United States
Environmental Protection
Agency
Environmental Research
Laboratory
Athens GA 30613
Research and Development
EPA/600/S3-86/030 Nov. 1986
SEPA Project Summary
RECEIVED
MOV 2 5 1986
Development of Land Disposal
Decisions for Metals Using
MINTEQ Sensitivity Analyses
David S. Brown, Roger E. Carlton, and Lee A. Mulkey
A metal speciation modeling ap-
proach was developed for evaluating
potential mobilities of arsenic, barium,
cadmium, chromium, lead, mercury,
nickel, selenium, silver and thallium in
ground waters under conditions reflect-
ing leachate contamination from a
failed land disposal facility. A modified
version of metal speciation model
MINTEQ was used in combination with
a set of generic ground-water specifi-
cations and an En/pH uncertainty win-
dow to determine metal solubility limi-
tations. Metal speciation results were
interpreted in combination with an
infinite-source, steady-state advective
dispersion model used to estimate dilu-
tion during transport to a down-
gradient exposure point.
The ten metals were divided into
"mobile" and "relatively immobile"
groups. The mobile group included ar-
senic, barium, cadmium, lead, nickel
and thallium. At least one Eh/pH combi-
nation within the uncertainty window
was found for which each of mobile
metals was dissolved up to concentra-
tions sufficient to exceed health-based
thresholds at the hypothetical down-
gradient exposure point. The relatively
immobile group included chromium,
mercury, selenium and silver. These
four metals had limited solubilities at
all Eh/pH combinations within the un-
certainty window. After allowing for
dispersion, predicted concentrations at
the down-gradient exposure point
were below the drinking water stand-
ards. Chromium, mercury and selenium
were, however, more soluble under
conditions more oxic than those re-
flected by the uncertainty window. Ad-
ditional sensitivity tests were per-
formed to establish the Eh range above
which the respective solubilities were
predicted to increase significantly. The
order of dissolution with increasing Eh
was selenium > mercury > > chromium.
This Project Summary was devel-
oped by EPA's Environmental Research
Laboratory, Athens, GA, 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).
Background
The U.S. Environmental Protection
Agency often has the responsibility for
dealing with pollutants in situations
where the specific environment avail-
able for interaction with the pollutant at
each individual site is unknown. In these
cases, the decision process must in-
volve a generic component. The full re-
port outlines a generic approach for
evaluating the speciation and transport
of metals in ground-water environ-
ments. The primary emphasis was on
the simultaneous solution of metal spe-
ciation equations that determine dis-
solved phase metals concentrations.
Speciation and transport of arsenic,
barium, cadmium, chromium, lead,
mercury, nickel, selenium, silver, and
thallium in ground waters are consid-
ered in an example setting designed to
mimic the behavior of the leachate from
a failed land disposal site. The proce-
dures used are designed to be compat-
ible with recently proposed methods for
treating organic leachates. Expected
variations in site-specific chemical back-
grounds relative to the presumed
-------
generic ground-water chemical envi-
ronment were accommodated using an
Eh/pH uncertainty window approach.
The uncertainty window was developed
from statistical data on existing wells
around the country.
Potential transport of equilibrium-
dissolved-phase metal species resulting
from solution of the speciation equi-
libria are estimated using a back-
calculation screening procedure. Sim-
ply stated, the goal of the screening
procedure was to establish health-
based concentration thresholds for dis-
solved metal species that would be ap-
plicable to extracts of raw waste and
that were assured of being protective of
human health and the environment at a
down-gradient exposure point (possible
drinking water source). Health-based
concentration thresholds (CAoi) applica-
ble to the point of exposure served as
the starting point. These were appor-
tioned such that exposure to other
sources (surface water, air) were taken
into account. Elaboration of the detailed
conceptual approach used in establish-
ing the CAD|'s was outlined in the Fed-
eral Register, 40 CFR, Part 260, January
14, 1986. The CADI'S were used in con-
cert with a steady-state advective-
dispersion model and the metal specia-
tion model (MINTEQ) results to
back-calculation allowable concentra-
tions in leachates from the landfill.
