EPA 600/7-86-035
NTIS PB87-120507
LEACHING AND HYDRAULIC PROPERTIES OF RETORTED OIL SHALE
INCLUDING EFFECTS FROM CODISPOSAL OF WASTEWATER
David B. McWhorter
Deanna S. Durnford
Agricultural and Chemical Engineering Department
Engineering Research Center
Colorado State University
Fort Collins, Colorado 80523
Cooperative Agreement CR-807668
Project Officer
Edward R. Bates
Air and Energy Engineering Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Prepared for:
U.S. Environmental Protection Agency
Office of Research and Development
Washington, DC 20460
-------
NOTICE
The information in this document has been funded wholly or in part by the
United states Environmental Protection Agency under cooperative agreement
CR80766S to Colorado state University, it has been subject to the Agency's
peer and administrative review, and it has been approved for publication as an
EPA document. Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.
ii
-------
FOREWORD
When energy and material resources are extracted, processed, converted,
and used, the related pollutional impacts on our environment and even on our
health often require that new and increasingly more efficient pollution
control methods be used. The Air and Energy Engineering Research Laboratory,
Research Triangle Park, assists in developing and demonstrating new and
improved methodologies that will meet these needs both efficiently and
economically. This study investigated the nature and quantity of potential
leachates from retorted (spent) oil shale disposal sites. Synergetic effects
from codisposal of wastewater with retorted shales were investigated. Leachate
quality along with hydraulic propeties were determined for retorted shale from
six retorting processes. Two new laboratory methods were developed which have
application beyond retorted shale characterization. An Equilibrated Soluble
Mass (ESM) column leaching test was developed to characterize the chemical
quality of the first leachate that would issue from a large disposal pile.
A new method was also developed utilizing iodine solution and a dual source
of gamma rays to assess moisture that moves by vapor diffusion from that which
moves as bulk liquid flow. ;
Frank Princiotta
Director
Air and Energy Engineering Research Laboratory
iii
-------
ABSTRACT
The purpose of this project was to develop methods and data on the
leaching and hydraulic properties of solid residues resulting from the
processing of oil shale. A column test, called the ESM (Equilibrated Soluble
Mass) test, was developed as an aid to characterization of the chemical
quality of the first leachate that would issue from a disposal pile of spent
oil shale. Water added for cooling, compaction, and dust control will develop
a chemical composition dictated by chemical reaction between the solution and
solid phases. In the ESM column test, this process is simulated by
moisturizing the solid to the expected field water content, followed by an
equilibration period. After packing the moistened material into a column, the
antecedent pore solution is displaced by injection of distilled water. Both
theoretical and experimental results indicate that the first effluent from the
column is displaced antecedent pore water, the chemical composition of which
has been unaffected by the displacement process. The chemical characteristics
of the first effluent are expected to be a reasonable index to the quality of
first leachate generated from a disposal pile. ',
The ESM test was used to assess the effect on leachate quality of
cociisposal of process water with the solids. This was accomplished by
conducting one set of tests with distilled water as the moisturizing fluid and
one set of tests with process waters as the moisturizing fluid. These tests
indicate an overall negative effect on leachate quality as a result of adding
process water to the solids. This is particularly true with respect to the
concentrations of organics in the column effluent. However, the data indicate
significant adsorption of organics on the solid and the leachate did not
contaj-n as much organic carbon as was present in the process waters. Organic
carbon in the leachate from tests utilizing process water was observed at
higher concentrations than in leachate from the same materials moisturized
with distilled water.
A variety of hydraulic properties were measured in addition : to leachate
quality for the spent shales tested, spent shales tested included samples
from the Lurgi-Ruhrgas, TOSOO II, Allis Chalmers Roller Grate, Paraho Direct
Mode, Chevron STB, and EXTORT retorting processes. A comprehensive data set,
including measurements of the vapor diffusion coefficient, was developed for a
Lurgi spent shale. This was accomplished by a new technique that utilizes an
iodine solution together with a dual source of gamma rays to measure
simultaneously the iodine and water distributions. These data permit one to
distinguish water that has been transported by vapor diffusion from that which
moves as bulk liquid flux.
IV
-------
1. INTRODUCTION
CONTENTS
NOTICE ............... i:L
FOREWORD ......... . ......... "!"."'.".*."." iii
ABSTRACT . ....... ...... " ...... iv
FIGURES; .... ............ ! 1 *. I '. ! ! ! '. '. ' ! viii
TABLES ......... ......
SYMBOLS; . . ............ .*''''"
2. SUMMARY AND CONCLUSIONS ................. . : 5
2..1 Solid Leaching .............. !!!!!! 5
2,,2 Hydraulic Properties ............ ! ! ! ! ! 7
2.3 Codisposal ................ .!!!!! 9
3. SAMPLE SOURCES AND IDENTIFICATION ..... . . ..... . n
3,,1 Lurgi .......... ............. \ '. ±1
3,2 Paraho ............... * 12
3 ,,3 TOSCO ........... ....... I I I I !! I 12
3«,4 Chevron .................. . . ! " ! 13
3 ,,5 Allis Chalmers ........ ...... ...!'" 13
3. ,6 H2TORT ... ........... .....!!!!! l 14
4. TEST PROCEDURES AND ANALYTICAL METHODS . ......... 15
4.1 Physical and Hydraulic Properties Test Procedures . . 15
4.1.1 Particle Size and Density ...... ..... ' 15
4.1.2 water Retention Characteristics ........ 15
4.1.3 Saturated Hydraulic Conductivity ....... 17
4.1.4 Unsaturated Hydraulic Properties . ...... 19
4.2 Leaching Test Procedures ....... ....... 22
4.2.1 The Equilibrated soluble Mass (ESM) Test* '. '. '. • 23
4.2.1.1 The ESM Procedure .......... 23
4.2.1.2 ESM Theory . . . ........ . . 25
4.2.2 The Codisposal ESM Test Procedure ....... 28
4.2.3 The Instantaneously Soluble Mass (ISM) Test . . 29
4.2.3.1 The ISM Procedure .......... 29
4.2.3.2 ISM Theory ... ....... .... 29
4.2.4 ASTM Water Shake Test ......... ! ! ! ! ' 31
4.2.5 RCRA Extraction Tests ....... !!""*.* 31
v
-------
4.3 Analytical Methods .................. : 32
4.3.1 inorganic Analysis ......... !!!!"' 32
4.3.2 Organic Analyses ........... ! I ! ! 32
4.3.2.1 Total Organic Carbon .....!!"*; 32
4.3.2.2 Volatile Organic Analysis ...... 32
4.3.2.3 Semi-Volatile Analysis ...... | 36
5. TEST RESULTS ............ 40
***
- ; • ? • • • ......... ... 4o
5.1.1 Physical and Hydraulic Properties ..... 40
5.1.2 ESM Test Results ............. \ " 50
5.1.3 ISM Test Results .......... !!!!!' 53
5.1.4 Codisposal Test Results .......!] 1 '. * 53
5.1.4.1 Inorganic Analyses ...... ! ! ! ! 53
5.1.4.2 Organic Analyses ....... . $2
. 5.1.5 ASTM and RCRA Test Results ....... <<
5.2 TOSOO II ........ . ........ !!.;.*.* 69
5.2.1 , Physical and Hydraulic Properties !!!!!** 69
5.2.2 ESM Test Results ....... . 75
5.2.3 ISM Test Results .... ...... ! ! ! ! ! 75
5.2.4 ASTM and RCRA Test Results . . . . ! ..... : 77
5.3 paraho ............. . ....... ! ! ! 79
5.3.1 Physical and Hydraulic Properties ....!!*• 79
5.3.2 ISM Test Results . ........... . . 81
5.3.3 Codisposal ESM Test Results ........ " .* 83
5.3.3.1 Inorganic Analysis Results ......' 83
5.3.3.2 Organic Analyses Results , ...... 33
5.3.4 ASTM and RCRA Test Results ...... .... 89
5.4 Allis Chalmers ............. .....! I 89
5.4.1 Physical Properties . ...... !!!!!"* 89
5.4.2 ESM Test Results .... ..... * • 9t
5.5 HYTORT ..................... ! ! ! 92
5.5.1 Physical and Hydraulic Properties .....!* 92
5.5.2 ASTM and RCRA Test Results ..... * 92
5.6 Cheyron .......... . .......... [ * 94
5.6.1 Physical and Hydraulic Properties ....!!! 94
5.6.2 Codisposal ESM Test Results .......... 98
.5.6.2.1 Inorganic Analysis ... ..... '. '. 99
5.6.2.2 organic Analysis ........... 99
6. DISCUSSION AND VERIFICATION OP THE ESM AND ISM
COLUMN LEACH TESTS ............. 103
6.1 The ESM Test .............. i ." [ \ ] [ 103
6.1.1 Reproducibility of the ESM Test ...!!!!! 103
6.1.2 Results as Support for the ESM Theory .!!!!-
6,2 The ISM Test ..... :
-------
GOALZPY ASSURANCE AND QUALITY CONTROL .... ; 119
7.1 Inorganic Tests . . . ............ Ill*1 119
7.2 Organic Tests ....... ........ I I I I I 124
7.2.1 Calibration and standards .. I I I I I I I "I I' ' 124
7.2.2 Duplicate Analyses ...... ....... I 125
REFERENCES
APPENDICES
A CODISPOSAL ESM RESULTS ............. 132
B ASTM AND RCRA. RESULTS ........ I I I I I I I 'I 149
Vll
-------
FIGURES • :
Figure ;
4.1 Schematic of permeameter 13
4.2 Schematic diagram of sectioned column used in
hydraulic diffusivity measurement ± . 21
4.3 Schematic of leaching column ; 24
4.4 Schematic representation of the location of the :
solute plane z* 26
5.1 Particle size distribution for Lurgi OLG retorted
shale 41
5.2 Particle size distribution for Lurgi RB-II retorted ;
shale 41
i
5.3 Particle size distribution for Lurgi RB-I retorted
shale 42
5.4 Pressure-saturation data for Lurgi OLG ........ 46
5.5 Hydraulic diffusivity for Lurgi ULG . 46
5.6 Hydraulic conductivity for Lurgi OLG .......... 48
5.7 Water retention characteristics for Lurgi RB-II
spent shale 49
5.8 Diffusivity for Lurgi RB-II retorted shale ; 51
5.9 Hydraulic conductivity for Lurgi RB-ii :
retorted shale 51
5.10 Normalized EC for Lurgi ISM tests 58
5.11 Major ion composition of effluent - Run 2 (ISM)
(Lurgi OLG)
Vlll
-------
Figure
5.12 Particle size distribution for TOSCO II retorted
........................ . , 70
5.13 Pressure saturation data for TOSCO II ....... ... 73
5.14 Hydraulic diffusivity for TOSCO II .... ...... :. 73
5.15 Hydraulic conductivity for TOSCO II ... ...... '. 74
5.16 Particle size distribution for Paraho II . ...... . so
5.17 Particle size distribution f or Allis Chalmers retorted
shale ............ .... ........ . 90
5.18 Particle size distribution for Chevron retorted shale . 95
5.19 Water retention characteristics for Chevron retorted
shale ...
5.20 Diffusivity for Chevron retorted shale . ....... . 97
5.21 Hydraulic conductivity for Chevron retorted
shale ............ . .......... < 97
6.1 Electrical conductivity of Lurgi ULG effluent samples . 104
6.2 Chloride concentration in Lurgi ULG effluent samples . . 104
6.3 Sulfate concentration in Lurgi ULG effluent samples . . 105
6.4 Relative concentrations in initial effluent samples . . 106
6.5 Theoretical breakthrough curve compared to Cl
data (Lurgi RB-I) .............. ..... m
6.6 Theoretical breakthrough curve compared to Na and '
SO4 data (ISM) leach of Lurgi RB-I) .......... . 113
6.7 Electrical conductivity breakthrough curves for
Runs 4 and 7 (ISM leach of TOSCO II) .......... H4
6.8 Electrical conductivity breakthrough curves for
Run 6 ' •' ....................... 116
6.9 Electrical conductivity breakthrough curves for ;
Run 8 (ISM leach of Paraho)
6.10 Comparison of sulfate breakthrough curves for
Runs 5 and 6 ..................... ; 11?
ix
-------
TABLES
TABLE ; page
1.1 LOCATION OP TEST RESULTS FOR EACH SPENT SHALE . . . . 4
2.1 FIRST EFFLUENT CONCENTRATIONS FOR ESM TESTS 8
2.2 SUMMARY OF RESULTS OF CODISPOSAL TESTS 10
4.1 METHODS USED FOR INORGANIC ANALYSES IN ISM AND
BATCH TESTS 33
4.2 METHODS USED FOR INORGANIC ANALYSES IN ESM
LEACH TESTS .' 34
4.3 METHODS USED FOR INORGANIC ANALYSES IN I
CODISPOSAL TESTS 37
4.4 VOLATILE ORGANIC HAZARDOUS SUBSTANCE LIST AND
PRIORITY POLLUTANT COMPOUNDS, GC/MS DETECTION LIMITS . . 37
4.5 SEMI-VOLATILE ORGANIC HAZARDOUS SUBSTANCELIST ANDd
PRIORITY POLLUTANT COMPOUNDS, GC/MS DETECTION LIMITS . . 38
5.1 SUMMARY OF HYDRAULIC CONDUCTIVITY MEASUREMENTS FOR
LURGI SPENT SHALES (m/s) 44
5.2 WATER RETENTION CHARACTERISTICS OF LURGI SPENT
SHALES DURING DESATURATION 45
5.3 ESM TEST PARAMETERS FOR LURGI ULG 50
5.4 EFFLUENT CONCENTRATIONS FOR RUN NO. 33 (LURGI ULG) ... 52
5.5 EFFLUENT CONCENTRATIONS FOR RUN NO. 34 (LURGI ULG) . . . 52
5.6 SUMMARY OF LURGI ISM TEST PARAMETERS 54
5.7 MAJOR ION COMPOSITION OF LURGI (ULG) EFFLUENT -
RUN NO. 2 (ISM) 54
X.
-------
TABLE Page
5.8 CONCENTRATION OP SELECTED TRACE ELEMENTS IN
LURGI (UD3) - RUN ED. 2 (ISM) 55
5.9 MAJOR ION COMPOSITION OF LURGI (DIG) EFFLUENT -
RUN NO. 3 (ISM) . . . 56
5.10 . SELECTED MAJOR ION CONCENTRATION IN LURGI (RB-I) - '
RON NO. 5 (ISM) . . . 57
5.11 PARAMETERS FOR LURGI RB-II CODISPOSAL TESTS . 60
5.12 RESULTS OF THE INORGANIC ANALYSES FOR THE LURGI RB-II
CODISPOSAL ESM TESTS 61
5.13 MAJOR ION COMPOSITION OF LURGI RB-II EFFLUENT .... . 63
5.14 AVERAGE TOTAL ORGANIC CARBON FOR LURGI RB-II '•
CODISPOSAL TESTS 64
5.15 RESULTS OF VOLATILE ORGANIC ANALYSIS FOR LURGI TESTS . '•. 65
(EPA Method 624)
5.16 RESULTS OF SEMI-VOLATILE ANALYSIS FOR LURGI TESTS . . . 67
(EPA Method 625)
5.17 ESTIMATED CONCENTRATIONS FROM LIBRARY SE2RCH 68
(Base Neutral compounds)
5.18 ASTM BATCH LEECHING TEST-LURGI . . 69
5.19 LURGI RCRA TEST RESULTS 69
!
5.20 WATER RETENTION CHARACTERISTICS OF TOSCO II SPENT :
SHALES DURING DESATURATION :. 71
5.21 SUMMARY OF HYDRAULIC CENEQCTIVIH (m/sec) FOR
TOSCO II 72
5.22 ESM TEST PARAMETERS FOR TOSCO II 75
5.23 EFFLUENT CONCENTRATICNS FOR RUN NO. 29 (TOSCO II) . . . 76
5.24 EFFLUENT CONCENTRATIONS FOR RUN NO. 35 (TOSCO II) .... 76
!
5.25 TOSCO II ISM TEST PARAMETERS 77
5.26 EFFLUENT CONCENTRATIONS IN TOSCO II - RUN NO. 7 .... 78
5.27 TOSCO II ASTM AND RCRA TEST RESULTS ;. 79
xi
-------
TABL% . • . ;
5.28 SUMMARY OP HYDRAULIC ODMSJCTI7ITY (m/sec) FOR
PARAHO-I 81
5.29 PARAEO ISM TEST.PARAMETERS 81
5.30 EFFLUENT CCiNCENTRATIQNS FOR PARAHO (0.4-1.2nm) -
Run No. 6 . i 82
5.31 PARAMETERS FOR PARAHO-II CODlSPOSAL TESTS * 84
5.32 RESULTS OF THE INORGANIC ANALYSIS FOR THE PARAHO-II :
CODlSPOSAL ESM TESTS 85
5.33 MAJOR ION COMPOSITION OF PARAHO-II EFFLUENT ..... 1 86
5.34 AVERAGE TOTAL ORGANIC CARBON FOR PARAHO-II
CODlSPOSAL TESTS 87
5.35 RESULTS OF VOLATILE ORGANIC ANALYSIS FOR PARAHO-II TESTS
(EPA Method 624) . 88
5.36 PARAHO-I ASTM AND RCRA TEST RESULTS 89
5.37 MAJOR ION COMPOSITION OF ALLIS CHALMERS EFFLUENT .... 92
5.38 HYDRAULIC AND MOISTURE RETENTION CHARACTERISTICS OF :
HYTORT RETORTED SHALE . 93
5.39 HYTORT ASTM AND RCRA TEST RESULTS ; 94
5.40 PARAMETERS FOR CHEVRON CODlSPOSAL TESTS 98
5.41 RESULTS OF THE INORGANIC ANALYSES FOR THE CHEVRON
ESM TESTS i 100
5.42 AVERAGE TOTAL ORGANIC CARBON FOR CHEVRON
CODlSPOSAL TESTS . . . 101
5.43 RESULTS OF VOLATILE ORGANIC ANALYSIS FOR CHEVRON .... 102
(EPA METHOD 624) Concentrations (ng/1)
6.1 INITIAL EFFLUENT O^CENTRAT3DNS (mg/L) 106
6.2 MASS OF SPECIES LEACHED PER UNIT MASS OF POROUS
MEDIUM (mg/g) 109
- 6.3 ANTECEDENT MOISTURE VOLUME AND CUMULATIVE LEACHATE
VOLUME AT 0.5 RELATIVE CHLORIDE CONCENTRATION OmL) . . 110
Xll
-------
TABLE
7.1 WATER POLLUTION STUDY NUMBER WP013 .......... ,., 120
7.2 WATER POLLUTION STUDY NUMBER WP013 .......... . 121
7.3 • WATER POLLUTION STUDY NUMBER wpoi4 ...... ..... 122
7.4 WATER POLLUTION STUDY NUMBER WP014 .......... .. 123
7.5 SURROGATE SPIKE PERCENT RECOVERY - METHOD 624 ..... 126
7.6 RESULTS OF DUPLICATE ANALYSIS, VOLATILE ORGANICS . . . , 127
7.7 RESULTS OF DUPLICATE ANALYSIS, SEMI-VOLATILE ORGANICS ;. 129
j
A.I SAMPLE IDENTIFICATION ................ i. 133
A.2 CODISPOSAL ESM TESTS: INORGANIC CONSTITUENTS ....;. 135
A. 3 LIBRARY SEARCH RESULTS FOR SEMI-VOLATILE, BASE
NEUTRAL COMPOUNDS ..... . ............ '. 144
B.I CONCENTRATIONS IN ASTM WATER SHAKE TEST EXTRACTS - '
SPENT SHALES ...........
.
B.2 CONCENTRATIONS IN RCRA TEST EXTRACTS - SPENT SHALES . . 149
Xi'll
-------
a'
A
cum.V
C
Ceq
Ci
D
D,
D^
D_
h
SYMBOLS
Description
unsaturated-flcw parameter in breakthrough
equation,
-------
S mass of a solute species leached during the MM""1
test per unit mass of dry material
t time, usually since injection began '• T
v interstitial pore-solution velocity or Li"1
seepage velocity
vm v at 8 = 6m or saturated seepage velocity ' laT1
x horizontal coordinate L
z vertical coordinate L
z* elevation of solute plane for conservative species L
0 volumetric solution content of porous medium
0£ antecedent volumetric solution content
9m solution content at maximum saturation
PJ.J dry bulk density of the porous medium ML~3
PS particle density of the porous.medium ML~3
-------
Chapter 1
'Ehe total identified shale oil resources in the United States are
estimated by the United States Geological Survey to be over 270 billion metric
tons, with a major source in the Green River formation of Colorado, Wyoming
and Utah. Possible future commercial exploitation of this vast resource
warrants careful consideration of the effects that a commercial oil shale
operation might have on the other natural resources of an oil ; shale region.
Ihe purpose of this report is to provide information pertinent to the
estimation of the quantity and quality of leachate from a spent shale disposal
pile.
Oil shale facilities will produce large volumes of solid wastes and the
potential for recycling these wastes is small. Therefore* establishing
environmentally acceptable methods for disposing of spent shale residues is an
important objective for the oil shale industry. Agarwal (1985) estimates that
if even one half of the planned oil shale production in the U.S. occurs,
approximately 245 million metric tons of retorted shale will be disposed of
each year. Mong the options for disposal are construction ' of compacted
embankments as valley fills, side-hill fills, contoured mounds and mine
backfills. Ihe size of individual disposal piles may cover many hectares and
extend hundreds of meters in depth.
In addition to the solid wastes, there will be a significant amount of
liquid waste generated by the various retorting processes. One method for
disposing of this liquid waste is to codispose it with the solid waste. Spent
shale leaves the retort at elevated temperatures and various liquid waste
streams could be used for cooling. Moisture is also needed for compaction at
the disposal site and for dust control. Oodisposal of the liquid and solid
wastes, however, may change the impact of the disposal pile on '. the environs
compared to using higher quality water for moisturization.
A particular concern in the proper disposal of the oil shale process
residues is the potential for the spent shale disposal pile to release
chemical species that might adversely affect the quality of the surface or
subsurface waters receiving the leachates. Ihe potential for degradation of
the receiving waters depends on both the quantity and quality of the leachate.
Hie movement of the moisture within the disposal pile and the quantity of
leachate generated depend on the initial moisture content at placement, the
hydraulic properties of the spent shale, the storage capacity of; the material,
and the quantity of net infiltration. The quality of the leachate depends on
the chemical-miner alological composition of the retorted shales and the
hydrogeochemical environment in the disposal pile. Hence, the natural
-------
environment is extremely complex and the leaching phenomenon is difficult, to
replicate in a laboratory. Yet a laboratory procedure capable of providing at
least an index of the quality of leachate is imperative if ajy reasonable
planning or comparisons are to be made for the disposal of the ; spent shale
residues. Data on the hydraulic and physical properties of retorted shales is
also needed for predicting the quantity of leachate. ;
The project reported in this paper addresses the components needed for
the prediction of disposal pile leachate quality and quantity. The primary
objectives of the research are to provide a column leaching methodology and
laboratory data base that will contribute to the eventual prediction and
assessment of the quantity and quality of leachates from retorted shale
disposal piles. Specific objectives are to:
l. Develop a viable column leaching test for spent shales,
2. Develop and verify the theory for the column leaching test
so that results can be compared and generalized,
i
3. Compare the chemistry of leachate from the column test with
the results from batch tests using the water shake test pro-
posed by the ASTM and the RCRA. test acetic acid test of
the OSEPA,
4. Conduct a study of the hydraulic and physical properties of
spent shales, including both its saturated and unsaturated
properties,
5. Quantify the effects on leachate quality of codispqsing the
wastewaters from the retort process with the solid wastes.
This report describes (a) two column leach test procedures, (b) the
mathematical theory required for interpretation of the leach test data, (c)
the results of chemical analyses of column leachate and comparison of these
with the results of batch tests, (d) the hydraulic and physical properties of
retorted shales and (e) the chemical differences in leachate quality if
process waters, instead of higher quality water, are used for moisturizing the
retorted shales.
It is emphasized the major purpose of this work was to develop laboratory
procedures for the measurement of leaching and hydraulic iproperties of
retorted shale. As a consequence, our methods and procedures changed as the
work progressed. During the course of the project, the retort residues and
process waters available to us also changed. For example, column leaching
studies utilizing saturated columns were attempted at the beginning of the
project (Bryant, 1982). Oiais procedure was abandoned eventually in favor of a
leaching test for which the salient feature was injection into an initially
dry column (the ISM test) (McWhorter and Nazareth, 1984a). Finally the ISM
test was replaced by column test in which an antecedent pore solution,
equilibrated at an appropriate liquid-to-solid ratio, was displaced by the
injected leachant (the ESM test) (Nazareth, 1984).
-------
Similarly, the measursnent .of hydraulic properties evolved during the
course of the project. 2he objective, initially, was to measure only the
hydraulic conductivity at saturation and the water holding capacity (McWhorter
and Nazareth, I984b). it became apparent that the hydraulic properties at
water contents less than saturation were the more important data and the
emphasis was shifted to those measurements (Nazareth, 1984; :McWhorter and
Brown, 1985a).
\
The evolution of the laboratory procedures, coupled with ,the changing
availcibility of residues on which to make the tests, resulted in a distinctly
nonuhiform coverage for any particular material or test. For '. convenience,
Table l.l summarizes the tests performed and gives the location of the results
in this report. No single material was subjected to all the tests and no
single test was performed on all of the materials. Thus, it is not possible
to draw general comparisons of the leaching and hydraulic properties across
all materials. It is believed that such a comparison would not be warranted
in any case. All of the materials tested in this work came from , experimental
or pilot retorting operations. Little or no information on feed source or
retorting conditions was available. We know that the leaching and hydraulic
properties of the solid residues are sensitive to such conditions. While the
date reported herein are believed valid for the particular material tested,
the relationship of those data to that for materials produced by the same
process under different operating condition (e.g., at a different temperature
or at commercial scale) remain largely unknown.
-------
eg
X
CO
D3
Oi
CQ
33
1
S
1
EM
EH
In
1
EH
»_4
.
t
i-l
ca
I-H
1
Mg
Q
S
1
H
01 t"1
|i
E
i
'O jj1
**•£
tj d Tj
32°
1 5?§
Saturated
Hydraulic
-s Conductivity (
O M
D -U u
: -u u
aj 10
6
* >,
|"|
g
Ol -H
'H Jj
O 0 g
3.23
i CO >-i
C^ CO
3
o* o\ c\ o% \o r-
*O «"l *H T-l m C4
, "> tn - , in in in* . | in
| 35 1 i '^o1
£ H i '^ g '^ ,-s'fe
:^
B a I « s i s s R .s
aa1 Q.OI! a i ci 'i
:
t^* 00 '
* « • • « ^^ '*"(
•" « 1 « « | « £ «' »,•
.S1 .S1 ' g -S1 ' .? o- w :cn
01
_ J-4 M l-(
ra 1 1 1 C2 M
ae $^55>-'1~IM>-1
ugc.«2S;D u
» 2 a -H .H .H .H 8 a o
3 §S cncncncnt3(9o
S^a^^^qSSS
^UBijjia^cnoj^;
-
i
1
•
3
1 1
i-l
S
s
J^
T3
d)
'S 02
cn
OJ X
CQ -i-i
-------
Chapter 2
i
SUMMARY AND CONCLUSIONS !
. _ .
This report summarizes the development of methods for determining the
leaching and hydraulic properties of solid residues resulting fron the
processing of oil shale. The activities, results,and conclusions can be
divided_ into three major categories: l) solid leaching, 2) hydraulic
properties, and 3) codisposal of solid and liquid wastes. The third of these
categories is actually an application of the methodology developed in the
first category.
2.1 iSolid Leaching
33oth column leaching and batch leaching tests were conducted. Tbe
production of chemicals into the pore solution and, hence, into the leachate
from a disposal pile, is a highly complex process that will occur in a
hydrogeochemical environment that cannot be rigorously simulated in the
laboratory. Thus, a laboratory test will be, at best, an index to the
chemical composition of the field generated leachate. A column test will be a
usefuIL index only if it is reproducible and if the effect on test results of
experimental parameters such as flow rate and column length can be assessed
quantitatively. Any features of the field evolution of leachate ^quality that
can be simulated in the column test will, of course, enhance its'usefulness.
Based on the above thinking, three different types of column tests were
perfonned. The first test involved saturating a column of spent shale, a long
period of equilibration, and finally, displacement of the pore fluid by
injection of distilled water. This test was attempted first because of the
availability of mathematical models that could be used to quantitatively
assess the effect of column length and injection rate on the shape of the
breakthrough curve. However, a condition for the applicability of the models
was a uniform chemical composition of pore fluid prior to initiation of the
displacement process. Such a condition was not satisfactorily achieved in the
experiments, even after many - weeks of equilibration. In the process of
establishing the initial saturated condition, a degree of nonuniform leaching
invariably occurred. Results from this initial work are reported by Bryant
(1982) but are not included herein.
The difficulty in establishing an initially saturated condition without
causing some leaching to occur led to a test in which the leachant (distilled
water) was injected at a constant rate into the bottom of an initially dry
column of solid. The salient feature of this test is incorporation of easily
dissolved chemicals into the advancing wetting front upon contact by the
leachant. An approximate mathematical model for this test was developed,
-------
applicable for conservative species instantaneously dissolved at the wetting
front., For this reason, the test was called the ISM (Instantaneously" Soluble
Mass) test. ;
Several data sets collected from the ISM test are presented in this
report. While it was determined that the test was reproducible and amenable
to a reasonably simple mathematical analysis (for conservative species),
important shortcomings remained. First, the injection of leachant into an
initially dry medium did not simulate the expected field conditions in which
the material would be disposed in a moist condition. Also, the results of the
test were highly sensitive to the length of the column, an undesirable feature
given that it was not possible to construct the columns at any length that
approached the depth of a field pile.
Ihese shortcomings were eliminated in the third type of :column test,
called the ESM (Equilibrated Soluble Mass) test. In this procedure, the solid
was moistened by distilled water (process water in the case of codisposal
tests) until the water content was that expected to be a reasonable value for
disposal under commercial operations. Ohe moisturizing process was carried
out by sprinkling the solid, spread on a plastic sheet, with frequent mixing
and re-spreading, The moisturized material was then placed in a closed
container and allowed to approach chsnical equilibrium. Ihe material, thus
prepared, was packed into the leach column and constant rate injection into
the bottom was performed. :
Both theoretical and experimental evidence was developed that indicates
the pore solution existing in the column antecedent to injection is displaced
ahead of the invading solution. Iliis is the salient feature of this test.
