Bench scale studies on retorting, oil recovery, and flow of heated shale

• NOTICE

The information contained in this report is regarded as con­ fidential and proprietary. It is provided subject to the provisions regarding confidential, proprietary information contained in the Research Agreement among the Participating Parties.

HOBIL OIL CORPORATION RESEARCH DEPARTMENT

TECHNICAL ~1EI>10RAtmur1 NO. 67-8

BENCH SCALE STUDIES ON RETORTING, OIL RECOVERY, AND FLOW OF HEATED SHALE

AlNIL POINTS OIL SHALE RESEARCH CENTER Rifle, Colorado

"April 13,1967

Author~ Approval:

D. Po Cotrupe _{I. -}

_A/.

(r~a/)"J-,,,/:A-. br~

_/0

R. H. Cramer /

• 2 ­

The primary object of the Anvil Points Oil Shale Research Center TECHNICAL I~MOruu~DUM is to advl!~ authorized personnel employed by the Participating Parties ) that various

activities are in progress or that certain significant data have been obtained within the Research Center.

These TECHNICAL MEMORANDA have been prepared to provide rapid, on-the-spot reporting of research currently in progress at Anvil Points. The conclusions drawn by project personnel

are tentative and may be subject to change as work progresses. The TECHNICAL r{EMORANDA have not been edited in detail.

(1) Mobil Oil Corporation, Project Manager Continental Oil Company

Huwble Oil and Refining Company Pan American Petroleum Corporation Phillips Petroleum Company

• •

- 3 ­

BENCH SCALE STUDIES ON RETOR.TIl"G, OIL RECOVERY, AND FLOW OF HEATED SHALE

TABLE OF Cm'TF.NTS

-

Page I. Introduction. • • • • • • • • •

.

. .

.

.

. .

.

5

II. Conclusions. • • • • • • • • • • • • • • • • • • • 6 A. Fischer Retorting Studies. • • • • • • • • • • 6

B. Recovery of Oil From Oil/Fines Mixtures. • • • 6 C. Shale Flow Studies ~ith Heated Shale • • • • • 6 III. Detailed Discussion • • • • • • • • • • • • • • • • 7

A. Fischer Retorting Studies. • • • • • • • • • • 7 1. Retorting at Atmospheric Pressure in the

Presence of Fines • • • • • • • • • • 7

2. Retorting at Reduced Pressure. • • • • • 9

3. Retorting at 10 mm Hq in the Presence of Fines • : • • • • • • • • • • • • • • • • 12 4. Effect of Primary Condenser Temperature • 12 B. Recovery of Oil From Oil/Fines ~~ixtures. • • • 13 1. Atmospheric Pressure Distillation • • 13 2. Reduced Pressure Distillation • • • • • • 15 C. Retorting Studies in the Mini-Retort • • • • • 18 1. Evaluation of the ~ini-Retort • • • • • • 18

2. Reduced Pressure Operation • • • • • • • • 20

3. Relative Rates of Formation of Liquid

and Gaseous Products. • • • • • • • • 21 4. Gas Flow Studies • • • • • • • • • • • • • 21 D. Shale Flow Studies With Heated Shale • • • • • 23

TABLES 1 Effect of Fines on Oil Yield

(10 mm Hg) Assay Hg

Assay Retorting Results

2 Comparison of Standard Fischer Assay With Vacuum

3 Effect of Retorting in the Presence of Fines @ 10 mm 4 Effect of Primary Condenser Temperature on Fischer

5 Recovery of Oil From Oil/Fines Mixtures

6 Recovery of Oil From Oil/Fines r,~ixtures (Pressure

=

10 mm Hg)

7 Comparison of Fischer J\.ssay and Run JI!R-7 Data 8 Effect of Pressure on f;Uni-Retort Pesul ts

9 Effect of Temperature on Flow of Standard Quality Shale

10 Effect of Temperature on Flow of Spent Shale

'!"" 4 ­

FIGt:RES

1 Distribution of Raw Shale Organic Carbon as a Function of Oil Yield

2 Yield Loss as a Function of Fines Type and Shale Richness

3 Oil Yield as a Function of Total Pressure

4 Distribution of Raw Shale Organic Carbon as a Function of Total Oil Yield

5 Bench Scale Experiment 2

6 Bench Scale Experiments 2A and 2P 7 Fixed Bed Petort

8 Comparative Time-Temperature Relationships 9 Production Rates at Atmocpheric Pressure 10 Production Rates at 10 mm Hq Pressure 11 Block Diaqram of Mini-Retort Process

12 Sketch of Mini-Retort as Used in Heated Shale Flow Studies

13 Effect of Temperature on Shale Flow, Oil Loss, and Benzene Solubles Concentration

14 Effect of Temperature on Shale Flow

- 5 ­

BENCH SCALE STUDIES ON RETORTInG, OIL RECOVERY, AND FLO,,"7 OF HEATED SHALE

Ie INTRODUCTION

A bench scale experimental program was initiated in April, 1966 to study various aspects of the retorting process. The overall objectives of the program were to broaden understanding of the Gas-Combustion Retorting process and to develop potential schemes

for improving oil yield and/or retort operability. The primary objective of the program was to establish ultimate retort yield level and to explore potential yield-loss mechanisms which have limited the pilot retort yield to a level slightly above 90 vol.%

Fischer Assay. In this connection, a fairly extensive study was

made to determine the effect of oil/fines interactions on oil yield. The purpose of this study was to explore the theory that oil and/or fines refluxing in the Gas-Combustion Retort is a

potential yield loss mechanism. In addition, the effect of

retorting pressure on oil yield was investigated. Finally, a limited study was made to determine the effect of temperature on the flow characteristics of shale. The latter study was made to provide some insight into shale flow problems which can

develop in the pilot retort at normal operating conditions. The bench scale program was terminated in February, 1967. This report is a detailed summary of all bench scale work completed by this date.

- 6 ­

II. CONCLUAIONS

A. Fischer Retorting Studies

1. Increased oil yields are obtained when retortincr at

reduced pressures~ however, product recovery is more

difficult at these conditions and more efficient re­

covery systems are required to realize t~e yield

advantaqe.

2. Fines promote cracking when retorting at either

atmospheric or reduced pres~ure. Yield losses re­

sult unless the efficiency of the recovery system is increased to recover the lighter products 'formed.

3. The degree of cracking (or yield loss) is a function

of oil/fines ratio, i.e. at a constant fines level,

cracking increases ~dth decreasing shale richness.

4. Fischer retort soent shale fines exhibit hiqr.er

activity with respect to crack"inq than either raw or pilot retort spent shale fines.

B. Recovery of Oil From Oil/Fines 1'IIixtures

Oil losses due to crackinq are obtained when recovering oil from oil/fines mixtures at either atmospheric or reduced pressure.

C. Shale Flow ~tudies with Heated Shnle

1. The rate at ~,'hich shale flows decreases with increasing

temperature reaardless of shale quality.

2. A marked drop in the rate of raw shale flow is ob­

served at temperatures above 650 F.

3. Static or slow moving srale beds ten~ to form shale

agglomerates (or Il cohesive masses") at the 800 F

level.

4. :Rich shale flo,,,-,s more slowly than lean shale at all

temperature levels, due primarily to differences in particle shape.

5. The flow of rich shale is more temperature sensitive

than that of lean shale. ,fj

- 7 ­

III. Detailed Discussion

A. Fischer Retorting Studies

1. Retorting at Atmospheric Pressure in the Presence of Fines

This series of experiements ~..,as part of an overall program designed to investigate potential yield losses resulting from oil/fines interactions in the retortina and oil revaporization processes in the ras Comrustion Petort. The specific objectives of this series ~rlas to determine the effect on oil yield ""hen retorting a standard ra'-t shale charge in the oresence of fines. The tests were conducted in the laboratory Fischer Assay equipment and the standard Fischer Assav procedure Nas follo,,,ed. The following materials ~"ere preoared 'for use in these tests.

a.

a -

14 mesh raw shale (RS). b. -65 mesh raN shale (RRD).

c. -65 mesh spent shale prepared by crushinq spent shale from the pilot retort (SSR).

d. -65 mesh spent shale prepared by crushing spent shale from the Fischer retort (SSFJ\~).

