Ecological Chemistry. St. Petersburg, Russia
SIMULATION OF HEAVY METALS MIGRATION IN PEAT DEPOSITS
I.L. Kharkhordin and F.G. Atroshchenko
St. Petersburg department of the Geology Institute of Russian Academy of Sciences,
196199 St. Petersburg
Interdepartmental Scientific Centre of Hydrogeoecology, St. Petersburg State University,
199178 St. Petersburg
(Accepted for publication September 30, 1998)
The problems of computer simulation of heavy metals migration in swamp deposits is considered taking as
an example a site at slime storage of the Arkhangelsk Thermal Power Station. Some principal results of the
simulation were found. Heavy metals in swamp and bog deposits are mainly in the sorbed state. The ratio of
sorbed to dissolved forms is 10
3
–10
4
. In low3lying swamps with near3neutral medium, peat exhibits consid3
erable buffer capacity with respect to both acid and alkaline solutions. In the initial migration state, heavy
metals concentration in water is virtually independent of their content in waste and is determined by the
composition of the exchange complex in peat. At low heavy metals concentrations in waste, interaction with
peat can lead either to their sorption or to their accumulation in solution. If the initial metal concentration
exceeds the first milligrams per litre, processes of self3purification predominate.
Key words: swamp deposits, heavy metals, slime storage, ion exchange, computer simulation.
Introduction
The prediction of changes in the chemical
composition of water in its interaction with
swamp deposits is necessary for solving many
practical problems. On the one hand, swamp
can accumulate heavy metals [1,2], radioactive
substances [3], and oil products. Therefore,
they can be used as slime storage spaces and
places of incompletely purified waste and drill3
ing water discharge. On the other hand, the
change in physico3chemical conditions in
swamp deposits under the effect of anthropo3
genic factors (drainage or flooding of swamps,
acid rains, etc.) can lead to desorption of toxic
elements previously accumulated in peat. The
aim of the present paper is to develop approach3
es to computer simulation of these processes
taking as an example a site of slime storage of
the Arkhangelsk Thermal Power Station
(ATPS).
Characteristic of the research location
Slime storage of ATPS is located in the delta of the
Northern Dvina River in the region of extensive swamp
deposits. It was constructed in the 1970s and is of a typ3
ical dike character. The discharged material differs in
chemical composition, volume, and degree of toxicity.
Therefore, the storage is divided in three sections: 1)
chemically purified water (CPW) about 130 thousand
m
3
in volume, 2) neutralized water (NW) — 50 thousand
m
3
in volume, 3) acid boiler washing water (AW) —
15 thousand m
3
in volume. At present only section CPW
and AW are used for discharging slime, and section NW
is temporarily closed. Taking into account the ratio of
solid to liquid phases in the discharged pulp, it was cal3
culated that in section CPW only 20% of capacity is filled
with slime, in section AW — about 10%, and in section
NW — not more than 2%. However, waste volume dis3
charged into these sections is such that if water leaks
from them is excluded, as is required the Norms and
Rules of designing these constructions [4], it would be
necessary to build a new slime storage every year. At
present the level in this storage is approximately con3
stant. This implies that leak volume corresponds to the
volume of discharged waste and atmospheric precipita3
tion. In order to evaluate the effect of leaks filtered
through the dike, extensive investigations were carried
out including testing of waste and pore solutions of
swamp deposits.
Chemical analyses were performed in the laboratory
of joint3stock society “Mekhanobr” by standard methods.
The averaged initial composition of swamp water and
discharge to sections CPW and AW taken for simulation
is given in Table 1. We can note the following changes in
chemical composition near sections CPW and AW found
from data on hydrogeochemical testing.
1. Section CWP. In underground water the concen3
trations of magnesium and sulphate ions increased twice.
Manganese concentration increased four times, and Ni,
Co, and Zn appeared in quantities exceeding LAC two and
66
I.L. Kharkhordin and F.G. Atroshchenko
three times, although their concentrations in the section
itself are not high.
2. Section AW. During filtering swamp deposits, wa3
ter from this section is transformed indo weakly acid and
neutral water and hydrobionts appear. The concentra3
tions of Mo, V, Zn, and Cu decrease to LAC and the con3
centrations of Ni and Co to background values charac3
teristic of water of swamp deposits. The content of Co,
Mn, and Al changes only slightly.
3. Section NW. The transformation of chemical com3
position of water of this section in the swamp deposits is
accompanied by increasing concentration of Mn and Zn,
whereas Ni, Co, Mo, and Cu content decreases to LAC or
to background values.
