United States
Environmental Protection
Agency
Industrial Environmental
Research Laboratory
Cincinnati OH 45268
Research and Development
EPA-600/S2-84-123 Sept. 1984
Project  Summary
The  Effect  of Underground  Coal
Gasification  on  Ground Water

Edward A. Walters and Thomas M. IMiemczyk
  The potential effect of underground
coal gasification (UCG) on groundwater
has been examined in a combined field
and laboratory study. The study was
directed at Fruitland Formation subbi-
tuminous coal of the San Juan Basin
and at the groundwater found in this
coal seam. A field site  for a possible
UCG test was selected  and two wells
were drilled into the coal seam which
lies about 500 feet below the surface.
Groundwater samples were collected
and used to establish baseline features.
Core samples of overburden, underbur-
den and coal were used for baseline
characterization of the minerals.
  A number of static  leaching and
sorption studies were performed.
Overburden was leached by both acidic
and basic water. This material acts as a
strong  buffer, holding the  water at a
high pH. The amounts  of  Cl~ and  F~
leached from  overburden  were about
10 ppm and show little dependence  on
the pH of the leaching solution.
  Ash was generated at several different
temperatures. The results of leaching
studies of these ashes are reported. The
amount of ash dissolved depends upon
the ionic strength of  the leaching
solution in addition to pH. Predominant
species dissolved are calcium, alumi-
num, sulfate,  hydroxide and carbonate.
Indications are that solubility of the ash
decreases with increasing ashing tem-
perature.
  Adsorption capacity experiments
were conducted in an effort to deter-
mine the extent to which mineral matter
in Fruitland Formation coal contributes
to sorption.  Results  from the study
indicate that coal minerals account for
about  95  percent of  the sorptive
capacity of metal ions while the carbon-
aceous matter accounts  for only about
five percent. This sorption appears to  be
largely ion exchange with the K* and
Ca 2 cations found on the local clay
minerals. Distribution coefficients, Kd,
for sorption on coal were determined
for  the cations and  anions under
conditions reasonably similar to those
in the  groundwater of the San Juan
Basin. The effect of the natural chela-
ting  agent humic  acid was generally
found to be minimal. Numerical values
for the Kd are reported as a function of
initial concentration in solution and the
data are presented graphically  as
isotherms. The isotherms  indicate
multilayer adsorption as initial ion
concentration  increases and generally
fit a modified Langmuir equation well.
  Phenol adsorption by clay,  raw coal
and demineralized coal  was studied. It
was found that phenol adsorbs strongly
to the  carbonaceous matter and only
slightly to the minerals.
  The analytical  problem of silicon
determination  in complex matrices  by
graphite furnace atomic adsorption
spectroscopy techniques has been
studied and improved methods are re-
ported here.
  This Project Summary was developed
by EPA's Industrial Environmental
Research Laboratory, Cincinnati, OH,
to announce key findings of the research
project that is fully documented in a
separate report of the same title (see
Project Report ordering information at
back).

Introduction
  The subbituminous coal of  the San
Juan Basin of northwestern New Mexico
is one  of the  largest  deposits of the
Southwest. Most of the coal lies in deep
deposits which will probably only  be
tapped by underground coal conversion.
The most attractive conversion method at
the  present time is underground coal

-------
gasification (UCG). The principal consid-
eration  in the ultimate viability of the
UCG process is one of economics;  how-
ever, environmental factors play a role in
the economics of the  process.
  There are several  potentially serious
environmental problems with UCG: sub-
sidence, air quality  impacts (including
SOX, NOx and particulates), and ground-
water contamination. The possibility of
groundwater contamination is the most
substantial and least understood of the
three. The purpose of this project was to
study the potential for groundwater con-
tamination in the San Juan  Basin as a
result  of  UCG. The  approach was to
establish baseline characteristics of
groundwater, coal and rock at a prospec-
tive UCG site in the San Juan  Basin; to
determine solubility and rates of solution
of metals from the residue of a UCG burn;
and to determine sorptive characteristics
of metal ions and organics on  the rock,
clay and coal.