Leachate concentrations were assumed
to be mimicked by pollutant concentra-
tions observed in extracts of the raw
waste material.
The primary emphasis of this work
was on evaluating the impact of metal
speciation equilibria on transport of
metals in a generically defined ground-
water environment.
Methods
A schematic representation of the
plume from a failed land disposal unit is
shown in Figure 1. Metals in the dis-
posal unit leachates are presumed to
equilibrate geochemically with the satu-
rated zone solution instantaneously at
the point of contact. Dissolved metals
concentrations in the resulting mixture
would be subject to reduction by precip-
itation of solid phases as well as dilution
and dispersion. Dilution due to disper-
sion was estimated using the model
represented in Equation 1 in conjunc-
tion with a Gaussian boundary condi-
tion. This model also was used to esti-
mate transport of organic leachates,
and it was designed to account for
Land Disposal
Unit
Down Gradient
Well
Saturated
Zone
o» • • • Ground Water
A I
Cross Section
Plume Boundary
Plan View
Figure 1. Schematic representation of plume from a failed waste disposal unit.
chemical-specific decay rates and veloc-
ity retardation due to sorption consis-
tent with several assumptions regard-
ing chemical behavior of the wastes and
characteristics of the transport medium.
Dxx^ + DVV 3y^+ Dzz a? ~*~~ =
where:
x,y,z = spatial coordinates in
the longitudinal, lat-
eral, and vertical direc-
tions, respectively (m)
c = dissolved concentra-
tion of chemical (mg/l)
x, Dyy, DH = retarded dispersion co-
efficients in the x, y,
-------
and z direction, respec-
tively (m2/yr)
V = groundwater seepage
velocity assumed to be
in the x direction (m/
YD
Rf = retardation factor (di-
mensionless)
t = elapsed time (yr)
X = effective first order de-
cay constant (yr~1)
I = dilution rate due to net
recharge (yr"1)
The retardation factor, Rf, and the ef-
fective decay constant, X, are defined as
follows:
R, =
(2)
and,
x = -
X2pbKd
!-PbKd
where:
Pb=bulk density of the porous
medium (g/cm3)
Kd = distribution coefficient (cm3/g)
9 = volumetric water content (cm3/
cm3)
XT = decay constant for dissolved
phase (yr1)
X2 = decay constant for sorbed phase
(yr1)
Because the behaviors of individual
waste constituents are highly depend-
ent on the chemical properties of the
compounds as well as on the input vari-
ables relating to the transport environ-
ment, selection of "reasonable worst
case" values would involve consider-
able uncertainty. To avoid this ambiguity,
a Fortran computer code, EPASMOD-P,
was developed for the dispersion model
that allowed monte carlo simulation of
the input parameter distributions. Back-
calculated concentration values were
determined for a large number of ran-
domly selected combinations of the input
variables (typically 5000). Results could
then be expressed as distributions of
back-calculated dispersion factors keyed
to relative probabilities of occurrence.
These dilution factors combined with
results of the metal speciation modeling
procedures described herein formed
the basis for back-calculating allowable
leachate concentrations. The dispersion
model is described in more detail in the
Federal Register, CFR 40, Part 260, Jan-
uary 14, 1986.
Two additional constraints on the dis-
persion model were employed in apply-
ing it to the transport of metals. The
degradation parameter was set to zero
(X = 0) in accordance with the assump-
tion that metals are conservative. This
also negated the consideration of sorp-
tion for metals because the retardation
effect would not change the concentra-
tion of a conservative pollutant ulti-
mately arriving at the down-gradient
exposure point.
The geochemical model MINTEQ is a
third generation equilibrium model that
uses the equilibrium constant approach
to solving the chemical equilibrium
problem. Its use in this work required
several modifications and additions.