The

a -

14 mesh ra,., shale was the base material retorted in all tests and the retort charqes were preoared bv mixinq this material with one of the

-65

mesh materials.

In the initial experiments, the Fischer Assay of the base material \\Tas 28.0 gallons oer ton ~\Thile that of the -65 mesh raw shale 'rras 27.? aallons per ton. Bot'l, pilot and Fischer retort spent shales tl7ere studie ~ because the material from the pilot retort "'las subjected to hiqr tem­ peratures and burning ~,rhile passing throtlah the combustion

zone, whereas t~at from the Fischer retort is heated only

to .#....,.. 935 P. It "Tas thought that this difference in

exposure might have a significant effect on the activity of the materials. Chemically, the spent shales differed mainly in organic carbon content, the Fischer retort material containina ..."..,,'. 15% more orqanic carbon than

that from the pilot retort. ~his is to be exnected since the coke formed in the Fischer retort is not burned.

The mixtures studied "rere prepared bv m1Xl.ng 75 q of

the base materials Nith 25 q of one of the -65 mesh materials. Fxtreme care '~'as exercised in each case to insure uniform dispersion of the fines throughout the mixture. The n'ixtures ,.rere charqed to the Fischer

.... 8 ­

retorts and retorted according to the standard Fi~cher

Assay procedure. In addition, samples of the base

material and the -65 mesh raw shale ~'1ere retorted

separately to establish the oil content of these materials. The data obtained on the base material

served as a control for comparison of results ob­

tained on the mixture9. Also, in order to allow m~re

thorough evaluation of these studies, all materials charged to the retorts anli. the spent shale residues remaining after retortinq were analyzed for total car­ bon, hydrogen and mineral C02 content. The snent shale residues were also analyzed for benzene extractables. All runs were made in duplicate and the precision of results ,,'as excellent.

The data generated by these studies indicate that the

presence of either raw or spent shale fines decreases the Fischer Assay yield. The data show further th8.t

the most significant decrease (rv 4%) occurred '-Then

retorting in the presence of spent shale fines from

the Fischer retort. Py comparison, tests "rith spent

shale fines from the pilot retort or with raw shale

fines show a yield loss of ~I 1.5%. These data appear

to confirm the hypothesis that the activity of snent shale which has been retorted only is different than that which has been exposed to hiqh temperatures and

burning. It is highly probable that the nature of the

surface of the spent shale from the pilot retort is sig­

nificantly altered by sintering in the co~bustion zone ano

perhaps the difference in act.ivity or~p.rved in these tests

is proportional to the difference in surface area. The

oil recovery data obtained ~roro these studies are tabu­

lated belo~." ~

TAPLE 1

FFFFCT OF FntES ON OIL YIFLD

RS Assay: 28 gallon per ton

rJfixture Oil Yield, Vol % Pischer Assay

RS Only (Control) _100.0(1)

RS + RPD 98.6

RS + SSR 98.6

RS + ~SFAR 96.1

(1) Corrected for oil in RSD.

In order to substantiate t~e recovery data shm..rn a.bove,

a detailed analysis was made of raw shale orqanic car­

- 9 ­

to maJ-e two assumptions to complete this analysis due to the lack of gas recovery facilities on the Fischer A.ssay apparatus. A lOOt material balance was assumed. That is, all material unaccounted for ~y the recovered oil and the spent shale ",as assumed to be offgas.

Also, the gas composition was assumed to be the same as the average of 16 Fischer Assay runs reported in the Bureau of Mines ~eport of Investigatiol'l 4825. Orqanic carbon balances calculated on this hasi~ averaged"99.8% with an average deviation of ± 0.4%. The calculated dis­

tribution data correlated extremely ~I,'ell with yield as shown in Figure 1. This correlation suggests that the assumptions made \l7ere not unrealistic and the data offer strong support for the reported recoveries.

These results prompted thought concerning the role of shale richness in such a studv. Pe~errinq to the data obtained during the shale richness studv in ~etort ~o. 1,

yield increased ~~ith increasing shale r:i.chness. HO'flever, quench data obtained from those runs sho't-1ed that the fines concentration also increased with richness. On the

basis of these data is was hypothesized that the ratio

of oil to fines was the controlling factor in the yield loss rather than the absolute concentration of fines.

To evaluate this hypothesis, bench scale experiments ,~ere made retorting rich (37 gallon per ton) and lean

(20 gallon per ton) shales in the presence of raw and spent shale fines. The experimental procedure was iden­ tical to that used for the 28 gallon per ton shale and the results were similar in that the most significant yield losses occurred when retorting in the presence of Fischer retort spent shale fines. Equally significant is the trend of increasing yield loss with decreasing shale richness exhibited by the data for all cases as shown in Figure 2. These results strongly support the hypothesis stated above.

2. Retorting at Reduced Pressure

In order to fully evaluate the significance of the results obtained when retorting in the presence of fines, it was necessary to also determine the effect of pressure since the oil partial pressure in the pilot retorts is in the order cf 10 rom Hg. Establishment of the effect of pres­ sure on yield would allow a more intelligent evaluation of the results obtained from a study of oil/fines inter­ actions at reduced pressure. Therefore a series of experi­ ments were made to determine the effect of pressure on oil yield when retorting a standard raw shale charge.

FIGURE 1

DISTRIBUTION OF RAW SHALE ORGANIC CARBON AS A FUNCTION OF OIL YIELD

Assumptions:

1. 100% overall material balance.

2. Gas composition same as average of 16 runs reported

in Bureau of Mines R.I. 4825.

75 74

I

.,

_{. i} r

---­

'N 'S?E t-!T S}-l ALE -

! .

I

I

i ',' i 1 . , . '-= • -­ I " ' . I f 20 ...:. ---­

-i---,·___

i. - ----:--:T

----~-I-:···-'----I----:----d

I I - t;j. '­ - I .. - I .. 19 ; 1-·! ; , - ' El I -,­

I ..

.--' -""- .,--- (---:-.--- .-' - - ­ --1--- -:-- ­ [D , 18

J

I

\ ' 1 ; :

'

I

• 1 • 9 8 7 92 \H ---­ ~ --~ ~

_.._.i...

_'

J __..

L--12.l-J-...

_{l ' ' !}

L--,

'~·:·-··.I-·.:l-i·'-·:·I'·'!l·11 . I ... ­ I - . , ' : . . I. f · ... 1 , . . . • . , ' ! .... _.. :.

\~

.... ----;---..

--'1"--

.@~.-'~-

.. L:­ __

:_:--I,

;

-I

I­ I . ­ I ' - , I -1 -! ­ - !' .. [ ' - I - f _'.. 1 .; - [ - i i i -\ ~ ..., rq:;o1, 94 - ~~lO....-~. .IIII'~....- - - ­ _ _ _ _~_-1 • ! , • I ! 96 98 100

Yield, Vol.' Raw Shale Assay

DPCotrupe 4/5/67

102 100 98 96 94 Legend: -FIGURE 2

YIELD LOSS AS A FUNCTION OF FINES TYPE AND SHALE RICHNESS

o -

Raw Shale Fines

m -

Pilot Retort Spent Shale Fines

A - Fischer Retort Spent Shale Fines

Note: Standard Raw Shale/Fines = 3/1 By Weight

I I r ! w ( . "r " _' "I _{" I} I " i I J

i

I

,__,I

" ~.*.-~----~.-- -~·1 , I " 1 92 1­_ _ _ _ _ _- - - ­...,---~--.---~ _ __

I

16 20 24 28 32 36

"Raw Shale Fischer Assay, Gallons Pe~ Ton

DPCotrupe 4/5/67

- 10 ­

In preparation for these experiments, two Fischer retorts were modified for vacuum operation. The modification in­ volved machining of the top and cap of the retort together to provide a flat sealing surface, and fabrication of a full cap, which was bolted to the top of the retort and sealed with an asbestos gasket. This design proved to be very satisfactory and all of the reduced pressure runs

were made in these units. In addition to the modifications made to the retort, the vent systems of the vacuum retorts were equipped with -80 C dry ice traps in order to recover light ends which might escape the 0 C condensing system in the standard Fischer Assay apparatus.