Process of interaction between metal
ions and swamp deposits. Evaluation of
model parameters
For quantitative description and prediction
of directions of further changes in chemical
composition of underground water, computer
simulation of migration of solutions contain3
ing heavy metals in swamp deposits was devel3
oped.
In spite a long history of studying the proc3
esses of solutions interaction with humus ma3
terial, an effective procedure for their mathe3
matical description is not yet available. This
mainly due to complex structure of humic com3
pounds, which contains various functional,
groups [5]. It is also important that the com3
position of these compounds is variable. There3
for, it is impossible in principle to isolate indi3
vidual humic compounds and one must work
with these complex mixtures. For instance,
according data in ref. [6], fulvic acids in natu3
ral waters migrate in the form of associates
with molecular weight from 300 to 60000 (de3
pending on their concentration and pH of so3
lution). These authors suggest that monomers
of fulvic acids contain two or three carboxylic
(K
1
= 2
⋅
10
–3
, K
2
= 5
⋅
10
–5
) and one or two phenol
groups. The study of fulvic acids fractionation
on ion3exchange materials has shown that the
fraction relatively enriched with nitrogen is
retained by cation exchangers at pH = 2 [7].
This fact indicates that nitrogen is predomi3
nantly contained in amino3groups. Cationic
groups in natural organic compounds are prob3
ably sorbed by clay minerals [8]. It is possible
that in this process hydrogen bonds and bonds
with the participation of metal ions are formed
It is established that mercapto3groups are
present in peat. Their concentration depend on
redox conditions [10].
The complexation model is usually applied
for the mathematical description of interaction
of dissolved fulvic and humic acids with metal
ions. However, in this case equilibrium con3
stants are tentative and depend on the solution
composition, in particular on pH [6]. This
makes it difficult to calculate the composition
of multicomponent systems. On the whole the
stability constants of hydroxyfulvic metal
complexes range from 10
3
to 10
11
increasing in
the following series:
Sr(II) < Ca(II) < Fe(II) < Ce(III) < Y(III) <
Cu(II) < Ru(IV) < Fe(III) < Sb(III) < Au(III) <
Hg(II) [11].
The model of cation exchange is generally
used to describe metal sorption on humic com3
pounds in the solid state of swamp deposits.
Either the ternary system Me
I
3Me
II
3H
+
is con3
sidered or the cation3exchange system is re3
garded as a function of pH and solution com3
position. The principal investigation methods
are potentiometric titration of humic materi3
al in salt solution and experiments on ion ex3
change. It is observed that the mechanism of
chemical bonding of metal with humic acids
and with peat are not very different [12]. More3
Table 1
Model compositions of solutions
Waste waters
Index
Measuremen
t units
CWP
AW
Swamp
waters
(background)
t,
°
C
6.0
6.0
6.0
pH
9.0
3.6
7.03
HCO
3
–
mg
⋅
l
–1
206
2
206
Cl
–
–“–
74
500
27
SO
4
2–
–“–
399
750
20
Ca
2+
–“–
112
180
43
Mg
2+
–“–
25
45
2.5
Na
2+
–“–
148
364
34
Mn
2+
–“–
0.05
1.87
0.26
Ni
2+
–“–
0.002 17.37
0.001
Zn
2+
–“–
0.001
7.17
0.5
Cu
2+
–“–
0.001
0.71
0.0003
Pd
2+
–“–
0.002 0.086
0.002
Cd
2+
–“–
0.015 0.028
0.005
67
Simulation of Heavy Metals Migration in Peat Deposits
over, most metal ions retain their hydrate
shells and the bond is of electrostatic charac3
ter, whereas copper ions form predominantly
more stable covalent bonds. The absolute rate
of metals adsorption on peat decreases in the
following series: Pb
2+
> Cu
2+
> Cd
2+
> Zn
2+
>
Ca
2+
and the time of half3reaction of sorption
(desorption) ranges from 5 to 15 s [13].
In this work for the description of metals
migration in swamp deposits, the model of cat3
ion exchange is also used. Note that the model
of surface complexation is probably better suit3
ed to the natural process but at present in the
published literature the quantity of experi3
mental data is not sufficient to evaluate the
required constants.