Thermodynamic data for chromium,
mercury, thallium, and selenium were
not present in the original data base and
had to be added. Also, because the orig-
inal model would not handle more than
20 input components, the code was
modified for matrix expansion up to 50
components. Numeric underflows,
which caused the program to abort oc-
casionally when individual component
concentrations became very low
(=10~30 mol/1), were eliminated by
modifying the program to set the con-
centration to a lower limit of 10 30
whenever this level was reached. This
allowed the executions to continue.
Because chemical properties of
ground waters underlying existing and
future waste disposal sites were not
known, generic specifications were sub-
stituted. These were derived from large
volumes of well data extracted from the
EPA STORET data retrieval system. Be-
cause of the undue influence of extreme
values on the means, the medians for
each variable were selected for use.
These are tabulated in Table 1. All speci-
ation model runs were performed in the
chemical environment indicated, with
the exception of the Eh and pH.
Clearly, a single set of generic chemi-
cal specifications could not be derived
that would accurately reflect conditions
existing at all waste sites. To accommo-
date uncertainties associated with the
generic specifications, an uncertainty
window encompassing several combi-
nations of Eh and pH values was de-
fined. The window spanned the range of
Eh and pH variations one standard devi-
ation (a) either side of the median
Table 1. Median Ground- Water Chemical Specifications
Analytical Concentrations (mg/l)
Component
Arsenic
Barium
Cadmium
Chromium
Lead
Mercury
Selenium
Silver
Thallium
Mean
.021
.235
.030
.044
.035
.0006
.007
.019
.068
Stand. Dev.
.174
.251
.207
.121
.078
.0008
.008
.030
.037
Median
.010
.200
.005
.200
.010
.0005
.005
.010
.010
Aluminum
Calcium
Iron
Magnesium
Manganese
Potassium
Sodium
Bicarbonate
Bromide
Carbonate
Chloride
Nitrate
Phosphate
Sulfate
Sulfide
Organic carbon
1.134
79.000
2.780
51.240
.242
5.670
182.610
189.730
3.400
22.300
161.470
2.660
1.450
153.810
.471
12.762
2.620
118.478
7.365
134.840
.615
11.390
642.780
111.000
12.800
30.037
778.530
5.100
5.210
347.350
1.334
15.603
.200
48.000
.200
14.000
.040
2.900
22.000
190.000
.300
0
15.000
1.000
.090
25.000
.200
7.2
Temperature (°C)
pH
Eh (mv)
14.403
6.621
-10.535
5.287
1.285
155.786
14
6.8
-50
-------
values obtained from the STORET data.
The resulting window is illustrated by
the small rectangle in Figure 2. Shifts in
metal speciation equilibria possible
within the window were mapped by ini-
tiating MINTEQ model calculations
using nine different Eh/pH combinations
selected from within the window. The
nine combinations were formed by
combining each of three Eh values with
each of three pH values. The three val-
ues of each variable consisted of the
median value, the median value plus
-------
for chromium were at low pH (5.5) and
high Eh (106 mv) and the solubility was
limited (0.08 mg/l) by formation of the
Cr2O3 solid. Additional tests at Eh levels
above the Eh window (Figure 4) indi-
cated dissolved chromium concentra-
tions increased substantially above 700
mv. Values this high are however, ap-
5.5
parently, unlikely to be based upon the
STORE! data base.
