preheated b efore en terin g the furnace.
Distillation separates the FT liquids in to th ree differen t pro duct streams according to molecular mass. The liquid fraction with b oiling p oin t of less than 180 Q C is sen t out as naph tha, from 180 Q C to 320 Q C as distillate, and the remainder as a hea vy w ax stream [90]. Distillation is a standard p ro cess in c hemical engineering and design of a distiller is relativ ely simple compared to other units in the whole design. Figure 52 sho ws a simple sc hematic of a generic distiller design. The v essel consists of tra ys for condensing se p arated pro ducts an d heat exc hangers. The three streams of separated FT h ydro carb ons are then sen t to refinery unit.
The refin ing pro cess will closely follo w a Lo w T emp erature Fi s c her-T ropsc h refinery design presen ted b y Betc hel [159] in Figure 53.
F our separate segmen ts of the design will require h ydrogen gas input from the h ydrogen facilit y of resp ectiv e m ass flo w r ate s, m ˙ H 2 , sho wn b y the follo wing equations. Eac h equation has a fudge-factor included dep ending on the pro cess .
F or naph tha h yd rotreatme n t [53],
© ASME. All rights reserved. This content is excluded from our Creative Commons license. For more information, see http://ocw.mit.edu/fairuse .
Figure 53: Betc hel Lo w T emp erature Fisc her-T ropsc h Refinery Design
m ˙ H 2
= 60 ϕ H 2
ϕ N aphtha
(40)
|
F or distillate h ydrotreatmen t [53], |
||||||
|
m ˙ H 2 |
= |
140 ϕ H 2 ϕ diesel |
(41) |
|||
|
F or w ax h ydrotreatmen t [93], |
||||||
|
m ˙ H 2 |
= |
1200 ϕ H 2 ϕ w ax |
(42) |
|||
|
F or C 6 /C 5 isomerization [51], |
||||||
|
m ˙ H 2 |
= |
32 k g m ˙ C 4 ton C 4 |
ton 907 k g |
(43) |
||
Using Equations 40-43 along with the p ercen t yield of eac h pro cess, the input of h ydrogen in to the refining pro cess w as able to b e calculated. Platin um/ alumin um o xide catalytic reforming p ro cess will b e utilized to optimize the con v ersion of naph tha to dies el. The refin ing pro cess in its en tiret y fo cuses on four main ob jectiv es: h ydrogenation of olefins, remo v al of o xy ge n-con taining comp ounds, h ydroisomerization (to increase the o ctane n um b er of the diesel fuel) and h ydro c r ac king of n-paraffins. This refining pro cess has b een sho wn to increase the researc h o ctane n um b er as high as 95.2 (the minim um rating in the United States is 87).
The fractionated distillation pro ducts are r e f orme d using a v ariet y of metho ds, including h ydrogenation of olefins, remo v al of o xygen and sulfur con taining com p ounds, h y droisome r iz ati o n to impro v e the o ctane rating, and h ydro crac king of n-paraffins to isoparaffins. As c an b e s een in Figure 53, H 2 gas is emplo y ed in most of these refining pro ces ses. An H 2 input of 7.9 kg/s is needed to refin e our ra w bi ofuels in to consumer-ready biogasoline and bi o diesel. Our estimated plan t outp ut is 662.4 tons/da y or 4637 barrels/da y of biogasoline and bio diesel, whic h is enough fuel to fill ab out 26,000 cars/da y at 10 gallon s /car.
9.7 Biofuels Results Summary
As part of a n uclear p o w er plan t coupled to h ydrogen and biofuels pro duction, a biofuels plan t w as designed with the capabilit y to handle 2,903 metric ton s /da y of switc hgrass feedsto c k at 20% moisture and to output
662.4 metric tons/da y or 4637 barrels/da y of biogasoline and bio diesel, along with other b ypro ducts suc h as syn th e tic natural gas. In termediate steps in the pro c ess include gasification and com bustion, particulate and tar remo v al, acid gas clean up, FT reactions, distillation, and refinin g. The plan t will b e lo cated in Harris Coun t y , T exas, and will utilize a significan t air input, natural w ater reserv oir, pr o cess heat from a n uclear p o w er plan t, and h ydrogen from a h ydrogen pro duction facilit y in order to ac hiev e its goals. Carb on dio xide outp uts will b e s equestered and all other w astes will b e di s p osed of as resp onsibly as reasonably ac hiev able. The plan t will create significan t jobs in the lo cal comm unit y through hir ing of switc hgr as s farmers, switc hgrass transp ortation and handling p ersonnel, and other biofuels plan t w ork ers. Switc hgrass exp enses are estimated to b e $ 120k/da y , while plan t rev en ue at curren t p rice s is estimated at $ 850k/da y .
P art IV
Conclusions
The final design calls for a 3,575 MWt lead-bism uth eutec ti c (LBE) co oled fast reactor with a secondary sup ercritical CO 2 system. The core utilized uranium mononitride (UN) as its fuel and a ferritic/martensitic steel (T91) [67] with a 100 micron corrosion resistan t la y er as its cladding material [138]. The primary heat exc hangers are lo cated at the top of the reactor v essel and u s e s h e ll - an d-tub e tec hnology . The outlet temp erature is 650 C C and the inlet is 484 C C. The secondary lo op consists of a sup ercritical carb on dio xi de Bra yton cycle, whic h e x trac t s heat from the primary LBE lo op b y means of three h e at exc hangers. The total cycle efficiency is calculated at 41.7% and pro duces at least 1 GW e for the grid. A zigzag c hannel configuration will b e used for the first prin ted circuit heat exc hanger (P C HE ) and a straigh t c hannel c on figuration w i ll b e used for the second one in order to minimize pressure drop s on the w ater side. Hydrogen will b e pro duced via the UT-3 pro du c tion pro cess and high temp erature steam electrolysis (HTSE). As part of a n uclear p o w er plan t coupled to h ydrogen a n d b iofuels p ro duction, a biofuels plan t w as d e signed with the capabilit y to handle 3,500 metric tonnes p er da y of switc h grass feedsto c k at 20% m oi s tu re and to output 1,450 metric tonnes p er da y of biogasoline and bio diesel. In termediate steps include gasification and com bustion, particul ate and tar remo v al, acid gas clean up, FT reactions, distillation, and refining.
10 F uture W ork
10.1 Short- T erm F uture W ork
10.1.1 Pro cess Heat
The mass flo w rate of the LBE should b e ca l c u late d as a function of time, so that the c hange in temp erature of the LBE alw a ys sta ys at a constan t 10 C C at the LBE/Helium heat exc hanger. Once the function is kno wn, the pumps can b e v ar ie d during sh utdo wn to matc h the mass flo w rate of the LBE to the function.
Insulation m ust also b e considered. In steady state, th e temp erature of the helium going through the storage device cannot drop b elo w 605 C C, or else it will start dra wing laten t heat from the helium. The helium en ters the storage device at 606.5 C C; therefore, th e losses to the en vironmen t cannot exceed 1.5 C C, or 1.016 MW (with a helium mass flo w rate of 132 kg/s). The storage device m ust therefore b e prop erly insulated to ensure that it loses less than this amoun t of heat to the en vironmen t.
Insulation will also b e imp ortan t to consider for the p erio d b et w een s h utdo wn and when the stored energy will b e used. The deca y heat of the reactor should k eep the LBE molten for appro ximately t w o w eeks; during this time , there w il l b e no helium flo wing through the storage device to dra w heat out, but it m ust still b e protected against en vironmen tal losses .
Because th e slabs of LiCl are so large, it is imp ortan t to ensure that t he y are structurally se cure. The effects of a s u pp ort system on the flo w prop erties of helium should b e examined, in order to increase the stabilit y of the system.
Adding another salt to the LiCl to reduce the melting temp erature sligh tly should b e examined. This w ould create a l arge r margin b et w een the op erating temp erature and the m eltin g p oin t, s o that if the op erating temp erature drops, the system do es not s ol idify . This could also help reduce the cost of the LiCl, as it w ould result in less LiCl n e eded.
10.1.2 Biofuels Plan t
It ma y b e most viable for the biofuels plan t to b e scaled-up to maximize gasoline and diesel pro duction. A larger plan t w ould generate more jobs for f arme r s , dr iv ers, and plan t w ork ers. It could also p oten tially increase our estimated profit of $ 1.4 million p er da y . This w ould require more h ydr oge n in put. Also, the biofuels plan t could utilize o xygen from the h ydrogen p lan t for the gasification step in the biofuels pro ce ss. Another pro ces s that needs to b e lo ok ed in to more is recycling flue gas, h ydrogen sulfide, carb on dio xide, and other p oten ti al w astes.
10.2 Long- T erm F uture W ork
10.2.1 Core
With the curren t design of the core, a n um b er of optimizations and more in-depth analysis w e r e left out. On the core side, up on c ompletion of the thermal analysis, it b ecam e apparen t the curren t cladding material w ould not b e suitable for the goals of the reactor. A t th e op erating temp eratures, the clad, T91 s tai nles s steel, cannot last longer then t w o y ears. F urthermore, there is not m uc h margin to clad failure in case of an acciden t scenario. A w a y to address the longevit y of the clad w ould b e to lo w er the op erating temp erature, ho w ev er, this still do es not lea v e m uc h margin to failure. As suc h, researc h will need to b e done to find alternativ e cladding materials that can b etter withstand the temp e r atures and last longer. A ma jor adv an tage of the design c hosen is its breeding capabili tie s, but withou t suitable cladding material, the lifetime of the core is not limited b y fuel a v ailable but instead b y structural issues. A sim i lar t w o la y er clad material to the one c hosen in this design could b e c hosen as these class of materials ha v e sho wn the abilit y to withstand LBE corrosion. Another p oten tial option are sili c on carbides.
Other issues on the core side of the reactor th at require more study are the natural circulation during sh utdo wn, alternativ e fuel t yp es, re-arranging con trol ro d p ositions to b etter balance p o w er across core, and altering BOL zoning to aid in depletion effects. During sh utdo wn the natural circulation of the core v aries dep ending on the difference in temp erature across th e core and the actual v alues of temp eratures themselv es (due to viscos it y effects). Dev eloping a mo del to correlate ho w deca y heat affects b oth of these parameters will require more analysis. While for an y giv en difference and temp eratures the natural circulation has b een calculated, it is difficult to sa y ho w those will c hange o v er time and whether or not the natural circulation will main t ain sufficien t ho w to co ol the core after sh utdo wn.
The c h oic e to switc h t o UN as the fuel material instead of UO2 w as done b ecause of the sup e r ior thermal c haracteristics of the UN fuel. P oten tial alternativ e fuel materials could pro v ide the same b enefits as UN but without the dra wbac ks to using this material, notably the requiremen t to enric h the nitrogen. Suc h materials include uranium carbides. Uranium carbide pro vides a b etter thermal conductivit y (23 W/mK vs. 21 W/K) but has a lo w er melting p oin t (2300 C C vs 2800 C C). Analysis should b e to see if UC’s thermal p erformance is sup erior or comparable to UN’s. If that is the case then switc hing to UC w ould b e preferred, ho w ev er if w orse than UN the cost of the fuel w ould ha v e to b e c on s id e r e d to determine if using UN w as w orth i t.
A depletion an alysis will sho w ho w the core reactivit y v aries o v er time. If the effects of these c hanges in reactivit y are un des i re d it ma y b e necessary to alter the b eginning of life core zoning. U n des i re d effects could include burnout of the fuel to o quic kly or con v ersely breeding could add significan t reactivit y t o the p oin t where it is difficult to main tain the abilit y to co n trol it. F ur ther criticalit y considerations w ould b e to reconsider the lo cation of the con trol ro ds in the core. Rep ositioning the ro ds ma y aid th e core in main taining a smo oth p o w er distribution. F urthermore, i f reactivit y increases b ecome an iss u e , placing them in more reactiv e areas of the core (closer to the cen ter) w ould increase their w orth and help alleviate these issues.
10.2.2 Pro cess Heat
The ideal configuration for the second heat exc han ge r w ould b e a zigzag c hannels for He and straigh t c hannels for H 2 O. The design group w as unable to mo del a h e at exc hanger of this design due to limitations of the computational mo del. F uture w ork should explore this option and also that of carrying out the heat exc hange b et w een He and H 2 O i n m ultiple stages of PCHEs or shell-and-tub e heat e xc hangers. A Matlab mo del for sizing a coun ter flo w shell-and-tub e HX with He in the shell and H 2 O in tub es is b eing dev elop ed and will b e a v ailable for future stud ie s.
10.2.3 Biofuels Plan t
One problem facing the biofuels pro duction is where to place the carb on dio xide formed from the gassification pro cess to minimize carb on dio xide w aste. Some options are to recycle, sell, store underground, or dissolute in the deep o cean. Curren tly General Electric has a design published for underw ater and underground carb on sequestration that should b e lo ok ed in to for future w or k.
11 Economics Of Design
11.1 Exp ected Rev en ue
Some rough calculations to predict the amoun t of profit deriv ed from selling electricit y to the gri d. Af te r giving ab out 425 MW e to differen t pro cesses (mostly h ydrogen) an d giving ab out 315 MWt (133 MW e) to pro cess heat a total of 558 MW e is lost from electricit y pro d uction. Assuming electricit y can sell for
$ 0.095/kWh, the predicted rev en ue is $ 1.272 M/da y . This do es not tak e in to accoun t op erations costs or the
cost to man uf ac tu re the electricit y since it is b ey ond the scop e of this course. Ho w ev er, this pl a n t could b e quite profitable if costs are k ept lo w and the exp ected rev en ue from e lectricit y sales exceeds $ 1 million a da y . With mass flo w rates of 1,874 barrels p er da y , 4,780 barrels p er da y for diesel and gasoline resp ectiv ely , the exp ected rev en ue from the biofuel plan t s h ould b e at least $ 1.7 million p er da y . This is enough gasoline to fuel 18,500 cars p er da y , ass u m i ng a fifteen gallon tank. The curren t US demand for gasoline is 9.12
million barrels p er da y .
P art V
Ac kno wledge men ts
The en tire design team w ould lik e to ac kno wledge the follo wing p eople for their in v aluable help with the design of this facilit y: Dr. Mic hael Short, T yrell Armen t, Koroush Shirv an, Bry an Herman, Prof. Mic hael Gola y , Dr. Charles F orsb erg, Prof.Mic hael Driscoll, and Prof. Neil T o dreas.
References
[1] Asme viii - div.1.
[2] Epa federal w ater p ollution con trol act.
[3] Korea atomic energy researc h institute table of n uclides.