Because of the displacement process, the first effluent from the •column is the
antecedent pore solution and exhibits the chemical composition thereof.
Because the chemical composition resulted from equilibration of the water with
the solid at a liquid-to-solid ratio approximately that expected ; under field
disposal operations, the quality of the first effluent should be a reasonable
index to the quality of the pore solution in the field. Furthermore, the
first leachate generated in the field will be the antecedent pore solution,
regardless of whether the leachate results from drainage or from net
infiltration. Net infiltration will displace the antecedent pore solution in
the field, just as does the injected water in the column test.
3her-e exists a degree of mixing between the injected water and the
displaced antecedent water due to hydrodynamic dispersion. When this mixing
zone reaches the outflow end of the column, the chenical composition of the
effluent is no longer that of the antecedent pore solution. Ihe subsequent
breakthrough curve is affected by such experimental parameters as column
length, injection rate, particle size, and initial moisture content. Nazareth
(1984) gives a detailed development that permits quantitative assessment of
these parameters. Nazareth's analysis permits one to reduce breakthrough data
from ESM tests to a common basis so that data from two or more tests can be
realistically compared. We emphasize, however, that the breakthrough data are
not ejcpected to closely approximate that which will be observed in the field
because of such considerations as residence time, column length etc. On the
other hand, the chemical composition of the first effluent from the column is
independent of column length and -injection rate, provided a certain minimum
-------
column length is achieved. It is largely for this reason that the ESM column
test has a much greater practical utility than other types of column tests.
Examples of the chemical composition of first effluent from ESM column
tests are summarized in Table 2.1. Such data are expected to be reasonable
indices to the quality of the first leachate that would occur from these
materials. The two experiments on Lurgi DLG and on TOSCO II indicate the
reproducibility of the test. Additional data relating to the assertion that
the antecedent pore solution is displaced and to reproducibility are contained
in the text of this report. ;
Batch leaching tests, the details of which are presented later in the
report, were also performed. The salient features of the batch tests, both
RCRA and ASTM, are the very large liquid-to-sol id ratios and ; the violent
agitation that are utilized. Neither of these conditions are remotely similar
to field conditions. Nevertheless, batch tests can be used to assess the
quantity of extractable chemicals in a given quantity of raw or retorted
shale,. Concentrations of most chemical components of leachate from disposal
piles are expected to greatly exceed the concentrations observed in the batch
tests.. This is because the liquid-to-sol id ratios that will exist in the
waste piles will be much smaller than in the batch tests.
2.2 Hydraulic Properties ;
The permeability at saturation, the moisture characteristic, and the
permeability as a function of water content were measured on materials by
well-known and recognized methods. These data are presented in ;the body of
the report. During the course of the project, it became apparent that water
transport at very low water contents would be an important consideration in
the question of leachate generation. Extensive measurements of the hydraulic
properties were made for one Lurgi retorted shale provided by the Gulf
Research and Technology Company (McWhorter and Brown, I985a). : A method was
developed for measuring the hydraulic diffusivity down to practically zero
water content., In this method, a horizontal column of the spent shale is
injected at a very low rate with a syringe pump. The water content
distribution is measured as a function of time and position ;in the column
using gamma attenuation. These data permit the direct calculation of the
hydraulic diffusivity.
As expected, the data indicated a range of water contents or> the dry end
of the water-content scale in which flow was doninated by vapor transport.
The measurements were repeated using an iodine solution for injection and a
dual source of gamma rays which permitted the simultaneous measiirement of the
concentration distribution of iodine and the water content distribution. It
was observed that iodine-free water indeed moved ahead of water containing
iodine. The iodine-free water moved by vapor diffusion. Further, it was
observed that the plane separating the iodine-free water from the other
liquid, while moving progressively farther into the medium, always occurred at
a constant, characteristic water content. This characteristic;water content
was interpreted as being the value below which bulk, Darcian-type flow of
liquid water could not occur. It is likely that liquid water at or below this
characteristic value existed in adsorbed films not capable of bulk flow
(McWhorter and Brown, I985a). ;
-------
1
8
ST*
03
£3
1
CQ
1
£H
1
1
g
&
1
i*4
03
L_J
s
_ ^
•
a
£y
^4
EH
i
8) °*
6
M
M
O
*S ^**
2
<3
ft
M
M
8>
CO
•H
0>
^
w
C3
|j
W
•H
3
[ | CO
M
03
8
0\
CN
,
§CO
•H
cn
u
T . . .
*™"^ Crt
\«J
o
«.
CO CN
o co
TH
0 I
CO
"*•
VO |
CN
0 1
CO
O ON
«o •
VO CN
CO
O CM
. ON
CO CN
CO
TH O
O\
• CO
^^ TH
TH
TH CO
CO •
• C"*
VI TH
TH
1 5
'O gi
§ "fa
^r o
co O
rM
tN
O 0
TJ- O
CN CO
*n
CO
in o
"* CO
TH ^J-
vo
co o
ON CN
(N CO
•*
vo
CO O
TH TH
CN >0
CO
CN
1 0
o
CO
*
co
TH O
*5j» *^
CO CO
•»
o\
•o o
o
p
u
s
CO
r-|
O
0
O
rH
CN
CO
0
o
o
CO
CO
«*
rH
ON
VO
CO
CO
CO
•^
•
o
ON
TH
vo
0
P
i1
5
CO
o
tH
*
0
rH
rH
rH
rH
•«*
*
rH
CO
vo
Tf
vo
to
rj-
*
t-
CO
•o
*
"O
p
I
s
O t— CO ON -VO
CO O O : • ON
rH • • CN •
O O CO rH
O •* 10 | CO CN
•* rH . 0 • ON
*» o o ! t
rH
i
O\ CN TH '. O
•«*• CN ON rH VO
CO * • * •
CN TH . Tf 10
CO CO :
,
o
rH .
,
!
'
— P P , P ^
"2; ff1 S1 G> •**
gl ^S ^S : S gl
^ &4 ^7 ' 03 CQ
-------
In the particular material on which these measurements were made, the
critical water content was about 7 percent (by volume) . This means, for
example, that water contents up to about 7 percent in this material can be
regarded as being nondrainable. Other materials are expected to have
different values. Detection of the critical water content permitted
separation of the diffusivity data into regions of vapor dominated and,
liquid dominated contributions. Hie method permits measurement of vapor
diffusion coefficients.
\
2.3 Codisppsal :
ESM column test was used to assess the effect of codisposing process
water with the solid. This was accomplished by utilizing distilled water as
the moisturizing fluid in ore set of tests and process Water as the
moisturizing fluid in an otherwise identical set of tests. '.
£>
Table 2.2 summarizes the results of these codisposal trials. It is
evident that moisturizing these particular materials with these particular
waters tended to result in greater IDS concentration in the first effluent as
compared to the tests in which the material was moisturized with distilled
water.. Even in the tests with distilled water moisturization, the TDS
concentrations were quite large and, therefore, the additional increment due
to the process water may not be too significant. •
Hie total organic carbon concentration in the first effluent from
materials moisturized with process water is markedly greater than f ran
materials moisturized with distilled water. Olie data indicate significant
adsorption of organic carbon by the solid, but adsorption is not sufficient to
reduce the TOG in the first effluent to that observed in tests with distilled
water moisturization.
-------
I
H
EH
1
a
§
&4
o
03
EH
1
OJ
§
>j
g
1
i
w
Q
I
S
5
&4
g
•H
N
"H
y
-U
03
•H
S
J->
^
rft
S
&!
Cn
•S
-S
•H
o
s
foisturizing
*=«
—
i,
4J
c
s
iH
M t
LI t
4-1
I
rH
til
ll_|
f^
•H
r-y
Si
ti
CD
Lit
111
4-)
CU
3
It j
t [ t
P^j
na
1
fQ
•d
si
•
o
vo
CO
o
CO
CO
"*
o
t—
^»
«
o
0
CO*
o
O
•>
CO
o
o
00
u
"w s-i
£^ ^J
s as
•H
IH
M
1
•H
,3
V^ V
0 V»
vo
«
o
0
p» 1
en
O O
o co
^ *
•* CO
o o
0 O
CS CM
co en
r-l rH
O
CO
t 1
co
4J U
8 5
4J Cd
-------
Chapter 3 • ',
SAMPLE SOURCES AND IDENTIFICATION ' ;
The spsnt shale samples used in this study are f ran the Lurgi-Ruhrgas,
Paraho-Direct, 1DSCO> Allis Chalmers Roller Grate, Chevron SOB and HXTORT oil
shale retorting processes. Fox (1983) and Agarwal (1985) discuss details of
the processes and summarize the physical and chemical properties of retorted
shales determined in other studies. In this chapter, the designations used
for the various samples from the processes above are listed and the origins of
the particular spent shales are given when possible. The ;physical and
hydraulic properties determined during this project are given in:Chapter 5 and
will not be repeated here. ;
3.1 Lurgi
Several Lurgi combusted shales are used in different phases of this
project. All of these spent shales were retorted using the Lurgi-Ruhrgas
process initially developed in Frankfurt, Germany but the samples differ in
particle size distribution and other characteristics. In. the Lurgi process,
the retorted shales are combusted to remove residual, carbon compounds. Ihis
produces a gray colored residue that is a characteristic of the Lurgi samples.
A sample of a fine grained Lurgi combusted shale, similar to a flyash was
used early in the project and is designated Lurgi KB-I. No reference number
is available for this sample. It was acquired fran the Rio Blanco Oil Shale
Corporation-.
A second Lurgi sample containing approximately two-thirds fine material
and one-third coarse material is a retorted shale from a pilot plant operated
by Gulf Research and Development Corporation at Hamerville, Pennsylvania. The
material was obtained from Gulf with the designation 10/31/83-2100-Run 108.9C
indicating the date and run when the sample was pulled. The specif ication
used in this study for this sample is Lurgi RB-II. The larger particles in
this sample were slightly darker in color than the smaller particles, because
of either different mineralogy or a different amount of residual kerogen. In
the pilot plant, the raw shale was crushed to about -3 mm and retorted to
remove the kerogen. It was then combusted to remove the residual- carbon.
Because of the high process temperatures that the Lurgi retorted shale is
subjected to, the combusted shale is chemically active. When. hydra ted in a
closed environment, a strong ammonia odor is evident and cementation occurs.
Cements are, in general, formed by the burning of natural material at high
temperatures to produce a clinker which is ground to a powder and, when
wetted, first becomes a paste and then sets or becomes rigid., This process
11
-------
occurs to some degree with the Lurgi combusted shales. When hydrated,
cementation occurs, apparently more than in other combusted shales that also
exhibit this same property. McWhorter and Brown U985a) discuss'the chsnistry
of the cenenting property of the Lurgi KB-II sample and discuss the
implications for hydraulic testing of this material. Their' results are
important in terms of the leaching tests conducted in this study. McWhorter
and Brown (1985a) also investigated the physical and hydraulic properties of
this retorted shale in detail. Some of their results are: summarized in
Chapter 5. Two Lurgi process waters, one steam stripped and the second
unstripped, were used in the codisposal tests with the Lurgi RB-II sample.
Two other Lurgi spent shale samples were used in the ISM and ESM colcsm
tests. These are designated by RG-I and ULG. The Lurgi ULG was used
extensively 'in the development of the leaching tests. Reference' numbers for
these samples are not available but the hydraulic and physical properties are
summarized in Chapter 5 for these materials. The Lurgi ULG is ; a relatively
uniform material with almost all particle sizes between 0.1 and 5 mi in
diameter. •
3.2 Paraho
TWo samples of shales retorted using the Paraho-Direct process were used
in this study. The Paraho-i sample is from a pilot plant operated by
Development Engineering, Inc. at Anvil Points, Colorado. This Anvil Points
facility is a direct-mode 23 m high semiworks retort. The Parah0-I sample was
separated by sieving into three size fractions (0.420-1.190 mm, 1.190-2.00 ESS
and 2.362-3.327 mm) and the hydraulic and physical properties of each fraction
were determined separately. The results are given in Chapter 5.' One property
of the Paraho-l material which affected results from the ISM tests is that the
Paraho material has a large microspace which affects the shape of the effluent
breakthrough curves. ;
A second Paraho sample, Paraho-ll, was obtained later in the project frcm
Battelle Pacific Northwest Labs in Richland, Washington. This second sample
had been retorted at the Anvil Points semiworks plant and stored by Battelle.
Two process waters, a gas condensate and a retort water, were also obtained at
the same time as the spent shale sample. The Paraho-II spent shale and these
waters were used in the codisposal tests. The water samples were sealed but
some degradation of the samples occurred.
3.3 TQ&& :
The TOSCO II material is frcm TOSCO's retort at Rocky Flats, Colorado and
was obtained under an agreenent with TOSCO, Inc. in November, 1981. The TOSCO
reference number is 30049 under TOSCO project number 18320. The TOSCO sample
is a. black, carbonaceous retorted shale that is hydrophobic and produces a
pungent odor during wetting. The black color results from residual carbon.
The TOSCO sample was used in the development of the ESM test. Nazareth (19S4)
reports the physical and hydraulic properties of this sample in detail.
12
-------
3.4 Chevron
A sample of Chevron STB spent shale was obtained from: the Che/ron
Research Company, Salt Lake City, Utah in July, 1985. This sample is
representative of Chevron's Salt Lake City Shale Oil Semiworks. : Plant, which
uses Staged Turbulent Bed technology. The retort is a fluidized-bed system
where pyrolysis is achieved through direct contact between the hot spent shale
and the raw shale.
The Chevron spent shale sample is from a nominal 26-27 gallon/ton fresh
shale feed run. The material is representative of that disposed of at the
semiworks plant but is expected to vary somewhat from that obtained in a
commercial operation. Two expected differences are that the sehiworks sample
is higher in organic content (l percent by weight versus 0.2 percent by weight
expected for a commercial operation) and finer in size, distribution
(approximately 80 percent from the semiworks operation passes, a 400 mesh
sieve). •
A sample of a moisturizing fluid was also obtained f ran Chevron at the.
same time as the spent shale sample and this fluid was used in the codisposai
tests,. The moisturizing fluid was not chemically fixed and therefore degraded
between the time the sample was taken and when it was analyzed. 'The water had
been stripped in a recoiled stripper column. Commercially, it would also be
treated with Chevron's WWT process. The stripper feed water also contains
spent S02 scrubber liquor that would not be present commercially. Other
species such as ammonia and hydrogen sulfide could be in different
concentrations for a commercial operation because fractionator reflux drain
partial pressures and steam balances could be different. ,
The hydraulic and physical properties of the Chevron STB spent shale
sample are given in Chapter 5. This sample, with the process water for
moisturizing, was used in the codisposai tests. '
3.5 Allis Chalmers
r
The Allis Chalmers combusted shale used in this study was, obtained in
March, 1985 from the Oak Creek, Wisconsin Advanced Technology Center. The
material is typical of that produced in Roller Grate pilot plant tests and
should be similar to the spent shales expected from the Allis Chalmers
commercial Roller Grate process.
j
The original raw shale was a New Albany eastern shale from Clark County,
'Indiana. It has a Fischer assay of 12.5 gallons/ton and contained 4.9 percent
total sulfur of which approximately 80 percent was in the' form of pyrite
(FeS2). Carbon dioxide contained in mineral carbonates was 2 percent of the
raw shale weight and the original organic carbon content of the shale was 11.5
percent. The combusted shale sample used in this project was not analyzed by
Allis Chalmers. However, similar combusted shales tested by Allis Chalmers
contained about 3.5 percent residual carbon and 2.8 percent sulfur, primarily
in the form of FeS.
13
-------
The only expected difference between the pilot plant spent shale and that
expected from the conmercial Roller Grate process is that the spent shale fran
titie commercial process will be more fully combusted and, hence, the levels of
contained carbon and sulfur will be lower. The reason for this difference is
that eastern shale has a higher residual carbon content and lower porosity
than western shale. Therefore, a more controlled combustion including recycle
of gases is required. In the batch process development unit, recycle of gases
was not possible. :
3.6 HHQQBI :
A sample of HYTORT retorted shale was otained from IGT; in Chicago,
Illinois. This sample is from process development unit (PDU) run 81 HYC-3
conducted in June of 1981. The shale used in this run was a Kentucky Kew
Albany Shale. :
'Che FDD is a two-stage 1 ton/hr unit developed to examine the potential
for shale gasification. As such, run 81 HYC-3 developed temperatures of up to
1240°P in stage 2 of the FDD which is the gas producing range for the EXTORT
process. The currently envisaged HYTORT process will operate1 in the 1000CF
range,. Therefore, these results may not accurately reflect the leachate to be
expected from a commercial HYTORT operation. The sample is: a very black,
carbonaceous spent shale with granular particles, the largest approximately 12
mm in size.
14
-------
Chapter 4
TEST PROCEDURES AND ANALYTICAL METHODS
Test Procedures :
4.1.1 Particle^Size and Density
Particle size analyses for all samples follow the procedure given in ASTM
D422. This standard describes a method for the quantitative determination of
soil particle sizes by dry sieving of the larger fraction and tjy hydrometer
analysis of the smaller fraction. Samples used in this study were subjected
to the dry sieving analysis or to both components of the procedure depending
on the characteristics of the sample.
The apparent particle density was determined by standard methods (ASTM
D854) using 500 ml volume flasks. A sample of at least 25 grams is placed in
the volumetric flask which is then filled three-fourths full with distilled
water. Mr is removed from the solid-water mixture by subjecting it to a
partial vacuum. The vacuum is left on the sample for at least 24 hours, after
which the flask is filled and. the mixture weighed. Prom the weight of an oven
dried sample, weight of the mixture, weight of distilled water in the flask
and temperature, the apparent particle density is determined. The term
"apparent" is used because the micro-porosity of some of the retorted shale
samples is high and, therefore, the specific value determined for a particular
sample may not be unique but depend on how long the vacuum was: applied and
whether the material was crushed before testing. i
4.1.2 Water Retention Characteristics
Water retention characteristics were determined for most of 'the retorted
shale samples used in this study. The procedure for determining the
desaturation pressure-water retention curve is an extension of the procedure
given in ASTM D2325. Duplicate or triplicate samples are packed to a
specified dry bulk density in small sample rings. These samples!are saturated
and placed on a saturated porous ceramic plate (bubbling pressure = 15 bars)
which is then placed in a pressure chamber. A neoprene bladder under the
plate is sealed along the circumference of the plate with'the;space between
the bottom of the plate and the bladder connected to the atmosphere with a
drain line. The samples on the porous plate are put in a pressure chamber and
the chamber is sealed. As the air pressure in the chamber is increased, part
of the water is displaced and passes from the sample, through'the plate'and
out the drain line. After mechanical equilibrium is reached, the pressure is
released and the samples are removed and weighed to determine the water
content corresponding to the applied pressure. The samples are then replaced
15 :
-------
on the plate and a higher air pressure applied. In this study, the process
was repeated over a range of pressures, usually from l to 14.2 bars (10.33 to
145 m of water) to determine the desaturation curves. ;
Additional water retention information was obtained for the Lurgi ULG,
TOSGO II and Lurgi KB-II samples. The procedure described above determines
water holding capacity during desaturation of the sample. However, hysteretic
effects occur in the relationship between the moisture content and capillary
pressure. In the column leach test and, similarly in the field case, the
shale has an initially low moisture content. This is increased by injecting
moisture in the leach column or by precipitation or irrigation in the field.
therefore, knowing the relationship between pressure and moisture content on
the wetting curve is as important as knowing this relationship for the drying
or desaturation curve.
The test cell used for the pressure-saturation experiments for Lurgi ULG
and TOSGO II is a brass ring 53.8 mm in diameter and 30.2 mm deep. This ring
rests on a small ceramic plate and is sealed in the end fixture by a rul±ser
0-ring. The ceramic plate is saturated with water and is placed in hydraulic
contact with a pressure transducer. Hie ceramic plate acts as a tensiometer
measuring the pressure potential of the solution in contact;with it. The
pressure in the system is shown on the transducer indicator. Pressure-
saturation data is obtained by adding distilled water to the sample using a
syringe via holes provided for this purpose in the retainer plate, The added
moisture is allowed to redistribute in the sample and equilibrate with the
ceramic plate. Equilibrium between the ceramic plate and the; sample pore
solution can be detected by watching the transducer indicator needle drop and
remain steady at a lower suction. Moisture content is then measured by noting
the change in weight of the test cell - transducer system. The process of
wetting is continued until the sample attains a maximum saturation at zero
pressure.
The water retention characteristics were measured for both: the wetting
and drying processes for Lurgi EB-II (McWhorter and Brown, 1985).
Desaturation (drying) data were obtained as described using the large pressure
plate apparatus above. Saturation (wetting) data was obtained using the same
pressure plate apparatus but water was added directly to the samples and the
air pressure was decreased in increments to obtain the wetting curve
characteristics.
Water vapor adsorption and desorption were used to determine the water
holding capacity at large values of apparent capillary tension for the Lurgi
HB-II sample. Capillary tension is approximated using the relationship
h_ = -1.39 x 106 In RH ' (4-1)
c
i
where h_ = capillary water tension in cm of water
C
• RH = relative humidity '
16
-------
This relationship is only approximate because the dissolved salts are not
accounted for but values reported are adjusted for the osmotic potential. -
In the vapor adsorption tests, approximately five grams of oven dried,
hydrated and processed Lurgi retorted shale were weighed into sample cans.
Duplicate samples were then equilibrated in closed chambers over various
saturated salt solutions with known vapor pressures. The chambers were kept
at atmospheric pressure and were maintained at 22°C ± l°C by a constant
temperature bath. Qice a day, the samples were weighed to an accuraqf of
0.0001 gm. The sample cans were covered during weighing1 to minimize
evaporation or condensation. The samples were considered to reach equilibrium
if their weight did not change by more than o.ooi grams over 24 hours.
Vapor desorption tests were conducted as above, but a "wetv sample was
used initially. Water was removed from the samples by evaporation until the
vapor pressure of the adsorbed water was equal to that of the salt solution.
To speed the desorption process, the samples were placed in vacuum desiccators
between weighings. The vacuum in the chambers was about 20 in. of Hg. The
chambers were kept in a constant temperature room at 23.5 ± l°C. '
4.1.3 Saturated Hydraulic Conductivity
The saturated hydraulic conductivity of a sample can be measured with
either a falling head or constant head permeameter. The sample is carefully
packed into a column to a bulk density similar to that used in the column
leaching tests. The size of the column depends on the characteristics of the
sample but is large enough so that edge effects are minimized. Upflow through
the sample occurs because of a differential head across the sample. Filters
at ^ the top and bottom prevent the movement of fines and the flow • gradient is
maintciined at a value low enough to prevent the possibility of piping. The
saturated hydraulic conductivity is calculated by application of Darcy's law
with boundary conditions appropriate for each test. In these standard tests,
no overburden loads are applied.
Tests for the hydraulic conductivity of some spent shales as a function
of moisture content arid bulk density were also completed with various
overburden loads. This procedure for measurement of the hydraulic
conductivity at saturation is given in Designation E-13 in ;the Bureau of
Reclamation^Earth Manual (DSDI, 1974). The test is a constant head test using
a 203 mm diameter permeameter with an overburden load simulated by applying a
load to the top plate. Figure (4.1) is a schematic of the experimental
apparatus.
In commercial, full scale operations, the initial moisture content of the
spent shale disposal piles will depend on the amount of water needed for
compaction, cooling and dust control. In the test procedure used for
determination of the hydraulic conductivity, both this moisture content and
the dry bulk density at which the permeability is desired are specified. A
sample of the material is brought to the prescribed moisture ;content using
standard procedures and the quantity of material required for the prescribed
bulk density is weighed out. The permeameter is packed in lifts to form a
' 17
-------
Tension rods
Dial gage
indicator
Inflow
Top load
plate
Bottom load
plate
Load spring
Gage arm
Load cell
Hydraulic jack
Permeameter
b!ody
i
Overflow
Piston
Porous stone
Sample
Bottom plate
Figure 4.1. Schematic of permeameter.
18
-------
test thickness of 100 mm, using a standard weight hammer. For the Paraho
material, the sample was packed inside an inner sleeve that was j raised as the
material was packed. In the annular space between the inner cylinder and- the
wall of the permeameter, minus 10 material was vigorously packed to prevent
channeling adjacent to the wall, a potential problem with the coarser grained
Paraho residue. . •
Af ter packing is complete, the top plates are placed and the simulated
overburden load applied. A load cell is attached to an extension of the shaft
of the jack and is calibrated to indicate the total force applied. She
desired force is maintained by a large spring compressed during the loading.
Upward flow slowly saturates the material. A gradient of about one is
maintained during saturation to minimize the possibility of piping. A
calibrated Mariotte siphon is used to determine the flow rates through the
sample and also to maintain a constant head on the inflow side. The testing
continues for periods of up to several weeks. The hydraulic conductivities
were also determined by the falling head method with overburden loads for
comparison. In this later case, the difference in elevation; between the
falling water level in the head tank and the constant outflow level is
monitored with time. ;
I
4.1.4 Unsaturated hydraulic Properties
Partially saturated flow prior to effluent production affects the shape
of the breakthrough curves in the column tests and, in the field, affects the
beginning of leachate production. In order to quantify these effects and
assess their importance, the hydraulic conductivity, K(6) , and the hydraulic
diffusivity, D(6) , functions need to be determined.
The hydraulic conductivity function can be derived for unsaturated flew
by assuming a unique, equilibrium relationship between the water pressure head
and the water content so that
^c (4-2)
DO) = -KO) -
where D (0) = the liquid phase dif fusivity (L /L - T)
0 = the volumetric solution content (L /L )
K(6) = the hydraulic conductivity as a function of water content
(L/T)
hc = capillary pressure head (L) .
From the diffusivity function and slope of the water retention curves,
the hydraulic conductivity as a function of 0 can be calculated from Eqn.
(4-2).
The hydraulic diffusivity function was obtained for Lurgi DLG . and TGSGQ
II by the Bruce and Klute (1956) method. In this method, the nonlinear
diffusion equation for horizontal flow is transformed to an ordinary nonl jnear
differential equation by introducing the Boltzman variable a. = xt and
integrating to get the following equation for diffusivity ;
19 '•
-------
D
-------
t3
a
o
o =
o o
ow
V_J
11
vE <»
^~ to
2 2
0
co
0 «
g-c
.2 «
Q. '
O
v.
'•».
* • **. *o*CVJ * * •*,"»
^^^
•'/.•.*»>•. •.'•'•'.
•„ »• V.O ' .' . •'
;^;>;^V
/•*.".*•»••' *"•*,--••"
'.-.•••,'" /•'.'•"
••'.0*.«.r- '. *'.8*. '
t.V«?-;'.iO> .;'••?:''
•V^2;/^
'•-.'»• -^ * .• «•
•',"• .•*»•' —•*;'• V-
•°- '/;('!:— .'a'-lA^
'•"t T«v22"tf''V"o,
**»"*,» «.«.*,
'.'.'« •••.''~".'.*,"".o
££#2g?S
••* ••".a?"."!*. ••.•V.' ;
.' .'o'.' •.'«•.•".'.•''.*>
-1 1_1 LJ 0 Ll U l_l U L_,'
92
s.
tn
0>
w S
a>O
!S
§ o
ro «
o
E
o
•H
(3
TD
(U
0?
1
O
o .
-------
attentuation is that water .contents can be measured without destroying the
sample and hence more data can be obtained from the same sample. : 'Sao separate
tests were performed. In each test, an oven dried sample is packed into a
lucite column with an internal diameter of, 38 mm and a length of 150 mm to a
dry bulk density of 1400 kg/m . Hie column is placed between a gammarray
detector and two different energy gamma ray sources ( 4M, 60. ipsv and 7Cs,
600- KEV). A 0.1% by weight sodium iodine solution is injected by syringe pump
into one end and traced by attenuation of the two energy gamma rays.
Since attenuation coefficients for the water and iodine are,different for
each ganma source, it is possible to calculate both the water and iodine
concentrations along the column as the water moves through it. Ab intervals
ranging from 12 to 48 hours over a seven day period, the columns are scanned
and the water content and iodine concentrations determined at several
positions. Between scans, the injection rate is held constant but this rate
is adjusted at the end of scans if necessary to produce adequate profiles.
Data are read from the ganma system into a computer where they are stored and
reduced. After the second scan, the combined liquid-vapor diffusivity is
calculated by a central finite difference approximation of the equation for
the combined liquid-vapor diffusivity function (Eqn. 4-4). Hie unsaturated
hydraulic conductivity as a function of water content can then be calculated
from the water retention curve and Eqn. (4-2).
4.2 Leaching Test Procedures
An ideal leach test would yield data that are directly translatable into
predictions of the chemical constituents in field generated leachate.
However, given that the chemistry of field, leachate will depend on the complex
interactions of flow rate, moisture content, residence time, temperature,
weathering, and pore-gas composition, it seems unrealistic to try to
rigorously simulate the field hydrogeochemical conditions in the laboratory.
Rather, tests were developed that retained certain features thought to be
particularly relevant to field conditions and that provide consistent,
reproducible data so that comparisons of different materials can.be made.
A characteristic feature of field leachate generation is the invasion of
a leachant into a body of porous medium that may contain some degree of
moisture content. Shis antecedent moisture content may range from practically
zero to perhaps 20 percent by weight. For nonzero antecedent moisture
content, the chemical composition of leachate will be strongly • dependent on
the ^chemical composition of the antecedent moisture. Hie chemistry of the
antecedent moisture, in turn, is expected to reflect chemical equilibrium with
the solid material achieved at small liquid-to-solid ratios. '
The mechanics of leaching initially dry retorted shales are fundamentally
different from the mechanics of leaching initially moist shales and invasion
of leachant into a dry body of shale does not appear to be analyzable as a
limiting condition of invasion into moist material. Instead, it is necessary
to treat leaching of dry material by a completely different, and more
approximate, procedure. Therefore, tests using these two different conditions
are given different names. Hie column leach test using initially;dry retorted
shale is referred to as the Instantaneously Soluble Mass (ISM) test. Hiis
22
-------
name was selected because the concentration of dissolved species in the
initial effluent from the column is largely dependent on the mass of- species
capable of being rapidly (instantaneously) dissolved on contact with the
wetting front of the invading leachant.