Baseline Studies
  A field site for a possible UCG test was
selected about ten miles northwest of
Farmington, New Mexico. Fruitland
Formation subbituminous coal  lies at a
depth of 500 feet at the test site and the
seam is 16 feet thick. The overburden is
part of the Kirtland Shale Formation and
the underburden is Pictured Cliffs
Sandstone.
  Two wells were drilled at the test site
and cores were taken from about 15 feet
above the principal coal seam to about 15
feet into the underburden.
  Water is found in  the Kirtland Shale
and typically fills wells to within 130 feet
of the surface in the  vicinity of the test
site. The  water is part of a  confined
aquifer which slowly drains in the
direction of the San Juan  River. The
aquifer is charged from exposed sections
of Kirtland shale and Fruitland Formation
coal at the mesas. The Fruitland Forma-
tion coal is water saturated and lies well
below the water table. The water migra-
tion rate appears to be one to two feet per
year, so that little flow occurs through the
UCG site. The flow  rate suggests that
equilibrium will be established between
dissolved ions in  the water  and any
adsorption or ion exchange sites in the
rock, clay, or coal. Permeability of the coal
seam is greater than that of the overbur-
den and underburden, so the flow occurs
through the coal seam.
  Water samples were collected from the
wells in the overburden region and in the
coal seam and they were analyzed using
standard techniques. Analyses showed
that the principal cations in the water are
Na+, K+, Ca*2, Mg+a and Sr+2; the major
anions are Cf, SO*"2, and HCCTs. These
species account for essentially all of the
mass observed  in the total dissolved
solids analyses. The pH of the samples
ranged between 7.8 and 10.7.
  Baseline determinations of the nature
of the Kirtland  shale  overburden  and
Pictured  Cliffs sandstone underburden
were made from rock samples taken from
the cores. The rock is composed almost
entirely of quartz, feldspar and clay. Clays
make up almost 30 percent of the  rock
with kaolinite and montmorillonite as the
principal  clay minerals. Cation exchange
capacities of 2.7 meg/100 g overburden
and  2.2 meg/100 g underburden were
determined.

Leachate Studies

Overburden
  Overburden samples were pulverized
and then the leaching characteristics as a
function of pH were studied using a batch
technique. Most of the metals (Ag, Ba, Cd,
Co, Cr, Cu, Mn, Ni, and Pb) showed no sig-
nificant leaching at pHhigherthanfive. In
general there is a decrease in the amount
of metal in  solution from a high value
near pH  1 to a much lower value as pH
increases. Iron, zinc and aluminum show
a general decrease in concentration to a
minimum in the vicinity of pH  6-7  with
increases as the  pH  approaches 10.
Sodium, calcium, magnesium and lithium
concentrations decrease as pH increases.
  The overburden has considerable
buffer capacity. Solutions of all but very
high and very low pH were buffered to a
final  pH of  8.3 to 8.4. This  may  help
explain the  groundwater basicity.  The
concentrations  of CI"  and F" in the
leachates were very low and showed no
trend with pH. The specific conductance
of the leachate was significantly higher at
the  extreme  ends  of the pH scale
indicating that more  of the  rock  was
dissolved at very high and low pH's.

Underburden
  Samples  of the underburden were
equilibrated with solutions  over the pH
range of 1-12. The rock has strong buffer
action as evidenced by pH's of the 7.1 to
7.3 for all the leachates. It is not clear why
the  pH of the underburden  leachates is
considerably lower than the groundwater
pH or the leachates from the overburden.
A comparison of the total dissolved solids
concentrations  of the leachates and
initial pH of the leachating solutions
shows a minimum total dissolved solids
concentration in the vicinity of pH 7 with
large amounts of rock dissolving in either
strongly acidic or strongly basic solutions.
Sulfate is the most abundant anion with
over ha If of the mass of the total dissolved
solids from the  pH 7 leach comprised of
sulfate. At the extremes of the  pH scale
sulfate makes up a small fraction of the
total dissolved solids mass. Concentra-
tions of CI", F~and CN~ were so  low at all
pH values that variations are insignificant.