The PHREEQC program [14] freely distrib3
uted by the USA Geological Service is used for
the simulation. This program makes it possi3
ble to calculate physico3chemical equilibrium
in multiphase and multicomponent systems
taking into account complexation in solution,
precipitation (dissolution) of minerals and ion
exchange. The user can introduce several types
of ion exchangers with different values of ex3
change constants. Particular attention should
be paid to the form of expression of constants
for ion exchange equilibrium in this program:
interactions are written in the form of half3
reactions:
nX
–
+ Me
n+
= X
n
Me
and the law of mass action is corresponding3
ly expressed as
+
−
⋅
=
]
Me
[
]
X
[
]
Me
X
[
K
n
n
n
where [X] is the fictious activity of positions
in the exchange complex (value analogous to
electron activity in recording redox reactions
in the form of half3reactions), [Me
n+
] is the cat3
ion activity in solution, and [X
n
Me] is the ac3
tivity of the cation bound in the exchange com3
plex. This form of evaluating ion3exchange
processes is more convenient for mathemati3
cal simulation and makes it possible to avoid
restrictions appearing when usual constants
for exchange of one cation by another are ap3
plied. For example, it is necessary to enter a
certain non3zero concentration for the “basic”
cation through which the expressions of ex3
change constants are written for other ions.
Exchange constants were evaluated by us3
ing the experimental results published in lit3
erature. Andre and Pijarovski [15] have car3
ried out the most complete investigation of
exchange equilibrium in peat at different pH
values with the participation of K
+
, NH
4
+
, Ca
2+
,
and Mg
2+
[15]. Exchange constants calculated
on the basis of these data are given in Table 2.
It was possible to obtain satisfactory coinci3
dence of calculated and experimental data by
assuming the existence of three types of ex3
change parts corresponding to groups with dif3
ferent degree of acidity (real distribution of
exchange positions with respect to density of
bonding to hydrogen atoms is more complicat3
ed). Capacity ratio of strongly acidic, acidic
and weakly acidic groups is 7 : 12 : 11.
These ratios were used in estimating the ex3
change constants for heavy metals on the basis
of published experimental data [12,16–18, etc.].
Ion exchange constants on peat accepted in fur3
ther calculations are listed in Table 3.
The model is a series of 20 cells. At the zero
step, solution composition in cells corresponds
to that of water of swamp deposits, and the
composition of cations in the exchange com3
plex is calculated on the basis of the condition
of equilibrium with it. It should be pointed out
that the ratio of sorbed to dissolved forms for
different metals is 10
3
–10
4
. In other words,
under natural conditions heavy metals in
swamp deposits are mainly in the sorbed state.
These results are in good agreement with data
of geochemical peat tests. Further, it is as3
Table 2
Ion exchange constants for swamp deposits
calculated from experimental data [15]
Groups with different degrees of
acidity
Cation
Strongly
acidic
Acidic
Weakly
acidic
H
+
1.75
4.48
7.41
K
+
0.00
0.00
0.00
Ca
2+
1.0
2.27
1.7
Mg
2+
0.65
1.93
1.3
68
I.L. Kharkhordin and F.G. Atroshchenko
7.0
6.5
6.0
5.5
5
4.5
0
4
8
12
16
20
pH
Cell number
Fig. 1. Change in pH during waste filtration from the AW section in swamp deposits at the 10, 50, 300, and 10003th
calculated steps
10
50
300
1000
sumed that the solution corresponding to wa3
ter composition in the slime storage begins to
be discharged into the first cell. During one
calculated step, the solution passes to the next
cell. After each passage solution composition
was brought into equilibrium with that of the
exchange complex. In the simulating brine
migration from section AW, the possibility of
carbon dioxide dissolution with the formation
of hydrocarbonate ions absent in the initial so3
lution was also taken into consideration. Sat3
isfactory agreement between calculated and
measured concentrations of hydrocarbonate
ions in solution was obtained at lg [CO
2
] = – 1.5.
Simulation results and their discussion
Let us consider in greater detail the pecu3
liar features of transformation of the chemi3
cal composition of waste in swamp deposits.
Section AW.