Tests in which the retorted shale is initially moist are called
Equilibrated Soluble Mass (ESM) tests. This name is used because the
concentration of dissolved species in the initial effluent is that achieved in.
the antecedent moisture under equilibrium conditions. The oil ; shales tested
during this study were moisturized with distilled water in the development of
the ESM tests and with distilled and process waters in the Codisposal tests.
4.2.1 The Equilibrated Soluble Mass (ESM) Test
4.2.1.1 The ESM Procedure '•
This test is conducted by injecting distilled water at a constant
rate into the bottom of the vertical column shown in Figure (4.3). The column
is chosen large enough so that leachate sample volumes are only a small
fraction of a pore volume. This assures that individual samples of effluent
can be regarded as "point" data and edge effects during leaching will be
negligible. The bottom and top end-plates have 2 mm conical disks machined
onto the surfaces that promote uniform flow over the cross section of the
column and insure mixing of the effluent as it enters and exits the column.
The bottom end-plate is glued to the column body while the top plate is
fastened with screws after packing. The perforated rigid plates support the
material and filter disks (Kendall, non-gauze milk filters) prevent the
movement of fines through these plates. A rubber ring seals the top plate to
the column body.
Preparation of the sample includes first mixing the material to be tested
on a plastic sheet and then adding moisturizing fluid while maintaining
thorough mixing. When enough moisture has been added to result in a
predetermined desired water content, the moist sample is placed in a double
plastic bag and closed. An equilibration time of 24 to 72 hours is allowed
before the column is packed with the equilibrated sample and moisture.
Careful packing is required so that the porous media does not .free fall and
particle segregation does not occur. Columns are packed in layers to assure
uniformity to a specified dry bulk density, usually about 1400 kg/m .
The leaching test is conducted using a positive displacement pump so that
the injection rate remains constant under the conditions of a variable pumping
head that prevails before the beginning of effluent production. ;Inflow at the
bottom of the column minimizes air entrapment and piping. Effluent from the
top of the column is routed through an electrical conductivity probe and pH
meter and into a graduated cylinder. A record of the cumulative volume of
effluent, flow rate, electrical conductivity, pH, and taiiperature is
maintained and effluent is saved at predetermined intervals for analysis of
the chemical constituents. Handling and storing of the effluent varies
depending on the type of chemical analyses to be performed. :
23
-------
outflow
.column top plate
^S__ perforated top plate
•filter disk
.column body
.filter disk
perforated bottom piate
inflow
Figure 4.3. Schematic of leaching column.
24
-------
4.2.1.2 ESM Theory :
The mechanics of the invasion of leachant into la previously
moistened porous medium are important in establishing the salient features of
the ESM test. The velocity of fluid parcels on a cross section, :averaged over
the cross section, is the familiar seepage velocity
-- • (4-5)
v = q/e :
where q is the volume flux, and e is the volumetric moisture content. She
seepage velocity v, is also the average time rate of displacement of all fluid
particles on the cross section. Thus,
_ ...
\ (4-7)
qdt-6dz = 0 " - .
i
where z is the position of the fluid particles and t is time.
liquation (4-7) can be integrated (Nazareth, 1984; Wilson; and Gelhar,
1981; and Smiles, et al. , 1981) to find ''
fz* (4~8)
£ 6dz=qot. :
In Eqn. (4-8), z* is the average distance from the bottom of the column to the
plane of fluid parcels that entered the column at t=o. The parameter q is
the volume flux at the bottom of the column (i.e. the injection rate divided
by the cross-sectional area of the column) and is a constant in the leaching
experiments.
The physical interpretation of Egn. (4-8) is that the volume of Ijquid in
the column up to position z* at time t is equal to the total volume injected
to time t. It also shows that z* is the maximum distance to which leachant
has penetrated the column at time t. Any increase in moisture content beyond
z* (iie. for z > z*) .must result from displacement of antecedent moisture.
Note that the right side of Bgn. (4-8) is the total volume of injected
leachant and must equal the volume of leachant in the column minus the volume
of antecedent moisture. Therefore,
jj* e dz » jr (e - e^dz = j£ * e dz - £* e^ + /z* (e -je^az (4-9)
r2* L • (4-10)
or £ ejdz = /z»(6 - e^dz
in which 6- is the antecedent moisture content and L is the length of the
column. The physical interpretation of Bgns. (4-8) and (4-10) is shown
graphically in Figure (4.4). The variable 6m is the maximum obtainable
moisture content or "natural" saturation. ' ,
25
-------
•CO
CD
o
•H
rH
-k
N
O
g
-------
The significance of Bjn. (4-10) is important to the interpretation of
leaching data. The left side is the volume of antecedent moisture originally
existing between the bottom of the column and position z*. The tight side is
the volume of moisture in excess of the antecedent volume that exists beyond
position z*. These two volumes are equal in Bqn. (4-10). Since the masciaasa
distance to which leachant has penetrated is z*, it is concluded that
antecedent moisture has been displaced by the invading leachant. Mixing of
the leachant and the antecedent moisture occurs only in a narrow zone near the
interface at z*.- Beyond this zone, the moving front contains the displaced
and undiluted antecedent moisture only.
Based on the above analysis, the first effluent from the column consists
entirely of antecedent moisture. Thus, the test provides a means of measuring
the chemical composition of the antecedent moisture that has been equilibrated,
with the solid at a liquid-to-solid ratio corresponding to that expected in
the field. Furthermore, the assumption that the invading liquid: displaces tbe
antecedent moisture is also valid in the field situation. The displacement
mechanism is a general phenomenon. If the mixing zone at the interface is
sufficiently small, the first leachate generated in the field is expected to
be undiluted antecedent moisture and the ESM test provides |a means for
assessing the chsnical composition of this antecedent moisture equilibrated
under conditions reasonably similar to field conditions.
i
The theory given above is extended by Nazareth (1984) to include the
effects of hydrodynamic dispersion, linear equilibrium adsorption and
unsaturated flow on the shape of the effluent breakthrough curves. Based on
the same antecedent moisture displacement assumption, Nazareth develops theory
that predicts that the relative concentration of a species in the test
effluent is given by
C,
= 1 - 1/2 erfc I
LV4Dsn,
(4-11)
where C. is the concentration of the species in the initial effluent.
parameter L is the column length, t is time from the beginning of leachant
injection, D is the hydrodynamic diffusion coefficient ' at maximsa
saturation am vm is the seepage velocity at the maximum volumetric moistare
content. The term v t/L is, therefore, a time variable that represents the
number of pore volumes injected. The term Lvm/4D_ is called the colusn
Feclet number and represents the relative importance of convection over
dispersion for the length of the column. The parameters af and It are,
respectively, a constant that accounts for the effects of partially saturated
flow on the breakthrough curves and a retardation factor at maximum saturation
that accounts for the retarding effect of adsorption/desorption processes on
the advance of the center of mass of a reactive species. !
Equation (4-11) provides a means of normalizing observed breakthrough
data that, in turn, permits assessment of the effects of the column length,
injection rate and hydraulic properties of the retorted shale on the observed
27
-------
results. Thus, it is possible to compare data from different materials and
tests. Because the derivation of Eqn. (5-7), the analysis of its terms and
the procedures for determination of the variables are extensive, the reader is
referred to the original publication (Nazareth, 1984) for more details.
Howevesr, data presented in Chapter 6 of this report will be compared to Eqn.
(4-11) as evidence that the assumptions used in its development, particularly
the antecedent moisture displacement assumption, are valid. For a
conservative species, the equation simplifies considerably because K^ equals
unity. therefore, the theory predicts that the 0.5 relative concentration of
a conservative species will occur after one pore volume of liquid is injected.
Comparison of this theory and effluent data for a species such as chloride
will be used as evidence for support of the theory.
4.2.2 The Codisposal ESM Test Procedure !
Ihe ESM test provides a convenient procedure to study the effects of a
codisposal scenario on the quality of leachate that would be expected from a
commercial spent shale disposal pile moisturized with process | water. Bie
procedure described for the ESM test is the same as that used in this section
except that the moisturizing fluid is process water rather than distilled
water. Distilled water is still used as the leachant, however. Since, as the
theory earlier in this chapter suggests, the first effluent from ;the column is
undiluted, equilibrated, moisturizing fluid, the ESM test .will provide
information about the chemical composition of the fluid in a retorted shale
disposal pile. The quality of the first leachate from the column simulates
the quality of the first leachate from a disposal pile. Subsequent leaching
with distilled water' is similar to the infiltration and flow of rain or
irrigation water into and through the disposal pile. However, the flow rates
and residences times in the field are not approximated by the laboratory
tests,, and, therefore, the ESM breakthrough curve is not expected to simulate
the field breakthrough. In addition, the effects of differences such as
compaction effort and length of time until placement are not known at this
time.
In all cases, the moisturizing fluid and spent shale were allowed to
equilibrate for 72 hours in a closed environment. Columns were then packed in
lifts to minimize particle segregation. Antecedent moisture contents were
determined from weighing, oven drying and reweighing at least three shale
samples collected during packing.
In each series of codisposal tests, distilled water was also used for the
moisturizing fluid in one test.' This is included for two reasons. First, the
tests that use distilled water for moisturizing provide a base for comparing
the codisposal leachate composition with the leachate composition expected if
higher quality waters are used for initial moisturizing of ; the disposed
retorted shales. Secondly, leachate quality from this test should provide
insight into the origin of chemical constituents found in the leachate f roa
the other tests that use process waters for moisturizing. In other words,
comparison of leachate constituency from the Codisposal ESM tests with the ESM
test using distilled water for moisturizing should provide seme; insight as to
whether compounds that appear in the leachate are from the shale; or from the
moisturizing water.
28 '
-------
During the Codisposal column leach tests, a record of cumulative volume
of leachate, time, electrical conductivity (EC), pH, flow rate and temperature
is maintained. Effluent is saved at intervals corresponding approximately to
O.l, 0.5, 1.0, 1.5 and 2.5 pore volumes for analysis of the mineral and trace
metal constituents. Alkalinity, ammonia, pH, EC and the Total Dissolved
Solids (IDS) are also determined. Ihe analyses of inorganic compounds were
performed by the Colorado State University Soil Testing Laboratory. In all
cases, the values of pH and EC determined from the lab samples compared well
with the values determined during the column test.
Samples of the leachates were also subjected to analyses; for organic
constituents. The organic analyses were performed by the Western Research
Institute in Laramie, Wyoming. Samples of the first and last effluent frou
the leaching columns were saved for analysis, as well as samples of the
process waters used for moisturizing the spent shales. A blank sample of
distilled water that was routed through an empty column was subjected to all
testing procedures to ensure the integrity of the columns and make sure the
leaching columns were uncontaminated.
The average Total Organic Carbon (TOG) was determined for the
moisturizing waters and the first and last effluents using EPA method 415.1
for all column tests. These samples were also analyzed using EPA method 624
for volatile Organic Priority Pollutants. In addition to, the volatile
organics listed on the EPA Hazardous Substance and Priority Pollutant Compound
List, some additional compounds were checked. The list \ of priority
pollutants, along with these additional compounds, is given in Section 4.3,2.
The Lurgi column leachates were also analyzed for Semi-Volatile Organic
Priority Pollutants using EPA method 625. Because of high concentrations of
some organic compounds found, a computerized library search was used to screen
the samples. First and last effluents, as well as the unstripped and stripped
process waters, were analyzed by these procedures for all the Lurgi column
tests. A sample of the distilled water used in the tests was also checked for
contamination.
4.2.3 The Instantaneously Soluble Mass (ISM) Test
4.2.3.1 The ISM Procedure
The experimental procedures used in the ISM test are Identical to
those used in the ESM test except for the moisturizing step, in :the ISM test,
dry material is initially placed in the column. Readily soluble species are
incorporated into the flowing liquid phase on contact with the invading fluid.
4.2.3.2 ISM Theory
The mathematical analysis of the ESM leaching test cannot be
extended to the limiting condition of zero antecedent moisture content.
Nevertheless, a more approximate analysis can be used to assess the influeixs
of the experimental variables.
29
-------
•33ie invasion of the leachant is idealized in the ISM test as plug flow in
which a sharp wetting front moves into the dry retorted shale. Ihe volumetric
moisture content at the front is 6 the maximum value obtainable consistent
with air entrapment. Chemical species that are capable of being
"instantaneously" solubilized on contact with the leachant are considered.
It is assumed that such species have no further interaction with the solid
phase or other species in solutions.
Let S be the mass of an "instantaneously" soluble species per unit mass
of porous medium. Ihen the mass of solubilized species per unit volume of
pore solution is Sp^/e^ where p^ is the dry bulk density of the packed
column. As the wetting'f ront moves into the column, the rate that the species
come into solution is controlled by the rate of advancement of : the wetting
front and is given by
(4-12)
em
where z* is the position of the wetting front. Once solubilized, the
convective-dispersion equation for a conservative species describes the
concentration of the species in solution. Solution of this equation, subject
to the boundary condition given by Bgn. (4-12) applied at the moving front,
gives (McWhorter, 1982) j
where a
and
1 e*«
(4-14)
In the above equations, C. is the species concentration in the effluent at
the moment effluent production begins, L .is the column length, v is the
seepage velocity and D is the effective coefficient of hydrodynauic
dispersion.
'Che above equations permit one to approximately assess how such
parameters as the length of the column and the flow rate, influence the
observed concentrations in the ISM test. Chloride is again probably the
species that comes closest to the idealization inherent in the analysis.
These equations can also be used to provide sane insight into the differences
between results of the ESM and ISM tests.
30
-------
4.2.4 ASTM Water Shake Test
The water shake test described in this section is a procedure proposed by
the Merican Society for Testing and Materials for the leaching of waste
materials (ASTM, 1978). The ASTM tests for leaching of waste materials
include both a water shake extraction procedure and an acid shake extraction
procedure. Oily the water shake extraction procedure was used in this study.
This test is intended "to determine collectively the .immediate surface
washing and time-independent, diffusion-controlled contributions to leaching
f ran the waste" (ASTM, 1978). :
A known dry weight of each material to be tested (700 g in these tests)
is placed in a glass vessel with 2800 cm of distilled water. A modified
paint-shaker apparatus that imparts both lateral and vertical reciprocating
motion to the vessel agitates the closed vessel for 48 hrs (±0.5 hrs). After
agitation, the solid-liquid mixture is allowed to separate by' gravity for
about l hr. The solution is decanted into a compressed nitrogen barrel
filtering apparatus and filtered through a 0.45 micron filter. |The filtered
solution is preserved for chemical analysis. All tests reported were
conducted at 2l-23°C, ;
4.2.5 RCRA Extraction Tests
The fourth type of test performed on the retorted shales 'is the RCRA
extraction test as detailed in the Federal Register, Dec. 18, 1978.
1 . i
The RCRA test procedure calls for agitation of the waste material by a
stirring mechanism. An agitator was constructed to the specifications in the
test procedure referenced above. The agitator stirs a mixture of 100 g of
solid in 1600 ml of deionized water for 24 hrs (±0.5 hrs). !The pH of the
solution is measured after 15 minutes of agitation and adjusted ;to 5.0+0.2
with 0.5 N acetic acid. If more than 400 ml of acid are required to adjust
the pH to 5.0 ±0.2, then only 400 ml are added and no further pH control is
exercised.
Agitation is continued for another 3 0 minutes and the i pH is again
measured. If the pH changes by more than 0.5 units, the pH is again adjusted
to 5.0 ± 0.2 by the addition of more acetic acid. This process :is continued
for a period of 6 hrs or until a total of 400 ml of acid have been added.
Agitation is then continued for 24 hrs. At the end of the 24 hour agitation
period, the pH is checked. If the pH is in the range 4.8 - 5.2 or 400 ml of
acid have been added, the extraction is complete. Otherwise, additional acid
is added to adjust the pH to 5.0 ± 0.2 on an hourly basis for a period of not
less than 4 hrs.
Once extraction is complete, the liquid-solid mixture is filtered through
a 0.45 micron filter using a compressed nitrogen barrel filter apparatus. The
volume of filtrate is brought to 2000 ml by addition of deionized water and
split for appropriate preservation.
31
-------
4.3 Analytical Methods: :
4.3.1 Inorganic Analysis• •
'Hie inorganic constituency of the leachates was determined by the
Colorado State University Soil Testing Laboratory. Samples'of the process
waters or the leachates from the column leach tests or batch tests were
analyzed for a number of different parameters by standard methods. PH was
measured using a Sensorex pH electrode (5200C) and an Orion (701A) pH meter.
Electrical conductivity was determined using a Beckman probe (Cel-Gl) and a
Wheatstone Bridge (YSI Model 31). Total alkalinity was measured by manual
titration. This value was used with pH to determine H2C03, HC03 and C03
concentrations. :
Anions like F, Cl, N03 and S04 were measured by ion chromatography
(Dionex, System 10). Concentrations of metal ions were usually determined by
the inductively coupled plasma instrument (Plasma-Therm) although, in some
cases, atomic absorption spectroscopy (Varian, 275) was used. : Tables (4.1),
(4.2) and (4.3) show the specific parameters tested and the specific methods
used for the inorganic analysis of leachates from the ISM, ESM and Codisposal
tests, respectively. The batch tests were also analyzed by the methods given
in Table (4.1). :
4.3.2 Organic Analyses
4.3.2.1 Tot^l Organic Carbon
The average total organic carbon (TOC) content in ppm was measured
for each process water and each first and last effluent sample from the column
tests. Carbon is a major element of organic matter and it is easily measured
and quantified. Therefore, it is often used as an indicator of the total
organic matter in a sample. The procedure used for determining TOC is a
modification of EPA method 415.1. The organic carbon is measured using a
carbonaceous analyzer. Organic carbon in the sample is converted to C02 and
titrated in a coulometer cell. Inorganic carbonate is first removai by
acidification of the sample with 5N HC1 and purged with nitrogen.
4.3.2.2 Volatile Organic Analysis
First leachate effluents, last effluents and process water samples
were also subjected to analysis for the Volatile Organic Priority Pollutants
using EPA method 624. This is a purge and trap gas chromatographic/mass
spectrometer method. An inert gas is bubbled through a 5 ml sample in a
purging chamber. The puirgeables are transferred from the aqueous to the vapor
phase. The vapor is swept through a sorbent column where the purgeables are
trappad. After purging, the sorbent column is heated and back flushed with an
inert gas to desorb the purgeables into a gas chromatograph column. The gas
chromatograph is temperature programmed to separate the purgeables.
No major problems were encountered in the Volatile Organic Analysis (VOA)
determinations. However, nine samples required dilutions of from 1:10 to
32
-------
TABLE 4.1. METHODS USED FOR INORGANIC ANALYSES IN ISM AND BATCH TESTS
Parameter Method
pH Electrode
EC Wheatstone Bridge
Alkalinity as Ca ^ Titration
TDS At 1800 gravimetric
F Ion chromatography
Cl Ion selective electrode
H03 Ion chronatography ;
SO, Ion chromatography ;
Zn Atonic absorption ;
Fe Atomic absorption i
Co - Atomic absorption
Li Atonic absorption '•
V Flameless atomic absorption
NH3 Ion selective electrode i
B Inductively coupled plasma
Cd Atonic absorption
Be Inductively coupled plasma
Mg Atonic absorption ;
p Inductively coupled plasma
Si Inductively coupled plasma
Mo Inductively coupled plasma
M Inductively coupled plasma
Ni Atomic absorption
Na Atonic absorption
Cu Atomic absorption
Al Inductively coupled plasma
Ca Atonic absorption
Ba Inductively coupled plasma
K Inductively coupled plasma
Cr Atonic absorption
Sr Inductively coupled plasma
:pb Atonic absorption
Ag Atomic absorption
Tl Flameless atonic absorption
Se Hydride generation
.as Hydride generation
Hg Cold vapor cell
33
-------
TABLE 4.2. METHODS USED FOR INOK3ANIC ANALYSES IN ESM LEACH TESTS
pH pH Electrode
EC Wheatstone Bridge
Alkalinity as CaC03 Manual Titration
F Ion chronatography . ;
Cl Ion chronatography
S04 Ion chromatography ;
Na Atonic absorption '
Ca Atonic absorption
Mg Inductively coupled plasma1
B Inductively coupled plasma1
M0 Inductively coupled plasma1
K Inductively coupled plasma
Al Inductively coupled plasma2
. Be Inductively coupled plasma2
P Inductively coupled plas(na
Si Inductively coupled plasma
Mn Inductively coupled plasma
Ba Inductively coupled plasma
Sr Inductively coupled plasma
Atonic Absorption Spectroscopy used for TOSCO II, run 29;
2
Atonic Absorption Spectroscopy used for TOSCO II, run 35:
34
-------
TABLE 4.3. ME1HCDS USED FOR INOEGANIC ANALYSES IN CODISPOSAL !TESTS
pH Electrode
EC Wheatstone Bridge l
CaG03 titration
CL ion selective electrode .. ;
S04 Turbidimetric method
^4 Automated sal icy late (colorimetric)
Ca Inductively coupled plasma
Mg Inductively coupled plasma ;
Na Inductively coupled plasma
K Inductively coupled plasma
P Inductively coupled plasma
Al Inductively coupled plasma ;
Fe Inductively coupled plasma ;
&} Inductively coupled plasma
Ti Inductively coupled plasma
Cu Inductively coupled plasma
Zn Inductively coupled plasma
Ni Inductively coupled plasma
Mo Inductively coupled plasma
Cd Inductively coupled plasma i
Cr Inductively coupled plasma
Sr Inductively coupled plasma ;
B Inductively coupled plasma
Ba Inductively coupled plasma ]
Pa Inductively coupled plasma
Hg Cold vapor cell
As Hydride generation
Se Hydride generation
Summation of dissolved constituents
35
-------
1:100 before analysis. After a highly concentrated sample is analyzed,
contamination of the column can occur. To verify that the column is clean and
no carryover takes place between samples, a reagent blank is analyzed. This
blank is a volume of deionized, distilled water that is carried through the
entire analytical process. For the volatile analysis, a blank wa;s analyzed at
least every day if the column was used less than 12 hours and every 12 hours
if the column was used more. • '-
Table (4.4) lists the Hazardous Substance and Priority Pollutant
Compounds and detection limits for the volatile organics. Detection limits
for each compound are also shown. In the results given in Chapter 5, if the
concentration found is near these detection limits, the value given should be
viewed as an approximation. In addition to the compounds shown in Table (4.4)
which are priority pollutants, there are several that are added but are not
priority pollutant compounds. These compounds were found in a previous study
and were included in the analyses made during this study to determine if they
would again appear in column leachate. ,
4.3'.2.3 Semi-Volatile Analysis '
The Lurgi samples were analyzed for semi-volatile organics using EPA
method 625, a gas chronatographic/mass spectrometer (GS/MS) method. A
measured volume of sample (approximately one liter) is serially extracted with
methylene chloride at a pH greater than ll and again at a pH less than 2 using
a separatory funnel. In method 625, the methylene chloride extract is dried
and concentrated to a volume of 1 ml* Mass spectrometry is used to separate
and measure parameters in this extract. Semi-Volatile Organics ! included in
this analysis with their detection, limits are shown in Table (4.5).
36
-------
TABLE 4.4. VOLATILE ORGANIC HAZARDOUS SUBSTANCE LIST AND PRIORITY
POLLDTANT COMPOUNDS, GC/MS DETECTION LIMITS
Volatile Organics
Chloronethane
Brononethane
Vinyl Chloride
Chloroethane
Methylene Chloride
Acetone
Carbon Disulfide
1 , l-Dichloroethene
trans-l ,2-Dichloroethene
Chloroform
l ,2-Dichloroethane
2-Butanone
1 ,1 , l-Trichloroethane
Carbon Tetrachloride
Vinyl Acetate
Brcmodichloromethane
1,1,2, 2-Tetrachloroethane
1 , 2-Dichloropropane
trans-l , 3-Dichloropropene
Trichloroethene
Dibronochloromethane
1,1, 2-Trichloroethane
Benzene
cis-1 , 3-Dichloropropene
2-chloroethyl Vinyl Ether
Bronoform
2-Hexanone
4-Methyl-2-pentanone
Tetrachloroethene
Toluene
Chlorobenz ene
Ethyl Benzene
Styrene
Total Xylenes1
Thiophene
Tetrahydrothiophene
Trimethylborane
Detection l^imits*
Low Water
CAS Number ng/L • ,
74-87-3
74-83-9
75-01-4
75-00-3
75-09-2
67-64-1
75-15-0
75-35-4
156-60-5
67-66-3
107-06-2
78-93-3
71-55-6
56-23-5
108-05-4
75-27-4
79-34-5
78-87-5
10061-02-6
79-01-6
124-48-1
79-00-5
71-43-2
10061-01-5
110-75-8
75-25-2
591-78-6
108-10-1
127-18-4
108-88-3
108-90-7
100-41-4
100-42-5
110-02-01
110-01-0
10.
10
10
10
5
10
5
5 ;
5
5
5
10
5 ;
5
10
5 :
5 ;
5
5
5
5
5
5 :
5
10
5
10
10
5 ;
5
5
5 :'
5 • '•
5
10
10 :
25 J
not Priority Pollutant Compound
J estimated value, no standard available
37
-------
TABLE 4.5. SEMI-VOLATILE ORGANIC HAZARDOUS SUBSTANCE LIST AND
PRIORITY POLLUTANT COMPOUNDS, GC/MS DETECTION' LIMITS
Semi-volatile Organics
2 , 6-Dinitrotoluene
Diethylphthalate
4-Chlorophenyl Phenyl
ether
Fluorene
4-Nitroaniline
4 , 6-Dinitro-2-methylphenol
EhnitroscxSiphenylamine
4-Bromophenyi Phenyl ether
Hexachlorobenz ene
Pentachlorophenol
Penanthrene
Anthracene
Di-n-butylphthalate
Fluoranthene
Benzidine
Pyrene
Butyl Benzyl Phthalate
3 ,3 '-Dichlorobenzidine
Benzo ( a ) anthracene
bis (2-ethylhexyDphthalate
Chrysene
Di-n-octyl Phthalate
Benzo (b ) f luoranthene
Benzo ( k ) f luoranthene
Benzo ( a ) py r ene
Indeno (l,2,3-cd)pyrene
Dibenz ( a, h) anthracene
Benzo (g , h, i ) perylene
(continued)
l not a Priority Pollutant
*• fif!/MR detection limits fc
CAS Number
606-20-2
84-66-2
7005-72-3
86-73-7
100-01-6
534-52-1
86-30-6
101-55-3
118-74-1
87-86-5
85-01-8
120-12-7
84-74-2
206-44-0
92-87-5
129-00-0
85-68-7
91-94-1
56-55-3
117-81-7
218-01-9
117-84-0
205-99-2
207-08-9
50-32-8
193-39-5
53-70-3
191-24-2
Compound
5r water are ha serf <
Detection Limits*
Low Water
ng/L i
10
10 i
i
10 ;
10
50 :
so ;
10
10
10
50 ;
10
10
10 .
10
80
10 !
10 !
20
10
10
10
10
10
10
10
10
10
10 .
in a one liter
sample, extracted and concentrated to a volume of 1 ml.;
i
Specific detection limits are highly matrix dependent. ; The
detection limits listed herein are provided for guidance
and may not always be achievable. :
38
-------
. TABLE 4.5. (continued)
Semi-volatile Organics
N-Nitrosodmethylamine
Phenol
Aniline
bis (2-Chloroethyl) ether
2-chlorophenol
l ,3-Dichlorobenzene
1 ,4-Dichlorobenzene
benzyl alcohol
1,2, -Dichlorobenzene
2-Methylphenol
bis (2-Chloroisoprophyl)
ether
4-Methylphenol
N-Nistroso-Dipropylamine
Hexachloroethane
Nitrobenzene
Isophorone
2-Nitrophenol
2,4-Dimethylphenol
Benzoic Acldi
bis (2-chloroethoxy)
methane
2 , 4-Dichlorophenol
1 , 2 , 4-Trichlorobenzene
Naphthalene
4-Chloroanil ine1
Hexachlorobut adi ene
4-Chloro-3 -methy Iphenol
( para-chloro-meta-cr esol )
2-Methylnapthalene1
Hexachlorocyclopentadiene
2,4, 6-Trichlorophenol
2,4, 5-Trichlorophenol1
2-Chloronaphthalene1
2-Nitroaniline1
Dimethyl Phthalate
Acenphythylene
3 -Nitroaniline1
Acenaphthene
2 , 4-Dinitrophenol
4-Nitrophenol
Dibenzofuran
2 , 4-Dinitrotoluene
Detection ! Limits*
Low Water
CAS Number ng/L
62-75-9
108-95-2
62-53-3
111-44-4
95-57-8
541-73-1
106-46-7
100-51-6
95-50-1
95-48-7
39638-32-9
106-44-5
621-64-7
67-72-1
98-95-3
78-59-1
88-75-5
105-67-9
65-85-0
111-91-1
120-83-2
120-82-1
91-20-3
106-47-8
87-68-3
59-50-7
91-57-6
77-47-4
88-06-2
95-95-4
91-58-7
88-74-4
131-11-3
208-96-8
99-09-2
83-32-9
51-28-5
100-02-7
132-64-9
121-14-2
10;
10
10
10
10
10
10
10 '. .
10
10
10
10!
10
10 :
10^
10;
10
10
50
1
10 ;
10
10
10
10
10
10 :
10
10 :
10
so ;
10
50
10 ,
10
50
10
50
50
10
10 ]
39
-------
Chapter 5
TEST RESULTS j
This chapter presents the physical, hydraulic and leaching test results,
both column and batch, for the experiments completed during the project. The
retorted shales are from the Lurgi, TOSGO II, Paraho, HYTORT, i Chevron and
Allis Chalmers processes. '
It is expected that the data presented in this chapter will have some
intrinsic value to the oil shale industry and other researchers. Therefore,
with the goal of making information about each retorting process as accessaide
as possible, this chapter is organized according to the retorting process
used. However, the primary objective of this research was to develop an
appropriate leaching test for determining the quality of leachate fran
retorted shale disposal piles. Therefore, the research focus was the leaching
tests,, not the individual test results or retorted shale properties.