Ash
  Coal from the Fruitland Formation was
used to prepare a series of ash samples to
simulate the material which will be left
after an underground coal gasification
experiment. Ash for a dynamic  leaching
study was prepared using ASTM proce-
dure D-3173 (1) which involves a final
ashing temperature of 750°C.  Distilled-
deionized water was gravity-fed through
a column filled  with the pulverized ash.
Analyses at the leachate showed Ca+2,
Na+, Mg*2and K+ present as the  principal
cations and Hg+2, Al+3, Pb+2,Zn+2, Fe+3and
Cu+2 present as trace constituents.
  Ashes were also produced using ASTM
procedure D-3173(1)  modified to give
final  ashing temperatures between
700°C and 1050°C, in an effort to study
the effect  of ashing temperature  on
leaching characteristics. Batch  leaching
tests using distilled-deionized water and
a simulated ground water were performed
on each ash. The simulated ground water
approximated the characteristics  of the
San Juan Basin in pH, conductivity and
primary ion concentration. A comparison
of the two  leaching solutions  indicates
that the simulated ground water  dissolves
more of the ash than the distilled-
deionized water does. This may  be due to
ionic strength or pH effects. The  solubility
of the ash  is also affected by ashing
temperature with solubility decreasing as
temperature increases.

Distribution  Coefficients
(Overburden and Underburden)
  Ultimately the rate of migration  of an
ion is determined by a heterogeneous
equilibrium  of the  ion between the
solution and the solid phases. The
equilibrium is characterized by a distribu-
tion coefficient  Kd which is defined as:
K -(activity of ion A in the solid phase)
    (activity of ion A in the liquid phase)
Since the  activities  are defined  in
concentration terms, Kd becomes the
ratio of the concentration of the ion (A) in
the solid phase to the concentration in the
liquid phase. It is inversely proportional to
the pollutant transport rate and directly
proportional to both the exchange capacity
and selectivity of the exchanger.

-------
  Distribution coefficients are dependent
upon the specific experimental conditions
for which the values are obtained. In an
effort to understand how some of the
experimental factors  influence Kd, we
measured the Kd of Cd+2 on the overbur-
den  and underburden  as a function of
Cd+z concentration and rock particle size.
  Samples of Kirtland shale overburden
and  Pictured Cliffs sandstone underbur-
den  were crushed and sorted into three
particle size  ranges. The samples were
equilibrated with solutions containing a
range of Cd+z concentrations. The concen-
trations of Cd+2, Na+, K+ and Mg+2 were
measured in the equilibrium solutions.
  The Ko's of Cd+2 on the overburden as a
function of Cd+2 concentration and rock
particle size is  shown in  Figure  1. A
similar plot was obtained for the under-
burden. For all of the cases studied, Kd
values increase with increasing Cd+2 ion
concentration  to a maximum  in  the
vicinity of 100 mg/l. This is followed by a
rather abrupt decrease which may  be
explained  in  two ways.  Either  the
maximum indicates saturation of the
surface sites by the Cd+2 added from the
solution or ion exchange occurs between
the  cadmium ions in solution and the
abundant metal ions bound to the clay
minerals. If the latter  is occurring, the
displaced ions (Na*, K+ and Mg*2) should
appear  in solution. Analyses show that
for each  Cd+2 concentration range the
amount of desorbed cation remains low
until the concentration of  added  Cd+2
reaches 1000 mg/l, at which point there
is an abrupt increase, by a factor of three
or more, in the amount of desorbed metal
ion.
  When the number of moles of Cd+z
absorbed is compared with  the  total
number of moles of Mg+2,  Na+ and K+
released, a ten-fold increase in Cd+z only
increases the moles of Mg*2, Na+ and K*
released by a factor of five. At the same
.time Kd decreases dramatically.  This
suggests that  ion exchange  is  the
dominant sorption  mechanism. As the
Cd*z concentration reaches very  high
levels the readily accessible ions in the
clay  matrix become saturated and the
amount of ion exchange does not keep up
with the increased Cd+z concentration.
Sorption Studies on Coal

