Batch bomb tests indicated the relative activity of twenty-nine catalysts; five catalysts were used in the operation of a continuous reactor.

Transcription

1 Catalytic hydrogenation of PAMCO solvent-refined coal by Lee Morgan Henton A thesis submitted in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE in Chemical Engineering Montana State University Copyright by Lee Morgan Henton (1978) Abstract: Catalytic hydrogenation of Pittsburgh and Midway Coal Company's Solvent-Refined Coal was accomplished using commercial and lab-prepared catalysts. Batch bomb tests indicated the relative activity of twenty-nine catalysts; five catalysts were used in the operation of a continuous reactor. Liquid hydrogenation products were analyzed by ASTM distillation quartz tube method for sulfur content, and Macro Kjeldahl method for nitrogen content. The Ketjen HC-5 catalyst (21% WO3, 6.5% NiO on alumina) did the best desulfurization, 88%. The Harshaw HT-400 (15% MoO3 3% CoU on alumina) did the best denitrogenation, 64%. Temperature, liquid hourly space velocity, and hydrogen-to-oil flow rates were studied with the continuous reactor; pressure and catalyst amount were held constant at 1000 psig and 70 ml, respectively As temperature increases, hydrocracking and heteroatom removal increase. Optimum space velocity varied between 1 and 2, depending on the catalyst. Heteroatom removal decreased as space velocity increased A hydrogen-to-oil ratio of 10,000 SCF/bbl gave the best hydrocracking and heteroatom removal results at 445 C and liquid hourly space velocity equals 1.

2 - STATEMENT OF PERMISSION TO COPY In p re s e n tin g t h i s th e s is in p a r t i a l f u l f i l l m e n t o f the re q u ire m e n ts f o r an advanced degree a t Montana S ta te U n iv e r s ity, I agree th a t the L ib r a r y s h a ll make i t f r e e ly a v a ila b le fo r in s p e c tio n. I fu r th e r agree th a t p e rm issio n fo r e x te n s iv e copying o f t h is th e s is fo r s c h o la rly purposes may be gra n te d by my major p ro fe s s o r, o r, in h is absence, by th e D ire c to r o f L ib r a r ie s. I t i s understood th a t any copying or p u b lic a tio n o f t h is th e s is fo r f in a n c ia l g a in s h a ll not be allow ed w ith o u t my w ritte n - p e rm issio n. S ig n a tu re Date li j / ^ " 7 3

3 CATALYTIC HYDROGENATION OF PAMCO SOLVENT-REFINED COAL by LEE MORGAN HENTON A th e s is s u b m itte d in p a r t i a l f u l f i l l m e n t o f the re q u ire m e n ts fo r th e degree o f MASTER OF SCIENCE in Chemical E nginee ring Approved: Chalrjman, Gradua4^e-jp mmrttee Head, M ajor Departipefrt L be VD GraduatevDean MONTANA STATE UNIVERSITY Bozeman, Montana March, 1978

4 iii ACKNOWLEDGEMENT The au th o r wishas to acknowledge the f in a n c ia l su p p o rt given t h is p r o je c t by the Energy Research and Development A d m in is tra tio n ; N a tio n a l Science F oundation, and Chemical E n g in e e rin g Department o f Montana S ta te U n iv e r s ity. The au th o r wishes to thank Dr. L lo yd Berg fo r h is encouragement and guidance d u rin g t h i s p r o je c t. S p e c ia l th anks go to fe llo w re s e a rc h e rs Steve Kujaw a, Gary Haas, and Ken Runnion. fo r the exchange o f id e a s and in fo rm a tio n th ro u g h o u t t h i s re s e a rc h.

5 TABLE OF CONTENTS Page. VITA... ACKNOWLEDGEMENTS * LIST OF TABLES.. ' i i i i i v i i LIST OF FIGURl S c * * * * * * * * * * * * * * * * * * * * * * * v ii. I ABSTRACT x INTRODUCTION 1 RESEARCH OBJECTIVE 3 BACKGROUND AND THEORY... 4 Coal L iq u e fa c tio n Processes D iffe r e n t Types SRC I S R C Il F a cto rs in L iq u e fa c tio n Coal S tru c tu re and C om position SRC S tru c tu re and C om position D is s o lu tio n o f Coal R e action to "P rom pt In te rm e d ia te s Slower P ro g re s s iv e and R egressive R eactions D e s u lfu r iz a tio n R eaction Steps D e n itro g e n a tio n R eaction Steps Cracking. R e a ctions E ffe c ts o f V a ria b le s on C onversion and S u lfu r-n itro g e n Removal 21 S o lve n t Hydrogen E ffe c ts Tem perature E ffe c ts P ressure E ffe c ts ' Hydrogen S u lfid e and Other E ffe c ts C a ta ly s is F a cto rs in C a ta ly s t A c t iv it y, S e le c t iv it y, and A c t iv it y M aintenance # *».. 28 H

6 V Table o f C ontents c o n t'd Page MATERIALS, EQUIPMENT, AND PROCEDURES Coal C om position SRC C om position « »» # 54 S o lve n t C om position e e «# e e e 6 e e «o * e e 57 Feed to the C ontinuous Reactor P r e s u lfid in g Reactor Rocking Bomb Reactor. C ontinuous R eactor.. C ontinuous Runs A n a ly tic a l Procedures RESULTS AND DISCUSSION # # «6 6 6 * 6» « * »» 6, * * P re lim in a ry Runs... 6 # e * o e e e «e «# Feed S o lu tio n,.. #» # # * # A n a ly tic a l D e fin itio n s e o * o e «o «# o o 6 # 6 P re lim in a ry C ontinuous Run.... Bomb Runs S tu d yin g V a ria b le s Other Than C a ta ly s t Type C a ta ly s t Screening Bomb Runs Comparisons o f Some Ni-W, Ni-Mo and Co--Mo C a ta ly s ts. «M etal C h lo rid e s Fe and Fe-Cr C a ta ly s ts Other E ffe c ts. C onclusion s o f Bomb T ests ' C ontinuous Runs Summary o f R e s u lts and C ontinuous Runs H yd ro cra ckin g Heteroatom Removal CONCLUSIONS P

7 vi Table o f C ontents c o n t'd - 2 Page RECOMMENDATIONS S hort Ter m Eong lerrn * LITERATURE CITED APPENDICES Appendix A. D e s c rip tio n o f C a ta ly s ts Used Appendix B. Batch Run Data Appendix C. C ontinuous Runs: C onversion o f SRC to L iq u id 1gQ Appendix D, Average Pore Diam eter C a lc u la tio n s Appendix E L is t o f Terms e.» e e ' e» e o e e e e e e 103

8 vii LIST OF TABLES Table. Page 1. Mean A n a ly tic a l Values fo r Chemical Elements in Coal T y p ic a l H yd ro g e n o ly s is M e ta ls Coal Analyses to SRC Process SRC U ltim a te A n a ly s is T y p ic a l Compounds in a C o a l-d e rive d L iq u id S o lve n t C om position ,D e s c rip tio n, C o n c e n tra tio n, and Makeup o f M o b il's GEC F ra c tio n A n a ly s is o f SRC S o lve n t R a tio Test Runs Batch C a ta ly s ts Tested by M etal Type Batch C a ta ly s ts Ranked by D is t illa t e, C u r v e Q u a lity Best Batch D e s u lfu r iz a tio n s Best Batch D e n itro g e n a tio n s... 55

9 vlii LIST OF FIGURES Figure Page 1. SRC Process Schematic 6 2. A R e p re s e n ta tiv e o f B itu m in o u s Coal S tru c tu re Coal D is s o lu tio n and R eaction Model C onversion o f B itu m in o u s Coal w ith Rapid H eating a t 400 C Top in T e t r a lin... Bottom in N aphthalene A u to cla ve 1 and H earer» * C ontinuous R eactor P roduct 1 Compared to 7 :2 S o lu tio n Feed D i s t i l l a t i o n o f HT-400 P roduct vs. 3 :2 S o lu tio n Feed. 52 v9. D i s t i l l a t i o n o f HT-500 P roduct vs. 3 :2 S o lu tio n " Feed E ffe c t o f Tem perature on H yd ro cra ckin g by STK E ffe c t o f LHSV on H yd ro cra ckin g b y HT ] 12. E ffe c t o f H g,r a tio on H yd ro cra ckin g by STK P ercent D e s u lfu r iz a tio n vs. Tem perature P ercent D e n itro g e n a tio n vs. Tem perature P ercent D e s u lfu r iz a tio n vs. LHSV P ercent D e n itro g e n a tio n vs. LHSV ^7 17. P ercent D e s u lfu r iz a tio n vs. Hydrogen to O il R a tio P ercent D e n itro g e n a tio n vs. Hydrogen to O il R a tio. 79

10 ix L is t o f F ig u re s c o n t'd F igure '.. Page 19. P ercent D e s u lfu r iz a tio n vs. Hydrogen to O il R a tio a t D iffe r e n t L iq u id H o u rly Space V e lo c itie s.... gg 20. P ercent D a n itro g e n a tio n vs." Hydrogen to O il. R a tio at D iffe r e n t L iq u id H o u rly Space V e lo c itie s

11 X ABSTRACT C a ta ly tic h yd ro g e n a tio n o f P itts b u rg h and Midway Coal - Company's S o lv e n t-r e fin e d Coal- was accom plished using com m ercial and la b -p re p a re d c a ta ly s ts. Batch bomb te s ts in d ic a te d the r e la t iv e a c t i v i t y o f tw e n ty -n in e c a ta ly s ts ; f iv e c a ta ly s ts were used in the o p e ra tio n o f a co n tin u o u s r e a c t o r. L iq u id h y d ro g e n a tio n p ro d u c ts were analyzed by ASTM d i s t i l l a t i o n q u a rtz tube method fo r s u lfu r c o n te n t, and Macro K je ld a h l method fo r n itro g e n c o n te n t. The K e tje n HC-5 c a ta ly s t (21% VJO, 6.5% NiO on alum ina) d id the b e st d e s u lfu r iz a tio n, 88%. The Harshaw HT-400 (15% MoO, 3% CoLI on alum ina) d id th e b e st d e n itro g e n a tio n, 64%. Tem perature, l i q u i d h o u rly space v e lo c it y, and h y d ro g e n -to -o il flo w r a te s were s tu d ie d w ith the c o n tin u o u s r e a c to r ; p re ssu re and c a ta ly s t amount were h e ld c o n s ta n t a t 1000 p s ig and 70 ml-, re s p e c tiv e ly As tem p e ra tu re in c re a s e s, h y d ro c ra c k in g and heteroatom rem oval in c re a s e. Optimum space v e lo c it y v a rie d between 1 and 2, depending on the c a t a ly s t. Heteroatom rem oval decreased as space v e lo c it y in c re a s e d A h y d r o g e n - to - o il r a t i o o f 10,000 SCF/bbl gave the b e st h yd ro c ra c k in g and heteroatom rem oval r e s u lt s a t 445 C and l i q u i d h o u rly space v e lo c it y equals I.

12 INTRODUCTION As o f 1974, p e tro le u m and n a tu r a l gas accounted fo r 15% o f the t o t a l energy consumed in the U n ite d S ta te s ( I ). Since.o il i s a non- renewable re so u rce and d is c o v e ry o f new re s e rv e s are la g g in g behind consum ption, the end o f a ffo rd a b le o i l is 'a p p ro a c h in g. Since e x o tic la rg e sca le re so u rce s l i k e fu s io n are not expected to be com m ercial in the next f i f t y y e a rs, an in te r im source o f energy i s needed. Coal i s a v ia b le c h o ic e. Proven re s e rv e s o f c o a l in the U nited S ta te s h o ld 3 tim e s as much energy as known dom estic re s e rv e s o f crude o i l, n a tu r a l gas, shale o i l s, and bitum ens ( 2). The n a tu re o f co a l makes i t u n s u ita b le fo r many uses. Much tr a n s p o r ta tio n fu e l must be l i q u i d to f i t e x is tin g system s; m in e ra l ash accounts fo r up to 10% o f c o a l. Many c o a ls c o n ta in more s u lfu r and n itro g e n than com bustion fu e ls made from crude o i l. P rocessing te c h nology w i l l th e re fo re have to be developed in o rd e r to make d i s t i l l a t e fu e ls from co a l e c o n o m ic a lly fe a s ib le.. L iq u e fa c tio n o f c o a l i s not new. German l i q u i f i e d co a l p la n ts o p e rated in the. 1920's and 1930 s. The Bureau o f Mines operated a synt h e t ic fu e ls d e m o n stra tio n p la n t from 1949 t o 1953; Union Carbide operated a co a l h yd ro g e n a tio n p la n t from 1952 to But c o a l- d e riv e d liq u id s were not e c o n o m ic a lly com petative in l i g h t o f la rg e crude o i l d is c o v e rie s (1, 3 ). Coal liq u id s te ch n o lo g y i s again being developed to meet a n t i c i pated need: FMC COED p y r o ly s is o i l, H -C oal, S y n th o il, and S o lv e n t-

13 2 R efined Coal are p ro d u c ts o f c u rre n t te c h n o lo g y. These s t i l l do not g iv e la rg e y ie ld s o f h ig h q u a lit y l i q u i d fu e l, however. -. T h e re fo re, the o b je c t o f t h is re se a rch i s to,upgrade P itts b u rg h and Midway Coal Company's SRC (S o lv e n t-r e fin e d Coal) in t o a clean d i s t i l l a t e f u e l.

14 RESEARCH 0B3ECTIVE The o b je c t o f t h is re se a rch i s c a t a ly t ic c o n ve rsio n o f PAMCO's S o lv e n t-r e fin e d Coal (an upgraded co a l p ro d u c t) to cle a n d i s t i l l a t e fu e ls.. A number o f com m ercial and la b -p re p a re d c a ta ly s ts were used to h y d ro tre a t a s o lu tio n o f S o lv e n t-r e fin e d Coal (SRC) a t high tem p e ra tu re and p re s s u re. The p ro d u c t s o lu tio n should have more lo w - b o ilin g f r a c t io n s than the feed s o lu tio n ; reduced s u lfu r and n itro g e n p ro d u c t c o n te n t i s a lso d e s ire d. The re se a rch p la n in c lu d e d a u to cla ve te s t scre e n in g o f h yd ro - t r e a t in g c a ta ly s ts. P rom ising c a ta ly s ts were employed in the o p e ra tio n o f a c o n tin u o u s t r ic k le - b e d r e a c to r. E ffe c ts o f li q u i d h o u rly space v e lo c it y, h y d r o g e n - to - o il r a t i o, and tem p e ra tu re were a lso s tu d ie d in the co n tin u o u s re a c to r.