Characterization of the retorted shales' physical and hydraulic properties and
evaluation of leachate quality are significant, but peripheral, benefits which
accrued during the development of the test methodology. The information was
needed for the comparison of methods, validation of the .theories and
quantification of the relative importance of pertinent processes in the
leaching tests. If the information presented in this chapter is not complete,
therefore, it is because either all samples were not available during all
phases of the research or, as development of a workable column test evolved,
the experimental procedures and supplementary data requirements also evolved.
5.1 Lurgi :
The Lurgi combusted shale samples referred to in this • section are
designated by the specifications KB-I, RB-II, RG-I and DLG. '.
5.1.1 Physical and Hydraulic Properties
Particle size analyses of spent shale samples were performed by dry
sieving or by a combination of dry sieving and hydrometer (ASTM D422). Figures
(5.1-5.3) show gradations for some of the Lurgi samples. The Lurgi RB-I is a
fine grained, flyash-like material. The Lurgi DLG is also a uniform size with
almost 95% of the material between 0.1 mm and 5 mm. A third Lurgi, RB-II, has
about thirty percent material by weight smaller than 0.045 mm, 30 percent
between 1.0 and 2.0 mm, with the largest particle size being 5.0 mm and the
median diameter (d5Q) about 0.5 mm.
The apparent particle density was determined by the ASTM |D854 method.
For the Lurgi materials tested, values ranged from 2700 kg/nrfor the Lurgi
40
-------
ICO
80
60-
c
C
-------
(00
80
$3 60
c
iZ
§ 40
20
[_
I
O.OC01 0.00! 0.01 O.'l i.C 10.0 100.0
Porfide Size (mm) ;
Figure 5.3. Particle size distribution tor Lurgi FB-I
retorted shale.
42
-------
DLG to 2760 kg/m3 for the Lurgi fines (Lurgi RB-I). Values for '• Lurgi m-II
averaged 2740 kg/m3. Material finer than 0.045 mm had an average of 2743
kg/m3 while the Lurgi RB-II material larger than 2 mm had an average value of
2728 kg/m3. The values are all higher than the 2610-2680 range reported by RLo
Blanco (1981) for Lurgi spent shale material.
The water retention characteristics and hydraulic conductivities of spent
shales are needed for predicting the movement of liquid within ;a spent shale
disposal pile. In addition, from a research perspective, these properties are
important for quantifying the mechanisms occurring in the columns during the
leaching tests. Therefore, a significant amount of attention was focused in
this study on determining these properties. Both saturated and unsaturated
conductivities were determined, pressure saturation and desaturation curves
were measured and, for the Lurgi IB-II sample, vapor transport mechanisms at
very low moisture contents"were investigated. -
Table (5.1) shows saturated hydraulic conductivity measurements for the
Lurgi samples. For the Lurgi KB-I, various overburden loads were applied and
dry bulk density was varied from a minimum of 1200 kg/m3 to a maximum equal to
a standard Proctor density of 1400 kg/m3. At the Proctor density, the optimum
moisture content was found to be 30 percent by weight.
Standard falling head and constant head methods without overburden loads
determined hydraulic conductivities for Lurgi EB-I, DLG and EB-II samples.
The test column for the Lurgi ULG sample was a lucite cylinder 50 mm in
diameter with a large inlet at the bottom of the column. Screens at the top
and bottom of the sample prevented movement of the fines and piezometer taps
were placed 5 cm apart in the test section to measure the head loss in the
column during the constant head test. Ihe constant head test for the Lurgi
ULG spent shale material, however, proved to be unreliable with inconsistent
values obtained for K. Clogging of the screens behind the piezometer taps is
a possible reason for the unreliable readings. For the falling head test,
these piezometer taps were sealed and results were more satisfactory.
The saturated hydraulic conductivity of the Lurgi EB-II material is from
a falling head permeameter test.' A column 69.5 mm in diameter with a test
section of 530 mm was packed to a dry bulk density of 1370 kg/m3 and allowed
to saturate overnight fay upflow from the bottom. Hie bulk density was held
constant during the test and no additional normal load was applied. During
the test, the hydraulic gradient varied from 3.5 to 6.6. Hie range of results
made over a three day period is 9.13 x 10~8 to 9.69 x 10~8 m/sec.,
i
The water retention characteristics of Lurgi KB-I and RG-I .found during
desaturation of the test samples from natural saturation in a pressure plate
apparatus (ASTM 2325) are given in Table (5.2). Shown are results for several
levels of compaction and pressures in the pressure plate apparatus ranging
from 1 bar to 14.2 bars. . :
Two pressure-saturation experiments were conducted on the Lurgi ULG
material. In these tests, small sample rings were packed with the retorted
shale and placed on a ceramic plate. Since the ESM leaching test is conducted
by injecting into an initially moist sample, the relationships between
43
-------
3
CO
1.4
Sj
3^
WJ
Pj
H
CQ
M
i
1-3
1
CQ
g
i|2
E!
s
p
*E
Ed
57*-
«-«
>»
M
CONDUCPIV
o
M
1
*^
&4
O
1
i
S
•
iH
>0
3
1
cc
or
*™
0
ON
VO
S
£
'S
s
•rt
•«*•
O
Method
«£•«
n
• • • • «
iH CO Ol TH OO
P-
1
O
TH
X
C*l
»-i
't^ (rt Q|j (rt f^™j
IB 03 03
-U 4J
CH C C7) C C7)
C (0 C "3 CJ
•H jj -H jj -H
rH 03 o
VO O O
CS TH iH
M M
S k
2 $
a a
eo oo
r- vo ||
1 1 0 O
O O TH iH
X X
XX
CO G\
CO O r-t VO
• • • *
oo d o\ os
33 H3 EG
.g1.!1 I
rH (H rH
Co cS
M 1-1
1 -S
3 3
44
-------
TABLE 5.2. WATER RETENTION CHARACTERISTICS OP LOTGI SPENT SHALES
BORING DESAKJRATION*
Sample | Compaction |
Pressure (bars)
Lurgi RB-I
Lurgi RG-I
No Compaction
Pb = 1300 kg/m!
Pb - 1450 kg/m!
Pb = 1600 kg/nf
No Compaction
Pb = 1600 kg/m3
1
73.6
62.4
60.2
47.2
27.5
20.7
3
62.0
62.3
58.7
46.3
27.6
20.2
5.1
64.5
62.2
56.6
45.8
26.9
19.8
10.2
63.1
. 62.0
55.5
44.4
25.3
19.8
14.2
59.5
61.7
55.2
43.7
15.5
19.0
* Table entries are moisture contents (w) expressed on a % weight basis:
weight of water per weight of dry solids times 100. .
pressure and moisture content were found on the wetting cycle. Ah exponential
function was found to fit the data well and this function can be used to
characterize the pressure-saturation data for this retorted shale sample.
Results of the tests are shown on a semilog plot in Figure (5.4) with the
visual best fit lines. The equations for these lines are
hc = 2784 exp (-15.8 6)
6 1 .337
hc = 2.67 x 1049 exp (-330 0) 6 > .337
i
where hc = the capillary tension head in cm ;
6 = volumetric water content ;
The Lurgi ULG diffusivity function was determined fay the Bruce and Klute
(1956) method as described in Chapter 4. The function D(6) is obtained fron
solution of Eqn. (4-3) by integrating a best fit curve through e vs xt~I/2
data points. Clothier, et al., (1983) presented a procedure that fits the
primary data with a power function of a normalized volumetric moisture content
multiplied by a scaling factor (see Clothier, et al., 1983). Both of these
evaluation techniques were used to fit the Lurgi ULG data and obtain a
dif f usivity function. It was found that, for Lurgi ULG, the Clothier, et al.,
technique for fitting a functional to the absorption profile (e: vs xt~172)
resulted in a poor fit. However, the diffusivity function derived by both
methods agreed fairly well in the mid range. Figure (5.5) shows , averages of
four tests with the test data evaluated by both methods. The results are
generally comparable over most of the range of e values. Near 6^ and 8™, the
Bruce and Klute method shows considerable scatter and anomolous behavior.
This was because of the steep slopes on the absorption diagrams in these
areas.
From the diffusivity function and, in this .case, pressure-saturation data
(Figure 5.4), the hydraulic conductivity function for unsaturated conditions
45
-------
in
to •
n _ . — . ~|
in - — -
oa 2 S3 • .
o a » C.
ao 5 o
"O S
m .2
CD o -=
= o
DO CO u
a o ° D
a o
a o
a o
ao
o a
o ° D a
•
1 i I r t r I t I 1 r i i i I i i t 1 i i r i I I i i ! t i t i i t i t
bT »l K> f
o
o
to
d
i
M ,
O
O
PJ
1
l£5
O ;
o
d ;
m !
o ,
d
'
o
2 "2 'o '2 ;
•H
D^
3
O
'W
r?
"d
r-1
C*3
3
It!
T3
O
"3
(3
•8
^*i
33
*
*rt
«
0)
3
cn
•H
Ed
-r— (3as/3ujo)
CP
o
-w
§
. (UI3)
3
01
9
(U
cn
46
-------
can be determined (Bgn. 4-2). Figure (5.6) shews unsaturated hydraulic
condmptivity for the Lurgi ULG spent shale material. Values near Q. where
the diffusivity function is uncertain, were ignored. The mid range values of
6, where the two methods for evaluating diffusivity agreed, result in a~K(6)
function that is fitted very well by a straight line on the 'semilog plot
shown,. Near e where again the diffusivity function showed uncertainty, the
calculated K values were higher than KS evaluated by standard saturated
pemeameter methods. Therefore, the saturated hydraulic conductivity was used
as a cut off level for the K(6) curve. Nazareth (1984) provides :more details
on the tests for Lurgi OLG.
She physical and hydraulic properties of Lurgi IB-II were i explored in
detail during this project and MoWhorter and Brown (I985a) report the results
of these investigations. The emphasis of their report is the liquid and vapor
transport properties at very low water contents in the Lurgi sample. However,
the physical, hydraulic and chonical properties of the sample are also
reported. Tests conducted include material tests (size distribution, batch
leach tests), cementing properties (x-ray diffraction, hydration test,
scanning electron microscope), water holding capacity (pressure plate, vapor
sorption) and hydraulic properties (diffusivity and conductivity)!. Results of
the water holding capacity and hydraulic properties will be briefly summarized
here. !
• i •
Figure (5.7) shows the moisture characteristic curve of Lurgi EB-II. At
capillary tension heads of 88 to 1.5 x 104 cm of water (.09 to 14.7 bars), the
data for both the desaturation and saturation cycles were measured using a
pressure plate apparatus. Approximately 66 grams of oven dried, hydrated
material were placed in sample rings at a bulk density of 1400 kg/m3. Samples
used for determining the drainage curve were wetted overnight to obtain
natural saturation. Triplicate samples were placed in the pressure cells at
various pressures and allowed to equilibrate for 3 to 5 days. The samples
were then removed, weighed and oven dried to determine the moisture content.
For the desaturation curve, pressures in the pressure plate ^apparatus are
increased by intervals. When the pressure is increased, a pressure gradient
is induced across the sample and flow occurs out of the sample. For the
saturation curve, pressures in the pressure plate apparatus are decreased,
rather than increased. Liquid is added directly to each sample, !the sample is
allowed to equilibrate at the new pressure level, and the equilibrated
moisture content is determined. If the triplicate samples are not in
agreement, the test is rerun.
Vapor sorption data for Lurgi IB-II are also plotted on Figure (5.7).
The values of the capillary tension heads corresponding to the vapor sorption
date have been reduced by 2.5 x 104 cm to account for osmotic potential. The
98 percent SE adsorption water content was matched to the pressure plate data
to determine this value. In any case, the curves are insensitive to the
estimated osmotic potential.
The diffusivity and hydraulic conductivity as a function: of moisture
content were also determined for the Lurgi RB-II retorted shale. Figure (5.8)
is a graph of the combined liquid-vapor diffusivity function. Gamma ray
attentuation was used for determining the 6 and x data needed for evaluation
47
-------
I.-
o
0)
E
o
ICT°
10'
,-6
10'
Ks = 2.0x!0"4cm/s
l~ I
0.10
0.15 0.20 0.25 0.30 0.35 Q.4O
n ;
Figure 5.6. Hydraulic conductivity for Lurgi ULG.
48
-------
o vapor adsorption (adjusted)
* vapor tiesorption (adjusted)
O pressure eel/ weffing
• pressure cell drainage
drainage
Figure 5.7.
Water retention characteristics.for
Lurgi IB-II spent snale.
49
-------
of D(6) by Eqn. (4-4). Two separate tests are shown in Figure (5.8). As this
figure shows, the function exhibits the expected local maximum in the region
where vapor transport dominates. The minimum value of D „ occurs at the
critical water content of 0.066. £v ;
The unsaturated hydraulic function can be calculated f ran the water
retention curve (Figure 5.7), and the diffusivity function using Bgn. (4-2).
Figure (5.9) shows computed values of hydraulic conductivity for the Lurgi
RB-II material. The conductivity for this sample varies over eight orders of
magnitude from 10~7 m/sec to 10 1>s m/sec at the critical water content.
I
5.1.2 I^vf Test Results •
Table (5.3) presents a summary of the relevant test parameters for the
two Lurgi ESM tests performed in the development stage of this study. The
same column was used for both tests so that the packed lengths and cross-
sectional areas are the same and test parameters are nearly identical. These
data are used in Chapter 6 to investigate the reproducibility: of the ESM
column test.
TABLE 5 .3 . ESM TEST PARAMETERS FOR LOB3I ULG :
Run No.
Packed length, L(mm)
Cross-sectional area, A(mm2)
Bulk density, pb (kg/m3>
Porosity, p
Initial volumetric
moisture content, 6^
Max. moisture content, 6m
Inflow Darcy flux, g^dnm/h)
33
298
8012
1609
0.404
0.159
0.354
8.57
34
298
8012
1606
0.405
0.162
0.345
8.61
Leachate samples from the leach tests were chemically '.analyzed for
several inorganic constituents. Concentrations of major ions F, d, S04, sa,
Ca, J*j and Mo were measured for all leachate samples. Tables (5.4) and (5.5)
present results for the major-ions. for-, each leach test. The column listing
the cumulative volume of leachate in each table represents the total volume of
leachate collected up to the mid-point of the sample taken. ;These results
show that leachate concentrations remain fairly constant for the first few
samples and then decline. Elevated levels of EC and sulfate concentration at
the end of the test indicate long term equilibrium between the leachate and
solid., High concentrations of sulfate and calcium suggest equilibrium with
solid phase calcium sulfate. Ions like chloride and sodium, ^however, are
readily dissolved and leached. Dashed lines in these tables indicate missing
data either because of insufficient sample volume or obvious analytical error.
50
-------
M
i-t
cn
•~i
'o
g
(S/UJ3) X
S-l
O
4-4
-U
O
3
-a
0)
a
51
-------
TABLE 5.4. EFFLUENT CONCEMTRailCNS FOR HJN NO. 33
(LDK3I DLG)
Cum V
ml
11.5
84.5
152.9
221.1
300.7
368.8
436.5
504.5
584.7
653.3
727.7
797.9
1121.7
Cum V =
EC-25°C
EC-25°C
dS/m
15.81
15.81
15.91
15.71
14.52
12.03
9.47
7.36
5.59
4.71
4.18
3.91
3.48
cumulative
F Cl SO. Na Ca Mg Mo Vt/L
mg/L . ;
17.8 355 9,650 3,830 519 0.619 5.53 '0.013
13.8 329 9,620 3,820 446 0.444 5.74 0.100
11.6 329 10,940 3,530 450 0.367 5.27 0.181
10.9 336 9,760 3,780 469 0.419 5.46 0.262
9.65 289 8,870 3,380 507 0.381 3.30 0.356
9.44 236 7,160 2,450 503 0.383 3.48 0.463
8.52 176 5,200 1,850 514 0.427 1.93 0.516
7.72 84.9 4,060 1,090 536 0.430 0.758 0.597
6.66 79.1 2,980 659 565 0.455 1.59 0.692
7.29 63.4 2,560 413 560 0.479 0.674 0.773
6.34 36.6 2,290 400 576 0.393 0.395 0.861
5.57 28.3 2,130 239 637 0.422 0.724 0.944
5.00 14.4 2,010 151 619 0.561 <0.05 1.327
volume of leachate
= electrical conductivity standardized at 25°C
1 deci siemens/raeter (dS/m) = 1 mmho/cm ,
TABLE 5.5.
Cum V
mL
12.4
83.9
156.4
227.1
309.9
380.2
451.0
520.4
600.3
668.8
738.0
806.2
1105. 6
EC-25°C
dS/m
14.91
14.81
15.50
15.40
14. 09
11.04
8.26
6.44
5.11
4.51
4.17
3.88
3.73
EFFLUENT CONCENTRATIONS FOR RUN NO. 34
• (LUEGI ULG)
F Cl SO. Na Ca Mg Mo vt/L
mg/L ;
13.0 341 9,840 3,680 466 0.478 7.24 O.'OIS
12.5 250 10,000 3,830 469 0.558 9.02 0.102
11.7 326 8,180 3,780 483 0.487 7.46 0.190
11.0 326 8,080 3,770 477 0.449 5.60 0.276
10.0 296 7,050 3,400 490 0.394 6.44 0.376
8.73 217 6,410 2,630 460 0.482 5.57 OJ461
6.95 140 4,730 1,500 518 0.518 2.27 0.:548
6.19 88.3 3,900 1,170 554 0.554 2.06 0.632
10.8 — 2,160 604 565 0.550 1.02 0.729
4.83 .84.8 2,130 498 383 0.616 0.884 0.810
6.14 41.9 2,440 312 576 0.649 <0.05 -O.J396
4.75 22.1 1,990 215 584 0.712 0.979
4.24 8.08 2,100 126 597 0.756 0.446 1.342
Cum V = cumulative volume of leachate
EC-25°C = electrical conductivity standardized at 25°C
1 decisiemens/meter (dS/m) = 1 mmho/cm
52
-------
5.1.3 ISM Test Results
Three ISM tests were-run-using two. different Lurgi spent shale samples.
Table (5.6) summarizes the leaching test parameters. Lurgi ULG was packed, to
about the same. .density in.. Runs 2 .and 3 (1806 kg/m and 1794 kg/m,
respectively) but the column size and flow rate were varied. In run number 2,
the column was 628 mm long and the flow rate was 13.4 mm/hr. In run number 3,
the column was only 442 mm long and the flow rate was increased tp 19.4 mm/hr.
These changes resulted in a residence time in run 2 over twice as! long as that
for run 3 (46.9 and 22.8 hrs, respectively). ;
Tables (5.7) to-(5.10) show the results for the three ISM tests. The EC
breakthrough data for these runs are given in Figure (5.10). The EC values
have been normalized by dividing through by the maximum observed EC value in
each run. The EC breakthrough data show a sharp decrease to a nearly
constant, low fraction of the maximum value. This very sharp decline in EC is
believed largely controlled by hydrodynamic dispersion. Figure (5.11)
illustrates seme of the differences in concentrations of individual species in
the effluent using data from run no. 2. Note that the relative concentration
of chloride declines most rapidly followed in order by sodium, sulfate.
carbonate and-calcium;- ^3Sae concentration of calcium, while showing a small
decline initially, maintains a nearly constant value consistent with the
concentration of a solution .in equilibrium with solid phase CaSO.. The
concentrations of sulfate are much larger than expected if solid phase CaSO.
were the only source, however. The data suggest that the minimurti
concentration of sulfate observed may be controlled by solid phase CaSO
solubility. This was also found in the Lurgi ULG ESM tests and concentration
values of Ca and SO..are nearly the same in the two different test effluents.
The data for Runs 3 and 5 show the same characteristics as those;observed for
Run 2. The consistent and orderly difference between the breakthrough curves
for Cl, Na, SO4, C03,""ahd Ca suggests a "progressively more important control
exerted by chemical reactions.
5.1.4 Codispos^1 Test Results :
The Lurgi RB-II spent shale material was used in the codisposal ESM
tests. Three tests were run. Moisturizing solutions used were distilled
water, a stripped and an unstripped process water. Table (5.11) summarizes
relevant test parameters. | • . .
5.1.4.1 Inorganic Analyses '
Table (5.12) summarizes the results of the inorganic analyses for
the Lurgi material. Complete results are given in Appendix A. The
constituents of the process water and the concentrations found in the first
and last effluents 4at-about ..0.1..and 2.25 pore volumes average cumulative
effluent) are given. Note that a significant difference in leachate
composition between runs l and 2 is the amount of ammonia in the process water
and initial leachate. The unstripped moisturizing fluid has :much higher
ammonia and alkalinity levels. However, the stripped water has higher sodium
and sulfate concentrations. The difference in ammonia concentration affects
53
-------
TABLE 5.6. SUMMARY OF LURGI ISM TEST PARAMETERS!
Run No.
2
3
5
Material
L
(mm)
Lurgi ULG 628
Lurgi ULG 442
Lurgi RB-I 432
(kg/m3)
1806
1794
1225
0 V
(mm/hr)
0.303
0.307
0.538,.
13.4
19.4
11.2
Maximum
EC (ds/ia)
1 24.5
: 22.8
55.3
-,--.- • - . - .-.._- . - • . i . - •
TBBLE
5.7.
MAJOR
ION
i
COMPOSITION OF LUEGI (ULG) i
EFFLUENT - HJN NO
vt
L
0.020
0.058
0.092
0.186
0.236
0.292
0.350
0.378
0.452
0.535
0.590
0.652
0.737
0.860
1.06
Ca
575
560
540
520
530
540
560
570
580
600
590
590
600
610
590 .
Mg
' 0.3
0.2
0.4
0.6
0.7
0.7
0.8
0.7
0.7
0.7
0.6
0.7
0.7
0.7
0.8
Na
11970
11270
10590
5410
3150
1660
1000
795
600
575
535
405
530
385
365
K
950
760
830
300
220
160
120
110
110
100
110
100'
110
80
120
a
mg/t
1360
1300
1150
450
230
131
75
65
45
48
. 39
30
41
26
23
. 2 (ISM) ;
HC03
136
67
57
29
25
26
26
22
21".
21
20
18
23
20
20
C03
393
351
328
208
149
123
101
111
105
118
90
93
105
136
137
SO,
23900
23400
21900
12600
5590
5450
3390
3480
3180
3110
3040
2760
3180
2770
2750
pH
10,83
11.09
11.13
11.22
11.15
11.05
10.96
11.07
11.08
11.12
11.02
11.07
11.02
11.21
11.20
54.
-------
TABLE 5.8, CONCENTRATIONS OF SELECTED TRACE
LDB3I
-------
TBBLE 5.9. MAJOR ION COMPOSITION OF LUHGI (ULG)
EFFLUENT - HJN NO. 3 (ISM)
vt
i.
O.QIO
0.028
0.049
0.070
0.092
0.114
0.132
0.152
0.170
0.189
0.208
0.226
0.245
0.263
0.282
0.300
0.318
0.337_r
0.355
0.531
0.679
Ca
' SfflT.
580
575~
555
560
555.
540
540
545
540'
530
535
545
560
560
580
570
: r.590 '
590 .
615
630
Hg
' 0.-5- ~
0.4
0.4
0.3
0.4 .
0.4
0.3
0.4
0.5
0.5
0,5
0.5
0.5
0.6
0.6
0.6
0.6
. 0-.6--..
0.5
0.6
0.7
Na
-" 12850 -
12980
12475
11015
10115
8380
7310
6300
5240
. 4745
4065
3255
2790
2245
1915
1790
1400
11280-,..
1085
410
235
• K Cl
mg/t
• -•- 900 '--• 1685"
820 1635
870 1530"
860 1320
1120
935
740
605
495
435
345
290
240
190
160
110
90
---r. -\: .80:--.
63
21
12
HC03
'-89
94
68
91
58
68
49
40
43
38
33
34
39
29
40
30
33
.28.
28
25
26
co3
" 439 '
410
363
346
293
260
235
213
208
179
164
144
132
122
108
103
88
V_-:92"
99
80
79
so4
30800
31200
30000
27300
25400
22700
19800
17400
15000
13200
11900
10300
9100
7420
6500
4810
4160
3910:.- '
3990
2335 .
2075 =•
PH
ii.be
11.01
11.10
10. bs
11.07
10.95
11-05
11.09
11.05
11.04
11.06
10.99
10.90
11.00
10.80
10.90
10.80
10,88
1
10.91
10.87
10.35
56
-------
TfiBLE 5.10.
SELECTED MAJOR ION CONCENTRATIONS IN
LDBGI RB-I - HJN NO. 5 (ISM)
vt
I
0.005 •
0.016
0.032
0.048
0.060
0.075
0.090
0.102
0.116
0.130
0.144
0.158
0.168
0.183
0.197
0.209
0.220
0.234
0.245
0.258
0.272
0.284
0.291
0.329
0.384 ,
0.436
0.621
Ca
535
510
525
520
510
495
495
490
510
490
485
475
465
490
485
485
495
495
510
515
505
525
530
545
545
560
575
Na
18770
17890
16400
14000
11750
10040
8270
7370
6220
5475
4480
3835
3145
2720
2380
2060
1875
1605
1410
1255
1110
1050
920
750
540
; ,450
325
C1
2250
2010
1690
1260
1080
960
750
580
480
390
. 310
270
210
180
150
130
. 110
90
72
57
49
48
41
36
26
23
15
co3
775
733
717
613
567
517
473
427
388
375
336
317
298
291
276
255
251
242
237
233
227
221
223
214
212
207
203
»,
34000
32400
30500
25800
24600
18100
16100
14300
13200
11300
10100
9310
7530
6980
6200
5400
4960
4610
4220
3900
3790
3440
2090
2750
2440
2320
2070
PH
12.24
12.17
12.20
12.17
12.12
12.18
12.07
12.05
12.04
12.01
11.99
11.96
11.94
11.92
11.92
11.87
11.90
11.83
11.84
11.83
11.83
11.81
11.32
11.78
11.79
11.77
11.83
57
-------
OS
O OGr
LU
o
UJ
0-4
0-2J-
Run no. 2
Oi
It I I
0 0-2 0-4
0-6 0-8
Vt/L
1-2
l-O
0-3
o 0-G
CJ
U)
0-2
1-0
0-8-
o
UJ
0-4
0-2
Run no. 3
O 0-2 0-4 O-6 0-3 1-0 1-2
Vt/L
Run no. 5
0-2 0-4 0-G 0-8
Vt/L
1-0 - 1-2
Figure 5.10. Normalized EC for Lurgi ISM tests.
58
-------
I-W
0.8
0.6
JL
Ci
0.4
0.2
o,
_
• § .«'»•••
* • »
d 3 •
g »• " RUN No. 2 i
0 CI , c; = 1360 mg/l
• S04, c; =23900 mg/l
A No , Cj = 11970 rhg/l
• Co, cf = 606 m:g/l
° C03 , cj = 393 mg/l
9
A
a :
O D °
U ,
o
a o a n :
• Q a :
o •
0 AA *"*'"• •!
n ^ •
°' ° A , ^ , A , *
0.6
yf
L
0.8 1.0
Figure 5.11.
Major ion composition of etfluenfr
Run No. 2 (iSMXLurgi ULG).
59
-------
TABLE 5.11. PARAMETERS FOR LURGI RB-II CODISPOSAL TESTS
Run No.
Moisturizing
Fluid
Packed Length, L(mm)
Cross-sectional Area, A(mm2)
Bulk Eensity, p. (kg/m )
Porosity, p
Initial volumetric moisture
content, Q^ (% by weight)
Maximum volumetric
moisture content, e
m*
Flow Rate (ml/hr)
Inflow Darcy flux,
q^cm/hr)
Total time of test, hrs
Time after beginning of
outflow, hrs
1
Unstripped
Retort Water
578
20600
1410
.49
.198
(14.0%)
.424
447
21.7
41.7
27.3
2
Stripped
Retort Water
578
20600
1430
.48
.200
(14.0%)
.435
452
21.9
42.0
27.6
3
Distilled
Water
578
20600
1410
.49
.204
(14.5%)
.419
399
19.4
43.7
30!.2
the EC levels measured for run number 1 and the relationship between EC and
TDS shown for the unstripped process water and leachates. The retort waters
used iia the Lurgi tests have pH levels of 9.0 and 9.5 for the unstripped and
stripped samples, respectively. The leachates, however, all have pH values
over 11. The high concentration of NH found in the unstripped process water
at a pH of 9.0 will be in the form of4NH if the pH is over 11. Since NH is
neutrally charged*- it does not— "contribute - to the -electrical ^conductivity
value. This is shown by comparing the EC and TDS values for each run, keeping
in mind that most of the ammonia is removed in the stripping process.
All the Lurgi leachate samples had significant boron arid strontium
levels, even though concentrations in the process waters were low. Levels for
strontium are similar in all three runs and seem to approach a constant value
of about 45 to 50 mg/1 in the final effluent, regardless of the moisturizing
fluid used. Boron follows a somewhat similar pattern except that the
distilled moisturizing fluid appears to have caused more boron to solubilize
during the equilibration period, resulting in a slightly earlier leaching
60
-------
01
G
EL
g
K
0)
JJ
ra
u.
s
1
to
OVJOOtncnOOOCOrJ-Ot^rHrtO
•T-t-st-TC iH C» • • • o O 80
CO
«r>«O
• •»
1-1 1-1
CSlrH
O O\ 4
o
Wl
O O CO
O O CO 00 ts O
^^ • • «
O O OO
•"D
8
61
-------
curve. In both cases, however, it seems clear that the source of these
constituents is the spent shale, not the process water, and concentrations are
governed by solid phase solubility constraints. :
Elevated levels of total dissolved solids were found at the end of all
three tests shown in Table (5.12). Hie almost equal levels of about 4000 rag/1
suggest that a long term equilibrium state is occurring between ^the leaching
fluid (i.e. the distilled water) and the solid material. For example, the
similar, high concentrations of sulfate and calcium found at the; end of all
the tests suggest that solubility constraints are occurring with the solid
phase calcium sulfate. Most of the other parameters measured; also showed
elevated levels in the last effluents, suggesting solubility constraints.