Sorption Capacities
  An adsorption capacity experiment was
conducted in an effort to determine the
extent to which the mineral matter in
Fruitland Formation coal contributes to
sorption. The experiment was designed in
   JO4
   10*
   ro1
O Passed 4.76 mm Sieve

D 2.83-1.68 mm

A <149 iim
     ;o'1        10°         /o1         /o2

             Initial Cd Concentration (mg/l)
                                                                        10*
Figure 1.
Ka for Cd*1 on Pictured Cliffs sandstone as a function of initial concentration of
CcT*.
such a way as to distinguish between ion
exchange and simple solution.
  The coal was  demineralized by diges-
tion in a series of acids. This coal typically
contained 0.5 percent ash. The exchange-
able ion sites on the demineralized coal
and on the raw coal were loaded with Na*
by equilibrating  the coal with a sodium
chloride solution. Sorption  capacities
were determined by equilibrating the
Na-loaded coals with a solution of 1  N
ferrous ammonium sulfate. The solutions
were analyzed for Fe*2, Na+ and Ca*2.
  After blank correction for leachable Na+,
ion exchange and  sorption  capacities
were calculated for both raw and demin-
eralized coal. The total amount of Na+and
Ca*2 released into solution was interpreted
as the  ion exchange capacity of the coal,
while  the total  sorptive capacity  was
taken to be the amount of Fe+z adsorbed
by the coal, calculated by subtracting the
final Fe+2 concentration from the initial
concentration. The  difference between
                             total sorption  and sorption due to ion
                             exchange is simply called adsorption.
                               Table 1 shows that  inorganic matter
                             accounts for approximately 95 percent of
                             the sorptive capacity of Fruitland Forma-
                             tion coal while the carbonaceous material
                             contributes only about five percent to the
                             adsorption of iron. It can also be seen that
                             from one-third to one-half of the sorptive
                             capacity for iron may be directly attributed
                             to ion exchange.


                             Distribution Coefficients
                               Since distribution coefficients are
                             influenced by such factors as the compo-
                             sition of the liquid phase, the chemical
                             nature of the solid phase and pH, we tried
                             to approximate field conditions when we
                             determined the Kd's. The Kd's for each ion
                             were determined  in  the presence of all
                             the other ions selected for study and the
                             background  ionic  strength was adjusted
                             to a constant value with the  predominant
Table 1.   Sorption Capacities of Raw and Acid-Washed Fruitland Coal

                           Raw Coal                         A cid- Washed Coal
Total Sorption
  Capacity

Ion-Exchange
  Capacity*
Adsorption
  Capacity
  (by diff.)
                155±15meqFe+2/kg


                55±5 meq Na+/kg


                10±meq Ca^/kg

                65±4 meq/kg

                90±11 meq Fe*2/kg
8.1±0.8 meq Fe**/kg


2.2±0.2meqNa*/kg


0.5±0.1 meq Ca^/kg

2.7+0.2 meq/kg

5.4±0.6 meq Fe^/kg
*Small amounts of trace elements, such as K* or Mg*2, not replaced by Na* may have been displaced
 by Fe*2. Therefore, reported ion-exchange capacities are considered to be minimum values.