15 BACKGROUND AND THEORY Coal L iq u if a c t io n Processes D iffe r e n t Types - Four typ e s o f liq u e f a c t io n processes are being developed in the U.5.: p y r o ly s is, d ir e c t h yd ro g e n a tio n, F is c h e r- Trdpsch co a l g a s - t o - liq u id s y n th e s is, and s o lv e n t h y d ro g e n a tio n. Examples o f the p y r o ly s is type a r e : FMCs COED process and G u lf O i l 's coker d i s t i l l a t e p ro ce ss. P roducts are a ta r o i l fo r h y d ro - t r e a t in g and a char p ro d u c t ( I ). D ire c t h y d ro g e n a tio n processes in c lu d e Hydrocarbon Research I n c. 's H-COAL, P itts b u rg h Energy Research C e n te r's S y n th o il, Conoco Coal Development Company's Zinc C h lo rid e process and I n s t it u t e o f Gas T e ch n o lo g y's R ise r C racking o f c o a l. Types o f c o n ta c tin g and c a ta ly s ts vary w id e ly. Coal s lu r r y and hydrogen gas pass through an e b u lla tin g bed o f c a ta ly s t in the H-COAL p ro ce ss; co a l s lu r r y and hydrogen pass t u r b u le n t ly th ro u g h a fix e d c a ta ly s t bed fo r S y n th o il. Conoco passes re a c ta n ts th ro u g h a m olten s a lt bed; re sid e n ce tim e fo r p u lv e riz e d co a l p a r t ic le s i s le s s than 10 seconds in the R iser C racking tu b u la r r e a c t o r. F isch e r-t ro p a ch s y n th e s is in v o lv e s re a c tin g ste'.am and co a l at h ig h te m p e ra tu re to make CO and H^ gas. The gas i s f i r s t p u r if ie d o f s u lf u r and n itr o g e n compounds, then i t re a c ts over a c a ta ly s t bed to make s h o rt ch a in a lc o h o ls and hydrocarbons ( 'I). S o lu tio n h y d ro g e n a tio n processes in c lu d e PAMCO1s S o lv e n t-

16 5 R e fin e d Goal ( SRC), C o n s o l's Cresap p r o je c t, E xxon's donor s o lv e n t process, Morgantown Energy Research C e n te r's "T w o-s tep," and the "Clean Fuels West" process sponsored by Southern C a lif o r n ia Edison, M o b il O il, EPRI, and Conoco C oal. The method o f hydrogen a d d itio n to the re c y c le s o lv e n t and/or co a l s o lu tio n v a rie s a cco rd in g to the process ( 4 ). SRC I - The SRCI process i s th e one o f in t e r e s t in t h is r e p o r t. The s o lv e n t- r e fin e d co a l major p ro d u c t i s a s o lid a t room te m p e ra tu re, w ith s u lfu r and ash c o n te n t s u b s ta n tia lly reduced from th e o r ig in a l c o a l. However, the o r ig in a l n itro g e n o f the co a l rem ains in the p ro d u c t, and the h y d ro g e n -to -c o a l r a t i o o f the p ro d u c t i s s l i g h t l y le s s than th a t o f the o r ig in a l c o a l. The upgrading o f the s o lid s o lv e n t- r e fin e d co a l p ro d u c t from SCR I i s th e o b je c t o f t h i s th e s is re s e a rc h. F igure 1 shows a s im p lif ie d process schem atic o f P itts b u rg and Midway Coal Company's S o lv e n t-refin e d Coal I process ( 5 ). The process s t a r t s w ith p u lv e riz e d c o a l being s lu r r ie d w ith re c y c le s o lv e n t in a mix ta n k. Hydrogen and s lu r r y are mixed,,heate.d, and pumped to th e h ig h p re ssu re d is s o lv e r. The p ro d u c t from the d is s o lv e r i s fla s h e d to separate th e raw gases from the c o a l s lu r r y. The raw gas i s tre a te d to remove im p u r itie s and hydrogen i s re c y c le d. The c o a l s lu r r y le a v in g the s e p a ra to r is at h ig h te m p e ra tu re, as most o f th e c o a l m a tte r i s s t i l l l i q u id. F i l t r a t i o n separates the u n d is s o lv e d s o lid s (ash and co a l re s id u e ) from the h o t l i q u i d. The &

17 HYDROGEN M Y M O C A R 3 0 N OAS SLURRY MIX TANK FIRED PREHEATER S00 F DIS SOLVER 50 F 100Mm COAL SEP- ARATOR SULFUR HYDROGEN SOLVENT RECYCLE SOLIDS SLURRY COAL' GASIFIER AND SHIFT CONVERTER,FILTEF,* STEAM SOLVENT RECOV ERY U N IT COAL SRC I Process Schematic LIQUID FUEL TO SOLIDIFICATION

18 7 s o lid c o a l re s id u e i s re a c te d w ith steam and a ir to y ie ld make-up hydrogen fo r th e pro ce ss. The hot l i q u i d from the f i l t e r e n te rs a s o lv e n t re c o v e ry u n it. The low er b o ilin g m a te ria l i s separated from the re c y c le s o lv e n t. T h is lower b o ilin g m a te ria l y ie ld s a sm a ll amount o f l i q u i d p ro d u c t and a la rg e amount o f s o lid m a te ria l at room te m p e ra tu re. The s o lid i s " s o lv e n t- r e fin e d c o a l" (5, 6). SRC I I - More liq u id s and gases, w ith le s s s o lid p ro d u c t, can be made by a m o d ifie d s o lv e n t- r e fin e d c o a l process c a lle d SRC I I. The two main changes from th e SRC I process are t h a t : a) ash i s not r e moved, but re ta in e d in th e f i n a l s o lid p ro d u c t; b) i t does not use a l i g h t " r e c y c le s o lv e n t." In s te a d, some o f the h o t p ro d u c t stream is d iv e rte d for. use as th e process s o lv e n t ( 6). I f the SRC I I l i q u i d y ie ld s are high enough, upgrading the SRC 11 liq u id s may be more fe a s ib le th a n upgrading th e a s h -fre e s o lid from SRC I. F a ctors in L iq u e fa c tio n A good many fa c to r s in flu e n c e c o a l- t o - liq u id c o n v e rs io n. The rank o f c o a l in flu e n c e s what p ro d u c ts are p o s s ib le. Coals vary in s u lf u r, n itro g e n, oxygen, ash, and tra c e m etal c o n te n t. Some o f those co n ta m in a n ts may be c a t a ly t ic, but some i n h i b i t c a ta ly s is by re a c tin g w ith th e c a ta ly s t or p lu g g in g heterogeneous c a ta ly s t po re s. A m yriad o f o th e r fa c to r s in flu e n c e c o a l co n v e rs io n and typ e s

19 8 o f p ro d u c ts. Length o f p ro c e s s in g tim e and typ e o f process appear to be ve ry im p o rta n t. The in flu e n c e o f s o lv e n t chem ical and p h y s ic a l n a tu re, s o lv e n t r a t i o,- amount and type o f re d u cin g agents, p re ssu re -a n d tem p e ra tu re o f s o lu tio n and r e a c tio n, c a ta ly s t n a tu re, and re a c to r c o n fig u r a tio n v a rie s, depending on the tim e and type o f p ro c e s s in g. Coal S tru c tu re and C om position - F igure 2 shows a re p re s e n ta t iv e s tr u c tu r e fo r b itu m in o u s c o a l ( 7 ). Coal s tr u c tu r e " c o n s is ts o f a ro m a tic r in g c lu s te r s o f no la r g e r th a n 6 or 7 r in g s t ie d to g e th e r by b rid g e lin k a g e " ( 8). C o a l's H/C r a t i o i s low er than most f u e l m a te ria ls d e riv e d from p e tro le u m. The H/C r a t i o o f I l l i n o i s No. 6 i s 0.8 6, w h ile r e s id u a l h e a tin g o i l No. 6 ' s r a t i o i s 1.55 and a r e f in in g naphtha f r a c t io n i s 2.20 ( 9 ). C o a l's low H/C r a t i o i s another in d ic a tio n o f th e a r o m a tic ity o f co a l compared to p e tro le u m. Coal c o n ta in s s ig n if ic a n t amounts o f oxygen, s u lfu r and n itro g e n. Compounds c o n ta in in g these im p a rt u n d e s ira b le q u a lit ie s to fu e l p ro d u c ts, p o iso n c a ta ly s ts, or become com bustion p o llu ta n ts. Removal o f oxygen i s le a s t d i f f i c u l t because oxygen atoms are u s u a lly in phenol s tr u c tu r e s and not h e te r o c y c lic. In uses o f upgraded co a l when th e presence o f oxygen i s a c c e p t- ' able (lo w grade f u e ls ), " i t i s d e s ira b le to remove o n ly s u lfu r and n itro g e n to m inim ize hydrogen consum ption" ( 10), S u lfu r i s le s s d i f f i c u l t than n itro g e n to remove (1 0 ). O rganic and in o rg a n ic s u lfu r are found in c o a l. Most in o rg a n ic s u lfu r is ir o n s u lf id e, FeS^, and a sm a ll amount i s p re se n t as s u lfa te io n.

20 H H I H I 0 S 1 I H H H z^. / H H H H I C =O,H F ig. 2. A R e p re s e n ta tiv e o f B itu m in o u s Coal S tru c tu re (?)

21 10 O rganic s u lfu r i s m o stly h e t e r o c y c lic ; t o t a l s u lfu r c o n c e n tra tio n in c o a ls ranges from s e v e ra l p e rc e n t to 10~15/. N itro g e n, l i k e s u l f u r, i s a c a ta ly s t poison and p o llu t a n t ; i t i s p re s e n t in c o n c e n tra tio n s o f 1-2 w e ig h t p e rc e n t in most c o a ls ( 11). An a n a ly s is o f 100 c o a ls "re v e a le d the presence o f w e ll over h a lf o f th e elem ents o f th e p e r io d ic ta b le " ( 11), Table I shows a ve r age c o a l m etal c o n c e n tra tio n s as compared to va lu e s in PAMCO1s SRC I from Kentucky No. 9 c o a l. Trace m etals may enhance or i n h i b i t liq u e fa c tio n c a ta ly s ts and re a c tio n s. " Sm all amounts o f TiO^ on a s i l i c a or alum ina su p p o rt can cause a marked in c re a s e in s u rfa c e a c id it y " (10) Aged c a ta ly s t p e lle t s from th e SYNTHOIL process were contam inate d w ith fe rro u s s u lf id e, aluminum, and s ilic o n ; tra c e m etals were v a r ia b ly. d is t r ib u t e d th ro u g h th e c a ta ly s t ( 12). U ltim a te analyses o f s o lv e n t- r e fin e d co a l and i t s source c o a l, Kentucky No. 9 are g iv e n in th e M a te ria ls s e c tio n ta b le s. SRC S tru c tu re and C om position - A lthough no compound breakdown o f s o lv e n t- r e fin e d c o a l has been p u b lis h e d in th e l i t e r a t u r e, charact e r iz a t io n by s o l u b i l i t y f r a c t io n is p ra c tic e d by s e v e ra l people (see M a te ria ls s e c tio n ). The s u lfu r and ash le v e ls o f SRC are much le s s th a n i t s p a re n t c o a l, but th e n itro g e n le v e l is not reduced. D e s u lfu r iz a tio n is about 80% and de-ashing about 97L/a.

22 11 Table I M ean analytical values for chemical elements in coal ( 1 1) V a lu e In o s h -fr e e, s o lv e n t- re fin e d C onsti M ean, Std. K entu cky tu e n t p p m d e v. N o. 9 c o a la As B Be Br Cd Co Cr Cu F ;.. / / c Ga Ge Hg Mn Mo Ni :.... ^.. P Pb Sb Se Sn V Zn Zr < - Al Ca Cl Fe K , * Mg Na ; Si Ti moisture 9.05% 5.05% 0.00% ash 11.44% 2.89% 0.051% Source: Illinois State Geological Survey a Data obtained by Manahan's research group, primarily by neutron activation analysis. ' '.. -N.".--- V. v;...-wj'* -- f

23 12 The H/C atom r a t i o o f SRC i s low er than th e p a re n t c o a l, 0/769 vs. 0,8 3 7, even though hydrogen i s consumed in th e p ro ce ss. Appare n tly hydrogen from the c o a l goes to d e s u lfu r iz a tio n (making H^S) and making hydrocarbon gases. A g ra d ie n t e lu tio n chrom atograp hic a n a ly s is o f SRC, SYNTHOIL, and H-COAL p ro d u c ts was perform ed by a M o b il O il team (1 3 ), The a n a ly s is c o n s is te d o f s e p a ra tin g th e liq u id s in to t h ir t e e n d if f e r e n t s o l u b i l i t y f r a c t io n s. S o lv e n t-re fin e d co a l has much more asphaltene c o n te n t than do SYNTHOIL, H-COAL, or p e tro le u m residuum. E ig h ty -o n e w e ig h t p e rce n t SRC i s a sp h a lte n e, w h ile 12 w eight p e rc e n t L ig h t A rabian Residuum is asphalte ne (1 3 ). S everal g e n e ra l tre n d s were le a rn e d from c h a r a c te r iz in g each o f the t h ir t e e n f r a c t io n s them selves: a) The p o la r, heavy fr a c tio n s "are more a ro m a tic than the l i g h t f r a c t io n s, b) The m o le cu la r w e ig h ts (and th e re fo re s iz e s ) o f c o a l l i q u i d components are s i g n if ic a n t ly s m a lle r than p e tro le u m re s id u a components o f the same f r a c t io n. (c ) N itro g e n and oxygen c o n te n t o f SRC is h ig h e s t in th e asphaltene f r a c t io n s, but s u lfu r c o n te n t i s n e a rly c o n s ta n t in a l l f r a c t io n s (1 3 ).

24 13 D is s o lu tio n o f Coal - A ccording to W e lle r 's t r a d i t i o n a l model o f s o lv e n t co a l - I iq u e f a c tio n, o i l s appear la s t in a sequence o f re a c tio n s : co a l > asphaltene- ' ^ o i l ' (4 ). I t i s known th a t " dep o lym e riza t io n o f co a l ta ke s p la ce at h ig h te m p e ra tu re s (up to 420 D and 80 atmosphere i n i t i a l l y in an a u to c la v e ) in the presence o f s o lv e n ts " (14- ) f The slo w e st re a c tio n i s from a s p h a lte n e > o i l, w ith the c o a l asphalte ne being 25 tim e s as fa s t ( 8). However, re c e n t s tu d ie s show th a t co a l can sometimes be conv e rte d to o i l very q u ic k ly, w ith o u t going through- the ve ry slow a s p h a ltene to o i l s te p. Rapid h e a tin g and s h o rt re sid e n ce tim e experim ents have le d to a new model fo r ; c o a l- t o - o il c o n v e rs io n. F igure 3 d is p la y s the new model (4 ). When co a l i s heated in a s o lv e n t to 4D0 C, the e a s ily -b ro k e n bonds between r in g c lu s te r s are r a p id ly broken. Four ty p e s o f prompt in te rm e d ia te s appear s a) o i l s, s o lu b le in p e n ta n e ; b) a sp h a lte n e s, s o lu b le in benzene but not pentane; c) a s p h a lto ls, s o lu b le in p y rid in e but not benzene; and d) re s id u e, not in c lu d in g ash. The f i r s t th re e typ e s o f sp e cie s then proceed w ith r e l a t i v e l y slow p ro g re s s iv e (to w a rd o i l ) and re g re s s iv e ( to sem i-coke) re a c tio n s. Semi-coke a n d 're s id u e do n o t re a c t p ro g re s s iv e ly (4 ). R eaction to "Prom pt" In te rm e d ia te s - Time i s th e im p o rta n t fa c to r in the f i r s t, c o a l r e a c tio n, break-up in to the fo u r species o f "prom pt" in te rm e d ia te s. Only, w ith ra p id h e a tin g and quenching are the 0

25 14 PRODUCT OIL COAL SEMI - COKE RESIDUE A working model of reaction paths in donor solvent liquefaction of a bituminous vitrinite. The curved lines at the left represent rapid cleavage of scissile bonds to yield prompt liquefaction intermediates. These comprise a large number of distinct species forming a hierarchy of increasing polarity toward the bottom of the figure, and also a prompt residue. Relatively slow progressive reactions, indicated by solid lines, convert the prompt intermediates to oil. The progressive reactions compete with relatively slow regressive reactions, indicated by dot-dash lines, that lead to semi-coke. Fig. 3 Coal D is s o lu tio n and R eaction Model (4 ).