[4] The Westinghouse sulfur pr o c ess for hydr o gen pr o duction , 2003.
[5] Cle an Water A ct NPDES Permitting Determinations for Thermal Dischar ge and Co oling Water Intake fr om Mir ant Kendal l Station in Cambri d ge, MA , 2004.
[6] Hydrogen pro duction form n uclear e n e r gy via high te mp erature electrolysis. In Pr o c e e dings of the International Confer enc e on A dvanc es in Nucle ar Power Plant , 2004.
[7] Hydrogen pro duction form n uclear e n e r gy via high te mp erature electrolysis. In Pr o c e e dings of the International Confer enc e on A dvanc es in Nucle ar Power Plant , 2004.
[8] Challenges in generating h ydrogen b y high te mp erature electrolysis u s i ng solid o xide cells. T ec hnical rep ort, Idaho National Lab oratory , 2008.
[9] P eterbilt motors compan y presen ts: a w h ite pap er on truc k aero dynamics and fuel efficiency , Ma y 2008.
[10] Emission factors for lo comotiv es, April 2009.
[11] Allo y 20, No v em b er 2011.
[12] Allo y 20 fasteners with w orld wide shipping, No v em b er 2011.
[13] Amine plan ts , 2011.
[14] Ha ynes 230 tec hnical d ata, Decem b er 2011.
[15] Ha ynes 617 allo y , No v em b er 2011.
[16] Heatric tm, 2011.
[17] Last federal helium reserv e, near amarillo, is ru nning out, No v em b er 2011.
[18] Lithium c h loride condensed phase thermo c he mistry data, No v em b er 2011.
[19] Marine plate heat exc hanger, Dec em b er 2011.
[20] P ermissible w eigh t table, Decem b er 2011.
[21] T exas observ ed fire danger, Decem b er 2011.
[22] Thermal p ollution, Decem b er 2011.
[23] Unified: Thermo dynamics an d propulsion, No v em b er 2011.
[24] W olfram alpha, No v em b er 2011.
[25] NNF CC Pro ject Num b er 10-035. P ath w a ys to u k biofuels. T ec hnical rep or t, National Non-F o o d Crops Cen ter and Lo w Carb on V ehicle P artnership, June 2010.
[26] Isao Ab e. Alk aline w ater electrolysis. Ener gy Carrier s and Conversion Systems , 1.
[27] A. Abhat. Lo w temp erature laten t h e at thermal energy storage : Heat storage m aterial s . Solar Ener gy , 30(4):313–332, 1983.
[28] S. Adhik ari and S. F ernand o. Hyd roge n mem b rane separation tec hniques. Industrial and Engine eri n g Chemistry R ese ar ch , 45:875–811, 2006.
[29] Avinash Kumar Agarw al. Biofuels(alcohols and bio diesel) appli c aiton s as fuels for in ternal com bustion engines. Pr o c ess in Ener gy and Combustion Scienc e , 33:233–271, 2007.
[30] US En virnomen tal Protection Agency . A comprehensiv e analysis of b io diesel impacts on exhaust emissions. Epa420-p-02-001, US EP A, Octob er 2002.
[31] A. Ao c hi, T. T adok oro, K. Y oshida, H. Kamey ama, M. Nobue, an d T. Y amaguc hi. Economical and tec hnical ev aluation of ut-3 thermo c hemic al h ydrogen pro duction pro cess for an industrial scale plan t. Int. J. H y dr o gen Ener gy , 14(7):421–429, 1989.
[32] A. Ao c hi, T. T adok oro, K. Y oshida, H. Kamey ama, M. Nobue, an d T. Y amaguc hi. Economical and tec hnical ev aluation of ut-3 thermo c hemic al h ydrogen pro duction pro cess for an industrial scale plan t. Int. J. H y dr o gen Ener gy , 14(7):421–429, 1989.
[33] James G S p eigh t Mustafa Balat Ayhan Demirbas, Mrinal K Ghose. The Biofuels Handb o ok . RSC publishing, 2011.
[34] Karin Bengtsson. Twin-b ed gasification concepts for b io-s ng pro d uction. T ec hnical rep ort, Lund Univ ersit y , Lund, Sw eden, 2007.
[35] BNSF. Bnsf grain elev ator directory: A ttebury gr ain, llc - tulia,tx (kim ble), No v em b er 2009.
[36] Bry an K. Boggs, Reb ecca L. King, and Gerardine G. Botte. Urea electrolysis: direct h y drogen pro duction from ur ine. Chemic al C ommunic ations , pages 4859–4861, 2009.
[37] Ulf Bossel. Do es a h ydrogen econom y mak e sense? In Pr o c e e dings of the IEEE, V ol. 94. , 2006.
[38] T on y Bo wdery . Lng applications of diffusion b onded heat exc hangers. In A l ChE Spring Me eting 6th T opic al Confer enc e on Natur al Gas Utilization , Orlando, F L , April 23-27 2006.
[39] Da vid I. Bransb y . Cellulosic biofuel tec hnologies. T ec hnical rep ort, Auburn Univ ersit y , F ebruary 2007.
[40] Jacop o Buongiorno. 22.06 engineering of n uclear systems. Lecture 10, F all 2011.
[41] National Climatic Data Cen ter. Av erage ann ual temp eratures in texas.
[42] Dongho on Kim Injin Sah W o o-Seog Ryu Y oung-sung Y o o Changheui Jang, Daejong Kim. Oxidation b eha viors of wrough t nic k el-based sup erallo ys in v arious high temp erature en vironmen ts. T r ansactions of nonferr ous metals so ciety of China , pages 1524–1531, 2011.
[43] J. Chaurette. Pip e roughness v alues, F ebruary 2003.
[44] C.K. Cho w and H.F. Khartabil. Conceptual fuel c hannel designs for candu-scwr. Nucle ar Engine ering and T e chnolo gy , 40(2):139–146, 2008.
[45] Y usuf Christi. Bio diels from microalgae. Biote chnolo gy A dvanc es , 25:294–306, F ebruary 2007.
[46] Meric hem Compan y . Lo-cat pro cess for cost effectiv e desulfurization of all t yp es of gas streams. Sc haum burg, IL.
[47] Sandmey er S te el Compan y . Allo y 304.
[48] Leanne M. Crosbie. Hydrogen pro duction b y n uclear heat. GENES4/ANP2003 , 2003.
[49] Da vit Daniely an. Sup ercritical-w ater-co oled reactor system - as one of the most promising t yp e of generation iv reactor systems, No v em b er 2003.
[50] Debabrata Das and T. Nejat V eziroglu. Hydrogen pro duction b y biological pro cesses: a surv ey of literature. International Journal of Hydr o gen Ener gy , 26:13–28, 2001.
[51] Arno de Klerk and Ph illip L. de V aal. Alkylate tec hnology selection for fisc her-tropsc h syncrude refining. A m eric an Chemic al So ciety , 2008.
[52] M. F aith Demirbas. Thermal energy storage and phase c hange materials: An o v erview. Ener gy Sour c es, Part B , pages 85–95, 2006.
[53] R.A Diaz-Real, R.S Mann, and I.S Sam bi. Hyd rotreatme n t of ath basca bitumen deriv ed gas oil o v er ni-mo, ni-w, and co-mo catalysts. Industrial and E ngi n e ering Chemistry R ese ar ch , 1993.
[54] V. Dostal, M.J. Driscoll, and P . Hejzlar. A sup ercritical carb on dio xide cycle for next generation n uclear r e actors. T ec hnical rep ort, Adv anced Nuclear P o w er T ec hnology P rogram., 2004.
[55] D.Southall and S.J. Dewson, editors. Innovative Comp act He at Exchangers , 2010.
[56] General Electric. Pro v en. tough: The 7fdl lo comotiv e diesel engine, June 2005.
[57] Rob ert J. Ev ans, editor. Nucle ar Hydr o gen Initiative . US DOE, AICHE Ann ual Meeting - Nuclear Energy and th e Hydrogen Econom y , 2007.
[58] Herman F eldmann. Biomass gasification system, 2 010.
[59] Charles W. F orsb erg and Mujid S. Kazimi. Nuclear h ydrogen using high - temp erature electrolysis and ligh t-w ater reactors for p eak electricit y pro duction. Center for A dvanc e d Nucle ar Ener gy Systems , April 2009.
[60] Williams M. C. Gemmen, R. S. and K. Gerdes. Degradation measuremen t and analysis for cells and stac ks. J. Power Sour c es , 184:251–259, 2008.
[61] G.Hewitt. He at Exchanger Design Handb o ok , v olume 2. Begell House Inc., 1987.
[62] H.D. Gougar. The V ery High T emp er atur e R e actor . Nuclear Energy Encyclop edia: Science, T ec hnology , and Applications. John Wiley and Sons, 2011.
[63] Carl Sto ots Brian Ha wks Gran t Ha wk es, Jame O’Brien. 3d cfd mo del of m ult-cell high-temp erature electrolysis stac k. International Journal of Hydr o gen Ener gy , (4189-4197), 2009.
[64] H.G. Gro ehn. Thermal h ydraulics of helical-t yp e helium/heliu m in termediate heat exc hangers (ih xs) for n uclear pro cess heat applications of high tem p erature gas-co oled reac tor s -fu ndamen tal researc h and large scale tests. Nucle ar Engine er ing and Design , 126:285–290, 1991.
[65] D.C. Gro enev eld and et al. The 2006 c h f lo okup table. Nucle ar Engine ering and Design , 237:1909–1922, 2007.
[66] Britta K. Gross, Ian J . Sutherland, and Henk Mo oiw eer. Hydrogen fuelin g infrastructure assessme n t. T ec hnical rep ort, GM Researc h & Dev elopmen t Ce n ter, 2007.
[67] G. Gun tz, M. Julien, G. Kottmann, F. P ellicani, A P ouilly , and J.C. V aillan t. The T91 Bo ok . V allourec Industries, 1990.
[68] Jin Haiming, Eric D. Larson, and F uat E. Celik. P erformance and cost analysis of future, commerically mature gasification-based electric p o w er generation from switc hgrass. Biofuels, Biopr o d, Bior ef , 3:142– 173, 2009.
[69] P atric k C. Hallen b ec k and John R. Benemann. Biological h ydrogen pro duction; fund am en tals and limiting pro cesses. International Journal of Hydr o gen Ener gy , 27:1185–1193, 2002.
[70] P aul N. Haub enreic h and J. R. Engel. Exp erience with the molten-salt reactor exp erimen t. T ec hnical rep ort, Oak Ridge National Lab oratory , 1969.
[71] Inc. Ha yn e s In ternational. Ha yn e s 625 allo y , 2001.
[72] An thon y E. Hec hano v a. High temp erature heat exc hanger pro ject. T ec hnical rep ort, UN L V Res earc h F oundation , 2008.
[73] A. Hoshi, D.R. Mills, A. Bittar, and T.S. Saitoh. Screening of high melting p oin t phase c hange materials (p cm) in solar thermal concen trating tec hnology based on clfr. Solar Ener gy , 79:332–339, 2005.
[74] IAEA. Thorium fu e l cycle - p oten tial b enefits and c h allenge s. T ec hnical rep ort, IAEA, 2005.
[75] Refactory Sp ecialties Inc or p orated. Gemcolite azs fg26-110.
[76] T. Ishizuk a, Y. Kato, and et al, editor s . Thermal Hydr aulic char acteristics of Printe d Cir cuit He at Exchangers in a Sup er critic al CO2 L o op , n u m b e r NURETH 11, 2005.
[77] N.P . Grandon J. Udaga w a, P . Aguiar. Hydrogen pro duction throu gh steam electrolysis: Mo del-based steady state p e rf ormance of a c at ho de-supp orted in terme d iate temp erature solid o xide electrolysis cell. Journal of Pow er Sour c es , pages 127–136, 2007.
[78] J.E.Hesselgrea v es. Approac hes to fouling allo w ances in the design of compact heat exc hangers. Applie d Thermal Engine ering , 22:755–762, 2002.
[79] J.Hejzlar. Computer co de for the analysis of prin ted circuit heat exc hangers w i th zigzag c hannels. T ec hnical rep ort, MIT, 2004.
[80] J.Hesselgrea v es. Compact heat exc hangers. Elsevier Scienc e , 2001.
[81] Nalladurai Kaliy an and R. V ance Morey . Natural binders and solid bridge t y p e bindi ng mec hanisms in briquettes and p ellets m ad e from corn sto v er and switc hgrass. Bior esour c e T e chnolo gy , 101(3):1082 – 1090, 2010.
[82] H. Kamey ama and K. Y oshida. Br-ca-fe w ater decomp osition cycles for h ydrogen pro du c ti on. Pr o c. 2nd WHEC. , pages 829–850, 1978.
[83] H. Kamey ama and K. Y oshida. Br-ca-fe w ater decomp osition cycles for h ydrogen pro du c ti on. Pr o c. 2nd WHEC. , pages 829–850, 1978.
[84] Seungmo Kang, Ha yri Onal, Y anfeng Ouy ang, Jurgen Sc heffran, and U. Deniz T ursun. Optimizing the biofuels infrastructure: T ransp ortation net w orks and biorefinery lo cations in illinois. In Handb o ok of Bio ener gy Ec onomics and Policy , v olume 33 of Natur al R esour c e Management and Policy , pages 151–173. Springer New Y ork, 2010.
[85] Ilgi K. Kap dan and Fikret Kargi. Bio-h ydrogen pro duction from w aste mate r ial. Enzyme and Micr obial T e chnolo gy , 38:569–582, 2006.
[86] W. M. Ka ys and M. E. Cra wford. Conve ctive He at and Mass T r ansfer . McGra w-Hill, 3rd edition, 1993.
[87] James B. Kesseli, Thomas L. W olf, F. W ells Ho dous, James S. Nash, Malcolm S. Child, Rob e r t S. Cherry , Ric hard L. Williamson, T ed R. Reed, and A. Joseph P almer. Conceptual design for a high- temp erature gas l o op test facilit y . T ec hnical rep ort, U.S. Departmen t of E nergy , August 2006.
[88] Ronald Allen Knief. Nucle ar engine ering: the ory and te chnolo gy of c ommeric al nucle ar p ower . Ameri can Nuclear So ciet y , 2nd edition, 2008.
[89] A.J. Konopk a and D.P . Gregory . Hydrogen pro duction b y electrolysis: Presen t and future. Iece c record, Institute of Gas T ec hnology , Chicago, Ill inois, 1975.
[90] T. G. Kreutz, Eric D. Larson, G. Liu, and R. H. Williams. Fisc her-tropsh f ue l s f rom c oal an d biomass. T ec hnical rep ort, Princeton Uni v ersit y , 2008.