The initial experiments were designed to provide a direct comparison of conventional Fischer Assay with vacuum assay. The vacuum assay runs were made at 10 mm Eg. The compara­ tive experiments produced very stimulating results. The data showed that when retortinq 28 gallon per ton shale at 10 mm Hg, oil product equivalent to 94 Vol % Fischer Assay was collected in the 0 C condensing system of the standard Fischer Assay apparatus. However, additional liquid product was collected in -80 C dry ice traps down­ stream of the 0 C system. Yields up to 116 Vol % Fischer Assay have been calculated on the total oil recovered. The specific gravity of the oil in the 0 C trap averages 0.956 compared with 0.916 for typical Fischer Assay oil. It was not possible to completely characterize the oil collected in the -80 C trap since significantly large losses occurred during the transfer operation. Some of the oil was lost as a residual film in the trap and an additional amount was lost by vaporization of the light ends fraction. However, reasonably accurate measurements were made of the total liquid product weight and of the water content. The oil weight was then obtained by dif­

ference. Specific gravity measurements of the recovered oil averaged ft₄ 0.86. Therefore, for the purpose of yield

calculations, an estimated specific gravity of 0.8 was assumed for the total oil fraction collected in the -80 C trap.

Comparative data obtained in these runs are summarized in Table 2. Runs A through D were preliminary shakedown runs and development of the recovery system and operating procedures continued through Runs 6 and 7. Good measure­ ments of the oil and water fractions collected in the

-80 C trap were obtained for the first time in Run 8. These measurements were obtained routinely in all succeeding

reduced pressure runs. Run 10 was a standard Fischer Assay run except that a dry ice trap was installed down­ stream of the 0 C condensing system, providing a recovery system identical to that used in the reduced pressure

e

TABLE 2

_e

COMPARISON OF STANDARD FISCHER ASSAY WITH VACUUM (10 mm Hg) ASSAY

Run # System Product Distribution, wt % Oil S. G. Oil Recoverx

oiI

H2O SS Gas + 60Q/600 F 00 _{C Total}

0

0 _C

_-80

0 _C

₀

0 _C

_-80

0

_C

_{Loss 0}0 _{C wt % Vol %FA wt % Vol %m.}

A, B Vac. 10.1 (1) 0.3 (1) 83.1 6.5 0.953 95.3 91.7 (1) (1)

c,

D Std. 10.6 1.2 86.1 2.2 0.916 (--100 (2)---:l> 6A, B Vac. 11.0 2.7(3) 0.2 84.3 1.9· 0.955 100.0 95.2 110(4) ( 100(2)~ 6C, D Std. 11.0 1.5 85.7 2.1 0.910 7A, B Vac. 10.6 2.7(3) 0.1 85.2 1.5 0.960 100.0 95.3 110.(4) 1C-F Std. 10.6 1.4 86.0

o

2.0 0.917 ( 100(2) ) 8A, B Vac 10.4 2.2 .0.2 1.5 84.8 0.9 . 0.957 97.2 93.3 118.0 116(5) lOA, B Std 10.9 0.0 1.5 0.0 85.4 2.2 0.919 ( 100 (2) ) _{100.0 100} (1) Not Obtained (2)By Definition (3)Oi1 Plus H2O (4) Estimated

(5)Ca1cu1ated using estimated S. G. of 0.8 for oil in -BOo C trap.

DPCotrupe 8/12/66

... 11 ­

in the -80 C trap indicating that the comparison of

standard Fischer Assay with vacuum assay is valid despite the difference in recovery systems.

It is concluded from these data that increased yields are obtained at reduced pressure: however, product re­ covery is more difficult and it is necessary to increase the efficiency of the recovery system to realize the potential yield advantage. Further, since the estimated oil partial pressure in the pilot retorts is in the order of 10 mm Hg, it is thought that perhans the retorting process in its current state of development is capable of generating yields comparable to those obtained in the bench scale studies. The fact that yields above the 92% level have not been obtained may be due to the inefficiency of the conventional recovery system for recovering

naphtha (or naphtha-like) vapor produced in the process. This position is strongly supported by the recycle gas analysis work. Analyses of the recycle gas stream con­ sistently show hydrocarbon concentrations equivalent to 20 to 25% yield in this stream.

The potential yield advantage indicated by these results generated considerable interest in the requirements for recovering naphtha from the gas stream. As a result, a program was initiated to devise and evaluate potential schemes for recovering this material. Schemes to be con­ sidered in this program included conventional and uncon­ ventional recovery schemes and/or process changes. In this connection, bench scale studies were carried out at 100 and 10 mm Ug to firmly establish the effect of pressure on oil yield and product recovery over a practi­ cal range of retort operating pressure.

The experimental procedure for these studies was identi­ cal to that followed in the 10 mrn Hg runs. Oil yields based on the 0 C condensate were slightly higher than

100% Fischer Assay at both pressures and the total yield averaged 113 and 109% Fischer Assay at 40 and

100

mm Hg respective1y. The experimental data are sho,m in Figure

3. It had been anticipated that increased oil yields would be obtained at reduced pressure because the lower pressure would facilitate distillation of the retorted

oil, thereby reducing the degree of cracking which occurred. The data offer strong support for this theory, especially in the range from atmospheric pressure to 40 mm Hg where a linear relationship between yield and log of pressure is obtained. The deviations observed at 10 rom Hg are believed to be due to the difficulty of recovering the

light ends product fraction at this low pressure. One indication is the shift in product distribution between the 0 C and the -80 C condensing systems. Further, the

•

•

e

_{FIGURE 3}

_e

OIL YIELD AS A FUNCTION OF TOTAL PRESSURE

120

tc~-t

1

_111_1111111_11_ I

_11_ _

11111_1

115

_1111111_1111...1_111_11

.

110

1 1 1 1 1 1 1 _ 1 1 _ ­

_1111._ _11"_ _1_1

.'

~:j:::

-A

IIIIIII_IIIIII_ _

1111_1

105

.IIIIPII

_lIlIlIllIIiI....

IIiIiiiiii...

IWlI

::

100

_ _1111111111

11111111._111

1i1l1_1I1_1I111I1I_ _ _

95

_ _ _ _ _1_ _11

_ 1 ._ _ _11111_111111-· .'

90 1 10 100 1,000

- 12 ­

coke carbon content of the spent shales fro~ these runs

correspond to yields of .cV 118% Fischer Assay, based on

the correlation shown in Figure 4.

The data points in Figure 4 represent all bench scale retorting studies made with 28 gallon per ton shale,

which include all runs made with and without fines present at atmospheric and reduced pressures. These data, pre­ sented here in support of the yield results, are entirely consistent and show that on the average very good organic carbon balances were obtained in these studies.