26 15 "prom pt" in te rm e d ia te s n o tic e d. The r a p id re a c tio n s are th e rm a l; th e y do not depend on: a) the presence or absence o f-h y d ro g e n ; b) th e presence or absence o f a s o lid c a t a ly s t ; c) hydrogen donor p r o p e r tie s o f the s o lv e n t (4 ). M o b il O il data on s h o rt r e a c tio n tim e y ie ld e d s e v e ra l a d d itio n a l fa c ts, "The i n i t i a l m o le cu la r w eight d is t r ib u t io n i s bim o d a l, w ith some fragm ents having w e ig h ts g re a te r than 2050, and the rem aining fragm ents are between 300 and 900" ( 4 ). The m o le cu la r w eight, o f most c o a ls va ry from 2000 to 12,000; ta r s and re s id u a l o i l s have m o lecular w e ig h ts ra n g in g up to the m illio n s (1 5 ). F o rty p e rc e n t o f the oxygen and s u lfu r in c o a l i s removed q u ic k ly, using hydrogen o n ly from th e c o a l. " E a r ly d is s o lu tio n p ro d u c ts are more a rom atic than the o r ig in a l c o a l"; since th e c o a l's hydrogen i s consumed in heteroatom, rem oval, the in c re a s e in a r o m a tic ity makes sense ( 4 ). "Prom pt liq u e f a c t io n in te rm e d ia te s c o n ta in r e l a t i v e l y few p o ly c y c lic a ro m a tic s tru c tu re s " (4 ). Slower P ro g re s s iv e and R egressive R eactions - S everal g e n e r a liz a tio n s can be made about the p ro d u c ts from slow re a c tio n s o f the prompt in te rm e d ia te s. "At lo n g e r tim e s, the a ro m a tic c o n te n t ( o f the p ro d u c t) in c re a s e s, as does a lso th e number o f polycondensed a ro m a tic r i n g s " ( 4 ). H igher te m p e ra tu re s also in c re a s e those two v a lu e s. 0

27 16 The H/C r a t i o decreases as r e a c tio n tim e and s e v e r ity in c re a s e. Though much hydrogen can be consumed, i t goes m ostly fo r the p ro d u c tio n o f gas and re -h y d ro g e n a tin g h y d ro a ro m a tic s tru c tu re s ( re c y c le s o lv e n t). L i t t l e hydrogen goes to heteroatom rem oval ( 4 ). ' A s p h a lto ls are the main source o f sem i-coke. "Coke fo rm a tio n r e s u lt s i f th e rm a l c ra c k in g 'g e ts ahead' o f h yd ro g e n a tio n o f c o a ly ( s ic ) m a tte r by hydrogen tr a n s fe r from th e donor s o lv e n t" ( 4 ). The hydrogen donor c a p a b ilit y o f the s o lv e n t is most im p o rta n t d u rin g the slow re a c tio n s. R egressive re a c tio n s predom inate i f a h yd ro gen donor s o lv e n t is not used.. A donor s o lv e n t a p p a re n tly suppresses th e re g re s s iv e re a c tio n s le a d in g to coke. F igure 4 i l l u s t r a t e s the p o in t ( 4). About 90/o o f a b itu m in o u s co a l was con ve rte d to p y r id in e - s o lu b le s when heated r a p id ly in e ith e r t e t r a l i n or n a p hth alene. That 5 m inute p e rio d i s th e regim e o f ra p id r e a c tio n to "p ro m p t" in te rm e d ia te s. A fte r th e f i r s t 5 m in u te s, the slow er p ro g re s s in g and re g re s s iv e re a c tio n s compete. In t h i s second tim e regim e, p y r id in e - s o lu b le s m arkedly decreased in na p h th a le n e, w h ile re m a in in g s ta b le in the h yd ro gen donor s o lv e n t t e t r a l i n. P ro g re s s iv e re a c tio n s to l i g h t e r p ro d u c ts c o n tin u e d in t e t r a l i n, but coke.was formed in naphthalene (4 ). D e s u lfu r iz a tio n R eaction Steps - hydro d e s u lfu r iz a tio n (HDS) o ccurs at le s s severe c o n d itio n s th a n h y d ro d e n itro g e n a tio n (HDN). Seve r a l in v e s tig a to r s have te s te d model compounds to determ ine step

28 17 Conversion, % of Coal Conversion, % of Coal (Ivloisfure-and-Ash-Free) (Moisture-and-Ash-Free) IO O PROMPT RESIDUE, ASPHALTENES ; ZL J I. i I s a,,..,., 0IL.,-ZLv' " X GAS Time, Minutes Time, Minutes F ig. 4. C onversion o f B itu m in o u s Coal w ith Rapid H e ating a t 400 C. Top - in T e t r a lin Bottom - in N aphthalene»

29 sequences in s u lfu r rem oval from a ro m a tic s ; th e y conclude a ro m a tic r in g s a tu r a tio n i s n o t necessary fo r C-S bond breaking- (10, 16). 18 Some th e rm a l rem oval o f s u lfu r o ccu rs. -D e s u lfu riz a tio n was observed f o r co a l o i l s passed over in e r t su p p o rt beds from C, though no n itro g e n rem oval o ccu rre d ('17,18). in th re e ste p s: D e n itro q e n a tio n R eaction S teps- - A rom atic n itro g e n i s removed 1) E q u ilib riu m h y d ro g e n a tio n o f th e.-n itro g e n -c o n ta in in g - - r in g to..hydrdarom atic s tr u c tu r e. 2) C racking o f r in g, w ith n itro g e n becoming an -NH. 3) E lim in a tio n o f amine group as ammonia, NH^. The second s te p, b re a kin g the h y d ro a ro m a tic r in g, i s the r a t.e -. d e te rm in in g step (1 0,1 9 ). D i f f i c u l t y o f n itro g e n rem oval in c re a s e s as th e number o f r in g s in a compound in c re a s e s. U n iv e r s ity o f Delaware found d e n itro g e n a tio n o f 3 - r in g compounds was slow compared to q u in o lin e, though hydrogenat io n -of the o u te r r in g s was s t i l l r a p id (1 9 ). Oklahoma S tate U n iv e r s ity re s e a rc h e rs lik e w is e found " r e a l" liq u e f ie d c o a ls much harder to t r e a t than a model raw anthracene o i l (1 8 ). C racking R e a ctio n s - The g o a l o f c o a l h y d ro c ra c k in g i s to s e le c t iv e ly cra ck th e co a l s tr u c tu r e w ith a minimum consumptiom o f hydrogen. Hydrogen g e n e ra tio n i s an expensive p a rt o f c u rre n t co a l l i q u i d te c h n o lo g y. "The fu n c tio n o f hydrogen i s ' t o 'h e a l' those carbon

30 19 bonds in v o lv e d in a C-C bond s c is s io n and a lso to form and H^S fo r heteroatom rem oval" (1 0 ). P o ly n u c le a r compounds are cracked by f i r s t h yd ro g e n a tin g one o f th e r in g s, then c ra c k in g i t. "The h y d ro a ro m a tic s tr u c tu r e most fa v o r able fo r h y d ro c ra c k in g r a th e r than h y d ro g e n a tio n i s th a t in which the nonarom atic r in g is f u l l y s a tu ra te d and in which hydrogen ation, causes lo s s o f an a ro m a tic r in g. Least fa v o re d i s a nonarom atic r in g which has a double bond" (20, 10). By therm odynam ics, cracked p ro d u c ts are more fa v o re d than h yd ro g e n a tio n p ro d u c ts a t te m p e ra tu re s above 350 C. But th e r a te s fo r h yd ro g e n a tio n p ro d u c ts are g re a te r. A c tiv a tio n e n e rg ie s o f th e h yd ro g e n a tio n step are ve ry low, but a c t iv a t io n e n e rg ie s fo r c ra c k in g steps are between 20 and 35 k c l/m o le (1 9,2 0 ). Coking o ccurs when cracked p ro d u c ts re a c t i r r e v e r s ib ly to high m o le c u la r w e ig h t, u n re a c tiv e s o lid compounds. Coking i s u n d e s ira b le since i t reduces l i q u i d p ro d u c t y ie ld and p lu g s or d e a c tiv a te s c a t a ly s t. H yd ro g e n -to -ca rb o n r a t io s o f cokes are low er than o f the coke p re c u rs o rs. " P y r o ly s is coking re a c tio n s are not sim ple p o ly m e riz a tio n s, as hydrogen i s e lim in a te d " (1 0 ). A lso, a c tiv a tio n energy o f co nden sation re a c tio n s i s u s u a lly la r g e r than fo r c ra c k in g re a c tio n s. Two p o s s ib le approaches may a vo id c o k in g : using low te m p e ra tu re s to m inim ize th e rm a l fre e r a d ic a ls ; and quenching th e r a d ic a ls formed by hydrogen donor or s u p e r c r it ic a l c a t a ly s is (1 0 ).

31 20. Other s tru c tu r e and c o m p o sitio n fa c to r s in flu e n c e c o k in g. U n s u b s titu te d a ro m a tics coke le s s than s u b s titu te d ones, p o s s ib ly because "Therm al f r e e - r a d ic a ls are in v o lv e d in the c o k in g -c o n d e n s a tio n re a c tio n " (1 0 ), S u b s titu tio n o f a n itro g e n atom in a p o ly n u c le a r a rom atic in c re a s e d the coking ra te by 10 tim e s in th e rm a l p y r o ly s is. Also p o s itio n o f the n itro g e n in the r in g a ffe c ts the r a te. Gas and l i q u i d phase re a c tio n s d i f f e r also (1 0 ). Anthracene' p y ro ly z e s more e a s ily in l i q u i d phase than v a p o r. Phenyl r a d ic a ls act d i f f e r e n t l y to o. In s o lu tio n, phenyl r a d ic a ls.a d d to a ro m a tic s, le a d in g to c o k in g ; in gas th e y ca p tu re a hydrogen atom from a s u b s titu te d m ethyl g ro u p..

32 21 E ffe c ts o f V a ria b le s on C onversion and S u lfu r-n itro g e n Removal S o lve n t - "The use o f condensed compounds having a rom atic and h yd ro a ro m a tic or h yd ro a ro m a tic and phenol r in g s in c re a s e s the liq u e fa c tio n o f c o a l" (8 ). The P o tt-b ro ch e s o lv e n t e x tr a c tio n o f co a l process was o p erated in Germany d u rin g the 19301s and 's; the s o lv ent was 20% c re s o l - 80% t e t r a l i n, at a 2:1 s o lv e n t- to - c o a l r a t i o, The s o lu tio n th e o ry was th a t co a l depolym erizes under the in flu e n c e o f c re s o l and h e a t; th e lo o s e ly bound hydrogens from th e t e t r a l i n ( 1,2,3,4 te tra h y d ro n a p h th a le n e ) are e a s ily given up to s t a b iliz e the polym er r a d ic a ls. Most c u rre n t co a l s o lv e n t- r e fin in g and li q u i d phase h y d ro g e n a tio n processes use m iddle fr a c t io n s from the process as a re c y c le s o lv e n t. R ecycle s o lv e n t c o m p o sitio n i s both s im ila r to coal fragm ents and cheaper than t e t r a l i n. S o lve n t n a tu re a ffe c ts l i q u i d co a l co n ve rsio n and p ro d u c t d is t r i b u t i o n. Though co a l r a p id ly breaks in to fragm ent m olecules re g a rd le s s o f the n a tu re o f th e s o lv e n t, subsequent slow er s t a b iliz a t io n re a c tio n s depend on s o lv e n t n a tu re (4 ). Some fa c to r s c o r r e la te w ith co a l d is s o lu tio n a c t i v i t y : - r e la t iv e ease o f r e a c tio n w ith a t y p ic a l fre e r a d i c a l, CC1^(20) - n o n -p o la r s o l u b i l i t y param eter o f th e s o lv e n t (21) - p e n e tr a tio n o f co a l fragm ents by s o lv e n t (2 0 ). C onversion, p ro d u c t d is t r ib u t io n, and d if f u s io n re s is ta n c e s vary w ith s o lv e n t- to - c o a l r a t i o. U n iv e r s ity o f N orth Dakota re s e a rc h e rs found hydro tre a tm e n t o f so Iv e n t - r e f in e d lig n it e, gave highest, conversions'

33 in a 1:1 r a t i o w ith t e t r a l i n. 22 C onversion decreased fo r an in c re a s e or decrease from th a t r a t i o, By NMR a n a ly s is, the a r o m a t ic / a lip h a t ic p ro to n r a t i o o f p ro d u c ts in cre a se d as s o lv e n t r a t i o decreased; p ro d u ct in which no s o lv e n t was used had th e h ig h e s t a r o m a tic /a lip h a tic r a t io (22).! V is c o s ity o f a SRC-solvent s o lu tio n i s very dependent on the SRC c o n c e n tra tio n in the / ra n g e, K in e m a tic v is c o s is ty v a rie s from 4 to 20 to 150 c e n tis to k e s fo r c o n c e n tra tio n s o f 30, 40, and 5 0 /, r e s p e c tiv e ly (2 3 ). - The use o f s u p e r c r it ic a l s o lv e n t media fo r c o a l d is s o lu tio n and r e a c tio n is an a t t r a c t iv e a lte r n a te to t r a d i t i o n a l p o ly a ro m a tic s o lv e n ts. S u p e r c r it ic a l f l u i d s are c h a ra c te riz e d by h ig h s o lv a tin g.p o w e r, high heat c a p a c itie s, h ig h e r d i f f u s i v i t i e s than liq u id s, and h ig h heat tr a n s fe r p ro p e r tie s (1 0 ). S ta n fo rd Research I n s t it u t e (20) conducted s u p e r c r it ic a l co a l d is s o lu tio n using a v a r ie ty o f common liq u id s. Thermal and chem ical changes o f the co a l were observed, not a d is s o lu tio n o f unchanged c o a l. That agrees w ith' the model o f co a l d is s o lu tio n S quires proposed (4 ). N e a rly a l l the s u p e r c r it ic a l f lu i d s gave a th e rm a l e f f e c t - - some o f the co a l was changed from p y r id in e - in s o lu b le m a te ria l to p y r i d in e -s o lu b le. Secondary a lc o h o ls gave a chem ical e f f e c t over and above th e th e rm a l e f f e c t, s im ila r to t e t r a l i n s i g n if ic a n t ly more co a l p ro d u c t was changed to p y rid in e -s o lu b le ' m a te r ia l. Is o p ro p y l a lc o h o l and t e t r a l i n each co n ve rte d about 5 0 / o f the co a l to p y r id in e - 0