[91] National Renew abl e Energy Lab oratory . Biomass resources. PDF, Septem b er 2007.
[92] Eric D. Larson, Jin Haiming, and F uat E. Celik. Large-scale gasification-based copro duction of fuels and electricit y from switc hgrass . Biofuels, Biopr o d, Bior ef , 3:174–194, 2009.
[93] Dieter Lec k el. Hydro crac king of iron-catalyzed fisc her-tropsc h w axes. A m eric an Chemic al So ciety , 2005.
[94] Mic hele A. Le wis, Man uela Serban, an d John K. Basco. Hydrogen pro duction a ¡500 c using a lo w temp erature thermo c hemical cycle. A r gonne National L ab or atory , lewism@cm t.anl.go v.
[95] James R. Lines. Helically coiled heat exc hangers offer adv an tages. T ec h nical rep ort, G r aham Man u facturing Co. Inc., No v em b er 2010.
[96] Lee R. Lynd, Eric Larson, Nathanael Greene, John Sheehan Mark Lase and, Bruce E. Dale, Sam uel McLaughlin, and Mic hael W ang. The role of biomass in america’s energy future: framing the analysis. Biofuels, Biopr o ducts and Bior efining , 3:113–123, Marc h 2009.
[97] R.R. Macdonald and M .J.Driscoll. Magnesium o xide: An impro v ed reflector for blank e t-f re e fast reactors. T r ansactions of the A meric an Nucl e ar So ciety , June 2010.
[98] H.G. MacPherson, editor. Fluid F uel R e a c t or s . Addison-W esley , 1958.
[99] S. Ma jumdar, A. Moisseytsev, and K. Natesan. Assessmen t of next generation n uclear plan t in terme diate heat exc hanger design. T ec h nical rep ort, Argonne National Lab oratory , 2008.
[100] C. Maretto and R. Krishna. Mo delling of bubb le column slurry reactor for fisc her-tropsc h syn th e sis.
Catalysis to day , 52:279–289, 1999.
[101] Hosein Shap ouri Steph e n P . Slinsky Marie E. W alsh, Daniel G. De La T orre Ugarte. Bio energy c r op pro duction in the united states: P oten tial quan tities, land use c hanges, and economic impacts on the agricultural sector. Envir onmental and R esour c e Ec onomics , 24:313–333, 2003.
[102] T. Matsuo, M. Uk ei, M. T ak ey ama, and R. T anak a. Strengthening of nic k el-based sup erallo ys f or n uclear h e at exc hanger applications. Journal of N u cle ar Materials , 22:1901–1907, 1987.
[103] S.Khela & S.J. Matthews. Ceramic heat exc hangers - materials and comp onen ts p erformance issues.
Materials Issues in He at Exchangers and Boilers Confer enc e , 1995.
[104] H. Mehling an d L. Cab eza. He at and c old stor age with PCM: an up to date intr o duction into b asics and applic ations . Springer, 2008.
[105] Dennis Y.C. Leung Meng Ni, Mic hael K.H. Leung. T ec hnological dev elopmen t of h ydrogen pro duction b y solid o xide electrolyzer cell (so ec). International Journal of Hydr o gen Ener gy , 33:2337–2353, 2008.
[106] Microtherm. Microtherm microp orous insulation vs. asp en p yrogel xt aerogel bl ank et. whic h is the b etter high p erformance insulation?
[107] An ton Moisseytsev and James J. Sienic ki. T ransien t acciden t analysis of a sup ercritical carb on dio xide bra yton cycle energy con v erter coupled to an autonomous lead-co oled fast reactor. Nucle ar , 238:2094– 2105, 2008.
[108] L.F. Mo o dy . F riction factors for pip e flo w. T r ansactions of the A . S . M.E , pages 671–684, 1944.
[109] H. Muller-Steinhagen, editor. He at Exchanger F ouling . PUBLICO Publ ic ati ons, 2000.
[110] N.E.T o dreas and M.S.Kazimi. Nucle ar Systems V olume 1: Thermal Hydr aulic F undamentals, Se c ond Edition . CR C Press, 2011. Equation 11-94.
[111] Chang H. Oh and Eu ng S. Kim. Design opt ion of heat exc hanger for next generation n uclear plan t. T ec hnical rep ort, Idaho National Lab oratory , 2008.
[112] N. Oh tori, M. Salanne, and P .A. Madden. Calculations of the thermal conductivities of ionic materials b y sim ulation with p olarizable in te raction p oten tials. The Journal of Chemic al Physics , 130, 2009.
[113] Haruhik o et al Oh y a. Separation of h ydrogen from thermo c hemical pro cesses u s in g zirconia-silica comp osite mem bran e . Journal of Membr ane Scienc e , 97:91–96, 1994.
[114] M. Ok azaki. High - temp erature strength of ni-base sup erallo y coating. Scienc e and T e chnolo gy of A dvanc e d M a ter ials , 2:357–366, 2001.
[115] Scott J. Olson. Gas clean up tec hnol ogie s suitable for biomass gasification to liquid fuels. ENergy T ec hnology Nexan t, Inc., 2006 AlChE National Meeting, San F rancisco, CA, 2006.
[116] Bulen t Ona y and Y asutoshi Saito. Corrosion b eha vior of fe-20c r and ni-20cr allo ys in ar-h2o-h br gas mixtures at 1000 k. Oxidation of Metals , 40, 1993.
[117] Union P acific. Co v ered hop p ers, 2011.
[118] M.A. P aisley . A promising p o w er option - the ferco silv agas biomass gasification pro cess - op eratin g exp erience at th e burlington gasifier. ASME T urb o Exp o , 2001.
[119] Cole P armer. Chemical resistance database, Octob er 2011.
[120] Mic hael A. P op e. Reactor ph ysics design of sup ercritical co2-co oled fast reactors. Master’s thesis, Massac h u s etts Institute of T ec hn ology , Septem b er 2004.
[121] Hideki Aita Kenji Sekita R. Hi no, Katsuhiro Haga. R & d on h y drogen pro duction b y high-temp erature electrolysis of steam. Nucle ar Engine er ing and Design , 223:363–375, 2004.
[122] S.T. Sie R. Krishna. Design and scale-up of the fisc her-tropsc h bubble column slurry reactor. F uel Pr o c essing T e chnolo gy , 64:73–105, 2000.
[123] U. Eb erle R. v on Helmolt. F uel cell v ehicles: Status 2007. Journal of Power Sour c es , 165:833–843, 2007.
[124] Prashaan th Ra vindran, Piyush Sabharw all, a n d Nolan A. Anderson. Mo deling a prin ted circuit heat exc hanger with relap5-3d for the next generation n uclear p lan t. T ec hnical rep ort, Idaho National Lab oratory , 2010.
[125] New Y ork State Energy Researc h and Dev elopmen t Authori t y . Hydrogen p ro duction - steam methane reforming (smr).
[126] M.W. Rosen thal, P .R. Kasten, and R.B. Briggs. Molten-salt reactors – history , status, and p oten tial.
Nucle ar Applic ations and T e chnolo gy , 8(2), 1969.
[127] Ajit Ro y , Lalit Sa v alia, Narendra Kothapalli, and Ragh unan dan Karamc heti. Mec hanical prop erties and crac king b eha vior of high-temp eratu re heat-exc hanger materials. In Pr o c e e dings of the ASME Pr essur e V essels and Piping Confer enc e 2005, V ol 6 , 2005.
[128] S. Sokhansanj S. Mani, L.G. T abil. Ev aluation of compaction equations ap plied to four biomass sp ecies.
Canadian Biosystems Engine ering. , 46:3.55–3.61, 2004.
[129] Piyush Sabharw all, Eung So o Kim, Mic hael McKellar, and Nolan Anderson. Pro cess heat exc hanger options for flu oride salt high temp erature rea ctor. T ec hnical rep ort, Idaho National Lab, 2011.
[130] M. Sakurai, E. B i lge n , A. Tsutsumi, and K. Y os h ida. Adiabatic ut-3 th e r m o c hemical pro ce ss for h ydrogen pro duction. Int. J. Hydr o gen Ener gy , 21(10):865–870, 1996.
[131] M. Sakurai, A. Tsutsumi, and K. Y oshida. Impro v emen t of ca-p ellet reactivit y in ut-3 thermo c hemical h ydrogen pro duction cycle. Int. J. Hydr o gen Ener gy , 20(4):297–301, 1995.
[132] B. K. et al Sea. F ormation of h ydrogen p ermselectiv e silica mem br ane for elev ated temp erature h ydro gen reco v ery from a mixture con taining steam. Gas Sep ar ation & Purific ation , 10(3):812–881, 1996.
[133] Ramesh K. Shah and Dusan P . Sekulic. F undamentals of He at Exchanger Design . John Wiley and Sons, 2003.
[134] J. S hee han , T. Dunaha y , J. Benemann, and P . Ro essler. A lo ok bac k at the u.s. departmen t of energy’s aquatic sp ecie s program–bio diesel from algae. Nrel/tp-580-24190, National Renew able Energy Lab oratory , Golden, CO, 1998.
[135] S. R. Sherman and Y. Chen. Heat exc hanger testing requiremen ts and facilit y needs for the nhi/ngnp pro j e ct. T ec hnical Rep ort WSR C-STI-2008- 00152, 2008.
[136] Y oic hiro Shimazu. Curren t situation of msr dev elopmen t in japan. Presen tation.
[137] K. Shirv an. The Design of a Comp act Inte gr al Me dium Siz e PWR: The CIRIS . PhD thesis, MIT, 2010.
[138] Mic hael Short. The Design of a F unctional ly Gr ade d Comp osite for Servic e in High T emp er atur e L e ad and L e ad-Bismuth (LBE) Co ole d Nucle ar R e actors . PhD th e sis, MIT, 2010.
[139] Mic hael S hort. Core group: Non-p wr/b wr reactors. Le ctur e , Septem b er 2011.
[140] J. J. Sienic ki, A. V. Moisseytsev, D. C. W ade, M. T. F armer, C. P . Tzanos, J. A. Stillman, J. W. Holland, P . V. P etk o v, I. U. Therios, R. F. Kulak, and Q. W u. The star-lm lead-co oled closed fuel cycle fast reactor coupled to a sup ercritical carb on dio xide bra yton cycle adv anced p o w er con v erter. In Glob al , New Orleans, LA, No v em b er 16-20 2003.
[141] E. M. Sparro w and Jr. Lo effler, A. L. Longitudinal laminar flo w b et w e en cylinders arranged in regular arra y . A.I.Ch.E. , 5:325, 1959.
[142] P .L. Spath and D.C. Da yton. Preliminary screening-tec hnical and economic assessmen t of syn thesis gas to fuels and c hemicals with emphasis on the p oten tial for biomass-deriv ed syngas. T ec hnical rep ort, National Renew able Energy Lab oratory , 2003.
[143] Mohamed S. El-Genk Stev en B. Ross and R. Bruce Matthews. Thermal conductivit y correlation for uranium nitride fuel b et w een 10 and 1923 k. Journal of Nucle ar Materials , 151:313–317, 1988.
[144] Hiroshi; P . C. P ari s T ad a. The Str ess A nalysi s of Cr acks Handb o ok (3 e d.) . American So ciet y of Mec hanical En ginee r s , 2005.
[145] Y. T adok oro, T. Ka jiy ama, T.Y am agu ic hi, N.Sak ai, H. Kamey ama, and K. Y oshida. T ec hnical ev alua tion of u t-3 thermo c hemical h ydrogen pro duction p ro cess for an industrial scale plan t. Int. J. Hydr o gen Ener gy , 22(1):49–56, 1996.
[146] Ha j im e T ak ey a T oshio Hirai T ak ashi Goto, Chen-Y an Guo. Coating of titanium carbide films on stainless steel b y c hemical v ap or dep os iti on and their corrosion b eha vior in a br2-o2-ar atmosphere. Journal of Material Scienc e , 27:233–239, 1992.
[147] Allison M. Thomson and Cesar R. Iz ar rualde. Sim ulating p oten tial switc hgrass pro duction in the united states. Pnnl-19072, US Departmen t of Energy , Decem b er 2009.
[148] Anna Nikiforo v a; P a v el Hejzlar; Neil E. T o dreas. Lead-co oled flexible con v ersion ratio fast reactor.
Nucle ar Engine er ing and Design , 239:2596–2611, 2009.
[149] N.E. T o dreas and M.S. Kazimi. Nucle ar Systems , V ol 1 . T a ylor and F rancis, 1989.
[150] P a v el Hejzlar; Neil E. T o dreas. Flexible con v ersion ratio fast reactor systems ev aluation final rep ort.
Nucle ar Ener gy R ese ar ch Initiative, Pr oje ct 06-040 , 2008.
[151] T. B. T ok are v a and V. V. Smolin . Corrosion resistance of structural materials in lithium c hloride.
Chemic al and Petr oleum Engine ering , 10:139, 1974.
[152] The Engineering T o olBo x. St re ss in thic k-w alled tub es or cylinders.
[153] Mic hael F elderhoff Ulric h Eb erle and F erdi Sc h uth. Chemical and ph ysical solution s for h ydrogen storage. A ngew Chem Int Ed , 48:6608–6630, 2009.
[154] US. Co de of federal regulations.
[155] U.S. Geological Surv ey U.S. Departmen t of the In terior . T exas precipitation. The Nation al A tlas of the United States of America, 2006.
[156] X. Vitar t, A. Le Duigou , and P . Carles. Hydrogen pro d uction using th e sulfur-io dine cycle coupled to a vh tr : An o v erview. Ener gy Conversion and Management , 47:2740–2747, 2006.
[157] X. Vitar t, A. Le Duigou , and P . Carles. Hydrogen pro d uction using th e sulfur-io dine cycle coupled to a vh tr : An o v erview. Ener gy Conversion and Management , 47:2740–2747, 2006.
[158] Qiu w ang W ang, Qiuy ang Chen, Guidong Chen, and Min Zeng. Numerical in v estigation on com bined m ultiple shell-pass shell-and tub e heat exc hanger with con tin uous helical baffle s. International Journal of He at an d Mass T r ansfer , 52:1214–1222, 2008.
[159] Eric D. Larson Xiangb o Guo, Guang jian Liu. High-o ctane gasoline pro duction b y upgrading lo w- temp erature fisc her-tropsc h syncrude. A meric an Chemic al So ciety , 2011.
[160] X.Li, D.Kinimon t, and et al. Allo y 617 f or the h igh tem p erature diffusion-b onded compact heat exc hangers. In ICAPP Pr o c e e dings , 2008.