3. Retorting at 10 mm Hg in the Presence of Fines

The study to determine the effect of oil/fines inter­ action on oil yield was continued with these experiments in which retorting was carried out at 10 mm Hg in the

presence of fines. Fischer Assay of the raw shale used

in these experiments was 28 gallons per ton and the effects of pilot and Fischer retort spent shale fines were deter­

mined. The experimental results, which appear in Table 3,

showed that the presence of spent shale fines decreases the oil yield based on the 0 C condensate and further, that the greater losses are obtained with the Fischer retort spent shale. These data are consistent with those obtained at atmospheric pressure except that the yield levels are considerably lower. However, it is interesting to note that yields calculated on total oil recovery are

in the range of 115 to 120 Vol % Fischer Assay, essentially

the same as, if not slightly higher than those obtained

in the absence of fines. This suggests that the fines pro­

mote cracking which changes the product distribution and

increases the difficulty of product recovery. The organic

carbon distributions calculated for these runs were con­ sistent with those obtained from previous runs made with 28 gallon per ton raw shale and the data are included in the correlation shown in Figure 4.

4. Effect of Primary Condenser Temperature

A short study was made to determine the effect of the pri­ mary condenser temperature on Fischer retorting results. The objective of the study was to expand the bases for com­ parison of Fischer retort results with those obtained in the pilot units. Auxiliary heating equipment was installed on the Fischer Assay apparatus to provide primary condenser temperatures equivalent to the average offgas temperatures in the pilot retorts. A condenser temperature of 130 F

was chosen for this study. Results of runs made at both

atmospheric and reduced pressures showed that the oil re­

cove~y was lowered by ~J 10% when the condenser temperature was raised from 32 F to 130 F. The experimental results

-: .

~

_u '::) p o D::: . £\.. 22 t!> Z o "2,.0 .~

'"

·Z . 18 o ~

~

'"

.~ ( j ' ) ' . 14 A Z ·0 tz

ru

~

FIGURE 4

DISTRIBUTION OF RAW SHALE ORGANIC CARBON AS A FUNCTION OF TOTAL OIL YIELD

0· . ! .:" i . I , , I j . i I I . ! i . ,.

.

I

1 I ·-~-·f - ­ --~-., 0

·---1

I I ! -i , .. I i

I

.~

j I . I_, . u . jOI---~---(J

:z

- ... <: .­ L ! c.b

tc

o B

,

0.. ... - - ­ . _ - " ' . . . •• 4

o

.0 ~---95' lOS .110 115 120 P.P. CO,,,VfoE' .• _ 9 - 1Z-G6

TABLE 3

e

e

EFFECT OF RETORTING IN THE PRESENCE OF FINES @ 10 mm Hg Raw Shale (8 -14 Mesh)/Fines (-65 Mesh)

=

3/1 By Weight nes system Product Distribution, wt % Oil S. G. 0 0 Oil H2O SS Gas + 600/60° F C Total 00 C -80° C 00 C -80 0 C Loss 0 0 C ~'Jt % Vol % ~A lilt % Vol % Flrzr ­ f--100 (3) Std. 10.7 0.0 1.4 0.0 85.8 2.1 0.918 > 100.0 100 Vac. 10.4 2.2 0.2 1.5 84.8 0.9 0.957 97.2 93.3 118.0 116 AR (4) Vac. 7.3 2.6 0.3 1.0 88.3 0.5 0.953 90.5 87.2 122.2 122 'AR (4) Vac. 7.1 2.4 0.0 1.3 88.3 1.0 0.957 88.2 83.6 118.6 118 AR (4) Vac. 7.2 2.1 0.3 0.9 88.5 1.1 0.954 90.2 85.8 116.5 115 AR(4) Vac. 7.2 2.4 0.2 1.1 88.4 0.9 0.955 89.6 85.5 119.1 119 (5) Vac. 7.7 1.9 0.3 0.9 88.8 0.4 0.956 95.7 91.4 119.1 118 (5) Vac. 7.3 !.9 0.0 1.3 88.3 0.9 0.958 91.3 87.1 117.3 116 (5) Vac. 7.7 1.9 0.3 0.9 88.5 0.8 0.953 94.7 91.2 117.9 118 (5) Vac. 7.6 1.9 0.2 1.0 88.5 0.7 0.956 93.9 89.9 118.1 117 e made in duplicate: average data are reported. sing estimated S. G. of 0.8 for oil in -800 C trap. n. rt spent shale fines. t spent shale fines. DPCotrupe 8/12/66 j /1/ _Ii

TABLE 3

e

e

EFFECT OF RETORTING I~ THE PRESENCE OF FINES @ 10 rom Hg Ra\,l Shale (8 -14 ~·1esh) /Fines (-65 }~esh)

=

3/1 By Weight Run ,(I) Fines system Product Distribution, wt % Oil S. G. 0 0 Oil H2O SS Gas + 60°/60° F C 0° C -80° C 0° C -80 0 C Loss 0 0 C \'7t % Vol % ='A 7&10 (Avg) Std. 10.7 0.0 1.4 0.0 85.8 2.1 0.918 ~100 (3) ~ 8 Vac. 10.4 2.2 0.2 1.5 84.8 0.9 0.957 97.2 93.3 9 SSFAR(4} Vac. 7.3 2.6 0.3 1.0 88.3 0.5 0.953 90.5 87.2 12 SSFAR (4) Vac. 7.1 2.4 0.0 1.3 88.3 1.0 0.957 88.2 83.6 18 SSFAR(4) Vac. 7.2 2.1 0.3 0.9 88.5 1.1 0.954 90.2 85.8 Avg (9,12 SSFAR(4) Vac. 7.2 2.4 0.2 1.1 88.4 0.9 0.955 89.6 85.5 18) 13 SSR (5) Vac. 7.7 1.9 0.3 0.9 88.8 0.4 0.956 95.7 91.4 1 15 SSR (5) Vac. 7.3 l.9 0.0 1.3 88.3 0.9 0.958 91.3 87.1 1 19 SSR (5) Vac. 7.7 1.9 0.3 0.9 88.5 0.8 0.953 94.7 91.2 1 Avg (13,15 SSR (5) Vac. 7.6 1.9 0.2 1.0 88.5 0.7 0.956 93.9 89.9 1 19) (1)A11 runs were made in duplicate: average data are reported. (2)Calculated using estimated S. G. of 0.8 for oil in -800 C trap. (3)By definition. (4)Fischer retort spent shale fines. (5)Pi10t retort spent shale fines. OF 81 11/ I:

- 13 ­

are summarized in Table 4. It is interesting to note

that when the pilot retort conditions of temperature and

pressure are imposed on the Fischer retort, that oil

re=

covery equivalent to only 82.6 Vol % Fischer Assay was

obtained. It should be noted also that the yield losses

were recovered by cold (-110 F) traps in each case, indi­ cating that the results are a true indication of the p.ffect

of condenser tero~p.rature and not of other yield loss

mechanism. ~

'1"}:\RT.F. 4

EFFECT OF PRIMARY CONDENSER TEf.~PE~,TURE ON

FISCHER RETORTING RESULTS

Run Pressure, Prim. Cond. Oil Recover:!, Vol % Fischer As sa:!

Number mm. Hg:. TemE· , 0 _{F. Primar,l Total}

34 620 32 100.0(1) 100

35 620 130 93.0 102

33 10 130 82.6 112

(2) 10 32 93.9 114

(l)By definition.

(2)Average data from earlier work, included for comparison.

B. Recovery of Oil From Oil/Fines Mixtures

1. Atmospheric Pressure Distillation

The second phase of the program studyinq the effect of oil/fines interaction on oil yield was focused on the oil revaporization process in the Gas Combustion Retort.

Bench scale simulation of this process involved distillation

of oil from prepared mixtures of oil and fines. The initial

series of runs was carried out in the laboratory Fischer

Assay equipment. Mixtures of dry retort oil (RO) and

fines were charged to the retort and the oil was recovered by distillation at atmospheric pressure. The materials

used in these studies included the following.

a. Glass beads.

b. -8 mesh raw shale (RS).

c. -8 mesh spent shale prepared by crushing spent

shale from the pilot retort (SSR).

d. -8 mesh Fischer retort spent shale (SSFAR).