Predicted dissolved mercury levels
were limited to 0.024 mg/l by the forma-
tion of Hg°(l) at the highest Eh available
from the uncertainty window. The tran-
sition to Hg(ll) ion began to occur at
about 300 mv (Figure 5). Above this
8.1
106 mv
ffe
1
Eh= 106 mv
pH - 5.5
IV
Eh = -50 mv
pH = 5.5
VII
Eh = -206 mv
pH = 5.5
II
Eh= 106 mv
pH = 6.8
V
Eh = -50 mv
pH = 6.8
1
1
VIII
Eh - -206 mv
pH = 6.8
1
1
III
Eh= 106 mv
PH = a.i
VI
Eh = -50 mv
pH = 8.1
IX
Eh = -206 mv
pH = 8.1
-206 mv
where: £nm = median £h = -50 mv
pHm = median pH- 6.8
0E = standard deviation of Eh= 156 mv
-------
Table 3. Worst Case MINTEQ Output Concentrations Compared to Total Metal Input tainty window. Considering the wide
Concentrations (mg/l)
pH,Eh
mv
Arsenic 8.1, 106
Barium 6.8, -206
Cadmium 6.8, 706
Chromium 5.5, 106
Lead 5.5, 706
Mercury 8.1, 106
Nickel 5.5, 106
Selenium 8.1, 106
Silver 6.8, -206
(Undeter-
minant)
Thallium 6.8, -026
DWS + back
Input/
/Output
0.060
3.58 x /O'7
7.70
7.70
0.070
0.070
0.140
0.00513
0.030
1.25 x 70~6
0.0035
0.0078
0.760
0.760
0.050
0.046
0.700
~9x 70~7
0.059
0.059
25 DWS
Input/
/Output
1.25
1.18
37.5
37.5
0.125
0.120
3.00
0.079
0.500
0.398
0.0750
0.0243
3.750
3.748
1.13
0.0808
2.25
=9x70~7
0.480
0.480
variability of reported groundwater Et,
50 DWS 700 DWS 10,000 DWS values, the ambiguity of interpreting E*
Input/ Input/ Input' data and the strong evidence for lack of
/Output /Output /Output redox equilibrium in many ground-
250 5.00 500 water systems, the expected mobilities
2.43 4.93 $00 °f these three metals remain very un-
certain. Better understanding of
75.0 750 75,000 ground-water redox processes is a criti-
75.0 750 13,967 cal need for future progress in predict-
ing subsurface geochemical phenom-
0.250 0.500 50.0 ena
0.250 0.500 50.0
6.00 72.00 7,200.0
0.079 0.079 0.079
7.00 2.00 200
0.887 7.67 799.7
0.7500 0.300 30.0
0.0243 0.0243 0.0243
7.500 15.00 1,500
7.497 14.97 1,500
2.25 4.50 450
0.0808 0.0808 0.0805
4.50 9.00 900
=9x70~7 =9x70~7 =9x70~7
0.960 7.92 792
0.950 7.29 81.3
Potential Chromium Mobility
3-
E 2.5-
.3
0
h*
•c 2 -
1
"5
8 1.5-
<3
-••.
1 ;-
0.5-
l
I B
o 4 r —
-200
pH = 5.5, 25 x DWS. Median Background
— B B —
I 1
0
B-
200
p
B — B— B— B-BD
1
400 600 500
Eh, Measured in mv
D In (mg/l)
Figure 4. Sensitivity of dissolved chromium concentration to Eh at pH = 5.5.
6
-------
0.08
0.07-
I
ta 0.06-
I
"5
.8
2 0.05
o
0.04-
0.03-
Potential Mercury Mobility
pH = 8.1, 25 x DWS. Median Background
\ I i
200 400
Eh, Measured in mv
D In (mg/U
600
800
figure 5. Sensitivity of dissolved mercury concentration to E h atpH=8.1.
1.00
0.00-
.1 -1.00-
c
1
-2.00-
-3.00-
-------
The EPA authors D. S. Brown (also the EPA Project Officer, see below), R. E.
Carlton, and L. A. Mulkey are with the Environmental Research Laboratory,
Athens, GA 30613.
The complete report, entitled "Development of Land Disposal Decisions for Metals
Using MINTEQ Sensitivity Analyses," (Order No. PB 86-233 186/AS; Cost:
$11.95, subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield. VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Environmental Research Laboratory
U.S. Environmental Protection Agency
Athens, GA 30613
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300
EPA/600/S3-86/030
'"- ?3^i:: n fi •> -
" - *.* ,-•> f. -
0000329 PS
U S ENVIR PROTECTION AGENCY
REGION 5 LIBRARY
230 S DEARBORN STREET
CHICAGO IL 60604
------- |