34 23 s o lu b le s in r e a c tio n a t 335 C, 90 m in u te s, and 1000 p s ig o f e ith e r N2 9r H2., The best co n ve rsio n was o b ta in e d when a c a t a ly t ic amount o f potassium is o p ro p o x id e was added to is o p ro p y l a lc o h o l. The co a l p ro d u c t was 97% s o lu b le in p y r id in e ; s u lfu r c o n te n t was reduced from 2.2% to 0.4% (2 0 ). N e ith e r m ethanol nor t- b u ta n o l e x h ib ite d the same e ffe c ts as is o p ro p y l a lc o h o l or secondary b u ta n o l. I t is p o s tu la te d th a t "some chem ical a c tio n by th e s o lv e n t, such as h y d rid e d o n a tio n, i s in c re a s in g th e d is s o lu tio n c a p a c ity " (2 0 ). Im p o rta n t to note i s th a t experim ents showed the th e rm a l change e f f e c t was not dependent on a tta in in g the s u p e r c r it ic a l s t a t e. A ls o, though p y r id in e s o l u b i l i t y o f the co a l i s a ffe c te d, "no major f r a c t io n o f th e co a l i s e x tra c te d by s u p e r c r it ic a l is o p ro p a n o l, benzene, or methanol under (th e r e a c tio n ) c o n d itio n s " (2 0 ). Hydrogen E ffe c ts - Although th e presence o f Hg i s not needed fo r th e ra p id th e rm a l f i r s t - s t e p re a c tio n s in S quires model ( 4 ), hydrogen i s u s e fu l in th e second-step slow er re a c tio n s. S o lv e n t-re fin e d co a l processes f in d " s o lv a tio n ( o f c o a l) in c re a s e s under hydrogen p re s s ure due to p a r t i a l h y d ro g e n a tio n " ( 8 ). PAMCO1s s o lv e n t- r e fin e d co a l b e n ch -sca le work found th a t " s o lv e n t e f f ic ie n c y d e c lin e d as i t was r e used in the absence o f Hg... (th e d e c lin e ) was a t t r ib u t e d to a tr a n s fe r o f Hg from the s o lv e n t to the co a l d u rin g the s o lu tio n process" (8 ). S everal o th e r s o lv e n t processes hydrogenate the s o lv e n t s e p a ra te ly from 0

35 I 24 the s o lu tio n pro ce ss, but " s a t is f a c t o r y r e s u lt s are o b ta in e d by adding Hg d ir e c t ly to the d is s o lv e r " (8 ). In h y d ro tre a tin g SRC f i l t r a t e under severe c o n d itio n s, derosset found hydrogen consum ption was S C f/b b l (2 3 ). He a lso concluded from h is data th a t h y d ro g e n a tio n was e q u ilib r iu m lim it e d, but c ra c k in g was k i n e t i c a l l y lim it e d (2 4 ), During C o n so lid a te d C o a l's work w ith co a l liq u e f a c t io n in m olten s a lts, r a te co n s ta n ts fo r c o n v e r tib le f r a c t io n s o f co a l were independent o f p re ssu re (2 6 ).. A lso, v a r ia t io n in H flo w ra te s may have l i t t l e e f f e c t. Rate changes from 8 to 12 MSCF/bbl d u rin g h y d ro g e n a tio n o f CDED t a r over a fix e d bed had no s ig n if ic a n t e f f e c t (2 7 ). Presence o f gaseous hydrogen enhances d e s u lfu r iz a tio n and de~ n itr o g e n a tio n o f co a l liq u id s (10, 18). During d e n itro g e n a tio n o f q u in o lin e., a model'compound fo r n itro g e n re m o v a l'fro m -c o a l s tu d ie s, h yd ro g e n a tio n steps e x h ib ite d a p o s itiv e o rd e r w ith re s p e c t to Hg p re s s u re ; c ra c k in g ste p s' show.a n e g a tiv e or zero o rd e r fo r Hg p re ssu re (1 6 ). Hydrogen gas generate d in s it u i s more a c tiv e fo r some, co a l form s than o th e rs. ERDA1s COSTEAM process makes hydrogen d u rin g liq u e fa c tio n by re a c tio n o f carbon monoxide and w a te r; th e COSTEAM process works b e tte r w ith l i g n i t e and low rank, m a te ria ls than w ith b itu m in o u s c o a l. Tem perature E ffe c ts - Tem perature a ffe c ts gas p ro d u c tio n, c o k in g, heteroatom rem oval, and c ra c k in g in co a l liq u e f a c t io n

36 25 processes. In PAMCD1s SRC p ro ce ss, "degree o f s o lu tio n i s p r im a r ily dependent on te m p e ra tu re and l i q u i d re s id e n c e tim e " ( 8 ). In a d d itio n, coke fo rm a tio n was observed above 440 C. G e n e ra lly, an in c re a s e in tem p e ra tu re a t a g ive n p re ssu re w i l l in c re a s e th e gas f r a c t io n o f t o t a l c o n ve rsio n p ro d u c ts (-28). In l i g n i t e liq u e f a c t io n, t o t a l c o n ve rsio n in c re a s e s a p p ro xim a te ly l i n e a r ly w ith te m perature from 575O-450 C,' but th e gas f r a c t io n in c re a s e s much more than the l i q u i d (2 2 ). At a g ive n p re s s u re, gas fo rm a tio n was 50/6 le s s. a t 4D0 C th a n a t 450 C fo r h yd ro g e n a tio n o f s o lv e n t- r e fin e d l i g n i t e. More m id d le o i l s and heavy ends were formed a t 4D0 C: "Heavy liq u id s are fa v o re d at the expense o f the gases" (2 9 ). H iteshue (28) notes O l iq u id c o n v e rs io n i s u s u a lly not in cre a se d above 450 C. N itro g e n rem oval from co a l liq u id s u s u a lly re q u ire s h ig h e r te m p e ra tu re s than s u lfu r rem oval. D e n itro g e n a tio n o f the model compound q u in o lin e in cre a se d.more from 342 to 363 C than from C (1 9 ). Both s u lfu r and n itro g e n rem oval u s u a lly in c re a s e w ith te m p e ra tu re o f re a c tio n. Pressure E ffe c ts - An in c re a s e in pre ssu re a t a g ive n tem peratu re w i l l g e n e ra lly not change the gas c o n v e rs io n, but the l i q u i d conv e rs io n u s u a lly in c re a s e s (2 8,2 9 ). However, batch h yd ro g e n a tio n o f s o lv e n t- r e fin e d l i g n i t e on Harshaw HT-400 c o b a lt m olybdate was found to y ie ld more l i q u i d p ro d u c t a t 1750 p s i g 'i n i t i a l c o ld p re ssu re than at 2500 p s ig or 1000 p s ig (2 9 ).

37 26 The h ig h e s t p re ssu re s a t a g ive n tem perature were found to y ie ld th e lo w e st amounts o f vacuum bottom s fo r h y d ro g e n a tio n o f s o lv e n t- r e fin e d l i g n i t e (2 9 ). Heteroatom rem oval in c re a s e d fo r co a l model compounds or m ix tu re s as p re ssu re in c re a s e d. S u lfu r and n itro g e n rem oval in c re a s e d w ith p re ssu re fo r raw anthracene o i l at 700 F and low space v e lo c it ie s over c a ta ly s t (3 0 ). N itro g e n rem oval from q u in o lin e in c re a s e d w ith p re ssu re up to 1500 p s ig, above which rem oval le v e le d o f f (3 1 ). Hydrogen S u lfid e and Other E ffe c ts - The presence o f sm all amounts o f H^S prom otes d a n itro g e n a tio n o f c o a l liq u id s ; in a d d itio n i t does not i n h i b i t h y d r o d e s u lfu r iz a tio n u n t i l i t reaches 2-3% mole c o n c e n tra tio n. H^S enhances"h y d ro d e n itro g e n a tio n (HDN) over a h e te ro geneous m etal c a ta ly s t in two ways: 1) S u lfid e d m etal c a ta ly s ts are more a c tiv e than the oxide form fo r HDNe 2) H^S h e lp s th e h y d ro c ra c k in g a c t i v i t y o f the c a ta ly s t ( I D ), Since s u lfu r is removed as H^S, s u lfu r in a fe e d s to c k d o e s n 't i n h i b i t HDN; however, h ig h n itro g e n c o n c e n tra tio n in a feed i n h i b i t s h y d r o d e s u lfu r iz a tio n (10, 18). Both s u lfu r and' n itro g e n rem oval re a c tio n s proceed by f i r s t - o rd e r k in e c t ic s w ith th e c o n c e n tra tio n o f each type o f compound (16, 17 ). The lo w e s t l i q u i d h o u rly space v e lo c it ie s, LHSV, g e n e ra lly g iv e th e b e st re p o rte d l i q u i d co n v e rs io n s and HDN. Best r e s u lt s were o b ta in e d by derosset fo r SRC a t th e..most severe c o n d itio n s : 450 C,

38 27 UHSV=^-, and 3000 p s ig H over 200 ml o f c a ta ly s t (2 4 ), Ram anthracene o i l showed in c re a s e d n itro g e n rem oval as LHSU decreased a t a p p ro p ria te te m p e ra tu re s (1 8 ). Coal rank in flu e n c e s liq u e f a c t io n r e s u lt s. In th e German p la n ts, the typ e s o f c o a l used determ ined the s e v e r ity o f the o p e ra tin g cond it io n s. Brown c o a l c o u ld be hydrogenated a t much low er p re ssu re than harder co a l (3 ). F in a lly, b e n e fic ia tio n o f co a l a ffe c ts c a t a ly t ic co a l conversio n s (3 2 ). U n b e n e fic ia te d co a l re a c ts b e s t: t h is ' is p ro b a b ly due to c a t a ly t ic FeS in the ro c k and ash mixed w ith the c o a l.

39 28 CATALYSIS C a ta ly s ts are im p o rta n t in co a l h y d ro tre a tin g. C a ta ly s ts work by a llo w in g a lte r n a te mechanism paths fo r h y d ro c ra c k in g, h yd ro g e n a tio n, -and heteroatom rem oval re a c tio n s ; the a lte r n a te.p a th s have low er a c tiv a t io n e n e rg ie s than u n ca ta lyze d p a th s, so c a ta ly s ts in c re a s e re a c tio n ra te s. H yd rocracking c a ta ly s ts must be dual fu n c tio n a l. crack h ig h m o le cu la r w eight hydrocarbons ;and. hydrogenate They must both the u n s a tu r- ates (coke p re c u rs o r) formed (3 3,3 5 ). F a ctors in C a ta ly s t A c t iv it y, S e le c t iv it y, and A c t iv it y M aintenance - Heterogeneous c a ta ly s ts, most o fte n used in h y d r o tr e a tin g, are..a ffe cte d by su p p o rt typ e and s tr u c tu r e, m etal lo a d in g, su rfa ce a c id it y, s u rfa ce area,, pore s iz e, p lu g g in g, and p o iso n s. Alumina or s ilic a - a lu m in a w ith or w ith o u t a lk a lin e and ra re e a rth prom oters are o fte n used fo r h y d ro c ra c k in g c a ta ly s t su p p o rts. S upports o fte n determ ine c ra c k in g, and loaded m etals a ffe c t hydrogenat io n (3 1 ). T y p ic a l c a ta ly s t m e ta ls used fo r h y d ro g e n o ly s is are li s t e d in Table I I (1 0 ). Table I I. T y p ic a l H yd ro g e n o lysis M etals B reaking o f M etal Bond Type' C-C Group V I I I Re W C-N " " " " C-O " " " " r q. 11 II Mo M Cu 1 1 Ag

40 29 Heterogeneous c a ta ly s ts are u s u a lly s u lfid e d ( tr e a te d w ith H^S) to be most a c tiv e. S u lfid e s o f tu n g s te n, molybdenum, c o b a lt, and n ic k e l are more a c tiv e -b n su p p o rts than t h e ir o x id e s. T h is i s fo rtu n a te since most m etal c a ta ly s ts go to s u lfid e s in a Yfo s u lfu r re a c ta n t conc e n tr a tio n (1 0 ). S urface a c id it y o f a c a ta ly s t is im p o rta n t. C a ta ly s t work done in Germany " le d to the. c o n c lu s io n c a ta ly s ts on b a s ic supports-w ere s u ita b le fo r... p re h y d ro g e n a tio n r e a c tio n s, w h ile those w ith a c id ic su pport are p re fe rre d as s p l i t t i n g c a ta ly s ts " (3 4 ). I t has been r e c e n tly found th a t p re fe rre d c e n te r r i n g ' c ra c k in g o f phenanthrene is p o s s ib le w ith a Cr^O^ - Al^O^ c a ta ly s t w ith low a c id ic s tre n g th. "T h e re fo r e, c o n tr o lle d a c id it y may be an approach to enhancing the s e le c t iv it y o f c e n te r r in g c ra c k in g re a c tio n s " (1 0 ). S tro n g ly a c id ic c a ta ly s ts fo r th e p e troleum in d u s tr y are "e x tre m e ly s e n s itiv e to n itro g e n o u s.poisons,; t'h s re fo re, i t appears d e s ira b le to p ro v id e a su p p o rt th a t c o n ta in s a la rg e number o f o n ly m o derately a c id ic s it e s " (3 5 ). S tu d ie s in upgrading SRC found th a t a c id it y o f th e c a ta ly s t s u rfa ce also a ffe c te d carbon d e p o s itio n (1 0 ). D e s u lfu r iz a tio n a c t i v i t y i s in flu e n c e d by a c id it y a ls o. "h y d ro g e n a tio n i re a c tio n s occur on s tro n g ly e le c t r o p h ilic s ite s, w h ile d e s u lfu r iz a tio n occurs on weakly e le c t r o p h ilic s ite s " (10.). High s u rfa ce area i s im p o rta n t to c o a l p ro d u c t c a ta ly s ts, as found by Montana S ta te U n iv e r s ity (3 6 ), and U n iv e r s ity o f N orth Dakota (3 7 ). In a s e rie s o f in -h o u se c a ta ly s t su p p o rts im pregnated w ith Ni-Mo

41 30 a t the U n iv e r s ity o f N orth Dakota, a c t i v i t i e s per u n it o f s u rfa c e area were h ig h e s t fo r c a ta ly s ts c o n ta in in g TiO^ ( 0^ and I j ) ; but a comm e rc ia l Ni-Mo c a ta ly s t on alum ina had much h ig h e r s u rfa c e area and b e t t e r o v e r a ll a c t i v i t y (3 7 ). Pore s iz e c o n s tr a in ts l i m i t heterogeneous c a ta ly s ts. Asphaltene m olecules are re p o rte d to be A b ig enough to p lu g sm a ll pores (1 0 ). E ngelhard In d u s tr ie s (35) i s te s tin g c a ta ly s ts on 100 A.. s u p p o rts ; W e lle r found 200 A pore s iz e to be optimum in h is co a l tre a tm e n t work (3 8 ); Exxon got the b e st asphaltene rem oval w ith c a t a l y s ts in the ' Arrange, as compared w ith A and 40 A (3 9 ). Although z e o lite s h e ld hope as c a ta ly s ts because o f t h e ir s u p e rio r re s is ta n c e to MH and H S ( l 0 ) j t h e ir pore size o f 5-13 A (33) is p ro b a b ly too s m a ll, Heterogenous c a ta ly s ts s u ffe r a c t i v i t y d e g ra d a tio n,from both ' p o is o n in g and f o u lin g. S u lfu r and n itro g e n poison some c a ta ly s ts when p re s e n t above p a rt per m illio n c o n c e n tra tio n s ; s u lfu r u s u a lly h u rts h y d ro g e n a tio n components and n itro g e n h u rts the c ra c k in g components (33) C a ta ly s t d e a c tiv a tio n has been, noted fo r p ro ce ssin g SRC, S y n th o il, and COED t a r, a l l which c o n ta in s u lfu r and n itro g e n a t about I co n c e n tra t io n. E a s ily s u lfu r-p o is o n e d c a ta ly s ts in c lu d e p la tin u m and p a lla d iu m ; s u lf u r - r e s is t a n t c a ta ly s ts are molybdenum, tu n g ste n, ir o n, n ic k e l- molybdenum, and n ic k e l-tu n g s te n (4 0 ). N itro g e n also causes' ir r e v e r s ib le a c t i v i t y lo s s fo r p la tin u m on s ilic a - a lu m in a (3 3 ).