[161] B. YildizI and M.S. Kazimi. Hydrogen pro duction using high-temp erature steam electrolysis supp orted b y ad v anced gas reactors with sup ercritical co2 cycles. Nucle ar T e chnolo gy , 2006.
[162] Seungmo Kang Y anfeng Ouy ang Y un Bai, T aesung Hw ang. Biofuel refinery lo cation and supply c hain planning under traffic congestion. T r ansp ortation R ese ar ch Part B: Metho dolo gic al , 45(1):162 – 175, 2011.
HoQ template courtesy of QFD Online. Used with permission.
P art VI
Figure 54: Core house of qualit y
App endix A: Core P arameter QFD
The QFD metho d is used to “transform user demands in to design qualit y” [QFD2012]. A house of qualit y is part of this metho d and h e lp s correlate what is desired from a tec hni c al and commercial standp oin t and can b e used to compare differen t pr o ducts (in this cas e differen t reactor designs). The h ouse of qualit y for the core design can b e seen in Figure 54.
The QFD metho d has b een used b y man y differen t designs across a wide arra y of fields. It helps sort through design parameters to iden tify whic h areas of a design are most imp ortan t to fo cus on. It do es this b y comparing pro duct goals, whic h can b e seen on the left in the house of qualit y , and design parameters, w h ic h can b e seen on the top of the house of qualit y . T h e pro duct goals are giv en a w eigh t on a scale from one to ten to determine ho w imp ortan t eac h is relativ e to the others. The pro du c t goals are then analyzed with resp ect to the design parameters. The goals an d parameters are judged to ha v e either a strong, mo derate, w eak, or no relationship. The diffic u lt y of optimizing the design parameter is then also judged on 0-10 scale. The com bination of the imp ort anc e of the design goals, relationship to design parameters, and difficult y to optimize a parameter are then used to pro duce relativ e w eigh t for eac h d e sign parameter. This w eigh t helps fo cus designers on the most imp ortan t features of a pro duct.
Other f e atu res of the house of qualit y are t he “ro of ” and the comp etitiv e analysis. The ro of compares differen t design parameters and sho ws ho w eac h parameter relates to another. The parameters are first lo ok ed at to see if th e goal is to maximize, minimize, o r hit a targe t for the p arame ter for optimization.
Then eac h is compared as to ho w optimizing those parameters effects the others. There is either a strong p ositiv e, p ositiv e, negativ e, strong negativ e, or neutral relationship. This to ol aids designers in iden tifyin g whic h par am eters will b e the most difficult or easiest to optimize based on ho w they effect other parameters. Finally , the comp etitiv e analysis compares differen t pro ducts, in our cas e reactor designs, based on the des i g n goals. The pro ducts ar e rated on a scale of 0-5 to see ho w th e y p erform in the design goal c ategories. The core designers to ok these p erformance rankings and m u ltiplied them b y the design goal imp ortance and w ere able to get a w eigh ted ranking of the differen t designs. The final des ign c hose, a lead c o oled fast reactor, w as the highest rank ed design and help ed justify the decision.
P art VI I
App endix B: Criticalit y Mo del
Crit i ca l it y Mo d el
Co d e ru n in M CN P 5 f o r crit ical ity dat a
2 2. 033 F al l 2011 C ore D es i g n v1 6_ 3_out
c ====== ====== ====== ========= ====== ====== = ========= ======
c =====> Cel l card s
c ====== ====== ====== ========= ====== ====== = ========= ======
c =====> Fuel p i n, cl ad and s urr oun d i ng s od i um ; E nri ch == 15% , 15%, 10%(g oing up ) c I nn er Fu el
|
101 |
71 - 12. 89 00 -10 00 |
u=1 $ l o w er f uel |
|
102 |
71 - 12. 89 00 -10 05 |
u=1 $ m i d d l e fuel |
|
103 |
7 - 12. 89 00 -100 6 |
u=1 $ t o p fuel |
|
104 |
3 -1 0. 192 0 1007 - 1001 |
u=1 $ g ap w / c oo l a nt |
|
105 |
21 0 . 0 429 1 1001 - 1002 |
u=1 $ p r otect i ve l a yer (m 21) |
|
106 |
2 0 . 04291 1002 - 1003 |
u=1 $ s t eel cl ad (m 2 ) |
|
107 |
21 0 . 0429 1 1003 - 1004 |
u=1 $ p r otect i ve l a yer (m 21) |
|
108 c 109 |
3 -1 0. 192 0 1004 I nt er m e d i ate F uel 72 - 12. 89 00 -10 00 |
u=1 $ c oo l ant u=2 $ f u el , |
|
110 |
72 - 12. 89 00 -10 05 |
u=2 $ f u el , |
|
111 |
71 - 12. 89 00 -10 06 |
u=2 $ f u el , |
|
112 |
3 -1 0. 192 0 1007 - 1001 |
u=2 $ g ap w / c oo l a nt |
|
113 |
21 0 . 0429 1 1001 - 1002 |
u=2 $ p r otect i ve l a yer (m 21) |
|
114 |
2 0 . 04291 1002 - 1003 |
u=2 $ s t eel cl ad (m 2 ) |
|
115 |
21 0 . 0429 1 1003 - 1004 |
u=2 $ p r otect i ve l a yer (m 21) |
|
116 c 117 |
3 -1 0. 192 0 1004 Out er F uel 72 - 12. 89 00 -10 00 |
u=2 $ c oo l ant u=3 $ f u el , |
|
118 |
72 - 12. 89 00 -10 05 |
u =3 $ fu el , |
|
119 |
71 - 12. 89 00 -10 06 |
u=3 $ f u el , |
|
120 |
3 -1 0. 192 0 1007 - 1001 |
u=3 $ g ap w / c oo l a nt |
|
121 |
21 0 . 0429 1 1001 - 1002 |
u=3 $ p r otect i ve l a yer (m 21) |
|
122 |
2 0 . 04291 1002 - 1003 |
u=3 $ s t eel cl ad (m 2 ) |
|
123 |
21 0 . 0429 1 1003 - 1004 |
u=3 $ p r otect i ve l a yer (m 21) |
|
124 |
3 -1 0. 192 0 1004 |
u=3 $ c oo l ant |
c =====> Pin co olant c hannel s
200 3 -1 0. 192 0 -100 4 u=4 $ c oo l ant
201 3 -1 0. 192 0 1004 u=4 $ c oo l ant c =====> Pin L att i ce
202 0 -2000 u= 5 l at=2
fi l l = -6: 6 -6 : 6 0:0
0 0 0 0 0 0 0 0 4 1 4 0 0 $R OW 1
0 0 0 0 0 0 4 1 1 1 1 4 0 $R OW 2
0 0 0 0 4 1 1 1 1 1 1 1 4 $R OW 3
0 0 0 1 1 1 1 1 1 1 1 1 1 $ R OW 4
0 0 4 1 1 1 1 1 1 1 1 1 4 $ R OW 5
0 0 1 1 1 1 1 1 1 1 1 1 0 $ R OW 6
0 4 1 1 1 1 1 1 1 1 1 4 0 $ R OW 7
0 1 1 1 1 1 1 1 1 1 1 0 0 $ R OW 8
4 1 1 1 1 1 1 1 1 1 4 0 0 $ R OW 9
1 1 1 1 1 1 1 1 1 1 0 0 0 $ R OW 1 0
4 1 1 1 1 1 1 1 4 0 0 0 0 $ R OW 11
0 4 1 1 1 1 4 0 0 0 0 0 0 $R OW 12
0 0 4 1 4 0 0 0 0 0 0 0 0 $R OW 13
c
203 0 -2000 u= 6 l at=2
fi l l = -6: 6 -6 : 6 0:0
0 0 0 0 0 0 0 0 4 2 4 0 0 $R OW 1
0 0 0 0 0 0 4 2 2 2 2 4 0 $R OW 2
0 0 0 0 4 2 2 2 2 2 2 2 4 $R OW 3
0 0 0 2 2 2 2 2 2 2 2 2 2 $ R OW 4
0 0 4 2 2 2 2 2 2 2 2 2 4 $ R OW 5
0 0 2 2 2 2 2 2 2 2 2 2 0 $ R OW 6
0 4 2 2 2 2 2 2 2 2 2 4 0 $ R OW 7
0 2 2 2 2 2 2 2 2 2 2 0 0 $ R OW 8
4 2 2 2 2 2 2 2 2 2 4 0 0 $ R OW 9
2 2 2 2 2 2 2 2 2 2 0 0 0 $ R OW 1 0
4 2 2 2 2 2 2 2 4 0 0 0 0 $ R OW 11
0 4 2 2 2 2 4 0 0 0 0 0 0 $R OW 12
0 0 4 2 4 0 0 0 0 0 0 0 0 $R OW 13
c
204 0 -2000 u= 7 l at=2
fi l l = -6: 6 -6 : 6 0:0
0 0 0 0 0 0 0 0 4 3 4 0 0 $R OW 1
0 0 0 0 0 0 4 3 3 3 3 4 0 $R OW 4
0 0 0 0 4 3 3 3 3 3 3 3 4 $R OW 3
0 0 0 3 3 3 3 3 3 3 3 3 3 $ R OW 4
0 0 4 3 3 3 3 3 3 3 3 3 4 $ R OW 5
0 0 3 3 3 3 3 3 3 3 3 3 0 $ R OW 6
0 4 3 3 3 3 3 3 3 3 3 4 0 $ R OW 7
0 3 3 3 3 3 3 3 3 3 3 0 0 $ R OW 8
4 3 3 3 3 3 3 3 3 3 4 0 0 $ R OW 9
3 3 3 3 3 3 3 3 3 3 0 0 0 $ R OW 1 0
4 3 3 3 3 3 3 3 4 0 0 0 0 $ R OW 11
0 4 3 3 3 3 4 0 0 0 0 0 0 $R OW 12
0 0 4 3 4 0 0 0 0 0 0 0 0 $R OW 13
c =====> I nne r Fu el A s s em b l y
300 0 -300 0 u=8 fi l l =5 $ as s em b l y f uel l att i ce
301 21 0 . 04291 0 300 0 -3001 u= 8 $ p r otect i ve l ayer ( m 21)
302 2 0. 042910 3001 -3002 u= 8 $ s t eel cl ad ( m 2)
303 21 0 . 04291 0 300 2 -3003 u= 8 $ p r otect i ve l ayer ( m 21)
304 3 -10 . 1920 3003 u=8 $ as s e m b l y g ap c =====> I nt er m e d i ate F uel A s s em b l y
305 0 -300 0 u=10 fi l l =6 $ as s em b l y fuel l att i ce
306 21 0 . 04291 0 300 0 -3001 u= 10 $ p r otect i ve l ayer ( m 21)
307 2 0. 042910 3001 -3002 u= 10 $ s t eel cl ad ( m 2)
308 21 0 . 04291 0 300 2 -3003 u= 10 $ p r otect i ve l ayer ( m 21)
309 3 -10 . 1920 3003 u=10 $ as s e m b l y g ap c =====> Out er F uel A s s em b l y
310 0 -300 0 u=11 fi l l =7 $ as s em b l y fuel l att i ce
311 21 0 . 04291 0 300 0 -3001 u= 11 $ p r otect i ve l ayer ( m 21)
312 2 0. 042910 3001 -3002 u= 11 $ s t eel cl ad ( m 2)
313 21 0 . 04291 0 300 2 -3003 u= 11 $ p r otect i ve l ayer ( m 21)
314 3 -10 . 1920 3003 u=11 $ as s e m b l y g ap c =====> R efl ect or As s em b l y
600 4 -3 . 58000 -300 0 u=12 $ refl ect or m ater i al
601 21 0 . 04291 0 300 0 -3001 u= 12 $ p r otect i ve l ayer ( m 21)
602 2 0. 042910 3001 -3002 u= 12 $ s t eel cl ad ( m 2)
603 21 0 . 04291 0 300 2 -3003 u= 12 $ p r otect i ve l ayer ( m 21)
604 3 -10 . 1920 3003 u=12 $ as s e m b l y g ap c =====> S hi el d A s s em b l y
700 6 -2 . 52000 -300 0 u=13 $ s hi el d m ater i al
701 21 0 . 04291 0 300 0 -3001 u= 13 $ p r ote ct i ve l ayer ( m 21)
702 2 0. 042910 3001 -3002 u= 13 $ s t eel cl ad ( m 2)
703 21 0 . 04291 0 300 2 -3003 u= 13 $ p r ote ct i ve l ayer ( m 21)
704 3 -10 . 1920 3003 u=13 $ as s e m b l y g ap c =====> Contr ol A s s em b l y
800 3 -10 . 1920 -800 0 u=14 $ c oo l ant channel
801 21 0 . 04291 0 800 0 -8001 u= 14 $ p r ote ct i ve l ayer ( m 21)
802 2 0 . 04291 0 8001 -8002 u= 14 $ s t eel cl ad ( m 2)
803 5 -2 . 52000 8002 -8003 u= 14 $ cont rol absor b er
804 2 0. 042910 8003 -8004 u= 14 $ s t eel cl ad ( m 2)
8 05 21 0 . 04291 0 800 4 -8005 u= 14 $ p r ote ct i ve l ayer ( m 21)
|
806 |
3 -10 . 1920 |
8005 |
u=14 |
$ as s e m b l y g ap |
|
807 |
3 -10 . 1920 |
-800 6 |
u=14 |
$ w i t hd ra w n C R b u nd l e p l en um |
|
808 |
3 -10 . 1920 |
8006 |
u=14 |
$ as s e m b l y g ap |
|
c =====> A s s em b l y coo l ant channel |
||||
|
900 |
3 -10 . 1920 |
-300 0 |
u=9 $ c oo l ant |
|
|
901 |
3 -10 . 1920 |
3000 |
u=9 $ co olant |
|
c =====> Core L att i ce
400 0 -4000 u=15 l at=2
fi l l = -14: 14 - 14: 14 0: 0
|
9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 |