Even -the chloride and sodium ions that were readily leached in : previous ESM
tests showed these-effects. :
A feature of the ESM test that is particularly important to the
codisposal tests is that the initial effluent from the column is the
equilibrated, undiluted antecedent moisture. That is, the first;erfluents are
not -the invading leachate but are moisturizing fluid that has equilibrated
with the solid material. A cursory glance at the concentrations: for seme of
the major ions lends support to this theory. Table '(5.13) shews
concentrations for some of the major constituents and the average cumulative
effluent volume when the sample was taken. As an example, consider run no. 1
that used unstripped retort water as the moisturizing fluid. Table (5.13)
shows concentrations for sane of the major constituents and the cumulative
effluent volume when the sample was taken. The total volume of; moisturizing
fluid, 0.AL, calculated from Table (5.11) for this case :is 2358 ml.
Therefore, the first three samples were taken before all the, moisturizing
fluid exited the column and the last three values shown are from samples of
the invading distilled leaching water. The differences in concentrations
before and after the moisturizing water is replaced in the column are clearly
evident. These same trends are also evident in the leaching patterns of other
species given in Table (5.13). ',
5.1.4.2 Organic Analyses
The average total organic carbon (TOC) content in ppm was measured
for each process water and each first and last effluent sample f ran the column
tests. These results are given in Table (5.14).
First leachate effluents, last effluents and process water ;samples were
also subjected to analysis for Volatile Organic Priority Pollutants using EPA
Method 624. Table (5.15) shows the compounds detected in the Volatile Organic
Analyses of samples from Runs 1, 2 and 3; the Lurgi column tests using
unstripped, stripped and distilled water for the moisturizing fluid. In seme
cases, a compound found in the analysis of the sample was found! in the blank.
In the tables, this is shown by a B following the uncorrected concentration
value in parentheses. The calculated concentration is shewn without
parentheses and is the difference between the measured sample concentration
and the concentration in the blank, adjusted by the dilution factor of the
sample.
62
-------
TSBLE 5.13. MAJOR ION COMPOSITION OF LDEGI RB-II EFFLUENT
Run.
No.
1
(e,v =
2358 ml)
2
(e.v =
2381 ml)
3
(e.v =
2430 ml)
cumul
volume
(ml)
60
1192
2292
4872
7655
11611
60
1199
2309
4910
7682
11866
615
2290
4410
7720
11200
zt
L
.01
.24
.45
.97
1.52
2.30
.01
.23
.45
.95
1.48
2.29
.12
.46
.88
1.55
2.25
CL
266
259
209
99
89
74
305
294
181
82
67
'74
145
110
64
53
53
so4
mg/L
4,180
4,460
3,740
1,870
1,870
1,580
8,060
7,490
5,180
1,870
1,440
1,580
6,480
4,320
1,730
1,440
1,440
NH4
8,320
7,270
3,690
41
22
19
251
186
73
16
16
17
92
34
19
21
23
Ca
655
713
819
1,280
1,210
1,200
643
735
832
1,210
1,240
1,200
. 683
861
1,160
1,150
1,120
Na
2,320
2,350
1,730
240
160
172
5,170
4,410
2,500
311
181
166
3,190
1,920
332
217
228
K
327
303
303
68
41
37
534
470
298
62
40
42
629
456
146
57
58
Si:
26.30
27.30
31.10
45.80
48.40
44.60
32.70
33.30
36.00
50.00
51.20
49.60
31.90
37.70
50.60
52.10
49.00
63
-------
TABLE 5.14. AVERAGE TOTAL ORGANIC CARBON FOR LORGI RB~II
CODISPOSAL TESTS
Run No.
1
2
3
Moisturizing
Fluid
Unstripped
Retort Water
Stripped
Retort Water
Distilled
Water
Total Organic Carbon
Moisturizing
Fluid
5,470
3,700
— —
First
Effluent
4,380
2,660
<5
in PPM
Last
•Effluent
36
;
54
l
<5
Some concentrations given are also followed by the letter J. This designation
means that the compound was found but accurate quantification was difficult.
In Table (5.15), the J designation means that the compound was found but, (a)
it was near detection limits, or (b) the compound was also found in the blank
and the nutiber is approximate because of the large dilution factors required.
For example, the first effluent sample for run no. 2 was diluted' 25 times and
2-hexanone was found in this diluted sample at a concentration 3.5 jig/1. The
detection limit for 2-hexanone is listed as 10 jig/1 (see1 Table 4.4).
Therefore, the concentration calculated for an undiluted sample (;88 ng/1) only
indicates that this compound was found in the effluent but the concentration
given is approximate. ;
In the case of methylene chloride in the unstripped retort water sample,
for example, the J designation is because of the large dilution factor and
because this compound was also found in the blank. At a dilution of 1:100,
the concentration of methylene chloride in the diluted sample was found to be
1050 ng/1 but a concentration of 12 ng/1 was found in the ' blank. The
difference of 1038 jig/1 multiplied by 100 is the value given in the table.
Therefore, methylene chloride is present in this sample at a concentration of
"about" 103,800 jig/1. The detection limit for methylene chloride is 5 ng/l
so the concentrations are not near detection limits.
!:
Note that the concentration of methylene chloride in the stripped retort
water is >l,000,000 ng/1 or >1000 ppm and appears to leach readily. In the
first effluent of run 2, the concentration of methylene chloride is >25,ooo
jig/1. Concentrations of acetone, 2-butanone and 2-hexanone were also
significant in this sample and the corresponding first and last effluents but
the t concentrations of these compounds were even higher in run no. 1.
Concentrations of the volatile organics in run no. 3 were found to be small as
would be expected. '
The Lurgi samples were analyzed for semi-volatile organics using EPA
Method 625, a gas chromatographic/mass spectrometer (GS/MS) method. In Method
1 . ". 64 '
-------
TABLE 5.15.
RESULTS OF VOLATILE ORGANIC ANALYSIS FOR LDRGJ TESTS
(EPA Method 624) ;
Run No. 1 (Lurai-Unstri
Compound
methylene chloride
acetone
2-butanone
1,1 ,.l-trichloroethane
2-hexanone
4-methyl-2-pentanone
Dilution
Concentrations (ng/L)
pped Retort Moisturizing
Unstripped
Retort Water
los^oouos.oooj.B)1'2
88, 800 (90, 000 J,B)
132,800(134,0006)
20(130J,B)
10,500
780
1:100
.
tfater) 1
First i
Effluent
72,OOOJ
322,000
37,100(39,4006)
2,000 •
1:100
Last:
Effluent
7
1,270
357(380B)
100
5
MA
Run No. 2 (Lurgi-Stripp
Compound
methylene chloride
acetone
2-butanone
2-hexanone
Dilution
ed Retort Moisturizing Ws
Stripped Retort
Water
>1, 000, 000
11,200(15,6006)
3,700(6,400B)
580
1:100
ter) :
First :
Effluent !
>25,000 I
9, 550(10, 650B)
1,185(1, 86 OB)!
88J
1:25
Last
Effluent
7.5
146U90B)
73UOOB)
8.7
NA
Run No. 3 (Lurgi-Distil
Compound
methylene chloride
acetone
2-butanone
Dilution
led Water)
Distilled
Water
NA
First
Effluent
5 ;
22
0(20B) ;
NA
Last
Effluent
14
2 (256)
NA
B indicates compound was detected in the blank. Numbers in parentheses are
uncorrected. Numbers outside parentheses have been adjusted.
J indicates concentrations near detection limits.
65
-------
625, the methylene chloride extract is dried and concentrated to; a volume of l'
milliliter. However, because of high concentrations, undiluted samples could
not be concentrated to the one milliliter in the final extract as specified by
EPA 625. Samples were initially concentrated to a factor based on original
sample volume. All sample extracts (base/neutral and acid) were then screened
by GC/PID (Flame lonization Detector). Based on this 'screening, final
dilutions were calculated. The acid fraction was analyzed separately from the
base neutral fraction to reduce the complexity of the chromatograms. Results
are shown in Table (5.16).
Computerized library searches using a combined wiley-KBS Library were
also performed on all Lurgi RB-II samples to identify major •non-Hazardous
Substance List compounds present in each sample, The results of! the library
search are shewn in Table (5.17), but these values should be regarded as
tentative since they are based solely on library matches. Quantification
reported for the tentatively identified compounds are based on:relative peak
areas of the nearest internal standard and assume a response factor of 1.
Absolute identification and quantification of these compounds would require
analysis of the appropriate standards. In some cases, the library search
showed significant concentrations of unidentified compounds. These are given
in Appendix A. ; '
5.1.5 ASTM and RCRA Test Results -•
Three of the Lurgi retorted shale samples were subjected 'to the ASTM
water shake and RCRA extraction tests. The results of the chemical analyses
of the filtrates for each test are given in Appendix B. This appendix lists
the ion concentrations determined by the laboratory analyses. 'Tables (5.18)
and (5.19) summarize these same results in terms of soluble mass, i.e.
milligrams of the ion per gram of material. In this form, it is possible to
compare the two batch tests used. From the results, it can be seen that the
RCRA test is a much more severe extraction test. Addition of the acid in the
RCRA test causes dissolution of more of the carbonates. !
i
A modified ASTM batch leaching test was also performed with the Lurgi
RB--II sample. A 150 gm sample of Lurgi spent shale was added to three liters
of distilled water (20:1 water to shale ratio) in a stainless steel container.
The mixture was gently stirred by a mechanical mixer to prevent:any cementing
of the particles. After 24 hours, a one liter water sample was filtered
through a 0.45 micron filter. The results of the chemical analyses of this
filtered sample are shown in Table (5.18). The total soluble mass was 2.1% of
the retorted shale. The ions shown account for 79% of the total dissolved
solids. Most of the dissolved material is calcium and sulfate. :
, !
The 20:1 liquid-to-solids ratio used for the Lurgi RB-II. material is
larger than the 4:1 ratio recommended in the ASTM procedure. It appears that
the higher liquid-to-solid ratio results in greater calcium concentrations.
This long term solubility of calcium was also seen in the results of the
leachjJig tests and it is unlikely that even the 20:1 ratio used in this last
batch test is sufficient to relax all solubility constraints.
66
-------
EH
M
a.
P
i-3
as
g
CO
M
a
1
5J -•••*
«
"5 2
.7] «^3
O -Ij
7-g
S s*
CM O-i
w 2
a~
03
EH
U-j
K
a
" •
vo
rH
v»
EH
CO
-
3
Q"l
SMI*
O
•H
2
§
8
rH
|
I
C
-p a
CQ ^
(U r-
j_n j^
jj
tii
*s
•g
II
" H_
K
•g
w S
T3
& ^
CQ
.y
4J §
tt 1
cy
"c
4-i rti
u -*
US
u
II
c
a
f(J
I
•|
^3 (^ C3 C5 Q /^
SsS «^ ^S ^E ^3 *2
SSiiii
o
^0 Q O O 0 m
OS gO r-j 3- «
"" •> rH CO
rH
O . O
O 2 0 0
<=> Q Q ° ° Q
•O g « CO CO g
•* 5 ^ «
§ 00
X Q o o o o
» g •<*• CO VO CO
s * ^ * *•
O 0
O Q O O O O
CO CS
0 0 * *
CO ^ rt 2 ^
rH
O
CU
e1"1 "<^ "S
S-5
§ fi'&'fi
O rH rH CD
C 0 rH ^^-S
"•"* -43 O 4J 4J Q
i^ M O4 CM •<*• CM
, O
rH ^
^ s •
rH °
I s
O O
O >O
rH 04
0 0
O U1
rH
-------
TABLE 5.17. ESTIMATED CONCENTRATIONS FROM LIBRARY SEARCH*
(Base Neutral Compounds)
Run No. l
Compound
Dnstripped
Retort Water
(ng/L)
First :
Effluent
(tig/L)
Last
Effluent
ng/L)
methyl pyridine
methyl-2-cyclopenten-l -one
dimethyl pyridine
methyl-2-cyclysenten-l-one
trimethyl pyridine
dimethyl-2-cyclopenten-l-one
dimethyl piperidine
55
74
61
128
126
37
27-28
26 i
25 !
28
35 :
41
28
45
n
Run No. 2
3 -methy 1-2-cyclopenten-i-one
d:unethyl-cyclopenten-l~one
dimethyl piperidine
2-pentenone
methyl-2-cyclopenten-l -one
Stripped First , Last
Retort Water Effluent: Effluent
((ig/L) (|ig/L) (ng/L)
125
58
36
102 |
74
18
* These are estimates only. In some cases, unknown ;
compounds were found at significant concentrations. Apr
pendix A contains more complete results.
68
-------
TABLE 5.18. ASTM BATCH LEACHING TESTS - LUBGI
Parameter
KG-I*
DLG*
RB-I*
RB-II**
F
ca
N03
so4
NH4
Si
Na
Ca
K
Sr
TDS
.008
.095
.035
10.18
.034
.022
1.296
3.00
.172
.056
14.80
.015
.038
.007
5.204
.025
.014
.720
1.608
.136
.052
8.560
.025
.444
.010
9.164
.022
1.100
2.856
.256
.080
13.40
__«_
: .is
.02
10.84
.
.44
5.62
; .04
:
21.8
•Values are mg of solubilized species per gram of solid.
**20:l liquid to solids ratio for RB-II, 24 hr leach.
TABLE 5.19. LDEGI RCRA TEST RESULTS
• • • • • 1
Parameter
a
NO-
NE*
B 4
Mg
P
Si
Na
Ca
Ba
K
Sr
TDS
RG-I*
.142
.031
13.80
.066
.010
5.80
.008
.012
.860
19.28
.064
.178
113.8
ULG*
.128
.010
8.400
.022
5.38
.014
.140
.560
25.60
.007
.100
.260
130.4
RB-I*
.378
.011
17.6 j
.029
8.60
.014
.158 !
1.10
29.58
:
.220
.260
170.4 !
*Values are mg of solubilized species per gram of solid.
5.2 TOSCO II ' ' I
5.2.1 Physical and Hydraulic
The TOSCO II spent shale was found to have an apparent particle density
of 2600 kg/m using the pycnometer method. The test was repeated until
cal.cul.ated values agreed to three significant figures. Figure .(5.12) shows
the particle size gradation curve for this sample. Particles larger than
0.061 mm (no. 250 sieve) were separated by dry sieving and a hydrometer
analysis was performed on the smaller fraction. '
69
-------
100
60
-------
The water retention characteristics for the TOSCO II sample are shown in
Table (5.20) for various capillary tension heads and bulk densities. These
values are equilibrium moisture contents for the main drying curve -from
natural saturation and were obtained by the standard pressure plate proce3ure
described in Chapter 4 and ASTM D23 25. \
TABLE 5.20. WATER RETENTION CHARACTERISTICS OF TOSCO II SPENT SHALES
DURING DESATORATION :
....
Compaction
None
pb -
j>-
pb =
1300
1450
1600
kg/m!
kg/m?
kg/in
l
48
42
36
34
.0
.2
.0
.6
Pressure (bars)
3 5.1 10
45
42
33
33
.8
.0
.8 -
.5
45
41
32
32
.9
.9
.9
.1
43
41
32
30
.2
.8
.6
.1
.8
14.2
: 44.7
: 41.4
30.5
30.5
r
* Table entries are moisture contents expressed oh a % weight basis
(weight of water per weight of dry solids times 100). j
The relationship between capillary tension head and moisture content on
the wetting curve was also investigated for the TOSCO II spent i shale sample.
This relationship-is particularly relevant to-interpretation-of the ESM column
tests because the leaching test is conducted by injecting into an initially
moist but unsaturated sample. Therefore, if hysteretic ;effects are
significant, the , .column.. tests__will.._ not be ..interpreted ..correctly using
desaturation data. In the field, a corresponding situation will occur (i.e.
infiltration of irrigation or rain water into an initially moist but
unsaturated disposal pile). Ideally, the pressure-saturation experiment would
be conducted on a sample initially packed to the same bulk density and at the
same initial moisture content as that used in the column leaching tests and
expected in the field. This, however, proved to be a major difficulty in the
procedure because the test cell for the saturation experiment is i considerably
smaller than the columns used for the leaching test. For this reason, a
sample of the material was wetted to a moisture content less than that used as
an initial moisture content in the ESM tests but packed to the '• same dry bulk
density. Moisture was added using a syringe via holes provided for that
purpose in the test cells. The procedure is described in more detail in
Chapter 4.
Results of the pressure saturation experiment are shown on a semilog plot
in Figure (5.13). The visual best fit lines are also shown and are given by
h = 1403.5 exp (-7.91 6) 61 0.497
j
h = 4.09 x 10~19 exp (-85.3 0) 0 > 0.497 ;
where h = capillary tension head in cm ;
6 = volumetric moisture content. i
' 71 ' '
-------
.an average value for the saturated hydraulic conductivity ;of TOSCO II
spsnt shale is 6.5 x 10 m/sec. This value was determined from several
standard falling head permeameter tests without overburden loads. Table
(5.21) summarizes the results for falling head and constant head tests with
various overburden loads and moisture contents. Results of• standard Proctor
tests are also shown. !
TSBLE 5.21.
SUMMARY OF HYDRAULIC COMXJCTIVITY MEASUREMENTS
-------
a.o • «•>
TT
°° , I?
00 ^Tc
—
0 , 2 "3
•o "S
' ' c
as ° - .
U '•=
a , 2 °
tn o
a o a
en
03
i . 03 '......
O
CD
O '
o a
-L ! I 1 "" 1 !..,-.. , |
!jj lo ' o—~^-o-i-i — : — : : . — _,
O
o -
b '
rO '
0
S '.
0
1 !
3 i
O ^S '
o
CM
O ;
2 '
C3 i
O
o
I— I
g
c-1
1-4
O
M-l
j><
•H
^
•H
CO
'-V-l
•H
O
•H
rH
3
cd
•0
•
1— 1
0)
i-)
(9)
8
U-l
O
I
U
3
4->
8-
3
W
CQ
(U
00
(U
3
73
-------
0.15 0.20 0.25 0.30 0.35 0.4O 0.45: 0.50
Figure 5.15. Hydraulic conductivity for TOSCO II.
74
-------
is used as an upper limit to the curve. Calculated K(6) values are larger
than the saturated hydraulic conductivity when 6 is near the maximum moisture
content because of uncertainty in the diffusivity and pressure-saturation-data
is this region. ;
5.2.2 ESM Test Results .
Table (5.22) summarizes the relevant parameters for the 'two TOSCO II ESM
tests conducted during this study. The same column was used for both
experiments. The change in packed length shown in the table is due to a
slight; change in the perforated end plate. The primary difference between.the
two runs is that the column was packed less densely in run 29. This results
in the larger porosity and smaller initial volumetric moisture content shown.
TABLE 5.22. ESM TEST PARAMETERS FOR TOSCO II ••
Run NO. 29 35
Packed length, L(ran)
Cross-sectional area, Adran2)
Bulk density, pfa (kg/m3)
Porosity, p.
Initial volumetric
moisture content. ei
Max. moisture content, 6
Inflow Darcy flux, q^ran/h)
300
SOI 2
1187
0.554
0.142
0.470
9.46
298
8012
1269
0.512
0.188
0.450
9.99
The results of the inorganic analyses of the effluents for the TOSCO II
experiments are shown in Tables (5.23) and (5.24). The EC concentrations are
fairly consistent in the first effluents in each run. This would indicate
that the initial effluent is the undiluted equilibrated liquid content.
Subsequently, concentrations in the effluents indicate that all the ions
measured leached readily except fluoride. There appears to be seme continued
dissolution of this ion into the leaching fluid but the very prevalent long
term solubility constraints seen in the Lurgi data for Ca and SO. are not
evident in the TOSCO II results. :
5.2.3 ISM Test Results '
TWo ISM tests were run with identical TOSCO II material (see Table 5.25).
The column in run no. 4 was packed less densely than the column in run no. 7
with a 80% higher seepage velocity and a, longer column.
Table (5.26) gives the results of the inorganic analyses of the leachate
in run no. 7. However, in both runs 4 and 7, atypical breakthrough curves
were observed. The TOSCO II material is hydrophobic until it is I in contact
with water for a considerable length of time. All other materials tested were
hydrophylic. The hydrophobic property of the TOSCO sample resulted in an
. 75
-------
TABLE 5.23. EFFLUENT ODNCE23TRATIONS FOR RUN NO. 29
(TOSCO II)
Cum V
mL
49.1
85.1
133.1
198.9
259.4
348.9
449.5
587.8
781.2
980.1
1106.0
Lab. EC
dS/m
38.
38.
38.
36.
31.
23.
14.
8.
4.
2.
2.
90
90
80
00
60
60
50
58
28
70
30
F
29.
—
27.
27.
26.
22.
22.
20,
22.
26.
28.
2
-
6
9
0
8
1
3
2
3
9
Cl
—
217
270
251
193
115
48.2
24.2
9.88
9.02
8.21
so4
23,300
—
10,830
30,180
27,070
6,840
11., 580
5,710
2,370
1,420
1,050
Na
mg/L
—
12,700
12,200
11,400
9,620
7,000
6,660
896
664
448
___
Ca
397
368
328
373
332
371
354
285
154
59.8
44.5
Mg
483
536
—
498
457 .
329
233
151 .
76.8
40.8
32.0
Mo
56.4
59.1 .
57.6
49.3
15.3
10.4
6.56
3.71
2.28
1.62
1.45
vt/L
0.'043
0.075
o. lis
0.176
0.230
0.309
0.398
0.520,
0.692
' 0.868
0.9/9
Cum V = cumulative volume of leachate
Lab.EC = electrical conductivity measured directly from test effluent
1 decisiemens/meter (dS/m) = 1 mmho/cm
TABLE 5.24.
EFFLUENT CONCENTRATIONS FOR RUN NO. 35 j
(TOSCO II) ;
Cum V
mL
12.4
76.5
141.4
205.9
273.0
337.9
413.0
489.7
565.2
640.7
716.0
791.6
865.9
938.5
1113.7
1233.6
Lab EC
dS/m
36.50
36.50
36.50
36.40
36.20
32.90
24.70
17.80
12.20
8.77
6.62
5.29
4.26
3.68
2.60
2.23
F
24.9
23.6
20.1
31.7
23.3
20.3
20.6
20.9
22.0
20.1
22.0
23.7
25.1
26.0
28.3
29.2
Cl
218
211
212
204
208
191
105
52.
24.
12.
8.
6.
5.
4.
4.
3.
8
1
l'
93
32
49
69
52
89
S04
28510
28380
30530
29670
27410
20380
16450
9761
7141
5120
3132
2513
2046
1685
1097
965
Na
mg/L
9910
10430
10300
10160
10100
8850
6380
3670
2470
1750
1160
968
772
678
436
371
Ca
368
342
322
348
354
318
274
169
112
79.5
61.7
49.7
40.6
34.9
22.3
15.7
Mg
695
687
680
696
687
599
416
260
164
116
83.3
73.6
56.0
47.2
35.5
33.2
K
109
108
125
162
136
91.5
86.6
43.8
30.6
23.5
15.4
14.8
11.7
11.9
7.66
5.84
Mo
26.3
26.0
25.0
26.0
25.8
24.2
19.2
13.2
9.44
6.97
5.36
4.20
3.29
2.99
2.29
1.73
vt/L
1
0.012
0 . 071
0. 132
0. 192
0. 254
d.314
0.384
0.456
0.526
0.596
0.666
0.737
0.806
0.874
1.037
1.148
Cum V = cumulative volume of leachate
Lab EC = electrical conductivity measured directly from test effluent
1 decisiemens/meter (dS/m) = 1 mmho/cm
76
-------
unstable wetting front during the wetting phase with the consequent nonuniform
arrival of the liquid at the top of the column. This caused an unusual' shape
or the breakthrough curve. Higher injection pressures are also needed,
compared to hydrophylic materials, to overcome capillary effects that axe
retarding rather than promoting upward flow. :
TABLE 5.25. TOSCO II ISM TEST PARAMETERS
.
Run No.
4
7
L
(mm)
625
278
Pb
(kg/m
1280
1460
0
.506
.438
V
(mm/hr)
21.0
11.8
Initial
EC^
(ds/m)
37.2
33.4
The differences in the breakthrough curves for the hydrophobic TOSCO II
material when compared to curves for hydrophyllic materials are discussed in
Chapter 6. However, the effects of this hydrophooic characteristic on tne
breakthrough curve is a major reason why the ESM test is preferable to the ISM
test, in predicting leachate quality from a field disposal pile. It is also
unlikely that the dry initial condition would occur at a field site.
5.2.4 ASTM and RCRA Test Results j
Table (5.27) gives the results of the analysis of filtrates from the ASTM
and RCRA tests for the TOSCO II material, uhese values have been converted to
units ot soluble mass or mg of the ion per gram of sample. The concentrations
in the samples are given in Appendix B.
Two ASTM tests were conducted with two different liquid-to-solid ratios.
The results with the recommended ASTM ratio of 4:1 are given in the first
column. In the second column, the ASTM procedure was repeated on a second
sample with a 20:1 liquid-to-solid ratio. As Table (5.27) shows,'more of the
ions dissolved with the larger liguid-to-solid ratio. It is prooable tnat
even the 20:1 ratio is not large enough to completely relax all solubility
constraints. However, the differences between the 20:1 and 4:1 ratios are
small except for NO3, Mg, p and Al. • !
The results rrom the RCRA extraction test are also given in Table (5.27).
As found in other cases, the addition of the acid causes more of the
carbonates to dissolve during the test. Here the soluble mass of |caj.cium is
two orders of magnitude higher than in either ASTM test. ;
77
-------
TSBLE 5.26. EFFLUENT ODNCENTRATIONS FOR (TOSCO II)
HJN NO. 7
vt
T
0.020
0.050
0.100
0.121
0.264
0.287
0.308
0.333
0.355
0.375
0.397
0.418
0.440
0.464
0.488
0.511
0.534
0.556
0.579
0.629
0.965
1.030
1.120
1.191
1.886
2.013
2.25
Ca
545
540
520
505
495
425
410
455
435
395
415
380
345
352
320
330
300
275
265
252
120
83
90
72
26
14
.13
Na
10095
10060
10540
10570
10520
9815
9355
9440
8590
7965
7420
6705
5770
5490
4885
4185
3744
3425
3075
2605
1280
945
1060
715
485
375
420
Cl
178
172
175
167
154
163
152
139
133
114
100
92
84
74
61
54
49
45
32
•29
19
13
16
12
7
6
7
S04
25000
25200
27400
26700
26800
26800
25500
24100
23400
21500
20500
17000
15900
14400
13500
12100
11200
10300
9400
7920
3210
2470
2820
1820
1010
660
840
F
27.0
29.0
30.0
30.2
39.2
36.3
34.0
23.2
28.4
15.9
17.2
37.3
32.6
33.3
17.7
17.7
18.2
19.7
18.3
19.1
29.2
29.5
29.1
30.5
33.5
38.5
31.1
pH
9.24
9.28
9.29
9.39
9.37
9.31
9.34
9.28
9.30
9.27
9.29
9.26
9.21
9.17
9.25
9.21
9.20
9.13
9.20
9.20
9.10
9.21
9". 24
9.27
9.35
9.38
9.25
78
-------
TABLE 5.27.
TOSCO II
ASTM AND RCRA TEST
RESULTS
, • • ' ,
Parameter
F
Cl
NCu
set
B
Mg
Si
MO
Ha
EL
ca
K
IDS
4:1
.081
.039
.064
4.52
.010
.140
.006
2.18
.009
.124
.032
7.88
ASTM*
20:1
.150
.038
.390
4.76
.010
.280
.006
.007
2.32
.075
.34
.034
10.2
RCRA*
.444
4.58
.013
1.62
1
.010 '
2.62 :
.054
37.44
!
163.6
.... . - • • •-..-. • •- • •.
"Values are rag of solubilized species per gram of solid.
5.3 Paraho :
Various size tractions of a Earaho spent shale were used in the
development stage of this project. This sample, designated by Paraho-I, is a
combusted shale from a pilot plant operated by Development Engineering, Inc.
at Anvil .foints, Colorado. A second sample, Paraho-II, obtained later in the
project rrom Battelle Pacific Northwest Labs, Richland, Wasnington; was used in
the codisposal tests. !
5.3.1 Physical and Hydraulic
The Paraho-I material was separated into three size fractions by dry
sieving. The size fractions are 0.420-1.190 mm, 1.190-2.00 mm; and 2.362-
3.327 mm. Three tests gave values of 2602 kg/m , 2589 kg/m and 2633 kg/m3
for particle density. Particle size is not expected to have any erfecu on
this value. Two exceptions to this might be if there are considerable
micropores in this material or if the different size fractions nave different
mineral or chenical compositions. The presence of non-interconnected
micropore space decreases the apparent specific yield determined. If the
non-interconnected pore space is decreasing the values obtained, 'the effect
would ce most pronounced tor the largest particle size. This relationship was
not observed for the Paraho-I material and, in fact, the largest; value was
obtained for the largest particle size range.
The apparent particle density of the Paraho-II material was found to be
2650 kg/m3. Tnis is the average of four tests witn results that; ranged f ran
2618 to 2679. The particle size distribution trom dry sieving is shown in
Figure (5.16). Holtz (1976) reported a value or 2600 kg/m3 for Paraho spent
snaie.
79
-------
100
80
_
-------
Table (5.28) summarizes the results for Paraho-I for a series of falling
head and constant head permeameter tests that varied dry bulk density and
overburden load. The sample material was packed in the permeameter to a
specified dry bulk density in 5 mm lifts to form a 10 mm test ;section.~ One
difficulty found in the permeameter tests for the coarse grained Paraho-I
material was the prevention of channeling adjacent to the permeameter walls.
To minimize this problem, the sample was packed into the permeameter inside an
inner sleeve that was raised as the packing was done. Between this inner
cylinder and the permeameter wall, minus 10 material was vigorously packed.
TABLE 5.28.
SUMMARY OF HYDRAULIC CDMXJCTB7ITY MEASUREMENTS (m/sec)
FOR PARAHO-I :
Proctor Results
Optimum Moisture Content: 20%
3
Proctor Density: 1700 kg/m \
Moisture
Content %
20.0
20.0
20.0
20.0
10.0
10.0
Dry Bulk
Density
(kq/m3)
1700
1700
1700
1700
1360
1360
Method
Load (kPa)
345
Constant Head
Falling Head
Constant Head
Falling Head -
Constant Head
Falling Head
4.7
5.0
1.9
8.1
9.8
9.8
X
X
X
X
X
X
10"9
10
10 J
10~9
10 °
10"6
4.2
4.6
1.3
8.1
7.0
690
X 10"
X 10
X .10 I
X-10~
X 10 6
1 1300
! 4.5
, 4.0
' 1.3
- 7.8
2.4
i
x
X
X
X
X
10~9
10
10 9
10'9
10 G
5.3.2 ISM Test Results \
Different size fractions of the same Paraho I material were used in two
ISM tests. A summary of test parameters is given in Table (5.29);. Run 6 used
material between 0.4" and 1.2 mm in size-and Run 8 used all particles smaller
than 3.96 mm. The results can be summarized by the effluent concentrations
given in Table (5.3.0) for run no..JL.'....It was observed that the Paraho material
leached at a slower rate than other materials. This is believed to be caused
by diffusion of dissolved species in the stagnant micropores into' the flowing
fluid in the macrospace. This is discussed further in Chapter 6.