42 31 A lk a li s a lts and tita n iu m compounds are p o isons in co a l p ro ce ssin g. For example, sodium io n s cause ir r e v e r s ib le s u rfa ce area lo s s d u rin g c a ta ly s t re g e n e ra tio n ; TiO^ d e p o s itio n on s ilic a - a lu m in a. in c re a s e s s u rfa c e a c id it y (1 0 ). F o u lin g r e s u lt s from e ith e r c o k e fo rm a tio n or tra c e contam inan ts c o v e rin g c a ta ly s t r e a c tio n s u rfa c e s. M e ta ls,- ash, and asphalte nes may a l l c o n trib u te to c a ta ly s t d e g ra d a tio n as observed in SRC upgrading (1 0,2 4 ). Homogeneous c a ta ly s ts, molten- s a lts, and s u p e r c r it ic a l c a ta ly s is are suggested as more e f f e c t iv e in a v o id in g mass tr a n s fe r lim it a t io n s o f th re e -p h a se heterogeneous c a ta ly s is. Lewis a c id -ty p e compounds are u s u a lly good homogeneous c a ta ly s ts : these in c lu d e m etal h a lid e s, e s p e c ia lly c h lo rid e s..z in c c h lo rid e i s. being used fo r co a l liq u e f a c t io n re se a rch in b o th -s m a ll and la rg e c o a l- t o - c a t a ly s t r a t io s. C a ta ly s t re c o v e ry and c o rro s iv e p r o p e r tie s o f c h lo rid e s are two r e s t r a in in g fa c to r s in m etal c h lo rid e use. Zinc "c h lo rid e re a c ts w ith H S and NH^ d u rin g liq u e f a c t io n re a c tio n s, but c u rre n t processes have been developed ;fo r 99.9% c a ta ly s t re co ve ry (3, 8, 26). Much work i s c o n tin u in g in heterogeneous and homogeneous c a ta ly s is o f co a l c o n v e rs io n. Good sources in c lu d e Rapoport (4 0 ), S ta n fo rd Research I n s t i t u t e (2 0 ), U n iv e r s ity o f N orth Dakota (2 9 ), Conoco Coal Development (2 6 ), H ite sh u a (2 8 ), E nglehard In d u s tr ie s (3 5 ), Montana S ta te U n iv e r s ity (3 6 ), and p a te n ts by M o b il O il (4 1 ), Exxon (3 9 ),

43 U n iv e r s a l O il Products- (4 2 ), and A t la n t ic R ic h fie ld (4 3 ). 32 Given any one p a r t ic u la r c a ta ly s t system, d if f u s io n processes may be ra te -d e te rm in in g. The paths.for r e a c tio n u sin g heterogeneous c a ta ly s ts are as fo llo w s : 1. D iffu s io n o f re a c ta n ts to c a ta ly s t e x te r io r 2. D iffu s io n o f re a c ta n ts in pores 3. A d s o rp tio n on c a ta ly s t s u rfa ce 4. R eaction 5. D e so rp tio n o f p ro d u c ts 6. D iffu s io n o u t o f pores 7,. D iffu s io n away from c a ta ly s t e x t e r io r. F ilm d if f u s io n re s is ta n c e can be decreased by in c re a s in g the re a c ta n t mass flo w r a te p a st the c a ta ly s t p a r t i c l e ; pore d if f u s io n re s is ta n c e i s decreased by using s m a lle r c a ta ly s t p a r t ic le s having le s s pore le n g th F ilm d if f u s io n i s te s te d by making two s e rie s o f ru n s, each a t. a d if f e r e n t c o n s ta n t'w e ig h t, W, o f c a t a ly s t. The l i q u i d feed r a te, F, i s v a rie d in each s e rie s o f ru n s. C onversion i s p lo tte d ve rsus W/F. I f the curve s c o in c id e, f ilm d if f u s io n i s n o t r a t e - c o n t r o llin g. Pore d if f u s io n i s te s te d by making two ru n s a t the-same c a ta l- y s t - charge, b u t using d if f e r e n t s iz e p e lle t s. The same c o n ve rsio n at a c o n s ta n t w eight o f c a ta ly s t/fe e d r a te r a t i o, W/F, means pore d if f u s io n i s not ra te c o n t r o llin g. I f pore re s is ta n c e i s c o n t r o llin g, then the co n v e rs io n w i l l in c re a s e w ith decreasing p e lle t siz e (4 4 ).

44 MATERIALS, EQUIPMENT, AND PROCEDURES Coal C om position _ T u jo 1 0 -g a llo n cans o f s o lid S o lv e n t-r e fin e d Coal mere re c e iv e d from P itts b u rg h and Midmay M in in g Company's F t. Lemis p i l o t p la n t. I t i s Sample 110, taken Oune 13, The co a l type used to make the SRC mas Kentucky #9 from the C o lo n ia l M ine. The co a l i s mashed and siz e d a t the m ine; th e size used i s i " x 0 d rie d to <( Yfo m o is tu re b e fo re use. The 1975 y e a rly average1pro xim a te and u ltim a te analyses as re c e iv e d are l i s t e d in Table 3

45 Table 3 Coal Analyses to SRC Process Proxim ate w t. p e rc e n t as re c e iv e d M o istu re 2.25 Ash V o la tile s ' Fixed Carbon U ltim a te wt. p e rc e n t as re c e iv e d M o istu re 2.25 C ' 70,10 H 4.89 N 1.37 S 3.51 Cl 0.04 Ash by d iffe re n c e SRC C om position The u ltim a te a n a ly s is o f the SRC sample as re p o rte d by PAMCO i s g iv e n b e lo w "in Table 4. Table 4. SRC U ltim a te A n a ly s is C 67.33% H 5,60% N' 2.28% S 0.68% % Ash 0.3 3%

46 35; I No com plete compound a n a ly s is i s a v a ila b le in th e l i t e r a t u r e fo r S o lv e n t-r e fin e d C o a l, C o a l-d e riv e d liq u id s are known fo r t h e ir la rg e number o f compounds p re s e n t, none o f which u s u a lly exceeds 2% c o n c e n tra tio n. Table 5 l i s t s compounds in the p ro d u c t from Union Carbide c o a l hydrogenat io n pro ce ss o f the 19401s and 1950's, Most c u rre n t co a l p ro d u c ts are c h a ra c te riz e d by typ e s o f s o lu b i l i t y f r a c t io n s. A.3. de Rosset o f UOP c h a ra c te riz e d an SRC sample as fo llo w s (2 5 ): C in s o l. ' 89.4/a Benzene in s o l ^ DMF/Xylene in s o l ^- Ash c o n te n t 0.11%

47 36> \ Table 5. T y p ic a l Compounds in a C o a l-d e rive d L iq u id (13) Arom atic Indane and H yd ro a ro m a tic: T e tra h yd ro n a p h th a le n e Naphthalene 1- M ethy!naph thalene 2 - M ethy!naph thalene 2, 3 - D im ethylnaph thalene D im ethy!naphthalene D im ethylnaph thalene Acenaphthene D iphenyl Fluorene 9, 1 0 -D ih ydroa nthracene Anthracene Phenanthrene and the fo llo w in g bases: P y rro le P y rid in e. 2 - M ethyl p y r i dine 3 - M e th y lp y rid in e. A -M e th y lp y rid in e 2 - E thylp y r id in e 3 - E th y lp y r id in e A - E th y lp y r id in e 2 - M ethy1-6 - e t h y lp y r idine 3 - M ethyi eth y lp y r id in e A -M e th y l-2 - e th y lp y r id in e 2, 3, A -T rim e th y lp y rid in e 2, 3,5 - T r im e th y lp y r id in e 2, 3, 6 - T r im e th y lp y r id in e 2, A, 5 -T rim e th y lp y rid in e. 2,A,6 - T r im e th y lp y r id in e A n ilin e 2 - M ethyi a n ilin e 3 - M ethyi a n ilin e A -M e th y la n ilin e 2 - E th y la n ilin e 2, 3 -D im e th y la n ilin e 2,A -D im e th y la n ilin e Arom atic and H ydroarom atic co n tin u e d : Chrysene Perylene Picene A lip h a t ic : ri-undecane _n-dode cane ni-tridecane B-Heptadecane H e te r o c y c lic : D ibenzofuran Benzothiophene D ibenzothiophene 2, 3 - D im e th y!p y rid in e 2, A -D im e th y lp y rid in e 2, 5 - Dim ethylp y r id in e 2, 6 - Dim ethyl p y r i dine 3, A -D im e th y lp y rid in e 3, 3 - D im e th y lp y rid in e 2 -ri-p ro p y lp y rid in e 2 -M e th y l-a -e th y lp y rid in e D im e th y la n ilin e 3,A -D im e th y la n iiin e D im e th y la n ilin e 1- Naphth y I amine N aphthylam ine In d o le 2 - M e th y lin d o le M e th y lin d o le Carbazole Q u in o lin e 2 - MethyI q u in o lin e 7 -M e th y lq u in o lin e Is o q u in o lin e 1-M ethyi i s o q u in o lin e 3 - M e th y lis o q u in o lin e

48 37 M o b il O i l (13) Lsed g ra d ie n t e lu t io n chrom otography to c h a ra c te r iz e SRC in t o 13 d if f e r e n t f r a c t io n s. Table 7 g iv e s and e x p la in s a c h a r a c te r iz a tio n o f a.w ils o n v ille, Alabama SRC. S o lvent C om position The s o lv e n t used in the p r o je c t i s one fo rm e rly used by M o b il O il re s e a rc h e rs (A ), The co m p o sitio n mas chosen due to i t s a p p ro x im a tio n o f a re c y c le s o lv e n t which mould be used in an SRC pro ce ss. Table 6. S o lve n t C om position (4 ) 43/ t e t r a l i n 38^ m ethyl naphthalene 17% m-p C resol 2% gamma p ic o lin e C o ntains.36% S and.59% N Feed to the C ontinuous R eactor ', The SRC s o lu tio n feed to the c o n tin u o u s re a c to r mas prepared by p re s s u re -te m p e ra tu re tre a tm e n t in batch a u to c la v e s. ft 1.5 :1 s o lve n t-to -S R C charge mas pre ssu re d to 2000 p s ig 0 to 400 C and h e ld th e re fo r o n e -h a lf h o u r. fo r 8 h o u rs, d e p re s s u riz e d -a n d em ptied o f s o lu tio n. i n i t i a l l y, heated The bomb mas a ir cooled

49 Table 7. D e s c rip tio n, C o n c e n tra tio n, and Makeup o f M o b il's GEC F ra c tio n A n a ly s is o f SRC Concn in GKC fraction (wt %) GEC JTncUon Description Hydrocarbon type Saturate M N A - DNA PNA oils PNA snft resin I. Satoratcs 2. M NA-DNAoil 3. I NA oil 4. PNAsoftresin Colorless oil or white wax. melting slightly above room temperature. Very soluble in paraffinic solvents. Very light yellow heavy oil, sometimes with waxy solids, easily melted. Very soluble in aromatic solvents. Bright yellow-orange oil. Usually a tacky orange-red semisolid. slow to turn fluid at elevated temperature, but pours readily at 200 0F. 5. Hard resin Bed brown powders which turn tacky 6. PoIarresin slightly above 100 F. Extremely viscous Iinuide at 240 ef. 7. Eluted asphaltenes Dark lirown powder. Behavior intermediate between resins and polar asphaltenes. Very soluble in methylene chloride Polar asphaltenes Black powder, very friable. Very high softening point. Poorly soluble in methylene chloride, very soluble in T i IF. 13. Nonelulcd Among others, can he strongly acidic. components (carboxylic acids), inorganic solids (carbon, catalysts fines, etc.). Paraffins Olefins a b b. b Naphthenes Cl.O Monnaromatics Diaromatics Triaromatics Tetraaromatics Pcniaammatics Aromatic sulfur compounds Unidentified aromatics G Total %Insufficient sample. h Analysis not applicable. GEC FRACTION VLflV1BER

50 P re s u lfid in g R eactor 39 Comrnerical c a ta ly s ts are d rie d a t IOQ0 C fo r a day, then s u l fid e d. S u lfid in g the" c a ta ly s ts in c re a s e s c ra c k in g and d e n itro g e n a tio n a c t i v i t y.. The s u lf id in g re a c to r i s a tiuo fo o t long one in c h I. D, s te e l p ip e, w ith a q u a rte r in c h th e rm o w e ll e xte n d in g a fo o t in t o i t. The p ip e i s loaded w ith o n e -fo u rth in c h i n e r t alum ina su p p o rt spheres, then c a t a ly s t, then more i n e r t spheres. A m ix tu re o f hydrogen and ^ hydrogen s u l f i d e flo w s th ro u g h the re a c to r fo r tw e lve hours a t 350.C and atm ospheric p re s s u re. The outgas i s scrubbed w ith w ater and sodium h y d ro x id e. The tem perature i s m a in ta in e d w ith a c o n t r o lle r and e le c t r ic p ip e h e a te r. The c a ta ly s t i s cooled w ith c o n tin u e d gas flo w. S u lfid e d c a ta ly s ts are s to re d in sealed ja r s to p re v e n t exposure to a ir and w a te r. Rocking Bomb Reactor A Parr s e rie s 4000 p re ssu re r e a c tio n apparatus was used to make batch ru n s (4 6. ), D e ta ils o f the h e a te r and bomb are shown in F igure 5 The P arr bomb was charged w ith s o lv e n t, s o lid SRC, and c a t a ly s t. Twenty ml. o f heterogeneous c a ta ly s t or 10 m l,o f powders and s a l t s. were used. A p re ssu re head and copper gasket f i t on the top o f the bomb. An e ig h t - b o lt screw cap secured the head; a pre ssu re gauge b lo c k was connected.to th e head.