$ R OW 1 |
|
|
9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 |
$ R OW 2 |
|
|
9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 |
$ R OW 3 |
|
|
9 9 9 9 9 9 9 9 9 9 9 9 9 9 13 13 13 13 1 3 13 1 3 13 13 13 13 13 9 9 9 |
$ R OW 4 |
|
|
9 9 9 9 9 9 9 9 9 9 9 9 9 13 12 1 2 12 12 12 12 12 12 12 1 2 12 13 9 9 9 |
$ R OW 5 |
|
|
9 9 9 9 9 9 9 9 9 9 9 9 13 12 11 11 11 1 1 11 11 11 11 11 1 1 12 13 9 9 9 |
$R OW 6 |
|
|
9 9 9 9 9 9 9 9 9 9 9 13 12 11 1 4 11 11 11 11 1 1 11 11 14 11 12 1 3 9 9 9 |
$R OW 7 |
|
|
9 9 9 9 9 9 9 9 9 9 13 12 11 11 1 1 11 11 11 11 11 11 11 1 1 11 12 13 9 9 9 |
$ROW 8 |
|
9 9 9 9 9 9 9 9 9 13 1 2 11 11 11 10 10 1 0 14 10 10 10 11 1 1 11 12 13 9 9
|
9 |
$ R OW 9 |
|||
|
9 9 9 9 9 9 9 9 1 3 12 11 11 11 1 0 10 10 10 10 1 0 10 10 11 11 11 1 2 13 9 9 9 $ R OW 10 9 9 9 9 9 9 9 13 12 11 11 11 10 1 0 14 10 10 10 14 10 10 1 1 11 11 12 13 9 9 9 $ R OW 11 9 9 9 9 9 9 13 1 2 11 1 1 11 14 10 10 8 8 8 8 10 1 0 14 11 11 11 12 1 3 9 9 9 |
$R OW 12 |
|||
|
9 9 9 9 9 13 12 11 11 11 10 10 1 0 8 8 8 8 8 10 1 0 10 11 11 11 12 1 3 9 9 9 |
$R OW 1 3 |
|||
|
9 9 9 9 1 3 12 11 11 1 1 10 10 10 8 8 8 8 8 8 10 10 10 11 1 1 11 12 13 9 9 9 |
$ROW 14 |
|||
|
9 9 9 13 12 11 1 4 11 10 10 14 8 8 8 8 8 8 8 14 10 10 11 1 4 11 12 13 9 9 9 |
$ R OW 15 |
|||
|
9 9 9 13 12 11 1 1 11 10 10 10 8 8 8 8 8 8 10 10 10 11 11 1 1 12 13 9 9 9 9 |
$ROW 16 |
|||
|
9 9 9 13 12 11 1 1 11 10 10 10 8 8 8 8 8 10 10 1 0 11 11 11 12 13 9 9 9 9 9 |
$ROW 17 |
|||
|
9 9 9 13 12 11 1 1 11 14 10 10 8 8 8 8 10 10 14 11 11 11 1 2 13 9 9 9 9 9 9 |
$R OW 18 |
|||
|
9 9 9 13 12 11 1 1 11 10 10 14 1 0 10 10 14 10 1 0 11 11 11 12 13 9 9 9 9 9 |
||||
|
9 9 $ R OW 19 9 9 9 13 12 11 1 1 11 10 10 10 1 0 10 10 10 10 1 1 11 11 12 13 9 9 9 9 9 9 9 9 $ R OW 20 9 9 9 13 12 11 1 1 11 10 10 10 1 4 10 10 10 11 1 1 11 12 13 9 9 9 9 9 9 9 9 9 $ R OW 21 9 9 9 13 12 11 1 1 11 11 11 11 1 1 11 11 11 11 1 1 12 13 9 9 9 9 9 9 9 9 9 9 |
$ROW 22 |
|||
|
9 9 9 13 12 11 1 4 11 11 11 11 1 1 11 11 14 11 1 2 13 9 9 9 9 9 9 9 9 9 9 9 |
$R OW 23 |
|||
|
9 9 9 13 12 11 1 1 11 11 11 11 1 1 11 11 11 12 1 3 9 9 9 9 9 9 9 9 9 9 9 9 |
$R OW 2 4 |
|||
|
9 9 9 13 12 12 1 2 12 12 12 12 1 2 12 12 12 13 9 9 9 9 9 9 9 9 9 9 9 9 9 |
$ R OW 25 |
|||
|
9 9 9 13 13 13 1 3 13 13 13 13 1 3 13 13 13 9 9 9 9 9 9 9 9 9 9 9 9 9 9 |
$ R OW 26 |
|||
|
9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 |
$ R OW 27 |
|||
|
9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 |
$ R OW 28 |
|||
|
c |
9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 =====> Full m od el |
$ R OW 29 |
||
|
500 |
3 -10 . 1920 -500 0 f i l l =15 $ l att i ce o f el em ent s |
|||
|
501 |
2 0. 042910 5001 -5002 $ cl a d |
|||
|
502 |
3 -10 . 1920 -500 3 $ up p er p l e num |
|||
|
503 |
3 -10 . 1920 -500 4 $ l ow e r p l e num |
|||
|
504 c |
0 5002 $ outer v oid ====== ====== ====== ========= ====== ====== = ========= ====== |
|||
|
c |
====== ====== ====== ========= ====== ====== = ========= ====== |
|||
|
c |
=====> s urface card s |
|||
|
c |
====== ====== ====== ========= ====== ====== = ========= ====== |
|||
|
c |
=====> p i n i nt eri or |
|||
|
1000 |
R CC 0. 0. 0 . 0. 0. 1 34. 0.5 00 $ A d jus t for axi al zoning reg i ons |
|||
|
1005 |
R CC 0. 0. 13 4. 0 . 0. 133. 0. 500 |
|||
|
1006 |
R CC 0. 0. 26 7. 0 . 0. 133. 0. 500 |
|||
|
1007 |
R CC 0. 0. 0 . 0. 0 . 4 00. 0.5 00 |
|||
|
1001 |
R CC 0. 0. 0 . 0. 0. 4 00. 0.6 14 |
|||
|
1002 |
R CC 0. 0. 0 . 0. 0. 4 00. 0.6 15 |
|||
|
1003 |
R CC 0. 0. 0 . 0. 0. 4 00. 0.6 24 |
|||
|
1004 |
R CC 0. 0. 0 . 0. 0. 4 00. 0.6 25 |
|||
c =====> p i n s hel l
2000 HEX 0 . 0. 0 . 0. 0. 40 0. 0. 1 . 0 .
c =====> as s em b l y i nne r
3000 HEX 0 . 0. 0 . 0. 0. 40 0. 0. 1 0. 000 0.
3001 HEX 0 . 0. 0 . 0. 0. 40 0. 0. 1 0. 010 0.
3002 HEX 0 . 0. 0 . 0. 0. 40 0. 0. 1 1. 790 0.
3003 HEX 0 . 0. 0 . 0. 0. 40 0. 0. 1 1. 800 0.
c =====> as ee m b l y s hel l
4000 HEX 0 . 0. 0 . 0. 0. 40 0. 0. 1 2. 0 .
c =====> react or v es s e l
5000 R CC 0. 0. 0 . 0. 0. 4 00. 290 . $ Press ure ves s el for 12 ri ng s
5001 R CC 0. 0. -5 0. 0 . 0. 500. 290 .
5002 R CC 0. 0. -5 0. 0 . 0. 500. 300 .
5003 R CC 0. 0. 40 0. 0 . 0. 50. 290 .
5004 R CC 0. 0. -5 0. 0 . 0. 50. 290 .
c =====> contr ol pin
8000 R CC 0. 0. 39 9. 0 . 0. 1. 1.5 00
8001 R CC 0. 0. 39 9. 0 . 0. 1. 1.5 10
8002 R CC 0. 0. 39 9. 0 . 0. 1. 2.0 00
8003 HEX 0 . 0. 3 99. 0 . 0. 1. 0. 10. 00 0 0.
8004 HEX 0 . 0. 3 99. 0 . 0. 1. 0. 11. 79 0 0 .
8005 HEX 0 . 0. 3 99. 0 . 0. 1. 0. 11. 80 0 0.
8006 HEX 0 . 0. 0 . 0 . 0. 3 99. 0. 11. 80 0 0.
c ====== ====== ====== ========= ====== ====== = ========= =====
c ====== ====== ====== ========= ====== ====== = ========= =====
c =====> d ata car d s fo r p ro b l em
c ====== ====== ====== ========= ====== ====== = ========= ======
c ====> run d es cri p t i on k cod e 10 000 1 . 0 40 13 0 R AN D G E N = 2
i m p : n 1 69r 0 .
c =====> s ource d es cri p t i on
s d ef erg =d 1 cel =d 2 ax s =0 0 1 rad =d 3 ex t =d 4
s p 1 -3 . 988 2 . 249 $ Wat t s p ect rum , t her m al u2 35 fi s s i on s p 2 D 1. 1r
s i 2 L 500: 400(0 0 0):30 0: 202(0 0 0): -1 01
500 : 400(1 1 0):3 00 : 202(0 0 0): -10 1
|
c |
500 : 400( 4 7 0): 30 0: 202(0 0 0): -1 01 |
|
|
c |
500 : 400( -3 -7 0) : 300: 202( 0 0 0): -101 |
|
|
c |
500 : 400( -6 0 0) : 3 0 0: 202(0 0 0): - 101 |
|
|
c |
500 : 400( 5 0 0): 30 0: 202(0 0 0): -1 01 |
|
|
s p 3 |
C |
0 . 1. |
|
s i 3 |
H |
0. 0. 5 |
|
s p 4 |
C |
0 . 1. |
|
s i 4 |
H |
0. 400 . |
c =====> m ater i al card s
m 1 92 238 2 . 6667e -1 $ UO2 fuel enri ch 2 0 p e rcen t 92235 0. 666 7e-1
8016 6 . 6667 e-1
m 11 9 2238 2 . 8333 e -1 $ fuel enri ch 15 p ercent 92235 0. 500 0e-1
8016 6 . 6667 e-1
m 12 9 2238 3 . 0000 e -1 $ fuel enri ch 10 p ercent 92235 0. 333 3e-1
8016 6 . 6667 e-1
m 13 9 2238 2 . 5000 e-1 $ fuel enri ch 25 p ercent 92235 0. 833 3e-1
8016 6 . 6667 e-1
m 14 9 2238 2 . 7472 5e -1 $ fuel enri ch 17 . 5 p e rcent 92235 0. 582 75e-1
8016 6 . 6667 e-1
m 15 9 2238 2 . 9137 5e -1 $ fuel enri ch 12 . 5 p e rcent 92235 0. 416 25e-1
8016 6 . 6667 e-1
|
24053 |
-0. 00 831337 5 |
|
24054 |
-0. 00 206937 5 |
|
24050 |
-0. 00 380187 5 |
|
14028 |
-0. 00 36892 |
|
14029 |
-0. 00 01868 |
|
14030 |
-0. 00 0124 |
|
42092 |
-0. 00 14098 |
|
42094 |
-0. 00 087875 |
|
42095 |
-0. 00 15124 |
|
42096 |
-0. 00 15846 |
|
42097 |
-0. 00 090725 |
|
42098 |
-0. 00 229235 |
|
42100 |
-0. 00 091485 |
|
25055 |
-0. 00 45 |
|
23050 -0. 0 000053 75 23051 -0. 00 215000 0 |
|
|
28058 |
-0. 00 272308 |
|
28060 |
-0. 00 104892 |
|
28061 |
-0. 00 00456 |
|
28062 |
-0. 00 014536 |
|
28064 |
-0. 00 003704 |
|
41093 |
-0. 00 08 |
|
6012 |
- 0. 000 989 |
|
6013 |
- 0. 000 011 |
|
7014 |
- 0. 000 697438 |
|
7015 |
- 0. 000 002562 |
m 2 24 052 -0 . 0733 1537 5 $ s t ai nl es s s t eel T 91
c
15031 -0. 00 02
16032 -0. 00 009502
16033 -0. 00 000075
16034 -0. 00 000423
c 16035 -0. 0 000000 2
|
26054 |
-0. 05 13619 |
|
26056 |
-0. 81 222646 |
|
26057 |
-0. 01 94821 |
|
26058 |
-0. 00 247954 |
m 21 2 4052 -0 . 100 548 $ s t ai nl es s s t eel outer i nner p rote ct i ve
|
24053 |
-0. 01 1401 |
||
|
24054 |
-0. 00 2838 |
||
|
24050 |
-0. 00 5214 |
||
|
14028 |
-0. 01 8846 |
||
|
14029 |
-0. 00 0934 |
||
|
14030 |
-0. 00 062 |
||
|
25055 |
-0. 00 01 |
||
|
28058 |
-0. 00 006807 7 |
||
|
28060 |
-0. 00 002622 3 |
||
|
28061 |
-0. 00 000114 |
||
|
28062 |
-0. 00 000363 4 |
||
|
28064 |
-0. 00 000092 6 |
||
|
6012 |
- 0. 000 00989 |
||
|
6013 |
- 0. 000 00011 |
||
|
15031 |
-0. 00 006 |
||
|
16032 |
-0. 00 005701 2 |
||
|
16033 |
-0. 00 000045 |
||
|
16034 |
-0. 00 000253 8 |
||
|
c 16035 26054 |
-0. 00 000001 2 -0. 04 986086 |
||
|
26056 |
-0. 78 848932 4 |
||
|
26057 |
-0. 01 891274 |
||
|
26058 |
-0. 00 240707 60 |
||
|
m 3 |
82 204 -0 . 0062 3 $ L B E |
||
|
82206 |
-0. 10 7245 |
||
|
82207 |
-0. 09 8345 |
||
|
82208 |
-0. 23 318 |
||
|
83209 |
-0. 55 5 |
||
|
m 4 |
80 16 |
0. 499805 |
$ refl ect or Mg O |
|
8017 0 . 0001 95 |
|||
|
c |
8018 0 . 0010 05 |
||
|
12024 0 . 3949 5 |
|||
|
12025 0 . 05 |
|||
|
12026 0 . 0550 5 |
|||
|
m 5 |
50 10 0. 1592 |
$ co nt rol rod s B 4C |
|
|
5011 |
0 . 6408 |
||
|
6012 |
0 . 1978 |
||
|
6013 |
0 . 0022 |
||
m 6 50 10 0. 1592 $ S hi el d B 4C
|
5011 |
0 . 6408 |
||
|
6012 |
0 . 1978 |
||
|
6013 |
0 . 0022 |
||
|
m7 |
7015 0. 5 |
$ Urani u m M ono ni t ri d e |
|
|
92238 0 . 45 |
$ 10% en ri ched |
||
|
92235 0 . 05 |
|||
|
m 71 |
7015 0. 5 |
$ 12. 5% e nri ched |
|
|
92238 0 . 4375 |
|||
|
92235 0 . 0625 |
|||
|
m 72 |
7015 0. 5 |
$ 15% enri ched |
|
|
92238 0 . 425 |
|||
|
92235 0 . 075 |
|||
|
m 73 |
7015 0. 5 |
$ 17. 5% e nri ched |
|
|
92238 0 . 4125 |
|||
|
92235 0 . 0875 |
|||
|
m 74 |
7015 0. 5 |
$ 20% enri ched |
|
|
92238 0 . 4 |
|||
|
92235 0 . 1 |
c ===== > t al l i es :
F 6: N 101
F 16: N 500
c F26 : N (1 01<202 [0 0 0]<300<4 00[0 0 0 ]<500)
F 7: N 101
F 17: N 500
P art VI I I
App endix C: In v estigation of PCHE thermal h ydraulics
In tro duction
This app endix explai ns the computational mo del that w as used to study thermal h ydraulics p erformance of PCHEs. The thermal h ydraulic paramete r s of 35 MW PCHEs w ere studied as a function of c hannel configuration, c hannel diameter and hot fluid mass flo w rate. The impact of these design parameters on PCHE v olume, heat transfer prop e r tie s and pressure drop are discusse d in the follo wing sections. The results of this study informed the design c hoices for the PCHEs in the Pro cess Heat System.