TABLE 5.29. PARAHO ISM TEST PARAMETERS ;
Run No.
Particle
Size (inn)
L
(mm)
(kg/m3)
v
(mm/hr)
6
8
0.4-1.2
<3.96
435.
278.
1090.
1203.
.569
.523
17.6
12.0
81
-------
TABLE 5.30. EFFLUENT ODNCENTR5TIONS FOR PARAEO
(0.4-1.2nm) - RUN NO. 6
vt
0.022
0.032
0.047
0.070
0.088
0.105
0.128
0.151
0.167
0.178
0.190
0.202
0.220
0.244
0.263
0.275
0.294
0.312
0.353
0.367
0.412
0.423
0.473
0.483
0.606
0.644
0.726
0.836
0.918
1.642
1.978
2.177
Ca
610
605
610
625
625
640
660
635
635
610
650
605
635
670
660
670
680
620
• 685
665
700
655
715
710
700
730
660
680
670
560
380
390
Ha
1500
1445
1385
1305
1280
1165
1095
1005
955
965
935
895
875
830
785
780
760
705
640
595
5.65
525
490
460
'430
370
340
305
285
140
100
105
Cl
mg/l
49
. 47
47
42
39
35
33
31
30
27
26
26
23
24
27
32
30
23
25
24
26
21
17
19
18
14'
16
13
14
9
6
6
so4
3840
3305
3735
369.5
3525
3485
3320
3270
3180
3185
3175
3085
3060
3065
2975
2930
2890
2890
2815
2605
2615
2530
2500
2395
2390
2290
2205
2105
2045
1500
1000
860
F
21.0
22.2
21.9
22.0
22.6
21.0
21.0
21.5
22.0
20.4
20.9
21.4
21.1
21.8
21.8
22.5
21.9
21.6
14.2
12.4
14.4
14.5
12.7
12.8
12.5
11.5
12.4
11.4
11.4
10.2
8.1
8.3
PH
. 11.55"
ir.eo
11.66
11.73
11.78
11.81
11.86
11.89
11.91
11.83
11.94 :
11.95
11.99
..12.02
12.01
12.03
12.06
11.96
12.06
12.10
12.13
12.07
. 12.14
12.17
12.21
12.24
12.27
12.31
12.35
12.46
12.57
12.60
82
-------
r
Table (5.31) lists relevant parameters for a series of three ESM tests
with the Paraho-Il spent shale material. Moisturizing waters used-were
distilled water, a process water and a gas condensate water. Thei spent shale
was equilibrated with these moisturizing fluids for 72 hours before being
packed into the leaching columns. Distilled water was used as the subsequent
leachant. ;
During the test, a record of pH, EC, cumulative volume out, time and
temperature was kept and the effluent at 5 intervals from .05 to 2.5 pore
volumes was analyzed for inorganic constituents. Ihe process waters and
distilled water were also analyzed. Samples of first and last effluents, in
addition to samples of the moisturizing fluids were taken and analyzed for
organic constituents. j
i
5.3.3.1 inorganic Analysis Results i
Tables (5.32) and (5.33) summarize the results for the Paraho tests.
The first table shows concentrations of some of the major constituents in the
process waters and the first and last effluents. In these tests, the gas
condensate had high ammonia and alkalinity levels and the retort water had
lower ammonia but higher sodium concentrations. PH levels were ! 8.7-8.8 for
the process waters with the EC and TDS values in the gas condensate as high as
68.70 ds/m and 79,200 mg/1, respectively. :
!
Ihe trace metal arsenic is found in both process waters and the majority
of Arsenate is retained by the shale. Other elements such as strontium,
calcium, molybdenum, selenium, potassium and magnesium only appear in the
process waters in very small quantities. However, higher concentrations are
found in the leachate. This would suggest that the source of these
constituents is the spent shale, not the process water. This is also seen by
comparing TDS values. The value of the total dissolved solids in the retort
water is 14,250 mg/L, whereas in the gas condensate, the value is 79,200.
Howeveir in the first effluents, the corresponding runs have TDS values of
59,700 and 65,900, respectively. The total dissolved solids for the first
effluent in run no. 7, which used distilled water for moisturizing, is almost
as high at 52,900. The major constituent is sulfate at levels of 37400, 36000
and 35300 mg/L, respectively. Sodium is also high for all , three first
effluesnts. Boron appears in the retort water at a high concentration but
there is also apparently some solid phase boron. The last effluents for all
three cases suggest that boron will occur in the leachate at a significant
level for at least 2.5 pore volumes. Long term solubility constraints are
evident for many of the compounds in Table (5.33).
The average total organic carbon (TOG) content of each \ moisturizing
water and each first and last effluent sample was measured for the Paraho-n
codisposal tests. These values are shown in Table (5.34) and the; results of
the VOA analysis performed on these samples are shown in Table (5.35).
Samples of the retort water and gas condensate were diluted by a factor of 10.
83
-------
TABLE 5.31. PARAMETERS FOR PARAHO-II CODISPOSAL TESTS
Run No.
Moisturizing
. Fluid
Packed Length, L(mm)
Cross-sectional Area,
A(irm )
Bulk Density, PK
(kg/m3) D
Porosity, 9
Initial volumetric
moisture content, 6^
(% by weight)
Maximum volumetric
moisture content, em
Flow Rate (ml/hr)
Inflow Darcy flux,
g^dim/hr)
Total time of test,
hrs
Time after beginning
of outflow, hrs
5
Retort
Water
413
8030
1510
.431
.218
(14.5%)
.375
232
28.9
17.0
14.1
6
Gas
Condensate
413
8120
1400
.472
.243
(15.1%)
.420
224
27.5
19.3
15.7
7
Distilled
Water
413
7950
!
1420
.465
.193
(13.6%)
.396
221
27.9
19.2
15.9
84
-------
1-H
M
M
Ol
H
e
£
S3
u
s
—"I
vo
in o oo >n
o i>
o o
rH O f-
io rP
O •
o
o
rHVOmtO
CO
o o
. TJ- O
CO • :O
o
o
<0
0) O
a -H
--! -U
«M (0
03 O
M C
VH r9
pt4 O
NO
"* rH «« CS rH rH
CO rH
(SVOrH
OOVOOOXUIVOOOOCO O O
CO
o
o
.00
o
o
ox
r-
o
o
o\
ca
§
es
CO
"8
r-i rH
r 8
85
-------
1
a
PARAHO-II EPEl
1
1
8
M
1
CO
CO
10
C-3
*
£
S1
3
•<*
^
a
•a*
,_, CS
(D"*""
O co es c— co
CO CS ON rH ON
es oo co cs
H
O O 0 O Tj-
rH rj> 10 CO •>»•
VO Tf CS rH
O O P- CO t-
co es v> o vo •
O 10 ON VO CS
* «k
CO CO T CS rH
ON 10 00 Tt rH
CS CS rH rH rH
O O O O OO
O CO P- VO CS
co 10 o o co
t* 10 co es
o o o o o
O O O rC rj-
O CS IO VO IO
vo 10 10 co es
CO CS rH
co 10 o -"t- o
f ON rH CS 10
10 CO CS rH
10 ON o r- T
O "*• 0 rj- Tf
rH rH CS
10 O C- VO VO
r- ON o vo co
VO Tf O •*
* <* •*
rH CS CO
VO > rH
•r-eo
O
O C— CO O\ t->
T»- CO rH O OS
I- vo co es
rH
O O O rH 00
O IO T rH IO
T rH ON I- CS
rH rj- rH
o r- es es co
o in co es co
i-H ON vo •* es
es
CO CS 00 OO C-
CO CO •*• c- O
•o 10 10 m \o
CO VO O\ rH f.
>O ON O ^f IO
10 CS CO rH
0 O O 0 0
O O VO VO rH
CO OO O VO CO
10 in oo rj- co
CO rH
T o in cs
eS rH
vo ON es O O
o in o vo vo
rH rH CS
CO rH 00 10 CO
p~ vo CS rH •*
C» CO rH TT
rH CS CO
P» ^ CO
6*° -
86
-------
The values in parentheses followed by the letter B indicate that the compound
was found in the blank. Numbers in parentheses are uncorrected for this.
Numbers not in parentheses have been adjusted. For example, consider run no.
6, analysis of the gas condensate. The process water was diluted by a factor
of 10. Therefore, the concentration of, for example, acetone found in the
sample analyzed was 3610 fig/L. However, the blank on that day was found to be
contaminated with acetone at a concentration of 25 ng/l. -The difference
multiplied by 10 (to account for the dilution) is reported as 35,850 |ig/l in
Table (5.35). A number followed by the designation J means that the compound
was found in the sample, but the concentration was near the detection limit so
quantification is approximate.
TSBLE 5.34. AVER2GE TOTAL OEGANIC CARBCN FOR PARMO-II ;
CDDISPOSAL TESTS
Run No.
5
6
1
Moisturizing
Fluid
Retort Water
Gas Condensate
Distilled
Water
Total Organic Carbon
Moisturizing
Fluid
2,960
4,140
First
Effluent
1,950
1,820
531
in PPM
Last
Effluent
;s6
118
12
- - - \
Considerable differences in leachate composition are shown in Table
(5.35). The highest concentrations of volatile organics were found in the gas
condensate moisturizing water. Acetone, 2-butanone, and 2-hexanone were found
at concentrations of 35850, 5170 and 1200 ng/L, respectively. The first and
last effluents also had significant concentrations of these compounds.
Benzene, 4-methyl-2-petanone, thiophene and toluene were also found at
significant levels in this moisturizing water. Tetrahydro thiophene and
ethylbenzene, although found in the sample, are in concentrations near
detection limits and were not found in the column effluent samples.
In run no, 5, acetone was found in the effluent, but not in1 the retort
water used for moisturizing. This was also true for run no. 7 where distilled
water was used for moisturizing. The source of this acetone could be -the
shale, contamination of the GC/MS system or contamination of. the leaching
column. Acetone was found in the blanks in each of these runs. Also, neither
sample was diluted and concentrations are not near detection;limits so it
seems probable that acetone was, in fact, present in the samples taken from
the columns. The leaching columns could also have been contaminated but a
sample of the distilled water was run through the test piping system and saved
for analysis. If the source of the acetone was a contaminated column or
contaminated distilled water, it should have been detected in the'analysis of
the distilled water. Therefore, it appears from the data that the shale
itself is the source of the acetone. However, the possibility of
contamination should be carefully examined again in further tests.
87
-------
TABLE 5.35.
RESULTS OF VOLATILE OKGANIC ANALYSIS FOR PARAHO-II TESTS
(EPA Method 624)
Concentrations (ng/L)
Run Ho. 5 (paraho with Retort Water as Moisturizing Fluid)
Sample
methylene chloride
acetone
2-butanone
1,1 ,1-trichloroethane
benzene
2-hexanone
4-methyl-2-pentanone
toluene
Dilution
Retort
Water
1U2
O(SIB)1
0(133B)
140(150B)
180
230
220
250
1:10
First :
Effluent" !
1.0<2.2J,B)
768(780B) :
54( 66B)
0<1.0J,B)
5J
— • —
6.0
NA
Last
Effluent
2.0(3.0J,B)
108(120B)
25(37B)
0(1.1J,B)
6.5
1.6J
2.2J
NA
Run No. 6 (Paraho with Gas Co-ndensate as Moisturizing Fluid)
Sample
methylene chloride
acetone
2-butanone
1,1, 1-trichloroethane
benzene
2-hexanone
4-methyl-2-pentanone
toluene
thiophene
tetrahydro thiophene
ethylbenzene
Dilution
Gas
Condensate
35,850(36,100B)
5,170(5,500B)
3(13J,B)
425
1,200.
120
165
280
14J
20J
1:10
First i
Effluent
7. 0(8. IB)
5,488(5,500B)
658(670B)
o(i.oj,B) ;
14
34 ;
14
— — — ' ,
___ :
— • — . i
NA
Last
Effluent
1.0(2.0J,B)
488(500B)
136(148B)
0(1.1J,B)
14
19
4.0J
— —
1.4
___
— —
NA
Run No. 7 (Paraho with Distilled Water)
Sample
methylene chloride
acetone
2-butanone
1,1, l-trichloroethane
1,1,2, 2-tetrachloroethane
benzene
4-methyl-2-pentanone
Dilution
Distilled
Water
- —
NA
First ;
Effluent ;
30J
84U09B) :
0(33B)
0(1.0J,B) 1
<1.0J
2.0J
1.5J
NA
Last
Effluent
1.6(2.6J)
8(20B)
2(14B)
0(1.2J,B)
2.8J
NA
B indicates compound was detected in ttie blank. Numbers in parenthesis
are uncorrected. Numbers outside parenthesis have been adjusted.
J indicates concentrations near detection limits. '
88
-------
5-3.4 ASTM and RCRA Test Results ;
The results of the chemical analyses of the filtrates from the ASTM.. and
RCRA tests for the Paraho-l spent shale sample are given in Appendix B and
summarized in Table (5.36). Values in this table are in terms of soluble
mass, that is, mg of the ion per original total weight in grams of the sample.
Sulfate levels are high in both test results. However, results 'fran the RCRA
tests included much higher levels of calcium, chloride, and mkgnesium than
determined in the ASTM test.
TABLE 5 .36. PARAHOI ASM AND RCRA TEST RESULTS ;
Parameter ASTM* RCRA* :
F •
' Cl
NO,
sol
ml
B
Si
Mg
P
Na
Ca
K
Sr
IDS'
.054
.029
.014
2.144
.009
.002
.009
.002
.580
1.064
. .124
.020
5.700
' j
""
.576 :
.035
4.40
— — '.
.007 ,
.080
9.68 ;
.010
.740 ;
14.48 '
.130
.168 j
124.4 :
*Values are mg of solubilized species per gram of solid.
5.4 Allis Chalmers • ;
The Allis Chalmers spent shale was obtained from Allis Chalmers, Oak
Creek, Wisconsin in March, 1985. The material is typical of;the material
produced in pilot plant tests and should be similar to spent shales expected
from the Allis Chalmers Roller Grate process. The only difference expected is
that the commercially processed spent shale will be more fully combusted so
that levels of contained carbon and sulfur will be lower than obtained in the
pilot plant samples. ;
5.4.1 Physical Properties
The size distribution for the Allis Chalmers retorted' shale was
determined by dry sieving. This sample was the coarsest sample1used in this
project and consisted primarily of large material with only a few fines.
Figure (5.17) shows the particle size gradation curve for the'. sample. The
sample was crushed and the pycnometer method used to determine the apparent
particle density. The average value obtained was 2650 kg/m . :
89
-------
100
80
60
40
20
1.0
10.0
Parfic/e Size (mm)
100.0
Figure 5.17.
Particle size distribution for Mlis Chalmers
retorted shale.
90
-------
5.4.2 ESM Test Results
The Allis Chalmers retorted shale was used in an ESM column test to
evaluate leachate quality. The moisturizing fluid was distilled water "that
was! added to the retorted shale and allowed to equilibrate for 72 hours.
After equilibration, the moistened sample was carefully packed in lifts into a
column and then the column was leached with distilled water. ' ; The average
initial moisture content, determined by oven drying, was 13% by weight and the
packed dry bulk density was 1230 kg/m3. The flow rate for this ESM test was
414 ml/hr, resulting in a total test time of 61.5 hours and a total time after
the beginning of outflow of 46.3 hours.
As Figure (5.17) shows, the Mlis Chalmers sample consists primarily of
particle sizes larger than 3.96 mm with-less than 10% smaller than this size.
Because the retorted shale may be disposed of in this form, the sample was not
crushed for the column test. However, even though care was taken to avoid
crushing during packing, the shale fractured easily and seme minimal
fracturing was observed. Therefore, a dry bulk density was used that was less
than the value for the other retorted shales. A larger column (89.5 cm in
length, and 16.3 cm in diameter) was used to minimize edge effects. The
possibility of significant wall effects should still be considered, however.
Generally, if the column diameter is 40 times the d5fl particle size, wall
effects are negligible (Franzini, 1956). in this case, the d50 particle size
is 12.1 mm so the ratio of column diameter to d-0 size is only 13Is. However,
because flow is up f ran the bottan and there was very little resistance to
flow through the sample" during the test (as measured -by pressure gages
connected to the inlet piping), it is unlikely that wall effects are important
in the results obtained. . '
The original raw "shale used" in the Allis Chalmers process' was a New
Albany eastern shale from dark County, Indiana. Tests done by Allis Chalmers
showed it had a Fisher assay of 12.5 gpt and contained 4.9% total sulfur.
About 80% of the sulfur was in the form of pyrite (FeS-). Carbon dioxide
contained in mineral carbonates was 2% of the raw shale weight. The original
organic carbon content of the shale was 11.5%
The spent shale sample used in this study was not analyzed by Allis
Chalmers. However, other similar combusted shales analyzed contained about
3.5% residual carbon and 2.8% sulfur, primarily in the form of FeS.
Table (5.37) lists the major ions found in the leachate samples from the
Allis Chalmers ESM test. The pH and alkalinity levels are considerably lower
than those found in the Lurgi tests. The results in Table (5.37) show that
the leachate samples for the Allis Chalmers ESM test contain iron;and sulfates
as expected. The results also show higher concentrations of magnesium,
manganese and boron than found in the leachates of other spent shales tested.
91
-------
TABLE 5.37. MAJOR ICN COMPOSITION OP ALLIS CHALMERS EFELDENT
Run
No.
4
(e.v =
JOOO ml)
cumul
volume
(ml)
63
2763
4630
6063
10753
Zt
L
.01
.28
.47
.6
1.10
CL
298
273
223
163
113
so4
mg/L
6,820
5,520
4,320
3,310
2,590
Mg
1,380
1,070
691
458
282
Fe
32.2
24.8
21.9
18.2
14.2
Ml
31.9
26.2
19.9
15.0
11.1
B'
5.64
5.37
4.57
4.0:6
3.77
i
TDS
9,750
8,040
6,260
4,850
3,840
The first and last effluents frqn this test were subjected to analyses
for TOC and priority list volatile organics. The measured total organic
carbon was less than 5 ppn and none of the volatile organics on the Priority
Pollutant list (see Table 4.4) were found in the effluents. :
5.5
The HYTORT spent shale was used early in this project for some analyses.
Batch leach tests were conducted on this sample but it was not included in the
column leach tests. However, although not complete, the information obtained
is included here. '
!
The moisture retention characteristics for the Hytort sample at several
compaction levels were determined by standard pressure plate methods (ASTM
2325) for capillary tension heads of 1 bar to 14.2 bars. Results: are shown in
Table (5.38) . Values given in this table were obtained by desaturating the
sample from "natural" saturation. ;
Table (5.38) also summarizes the hydraulic conductivity values obtained
for the HYTORT spent shale sample. Proctor density is 1700 kg/m3 at an
optimum moisture content of 20%. ;
5.5.2 ASTM and RCRA Test Results
The results of the batch leaching tests are given in Table (5.39).
Values in this table are the mass of the ion leached per mass of the original
sample in units of mg/gm. Concentrations determined by the laboratory in the
filtrate are given in Appendix B. Dashed lines in Table (5.39) mean that the
values are not significant in the third decimal place. As Table (5.39) shows,
the RCRA test is more severe than the ASTM water shake method. ;
92
-------
00>
fl
v»
o
o
s
*^
a
g
8
en
cffl'5)
o
I
d>
5
CO
(U
M
5
1 &
0)
O 0-J \O
rH en
en
O
•
co
o
o
93
-------
TABLE 5 .39. HYTORT ASTM AND RCRA TEST RESULTS
Parameter
F
Cl
NO,
4
MC
Si
Mo
Na
Ca
K
Mg
B
TDS
ASTM*
(mg/gm)
.014
.050
.021
.712
.066
.084
.008
.104-
1.000
.240
4.24
RCRA* ;
(mg/gm) •
— — ™
.179
.046
1.94
.033 ;
.074
.006 ;
.220
6.380
.44 i
.180
.007
34.80
*Values are mg of solubilized species per gram of solid.
5.6 Chevron
5.6.1 Physical and Hvdrauli<
The particle size distribution for the Chevron STB spent shale sample is
sncwn in Figure (5.18). This sample is f ran a representative operation or me
Salt Lake City Shale Oil Semiworks Plant. The size distribution shown (80
percent passing a 400 mesh sieve) is smaller than expected for a commercial
operation out is representative of the material disposed of at the Ssniworks
Plant. Particle density was measured as 2730 kg/m by standard methods.
'Ihe water retention characteristics ror the Chevron sample are snown in
Figure (5.19). This figure shows the material's ability to hold capillary and
adsorbed water at pressures less than atmospheric during wetting and uryiny or
the sample. A standard pressure plate apparatus was used for both curves. The
procedure tor drainage follows the method described in ASTM D2325. For the
wetting curve, the same pressure plate was used but water was circulated
beneath the plate to provide a source for wetting. :
Figures (5.20) and (5.21) show the hydraulic diffusivity and hydraulic
conductivity runctions for Chevron retorted snale. The Bruce and Klute method
described in Chapter 4 was used to determine the diffusivity function. Jie
unsaturated hydraulic conductivity was determined from tne diffusivity
function and the slope of the water characteristic curve. i
94
-------
100
80
60
-------
o Drainage
a WeMng
10
O.I
0.2 0.3
9
Figure 5.19. Water retention characteristics for Chevron
retorted snaie.
96 !
-------
M I I 1. I I 1 li t i i i i .
(O
'o
(03S/OIOJ
to
d
CM
O
00
'o
2
S-I
o
&
-H
b
li
o >J1
•H IU
r-l Ot
3 -W
RJ M
••4 O
VI
Q)
t-i
3
cn
fi
.J—: i__i i_
to
o
CJ
C5
2
fe
6
u .
O 01
0}
3
H
Q
O
(U
D»
ti
(03S/ LUO) Q
97
-------
5.6,2
Two ESM tests were run using the Chevron retorted shale: one vath
distilled, deionized water as the moisturizing fluid and tne second using a
process water for moisturization. The process water was not cnemicaily fixed
and theref ore probably degraded from the tame of sampling to tne time of
analysis. The water was stripped in a reboiied stripper column and would axso
be treated commercially with Chevron's WWT process. . '.
Taple (5.40) summarizes the test parameters tor the two Chevron ESM
tests. These are labeled run 8 using process water for moisturizing and run 9
using deionized distilled water. In addition, two other sets or parameters
are shown, runs 8a and 9a. During the test, there was a power failure
requiring a repeat of the moisturizing period with new samples and repacking
of the columns. Runs 8a and 9a are the columns for the initial runs. The
first erfluents rrom these initial tests were saved and analyzed. ' Comparison
of the erfluent quality provides information on tne reproducibility of the
results. :
T2BLE 5.40. PARAMETERS FOR CHEVRON CODISPOSAL TESTS
Run No.
8a
9a
Moisturizing Fluid
Packed length, L (mm)
Cross-sectional Area, A (mm2)
Bulk density, p (kg/m3)
Porosity, p
Initial volumetric
moisture content, p.
(% by weight) i
Flow rate (ml/hr)
Inflow Darcy rlux
q... (raa/hr)
Process Water
413
8123
1300
0.523
0.186
(14.3)
81.0
9.97
413
8123
1370
0.498
0.175
(12.8)
96.5
11.88
Distilled Water
412
8155 !
1290
0.527
!
0.188
(14.6)
!
79.6
!
9.76 !
!
412
8155
1330
0.514
0.191
(14.4)
95.8
11.83
98
-------
5.6.2.1 Inorganic Analysis
Table (5.41) shews the concentrations of inorganics in the process
water and the first and last effluents from the columns for the Chevron tests.
First effluent concentrations are significantly different for the two
tests. Run 8 had higner sodium and chloride concentrations. : The suifate
level was aiso higher. These higher concentrations reflect the high sodium,
suifate and chloride concentrations in the process water. The tnoisturizing
water used was stripped in a reboiied stripper column. The sttipper reed
water contains spent SO, scrubbing liquor (NaHSO,, NEUSO, and Na^SO,) which,
according to Chevron, would not be present commercially. However,: the results
obtained from the ESM test are relevant for the shale and water disposed of on
the site or the Semiworks Plant. '
The first effluent in run number 9 (using deionized ; water for
moisturizing) also has high concentrations of sodium and suifate, indicating
that in run 8, the process water is not the only source or these constituents.
The first effluent in this run, however, has higner levels of calcium than run
8, reflecting the equilibrium values of about 1450 mg/L found in tne last
erfluents from both columns. Calcium and suifate concentrations in the last
erfluents indicate long-term solubility constraints witua solid phase calcium
suifate. Strontium and potassium levels in the leachates are also significant
with -che source of these constituents being primarily the solid material
althougih potassium was also found in the process water. j
i
5.6.2.2 Organic Analysis
Samples or the moisturizing process water, distilled moisturizing
water, and 'first and last effluents were subjected to analysis for total
organic carbon and priority pollutant volatile organics. The results for TOC
are given in Table (5.42). The last effluents indicate anomalies with total
TOC and uydrophilic TOC. Duplicate analyses were performed on different days
(numbers shown in parenthesis) and higher TOC values after1 C^g SepPak
separation were again found. One possible cause may be sample stripping of
organics during the separation step. Also, both samples affected have TDC
values near the 10 mg/L quantitation limit. '.
Table (5.43) summarizes the results of EPA method 624 for the volatile
organics in each sample. Again, the process water was not chemically fixed so
degradation occurred before analysis. Analyses for runs 8a and 9a are also
shown. Numbers followed by a J are approximations because, although tne
compound was found, quantification was difficult because concentrations are
near detection limits. A number in parenthesis followed by a B indicates the
compound was found in the reagent blank run tne day of tne analysis. Numbers
in parentheses are uncorrected for the concentration found in tne blank.
Numbers shewn outside the parentheses have been adjusted. A duplicate
analyses for the last effluent in run 9 resulted in concentrations (adjusted
for the concentrations'found in the blanks) of 81 jig/L and 43 ng/L; f or acetone
and 2-butanone, respectively. These values confirm the numbers snbwn for this
sample in Table (5.43). '.
99
-------
TABLE 5.41. RESULTS OF THE INOEGANIC ANALYSES FOR THE CHEVRCN ESM TESOS
EC (dS/m)
pH
CaC03
co3
OH
F
Cl
so4
Ca
Ka
K
Fe
Zn
Mo
Sr
B
Ba
As
Se
TOS
Moisturizing
Fluid
Concentrations
(mg/L)
8
4.89
10.1
895
279
2.0
1,050
584
10
1,270
39
0.27
<0.01
0.33
0.34
0.30
0.04
0.088
0.028
3,790
First Effluent
Concentrations
(mg/L) (rog/L)
81
16.3
12.7
2,930
642
631
8.4
794
2,780
981
2,090
219
0.09
0.43
7.04
32.3
1.82
0.14
0.015
0.124
8,210
92
12.8
12.6
2,590
486
604
7.1
201
2,160
1,360
834
146
0.06
0.47
4.21
36.4
1.91
0.18
0.005
0.079
5,850
Last fcrrluent
Concentrations
(mg/L) j (mg/L)
•«* 92
9.3
12.6
1,990
168
580
4.3
52
1,820
1,460
69
22
o.o;8
0.23
1.20
25.6
1.75
0.17
0.002
0.025
4,230
9.3
12.6
1,990
192
568
4.4
24
1,540
1,430
68
24
0.20
0.28
1.21
24.9
1.91
0.20
0.001
0.033
3,900
Process water used for moisturizing.
'Distilled, deionized water used for moisturizing.
Ni, Cd, Hg and Pb were not detected, see Appendix A.
100
-------
TSBLE 5.42. AVERAGE TOTSL ORGANIC CARBON FOR CHEVRON CODISJOSAL TESTS
I
Run 8
Moisturizing water (run 8)
First effluent (run 8)
First effluent (run 8a)
Last effluent (run 8)
Run 9
Distilled deionized water
First etfluent (run 9)
First erfluent (run 9a)
Last effluent (run 9)
Total
TOC(ppn)
660
565
635
20(16)1
10
380
390
18U1)1
Hydropnilic
TOC(ppn)2
190
470
465
25(28)1
10
335
340
39(34)1
Hydrophobic
TCC(ppa)3
: 470
95
170
'•
10
45
i 50
Numbers in parentheses, are recheck analyses. '.
2
Values are corrected ror blanss run through tne SepPak C^
cartridges, sample volume run through the C- c was 3-5 ml.
3 • i«
Total TOG - hydrophilic TCC = hydrophobic TOG.
101
-------
TABLE 5 .43. RESULTS OF VOLATILE ORGANIC ANALYSIS FOR CHEVRON
(EPA Method 624) :
Concentrations (ng/1) ;
i
Run No. 8
acetone
2-butanone
benzene
toluene
2-hexanone
Dilution
Run No. 9
acetone
2-butanone
1 .1 .l-tricnloroethane
toluene
Dilution
Process
Water
Run 8
1139 (1150B)1
558 (570B)
6.8J2
9.6J
67J
1:10
Distilled
Water
Run 9
19 (SOB)
6 (18B)
U
U
NA
First
Effluent
Run 8
1769 (1780B)
138 (150B)
1.4J
NA
First
Effluent
Run 9
1190
NA
First ;
Effluent
Run 8ai
1460 !
10
j
First :
Effluent
Run 9a
960 ;
NA ;
Last
Effluent
Run 8
72 (83B)
76 (88B)
U
NA
Last
Effluent
Run 9
83 (94B)
46 (58B)
NA
B indicates compound was detected in the blank. Numbers in i
parenthesis are uncorrected. Numbers outside parenthesis have; been
corrected.