In the initial stage of filtration
of waste water from slime storage, as a result
of interaction with swamp deposits their chem3
ical composition is transformed drastically:
the pH of solutions increases (Fig. 1) and the
concentrations of micro3 and macrocompo3
nents change. High buffer capacity of peat
with respect to acid and alkaline solutions over
a wide pH range is due to the presence of dif3
ferent groups in the exchange complex. Even
after repeated washing with an acid solution,
the pH of swamp deposits are 5.5–6. The
changes in concentrations of macrocompo3
nents are of a smoother character than for
rocks with an exchange complex of a single
type. In the initial stage calcium content
(Fig. 2) increases with respect to that in the pri3
mary solution, whereas the content of sodium
Table 3
Ion exchange constants for heavy metals
in swamp deposits
Groups with different degrees of acidity
Cation
Strongly
acidic
Acidic
Weakly
acidic
Mn
2+
0.8
2.0
1.6
Cd
2+
0.8
2.0
1.6
Zn
2+
1.0
2.4
1.8
Ni
2+
1.2
2.9
2.2
Pd
2+
1.3
3.2
2.4
Cu
2+
1.4
3.5
2.8
10
50
300
1000
Cell number
Ca,
m
g
⋅
l
–1
600
500
400
300
200
100
0
4
8
12
16
20
and magnesium decreases, later gradual sta3
bilization takes place. The transformation of
the macrocomponent composition is virtually
completed at the thousandth calculated step
(fifty3fold change of solution over the whole
simulation range). Effective purification of
waste water from heavy metals takes place.
Fig. 3 shows changes in nickel concentration
the initial content of which in the AW section
is about 17 mg
⋅
l
–1
. After interaction with
swamp deposits it decreases by more than four
order of magnitude. Similar changes in concen3
Cell number
Fig. 3. Change in nickel concentration during waste filtration from the AW section in swamp deposits at the 10, 50,
300, and 10003th calculated steps.
4
8
12
16
20
0
20
16
12
8
4
Ni,
m
g
⋅
l
–1
10
50
300
1000
Fig. 2. Changes in calcium concentration during waste filtration from the AW section in swamp deposits at the 10, 50,
300, and 10003th calculated steps.
70
I.L. Kharkhordin and F.G. Atroshchenko
Fig. 4. Change in zinc concentration during waste filtration from the CWP section in swamp deposits at the 10, 50,
300, and 10003th calculated steps.
Zn,
m
g
⋅
l
–1
4
8
12
16
20
Cell number
0
0.02
0.016
0.012
0.008
0.004
0.009
Cell number
Fig. 5. Change in plumbum concentration during waste filtration from the CWP section in swamp deposits at the 10,
50, 300, and 10003th calculated steps.
tration are characteristic of zinc, copper, and
lead: their contents decrease from 7.5, 0.7, and
0.11 mg
⋅
l
–1
, respectively, to a few hundredths
and thousandths of milligram per litre. Man3
ganese content slightly increases at first, then
decreases and is stabilized at the level of the
initial solution.
Section CWP
. Waste water in this section
is also drastically transformed interacting
with swamp deposits: pH decreases to neutral
value and heavy metals concentrations change.
It should be pointed out that the solution is
purified from some metals and is contaminat3
ed with others. For instance, in the initial stage
zone concentration increases almost 203fold
with respect to that in the initial solution
(Fig. 4). It decreases only after the filtered so3
lution has been changed many times as zinc is
10
50
300
1000
Zn,
m
g
⋅
l
–1
0.008
0.007
0.006
0.005
0.004
0.003
0.002
0.001
4
8
12
16
20
10
50
300
1000
71
Simulation of Heavy Metals Migration in Peat Deposits
discharged from the peat exchange complex.
Manganese behaves in a similar way. Cadmi3
um content slightly increases in the initial
stage, then it decreases and stabilizes at its
concentration in the slime storage. Lead behav3
iour deserves particular attention (Fig. 5). In
the initial stage near the slime storage, high
lead concentration zone is formed in solution.
Subsequently it moves slowly downstream and
its maximum concentration increases. This
increase is caused by changes in migration
forms: near the storage the pH values of water
are higher and lead is bound into carbonate
complex compounds to a considerable extent
but with increasing distance from the storage
pH decreases and lead fraction in the ionic form
increases. Hence, ion exchange processes take
place more actively and cause lead bonding in
the exchange complex. Consequently, a mobile
geochemical barrier is formed in which the
process of concentration of lead in solution in3
creases.
On the whole the simulation results are in
good agreement with data on testing peat and
water of swamp deposits in the ATPS region.
Conclusions
The results of computer simulation proved
that heavy metals in swamp deposits are main3
ly in the solved state: the ratio of sorbed to dis3
solved forms is 10
3
–10
4
.
It was established that in low3lying swamps
at pH @ 7 peat exhibits considerable buffer
capacity for both acidic and alkaline solutions.