The glass beads were included in these studies to deter­ mine the effect, if any, of increased surface area on oil recovery. The other materials were studied for the

- 14 ­

same reasons discussed above in Section 1 of the Fischer Retorting Studies.

The retort charges comprised 75 g of fines or beads and 10 to 11 g of retort oil. Also, samples of retort oil only were charged and recovered to serve as a control. Following the analytical procedure established in the previous experiment, all materials charged to the retort and residues remaining after the oil was recovered were analyzed for total carbon, hydrogen and mineral C02 con­ tent. All runs were made in duplicate and the precision of results was excellent.

Runs made with the control samples indicated that only

'~89% of shale oil charged is recovered by simple dis­ tillation at atmosoheric pressure. The remainder of the charge is 10stNin the" form of coke or gas. These data are comparable to coking distillation data reported

in the Bureau of tltines Bulletin 210. Comparative data

are shown below. Source

OIl

Coke Lost as Gas H20

A.P. B.M. 89 87.3 7 9.2 3 3.5 1

The studies involving recovery of oil from mixtures indi­ cated that oil recovery is apparently unaffected by the presence of raw shale fines or inert material such as

the glass beads used in these tests. However, spent shale fines from both the pilot and Fischer retorts promoted cracking of the oil, resulting in reduced oil recoveries and increased coke and gas production. The oil recovery data are tabulated below.

TABLE 5

RECOVEFY OF OIL FROt1 OIL/FINFS rUXTURES

MixtUre Oil Recovered, %

RO Only (Control) 89.3

RO + Glass Beads _89.5(1)

RO + RS 89.8

RO + SSR 86.4

RO ₊ SSFAR 87.3

- 15 ­

2. Reduced Pressure nisti11ation

The reduced pressure experiments lrlere designed to approxi­ mate actual operating conditions in the oil revaporization zone of the pilot retort with respect to oil partial pres­ sure, oil/fines ratios and fines type. ~he primary objec­ tives of this study were to determine (1) overall oil loss and (2) the effect of oil/fines ratio on oil recovery.

Secondary objectives were to define oil cracking temperatures and establish gravity history of the overhead product.

The experiments were conducted in laboratory vacuum distil­ lation equipment and the runs were made at 10 mm Bg to approximate the oil partial pressure during normal retort operations. The distillation apparatus comprised the fol­

lowing components.

a. Shortneck distillation flask with thermowe11. b. Simple vapor take-off.

c. Warm water ( IJ""V' 100 F) condensers.

d. Fraction collector. e. Dry ice traps.

f. Vacuum pump.

g. Electric heaters.

The non-heated portion of the flask and the vertical section of the vapor take-off were insulated to minimize ref1uxing. The pressure was controlled with a manual bleed valve and was maintained at 10

±

1 mm Hg throughout each run. The

charge materials used for these studies were retort oil, produced in a recent pilot retort run, and -8 mesh Fischer retort spent shale fines.

The first run of this series was made with retort oil only, to establish a base data point and to check the experimen­ tal procedure. The oil used was free of water and had a specific gravity of 0.938. A 90 9 oil charge was distilled

and the overhead product was collected in approximately 10 m1 fractions. The distillation proceeded smoothly until the overhead temperature reached 650 F. At this point oil cracking occurred and the distillation was dis­ continued. Total oil recovery overhead was 92. 4 t·~t %

and the specific gravity of the collected fractions ranged from 0.851 to 0.990, averaging 0.935. The pot residue amounted to 2.2% of the original charge1 gas plus losses accounting for t~e remaining 5.4%. The oil content of the residue was 4.5%, equivalent to 0.1 Wt %

- 16 ­

of the oil charged. The distillation data and gravity

history of the collected fractions are shown in Figure 5. Data from a typical ASTM shale oil distillation are in­ cluded for comparison.

The next two runs, Fxperiments 2A and 2B, were made with 30:70 weiaht ratios of retort oil and Fischer retort

spent shaie fines. The experimental procedure described

above was followed for these runs with one exception. In

Experiment 2B an attempt ~"as made to continue running

past the cracking point to push the distillation to com­ pletion. However, the rate of cracking increased very rapidly and the increased gas production was too great to be handled by the vacuum equipment, forcing termin­ ation of the run.

The charge for Experiment 2A consisted of 92 g retort oil and 210 g -8 mesh Fischer retort spent shale fines. Oil cracking began at an overhead temperature of 503 F

and the distillation ~1as discontinued. Overhead oil re­

covery was 84.5 Nt % and the specific gravity of the col­ lected fractions ranged from 0.847 to 0.970, averaging

0.929. The pot material was analyzed and found to con­

tain residual oil equivalent to 1.4 Wt % of the original

oil charged. Coke production accountea for an additional

4.5 Wt % of the original oil charge, leaving 9.6% lost as gas. The analyses shm'led also, that 15.5% of the

carbonates in the fines were decomposed. This is sUr­

prising since the pot temperature never exceeded 835 F. Experiment 2B was similar to 2A except that the distil­

lation was not discontinued immediately at the oil cracking

point. The charge contained 90 g of retort oil and 210 9

-8 mesh Fischer retort spent shale fines. The overall temperature history 'Alas slightly hiqher than that of Experiment 2A and the observed oil cracking temperature was 568 F. An attempt was made to continue running past

this point: however, as discussed above, the rapid rise in the rate of gas production forced termination of the

run. Overhead oil recovery was 87.5 1;.1t %, ~V 3% higher

than that in 2A, and the specific gravity of the overhead fractions ranged from 0.846 to 0.990, averaging 0.928. Analysis of the pot material showed that only a trace of oil was left behind. Coke production accounted for 5.9% of the original oil charge and 6.6% was lost as gas.

Carbonate decomposition was 21.5% in this run compared

with 15.5% in 2A. This is not unexpected since the over­

all temperatures were higher in the latter run.

The distillation data for Experiments 2A and 2B appear in Figure 6. Since the gravity histories of the collected

N ~ r.r.:I ~ H ~ If) rz:I p.. X ~ rz:I :::> rx:t t:; ...:I H ~ ~ u tI) ...

....

U Z P:1 ~ t:TI ::t:

~

0 r-I ClV ...:I H 0 rz:I

~

::t: tI) ~ 0 Z 0 H

~

...:I ...:I H E-t tI) H 0 i' L. , .! ­ /009 A~!Ae~2 ~111~ads 0 g Ii! S In d 009 ~ 0: ~ 0:: ____ ,,____ ,__ ~,---,,--=~----..--,---, 0 ___ , _______ ,____ , __ 9 ___ ,_, ________ ,_:' ___________ , \ -\

\,

\---\

\

"

\ ~--\ \:

-\'

'\ \ ,,--­ " -_._,

\

-0 '(0

---g,

0 \£)rc It (J. .c

,.

<1

..

'" C ~cn (l IE :: -. ­ .­ ,.. I C ~ () j('C)

I

-i

FIGURE 6

... '\ ~

BENCH SCALE EXPERUiENTS 2F, F_ND 2B

RECOVERY OF OIL FRO~i OIL/FHJES !·~IZTUP.FS @ 10 mm HG

r­

!

!

i : _ _••_ _ _ _ _~_ _ _ _ _ _ _ _ • _ _ . ._._-- ._..._.- .._..._---­ 1200---•..-. .-' 1

r.

·

~. _ 1 IIO()-t'..-.·--·!--···---T~~-·· • .-.---­ . . . \ . ~ . ---, ! ~OOOi :

-!

.