J indicates concentrations near detection limits. :
102
-------
Chapter 6
i
DISCUSSION AND VERIFICATION OF THE ESM AND ISM COLUMN LEACH TESTS
The two column leach tests developed during this project differ
fundamentally in the physical and chemical processes that i occur in the
columns. In the Instantaneously Soluble Mass (ISM) test, leachant is injected
into an initially dry column and, as the name implies, this: test leaches
species that are rapidly dissolved on contact with the invading leachant;. She
second test, the Equilibrated Soluble Mass (ESM) test, injects distilled water
into a column that is initially moist. The antecedent moisture has been
allowed to equilibrate with the retorted shale and, as the theory: presented in
Chapter 4 predicts, the first effluent fran the column is this undiluted,
equilibrated antecedent moisture. Because this second test proyides a means
of determining the quality of the equilibrated moisture and also because this
test more closely approximates conditions in a field retorted shale disposal
pile, the ESM test holds more premise as a prediction of field generated
leachate quality. This chapter uses the results presented in: Chapter 5 to
validate and illustrate the theory given in Chapter 4 for these two column
tests. i
6.1 The ESM Test
6.1.1 Reproducibility of the ESM Test •
The results of the two leach tests using the Lurgi ULG spent; shale will
test reprcducibility of the results. Test parameters (see Table 5.3) are
nearly identical for ESM runs 33 and 34. Figures (6.1), (6.2) and (6.3) are
plots of electrical conductivity, chloride concentration : and sulfate
concentration versus cumulative volume of leachate, respectively.; Electrical
conductivity is directly related to the total dissolved solids and, as a gross
indicator of the leachate composition, displays excellent reproducibility.
Chloride is a conservative species and displays good reproducibility although
seme scatter in the results is observed. Much more scatter is seen in the
results for the sulfate ion. This scatter in the data is because of either
nonuniformities in the pore solution due to material and packing
nonuniformities or error in the chemical analysis of leachate samples.
Reproducibility appears to be better for conservative species such as chloride
that undergo little or no interaction with the solid matrix.
information frcra these tests,
test is reproducible.
albeit limited, it can be assumed that the leach
103
Based on the
-------
16
14
CM 10
B
Is
ui 6
"0 o
q
a
O Run No. 33
a Run No. 34
cn
ZOO
400 600 800
Cumulative Volume of Effluent (ml.)-
1000
1200
Figure 6.1. Electrical conductivity of Lurgi UIG effluent^samples.
350 ^
a
300
0-250
o ZCO
"c
«
o
o 150)-
v
I IC<>
so
O
a
O Run No. 33
a Run No. 34
°a
20O
40O 60O 800
Cumulative Volume of Effulent (mL) •
IOCO
1200
Figure 6.2. Chloride concentration in Lurgi ULG effluent
104
samples.
-------
10,000
} 8,000
o>
26,000
4.0CO
2,000
o
o
o Run No. 33
D Run No. 34
2OO
ado"
4OO 60O
Cumulative Volume of Effluent (mL)
1000;
I2OO
Figure 6.3. Sulfate concentration in Lurgi UIG effluent samples.
105
-------
6.1.2 Results as Support for the BSM Theory
The most important requirement in interpretation of the ESM test results
is that the antecedent pore solution is displaced during invasion of the
leachant. This is inherent in the theory in Chapter 4 and in Nazareth (1984).
The results from the initial runs using the TOSCO II and Lurgi spent shales
are used to provide support for this assertion. ' ;
At the moment of incipient effluent production, Eqn. (4-10) becomes
i (6-1)
6,z* » (6 - 8.) (L - z*>
i m i .
which, when rearranged, can be used to calculate the thickness, A ;= L - z*. in
terms of the column length and the initial and maximum moisture contents
(6-2)
m
In the absence of dispersion, the entire zone of thickness, : A, at the
outf low end of • the column is displaced antecedent pore solution and remains
unc»ntaminated by the injected solution. However, there will exist an
interval of mixing near z* due to hydrodynamic dispersion. The thickness of
the dispersed zone will be about (Dz*/v ) (Wilson and Gelhar, 1981) where D
is the dispersion coefficient and v is the seepage velocity of the pore
solution at 9 = 0 . Therefore, there exrsts reasonable assurance that the
first increment of effluent from a column will be uncontaminated antecedent
pore solution if • ;
m ' m
For TOSCO experiments 29 and 35 (see Table 5.22), values of (p.L/6 ) are
9.1 and 12.4, respectively. Corresponding values of (Dz*/4v ) ' are Om80 and
0.73. For Lurgi runs 33 and 34, O-jL/aj values are 13.4 anS 14.0 with values
of (Dz*/4y _) equal to 1.4 and 1.31, respectively. In all of .these tests,
the thickness of the displaced antecedent moisture content is an order of
magnitude larger than the. thickness of the dispersed zone. This provides
reasonable assurance that enough uncontaminated antecedent moisture can be
sampled. i
If the extent of mixing between the antecedent and the invading solution
is sufficiently small, one would expect the initial leachate effluent to have
the same chemical composition as the antecedent pore solution. It follows,
then, that the antecedent pore solution, since the wetted material was well
mixed and uniformly packed, will have a uniform composition. This uniformity
of the? initial effluent was evident in Figures (6.1-6.3) and is further
displayed in Figure (6.4) which plots relative concentration versjus effluent
for scare of the chemical parameters measured for all four TOSCO and Lurgi
tests. C. is the average concentration of the first three or four samples
when their concentrations are approximately equal or C. is the maximum
106
-------
1.1
EC
1.0
-CD A ao -
Cj '*
0.9
a—o» a
A Symbol
o
•
A
Chloride a
o
= -A— ^ — n
Run No.
' 33
<34
:29
;35
1
1.0
0.9
Sodium
-o a& _. __.
i.o
0.9
Calcium
0 50 IOO 150 200 250 30O
Cumulative Volume of Effluent (ml.)
Figure 6.4. Relative concentrations in initial effluent samples.
TABLE 6.1. INITIAL EFFLUENT CONCENTRATIONS (mg/L)
Material
Run No.
. F
C1
so4
Na
Ca
Mg
Ho
Uurgi
33
17.8
337
9680
3810
519
0.619
5.50
ULG
34
13.0
331
9025
3790
466
0.478
7.91
29
29.2
270
30180
12700
360
560
59.1
Tosco II i
.35 , ;
24.9
214
29300 |
10430 j
!
368
696
26.3
107
-------
concentration when there is scatter. The electrical conductivity and chloride
plots show good uniformity while some scatter is observed for calcium and
sodium. Values used for C^ are given in Table (6.1). ; -
i
If C- is, as the theory predicts, the concentration of the particular
species in the antecedent pore solution, then the total mass of this species
dissolved in the antecedent solution per unit dry weight of leached material
is given by the expression ©jCj/pi.. The leached mass of a species in the
effluent is estimated by integrating the species breakthrough curve
(concentration versus effluent volume) and dividing by the total !dry weight of
leached material. This estimate is denoted by S. For a highly soluble,
noninteracting species such as chloride, the two numbers ©jCj7pb and S are
equal if C-, determined fron the initial effluent, is indeed the antecedent
pore solution concentration. Further, values of QiCj/pu determined for
chloride should be independent of the antecedent solution content used.
Table (6.2) displays values of S and QjCj/pi., for the different tests am
chanical species. Good agreement for chloride between the estimates of
antecedent dissolved chloride mass and the leached chloride mass in the
effluent confirm that C. is indeed the concentration of chloride in the
antecedent solution. In turn, this is strong evidence for the ; displacement
model., •
Table (6.2) is also useful to contrast the behavior of conservative and
noncoriservatiye species. As expected, nonconservative species are observed in
greater quantities in the effluent than are apparently present in the
antecedent solution. This results from continued dissolution by the flowing
fluid as evidenced by the fact that the concentrations of most of these
species, e.g., sulfate and calcium, remain elevated at the end of the test.
The leached mass, S, would increase even further had the test been conducted
for a longer time. The concentrations of nonconservative species in the
antecedent pore solution are evidently limited by the complex solubility
relations and not by the availability of the species. The observations
involving nonconservative species provide no evidence, either for or against,
the displacement hypothesis.
iilquation (4-11) given in Chapter 4 predicts the relative concentrations
of species in the effluent at any time after the beginning of leachant
injection. For a conservative species, I* equals 1 and the equation shows
that a relative concentration of 0.5 should occur when the cumulative leachate
volume is equal to the volume of the antecedent pore solution or, in other
words, one pore volume of leachant has been injected. Therefore,:, a comparison
of observed effluent volumes when the relative concentration is 0.5 with the
known volume of antecedent moisture provides an additional check on the
theory. Values of the volume of antecedent moisture, e.fiL, and >the observed
cumulative effluent at the 0.5 relative chloride concentration for each t:est
are shown in Table (6.3).
108
-------
cn
1
^"^
a
i-H
1
co
Q
g
f*
§
CO
1
£-*
*— j
<%
§
a
s .
3
CQ
M
fr|
CQ
§
§
^t
es
•
\o
63
E-i
o
s
=3
«J
21
a
CJ
u_
o
21
3
Q£
(0
S-
0)
1
a
-H
o
U3
•3-
O
rH
X
r-i
"0
rH
^
O
rH
X
O
rH
1
a
rH
X
rH
IO
<3-
°0
X
in
•3"
'"
CM
1
O
i—4.
X
r-l
CM
_j 3
A
Q.
CO
1
3
r— 1
X
cn
m
CM
1
o
rH
X
O
0
ai
CM
i
o
X
CO
in
ai
o
o
r-4
X
r-.
i-H
o
o
pH
X
m
co
*
CM
1
O
X
ID
•3-
CO
CO
1
o
rH
X
r—4
CO
ai
«/•
CO
1
o
rH
X
o
r^
CM
1
O
-H
X
rH
*
CM
O
X
in
o
10
°o
r-4
X
CM
in
r- 1
o
o
r—4
X
r-t
VO
cn
CM
1
O
X
cn
CM
CO
cn
t
o
rH
X
cn
•3"
CO
CJ
0
cn
CM
.Q
a.
cn '
i
o i
X '
r-4 ,
•3- ;
rH
1
O
r— t •
X
rH'
rH;
CM
1 ,
O(
Xi
to
o
o
rH^
X
CO1
l~-_.
r-4
O
r— t'
X
CO
tn.
•3",
CM :
1
O
r—4.
X
CM
rH
cn
cn
0
r— t
X
m.
r^
cri
V)
cn
i
o
X
o
cn
CO
r-4
a
rH
X
cn
o
r-4
CM
'o
X
m
in
°o
t—l
X
in
Uf)
rH
a
o
r— t
X
•3*
CO
^
CM
t
O
X
r-t
CO
CO
O
i— t
X
lO
CO
u
•*•
-i
1-7 |
109
-------
TABLE 6.3 ANTECEDENT MOISTURE VOLUME AND CUMULATIVE LEACHATE VOLUME
AT 0.5 RELATIVE CHLORIDE CONCENTRATION (mL)
'
Material
Run No.
e£AL
(cum.V)n ^
U * Z)
Lurgi ULG
33 34
380 387
435 425
TOSCO II
i
2.9 ; 35
341 449
i
327 ' 419
The results show that the solute front is observed somewhat' later than
predicted by Eqn. (4-11) in the Lurgi experiments and slightly earlier in the
TOSCO tests. However, the largest discrepancy shown in Table (6.3) is 14.5%
for run No. 33. This is probably within experimental error.; However, if
incomplete displacement is occurring and some antecedent solution is immobile,
the solute plane would be observed earlier than expected. On the other hand,
if solute from the immobile phase diffused rapidly into the displacing water,
the concentration behind the front will be elevated to levels higher than
those expected by just hydrodynamic dispersion. This would cause: an extended
tailing effect and the 0.5 relative concentration would appear! later in the
effluent. The two effects of immobile liquid and tailing effect oppose each
other in the locating of the 0.5 relative concentration. Nevertheless,
regardless of these secondary effects, it appears that enough antecedent
moisture was displaced to provide evidence of the displacement theory and
provide assurance that enough sample of the equilibrated, uncontaminated
antecedent moisture is available for analysis.
Nazareth (1984) discusses the effects of all the parameters in Eqn. (4-
11) and the effects of the various parameters on the shape of the; breakthrough
curves in much more detail than is warranted here. \
6.2 The ISM Test
The data from run 5 (Lurgi RB-I, Table 5.6) are believed i:o be least
affected by an undesirably large mixing volume at the outflow end of the
column and can be examined in the perspective provided by the theory presented
in Chapter 4. It will be recalled that the theory was intended to apply to an
" instantaneously" solubilized species that undergoes no further chemical
interaction with the solid. . The chloride ion approaches this ideal more
closely than any of the other major ions. Figure (6.5) shows the relative
chloride concentrations as a function of effluent volume superimposed on the
theoretical breakthrough curve calculated from Eq. (4-13) with D/Lv = 0.00625.
The fit is very good and the corresponding value of D = 0.3 on /hr is a very
reasonable value for the coefficient of hydrodynamic dispersion. '< This implies
that the shape of the chloride breakthrough curve in this case is indeed
dominated by hydrodynamic phenomena. • :
The total weight of chloride ion leached from the column in run 5 can be
determined by integrating the area under the breakthrough curve. It was
determined that the mass leached per unit mass of solid was 0.080 mg/g. This
110 :
-------
0.8 h
0.6-
RUN No. 5
• Chloride data :
— Theory for —=0.00625
Lv
0.2 0.4 0.6 0.8
T"
1.0
Figure 6.5. Theoretical breakthrough curve compared to
Cl data (Lurgi EB-I). '._
Ill
-------
value was used for S in Bqn. (4-14) (together with L « 432 mm, v= 11.2 mm/hr/
9 - 0.538, and p^ =1225 kg/m ) to calculate a value of c =
agreement with the **S^^£jg££ ffJS?"?!^
parameters and the measured C-, is reduced by mixing at the top ojf the
Sodium and sulfate are the largest contributors to the total dissolved
K • L in ^luent from run 5. it is of interest to apply Bgn.< (4-13) to the
breakthrough data for these ions. A fit of Bjn. (4-13) with D/£v= 0.0? to
the breakthrough data is shown in Figure (6.6). Because the persistence of
elevated concentrations of S04 suggest that the long-time concentration of
ttiis ion is controlled 4by the solubility oF Sfid ^h2e S£? a
nonreactive" concentration was calculated by subtracting C = 1450 nn/1
SSL?6 ^doa^7f concentrations. It will be noted that thf^fit between the
theoretical and observed breakthrough curves for these ions is not as good as
tor enloride. KLso' the fit shown ^ Figure (6.6) corresponds I to D = 0.0048
mm /hr, a value significantly greater than the 0.003 mmz/hr value determined
for- chloride. Coefficients of hydrodynamic dispersion for Na and SO should
not deviate from that for a by so wide a margin. The breakthrough curves for
sodium and sulfate, although dominated by hydrodynanic : effects, are
Zu XT y other Processes. For example, the concentration of
the effluent is influenced by adsorption/desorption processes on the
exchange complex of the solid. re. un=
i< _.:tSM, runs 4 and 7 were performed on identical TOSCO II materials. Figure
rlllL f ?at the k^tkrough data show an initial increase, a period of
nearly constant concentration, and finally a sharp decline. The'shapTot the
curve at small tunes seems strongly influenced by the wettability of the
porous; medium. The material used in these runs is hydrophobic until placed in
£252; i^ S ^ /°r a considerab:!-e time. All other materials tested are
hydrophyllic. The hydrophobic property of the porous medium had two marked
?X?SLJTr:Ln;L the wetting phase of the leach tests. First, very high
(relative toBother media), injection pressures were .required to lovercome the
capillary effects that were acting to retard rather than to promote wetting.
Since the medium was hydrophobic, the nonwetting phase entry capillarv
pressure had to be overcome. This is in contrast to the^drophyllg
materials that exhibit a rapid uptake of water by capillary action/ A Sond
etrect. was to create an apparently unstable wetting front, while the wettina
front in other runs was observed to be uniform, the wetting front in runs 4
ana 7 was ragged with narrow fingers penetrating several centimeters bevond
cne average TV-ICT+-i ^f> /~*P 4-u^, -c—~_j_ ~j »***»*
The observed nonuniform wetting front causes a nonuniform arrival of
riuid at the top of the column. This nonuniform arrival of fluid mav explain
858 °f • S6 breakthrou* curve at small tine. -RirtheSrl, S2
4- reS1<^f Preferentially in the larger pore space until the
system reverts to a preferentially water wet system. This could result in a
SSlffi nonuniform distribution of dissolved constituents on any cross section
and further contribute to the unusual shape of the breakthrough curve at small
£S^ h^?nr£r?^et£2g P^OCeSS used in ^ ESM tests ^"ses the material to
become hydrophyllic before injection begins and problems encountered in the
ISM tests are not observed in the ESM tests.
112
-------
RUN No. 5
• Sulfate data
— Theory for — = o.Ol
Lv
RUN No. 5
• Sodium dafa
Theory for •- =0.01
0.6 0.8 1.0
Figure 6.6. theoretical breakthrough curve compared to Na and SO, date
(ISM leach of Lurgi BB-I). ; 4
113
-------
l-Or
a
08-
J 0-6 h
UJ
o
IU
0-4[-
0-2
0' - '
Run no. 4
0 0-2 0-4 0-6 0-8
Vt/L
-0
l-Or
0-81-
£
o
UJ
\
o
UJ
0-4-
0-2-
0
Run no. 7
1
0 0-2 0-4 0-6 0-3 i-0 \-2-
Vt/L
Figure 6.7. Electrical conductivity breakthrough curves for
Runs 4 and 7 (ISM leach of TQSCO II). \ .
114
-------
Runs 6 and 8 were conducted with different size fractions of the same
Paraho material (see Table 5.29). Run 6 used material between 0.4 and- 1.2 nrn
and run 8 was made with all material smaller than 3.96 mm. All 'breakthrough
curves previously discussed are characterized by a maximum EC I that is much
greater than the "equilibrium" value approached asympotically at large time
This is not the case for runs 6 and 8, and the result is an apparently much
slower leaching rate, -me leaching rate in runs 6 and 8 is less 'than in tie
previously discussed cases, but the disparity is not nearly as great as the EC
breakthrough curves suggest (Figures 6.8 and 6.9). This is demonstrated by a
comparison of sulfate concentrations in the effluent from runs 5 and 6 as
shown in Figure (6.10). AS before, the concentration of
.. , ncenraon o sulfate in
equilibrium with solid phase CaSO has been used as a i normalizdS
4
maz
concentration, providing values that can4be directly compared. A ; coSfiSent
of hydrodynamic dispersion much greater than considered a reasonable maximum
is required to explain the much greater normalized concentrations observed in
run 6,. An important factor in explaining the different leaching rates shown
is the presence of a significant microspace in run 6. Diffusion of dissolSS
species from the stagnant fluid in the microspace to the flowing fluid in the
rracrospace is believed to be largely responsible for the relativJ ! pSsiSenS
of high values of normalized sulfate concentrations in run 6. A similar
comparison of presumably nonreactive chloride in effluent from runs 5 and 6
show the same disparity in leaching rates. !
«.« <-? addjtion to the Afferent leaching rates for runs 6 and 8 as compared
to the other runs, the maximum EC values observed in runs 6 and 8 are men
iS T ."K^ured in the other experiments. The diffusion! hypothesis
advanced in the previous paragraph cannot account for such large differences
Tne much lower EC at initial breakthrough in runs 6 and 8 S ^roSS? a
reflection of correspondingly lower values of "instantaneously" solubiliz.=d
solids in contact with the fluid. The parameter S in Eqn. (4-14) is the mSs
of instantaneously solubilized species £ contact with the f luid pe? Sit
of solid and the concentration at initial breakthrough i? <
proportional to s. Thus, factors influencing s are Srectly reflected
maximum concentration (of EC in this case) observed. ^£eotiy rerj-fccecl
' A material property expected to greatly influence S is the
surface area of the packed solid per unit volume
as tar*s ^s^
«-r"ta« -
SS*^ ^S rS^° °f ^^ ic surface f or run 5 fco tSt :Sr ru? 6
is est.unated as-5.7. This ratio is roughly the same value as that formed bv
the maxrunum EC values for the same runs. The implication is Satthe ST of
instantcineously soluble solids per unit solid Surface area S coraSble S
the two materials. Apparently, it is the much smaller solid surf^lr uni?
!?SfHM hS^hro01^ ^at iS re^««^e t™ the much smaller value oT£
at initial breakthrough in runs 6 and 8. it follows that column tests should
115
-------
.^> Run no. 6
%
0-8- \
LU
X.
O
ui 0-4
0-2-
«•• o «« * e «
°o
' '
0-4 0-8 1-2 1-6 2-0 2-4 :2-8
Vt/L
Figure 6.8. Electrical conductivity breakthrough
curve for Run 6. :
•or-.. ;
., V Run no. 8
0-8
0eO-6
LU
\
o
LU 0-4
0-2
•
01 2 3 4 5 6 7
Vt/L
Figure 6.9. Electrical conductivity breakthrough :
curve for Run 8 (ISM leach of Paraho). ;
116 i
-------
I.Oo*
o •
0.8
0.6
C - Cpq
Ci -Ceq
0.4
0.2
Sulfate data
o Run 5 - Lurgi
• Run 6 — paraho
o
O-
0.2
0.4
0.6
_vf
- L
0.8
1.0
1.2
Figure 6.10. Conparison of sulfate breakthrough curves for
Runs 5 and 6. :
117
-------
be constructed to mimic the specific surface that will exist under disposal
conditions. A convenient indicator of similar values of specific surface is
similar values of hydraulic conductivity.
118
-------
Chapter 7
QUALITY ASSURANCE AND QUALITY CONTROL
i
The inorganic chemistry of the process water and leachate was determined
by the Colorado State University soil Testing Laboratory at'Ft. Collins,
Colorado. The organic analyses were done by the Western Research;Institute in
Laramie, Wyoming. The objectives of the project QA/QC were to vetify that the
results obtained from the analyses were sufficiently accurate to determine
which contaminants were present and at least the order of magnitude of the
concentrations. The analyses were of a screening characterization nature and
the QA/QC objectives were met. Details of the QA/QC are given in this
chapter. ,
7.1 Inorganic Tests ;
Msthods used by the CSU Soil Testing Laboratory for the inorganic tests
are given in Tables 4.1-4.3. Generally, only one sample of leachate was
provided for each set of analyses but duplicates were provided when possible.
Samples -were also checked for pH and EC during the column tests and these
numbers were compared to numbers reported by the Soil Testing Lab. Howewr,
no attempt was made to quantify the fit. The check was made primarily "to
reinforce confidence in our own testing procedures. !
As routine procedure, the CSU Soil Testing Lab checks internal standards
from the National Bureau of Standards every 10 samples. The accuracy required
is 5 percent. Restandardization occurs every 20 samples. As part of this
study the CSU Soil Testing Lab also participated in USEPA Water Pollution
Performance Evaluation Studies WP013 and WP014. Trace metals and mineratls
were tested. Tables (7.1) and (7.2) show the results for WP013. Trace metals
were all within acceptance limits except for mercury. Mercury sample number 2
was within acceptance limits. However, sample number 1 was reported low by
the CSU lab. This was probably because the value was near the lower limit of
detection. Several additional check samples for mercury were obtained from
the EPA quality assurance officer and the procedure for mercury w&s adjusted
and rechecked before the column test analyses were done. Mineral analyses
were* all within acceptance range. \
i
Tables (7.3) and (7.4) show the results for the WP014 study. In general,
the laboratory performance was good. The trace metals were all within warning
limits so these results are well within acceptance limits. Some of the
minerals, however, were reported slightly high. The chloride samples were
reported as 185 and 36 mg/L when true values were 167 and 30, respectively,
and acceptable upper limits were 179 and 34, respectively. Alkalinity and
sulfate were each reported high in one of two samples and TDS in1 the first
I
" -.- i
• ' 119
-------
TABLE 7.1. WATER POLLUTION STUDY NUMBER WP013
i
Laboratory: CSD Soil Testing Lab
Parameters
Trace Metals
Aluminum
Arsenic
Cadmium
Chromium
Copper
Iron
Mercury
Manganese
Nickel
Lead
Selenium
Zinc
True
Value*
Acceptance
Limits
Warning
Limits
Reported
Values
in Micrograms per Liter: i
98.1
26.2
289.
4.95
65.1
14.3
122.
27.0
260.
22.0
440.
1.70
7.43
22.8
309.
25.0
250.
43.3
338.
7.00
63.6
16.0
280.
56.8-211.
16.9-34.7
206 .-3 70.
2.95-6.47
49.3-74.0
8.75-20.1
91.9-154.
19.0-35.0
220. -2 90.
9.11-39.0
365. -514.
.825-2.60
4.46-10.2
12.1-30.8
265. -3 45.
14.9-35.7
207 .-2 92.
31.8-55.1
263 .-411.
2.94-9.78
32.6-83.6
7.80-26.5
23 6. -3 24.
77.3-190.
19.2-32.3
228. -349.
3.41-6.01
52.5-70.8
10.2-18.7
99.8-146.
21.1-33.0
229. -2 81.
13.0-35.1
3 84 .-4 94.
1.06-2.36
5.21-9.48
14.5-28.3
275.-3S5.
17.7-33.0
219. -2 81.
34.8-52.1
2 82 .-3 92.
3.85-8.87
39.4-76.8
10.2-24.1
247 .-3 12.
170.
* 28.
302.
1
1 5.
! 58.
i
i
; is.
126.
i 30.
258.
1 29.
437.
! <0.5
1 5.
23.
300.
,23.
244.
<50.
344.
7.
;74.
23.
272.
*Provided by EPA.
120
-------
TABLE 7.2. WATER POLLUTION STUDY NUMBER WP013
i
Laboratory: CSU
Parameters
Soil Testing
True
Value*
Lab
Acceptance
Limits
Warning
Limits
^Reported
: Values
Minerals in Milligrams per Liter: (Except as noted)
TDS at 180 C
Total Hardness
(as CaC03)
Calcium
Magnesium
Sodium
Potassium
Total Alkalinity
(as CaC03)
Chloride
Fluoride
Sulfate
pH-units
Spec. Cond.
(nmhos/cm at 25°
64.1
458.
29.5
205.
3.40
59.0
5.10
14.0
6.74
53.5
6.00
35.0
6.26
44.0
22.1
146.
0.702
2.50
12.0
105.
9.19
4.01
126.
C) 814.
25.5-00.5
409.-541.
23.7-35.7
189. -214.
2.24-4.70
47.5-67.4
4.06-5.98
11.5-15.9
5.22-8.30
45.4-60.8
4.70-7.23
28.3-40.3
3.19-10.3
39.2-49.2
18.8-25.8
13 3. -157.
.578-. 809
2.13-2.79
8.36-15.3
83.9-122.
8.78-9.43
3.93-4.12
109. -140.
715.-888.
35.0-90.1
426 .-524.
25.2-34.1
192. -211.
2.56-4.38
50.1-64.8
4.31-5.73
12.1-15.4
5.62-7.89
47.4-58.8
5.03-6.89
29.9-38.7
4.09-9.41
40.5-47.9
19.7-24.9
13 6. -154.
.608-. 779
2.22-2.71
9.27-14.4
88.7-117.
8.86-9.35
3 .95-4.10
113 .-136.
737 .-866.
! 63
i 530
i 26.5
205
1 2.8
; 60.5
i
4.7
13.5
: 6.2
i 50.8
5.4
33.2
6.0
! 42
: 24
145
1 0.6
2.4
! 12
111
9.20
j 4.00
127
i 742
*Provided by EPA.
121
-------
TABLE 7 .3. WATER POLLUTION STUDY NUMBER WP014
•
Laboratory: CSU Soil Testing Lab
Parameters
Trace Metals
Aluminum
Arsenic
Cadmium
Chromium
Copper
Iron
Mercury
Manganese
Molybdenum
Nickel
Lead
Selenium
Strontium
Titanium
Zinc
True
Value*
Acceptance
Limits ..
Warning
Limits
Reported
! Values
in Micrcgrams per Liter: ' ;
1074
84.1
463
391
90.4
66.2
916
694
110
1558
1120
7.19
24.0
803
252
3.75
53.8
283
121
250
556
60.6
45.5
52.5
2.10
198
41.2
201
1378
847 .-1310.
59.3-108.
3 27. -600.
3 45 .-441
72.7-110.
48.9-82.5
6 96. -1100.
606 .-7 64.
92.5-129.
1320.-1780.
947 .-1230.
4.47-9.77
15.5-31.9
704 .-83 8.
221 .-2 86.
D.L.-13.4
30.3-69.9
244. -330.
99.0-143.
198.-300.
443 .-663 .
31.5-78.9
23.4-59.3
37.5-67.6
D.L.-7.10
85.7-335.
6.71-83.7
168. -234.
11 80. -1560.
909. -1240.
65.8-102.
362. -565.
3 57 .-429.
77.4-105.
53.2-79.2
749. -1050.
6 26 .-7 44.
97.2-125.
1380 .-1700.
991 .-1240.
5.17-9.08
17.6-29.7
723 .-534.
229. -277.
D.L.-10.5
35.8-64.5
255. -3 19.
105 .-137.
211 .-287 .
471. -634.
37.8-72.6
28.2-54.5
42.0-63.1
D.L.-5.47
125 .-296.
18.7-71.7
176. -225.
1230. -1510.
! 1100
'] 94
' 490
381
94
•! 68
: 904
665
; 108
i 1535
1115
7.8
27
788
248
5
i 52
! 277
: 118
273
600
: 51
32
50
; 2
196
37
189
1320
*Provided by EPA.
122
-------
TABLE 7 .4. WATER PCLLUTION STUDY NUMBER WP014
.... .
Laboratory: CSU Soil Testing
Parameters
Minerals in
IDS at 180 c
Total Hardness
(as CaC03)
Calcium
Magnesium
Sodium
Potassium
Total Alkalinity
(as CaC03)
Chloride
Fluoride
-
Sulfate
pH-units
Spec. Cond.
(jomhos/cm at 25°
True
Value*
Milligrams per
538
67.0
235
40.3
48.0
5.40
28.0
6.50
86.3
4.73
15.0
4.00
65.8
4.07
167
29.9
1.30
0.802
131
7.00
6.00
7.00
922
C) 136
Lab
Acceptance
Limits
Warning
Limits
<
;Reported
Values
Liter: (Except as noted)
445 .-617 .