-------
cationic and anionic constituents of the
local groundwater. An attempt was made
to use realistic concentration ranges of
the ions of interest (Mn+2, Fe+3, Zn , Cd+2,
Al^ Cr+3, Co+2, Ni+2,  Cu+2,  Hg+2, Pb+2,
Be+2, CN", F", MoOF~>Mo04~2>Se03~2>As04~3.
  The shape of the cadmium isotherm,
shown in Figure 2, is representative of
the shape of the fluoride and molybdate
isotherms.  The isotherm of lead, shown
in Figure 3, is unique in shape compared
to the other ions  which were  studied.
Figure 4 shows the shape of the isotherms
for the remainder of the ions studied.
  Theoretically,  chelation of  a trace
element by humic material decreases the
amount of  free ion available for adsorp-
tion. Since a greater quantity  of trace
element  remains in  solution, Ka values
should decrease when chelation  by
humic substances  occurs. Our  data
generally show  small increases in the
solubility  of the metal  ions  in  the
presence of humic acid. Over the rather
limited range of conditions  explored, the
effect is not large. In the cases of Cd+2and
Pb+2 there may even  be a slight  increase
in Kd with the addition of humic acid.
  The effect of coal  particle size on the
adsorption  of trace metal ions was also
                     investigated. A simulated contaminant
                     solution consisting of Fe, Al, Mn and Zn
                     ions was prepared with the background
                     ionic strength adjusted to 800 ppm using
                     Ca+2 and Na+. The coal was separated into
                     four particle size ranges. Coal from each
                     size range was equilibrated with a known
                     amount of the contaminant solution and
                     then the  solution was analyzed for the
                     metals of interest.
                        Table 5 shows the results. The particle
                     size effects bear a close resemblance to
                     exponential functions, that is, Kd appears
                     to decrease exponentially with increasing
                     mean  particle size. The data could not be
                     made  to fit any function  which was
                     proportional to  r2 where  r is the mean
                     particle radius (r2 will be directly propor-
                     tional  to surface  area  whether the
                     particle is a sphere or  irregular figure).
                     We conclude, therefore, that some of the
                     adsorption  occurs at surface sites and
                     that there are additional sites inside the
                     coal  molecular network that become
                     more  accessible as the  particle  size
                     decreases.

-------
Table 3. The Adsorption of Selected Cations on Fruitland Formation Coal
No Humic Acid
Trace
Element
Mn