Computational mo del
A n o dal computational mo del implemen ted in F ortran, dev elop ed at M IT b y Hejzlar and Dostal [54] and impro v ed up on b y J.Hejzlar [79] and Shirv an [137], is b eing u s ed to optimize the PCHEs in the Pro cess Heat system. The mo del assumes that:
• The total mass flo w rate is uniformly d istributed among the c hannels,
• The w all c hannel temp erature i s uniform at ev ery axial no de and
• Cold and h ot plates ha v e the same n um b er of fl o w c hannels.
The fortran co de uses a coun terflo w c hannel configuration and can b e used to analyze b oth straigh t and zigzag c han nels. Giv en the the inlet temp eratures and pressures of the hot and cold fluids as w ell as the n um b er of no des in a c han ne l , the c o de iterates to find the length of eac h no de and sums them to calculate the length of eac h c hannel. Sho wn in Figure 55 is a s c hematic of the treatmen t of a single PCHE c hannel b y the no dal mo del.
The mo del use s a constan t Nu of 4.089 for the laminar regime (Re < 2300), linear in terp olation is used to find Nu for the transitional regime (2300 < Re < 5000) and the Gnielsinki correlation is used for the tur bulen t (Re > 5000) regime [54]. Single phase and t w o phase pressure drops are mo deled using exp erimen tal results from the T oky o Institute of T ec hnology [76] and the T a ylor correlation [78] resp ectiv ely .
Figure 55: PCHE no dalization [54]
Comparison of straigh t and zigzag c hannel PCHEs
PCHEs can either ha v e straigh t or zigzag c hann e ls. F or t w o PCHEs of the same heat rate, the zigzag c han nel PCHEs are more compact but ha v e higher pressure d rops. T able 25 compares a straigh t c hannel PCHE with
a zigzag c hannel PCHE. F or b oth PCHEs, S-CO 2 and He are the h ot and cold fl uids resp ectiv ely . Mass flo w rates of 22 kg/s and 90 kg/s w ere used for the cold and hot fl uid resp ectiv ely . A b end angle of 57.5 C , as recomm end e d b y Heatric and studied b y Shirv an [137], w as used for the zigzag c hannel. Both the zigzag and straigh t c hannels ha v e a c hannel diameter of 2 mm. As seen in T able 25, PCHEs ha v ing zigzag c hannels ha v e b etter heat transfer prop erties. In a zigzag c hannel, b ends increase the flo w turbulence and impro v e fluid mixing [79]. This in c reases the heat transfer co effic i e n t of b oth the hot an d the cold fluid and results in impro v ed heat transfer. Ho w ev er, increasing flo w turbulence using b ends, as seen in T able 25 , increases the pressure drop. F urther study w as carried out on the zigzag c hannel PCHE due to its significan tly sm al le r v olume.
T able 25: Straigh t c hannel and zigzag c hannel P CHEs
|
Channel |
Zigzag |
Straigh t |
|
V [ m 3 ] |
2.99 |
4.37 |
|
h cold [ W / K − m 2 ] |
2192.32 |
959.2 |
|
h hot [ W / K − m 2 ] |
3280.89 |
524.7 |
|
Re cold |
1209.28 |
450.68 |
|
Re hot |
10338.09 |
3855.18 |
|
ΔP cold [ P a ] |
75130.55 |
5532.84 |
|
ΔP hot [ P a ] |
52904.25 |
2945.744 |
Channel diameter study
In this p ortion of the study hot and cold c hannel diameters w ere v aried and their impact on the zigzag c hannel PCHE v olume w as studied for PCHE1. PCHEs are diffusion b onded whic h results in a high cost of fabrication. Both materials and fabrication costs are prop ortional to the v olume of the PCHE and the smallest p os sible v olume is desirable in order to reduce the capital cost of the Pro c ess Heat system. In order to study the effects of c han nel diameter on the PCHE v olume, b oth hot cold c hannel diameters w ere assumed to b e of the s ame size. A range of c han nel diameters from 1 mm t o 5 mm w ere studied and, as sho wn in Figure 56a, it w as fou nd that the PCHE v olume increases linearly with an increase in c hannel diameter. This increas e in v olume can b e attributed to a decrease in the heat transfer co efficien t as sho wn in F igure 56b. Ho w ev er, one of the disadv an tages of using smaller c hannels is a larger pressure drop. The pressure drop whic h is of the order of kP a, is small compared to the op erating pressure. The next sec ti on discusses the the impact of S-CO 2 mass flo w rate on the PCHE v olume.
(a) PCHE V olume
(b) Heat transfer co efficien ts
Figure 56: PCHE v olume and heat transfer co efficien t as a fun c tion of c hannel diameter
S-C O 2 mass flo w rate study
Previous s tu dies indicate that c hannel diameters less than 2 mm suppress eddies and reduce flo w turbul e n c e, th us adv ersely affecting heat transfer [79]. As a result, further optimization for PCHE1 w as p erformed b y v arying the S-CO 2 mass flo w rate for a zigzag c hannel coun terflo w configuration ha ving h ot and cold c hannel diameters of 2 mm eac h. The S-CO 2 mass flo w rate w as v aried from 90 kg/s to 150 kg/s and th e effect of v arying this mass flo w rate on the hot and cold fluid Reynolds n um b er, pressure drop, heat transfer co efficien t and PCHE v olume w ere studied. The cold fluid mass flo w rate w as fixed at 22 kg/s due to its
b eing constrained b y the required heat rate for PCHE1 (35 MW) and the outlet temp erature needed to meet the heat storage temp erature requiremen ts.
Reynolds n um b er and pressure drop
As seen in Fi gure 57a, the cold fluid is in the lam i nar flo w regime and its Reynolds n um b er is insensitiv e to the mass flo w rate of the h ot fluid whereas th e Reynolds n um b er of the hot fluid increases l inearly with an increase in the mass flo w rate. Figure 57b sho ws that the c old fluid pressure d rop decreases with an increase in the mass fl o w r ate of the hot fluid. As explained in the next section, this o ccurs b ecause a higher mass flo w rate of the hot fluid increases its turbulence and impro v es heat transfer whic h leads to a re d uction in the length of b oth hot and cold c hannels. F or the sam e m ass fl o w rate an d c hann e l diameter, a short e r c hannel results in a smaller pressure drop for the cold fluid.
The pressure drop of the h ot fluid first decreases and then increases with an increase in the hot fluid mass flo w rate. A c hange in the mass flo w rate from 90 kg/s to 100 kg/s , reduces the pressure drop b ecause the decrease in c hannel length is greater than the increase in the pressure drop due to an increase in frictional losses. Ho w ev er, for larger mass flo w r ate s, the latter dominates and the hot fluid pressure drop is seen to increase as a function of the mass flo w rate of the hot fluid.
(a) Reynolds n um b er
(b) Pressure drop
Figure 57: Reynolds n um b er and Pressure drop as a function of S-CO 2 mass flo w rate
Heat transfer co efficien t and PCHE v olume
As sho wn in Figure 58a, the heat transfer co effic i e n t of the cold fluid is insensitiv e to the m ass flo w rate of the hot fluid. Ho w ev er, the heat transfer co efficien t of the hot fluid increases with an increase in its mass flo w rate. This increases the total heat transfer co efficien t and, as seen in Figure 58b, reduces the PCHE v olume. Coun terin tuitiv ely , the total heat transfer co effi cien t (HTC) of a PCHE is not the a v erage of the hot and cold HTCs. The total HTC is calculated using th e expression in Equation 44 [137] in whic h h h and h c are the HTCs for the hot and cold fluids res p ectiv ely , c l is the conduction l e n gth, and k is the therm al conductivit y of the PCHE plates.
1
h = (
+ c l P h +
P h ) − 1
(44)
tot
h h k 2 P h c P c
(a) Heat transfer co efficien ts
(b) PCHE v olume
Figure 58: Heat transfer co efficien ts and PCHE v olume as a function of S-CO 2 mass flo w rate
Conclusion
This w ork indicates that zigzag c hannel PCHEs are significan tly smaller th an straigh t c hannel PCHEs ha ving the same op erating parameters. It w as also observ ed that th e PCHE v olume increases with an increase in the c hann e l diameter and dec r e ases with an increase in the mass flo w rate of the hot fluid.
P art IX
App endix D: Implemen tation of switc hgrass as feedsto c k for a industrial biofuels pro cess in a n uclear complex
Abstract
This article describ es new calculations that con tribute to this study of the use of sw itc hgrass ( Panicum vir gat u m ) in a n uclear-p o w ered biorefinery . Areas examined include lo cation, transp ortation metho ds and c osts, carb on emissions, and injection of feedsto c k as fuel for the pro cess.
Bac kground
The curren t dev elopmen t of the n uclear reactor complex b y the 22.033 design team consists of sev e r al c hallenges ro oted in the pro vision of enough resources to the differen t sectors at the appropriate rates. The biofuels refinery has to b e scaled accordin g to the amoun t of h ydrogen a v ailable for the refining pro c edu re . The gasification and distillation pro cesses c on tained within the biorefinery are dep enden t on the a v ailable p o w er from the pro ces s heat, and therefore, since h ydrogen is the ultimate limiting factor, the pro cess heat designers m ust engineer to sustain those infrastructures.
With the p o w er a v ailable from the lead-co oled fast reactor at 3. 6 GeV, the h ydrogen facilit y is able to pro vide 7.9 kg/s of h ydrogen gas to our refining pro cess. The biorefinery is th us allo w ed to op erate with an input of 2,903 t/d 1 (24.38 kg/s), whic h amoun ts to 9.2 kg/s of end-pro duct pro duced after pro cessing. T o mee t these needs, the pro cess heat group is pro viding a large heat source to the biorefinery , whic h will allo w for 13.7 MW of steam to b e used i n the gas ifi c ati on pro cess. The co oling of syngas from this pr o cess to the acid-gas remo v al pro cess will pro vide 19 MW bac k to the heat source. Based on curren t crude oil and natural gas prices, the bi ofuels facilit y should b e exp ected to ge n e rat e ab ou t $ 1.7 M in re v en ue on a dail y basis from end pro ducts, all the while con tributing to an American-based sustainable fuel source.
In tro duction
The design requi re men ts surrou nding the successful implemen tation of thes e flo w rates rev olv e around fa cilitating the injection of th e fee d s to c k in to the r e fi nery , ha ving a reliable heat dump, and co ordinating the op erating times of all sectors (i.e. the core life cycle and sh utdo wn p erio d of the h ydrogen facilit y .) A natural reserv oir is required for co oling the Fisc her-T ropsc h (FT) pro cess, and t his t re atmen t m ust meet EP A standards for en vironmen tal impact. Bac kflo w prev en ters m ust b e installed in all pip elines in v olving large amoun ts of flammable gas, suc h as the syngas and h ydrogen pip elines, in the case of large pressure w a v es. Multiple flares can also b e used to get rid of excess pressure.
The site m ust allo w ro om for temp orary switc hgrass (SG) storage, and this in turn m ust b e lo cated a safe distance fr om the hold ing tanks for the biofuels due to flammabilit y . Storage m ust b e climate-con trolled and ha v e a system of feeding biomas s in to a con v ey or system. If temp orary storage is at capacit y , there m ust b e op en grounds up on whic h excess material can b e dump ed to b e used later. This requires loading op erators to b e on hand on a con tin ual basis, and for the moistur e con ten t to b e monitored for optimal le v els. Storage of material should amoun t to holding three da ys w orth of feedsto c k in case of shortages caused b y traffic congestion, sev ere w eather or natural disasters. In an y c ase, this w ould allo w enough time for the engineers to co ordinate a refinery sh utdo wn with minimal impact.
Loading ba ys for SG m ust ha v e direct access to ma jor high w a y s y s tems and b e designed to reduce on-site traffic conditions. T h e distance b et w een the facilit y and the cropland pro viding the feedsto c k s h ould b e les s than 200 km, and SG densification should b e p erformed as close to the l and as p ossible to allo w for more efficien t transp ortation and few er c ar b on emissions. The grain truc ks used for transp ortation are limited
1 These v alues are based on an appro ximately 20 hr daily op erating p erio d.
Map produced by the National Renewable Energy Laboratory for the U.S. Department of Energy.
Figure 59: Map of biomass resources in T exas. Harris C ou n t y is notice ab le in dark green to the lo w er righ t. (National Renew able Energy Lab oratory)
b y the al lo w ed tonnage of the high w a y system used. They m ust feature an efficien t unloading m ec hanism, p ossible b y h ydraulic-p o w ered dumping, and the p elletized SG should b e con tained in suc h a fashion that minimizes op eration al losses. Ideally , some sort of grain e l e v ator will b e used to feed a lo c k-h o p p er, whic h will directly feed in to the gasifier.
Lo cation
Harris Coun t y , T exas is a pr im e lo cation for the reactor complex. Lo cated in the southeastern p ortion of the state, it is situated near a large natural b o dy of w ater and has considerable biomass resources [91] (see Figure 59). The region is in a lo w-risk area for wildfire [21] and receiv es around 65 inc hes in ann ual precipitation [155], and has a climate that is less harsh than more established gro wing regions in the midw est. It is considered part of the greater Houston metrop olitan area, and the nearb y T rinit y Ba y can b e used as a natural reserv oir for co oling the FT pro cess and as a general source of w ater for in tak e. (This en vironmen tal resource will b e sub ject to an EP A-limited 36 C C on temp eratu re increases.)