51 40 B re a th e r Tube / A utoclave body Gauge b lo c k Thermocouple hole i I F ig u re 5. A utoclave and H eater

52 41. The bomb mas p re s s u riz e d w ith hydrogen to about 2000 p s ig w ith a H askel gas compressor (4 7 ). I f no le a k s appeared in f iv e m inutes, th e bomb was p la ce d in the r o c k e r - h e a te r. A b o lt screwed th ro u g h the h e a te r base and in t o th e bomb bottom secured i t. An iro n -c o n s ta n ta n therm ocouple was in s e rte d in t o a h o le o f th e se c u rin g b o lt. The e le c t r ic d riv e ro c k e r was s ta r te d ; heat in p u t was c o n tr o lle d w ith a P ow erstat v a r ia b le tra n s fo rm e r; tem p e ra tu re was graphed by a H oneyw ell te m p e ra tu re re c o r d e r. I n i t i a l runs were conducted fo r o n e -h a lf hour a t C. Subsequent runs were a t & C sin ce no coking was observed. About 50 m inutes he a t-u p were re q u ire d to reach run te m p e ra tu re. A fte r the ru n, the bomb co oled in a ir to room te m p e ra tu re. Gas p re ssu re was b le d o f f g ra d u a lly over 10 m inutes to m inim ize lo s s o f l i q u i d p ro d u c t. P roducts were s to re d in g la s s b o ttle s u n t i l analyses were perform ed. C ontinuous R eactor. The re a c to r system i s shown in F ig u re 6. The re a c to r i s a o n e -in ch 0. D. th r e e - fo o t long In c o n e l p ip e. I t i s packed w ith a s e c tio n ' o f in e r t N orton Denstone su p p o rt in the top s e c tio n ; the mid s e c tio n h o ld s a mix o f a c tiv e c a t a ly s t and i n e r t s u p p o rt; the bottom s e c tio n is f i l l e d w ith in e r t s u p p o rt. A 1 /4 s ta in le s s s te e l th e rm o w e ll extends from the top down th ro u g h the packed c a ta ly s t and s u p p o rt. 0

53 42 Temp Recorder Thermowel I B a ll Check Heater Feed Reservoir & Feed Lines V ariable Transformers Micro, Metering Valve C atalyst' Inconel Reactor Aluminum Heating Jackets Hydrogen Gas I Outlet Gas-Liquid Separator = : Gas In le t Liq u id Catchpot '^ P r e s s u r e Release FIGURE 6 liq u id O utlet Continuous Reactor

54 ..4 3 The re a c to r p ip e f i t s in an aluminum b lo c k, which f a c i l i t a t e s is o th e rm a l o p e ra tio n. Three nichrom e h e a tin g w ire s wrapped around the b lo c k are c o n tr o lle d by manual v a r ia b le tra n s fo rm e rs. A sheet m etal c o n ta in e r w ith z o n o lite in s u la t io n in s id e encloses the aluminum b lo c k. Three chrom e-alum el therm ocouples in the th e rm o w e ll m o n ito r the to p, m id d le, and bottom p a rts o f th e a c tiv e c a t a ly s t s e c tio n ; the therm ocouples are connected to a Leeds and N o rth ro p tem perature re c o rd e r. T e c h n ic a l grade hydrogen flo w s through a Nupro m e te rin g v a lv e, a b a ll check v a lv e, and p a st a p re ssu re gauge in t o the top o f the r e a c t o r. A s o lu tio n o f s o lv e n t- r e fin e d co a l i s pumped from a g la s s r e s e r v o ir th ro u g h 1 /8 " s ta in le s s s te e l tu b in g and in t o the re a c to r- by a H ills-m cc a n n a chem ical p ro p o rtio n in g pump. The r e s e r v o ir, tu b in g, and pump are a l l e l e c t r i c a l l y heated to p re v e n t the fe e d s to c k from p lu g g in g lin e s or pump check v a lv e s. The SRC s o lu tio n and hydrogen flo w downward th ro u g h the c a t a ly s t bed and in t o the g a s - liq u id s e p a ra to r. Grove b a ck-p re ssu re r e g u la to r, are scrubbed, Gases pass through th e and vented in t o a hood. L iq u id s drop down in t o the l i q u i d c a tc h p o t. Ta empty the c a tc h p o t, the v a lv e between the g a s - liq u id sepa ra to r and c a tc h p o t i s c lo s e d. P ressure i s l e t dow n.through the p re ssu re re le a s e v a lv e and p ro d u c t i s c o lle c te d from the bottom o f the (I

55 44 c a tc h p o t. A fte r the p ro d u c t dump, th e c a tc h p o t i s re -p re s s u riz e d w ith n itro g e n th ro u g h th e gas i n l e t v a lv e. The v a lv e between the se p a ra to r and th e p a tc h p o t i s re -opene d, w ith no d is r u p tio n o f system p re ssu re. The c y c le i s re p e a te d fo r ecery l i q u i d sample. / C ontinuous Runs The fo llo w in g d e s c rib e s th e best procedure developed d u rin g the p r o je c t. One p r e lim in a r y ru n had m inor d iffe re n c e s in c a ta ly s t amount and therm ocouple placem ent, b u t th e r e s t o f th e c o n tin u o u s ru n s were conducted as o u tlin e d here. A r u le o f thumb is th a t 100 m l. o f sm all p e lle t s f i l l 9^-16" o f re a c to r w ith the th e rm o w e ll in p la c e. The re a c to r was f i l l e d as fo llo w s : 1 /4 " spheres in e r t N orton Denstone su pport 1 /8 " e x tru s io n s in e r t N orton Denstone su p p o rt 15 ml. 120 ml. 13.5" preheat s e c tio n mix o f c a ta ly s t 140 ml. and in e r t Denstone (70 m l. a c tiv e c a ta ly s t) 14" r e a c tio n s e c tio n 1 /8 i n e r t Denstone 80 ml. 8 " Rest', was f i l l e d w ith 1/4" spheres and s te e l screen. to h o ld packing m a te ria l in nlar.e Therm ocouples were p la ce d in th e th e rm o w e ll to measure tem peratu re s in th re e p a rts o f th e r e a c to r. The f i r s t was 10^ in t o

56 45 the packed m a te ria l (near the bottom o f th e p re h e a t s e c tio n ). The second mas 20 1/4" in t o the b e d.(c e n te r o f the re a c tio n s e c t i o n ).. The t h i r d mas 28 1 /4 " in t o the bed ( ju s t belom end o f re a c tio n s e c tio n ). The re a c to r mas placed in the aluminum b lo c k. Feed lin e s and p ro d u c t s e p a ra tio n and c o lle c t io n p ie c e s mere connected; the mhole system mas p re s s u riz e d and te s te d fo r le a k s. The re a c to r mas heated to steady s ta te tem perature in about 8 ho u rs. Power in p u t mas m anually c o n tr o lle d w ith th re e Pom erstat v a r i able tra n s fo rm e rs and in d ic a te d by th re e ampmeters. The Leeds and N orthrup tem p e ra tu re re c o rd e r p r in te d bed te m p eratures d u rin g a fo u r - m inute c y c le. Hydrogen flo w to the re a c to r mas c o n tr o lle d ' by a m ic ro - z m e te rin g v a lv e. E xperience showed th a t c a lib r a t io n o f the m ic ro - m e te rin g v a lv e i s u n r e lia b le. A chrom atograp hic a n a ly s is a tte m p t o f o ff-g a s e s from s im ila r re a c to rs a t Montana S tate U n iv e r s ity found the o ff- g a s to be ove rw h e lm in g ly hydrogen. T h e re fo re, a met t e s t meter- mas connected a f t e r the o ff- g a s scrubber to measure th e hydrogen flo w r a te. A fte r the hydrogen ra te mas s e t, pumping o f the SRC s o lu tio n mas begun. Flow mas measured by the le v e l in a graduated g la s s re s e r v o ir. Hydrogen and l i q u i d flo w r a te s were checked every 20 m inutes sin ce both tended to f lu c t u a t e. Flows co u ld u s u a lly be c o n tr o lle d to - 10% o f the d e s ire d h y d r o g e n - to - o il r a t i o.

57 46 Once both Flow ra te s and te m p e ra tu re were s te a d ie d o u t, the re a c to r was op e ra te d a t c o n s ta n t c o n d itio n s For a t le a s t 90 m inutes before th e F ir s t sample, p e rio d began. When o p e ra tin g param eters were changed, a 60 m inute t r a n s it io n p e rio d a t c o n s ta n t c o n d itio n s was a llo w e d. T r a n s itio n p e rio d p ro d u c ts were c le a re d From c a tc h p o t before sample p e rio d c o lle c t io n began. Each sample p ro d u c t was about 80 to 100 ml. P o te n tia l p lu g g in g of th e re a c to r system re q u ire d co n tin u o u s m o n ito rin g ; runs la s te d From tw e lv e to tw e n ty hours, w ith Four to nine sample p ro d u c ts per ru n. AFter the ru n, Feedstock s o lu tio n, hydrogen, and power in p u t were h a lte d. M otor o i l was pumped th ro u g h the system to h e lp clean lin e s and loosen re a c to r p a ckin g. The Feed lin e s and bottom c o lle c t io n p ie c e s were disconne cted From th e r e a c to r ; the re a c to r was removed From th e b lo c k w h ile s t i l l h o t. The th e rm o w e ll must be p u lle d w h ile th e re a c to r i s h o t, o th e rw is e i t and the packing u s u a lly become stu c k and the p ip e must be d r i l l e d o u t. Acetone i s pumped th ro u g h th e r e s e r v o ir and lin e s to c le a r them. A ll p ie c e s From th e r e s e r v o ir to th e o u t le t sid e of th e pump are d is m a n tle d and cleaned if necessary. The re a c to r p ip e, s e p a ra to r, and c a tc h p o t are cleaned w ith acetone; th e.b a c k p re ssu re r e g u la to r is d is m a n tle d and cleaned.

58 47 A n a ly tic a l Procedures P roducts from the batch and co n tin u o u s runs were analyzed to determ ine th e degree o f cracking., d e n itro g e n a tio n, and d e s u lfu r iz a tio n Water in the sample in t e r f e r e d w ith the d i s t i l l a t i o n a n a ly s is, so D r ie r it e ( anhyhdrous CaSO^) was used. The o n ly e f f e c t iv e method was to d ir e c t ly add Jf teaspoon o f D r ie r it e to 100 ml. samples in sealed ja r s. An in d ic a t io n o f c ra c k in g was determ ined by ASTM atm ospheric d i s t i l l a t i o n method D-86 ( 48 ). T h is method determ ines th e amounts o f d i s t i l l a b l e s a t te m p e ra tu re s between 70 and 700 F..'Comparison o f feed and p ro d u c t d i s t i l l a t i o n curves o f p e rc e n t d i s t i l l e d vs. tem pera tu re a llo w s c a lc u la tio n o f c ra c k in g. I W eight p e rc e n t s u lfu r in the samples i s determ ined by the q u a rtz -tu b e com bustion method, ASTM method D1551 ( 4 9,' 5 0 ). Samples from m etal c h lo r id e c a ta ly s t ru n s had hydrogen c h lo rid e i n t e r fe re n ce o f t i t r a t i o n measurement in ASTM D1551. A g ra v im e tric procedure which p r e c ip ita te d the s u lfa te using barium c h lo r id e was used fo r those samples ( 5 1 ). W eight p e rc e n t n itro g e n in samples was determ ined by the Macro K je ld a h l method, ASTM E 258 ( 5 2, 53 -). C hrom atographic a n a ly s is o f co n tin u o u s re a c to r o ff- g a s has not been s u c c e s s fu lly done.

59 RESULTS AND DISCUSSION The o b je c t o f th e experim ents mas to c a t a l y t i c a l l y c o n v e rt PAMCO1s SoIv e n t-r e fin e d Coal (SRC) to cle a n d i s t i l l a t e fu e ls. A v a ila b le com m ercial c a ta ly s ts were f i r s t te s te d, then c a ta ly s ts prepared here at Montana S ta te U n iv e r s ity. L iq u id c o n ve rsio n mas measured by d i s t i l l a t i o n o f'p r o d u c ts up to 700 F; s u lf u r c o n te n t o f th e p ro d u c ts mas determ ined by th e q u a rtz tube com bustion method; n itro g e n c o n te n t mas determ ined by th e Macro K je ld a h l method. T w e n ty -e ig h t c a ta ly s ts were used in t h ir t y - t w o batch ru n s; f iv e c a ta ly s ts were used in f iv e c o n tin u o u s ru n s. The e f f e c t s o f te m p e ra tu re, re s id e n c e tim e ( in term s o f l i q u i d h o u rly space v e lo c it y ), and h y d r o g e n - to - o il r a t i o were in v e s tig a te d in th e co n tin u o u s runs o f t h e, t r i c k l e bed re a c to r. M etal lo a d in g, s u p p o rt, s u rfa c e area, and pore volume data a v a ila b le on the c a ta ly s ts te s te d are l i s t e d in Appendix A. C onversion, heteroatom rem oval, and ru n c o n d itio n s f o r each batch t e s t are li s t e d in Appendix B. Appendix C c o n ta in s c o n v e rs io n s, heteroatom rem oval, and ru n c o n d itio n s fo r a l l c o n tin u o u s ru n s. P re lim in a ry Runs Feed S o lu tio n - 'PAMCO1s S o lv e n t-r e fin e d Coal is a d a rk, sh iny s o lid a t room te m p e ra tu re ; o n ly 3% o f i t w i l l b o il o f f up to 650 F (1 3 ). Such a s o lid c o u ld n 't be e a s ily pumped in t o a co n tin u o u s r e a c to r, so a s o lv e n t v e h ic le mas used.

60 49 A s u ita b le s o lv e n t fo r co a l work mas suggested a t th e 1975 ERDA U n iv e r s ity C o n tra c to rs C onference. I t i s d e scrib e d in M a te ria ls, Equipment and Procedures.- The use o f as l i t t l e s o lv e n t as p o s s ib le made e le v a te d tem peratu re s necessary. To r e t a in the n itro g e n compound c h a r a c t e r is t ic s o f a co a l o i l s o lv e n t, th e s o lv e n t co u ld n o t be heated p a s t 120 C a t.atm ospheric p re s s u re, where th e most v o l a t i l e component, gamma p ic o lin e, s t a r t s b o ilin g o f f. SRC m i l l d is s o lv e in a 7 :2 s o lv e n t- to - c o a l w eight r a t i o in th e s y n th e tic co a l o i l s o lv e n t a t IIQ 0C. T h is s o lu tio n was p icke d as th e fe e d s to c k. The r e s e r v o ir, pump, v a lv e s and lin e s o f the co n tin u o u s apparatu s were m o d ifie d w ith e l e c t r i c a l h e a tin g ta pes to handle the h o t s o lu tio n. A n a ly tic a l D e fin itio n s - L iq u id c o n ve rsio n data was o b ta in e d by com paring c a ta l y t i c a l l y - tre a te d p ro d u c t w ith the o r ig in a l 7 :2 s o lu tio n. D i s t i l l a t i o n o f th e 7 :2 co a l s o lu tio n showed th a t 79% o f i t b o ils below 475 Fj th e re m a in in g 21% o f n o n - d is t illa t e s decomposes above 475 F. D i s t i l l a t i o n o f th e s y n th e tic c o a l o i l s o lv e n t alone showed 98% d i s t i l l s below 47Q F; 2% decomposes above 470 F. Since e s s e n t ia lly a l l th e s o lv e n t b o ils below 700 F, but ve ry l i t t l e s o lid SRC d i s t i l l s below 700 F, l i q u i d c o n v e rs io n s fo r the p r o je c t were d e fin e d as fo llo w s :

61 l i q u i d _ % t o t a l d i s t i l l a t e ^ t o t a l d i s t i l l a t e co n v e rs io n ~ in p ro d u c t in feed % n o n - d i s t i l l a t e in feed L iq u id c o n ve rsio n i s th e -percentage amount o f n o n - d is t illa t e th a t becomes d i s t i l l a t e. p ro d u c t, the co n v e rs io n i s z e ro. the l i q u i d c o n v e rs io n i s 100/o. I f no in c re a s e o f d i s t i l l a t e i s noted in the v I f the t o t a l d i s t i l l a t e becomes IOO^o, P ercent d e s u lfu r iz a tio n i s d e fin e d as fo llo w s : % s u lfu r - /o s u lfu r in p ro d u c t in feed c^e ^ ~ /o s u lfu r in feed D e n itro g e n a tio n i s d e fin e d in the same manner. The 7 :2 c o a l s o lu tio n was 0.43$! s u lfu r and % n itro g e n. 1 P re lim in a ry C ontinuous Run - The f i r s t c o n tin u o u s ru n s e rie s was made a t 450 C, 1000 p s ig, 100 m l; o f Ni-4401 n ic k e l tu n g s te n c a t a ly s t, and hydrogen flo w o f 10,000 SCF/bbl o i l. The d i s t i l l a t i o n o f th e p ro d u c t i s compared to the d i s t i l l a t i o n o f the 7 :2 s o lu tio n feed in F igure 7. ; j P roduct 1, d is p la y e d on the d i s t i l l a t i o n graph, was o b ta in e d a t a l i q u i d h o u rly space v e lo c it y (LUSU) o f 2. The d i s t i l l a t i o n curves o f p ro d u c t I and the 7 :2 s o lu tio n feed are graphed w ith tem p e ra tu re versus volume p e rc e n t d i s t i l l a t e c o lle c te d. E ig h ty -o n e and a h a lf p e rc e n t o f th e p ro d u c t d i s t i l l e d below 700 F. The l i q u i d c o n v e rs io n I.. '

62 !.s (p ro d u c t d i s t i l l a t e s ) - 79/o ( s o lu tio n Feed d i s t i l l a t e ) 21/o ( s o lu tio n feed n o n - d is t illa t e s ) = 12#. For P ro d u c t.2, LHSV was changed to f and a l l o th e r c o n d itio n s rem ained th e same as fo r P roduct 1. The re a c to r never reached steady s ta te o p e ra tio n. The gas r a te surged the e n tir e p e rio d and d id not s e t t le to a steady v a lu e. The run was h a lte d and c le a n in g attem pted by pumping s t r a ig h t s o lv e n t. The re a c to r plugged a b ru p tly and pressure rose r a p id ly.. The e n tir e system was shut down and a llo w e d to c o o l. L a te r e xa m in a tio n found a hard carbon b lo c k had se t up in th e re a c to r p ip e. S everal im p o rta n t.c o n c lu s io n s co u ld be made about the ru n : 1) M o d ific a tio n o f th e re a c to r system had to be made. O u tle t liq u id s and gases both passed through th e back pre ssu re r e g u la to r v a lv e. P o s s ib ly liq u id s gummed up th e v a lv e and backed up in t o the r e a c t o r. The o th e r p o s s i b i l i t y was th a t th e r e a c tio n tem p e ra tu re o f 450 C was too high and the s o lu tio n. feed cracked to o f a r, to carbon. 2) Turn around tim e fo r c le a n in g and p re p a rin g the contin u o u s re a c to r fo r the next ru n was too lo n g. Batch bombs are f a s te r and s im p le r to p re p a re, o p e ra te, and c le a n. Batch bombs were used th e r e a fte r to screen poor c a ta ly s ts from p ro m is in g ones.