Fe



Zn



Cd



Al



Cr



Co



Ni



Cu



Hg



Pb



Be



Co
(mg/l)
0.186
0.963
5.00
21.1
1.08
5.72
27.3
103
0.210
1.07
4.74
20.6
0.032
0.063
0.123
0.490
20.2
39.3
75.3
154
2.52
4.48
9.10
17.35
2.16
4.21
8.16
16.45
1.60
3.10
6.36
13.6
2.02
4.08
8.12
16.5
11.7
23.5
49.4
97.7
6.85
11.9
21.6
40.8
0.64
1.12
2.31
4.31
Amt in So/n
(mg/ll
0.093
0.560
2.24
11.4
0.015
0.080
0.520
3.68
<0.02a
<0.02*
0.305
4.36
<7x10-<*
0.006
0.045
0.134
<0.04
<0.04
5.48
34.6
0.127
0.339
0.981
3.88
0.363
1.00
3.16
9.75
O.247
0.902
2.83
8.26
<0.003
0.071
0.179
1.08
<0.2
0.60
2.60
13.3
1.98
2.43
2.87
3.76
<0.002
0.14
0.81
2.33
Amt Ads
(mg/kg)
0.372
1.61
11.04
38.8
4.26
22.6
107
397
0.840
4.28
17.7
65.0
0.128
0.228
0.312
1.42
80.8
157
279
479
9.57
16.6
32.5
53.9
7.19
12.8
20.8
26.8
5.41
8.79
14.1
21.4
4.08
16.0
31.8
61.7
46.8
91.6
187
338
19.5
37.9
74.9
148
2.46
3.92
6.00
7.92
Ka
4.0±0.22
2.9±0.22
4.9±0.22
3.4±0.22
280±23
280±23
210±23
1 10±23
CO
OO
58±3.7
15±3.7
OO
38±3
7±3
11±3
OO
OO
51 ±3
14+3
75±4.4
49±4.4
33±4.4
14±4.4
20±5.9
13±5.9
6.3±0.37
2.8±0.48
22±5.8
10±3.9
5.0±0.96
2.6±0.89
CO
230±17
120±11
60±11
OO
153±4.3
72±4.3
25+4.3
10±2.4
16±2.4
26±2.4
39±2.4
CO
28±2.5
7.4±0.4
3.4±0. 1
Co
(mg/ll
0.187
0.963
5.03
21.4
1.18
5.67
26.3
106
0.210
1.07
4.75
20.8
0.032
0.063
0.123
0.490
20.2
39.3
69.3
153
2.36
4.42
4.94
17.6
2.08
4.22
8.10
16.9
1.50
3.06
6.40
13.8
2.03
4.11
8.03
16.6
11.4
23.1
48.9
95.8
6.85
11.9
21.8
41.4
0.64
1.18
2.38
4.18
Humic Acid,
Amt in Soln
(mg/l)
0.093
0.590
2.24
12.0
0.015
0.080
0.520
5.67
<0.02*
<0.02*
0.282
4.48
<7x10~t"
0.019
0.071
0.142
<0.04
<0.04
4.85
28.6
0.127
0.324
0.876
2.93
0.248
0.722
2.93
8.77
0.101
0.611
2.58
7.47
<0.003
0.084
0.178
0.908
<0.2
0.6
2.4
11.5
1.75
2.21
2.87
4.19
<0.002
0.17
0.78
2.30
5 ppm
Amt Ads
(mg/kgj
0.376
1.49
11.2
37.6
4.66
22.4
103
401
0.840
4.28
17.9
65.3
0.128
0.176
0.208
1.39
80.8
157
258
498
8.83
16.4
32.3
58.7
7.33
14.0
20.7
32.5
5.60
9.80
15.3
25.3
8.12
16.1
31.4
62.8
45.6
90.0
186
337
20.4
38.8
75.7
149
2.56
4.04
6.40
7.52

Ka
4.0±0.22
2.5±0.22
5.0±0.22
3. 1 ±0.22
310±23
280±23
200±23
70±23
CO
OO
63±3.7
15±3.7
OO
9±3
3±3
10±3
CO
OO
53±3
17±3
70±5.5
50±4.4
37±4.4
20±4.4
30±5.9
19±5.9
7. 1 ±0.37
3.7±0.48
55±5.8
16±3.9
5.9±0.96
3.4±0.89
00
190±17
180±11
70±11
CO
150±4.3
78±4.3
29±4.3
12±2.4
18±2.4
26±2.4
36±2.4
OO
24±2.5
8.2±0.4
3.3±0.1
    symbol < indicates the concentrations were below the given detection limit.
Adsorption of Organics

  The  potential  for  release  of  large
quantities of both inorganic and organic
contaminants  is one of the  principal
environmental concerns associated with
UCG.  A large  number of organics are
generated during the pyrolysis stage  of
UCG; these are transported in the vapor
phase to the region surrounding the burn
where they then enter the groundwater.
Comparison with wastewater effluents
from surface coal processing plants has
produced the estimate that concentra-
tions of phenolic compounds alone could
reach levels of 1000-3000 mg/l. On this
basis, we have chosen to use phenol as a
model organic and perform a study of the
interaction of phenol with  the UCG
surroundings.
  The adsorption of phenol on standard
clay (Na-Montmorillonite), overburden,
coal and demineralized coal was studied
by equilibrating these substances with
solutions of varying phenol concentration.
The liquid was  separated from the
suspended matter and analyzed for
phenol.
  The  adsorption characteristics  of
phenol on Na-Montmorillonite were
5