×
T o meet the daily reactor requiremen t of 2903 t/d (1 . 06 10 6 t/y r ), an d considering that SG is a p erennial crop that can b e cultiv ated at 14.6 t/ha and harv ested t wice a y e ar, the gro wing region will need to co v er at least 36,287.5 ha (140.1m 2 ). T o allo w for a consisten t input of feedsto c k, the engineers will need to set up a net w ork of suppliers (b oth primary and sec on dary/bac k-up) that can ensure a dep endable flo w of pro du c t. There are existing high w a ys in all d irec ti o n s from T rinit y Ba y that can allo w for m ultiple pro viders, and
Harris Coun t y is surrounded b y other regions con taining considerable biomass resources.
T ransp ortation
Road
×
Grain tru c ks can b e use d to transp ort densified SG from the farmland to the storage facilities on the reactor site. If the SG is compressed to a densit y of ab out 1300 kg/m 3 , a large 2 ft grain b ed with a 905 ft 3 (25.6 m 3 ) capacit y should b e able to haul at leas t 3 . 4 10 4 kg (33.3 t) of feedsto c k. T o meet the daily requiremen t for feedsto c k, this w ould imply around 85 trips to the reactor site on a daily basis (not ac kno wledging r e li e f from the amoun t a v ailable in storage.)
If College Station, TX w as tak en as the nearest gro wing lo cation, this w ould mean that 98.1 mi (157.9 km) w ould need to b e tra v eled on a 1.6 h r 2 one w a y trip. A t ypical P e terb ilt @ truc k has a fuel efficiency of ab out 10 miles p er gallon [ 9], so ab out 20 gallons of diesel are consumed on a round-trip, amoun ting to ab out $ 70 w orth of fuel. F or a full da y of refinery op e r ation, this w ould impl y an exp ense of $ 5,950.
The truc ks w ould n e ed to tra v el along US-290 W and TX-6 (T exas State High w a y), and could p ossibly resort to the I-45 N and TX-105 W to a v oid congestion (although this w ould mean a longer trip). If w e use a P eterbilt Mo del 388 Da y Cab (fron t/tand e m rear axle of 12,000/40,000 lbs) with a load of 66,600 lbs and a tare of 4,440 lbs, a tonnage limit of 80,000 lbs o v er 6 axles[20] can b e met.
Rail
T ransp ortation b y rail is highly feasible in Harris Coun t y due to existence of a w e ll - d e v elop ed railroad net w ork, although it in tro duces trade-offs in adjusting the amoun t of freigh t cars for a particular bulk load needed b y th e refinery . Ho w ev er, it allo ws larger amoun ts of feedsto c k to flo w un im p eded b y ground traffic patterns and for a more systematized unloading system, along with greater loads and capacities through freigh t c ar s .
The Un ion P acific (UP) Railroad w ould comprise most of the rail net w orks used. A large UP 65 ft (19.8 m) co v ered hopp er car with three compartmen ts and a capacit y of 5,200 ft 3 (147.24 m 3 ) can haul 191,425 k g of densified SG, whic h is just under the lo w er load limit for these cars [117]. T o supp ly the reactor on a daily op erating basis, that w ould amoun t to at least 14 cars at full capacit y deliv ering to the reactor site.
The site will need to h a v e lo cal rails that will ensure the cars can remain idle or in use for the prop er amoun t of time without in terfering with regular rail traffic. These hopp er cars will gra vi t y-feed the SG p ellets in to the unloading and con v eying system for temp orary storage. If storage is at capacit y , then the cars can use an alternativ e rail to dump on to op en storage space, although this presen ts the risk of exp osing the p ellets to v ariable amoun ts of moisture from the en viron m en t.
Storage
SG in its p elletized form will amoun t to 2,026 m 3 of v olume used p er da y of op eration. T o op erate with a guaran teed flo w of feedsto c k, storage s h ould s u pply for m an y da ys of op eration throughout the gro wing season. If w e get a full yi e ld t wice a y ear from 140 mi 2 of farmland, and if w e use a large grain elev ator of
×
× × ×
1 . 4 10 7 bu (4 . 8 10 5 m 3 ) [35], w e can supply for a maxim um of 230 da ys of op eration. With 4 . 8 10 8 kg as an upp e r estimate yield of 140 mi 2 of farmland in half a y ear, this w ould amoun t to 3 . 7 10 5 m 3 of densified feedsto c k, whic h is enough to fit en tirely in suc h a grain elev ator. Th us, economic factors will come in to pla y in to deciding whether to build suc h a facilit y on site, to build a smaller one, or t o emplo y on of the man y existing grain elev ators around Harris Coun t y (see [35] for examples).
Lo cal managers of grain elev ators w ould store p elletized feedsto c k un til they can b e transp orted to the site f or temp orary storage and implemen tation. T emp orary storage w ould amoun t to a small grain elev at or of appro ximately three op erating da ys capacit y sp ecifically engineered for con v eying feedsto c k in to the gasification pro cess .
2 The truc k can b e assumed to ha v e an a v erage sp eed of 60 mi/hr.
Emissions
Carb on emissions are limited b y the EP A to 15 g/bhp-hr (6.1 g/MJ) of nitric o xides and Non-Methane Hydro-Carb ons (NMHC). F ollo wing these standards, on a t ypical da y of round-trip site-to-reactor driving, there are 5,100 g/bhp of CO emissions and 850 g/bhp in NO x /NMCH. With a total w eigh t of 83,000 lbs (38,000 kg) tra v eling at 60 mi/hr (26.8 m/s) for 4 hours a da y , with 85 trips to supply the refinery , this w ould am ou n t to appro ximately 40 kg CO and 7 kg NO x /NMCH emitted in a single da y .
Emissions resulting from rail transp ort are regulated b y the EP A at 1.28 g/bhp-hr of CO and 4.95 g/bhp hr of NO x for a mo dern engine [10]. Since lo com otiv es t ypically consume 1 gallon of fuel p er 400 ton-miles [10], a General Electric 8-cylinder mo del 7FDL with a con tin uous p o w er out put of 2,045 bhp (1,525 kW) at 1,050 RPM [56] can hau l 2,903 tons from College Station in app ro ximately 5 hours, using ab out 700 gallons of fuel. If these engines meet EP A standards, they should th us b e exp e cted to emit around 13 kg CO and 51 kg NO x ev ery trip.
Conclusion
With the ge n e r al conditions in place for buildi ng the reactor complex in T exas, further m easures can b e tak en suc h as assessing the reactor safet y and disaster p rev en tion mec hanisms, a v ailabilit y of lab or, cost of distribution, and en vironmen tal regulation s , among other things. It will b e id e al in defending this c hoice of lo cation to use sup erimp ose d maps of climate, precipitation, etc, and to compare with GIS maps of p oten tial biomass gro wing regions. It is also necessary to establish a comm unication net w ork b et w een engineers at the complex and farmers in the region to ensure a steady flo w of fee d s to c k.
Figure 60: Silv a gas and FICBC for pro duct selectivit y
P art X
App endix E: Impact of gasifier design on FT pro duct selectivit y
A gasifier pro duces a syngas stream whic h is then fed in to FT reactor. Differen t gasifier designs will pro duce syngas streams of differen t comp ositions and H 2 /CO ratios. Since the H 2 /CO ratio critically impacts pro duct selectivit y , it is necessary to ev aluate ho w differen t gasifier d e sign c h oic es affect final pro duct comp osition. Dep ending on the final pro duct requiremen t, the condi tions of the reactor can b e mo dified to alter the comp osition of pro duct F-T liquid. The parameters of the reactor ha v e b een c hosen to maximize the pro duction of carb on n u m b e r 5-20 molecules whic h can b e used for biogasoline and bio diesel pro duction. Tw o designs of gas i fie r w ere prop osed and their final pro duct selectivities w ere analyzed using Equation 45 for the prob abilit y of c hain gro wth.
0 . 23
2
α = H /C O + 1 + 0 . 63 · [1 − 0 . 0039( T − 533 K )] (45)
T = 240 C C = 513 K is the op erating temp erature of the reactor and for Silv a gas gasification pro cess whic h is an American des i gn, the h ydrogen to CO ratio in the syngas s tr e am is H 2 /C O = 22 / 38 . 2 = 0 . 57 ≈ 0 . 5 . F or the Sw edish FICBC pro cess, H 2 /C O = 44 . 4 / 22 . 9 = 1 . 94 ≈ 2. W e can then compare the c hain gro wth probabilit y of the t w o differen t gasifier designs at the same temp erature b y using Equations 46 and 47.
α S il v a =
0 . 23
2
H /C O + 1 + 0 . 63 · [1 − 0 . 0039( T − 533 K )] = 0 . 84 (46)
α F I C B C =
0 . 23
2
H /C O + 1 + 0 . 63 · [1 − 0 . 0039( T − 533 K )] = 0 . 76 (47)
W e can then ap ply ASF d is t ribution to calculate mass fraction of molecules with carb on n um b er, n, u s in g Equations 48 and 49.
n
χ S il v a = 0 . 16 2 n · 0 . 84 n − 1 (48)
n
χ F I C B C = 0 . 24 2 n · 0 . 76 n − 1 (49)
Figure 60 sho ws mass fraction as a function of carb on n um b er for b oth pro cesses.
Naph tha is a term that refers to mixture of h ydro carb on molecules ha ving carb on n um b ers b et w een 5-12. The term distillate co v ers h ydro carb ons with n b et w een 12-20. W ax refers to carb on n um b er 20 or more
h ydro carb ons. Using this definition, w e can tabulate the mass flo w rate for naph tha, distillate and w ax streams. Silv agas is c hosen as it pro duces more hea vy diesel pro ducts an d is a design that is paten ted in the US. This will impact p ositiv ely on feasibilit y of our design.
App endix E: The UT-3 Hydrogen Pro duction Pro cess
Bac kground
UT-3 h ydrogen pro duction is a four-staged thermo c hemical w ater splitting pro cess using four separate reactor units connected in series, whic h undergo the follo wing reactions at the ind icate d desired temp eratures. Eac h set of calcium reactors and iron reactors are cyclically link ed suc h that th e pro ducts of one reaction b e come the reactan ts of the other, and can op erate con tin uously with a p erio dic rev ersal of th e flo w direction of gaseous com p ounds [83] .
C aB r 2 ( s ) + H 2 O ( g ) → C aO ( s ) + 2 H B r ( g ) (50)
C aO ( s ) + B r 2 ( g ) → C aB r 2 ( s ) + 0 . 5 O 2 ( g ) (51)
F e 3 O 4 ( s ) + 8 H B r ( g ) → 3 F eB r 2 ( s ) + 4 H 2 O ( g ) + B r 2 ( g ) (52)
3 F eB r 2 ( s ) + 4 H 2 O ( g ) → F e 3 O 4 ( s ) + 6 H B r ( g ) + H 2 ( g ) (53)
Multiple scalings of a UT-3 h ydrogen pro duction p lan t ha v e b een done previously using th e soft w are ASPEN-PLUS [130] and other means to determine th e thermal p o w er required for a particular h ydrogen pro duction rate [145]. Eac h reference used for the scaling of this UT-3 h ydrogen pro d uction plan t w as for a similarly designed UT-3 reactor system to ensure the most accurate extrap olations p ossible f or the h ydrogen pro duction rate of in terest in this study . F uture analysis with ASPEN or other similar soft w are w ould b e ideal to ensure the accuracy of the required thermal p o w er for the requested h ydrogen mas s flo w b y the biofuels pro duction plan t; ho w ev e r , e x trap olation us i ng studies that ha v e utilized suc h soft w are pro vides the most acc u rate v alue for the thermal p o w er requiremen t presen tly p ossible.
Reactions 50 and 51 in the UT-3 pro c ess o ccur in t w o differen t calcium reac tor units. P ellets con taining either CaBr 2 or CaO as the initial comp ounds will transform in to the other calcium reagen t cyclically during the con tin uous h yd roge n pro duction op eration of the UT-3 plan t. There is a 76% v olumetric difference in the structure of the t w o calcium comp ounds, whic h could cause fines to form and p ellet sin tering as cycling progresses [94]. Supp orts to stabilize the calcium reagen ts, and w ell as p orous configurations of CaO micro- p ellets to allo w space for expansion and con traction during the UT-3 cycle, w ere researc hed and implemen ted in eac h calcium reactor unit to ensure the structural in tegrit y and material s tab ilit y of the calcium reagen t p ellets.
The optimal temp eratures for eac h reactions w ere iden tified as 760 C C, 572 C C, 220 C C, and 560 C C
resp ectiv ely , as sho wn in Figure 61. Steam is used to b oth react with solid c hemicals to pro duce the desired pro ducts and mix with and transp ort gaseous pro ducts to the next reactor. Once the reactions ha v e run to completion in a forw ard progression, the flo w of the steam cycle wil l b e rev ersed, utilizing the pro ducts remaining in the reactors as the reactan ts for the corresp onding rev erse reaction. Figure 61 depicts a sc hematic of the UT-3 plan t in forw ard flo w. Heat exc hangers (denoted b y the orange, crosse d circles) are placed in b et w een the reactors to ensure that the gaseous reactan ts en ter the reactor at the correct temp erature. Tw o compressors (one for forw ard flo w, one for bac kw ard flo w) are placed in series with the reactors, creating the p re ssure differen tial to sustain the flo w progression of the gaseous pro ducts. H 2 and O 2 separators remo v e the pro ducts from the system; the H 2 will b e sen t to a bio fuels pro duction plan t while the exces s O 2 will b e sold or v en te d to the atmos p here .
UT-3 Plan t Design
Analytical Scaling of the UT-3 Hydrogen Pro duction Plan t
The analytic scaling of the UT-3 h ydrogen pro duction plan t w as cond ucte d using information from t w o pre vious studies of commercial scale UT-3 h ydrogen pro duction plan ts, is prese n ted in T able 26 for con v enience.