In the initial migration stage, the concen3
tration of heavy metals in the bulk of peat lay3
er in solution virtually does not depend on their
initial content in waste water and is determined
by the composition of the peat exchange com3
plex. At low concentrations of compounds of
heavy metals in waste water, interaction with
peat can lead both to ion sorption and to their
accumulation in solution. If the initial metal
concentration exceed the first milligrams per
litre, the processes of self3purification occur
actively.
This work is supported by Russian Founda3
tion for Basic Researches, project codes 973053
64419 and 96305364338.
1.
Casagrande D.J.$and Erchull L.D. (1976) Metals in
Okefenokee peat3forming environments: relation to
constituents found in coal. Geochim. et Cosmochim.
Acta.
40, 387–393.
2.
Cheshire M.V., Berrow M.L., Goodman B.A. and
Mundie C.M. (1977) Metal distribution and nature
of some Cu, Mn, and V complexes in humic and ful3
vic acid fractions of soil organic matter. Geochim. et
Cosmochim. Acta.
41, 1131–1138.
3.
Szalay A. (1964) Cation exchange properties of hu3
mic acids and their importance in geochemical enrich3
ment of UO
2
++
and other cations. Geochim. et Cosmo1
chim. Acta.
28, 1605–1614.
4.
Construction Norms and Rules 2.01.28.95
(1985)
Landfills for Purification and Burial of Toxic Indus1
trial Waste.
Moscow: Goskomstroi. (in Russian).
5.
Livens F.R. (1991) Chemical reactions of metals with
humic material. Environm. Pollut. 70, 183–208.
6.
Varshal G.M., Velyukhanova T.K., Koshcheeva
I.Ya., Dorofeeva V.A., Bauchidze N.S., Kasimova
O.G. and Makharadze G.A. (1983) Study of chemical
forms of elements in surface waters. Zhur. analit.
khim
. 38, 1590–1599 (in Russian).
7.
MacCarthy P. and O’Cinneide S. (1974) Fulfic acid:
I. Partial fractionation. J. Soil Sci., 25, 420–428.
8.
Hedges J.I. (1977) The association of organic mole3
cules with clay minerals in aqueous solutions. Geo1
chim. et Cosmochim. Acta.
41, 1119–1123.
9.
Stevenson F.J. (1982) Humus Chemistry: Genesis,
Composition, Reactions.
Willey, New York, 473 p.
10. Shemiakin V.N., Potapov A.A., Abramov V.Ya. and
Kiriukhin V.A. (1989) Biogeochemical Method of
Search for Hydrothermal Sulphide1Containing De1
posits.
Authors’ certificate, No. 1515920,
16.06.1989 (in Russian).
11. Varshal G.M., Velyukhanova T.K., and Baranova
N.N. (1984) Stability of fulvic acids in natural wa3
ter under hydrothermal conditions. Geokhim. No. 2,
279–283 (in Russian).
12. Bloom P.R. and McBrige M.B. (1979) Metal ion bind3
ing and exchange with hydrogen ions in acid3washed
peat. Geochim. et Cosmochim. Acta. 43, 687–692.
13. Bunzl K., Schmidt W. and Sansoni B. (1976) Kinet3
ics of ion exchange in soil organic matter. IV. Ad3
sorption and desorption of Pb
2+
, Cu
2+
, Cd
2+
, Zn
2+
, and
References
72
I.L. Kharkhordin and F.G. Atroshchenko
Ca
2+
by peat. J. Soil Sci. 27, 32–41.
14. Parkhust D.L. (1995) User’s guide to PHREEQC:
A computer program for speciation, reaction3
path, advective transport, and inverse geochemi3
cal calculation. U.S.G.S. Water Res. Invs. Rept. 951
4227.
143 p.
15. Andre J.P. and Pijearovski L. (1977) Cation ex3
change properties of sphangnum peat: Exchange be3
tween two cations and protons. J. Soil Sci. 28,
573–584.
16. Bloomfield C. and Sanders J.R. (1977) The complex3
ing of copper humified organic matter from labora3
tory preparations, soils, and peat. J. Soil Sci. 28,
435–444.
17. Bunzl K. (1974) Kinetics of ion exchange in soil or3
ganic matter. II. Ion exchange during continuous
addition of Pb
2+
ions to humic acid and peat. J. Soil
Sci
. 25, 344–356.
18. Van Dijk H (1977) Cation binding of humic acids.
Geoderma.
5, 53–67.