-------
Table 4. The Adsorption of Selected Anions on Fruit/and Formation Coal
No Humic Acid
Trace
Element
CAT



F



Se



Mo



As



Co
(mg/l)
1.25
2.58
5.63
11.2
48.4
75.4
121
203
16.9
42.1
77.2
176
21.4
44.3
95.4
182
21
45
86
164
Amt in Soln
(mg/l)
<0.03"
<0.03a
<0.03B
<0.03"
24.0
41.0
67.0
90.0
12.8
38.0
72.8
155
16.4
29.5
51.4
145
19
46
86
164
Amt Ads
(mg/kg)
4.96
10.3
22.5
44.8
97.6
138
216
452
12.8
16.4
17.6
84.0
20.0
59.2
176
150
=0
=0
=0
=0
*d
>1500
>1500
>1500
>1500
4. 1 ±0.23
3.4±0.23
3.2±0,23
5.0±0.23
0.93±0.02
0.43±0.02
0.24±0.02
0.52±0.02
1.2±0.60
2.0±0.60
3.4±0.60
1.0±0.60
=0
=0
=0
==0
Co
(mg/l)
1.20
2.48
5.25
10.7
47.6
74.6
120
202
16.9
42.5
76.4
172
22.0
43.6
90.6
180
21
45
91
163
Humic Acid,
Amt in Soln
(mg/l)
<0.03a
<0.03"
<0.03"
<0.03a
22.6
41.0
68.0
92.0
13.7
38.4
71.7
151
18.4
33.9
52.9
149
21
48
88
160
5 ppm
Amt Ads
(mg/kg)
4.80
9.92
21.0
42.8
100
134
208
440
12.8
16.4
19.2
84.0
14.4
38.8
151
125
=0
=0
=0
=0

Ka
>1500
>1500
>1500
>1500
4.4±0.23
3.3±0.23
3.1 ±0.23
4.8±0.23
0.93±0.02
0.43±0.02
0.27±0.02
0.56±0.02
0.78±0.06
1.1 ±0.60
2.8±0.60
0.8±0.60
=0
=O
=0
==0
"Below the detection limit.
examined because montmorillonite makes
up approximately half of the clay minerals
found in the overburden and San Juan
Basin  groundwater contains a  large
concentration of monovalent  cations.
Adsorption  on Na-Montmorillonite con-
forms to the linearized Langmuir isotherm
over the equilibrium phenol concentration
range  of 50-1000 ppm.  The observed
adsorption capacity is 3.52 mg phenol per
gram of clay. At concentrations greater
than 1000 ppm the amount of phenol
adsorbed  increases substantially; the
amount adsorbed accelerating for some
magnitude beyond  this concentration.
This behavior is characteristic of physical
adsorption  described by the Brunauer-
Emmett-Teller(BET) equation which
takes into account multiple layer forma-
tion of  physically adsorbed molecules (2).
  Phenol  adsorption on Kirtland Shale
overburden is quite similar to adsorption
on  Na-Montmorillonite. Adsorption fol-
lows the Langmuir equation up to about
1000  ppm phenol solution with  an
adsorption capacity of 4.82 mg phenol per
gram of overburden. This implies that
sorption sites on the overburden are likely
of the same nature as those in clays;  in
fact, sorption on the  Kirtland Shale  is
probably  due to sorption on the clay
components  of the  shale.  Sorption
appears to be physical adsorption with
multilayer effects showing up at high
phenol concentration.
  For adsorption on coal there are two
linear segments to the Langmuir isotherm
followed by the rapid increase in adsorp-
tion of equilibrium phenol concentrations
in excess of 4500 ppm. In the case of raw
coal the overall adsorption data may be
interpreted as a sum  of  contributions
from overburden clay minerals in the coal
plus that from the organic matter of coal.
The adsorption capacity of the organic
component has  a substantially higher
value  than the clay minerals. The signifi-
    1.5  •
                            cance of this is that a large quantity of
                            phenol can be adsorbed on coal so that the
                            coal remaining in the aquifer down-dip
                            from the burn site should retard phenol
                            migration quite effectively.