-.---,---,'._._.'-_.. - .!-

2A

o 1 r ! . .- '_."_'1 ! .. __ ... . ._ ... _. .. i' - ' T ' - ' , ~ 0: ' r - ' L----'----+---:..----t­ -90b;-"'- i ! o ... , . . . > 1 _C E ! 1..0 E '

...

"'­_o "'--"'--'~"' --~---o BOO' 1.000 0 '"""­ r-->­ 1.0 . 1.0 +l

~

.,..j > 1'0 I-l t::; ···-v.n.o U o,..j , _~ I o,..j U Q.I I .~,_____ .. __ 0­ w ---"~---- - . ­ - ·J.:/CO i 200' .. -~.-...---.-,---~-..----...--"--. -_. --.. - - -, ~~ ~ ~: ') 1 - 1---J'jt._-..., ,C~

.-.

J"'. .-;;" 10 20 ₃₀ 40 50 GO 70

eo

90 J00'-"

- 17 ­

fractions were almost identical in both runs, only the data from Experiment 2B are shown in the figure.

Experiment 2C was planned to study a low (1~9) oil/dust ratio. A mixture containing 23 q retort oil and 210 g fines was charged and the distillation was begun. Due to the difficulty of transferring heat through the rela­ tively dry mixture, excessively high pot temperatures were obtained. These temperatures quickly approached the maximum operating temperature of the equipment and it was necessary to abort the run.

In order to obtain data on a second oil/dust ratio, a substitute run, 20, was made in which a l~l ratio was studied. A charge containing 88 9 oil and 90 g fines was used. The distillation proceeded smoothly and

cracking began at an overhead temperature of 635 F4 One apparent anomaly was observed in this run. The over­

head temperatures measured throughout the run ,,,,ere hiqher than had been observed in any of the previous runs in­ cluding the control run. Therefore, upon completion of the run, the temperature measuring equipment was checked and recalibrated. The response of the instrument checked the original calibration and no indications of equipment malfunction were detected. In all other respects the

run appeared normal. Overhead oil recovery was 88.7 t>Jt %

and the specific gravity of the overhead fractions ranged from 0.842 to 0.981, averaging 0.930. The remainder

of the original oil charge was distributed as follows: 2.4% as residual oil on the spent shale, 3.4t as coke, and 5.5% as gas. Again, as in all previous runs, a rela­ tively high percentage (11.4%) of carbonates were decom­ posed.

Potential causes of the carbonate decornpsosition observed in these runs were briefly explored. The carbonate decom­ position results were so~ewhat surprising in that the measured pot temperatures never reached 900 F in any run and were well below normal decomposition temperature. However, since the pot was heated by an electric heating mantle, it is conceivable that the temperatures at the

wall were high enough to produce the observed decomposition. This situation could have become quite severe near the

end of each run when the bulk of the oil had been removed and the pot material 't';ras very dry.

A second conSideration was that the decomposition was caused by organic acids present ~n the original oil,or formed during the cracking react10n. To explore th1s pos­ sibility, an additional oil charqe was distilled fol- , lowing the procedure described above for Run 2. Neutra11­ zation numbers were obtained on samples of the overhead

- 18 ­

fractions and of the oil charged. The total acid deter­ mined by this procedure was equivalent to a maximum of

0.2% carbonates. These results suggest that the decom­ position was not caused by organic acids unless the presence of these materials catalyzed the decomposition reaction. On the other hand, if the latter were true, one would expect carbonate decomposition to occur during Fischer Assay tests and other bench scale retorting

studies. This is not the case. No further work t'\Tas done on this problem.

Summarizing the results of these experiments, it appears that the presence of fines does contribute to oil yield losses and further, that there is a relation between yield loss and oil/fines ratio. The experimental data ap~ear

in Table 6.

c.

Retorting Studies in the Mini-Retort

1. Evaluation of the ~Uni-Retort

The initial experimental program for the r'tini-Retort was designed to provide fundamental data with respect to the retorting characteristics of the fixed-bed unit. The specific objectives of these stUdies were to determine the efficiency of the unit compared \<1ith that of the

Fischer retort, and to develop data on the relative rates of formation of the liquid and gaseous products during the retorting process. The unit was equipped with a liquid product condensing system similar to that used in the Fischer Assay retort except that a dual (swing) sys­ tem was employed to permit collection of liquid product fractions. In addition, a wet test qas meter was installed in the vent line and the offgas was continuously metered throughout each retorting run.

The operating procedure for these runs was very similar to that of the Fischer Assay. The major deviation from the Fischer Assay procedure was that the ~Uni-Retort was heated before the shale was charged. It was necessary to preheat the unit in order to obtain time-temperature

relationships which approximated that of the Fischer Assay. The preheat requirement can best he explained by referring to a sketch of the unit (see Figure 7). As shown, the unit is heated externally with electrical heaters which are wound on an Alundum core. Further, note that the

retort block is a substantial mass of metal which severely limits the ability to significantly vary the unit tempera­ ture at a very rapid rate. Adding the fact that the heat transfer rate through a static shale bed is relatively

slow, it becomes clear that rather high preheat temperatures would be necessary to obtain reasonable retorting rates. In

---TABLE 6

RECOVERY OF OIL FROM OIL/FINES MIXTURES

Pressure

=

10 rom Hg Experiment No. 2 2A 2B 20 Char~e,

t.

Oil ( ) 90 92 90 88 Fines(2} 0.0 210_-. 210 90 Pot Residue, ₉ 2.0 210 203.3 92.7 Oil Distribution, Wt % Overhead 92.4 84.5 87.5 88.7 Residual 0.1 1.4 Trace 2.4 Coke 2.1 4.5 5.9 3.4 Gas(3} 5.4 9.6 6.6 .5.5 Cracking Temperature, of @ 10 rom Hg. ₍₄₎ 650 503 568 635 @ 760 rom Hg. 1000 800 882 965

s.

G. 600_/600 _{F Overhead Product 0.935 0.929 0.928 0.930} 11.4 'Carbonate Decomposition, % 15.5 21.5

(1) Retort oil: S. G 60/60 = 0.938; H20 content - Trace (2) Fischer retort spent shale fines (-8 mesh) . (3)Calculated by difference, includes physical losses (4) Reference: "Data on Hydrocarbons", J. B. Maxwell

DPCotrupe 4/5/67

FIGURE 7

FIXED BED RETORT

THERMOWELLS HEATER WINDINGS ~ ON ALUNDUM ... CORE

..

•

.

.

I

: I

_{Ll.. _}

_..

U

~ STAINLESS STEEL RETORT BLOCK ~MILD STEEL SLEEVE

•

- 19 ­

the studies where attempts were being made to duplicate the Fischer Assay time-temperature relationship, it was found that preheat temperatures of ,... 950 F were required. Shale bed temperatures were measured by making vertical traverses with thermocouple probes in the wells at the wall and in the center of the bed. Wall and center-bed temperatures were obtained at regular intervals through­ out the bed and the average temperature at each level

was calculated. Since the retort is tapered over ~ 60%

of its length, the temperatures at each level were weighted in accordance with the volume of shale present in each

zone. The weighted values were used to-calculate the

overall average bed temperatures. Rather high vertical

and horizontal temperature gradients were obtained in the early portion of each run, however, they continually

diminished as the run progressed.