T able 26: Hydrogen Pro duction Rates and Thermal P o w er Requiremen ts
|
H 2 Pro duction Rate (Nm 3 h − 1 ) |
Thermal P o w er Requiremen t (MW) |
H 2 Pro du ction Rate (Nm 3 h − 1 ) p er MW |
|
20000 |
176.7 |
113.2 [145] |
|
20000 |
157.8 |
126.7 [145] |
|
30000 |
225.4 |
133.1 [130] |
Though there is a range of 20 MW o v er all v alues analyzed, the plan ts yielding the v alues of 126.7 Nm 3 h − 1 MW − 1 and 133.1 Nm 3 h − 1 MW − 1 are most similar to th e design used in this study , and pro vides a roughly constan t h ydrogen pro d uction rate p er MW suggesting the assumption of a linear b eha vior of h ydrogen pro duction rate p er MW is a reasonable one. Originally , only 0.7 kg s -1 w as required b y the b iofuel pro duction plan t, whic h corresp onds to 30090 Nm 3 h − 1 and th us w ould required appro ximately 226.1 to 237.29 MW of thermal p o w er. This scaling w as tak en simply as an estimation for the thermal p o w er required for the desired h ydrogen pro duction rate, and due to the large uncertain t y of whether h yd rogen pro duction rate p er MW actually scales linearly up to the des i red pro duction rate of 30090 Nm 3 h − 1 , v alues of thermal p o w er requested from the pro cess heat design team w ere up w ards of 300-400 MW. Though this requiremen t ma y seem arbitrary , without a more sophisticated an alysis for thermal p o w er requiremen ts using a soft w are pac k age suc h as ASPEN, it is clear the fa v orable s i tuation for the h yd roge n p ro duction plan t w ou ld b e a surplus of p o w er rather than the alternativ e. Should the h ydrogen pro duction plan t require less than 300-400 MW of thermal p o w er, the pro cess heat te am can redistribute that heat to either the biofuels pro duction plan t, or bac k to the core group to feed in to the secondary turbine to pro duce additional electricit y .
Ho w e v er, after subsequen t c hanges in required biofuel pro duction rate, the required pro duction of h y drogen w as raised to 7.9 kg s -1 . This larger pro duction rate of 316351 Nm 3 h − 1 w ould requir e appro ximately 2376.8 to 2496.9 MW of thermal p o w er using t w o most conserv ativ e v alues f or p o w er requiremen ts p er T able 26 . This quan tit y of thermal p o w er is to o large to b e pro vided b y the pro cess heat syste m, and th us with this new h ydrogen pro duction requiremen t, the UT-3 cycle is no longer a viable approac h for this design. This large thermal p o w er requ irem en t motiv ated the tr ansition to high-temp erature s team electrolysis (HTSE) for this h ydrogen pro duction plan t. Nev ertheless, should suc h a large h ydrogen pro duction rate not b e re quired, the UT-3 pro cess is an attractiv e h ydrogen pro duction approac h that has b een studied thoroughly for implemen tati o n on the sc al e of 0.5 - 0.75 kg s -1 h ydrogen pro duction rates.
Plan t Sc he m atic
The optimal temp eratures for eac h reactions ha v e b een iden tified as 760 C C, 572 C C, 220 C C, and 560 C C resp ectiv ely . The steam is used to b oth react with solid c hemicals to p ro duce th e desired pro ducts and to serv e as a w orking fluid to transp ort gaseous pro ducts to the next reactor unit. Once the reactions ha v e run to completion in a forw ard progression, the flo w of the steam cycle w i ll b e rev ersed, utilizing the pro ducts remaining in the reactors as the reactan ts for the corresp onding rev erse reaction. Figure 61 depicts a sc hematic of the UT-3 plan t in forw ard flo w. Heat exc hangers (denoted b y the orange, crossed circles) are placed in b et w een the reactors to ensure that the gaseous reactan ts en te r the reactor at the correct temp erature. Tw o compressors (one for forw ard flo w, one for bac kw ard flo w) are placed in series with the reactors, creating the p res sure differen tial to sustain the flo w progression of the gaseous pro ducts. H 2 and O 2 separators remo v e the pro ducts from the system; the H 2 will b e sen t to a bio fuels pro duction plan t while the excess O 2 will either b e sold or v en ted to the atmosphere.
Materials and Comp onen ts
Calcium Reagen t Structures
A prop os ed design for a stable calcium p ellet structure whic h can endure the cycling e x pansion and con traction of the calcium reagen ts during the UT-3 pro cess has b een thoroughly researc hed and is presen ted graphically in Figure 62 [131].
Figure 61: Blo c k diagram of the UT-3 plan t.
Calcium p ellets are formed using smaller p ellets that con tain b oth the reactan t CaO and binder CaTiO 3 that main tains structural in tegrit y while the CaO expands to CaBr 2 , and vice v ersa. The CaTiO 3 agglomer ations are essen tially solid spheres, whereas the CaO p ellet con tains a smaller substructure of CaO primary particles arranged uniformly thr oughout the smaller sphere with considerable space b et w een eac h primary CaO particle b efore bromination. While CaO un dergo es bromination and transforms in to C aBr 2 , the space b et w een the pr im ar y parti c les decreases due to the v olumetric expansion from CaO to CaBr 2 , and the r e action go es to completion just as the pr im ar y particles b egin to exert s t re sses on one another. Due to the presence of v oids for CaO to expand in to during bromination, the expansion of eac h substructure is fairl y small, and th us the re l ativ ely small v olumetric c hange of an en tire calcium p e l le t du ring cyc l ing b et w een CaO and CaBr 2 coupled with th e structural supp orts of CaTiO 3 pro vide a stable calcium structure whic h can endure the UT-3 pro cess cycles. This prop osed design w ould b e implemen ted in b oth calcium reactors in this UT-3 plan t to ensure the structural in tegrit y of calcium reagen t p ellets.
Hydrogen Separator Mem brane
Hydrogen me m brane separation tec hnology has emerged as an attractiv e option to the energy-in tensiv e pro cesses of cry ogenic d istillation and p res sure swing adsorption [28] , and consequen tly has b een c hosen as the metho d of H 2 separation for the UT-3 cycle. V arious t yp es of mem branes exist, eac h offe ri ng b enefits when optimized for the s ystem’s temp erature. Of these c hoices, me t allic mem branes (optimized at ˜350 C C) and ce ramic mem branes (optimized at ˜500 C C) w ere iden tified as c on tenders. Ceramic mem branes w ere ultimately c hosen as they presen ted few er p oisoning concerns [28]. Lo oking at the p ermeance of the v arious t yp es of ceramic mem branes, a Zr Silica [113] mem brane w as compared to the c hemical v ap or dep osition tetra-eth yl-ortho-silicate (C VD TEOS)[132]. The results of the comparison are summarized b y the table 27. The Zr silica w as c hosen sp ecifically b ecause it w ould requi re 39200 m 2 of mem brane area to ac hiev e the 7.9 kg/s compared to the CVD TEOS whic h r e qu ired 98000 m 2 .
Oxygen Separator Mem brane
The thermo c hemical decomp osition of w ater also pro duces o xygen in addition to h ydrogen whic h m ust b e remo v e d b efore the steam can b e recycled. Since o xygen in its gaseous form is m uc h larger than h ydrogen gas and is comparable to or larger than the s team molecules it m ust b e separated f rom , o xygen separation is not as e asy . Ho w ev er, o xygen-ion conducting ceramic mem branes ha v e b een sho wn to pro duce o xygen of
Courtesy of Elsevier, Inc., http://www.sciencedirect.com . Used with permission.
|
CVD (TEOS) |
Zr Silica |
|
|
Diffusion |
Solution-diffusion |
Solution-diffusion |
|
P ermeance [ mo l ] m 2 s P a |
− 8 4 . 0 ∗ 10 |
− 8 10 . 0 ∗ 10 |
|
Pressure [ M P a ] |
2 . 0 |
2 . 0 |
|
Area Needed [ m 2 ] |
1 , 240 |
496 |
|
Stabilit y Concerns |
Phase transition |
Phase transition |
|
P oisoning Conc ern s |
H 2 S , H C l , C o |
H 2 S , H C l , C o |
Figure 62: Prop osed Calcium P ellet Design [131]
T able 27: Co mp a r ison of CVD (TEOS) and Zr Silica ceramic mem branes v ery h igh purit y , separating out the molecules from a mixed gas.
Electric v oltage driv en separators w ere considered but dee med to b e to o energy in tensiv e giv en that the separation rates w ere directly prop ortional t o the v oltage applied.
Muc h m ore promising w ere found to b e mixe d conducting mem branes. Mem brane electrons create o xygen ions whic h then pass through the mem brane due to temp erature an d pressure differences on the t w o sides of the mem br a n e . Th us no electro des are required. The only mem brane material found to w ork at our temp erature of 500 C w as C eO 2 dop ed with S mO to pro vide electrical conductivit y .
Corrosion and Bromination
The pro duction of high temp erature bromic acid in the UT-3 pro cess is of considerable concern f rom a material science p ersp ectiv e. The industry standard for de al ing with high-temp erature corrosiv e substances are the allo ys F e -20 Cr and Ni-20 Cr. It has b een sho wn that under exp osure to high temp erature br om ic acid b oth of these allo ys form c hromium scales whic h prev en t the o xidation of the underlying iron. The b est of these t w o materials w as sho wn to b e the F e-Cr, whose easier scale formation b etter prev en ted o xidation and bromination than its nic k el coun terpart [116]. O ther corrosion resistan t coatings suc h as titanium carbide films w ere also in v estigated. It w as sho wn that titanium o xides form along with v olatile titanium bromides making micro crac ks in the coating [146]. F or these reasons F e-Cr w ould b e preferred going forw ard.
Figure 63: The mixed conducting mem brane pro ces s. O 2 is separated from t he steam mixture and diffused across as ions with the aid of a pump.
P art XI
App endix G: Excess 0 2 and Hydrogen Storage
Excess O 2
Oxygen is a v ersatile gas with man y applications in the industrial, pharmaceutical, and medical w orld. The largest consumer of o xygen is the m o dern stee l industry . Oxygen is use d to enric h air and increase com bustion temp eratures in blast furnaces, com bine with un w an ted carb on in the steel-making pro ces s to form carb on o x ides , and allo w gr e ater use of sc r ap metal in electric arc furnaces. These applications apply to the man ufacturing of other metals as w ell (copp er, zinc, lead, etc. . . ).
Hospitals also require large stores of liquid o xygen (k ept at cry ogenic temp eratures) for v arious medical applications. After the o xygen has b een v ap orized and divided in to smaller con tainers, it is distributed throughout the hospital to b e used in h yp erb olic c ham b ers, surgeries, and increasing patien t comfort. Hos pitals require an o xygen purit y o v er 90% for use w i th patien ts whic h is easily ac hiev ed b y high temp erature steam ele ctroly s is.
Off-site o xygen is deliv ered to steel plan ts and hospitals through bulk liquid shipmen ts. Large o xygen tank ers can transp ort 26 m 3 (˜30,000 kg) of liquid o xygen using refrigerated storage systems. Consequen tly , hospitals and steel plan ts purc hase o xygen at a p rice that accoun ts for refrigeration costs, truc king costs and transfer losse s. A rev en ue will b e generated f rom sales of excess o xygen pro duced b y the high-temp eratures steam electrolysis, but this amoun t will ultimately b e dw arfed b y the rev en ue from biof uel sales. (Information pro vided b y Univ ersal Industrial Gases Inc.)
Hydrogen Storage
Since this h ydrogen pro duction plan t is part of a larger facilit y , it is imp ortan t that the h ydrogen pl a n t b e sh ut do wn as little as p ossible in order to k eep the other parts of the facilit y w orking as w ell. In initial design, h ydrogen storage w as included for one da y’s w orth of h ydrogen as a l iquid so that there w as a bac kup source of h ydrogen for the biofuels plan t s h ould the h ydrogen plan t need to b e sh ut do wn for a short amoun t of time.
Hydrogen can b e stored as a s olid , liquid, or gas . Examples of th e se can b e seen in 64. Solid h ydrogen storage mostly consists of metal and c omplex h ydrides suc h as M g H 2 and N aAl H 4 . In metal h ydrides, h ydrogen atoms o ccup y th e in terstitial sites; in complex h ydrides, h ydrogen is c o v alen tly b ound to a me t al to f orm a complex anion whi c h is then balanced to a cation [153]. While solid-state materials ha v e m uc h p oten tial, at this p oin t they ha v e v ery lo w storage capabilities, only 2- 8% h ydrogen b y w eigh t. In addition, the thermo dynamics and kinetics of solid h ydrogen systems are unkno wn at thi s p oin t. Solid h ydrogen storage will not b e a feasible c hoice un til it has a higher storage c ap ac i t y and more is kno wn ab out the system.
T o store h ydrogen as a gas, it m ust b e compressed to 35-70 MP a, though the temp e r ature is merely ro om temp erature. Compressed gas eous h ydrogen (CGH2) is a v ery w ell-dev elop ed tec hnology whic h is widely used, esp ecially b y fuel-c ell v ehicle man uf ac tu re r s . Ho w ev er, b ecause the v ery high pressures cause strain on the w alls, the w alls m ust b e made of high-strength materials suc h as carb on comp osites and b e v ery thic k. Because of this, CGH2 mak es the most sense f or small- to mid-scale pressure v essels, whic h w ould not b e enough to store the amoun t of h ydrogen the biofuels plan t requires.
Liquid h ydrogen (LH2) stored at 0.1 MP a and -253C has a v ery high mas s densit y , whic h is adv an tageous. The main concern with liquid h ydrogen is that heat flo wing from the en vironmen t in to the storage tank will cause the h yd rogen to ev ap orate. Ho w ev er, this m eans that larger tanks ha v e implicitly b etter thermal b eha vior than smaller ones b ec au s e they ha v e a lo w er surface-to-v olume ratio. F or large v olume storage, as this plan t plans to ha v e, liquid h ydrogen is the mos t promising option.
The am ou n t of h ydrogen to b e stored w as determined in conjunction with the biofuels team and the pro cess heat team. While a larger amoun t of h ydrogen w ould pro vid e a larger safet y net for the biofuels
Courtesy of Elsevier, Inc., http://www.sciencedirect.com . Used with permission.
Figure 64: Hydrogen storage options with corresp onding energy , op erating temp erature, and wt.% [123]
team, allo wing the biofuels plan t to run longer if the h ydrogen p lan t w as sh ut do wn, there w ere significan t size and safet y concerns. The first prop osed n um b er w as 26,000 kg of h ydrogen, whic h w ould b e enough for 3 da ys, but that w ould b e 375 m 3 of liquid h ydrogen, and the pro cess heat team calculated that the pl an t w ould need to b e 150 m a w a y from ev erything else for safet y purp oses. By decreasing the amoun t of h ydrogen to just one da y’s w orth, or 9,000 kg, it will only b e 125 m 3 and need to b e 40 m a w a y . When the amoun t of h ydrogen needed p er da y increased b y more than an order of magnitude, it w as determined that h ydrogen storage w as no longer a smart decision, and w as th us remo v ed from the design. Ho w ev er, it is includ e d here for completeness in case fu ture w ork requires h ydrogen storage.
MIT OpenCourseWare http://ocw.mit.edu
22.033 / 22.33 Nuclear Systems Design Project
Fa ll 2011
For information about citing these materials or our Terms of Use, visit: http://ocw.mit.edu/terms .