                            Conclusions
                              From  many points of view northwest
                            New Mexico represents a very good site
                            for an  underground coal  gasification
                                                      •— No Humic Acid


                                                      O— Humic Acid Added
Figure 2.
          O.05             0.10

                    Cd in Solution (mg/l)

Sorption isotherm of cadmium.
                                                        0.15
0.20

-------
   150
\

I
I
   100
    50
                                                        No Humic Acid
                                                     O—Humic Acid Added
                   1.0          2.0          3.0

                             Pb in Solution (mg/lj
                                           4.0
                                                        5.0
 Figure 3.   Sorption isotherm of lead.
     60
I   40

I
I
    20
                                                        No Humic Acid
                                      O—Humic Acid Added
   Figure 4.
       1.0            2.0

              Cr in Solution (mg/ll

Sorption isotherm of chromium.
                                                  3.0
4.0
               (UCG)  project.  The proximity of  major
               power plants and a power distribution
               network, as well as extremely large, deep
               coal  deposits make  such  a project
               attractive. The environment of the locale
               is high desert and therefore characterized
               by low human and animal  populations
               and sparse vegetation. Background
               studies  indicate the  groundwater is
               brackish.  The combination  of  these
               factors allow one to project a  minimum
               environmental impact of such a site.
                 Static leaching of  overburden and
               underburden indicates that pH is  a very
               important variable, with large quantities
               of rock dissolving at very high and very
               low pH. Baseline studies show that the
               groundwater is strongly alkaline and the
               most abundant dissolved anion issulfate.
               Trace metal measurements indicate that
               only at low pH values are  significant
               amounts extracted.  The  distribution
               coefficients determined here for several
               metals were found to be  functions of
               initial concentration of metal and particle
               size and when presented as isotherms
               the data fit the Freundlich equation.
                 Sorption studies on coal  indicate that
               the inorganic matter in the coal accounts
               for approximately 95% of the adsorption
               of trace metals. The amounts of metals
               adsorbed onto the coal and the distribu-
               tion function for the process have been
               measured in the presence of humicacid.
               The  results indicate that the affinity of
               humic acid for the trace metals studies is:
               Fe+2> AI+3> Pb+2>Cr*3>Nra> Mn+2>
               Cd+2 > Be+2. By fitting the sorption results
               to a  Langmuir isotherm the sorption
               capacity of Fruitland Formation coal was
               found  to be 1.5 g/kg for cations and 2.7
               g/kg for anions.

               Tables.   Effect of Particle Size on Adsorption
Trace
Element
Al



Mn



Fe



Zn



I



Co
(mg/ll
50



13



120



10



193



Particle size
range, mm

-------
  References
   1.  "Standard Methods of Laboratory
      Sampling and Analysis of Coal and
      Coke,"D-3173, American Society for
      Testing and Materials, 1976.
   2.  A.W. Adamson,  Physical Chemistry
      of Surfaces,  3rd Edn., John  Wiley
      and  Sons, New York, 1976, Chapter
      IX.
         Edward A. Walters and Thomas M. Niemczyk are with University of New Mexico,
           Albuquerque. NM 87131.
         Patricia S. Fair is the EPA Project Officer (see below).
         The complete report,  entitled "The Effect of Underground Coal Gasification on
           Ground Water," (Order No. PB 84-220 318; Cost: $13.00, subject to change)
           will be  available only from:
                National Technical Information Service
                5285 Port Royal Road
                Springfield, VA22161
                 Telephone: 703-487-4650
         The EPA Project Officer can be contacted at:
                Industrial Environmental Research Laboratory
                U.S. Environmental Protection Agency
                Cincinnati, OH 45268
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use S300

-------