Typical time-temperature data generated in the fixed bed unit are shown in Figure 8. The Fischer Assay relation­ ship is dashed in for comparison. As suggested by the earlier discussion on the heating capability of this unit, there was little flexibility in altering the shape of

the time-temperature curve. Changing the preheat tem­ perature or heating program resulted in either an upward

or downward shift of the curve. For exa~ple, if an attempt

was made to reach the Fischer Assay peak temperature more rapidly, the temperatures in both the early and latter portions of the run were considerably higher. The MR-7 curve shown here is as close an approximation of the Fischer Assay curve as any obtained with this eauipment. The initial runs in the fixed bed unit were made with a

bed height of 22 inches (,~ 3.5 pound charge). Yields

of 85 to 87 Vol % Fischer Assay were obtained and low oil specific gravities and high residual organic carbon on the spent shale indicated that considerable oil cracking

had occurred. It was thought that the large void space

above the bed and/or excessive wall temperatures were con­ tributinq factors to the relatively high degree of cracking. The bed height was increased to 31 inches (6.5 pound

charge), but the results were the same. Runs were made

in which the heating rate was varied from slow (run time

~2 hours, no measured temperatures above 960 F) to fast

(run time ,,'''vl.5 hours, wall temperatures ,,\,/ 1,000 F) with

no significant change in results. In fact the data from

all rUni-Retort runs were very consistent and oil yields in the 85 to 87% range were obtained in all of the runs made at atmospheric pressure •

Representative experimental data from Run MR-7 are pre­

~ 0 ₈₀₀

..

~ ='

.,.,

ItS ~ G.I 0 ₆₀₀ E G.I Ei ro G.I Il1 G.I ItS 400 ~ G.I

~

200

o

FIGURE 8

COHPARATIVE TIME-TEt1PERATURE RELATIONSHIPS

20 40 60

Time, Minutes

80 100

DPCotrupe

TABLE 7

COMPARISON OF FISC3ER ASSAY AND RUN MR-7 DATA

Run Fischer Assay 14R-7

Oil Yield

Gal/ton 2S.3 24.2 (85.5 Vol.% FA)

·wt.% _{,.". ,} , . 10.7 9.1 (85.1 Wt.% FA)

H20, wt.% , 1.3 1.5

Spent Shale, wt.% 85.7 85.S

Gas + Loss, wt.% '2.3 3.5

Oil Specific Gravity 60/600 _{~ , 0.911 . 0.902}

Raw Shale Organic Carbon On IS.3(1} 22.1

Spent Shale

(l)Estimated from correlation relating organic carbon on spent shale with oil yield (Fiollre 4).

MATERIAL BALAi~CES FOR RUi-1 MR-7

In,

g. Out, g. Wt. % Overall 2,987 2,969 99.4 . Ash' 2 r 004, 1,990 99.3 Min. CO 2 520 503 96.7 Organic Carbon , 'Oil 229 Spent Shale 80 43 .. Gas Total 363 352 97.0 H2O . " _{39 (Fischer} 46 118.0 ... Assay) ., " '. '. DPCotru'Oe 10/14/66

- 20 ...

relatively high degree of cracking is shown by the high gas make, low oil specific gravity and high raw shale

organic carbon on the spent shale (22.1%). The latter

value corresponds to an oil yield of 86.6% in the cor­

relation shown in Figure 4. Water production in all

Hini-Retort runs was in the 1.5 to 1.6% range, ,.,herE!as the Fischer Assay water results were in the 1.2 to 1.3%

range with the exception of that for Run ~R-6 which was

reported as 1.6%. Consequently the water balances were

high for all runs except 1\~R-6. The calibrations of the

centrifuge tubes used for the collection of the liquid product fractions were checked and found to be accurate. The apparent discrepancy in water make was studied in cooperation with Analytical Laboratory personnel, but the problem was not resolved.

2. Reduced Pressure O~eration

One of the primary objectives of the r~ini-Ret(")rt experi­

mental program was to study the effect of using preheated

sweep gas in the retorting process. In the proposed gas

flow studies it was planned to study estimated oil/gas ratios normally encountered in the pilot retorts. In order to evaluate the effect of sweep gas, it was neces­ sary to establish a hase point in the absence of sweep gas at the simulated oil partial pressure in the pilot

retort. Therefore a series of runs ~Tere made at 10 rom

Hg. The oil yields were about 14% higher than those

obtained at atmospheric pressure. The effect of pres­

sure is consistent with that observed in the Fischer retorting studies, although the yield levels obtained in the Mini-Retort were considerably lower than those ob­ tained in the Fischer retorts. For example, retorting

at 10 rom Hg pressure, average yields of 114 Vol % Fischer

Assay were obtained in the Fischer retort, compared with

about 100 Vol t in the Mini-Retort. The comparison holds

at atmospheric pressure also, where Hini-P-etort yields

averaged about 86 Vol % Fischer Assay. nevertheless,

the data were internally consistent and reproducible

throughout this study and a firm basis has been established for comparison and evaluation of results from any subse­ quent retorting studies which might be made in this unit. Typical experimental data, showing the effect of pressure, appear below.

... 21 ­

TABLE 8

EFFECT OF PRESSUPE ON MINI-RETORT RESULTS

Run Fischer AssaI f1lR-7 "1R-12

Pressure, mm Hg 620 (atm) 620 10

Oil Yield

Gal/Ton 2B.3 24.2 28.6

Vol %, Fischer Assay 100.0 85.5 101.2

H20, Wt % _1.3 1.5 1.5

Spent Shale, Wt % 85.7 85.8 85.0

Gas

+

Loss, t~t % 2.3 3.5 2.6

Oil Specific Gravity, 60/60 F 0.911 0.902 0.914 3. Relative F.ates of Formation of Liquid and Gaseous

Products

The Uswing'~ condensing system and wet test meter were in­ stalled on the unit to provide data on the relative rates of formation of oil, gas and water during the retorting operation. The oil and water production rates were

measured at both atmospheric and reduced pressures whereas the gas rates were measured at atmospheric pressure only. The unit was not adequately equipped to measure the gas production rates at reduced pressure. However, the rate of gas production was nearly linear in all of the atmos­ pheric runs and since the general shapes of the oil and water curves are similar at both pressures, it appears that the gas rate can be assumed to be linear at reduced pressure also.

The purpose of obtaining these data was to determine the average oil partial pressure during the batch retorting operation. The partial pressure calculated from these data were used as guides in establishing the sweep gas rates necessary for simulation of the oil partial pressure in the pilot retort. Typical rate data obtained at

atmospheric pressure and at 10 mm Hg appear in Figure 9 and 10, respectively.

4. Gas Flow Studies

The primary objectives of the proposed gas flow studies were to determine the effects of s,.reep gas on retorting and oil yield. It was planned to vary gas rate and gas preheat temperature in this work. Gas preheat and pro­ duct recovery modules were added to the unit as shown in the block diagram in Figure 11. As shown in the figure, the original product recovery system comprised a series of cooling stages and collection vessels. This system proved to be inadequate, therefore the demister was

FIGURE 9

PRODUCTION RATES AT ATMOSPHERIC PRESSURE

!I

ll. -t oj dP till 0 Il Il ~i

~l

Q) :>

~I

tJ 0 0 Q) ~ 8" :x.; 0:: +-I +l 0, :::I' 0 'JP • .j.l

el

:3:: 6(

o

~i

,..;'

~

Itl .:::: 0

....

+l 0 :> 0 0

&!

4(

~O E~

+l 0 :::I '0 0 ~ ~ 80 l4r~.~,~~~~~~~~~rr~~~~ , I I I I I I I I I ,'140

l21=W=1+1 I I I I I I I I I I I I I I I I I I 1-1:::w=+=+=t::W:W-I..

·I-l+l-lllllAstl:xL-tl-1

I I I I I I Ixrnru I I I I LJLtLtLilII I I I I I I I I i I I i I I I I I I I

~120

I I I 1 1 I I I I I

U±J

!, I I I I I I I I I I

l±ltktJittttjjjj~t'J::i:LLI±tHt

I I I I

1±U±J=ltltl

I I I ll±iilli±J±l±±±J-f±t±±-1 I I I.! H-±t-tl

.

i'.. _, I..-

..