63 52 Tem perature in :2 Feed P roduct Volume P ercent D i s t i l l e d F igure?. P roduct I Compared to 7 :2 S o lu tio n Feed

64 53 Bomb Runs S tudying V a ria b le s Other Than C a ta ly s t Type - E ig h t bomb ru n s were made which te s te d v a ria b le s o th e r than c a t a ly s t. The ru n s are l i s t e d in c h r-o n o lo g ica l o rd e r on the f i r s t page o f Appendix B, Batch Run Data. ' The e f f e c t o f i n i t i a l charge p re ssu re o f mas in v e s tig a te d. Two bombs were charged w ith 125 g r. o f h o t 7 :2 co a l s o lu tio n and 20 grams o f Ni-4401 c a t a ly s t, but one was p re s s u riz e d to 2100 p s ig and th e o th e r to 500 p s ig. Run tim e was o n e -h a lf hour a t C. Three o b s e rv a tio n s were made: 1) D uring a ru n, bomb p re s s u re s in c re a s e d w ith tem perature to a peak, then decreased re g a rd le s s o f the te m p e ra tu re. 400 C was the lo w e st tem p e ra tu re a t which the decrease was noted. T h is suggests th a t hydrogen is consumed, by re a c tio n and I th a t 400 C. i s about th e minimum re a c tio n te m p e ra tu re. 2) The p ro d u c ts were homogeneous, b la c k, e a sy-p o u rin g liq u id s a t room tem perature and p re s s u re. The c h a ra c te r o f the 7 :2 atm ospheric s o lu tio n feed is q u ite d if f e r e n t. SRC p r e c ip it a t e s from the 7 :2 s o lu tio n a t le s s than 100-1IQ0C. 3) The h ig h pre ssu re fa v o re d liq u i d c o n ve rsio n and m ild ly in c re a s e d heteroatom re m o va l. High versus low p re ssu re p ro d u c t data are 23% to 12% l i q u i d c o n v e rs io n,. 53% to 44/a d e s u lfu r iz a t io n, and 35% to 20% d e n itro g e n a tio n.

65 54 A run w ith no s o lv e n t, ju s t SRC and c a t a ly s t, was done. 38.9gr SRC and 20 g r. powdered c a ta ly s t-w e re ru n a t 725 p s ig and 420 C. A few drops o f l i q u i d form ed; a p p a re n tly s o lv e n t i s needed fo r good c a ta ly s t c o n ta c t. The q u e s tio n arose whether the homogeneous p ro d u c ts from the f i r s t two bomb runs were room te m p e ra tu re s o lu tio n s because o f tem pera tu re and p re ssu re e ffe c ts or c a ta ly s t e ffe c ts. The u n ca ta lyze d fo u rth ru n l i s t e d in Appendix B co n firm e d th a t one h a lf hour a t C,and 750 p s ig i n i t i a l p re ssu re w ill- d is s o lv e SRC in to a s o lu tio n th a t sta y s homogenous when reduced to room tem peratu re and p re s s u re. S o lu tio n -to -S R C r a t i o versus s o lu tio n c h a r a c t e r is t ic s were te s te d in th re e more bomb runs w ith o u t c a ta ly s t at 750 p s ig i n i t i a l and d n e -h a lf hour a t C. The r e s u lt s are l i s t e d in Table 8.

66 55 Table 8. S o lve n t R a tio Test Runs S olvent-to-s R C R a tio S o lu tio n C h a r a c te r is tic s H ig h e st Pressure During Run 7 :2 Homogeneous e a s y -p o u rin g l i q u i d 2:1 Darker and more v is c o u s th a n 7 :2, but s t i l l pours e a s ily 1:1 Very s t ic k y and v is c o u s t a r - l i k e m a te ria l 3: 2 Much more- l i q u i d than th e 1:1 About 1600 About 1600 About 1600 About 1600 Since h ig h v is c o s it y in c re a s e s mass d if f u s io n re s is ta n c e, the 3 :2 so lv e n t-to -S R C s o lu tio n.w a s chosen as optimum fo r a l l. fu tu r e c a t a ly s t t e s ts. The 3 :2 s o lu tio n w i l l be much more c o n c e n tra te d than the 7 :2 s o lu tio n. That should h e lp re a c tio n s and reduce e r r o r in c o n v e rs io n c a lc u la tio n s. Less severe te m p e ra tu re s and re s id e n c e tim e w i l l g iv e a homogenous 3 :2 s o lu tio n, too.. A 3 :2 bomb ru n a t 365 C fo r f iv e m inutes and 775 p s ig i n i t i a l p re s s u re y ie ld e d a p ro d u c t s im ila r to the lo n g e r, more severe ru n a t 420 C fo r o n e -h a lf h o u r.

67 C a ta ly s t S creening Bomb Runs - The r e a c tio n v a ria b le s 56 (te m p e ra tu re, p re s s u re, run tim e, s o lv e n t typ e and amount, and. s o lv e n t- to-src r a t i o ) mere a l l fix e d in th e c a ta ly s t scre e n in g bomb ru n s. The c o n d itio n s o f th e ru n s are l i s t e d belom: L iq u id charge gr t o t a l, in 3 :2 so lve n t-to -S R C r a t io I n i t i a l p re ssu re p s ig H eat-up p e rio d - about I hour Run te m p e ra tu re C The one e x c e p tio n i s CoMo , th e f i r s t one o f the s e rie s. I t mas. run a t Iorner te m p e ra tu re and p re ssu re than the r e s t. Tmenty m i l l i l i t e r s o f p e lle t-rty p e c a ta ly s ts or te n m i l l i l i t e r s o f homogeneous pomder c a t a ly s ts mere used in each ru n. Less pomder c a ta ly s t than p e lle t s mas used to p a r t i a l l y compensate fo r th e g re a te r c o n c e n tra tio n o f a c tiv e c a t a ly t ic m a te ria l in homogeneous pomders. The e n tir e c a ta ly s t must be c o n sid e re d in com paring th e a c t i v i t y o f one c a ta ly s t to th e a c t i v i t y o f a n o th e r. Many fa c to r s in flu e n c e the a c t i v i t y o f a g iv e n c a t a ly s t. F a c to rs in flu e n c in g heterogeneous c a ta ly s ts in c lu d e : - typ e and lo a d in g o f c a t a ly t ic m etals - su p p o rt co m p o sitio n and c r y s t a llin e s tr u c tu r e - su p p o rt pore diam eter and pore volume d is t r ib u t io n - a c tiv e s u rfa c e area The s iz e o f a g ive n p e lle t or. pomder can a lso a f f e c t pore d if f u s io n l i m i t a t i o n o f r e a c tio n r a te. A ll these p r o p e r tie s va ry among the

68 57 c a ta ly s ts te s te d, as l i s t e d in the D e s c rip tio n o f C a ta ly s ts Used, Appendix A. Because the c a ta ly s ts va ry in many re s p e c ts, com parison o f a c t i v i t i e s must be made between s p e c ific c a ta ly s ts. Comparisons between c a ta ly s ts w ith some s i m i l a r i t i e s can be made, b u t e x te n s io n o f g e n e ra liz a tio n s to broad c la s s e s o f c a ta ly s ts may not be p o s s ib le. A ll c a ta ly s t t e s t bomb ru n s are l i s t e d c h r o n o lo g ic a lly be g in n in g on the second page o f Appendix B, Batch Run Data. A g ro u p in g by m etal type o f te s te d c a ta ly s ts i s l i s t e d in Table 9..Most o f the com m ercial c a ta ly s ts were s u lfid e d ; u n s u lfid e d c a ta ly s ts are the two G ir d le r iro n -c h ro m a te s, Cr-0105 in one ru n, the iro n -c h ro m a te on CAS-3QKR s i l i c a alum ina, the p la in su p p o rt CA5-30KH, and a l l the m etal c h lo r id e s. Table 9. Batch C a ta ly s ts Tested by M etal Type Ni-Mo ---Harshaw H T Harshaw HT MSU STK-I Supported - N ic l on C e llit e 410 M etal C h lo rid e s - N ic l^ on CAS-30KR - CdClg on CAS-30KR Ni-W - Harshaw Ni Harshaw N i K e tje n HC-5 - K e tje n 330-E3 M etal C h lo rid e Powders Co-Mo - Harshaw Co-Mo Harshaw Co-Mo Harshaw HT-400 Cr and - Harshaw Cr-0105 Fe-Cr - Harshaw Cr G ir d le r G-3A t a b le ts - Fe-Cr on CAS-30KR O thers' Harshaw W-Q101 Harshaw M o-1201 Harshaw N i-3210 Sn powder CAS-30KR unloaded support

69 58 The c a ta ly s ts are ranked a cco rd in g to d i s t i l l a t e p ro d u c t q u a lit y in the fo llo w in g ta b le. Table 10 l i s t s the batch c a ta ly s ts, d i s t i l l a t e p ro d u c t d is t r ib u t io n, and o v e r a ll d e s u lfu r iz a tio n and d e n itro g e n a tio n. The N i-4401 c a ta ly s t i s l i s t e d t w ic e : the ve ry f i r s t batch run using.7:2 s o lu tio n, and a l a t e r run using the standard 3 :2 s o lu tio n fe e d. Comparison w ith o th e r c a ta ly s ts must be made u sin g th e low s o lv e n t r a t i o r e s u lt s. As can be seen, s u lfu r and n itro g e n rem oval were h ig h e r fo r th e le s s c o n c e n tra te d s o lu tio n. T h is in d ic a te s th a t heteroatom rem oval i s e a s ie r from the l i g h t s o lv e n t f r a c t io n than from the heavy SRC f r a c t io n. N ic lg i s a lso l i s t e d tw ic e : ' the -15^ N ic lg i s on 3M / g r, s u rfa ce area su p p o rt and the 24% N ic lg i s on 43M / g r. s u rfa c e area s u p p o rt. The s u p p o rts are noted in d e t a il in Appendix A. The h ig h s u rfa ce area su p p o rt N ic lg had b e tte r h y d ro c ra c k in g and d e s u lfu r iz a tio n. The f iv e b e s t h y d ro c ra c k in g c a ta ly s ts were: - HT-400, 3% CoO, 15% MoO^ on alum ina - HT-500, 3.2% NiO, 15.5% MoD^ on alum ina - N i , 6% NiO, 19% WO^ on alum ina - K e tje n 330-E3, 6.6% NiO, 19.8% VJO3 on alum ina and - A lc l3 powder

70 59 Table BATCH CATALYSTS - RANKED-BY DISTILLATE CURVE QUALITY P ercent P ercent P ercent P ercent P ercent Pera- C a ta ly s t Naphtha Fuel O il Gas O il Residue DeS ce n t I BP-425 F F F DeN SnCl2. ZH2O G ird le r. G-3 t a b le t ' o. 0 (Fe2O,) C r-0304, u n su l C r-0105, u n s u l Sn powder ' CuCl2. H2D % CdCl2 on s i l i c a alum ina ' W-OIOI o % N ic l2 on 86% s i l i c a CoMo Fe D -C r 0 on s i l i c a alum ina CoMo ' Ni C r s u l f i d ZnCl2, powdered STK-1 ' ' K e tje n HC-5 3 8, K e tje n 330-E3 and CdCl on s i l i c a alum ina n CAS-30KR s i l i c a alum ina M o ' Ni (lo w s o lv e n t r a t io ) G ir d le r G-3A Fe^O, powder

71 60 (T able 10 ~ Batch C a ta ly s ts c e n t1d) C a ta ly s t #Naptha #Fu e l O il #Gas O il ^Residue #DeS #DeN 24# N ic l on s i l i c a alum ina HT O A lc l^ Pomdered 0 25 O K e tje n 330-E N i O HT Ni-4401 (h ig h s o lv e n t r a t i o ) O HT O O

72 61 The fo u r top heterogeneous c a ta ly s ts have co m b in a tio n s o f molyb denum or tu n g s te n prom oted w ith e ith e r c o b a lt or n ic k e l. The p ro d u c t d is t r ib u t io n s o f the f iv e best, c a ta ly s ts v a rie d, though c o n ve rsio n was g e n e ra lly h ig h e s t in th e m iddle or low b o ilin g ranges. Three o f the 2 fo u r p e lle t c a ta ly s ts have s u rfa c e areas o f 200 M / g r ; a l l fo u r are mounted on alum ina. D i s t i l l a t i o n graphs d is p la y in g th e p ro d u c ts o f the two b e st c a ta ly s ts are shown in F ig u re s 8 and 9.. Both have q u ite d if f e r e n t curves when compared to th e 3 :2 s o lu tio n feed curve but both gave o v e r a ll l i q u i d co n v e rs io n o f 25%. The 3 :2 s o lu tio n feed has 64% d i s t i l l a t e s, and the HT-400 p ro d u c t has 73% d i s t i l l a t e s. L iq u id co n v e rs io n was = 25%. The d i s t r ib u t io n o f the 3 :2 s o lu tio n feed in the th re e b o ilin g ranges i s 55% in Naphtha (IBP F ), 9% in Gas O il ( F ).. An example o f the b o ilin g range d is t r ib u t io n o f c o n ve rsio n i s given..here fo r th e K e tje n HC-5 p ro d u c t. I t b o ile d as fo llo w s : Naphtha range -.'56% Fuel O il range - 12% G a s.o il range - 4% T o ta l 72% d i s t i l l a t e s O v e ra ll co n v e rs io n = *^77^' = 22% 36 to liq u id s C onversion to Naphtha C onversion to Fuel O il C onversion to Gas O il : 3% (rounded o f f ) = 8% =11% -..22%

73 62 Tem perature in Feed P roduct Volume P ercent D i s t i l l e d F ig u re 8. D i s t i l l a t i o n o f HT-400 P roduct vs. 3 :2 S o lu tio n Feed