31.8-111.
218. -246.
34.3-48.0
40.8-54.2
4.16-6.69
23.6-31.6
5.24-7.63
74.4-94.0
3.93-6.02
11.9-17.3
3.04-4.66
58.4-69.5
1.21-8.04
152. -179.
26.2-33.7
1.12-1.45
.670-. 921
111 .-148.
4.14-9.65
5.85-6.13
6.32-7.14
824. -1030.
119. -154.
467 .-595.
41.9-101.
221 .-242.
35.8-44.5
42.6-52.5
4.49-6.36
24.6-30.6
5.55-7.32
76.9-91.5
4.20-5.75
12.6-16.6
3.25-4.45
59.8-68.1
2.07-7.17
155. -175.
27.2-32.8
1.16-1.41
.703-. 889
116. -143.
4.86-8.93
5.39-6.10
6.86-7.10
850. -1010.
123 .-149.
;630
85
233
39
^45.2
^5.3
;29.3
6.2
82.6
;4.7
14.8
3.8
63.0
;13.0
i
185
36
i
1.2
0.8
177
6
6.00
7.00
'984
138
*Provided by EPA. •
i
i
123
-------
sample was reported as 630 mg/L when the true value was 538; and the upper
acceptance limit was 617. However, in WP013, all these minerals iwere reported
within 10% of the true values and well within warning limits with "the
exception of TDS which was reported 15% higher than the true value in one
sample but within warning limits. The laboratory was notified or the results
and the probable reasons for the discrepancies were discussed. No test
samples were obtained for further analysis and evaluation because it was felt
that there were specific analysis problems that occurred during the time "the
WFG14 samples were tested and these had been corrected. :
7.2 Organic Tests
The procedures used for analyzing leachate and water samples for organic
constituents are (1) EPA method 415.1 for Total Organic Carbon (TOG), (2) EPA
method 624 for volatile organic priority pollutants, and (3) EPA method 625
.for said-volatile organic priority pollutants. Tables (4.4) and (4.5) list
the volatile and semi-volatile organics included on the Hazardous Substance
and Priority Pollutant List with the detection limits for each compound.
The analytical data provided by the Western Research Institute (WRI) was
generated using procedures that include comprehensive quality control and
quality assurance considerations. Surrogate spikes, duplicate runs, frequent
calibration and checking were included in the analyses procedures. For the
volatile analyses, a reagent blank is analyzed each 12 hours or once a day if
the column is used less than 12 hours. For semi-volatile analysis, a reagent
blank is processed each day. A sample of the distilled water ;used in "the
leaching tests was routed through the leaching column set up arid analyzed to
insure that the columns and distilled water were uncontaminated. The
follcwing paragraphs report the results of other specific quality assurance
procedures reported by WRI for this study. :
7.2.1 Calibration, and ^Standards ;
The GC/MS is hardware tuned to meet ion abundance criteria for
decafluorotriphenylphosphine (DFTPP) in the semi-volatile procedure, or
bromofluorobenzene (BFB) for the volatile organic priority pollutant analysis.
After tuning and before sample analysis, calibration standards' are prepared
and analyzed to define the linear range of the GC/MS system. Standards are
prepared from pure compounds or purchased as prepared mixtures. jail standards
used for methods 624 and 625 are evaluated with EPA traceable standards to
verify both purity and concentration. A minimum of 2-3 calibration standards
are used. " :
Internal standards are also included in both volatile and semi-volatile
procedures. The multiple internal standard procedures adopted by the EPA
Contract Laboratory Program are used at WRI. Three volatile internal
standards and five semi-volatile internal standards are used. These compounds
also act as retention time markers and provide a way to better quantify
nonpriority pollutant compounds that do not have calibration standards.
Prior to extraction or analysis, all samples are spiked with surrogate
compounds to monitor extraction or purging efficiencies. Surrogate recoveries
124 ;
-------
are evaluated against performance windows established by the EPA. Three
compounds were used in the VGA. The percent recovery of thesfe compounds is
shown in Table (7.5). The EPA performance windows are shown in parentheses
under the name of the compound. In all the volatile organic1 analyses, the
percent recovered is within acceptable limits.
Surrogate compounds were also used for the sani-volatile ,analyses but
because of the large dilutions needed for GC/MS analyses* no surrogate
recovery data was obtained. \
7.2.2 Duplicate Analyses
All TOG tests were performed in duplicate as standard procedure. Values
reported are the average of the two analyses. Duplicate•analyses were
performed for selected samples in the volatile and semi-volatile analyses.
Table (7.6) shows results of duplicate analyses of sampled of the Lurgi
unstripped process water and the last effluent of Run No. 6, the column test
using Paraho gas condensate for moisturizing. The results marked with the
designation J are approximations, either because standards were not available
or the values are near detection limits, in Table (7.6), a library search was
used to quantify the compounds designated by a J except for the methylene
chloride found in the Paraho sample, in this case, the levels found were near
the lower detection limits. •
Some of the values shown in Table (7.6) vary considerably: in the two
analyses. However, only one sample was provided for analysis. Therefore, the
duplicate sample was taken from the same bottle which contained headspace.
.Most of the differences shown in the table can probably be explained by this
headspace. i
Duplicate analyses for method 625 (semi-volatile organics) 'are reported
in Table (7.7). The data are for analyses of the unstripped Lurgi process
water.
125
-------
TOBLE 7.5. SURROGATE SPIKE PERCENT RECOVERS - MEDSCD 624
Sample ID
L3-6
L3-1
Ll-2
Ll-1
L2-2
L4-8
L4-1
L2-1
L2-1 dup.
Stripped Lurgi
Unstripped Lurgi
Unstripped Lurgi dup.
PR5-7
PR5-7 dup.
PR6-7
PR6-7 dup.
PR7-7
PR5-2
PR6-2
PR7-2
GGW-Col. 2
IW-Col. 1
Toluene-dR
( 86-119)
106
94
101
110
98
100
95
113
108
100
101
102
.96
101
97
98
100
100
108
103
96
96
Bronofluoro-
benzene
(77-120)
103
99
101
105
102
96
92
102
103
100
91
97
—
101
92
94
95
98
103
100
92
97
1,2-dichloro
ethane-d.-
(85-i 21 r
87
i
86
i 86
; 90
! 51
94
; 87
I
\ """"
92
97
83
82
- 85
94
• 79
'. 80
79
87
95
90
80
81
126
-------
TABLE 7.6.. RESULTS OF DUPLICATE ANALYSIS, VOLATILE OKGANICS
Lurai Unstripped Process Water ;
Concentration (nq/L) ;
irsethylene
chloride
Acetone
2-butanone
1,1, i-trichloro-
ethane
2-hexanone
4-methyl-
2-pentanone
acetonitrile
propanenitrile
2-pentanone
cyclopentanone
hexane
pentanenitrile
metbylgyride
2-heptanone
1st Analysis
1 05000 J
>90000
134000
13 OJ
10500
780
14000J
4200J
3600J
900J
7100J
600J
770J
1200J
2nd Analysis
1 05000 J
>85000
132000
11 OJ
10400
770
15000J
4800J
3700J
880J
7600J
680J
890J
1300J
Average . Percent
i Dif f erence
105000 0
1
875000 ; 1
133000 i 1.5
120 | . 1?
10450
775
14500
4600
3650
890
7350
640
830
1250
1
1.2
6.8
13
2.7
2.2
6.8
12.5
14.5
8.0
.
•
J = estimated value
(continued)
127
-------
Table 7.6. (continued)
Paraho II - Last Effluent of Run 6
Concentration (iig/L)
Compound 1st Analysis
methylene chloride
acetone
2-butanone
benzene
2-hexanone
acetaldehyde
acetonitrile
propanenitrile
butanal
pyridine
2.3J*
500B
148B
14
19
13
27
10
6.3
5.0
2nd Analysis
1.4J*
440B
13 OB
12
178
5.4
23
9.1
6.1
4.9
Average '• Percent
1 Diff erenee
1.85*
470 ;
139 ;
is ;
18
1
9 .2J*
25J '
9.55J !
6.2J* !
4.95J '
48
13'.
13
15
11
83
16
9.4
3.2
2.0
1
*Estimated value, below detection limits.
128
-------
TABLE 7.7. RESULTS OF DUPLICATE ANALYSIS. SEHI-VOATILE !OB3ANICS
Base NeutraJ
.s - Lurai Dnstricfoed i
Concentration (ptg/L)
Compound 1st Analysis
Aniline
Isopfaorone
32,750
6,000
Acids -
2nd Analysis Average Percent
Difference
33,250 33,000
6,750 6,375
Lurai Dnstripced
Concentration (ug/L)
Compound
Phenol
2-methylphenol
4-methylphenol
1st Analysis
11,000
9,750
1,600
2nd Analysis
10,750
10,750
1,150
Average
I
10,875
10,250
1,375
1.5
12.0
Percent
)ifference
2.3
9.8
33.0
129
-------
REFERENCES ',
Agarwal, A. K., 1986. Assessment of solid waste characteristics and control
technology for oil shale retorting, EPA 600/7-86-019, NTIS PB 86 198371.
American Society for Testing and Materials, 1978. Proposed methods for leach-
ing of waste materials. ASTM, Philadelphia, PA. ;
I
American Society for Testing and Materials, 1980. Annual book of ASTM Standards,
Part 19, Soil and Rock; Building Stones. :
Bruce, R. R. and A. Klute, 1956. The measurement of soil-moisture diffusivity
Soil Sci. Soc. Am. Proc., Vol. 20, pp 458-462.
I
Bryant, M. A., 1982. Dispersion in bi-modal oil shales. Unpublished M.S. Thesis,
Colorado State University, Fort Collins, CO. 84 p.
Clothier, B. E., D. R. Scotter, and A. E. Green, 1983. Diffusivity and one-
dimensional absorption experiments. Soil Sci. Soc. Am. J.,:Vol. 47,
pp. 641-644. ,
Corey, A. T., 1977. Mechanics of heterogeneous fluids in porous media. Water
Resources Publications, Fort Collins, Colorado. '
Fox, J. P., 1983. Leaching of oil shale solid wastes: A critical review,
prepared for Center for Environmental Sciences, University of Colorado
at Denver, supported by U.S. Department of Energy.
Franzini, J. B., 1956. Permeameter wall effects. Trans. Amer. Geophys. Union,
Vol. 37, No. 6, p. 735-737. - . i
Holtz, W. G., 1976. Disposal of retorted shale from the Paraho oil shale project
report to U.S. Bureau of Mines by Woodward-Clyde Consultants1. OFR 27-77.
McWhorter, D. B., 1982. Leaching and hydraulic properties of retorted oil shales.
Interim Report for EPA Cooperative Agreement CR-807668, Apri,l, 78 p. un-
published. ;
McWhorter, D. B. and V. A. Nazareth, 1984a. Leaching and selected hydraulic
properties of processed oil shales. Env. Res. Brief, EPA 600/D-84-228,
October. •
130
-------
McWhorter, D. B. and v. A. Nazareth, I984b. Leaching and hydraulic
properties of processed oil shales. Interim report for EPA CR-807668
April, unpublished. i
McWhorter, D. B. and G. 0. Brown, I985a. Adsorption and flow of: water in
nearly dry Lurgi retorted oil shale. Department of Agricultural and
Chemical Engineering, Colorado state University, Fort1 Collins,
Colorado. Prepared for Standard Oil Company (Indiana) and EPA. 94 p.
McWhorter, D.B. and G.O. Brown, I985b. Liquid and vapor''. transport-
coefficients for retorted oil shales. Proceedings of the 18th OJJ
Shale Symposium, Grand Junction, CO, April, Colorado School"• of Mines
Press, Golden, CO. ;
Nazareth, V. A., 1984. A laboratory column leach test for oil shale solid
wastes. Ph.D. dissertation. Department of Agricultural and ChemicaJ
Engineering, Colorado State University, Fort Collins, Colorado.
U.S. Department of the Interior, 1974. Earth Manual: A water Resources
Technical Publication, Second Edition, Reprinted 1980. pp. 491-505.
Rio Blanco Oil Shale Company, 1981, Tract C-a, modification to
development plan, Lurgi Demonstration Project.
Wilson, J. L. and L. W. Gelhar, 1981. Analysis of longitudinal
in unsaturated flow: l. ihe analytical method. Water
Vol. 17, NO. 1, pp. 122-130.
the detailed
dispersion
Resour. Res.,
131
-------
Appendix A
CODISPQSAL ESM RESULTS |
i
The results of the codisposal ESM tests are presented in Chapter 5.
appendix supplements that chapter and includes more complete results fran the
inorganic analyses. In addition, results fran the computer search used for
screening the samples for semi-volatile organic compounds are included in this
appendix.
Sample identifications are given in Table (A.I). Table (A.|2) gives the
complete results of the inorganic tests completed for the codisposal tests.
i
Samples analyzed in the semi-volatile analysis (EPA method 625) were
diluted and analyzed by GC/MS methods. Because of the complexity of the
chrcmatographs, the acid fractions and base neutral fractions were analyzed
separately. These results are presented in Chapter 5. in addition, however,
a computerized library search using a combined wiley-tBS library was performed
on the Lurgi base neutral samples to identify major compounds that are not on
the Hazardous Substance List. The results of these library searches are given
in Table (A.3). Estimated concentrations given in these tables should be
regarded as tentative since they are based solely on library matches.
However, the results of the library searches give an indication of sane
probable compounds present.
132
-------
Table A.l. Sample Identification
Run No.
l
(Lurgi IB-ID
. 2
(Lurgi IB-ID
3
(Lurgi IB-ID
4
(ALlis Chalmers)
5
(Paraho-ll)
6
(Paraho-iD
7
(Paraho-iD
Sample #
1-1
1-2
1-3
1-4
1-5
1-6
UnSt WH31
2-1
2-2
2-3
2-4
2-5
2-6
St WJEG1
3-1
3-2
3-3
3-4
3-6
4-1
4-2
4-5
4-7
4-8
5-1
5-2
5-3
5-4
5-5
5-7
6-1
6-2
6-3
6-4
6-5
6-7
7-2
7-3
7-4
7-5
7-7
Average
cum volume 1
(ml)
60
1192
2292
4872
7655 ,
11611 !
Maisturizing fluid
j
60 !
1199
2309
4910
7682
11866 :
Moisturizing fluid
615 ;
2290 '
4410 ;
7720
11200 ;
63
2763 :
4630
6063
10753
Moisturizing fluid
82
700
1468
2006
3158 !
Moisturizing fluid
75 !
690 i
1407
2066
3436 ;
78
761 :
1328
2115 !
3443
133
-------
Table A.l. (continued)
Run No.
8
(Chevron)
Sample #
8-1 a
8-1
8-3
8-4
8-5
8-6
Average
cum volume
(ml)
87
87
835
1754
2883
5219
CIW Moisturizing fluid
9 9-la 85
(Chevron) 9-1 91
9-1 91
9-3 457
9-4 1696
9-5 2819
9-6 4674
134
-------
I
a tl
!
0}
•H
a
EH u
s "?
a I
•6
u
OcoO\irj
in t-. c— ^o «o cs oo ot— WTf^ooo »-403 cso\vo«o u->
'
co
Tj-t-.oOO
CO O
CO
ooooooooooooooooooooooooo
CJtSi-f
B
dt
=*fe
^}
I I I I I I I I I I I I I I I I I I I I I I .
THTHr^rH»HiHc'
-------
I
O
a
en
•H
5,
s-i
1
S-l
s
a
=*fe
(U
oooooooooooooooooooooop^co
OOOOOOOOOOOOOOOOOOOOOOOOO
OOOOOOOOOOOOOOOOOOOOOOOOO
OOOOOOOOOOOOOOOOOOOOOOOOO
OOOOOOOOOOOOOOOOOOOOOOOOO
OOOOOOOOOOOOOOOOOOOOOOOOO
\S ^X ^^ *^^ ^X ^^ i
O*OC3OOO*OOOOOOO
-------
TABLE A. 2 (continued)
Research Water Analysis Report
Sample #
l-l
1-2
1-3
1-4
1-5
1-6
2-1
2-2
2-3
2-4
2-5
2-6
3-1
3-2
3-3
3-4
3-6
4-1
4-1 Extra
4- -2
4-5
4-7
4-8
St WRG1
UnSt WEG1
B
1.37
1.88
1.70
2.03
2.17
2.20
1;11
1.39
1.19
1.78
1.85
2.05
2.25
1.75
1.76
2.33
1.87
5.64
5.48
5.37
4.57
4.06
3.77
0.44
0.37
Ba
0.13
0.14
0.16
0.21
0.14
0.10
0.14
0.13
0.13
0.17
0.16
0.11
0.12
0.15
0.17
0.19
0.21
0.09
0.08
0.08
0.07
0.07
0.06
0.02
0.02
Pb Hg
mg/L
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01.
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
AS
0.034
0.031
0.014
0.004
0.004
0.001
0.031
0.028
0.015
0.006
0.004
0.004
0.003
0.017
0.002
<0.001
<0.001
<0.001
0.001
<0.001
<0.001
<0.001
0.001
0.285
0.286
Se
0.028
0.022
0.072
0.048
0.031
0.029
0.096
0.111
0.066
0.022
0.020
0.017
0.068
0.056
0.025
0.018
0.021
0.015
0.010
0.006
0.005
0.004
0.011
0.057
0.024
Total
Dissolved
Solids
32,700
30,800
18,700
4,480
4,260
3,990
18,200
16,600
11,000
4,560
3,970
4,020
13,200
9,200
4,460
3,850
3,820
9,750
9,410
8,040
6,260
4,850
3,840
8,430
48,900
137
-------
O «n <•$ r— Or-iOcn
i-lv>i-|O\xmo»
-3- OO ! CO
O TH CO
oo TT TH co
VO
T-) rH rH
«o en
§
-U
I
as
UJ
•H
en
I
•6
3
8
£
a
ooooooooooooooooo
MOO «OOr-lf~OOO-*'!l-OO\O*OTH
8
en
£
-H
•H Wg
§
.. s
e-i
OOOOc»THTHTH
ooooooooooooooooo
5
QJ
=*fe
4!
V T T T T
? T T V
? T ^ T
138
-------
ooooooooooooooooo
ooooooooooooooooo
OOOOOOOOOOOOOOOOO
OOOOOOOOOOOOOOOOO
V N^S^NX V >X NX V V
-U
M
ooooooooooooooooo
T3
§ W
.S T»
•
O
«s
•
-l»HOO
••••••••••••••••«
OOOOOOi-le^OOOOOOOOO
OOOOOOO\Oi-IOOO-*iHi-4OO
OOOOOOOTJ-OOCSCSOGSOCSO
oooooocnoooooooooo
OOOOOOOOOOOOOOOOO*
OOOOOOOOOOOOOOOOO
ooooooooooooooooo
fi
ooooooooooooooooo
OOOOOOOOOOOOOOOOO
OOOOOOOOOOOOOOOOO
CQ
I I I I I I I I I I I I I I I |
139
-------
TABLE A. 2 (continued)
Research Water Analysis Report ;
Sample #
5-1
5-2
5-3
5-4
5-5
5-7
6-1
6-2
6-3
6-4
6-5
6-7
7-2
7-3
7-4
7-5
7-7
B
13.30
1.91
2.21
2.34
2.44
2.46
0.73
1.73
1.78
1.87
1.98
2.65
1.92
2.15
2.70
2.02
2.16
Ba
0.12
0.08
0.07
0.08
0.09
0.08
0.02
0.08
0.06
0.06
0.05
0.05
0.09
0.07
0.07
0.08
0.08
Pb Hg
mg/L
<0.1 <0.001
<0.1 <0.001
<0.1 <0.001
<0.1 <0.001
<0.1 <0.001
<0.1 <0.001
<0.1 <0.001
<0.1 <0.001
<0.1 <0.001
<0.1 <0.001
<0.1 <0.001
<0.1 <0.001
<0.1 <0.001
<0.1 <0.001
<0.1 <0.001
<0.1 <0.001
<0.1 <0.001
As
0.442
0.068
0.044
0.020
0.010
0.006
0.915
0.081
0.058
0.044
0.024
0.013
0.026
0.011
0.005
0.002
0.002
Se
0.010
0.465
0.252
0.095
0.033
0.012
0.033
0.075
0.277
0.157
0.107
0.044
0.441
0.176
0.075
0.023
0.012
Total
Dissolved
: Solids
14,250
59,700
31,000
i 14,600
8,360
5,020
79,200
65,900
47,600
: 28,900
: 17,000
i 5,150
52,900
i 23,600
': 12,500
7,050
'. 4, 770
140
-------
•3
•H
8
"""
c*
f^P
pfT
§2
CO
M
a
i
«3
I
B
1
§ i
&
3
3 |
r-4
•H
O CJ
§ |
4J co
i— ) QJ O^
Is
(y
P M w
Q
1
,g
s,
|U
O O O O O O Tf
O CO C4 CO -5f CM CO
in p- o vo vo co in
« C4 *o
!^H t"* "*!f C*1 ^H ^5
«* ^
rH rH
VO Tf CO Tj- rH CO O
o* co in ^4* *<3* ^f
o\ eo «n •* •* ^ co
o o o o o o vo
o .0 o o o o co
C^ CO VO VO CO CO CO
VO VO CM O O O\ ^
VO t— VO VO VO VO rH
CM CJ ) CM CS O
rH rH rH rH rH rH rH
rH rH CO -^ in VO te
CO CO 00 00 CO CO CJ
O 0 0 O 0 O
^2 VO OO OO Tj» "tj"
rf rH vo vo Tf «n
CM CM rH rH rH rH
•^f CO Cl CM rH rH
•<*• rH CO 00 O •*
co O m rH r- cs
-------
•a
i
•H
-P
^-
O4
<
1
Is
o
»
u,
S3
13
g
1
ctf
e3
<2
9
1
°3
CQ GI
g
i-l
•4 s1
s
o>s!
•H
§
rH
ro Pn
r-H
03 "CT
i
r-H
CQ CT
te
•H
r"! "™
'O rr1
O g
S "?J
~ 5
0
CO CO
>0
CM
o
0
o
o
o
o
o
o
o
NX
f-
T-l
o
IP,
,_(
vo
<0
CM
e-i
o
o
o
o
o
o
o
vo
Jo
CO
M
0
o
CO
00
o
0
0
o
o
o
••a-
o
o
o
en
o
Tf
o
T-4
•
O
NX
o
o
TO
TO
O
O
o
Ig
6
vo
CO
o
o
o.
o
o
o
o
o
NX
o
NX
o
o
VO
ON
,_(
ON
CM
TO
•*
r-
o
o
o
0
en
O
o
as
ON
ON
O
o
«o
o
o
o
o
o
0
NX
o
NX
00
T-l
o
T-l
ON
T-l
•«a-
VO
TO
T-l
VO
o
o
o
T-l
CM
•«*•
O
o
NX
T-l
ON
O CM
vo * o
CM CM
0 O
en TH
ON ON
O rH
ON ON
ON tf
Cl CM
CM TO
O O
0 0
o o
O 0
NX NX
•*• T-l
CM CM
CM iH
O 0
o o
NX NX
>O VO
ON ON
142
-------
CO
O O O O O O
co 5
o
ON p- t— \o O oo
co rj- e«» co rH e««
O O O O O O
rH rH O O O O O
O O O O O O O
oo vo r- f c^ IN
rH rH O O O O
O O O O O O
CO CO ^* ^* CO ^O vH
o o o o o o o
o o o o o o o
o o o o o o
o o o o o o
MkV j-.l -_% ^.% T~l
CN CS CO CO d CS e~*
O O O O O O ^J
000000 S
CO O* «S M C< CS
o o o o o o
o o o o o o
Cft ^^ VO f^* CO CO
O O rH O O O
r» ve vo c- co o
o o o o o es
OOOOOO O OOOOOO
I
•H
8
o
a
cr> ^f tn
O O O O O O
CO «S « « C» CO
o o o o o o
3:
5
CO CO CO CO CO C^i ^^
o o o o o o o
CO CO CO CO if Tj-
rH rH rH rH T-l rH
£ rH CO •<*> «0 VO
<» 00 OO OO OO OO
RJ
_ co •* vo
7IIII I
3v w\, w\ CA O\ vs
0\
143
-------
Table A.3. Library Search Results for Semi-Volatiles, Ease Neutral Compounds
Sample: Stripped process water, run no. 2.
Library Search
Number
Compound Name
Estimated
Concentration
fig/L
1
2
3
4
5
6
7
Sample: Unstripped
Library Search
Number
1
2
3
4
5
6
7
3 -methyl-2-cyclopent en-l-one
Dime thy 1-cyclopenten-l-one
Dimethyl piperidine
£3 subst. phenol (?)
Unknown compound
Unknown compound
Unknown compound
process water, run no. 1.
Compound Name
Methyl pyridine
Methyl-2-cyclopenten-l-one
Dimethyl pyridine
Methyl-2-cyclysenten-l-one
Trimethyl pyridine
Dimethyl-2-cyclopenten-l-one
Unknown compound
144
! 125
58
36
i
28
35
'.. 38
i
; 45
!
!
! Estimated
Concentration
: t^g/L
55
i
; . 74
: 61
i 128
: 126
i 37
i
119
-------
Sample: Run 1, first effluent
Library Search
Estimated
Concentration
Number
1
2
3
4
5
6
7
8
Sample: Run 1,
Library Search
Number
1
2
3
4
5
6
Comcound Name
Methyl pyridine
Methyl pyridine
Dimethyl pyridine
Methyl-2-cyclopenten-l-one
Trimethyl pyridine
Dimethyl-2-cyclopenten- 1-one
Unknown compound
Unknown compound
last effluent
Compound Name
Methyl pyridine
Dimethyl pyridine
Trimethyl pyridine
Dime thy 1-2-cyclopenten-l-one
Dimethyl piperidine
Unknown compound
: W/L
1 27
1
: 2S
: 25
26
'• 28
35
; 13
: 53
!
: Estimated
Concentration
(ig/L
1 41
28
45
; 49
1 76
i
26
145
-------
Sample: Run 2, first effluent
; Estimated
Library Search Concentration
Number , Compound Name uq/L
1
2
3
4
5
6
7
8
9
10 '
11
: 12
13
14
15
16
2-pentenone
Methyl-2-cyclopenten-l-one
Methyl -2-pyrrolidinane (?)
Unknown compound
Unknown compound
Unknown compound
C3 subst. pi peri dine
Unknown compound
C3 subst. pyrrol idinane
C2 subst. piperidine
C2 subst. piperidine (?)
C2 subst. piperidine (?)
C3 subst. piperidine (?)
Unknown compound
Unknown compound
Unknown compound
; 102
1
: 74
i
' 64
i
104
: 60
; 145
: 171
59
1
i 89 .
70
! 62
78
133
; 55
: 176
377
146
-------
Sample: Run 2, last effluent
, Estimated
Library Search Concentration
Number Compound Name |
1 Methyl-2-cyclopenten-l-one ; 18
2 Unknown compound ! 15
3 Unknown compound ; 7
4 Unknown compound -5.5
147
-------
Appendix B •
ASTM AND RCRA RESULTS (from MCWhorter, 1982)
TABLE B.I.. CONCENTRATION IN ASTM WATER SHAKE TEST EXTRACTS - SPENT SHALES
Parameter
PH
EC
ALK
H2C03
HC03
co3
TDS
F
Cl
P04
N03
so4
Zn
Fe
Co
Li
V
NH3
8
Cd
Be
Mg
P
Si-
Mo
Mn
Mi
Na
Cu
A1
Ca
Ba
K
Cr
Sr
Pb
Ag
Tl
Se
As
Hg
Units
---
pmhos/cm @ 25°
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
I mg/l
mg/l
mg/l
mg/l
mg/l.
TOSCO
4:1
8.69
2650
164
0.88
191
4.0
1970
20.2
9.8
<0.03
16.0
1130
0.020
0.018
<0.005
0.079
0.007
0.644
2.500
O.001
<0.0005
35
0.18
0.8
1.56
0.053
<0.005
545
<0.001
2.15
31
0.117
8
0.025
0.53
<0.010
<0.001
0.006
<0.020
0.018
<0.001
II
20:1
9.01
740
96
0;24
107
4.7
510
7.5
1.9
<0.03
19.5
238
<0.005
<0.005
<0.005
0.039
<0.002
0.067
0.530
-------
TABLE B.2. COKONTRATIONS IN RCRA TEST EXTRACTS - SPENT .SHALES
Parameter
pH
EC
AUC
H2C03
HC03
C03
IDS
Cl
P04
N03
so4
Zn
Fe
Co '
Li
V
NH3
B
Cd
Be
Mg .
P
Si
Mo
_ Mn;
Ni
'Ma
Cu
Al
Ca
Ba
K
Cr
Sr
Pb
Ag
n
Se
As
Hg
Units
......
ymhos/cm@ 25°
mg/t
mg/t
mg/t
mg/i
mg/t
mg/t
mg/i
mg/t
mg/i
mg/i
mg/i
mg/t
mg/i
mg/i
mg/i
mg/t
mg/t
mg/i
mg/t
mg/i
mg/t
mg/i
mg/t
mg/i
mg/t
mg/i
mg/t
mg/t
mg/t
mg/t
mg/t
mg/t
mg/i-
mg/t
mg/t
mg/i
mg/t
mg/t
TOSCO II
7.72
5710
2737
140
3325
7.5
8180
22.2
<0.01
2.0
229
0.078
<0.005
<0.005
0.084
0.006
0.60
0.640
0.003
0.0045
81
0.6
1.2
<0.05
1.260
0.055
131
0.014
<0.02
1872
0.780 -
3.9
0.007
16
<0.01
0.002
<0.005
<0.02
<0.01
0.075
HYTORT
4.94
1540
430
13278
525
0.002
1740
8.95
<0.1
2.3
97
0.477
0.078
0.120
0.019
<0.002
1.66
0.340
0.013
<0.0005
85
0.4
3.7
<0.05
8.98
0.971
11
0.023
0.44
319
0.210
22
<0.005
1.0
<0.01
0.003
<0.005
<0.02
0.010
<0.001
LURGI
RG-I
8.06
4150
1612
37.8
1948
9.5
5690
7.1
<0.01
1.53
684
0.138
<0.005
<0.005
0.173
0.008
3.28
0.520
0.004
<0.0005
290
0.4
0.6
------- |