" V ~ ,..l:J ~,.

no

[.j: ~ _{~ ~} , / , ' 7 r-~ I r-'~ -- r- " _i ~ b< _-~!::::t -f j _{'JI~ ,} ,

If- r-/ 1/ I 2

a

! If! I ! I I I I I ~'f ! I I I,~! I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I J I I ,~ I I I I I I I I I 1,.0: I~ I I I I I I I I I I I I I I I I I I I ! ! I I I ! ! I I I I I I I ~0 o 40 60 '80

o

20 40 60

Time, Minutes DPCotrupe 4/5/61

FIGURE 10

PRODUCTION RATES AT 10mmHg PRESSURE

140. 12

n

~ ~ I :> ~ r100 Ul

~llOt

,c( ,... ~. ~

~I

II If. ~ _".-1

..

,) rl 81. 80

_....

""

i C

"

-4 ~ ) • +l p

.

3; J ,. _6C

..

60 'C <IJ ,...

,

... <IJ 1) :> -I a (J 1) > )

~

J 4l I) 40 +l 0 ::J ~ J 'U :.> ::s

~

J :J '-l "., 2 I t ! i • ; i 1

11111 I I I 1 1 r-tTt1-n I 1 I I I 1 I I I I 1 1 1 I-T-l-ti-I---H-H+-j-ti-ti 1 1 1 1 I++-++;-T+H-I++ I I I I I I I I+~++++-H-f 1 1 I

f-+-f-HH+H-+-+-++H

I I 1 I ! I I I 71/ iI·~

rH-- r J.\L _0 L t--- _... V - i-- r- r t- f--t- 1 -r 1- - -. t-- -- _. f-_oJ_. - ---1/

I-i--f~

-J+- --

.... -t-- _-- --I-t-- r · I

-'''r

r·i

. -j

i - ; _---

--

_.

--- i- ~ ~... ! _-T~ _ 1-··· :,\~ r /.~ __l f.<:; F. J r 19 t-- --- ---- --- -~

---:--t-

-/ , _{+ f}

-LrttIt

t- - t-:: -- r- -e::L f-- i-V-:~::~ _-..;..l:=P, ·ttl-i--- i - i---- r t-. - t - --i .- --i-- 1----V- I

.i-

+

L __ _~ - - j -1-- .- _{f- 1-- --} I-Vi-- -l--f- -- - - -1-- - .-. r . - _{i - --} ....i-1-. I 1- l -.... ~ --1-- - 1·--_f- 1---- - _;;;i-- --,-- ,.- _II ~. ~ f---.- J

f-

t-- j.,

--\-

~-.--

::;

-,[<$"1: >1"

lZr

~- -- _r

Il

-- - - -+t

-={7

r --f •. I- i"'- - .... - r - . 1-- f--i::

-t

--1- , r _.- 1-- 1- --n _ _ 1_ , _i -

4~

1 , _v --j

-;-U~

_-0---- -- -!,.;

If

I

l l r '

f.

t-- -- f- --1- _iJ;;r~

,-

f-- 1-- r ... -f- -- --, . ---_{r -} i-I , I 1--

_fI

!iH

11 +I-j~

+t-++-I _{I i}

!

I , - i- - -

-f-o

20 40 60 80 100 20 40 60 80 Time,

~inutes

\ _\ ,. < -TO VENT 130°F RETORT _{CONDENSER/COOLER} <.

~

r'~~

_c..!

132°F C"~ COOLER <:.,(

, I

I

L..

J

DEMISTER KNOCKOl)T (GLASSWOOLJ 'KNOCKOUT -110°F TRAP. GAS CYLINDER FIGURE

II

- 22 ­

installed in an effort to improve the efficiency of the system. The vacuum pum p was installed in the vent sys-' tern to overcome the pressure drop in the recovery "'ysteIl'l in order that near-atmospheric conditions could be main­ tained at the top of the retort.

The initial runs in this program clearly indicated the inadequacy of the product recovery system. The major problem was mist carryover throughout the system, re­ sulting in plugging of lines and fittings connecting the various cooling stages due to accumulation of oil in these areas. Altering the temperature levels of the cooling stages served only to decrease the efficiency of the system or to increase the degree of plugging. It was obvious that an efficient mist recovery unit ~tas needed. Therefore, a glass wool-packed column was installed down­ stream of the primary condenser.

The experimental results obtained -..Tith the glass wool demister were mixed, though not entirely satisfactory. In runs where efficient mist recovery was obtained, high pressure drops (6 to 7 inches Hg) developed across the demister. In such cases, it was necessary to operate the low temperature condensers downstream of the demister under a vacuum of 6 to 7 inches Hg in order to maintain near­ atmospheric pressure at the top of the retort. The

efficiency of the low temperature condensers was signifi­ cantly reduced under these conditions, resulting in

losses of water and lighter oil products and in serious plugging of the lines in the -110 F condensing system. Several modifications were made on the packed column and alternate operating procedures were tested in an effort to reduce the pressure drop while maintaining recovery efficiency. None of these actions was completely success­

ful for in each case, although the pressure drop was re­ duced to an acceptable operating level, the efficiency of the unit was also reduced, resulting in mist carry­ over through the recovery train. It was apparent from this work that the modified recovery system was also inadequate for the proposed gas flow studies and would be so for any future bench scale retorting studies which might be contemplated. Therefore it was ~ecided that

work on gas flow studies be discontinued until an adequate mist recovery unit (e.g. a small electrostatic precipitator) was purchased or constructed. In a subsequent revie~7 of

this program, the decision was made to abandon gas flow retorting studies in view of the estimated time required to acquire and develop suitable mist recovery equipment.

- 23 ­

D. Shale Flow Studies with Heated Shale

An item of considerable interest, particularly with respect

to retort operability, is the effect of temperature on the IIflowability" of shale. Therefore, a short program was de­ signed to study the flow characteristics of shale at elevated

temperatures. The studies were carried out in the ~1ini-Retort

which was equipped with a slide valve as shown in Fiqure 12.

Briefly the operating procedure involved charging shale to

the preheated unit, heating the shale to a specified temperature,

discharging the shale from the bottom of the unit ~y JT\pans of

a slide valve, and measuring the time required to discharge

the shale. In practice, the retort was preheated to the desired

temperature level, the shale was charged and heated as rapidly as possible to this level. Extreme care was taken to insure that no measured temperatures exceeded the specified level.

When the entire bed had reached the operating temperature level, the center thermowell was removed, the slide valve opened and the discharge time measured.

The experimental program was designed to investigate temperature

increments of 200 F beginning at ambient temperature and con­

tinuing upward until a discontinuity in flow ~Tas observed.

Temperatures above this level would also be investigated to determine if and at what temperature continuous flow is reestab­

lished. The initial flow studies were made with 8 to 14 mesh,

28 gallon per ton raw shale and temperatures ranqing from 80

to 1,000 F were investigated. The experimental results appear

in Table 9. The flo'., times reported in the table are averages

of times obtained in at least three runs made at the same con­ ditions.

As indicated in Table 9, the flow at 200 and 400 F was extremely

smooth. In both cases the discharged shale was not visibly

different than that charged, although a weight loss (presumably

water) of-, 0.5% was obtained in the 400 F runs. At the 600 F

level, several differences were observed in addition to the significant loss in flow efficiency. Shale flow did not com­ mence immediately upon opening of the slide valver however continuous flow was readily established by tapping the bottom of the retort. Once started, the flO't-l was very smooth and the retort walls were clean after each test. A small amount of

water and oil, equivalent to . , . 1 l-Jt % of the charge, ",'as dis­

tilled off during the heating process and as expected, a large number of the discharged shale particles were blackened.

At 800 F, continuous flow could not be established. The bed

was apparently completely bridged since it was necessary to continuously tap the retort to discharge the shale. Slugs of shale were removed with each tap but at no time did the shale

flow without tapping. A large fraction of the discharged shale