74 63 P roduct Feed Temperati Volume P ercent D i s t i l l e d F igure 9. D i s t i l l a t i o n o f HT-50D P roduct vs. 3 :2 S o lu tio n Feed

75 64 T ables 11 and 12 shorn the f iv e c a ta ly s ts which accom plished the most s u lfu r and n itro g e n removal.- ve ry w e ll in h e te ro a to m.re m o v a l. As a group, the m etal c h lo rid e s d id A lso, note th a t fo u r o f the f iv e best h y d ro c ra c k in g c a ta ly s ts are a lso in a t le a s t one o f T ables 11-and 12. Table 10 shows t h a t none o f the "b e s t" h y d ro c ra c k in g c a ta ly s ts were poor heteroatom rem overs. Good h y d ro c ra c k in g seems to c o r r e la te w ith good heteroatom rem oval. The heteroatom rem oval data p o in t to one g e n e r a liz a tio n th a t was a lso in d ic a te d in th e l i t e r a t u r e search: n itro g e n i s c o n s is te n tly harder to remove than s u lf u r. Only one c a t a ly s t, A lc l^, gave a g re a te r p e rc e n t re d u c tio n in n itro g e n than s u lfu r (48% de-n vs, 31% d e -S ). Comparisons o f Some Ni-W t Ni-M o, and Co-Mo C a ta ly s ts - Three Ni-W c a ta ly s ts (K e tje n HC-5, K e tje n 330-E3, and Harshaw N i-4303) have s im ila r m etal lo a d in g s, b u t d iffe re n c e s u p p o rts and v a ry in g r e s u lt s. U n fo rtu n a te ly, each was on a d if f e r e n t s iz e p e lle t, so th e p o s s i b i l i t y o f pore d if f u s io n e f f e c t s makes g e n e r a liz a tio n s u n c e rta in. The Harshaw c a ta ly s ts HT-IOO and HT-500 are both about 3% NiO, 16% MoO on h ig h area s u p p o rt, but HT-IOO js - l / l 2". diam eter and s ilic a t e d ; th e HT-500 i s T/8" and ju s t alum ina. S u r p r is in g ly, ^ n itro g e n and s u lfu r rem oval fo r the two d i f f e r very l i t t l e. r the s ilic a t e d, s m a lle r p a r t ic le.. HT-400 (3%-CoO, L iq u id co n v e rs io n i s le s s fo r ' 2 15% MoO3 on alum ina, 220 M / f l r, and average

76 65 Table 11. BEST BATCH DESULFURIZATIONS ZnCl2 powdered 65^ HT-400!>% CoO,. 15^ MoO3 on alum ina 47# CuCl2. H2O f in e c r y s ta ls 46# HT # NiO, 15.5# MoO3 on alum ina 45# K e tje n 330-E3 6.6 # NiO, 19.8# WO3, 1.2# SiO 43# on alum ina Table 12. BEST BATCH DENI TROGENATIONS A lc lv powdered 48# ZnCl2 powdered. 35# K e tje n HC # NiO, 21# WO3 on alum ina 28# HT-400 3# CoO, 15# MoO3On alum ina 23# CuCl2. HJD f in e c r y s t a ls 22#

77 66 pore diam eter 100 fl ) i s s im ila r to Co-Mo 0603 ( CoO, 12/^ MoO^ on alum ina, 166 M ^/g r, 106 fl ). Both are l/ 8 " p e lle t s, and the high s u rfa c e area HT-400 d id about 1/3 b e tte r in heteroatom rem oval. M etal C h lo rid e s - The unsupported c h lo rid e s (Sn, Cu, Z n,. and Al). a l l m e lt below the r e a c tio n te m p e ra tu re ; the supported c h lo rid e s (N i and Cd) do not m e lt below re a c tio n te m p e ra tu re. A lc l^ was the b e st o f a l l the c h lo r id e s fo r c ra c k in g and best o f a l l c a ta ly s ts fo r d e n itr o - g e n a tio n (48%). The supported c h lo r id e s were f a i r l y average c a ta ly s t. in both c ra c k in g and heteroatom re m o v a l; the unsupported c h lo rid e s ranged from poor to good h y d ro c ra c k e rs, b u t as a group t h e ir h e te ro - atom rem oval was ve ry good. Copper and z in c c h lo rid e were both among the b e st f iv e n itro g e n and s u lfu r rem overs. Fe and Fe-Cr C a ta ly s ts - The f i r s t iro n -c h ro m a te te s te d, G ir d le r G-3A powder, gave encouraging r e s u lt s : 22% o v e r a ll c o n v e rs io n, 14% de-s, and 19% de-n. But th e p e l l e t form o f G-3A proved in e f f e c t iv e, w ith low c o n ve rsio n and no heteroatom rem oval. Pore d if f u s io n l i m i t a tio n s a n d /o r low er s u rfa c e area p ro b a b ly caused the poor r e s u lt s. The in te rm e d ia te area s ilic a - a lu m in a im pregnated w ith Fe-Cr d id not g iv e good r e s u lt s e it h e r. ' O ther in v e s tig a to r s a t Montana S ta te had b e tte r r e s u lt s w ith u n s u lfid e d chrom ates on su p p o rt than w ith the same c a ta ly s ts s u lfid e d., Two u n s u lfid e d Cr c a ta ly s ts d id n o t g iv e good r e s u lt s w ith the SRC 3 :2 fe e d s to c k, though. The e f f e c t o f s u lf id in g was te s te d by ru n n in g e

78 67 one o f the Cr c a ta ly s ts ag a in, o n ly s u lfid e d. H yd ro cra ckin g and heteroatom rem oval a l l in c re a s e d w ith s u l f i d i n g. O ther E ffe c ts - The unloaded su p p o rt CAS-30KR s i l i c a alum ina, 43 M / g r, was te s te d to help determ ine the e f f e c t o f s u p p o rt. The p la in su pport cracked f a i r l y w e ll ( i t i s midway in the t a b le ), b u t had n e g -l l i g i b l e heteroatom re m o v a l. One in t e r e s t in g s id e lig h t the p la in su p p o rt p ro d u c t was th e o n ly c a t a ly s t which had le s s d i s t i l l a t e in the Fuel O il range than the S o lu tio n Feedstock. C o nclusion s o f Bomb T e s ts ' - Batch screening -of the c a ta ly s ts showed s e v e ra l th in g s : - The b e s t heterogeneous h y d ro c ra c k e rs were co m bination s o f Mo or W promoted w ith N i or Co on h ig h area alum ina. - The b e st m etal c h lo r id e c a ta ly s ts were A lc l,, ZnCl, and N ic l2. Since HT-400 (Co-Mo on alum ina) was in each o f the th re e c a te g o rie s o f "b e s t" c a ta ly s ts, i t was chosen fo r co n tin u o u s mode t e s t in g. K e tje n HC-5 (Ni-W or alum ina) was chosen f o r c o n tin u o u s te s tin g since i t had the b e s t d e n itro g e n a tio n o f the heterogeneous c a ta ly s ts. A lc l^ was not chosen f o r t e s t in g because o f c a ta ly s t re c o v e ry problem s in h e re n t w ith m etal c h lo r id e s, C ontinuous Runs The e ff e c t s o f te m p e ra tu re, l i q u i d h o u rly space v e lo c it y, and hydrogen f l o w - t o - o i l r a t i o were in v e s tig a te d, using the co n tin u o u s

79 re a c to r d e s c rib e d in the M a te r ia ls, Equipment, and Procedure S e c tio n. 68 See Appendix C, C ontinuous Runs: C onversion of- SRC to L iq u id, fo r a l l c a ta ly s ts, run c o n d itio n s, ond p ro d u c t r e s u lt s o f the continuous, ru n s. Fixed v a r ia b le s mere: - p re ssu re p s ig - feed - 3 :2 s o lu tio n fe e d, prepared in bombs a t 400 C - Amount o f - 70 ml. c a ta ly s t C a ta ly s ts te s te d mere HT-400 ( CoMo on a lu m in a ), K e tje n HC-5 (N i - W on a lu m in a ), MSU STK-7 ( Co-Ni-Mo on a lu m in a ), and MSU STK ( Co-Ni-Mo on a lu m in a ). Complete c a ta ly s t d e s c r ip tio n s are l i s t e d in th e second p a r t o f Appendix A. The p ro d u c ts from the co n tin u o u s runs mere b e tte r than the p ro d u c ts from th e Bomb ru n s. L iq u id c o n v e rs io n s fo r HT-400 ranged from %.compared to 25% in th e bomb; l i q u i d co n v e rs io n fo r K e tje n HC-5 mere 44 to 61%, compared to 22% in the bomb. Summary o f R e s u lts from C ontinuous Runs -.Of th e.fo u r c a ta ly s ts te s te d, th e b e st d e s u lfu r iz a tio n, 88%, mas done by HC-5 (N i-w ). The b e st d e n itro g e n a tio n, 64%, mas o b ta in e d by HT-400 (CoMo). The tre n d mas observed th a t n itro g e n i s more d i f f i c u l t than s u lfu r to remove. A ll o f th e c a ta ly s ts te s te d gave a 55-60% c o n v e rs io n to liq u id s d u rin g at. le a s t one se t o f c o n d itio n s te s te d. O v e ra ll l i q u i d c o n v e rs io n tends to :

80 69 1) in c re a s e w ith in c re a s in g te m p e ra tu re 2) be b e st a t LHSV from 1 to 2 3) maximize a t.10,000 SCF/bbl fo r LHSV = I and 445 C. P roduct d is t r ib u t io n s va ry in no d e fin a b le tre n d w ith te m p e ra tu re, LHSV, and H^ r a t i o. ' H eteroatom rem oval tends to : 1) in c re a s e as tem p e ra tu re in c re a s e s 2) decrease as LHSV in c re a s e s 3) maximize about 10,000 SCF/bbl fo r LI-ISV = I and 445 C. H yd ro cra ckin g - The e f f e c t s o f te m p e ra tu re, l i q u i d h o u rly space v e lo c it y, and H^ r a t i o on h y d ro c ra c k in g are i l l u s t r a t e d in th e fo llo w in g th re e fig u r e s. u sin g STK-7. F ig u re 10 d is p la y s 'th e e f f e c t o f tem p e ra tu re on h y d ro c ra c k in g, The 420 C p ro d u c t has le s s l i q u i d co n v e rs io n and i s more 0 s im ila r to th e s o lu tio n feed than i s th e 445 C p ro d u c t. F ig u re 11 shows th e HT-400 p ro d u c t d i s t i l l a t i o n s fo r th re e d if f e r e n t space v e lo c it ie s. LHSV=2 p ro d u c t has th e most naphtha range m a te r ia l, b u t th e LHSV=I p ro d u c t has th e most o v e r a ll c o n v e rs io n. The LHSV=3 p ro d u c t i s n o t much d if f e r e n t than the fe e d s to c k, b u t somewhat extended. From th e Appendix, i t can be seen th a t LHSV=2 g iv e s the most naphtha fo r HC-5 a ls o, b u t n o t fo r the o th e r two c a ta ly s ts.

81 C Temperature in F eed Volume P ercent D i s t i l l e d F ig u re 10. E ffe c t o f Temperature on H yd ro cra ckin g by 5TK-7

82 Temperature in 300- LHSU 445 C;. 10,000 S C r/b b l Volume P e rcent D i s t i l l e d F ig u re 11. E ffe c t o f LHSU on H yd ro cra ckin g by HT-400

83 72 F igure 12 shows th e e f f e c t o f H '^ - to - o il r a t i o on h y d ro c ra c k in g by STK The 10,000 S C F /b b l.p ro d u c t shows improvement over the 5,000 SCF/bbl p ro d u c t. In c re a s in g th e r a te to 15,000 SCF/bbl in the HC-5 and STK-7 run d id not im prove th e l i q u i d c o n ve rsio n nor naphtha range c o n v e rs io n. L im ite d data i s a v a ila b le on th e e f f e c t o f r a t i o a t a.space v e lo c it y o th e r than I. Appendix A re v e a ls th e best o v e r a ll co n ve rsio n fo r STK a t LHSV=2 and 445 C came a t a r a t i o o f 7500 SCF/bbl. H eteroatom Removal - E ig h t fig u r e s d is p la y th e e ff e c t s o f te m p e ra tu re s, LHSV, and r a t i o on s u lf u r and n itro g e n rem oval. N itro g e n and s u lfu r rem oval both in cre a se d w ith te m p e ra tu re. F ig u re s 13 and 14 d is p la y data p lo ts o f s u lfu r and n itro g e n rem oval versus te m p e ra tu re fo r two d if f e r e n t c a ta ly s ts. Both d e s u lfu r iz a tio n and d e n itro g e n a tio n in c re a s e as L-HSV decreases. F ig u re s 15 and 16 d is p la y th e te n d e n c y. One data p o in t in F ig u re 16 does n o t fo llo w the tre n d. V a r ia tio n o f e ff e c t s o f H - t o - o i l r a t i o is d is p la y e d in F ig u re s 2 17 and 18 fo r heteroatom rem oval a t LHSV=I and 445 C. D e s u lfu r iz a tio n f o r HC-5 and STK-7 peaked or le v e le d o f f a t H^ r a t io s o f about M SC F/bbl. S u lfu r rem oval d id not va ry much fo r STK D enptroge nation r e s u lt s c o n f lic t e d f o r HC-5 and STK-7, as seen in F ig u re 18. I qw H r a t io s. The STK run was made to determ ine th e e f fe c t o f A p o in t a t 10,000 SCF/bbl was o b ta in e d fo r both the f i r s t

84 Z T Feed F- 400 Volume P ercent D i s t i l l e d F ig u re 12. E ffe c t o f H R a tio on H yd ro cra ckin g by STK

85 74 D e s u lfu riz a tio n O HT-400 i STK-7 Tem perature Figure 13 Percent Desulfurization vs. Temperature

86 75 % D e n itro g e n a tio n HI-400 STK C 425 C 450 C Tem perature Figure 14 Percent Denitrogenation vs. Temperature

87 76 O HT-400 HC-5 STK-7 D e s u lfu riz a tio n 75 _ STK L iq u id H o u rly Space V e lo c ity ( V o l, Feed per h o u r/v o l. C a ta ly s t) F ig u re 15. P ercent D e s u lfu r iz a tio n vs. LHSV

88 O ht-400 O HC" 5 75 _ A STK-7 % D e n itro g e n a tio n 50 _ x STK BBL 25 - L iq u id H o u rly Space V e lo c ity (V o l. Feed per h o u r/v o l. C a ta ly s t) Figure 16. Percent Denitrogenation vs. LHSV

89 78 HC-5 STK-7 STK % D e s u lfu r iz a tio n H2 to O il R a tio, ^ "'2 F ig u re 17. P ercent D e s u lfu r iz a tio n vs. Hydro g e n -to -O il R a tio

90 79 HC-5 5TK-7 STK % D e n itro g e n a tio n SP F H H2 to O il R a tio, ^ 2 F ig u re 18. P ercent D e n itro g e n a tio n vs. Hydrogen - t o - O il R a tio

91 80 X LHSV=I $ LHSV=2 D e s u lfu r iz a tio n C a ta ly s t STK ; 445 C to O il R a tio F igure 19. P ercent D e s u lfu r iz a tio n vs. Hydrogen to O il R a tio a t d if f e r e n t l i q u i d h o u rly space v e lo c it ie s

92 81 LHSV=I C a ta ly s t STK ; 4A5 C to O il R a tio Figure 20 P ercent D e n itro g e n a tio n vs. Hydrogen to O il R a tio a t d if f e r e n t l i q u i d h o u rly space v e lo c it ie s