Add Add to Quick Collection

Please use this identifier to cite or link to this item:

OCR Text

The Southern Aican Institute of Mining and Metallurgy
Water in the Southern African Minerals Industry 2011

R Huberts


R Huberts
University of Johannesburg, South Africa

Metal sulphate containing acid mine water decantation is generally expected to be an ever
increasing challenge as mining activity winds down on the Witwatersrand. The water is
contaminated due to by-products of aerobic bacterial action, which catalyses the oxidation of
sulphide minerals in the presence of oxygen and water. During the ooding of an abandoned
mine, air is replaced with ground water, limiting the access of oxygen as this gas is only
sparingly soluble in water. This paper is a theoretical study on how the quality of decanting
mine water at a given location may be expected to vary in the longer term.

Key Words

acid mine drainage, AMD, acid mine decantation, iron oxidising bacteria, mine voids, central
rand basin, central basin, model

1 Introduction

The mine voids in the Central Rand Basin beneath the city of Johannesburg, formed due to
mining activity, are relentlessly lling up with water. Based on the West Rand Basin, where
decanting commenced in 2002, the water is expected to have typical Acid Mine Drainage
(AMD) characteristics (Carte Blanche, 2010). The decanting water will have to be treated to
remove dissolved metals and acid in order to protect other water resources and the

About 12- to 50MB of AMD decants from the West Rand Basin per day, which is estimated
to have a mine void volume of 45million m3 (Carte Blanche, 2010, miningweeklycom,
2010). For the Central Rand Basin, these gures are 32- to 70MB AMD (estimated) and 400
million m3 respectively (Carte Blanche, 2010).

The AMD on the West Rand is treated with lime as and when money and reagent is available
or left running into the surroundings (Carte Blanche, 2010). In the public domain, most
emphasis is understandably placed on the current crises, and no clear picture is presented to
the general public what will happen to the AMD over the next several years. The implication
is that huge amounts of lime will be required forever.

This paper investigates what could reasonably happen in the next several years by considering
the requirements of the bacteria that give rise to AMD, and the changes in their environment
due to the cessation of underground mining activity. The study focuses on cases where the
mine voids are totally ooded with water.

In the next section, a model is set up to predict the variation in AMD quality over a time
period of several decades. Based on this, treatment requirements are established over the

Page 159

The Southern African Institute 0f Mining and Metallurgy
Water in the Southern African Minerals Industry 20] 1

R Huberts

same time period. The results are then interpreted and suggestions for further work are made
at the end of the section and in the conclusions section.

2 Theory and Results

Bacteria such as T hiohacillusferrooxidans and Leptospirillumrrooxidans, present in water
films covering rock surfaces exposed by mining, catalyse the oxidation of pyrite. The
bacteria are proposed to do this indirectly by the oxidation of ferrous iron in solution in the
presence of dissolved oxygen (Verbaan and Huberts, 1988, Huberts, 1994), which diffuses
into the water lm from the air that is circulated inside the mine:

Fe-�-*+0.25o2+u1�1=e3*+0.51120 (1)

In this process, the bacteria obtain energy for living and multiplying. The ferric iron fonned
is an oxidising agent for sulphide minerals as follows using pyrite as an example (Huberts,
1994, Moses, Nordstrom, Hermana, and Mills, 1987):

Fesnitir�sngoaisrezwzsoqzvisut (2)

This re-generates ferrous ions that are oxidised by bacteria. The net reaction for (1) and (2) is
as follows:

Fes2+3.75o2+0.sn2o->Fe3*+2so42-+n* (3)
The supply of oxygen in Equation 3 is mainly ensured by ventilation of the mine.

Consider-a mine void of 106mg. In Johannesburg it contains air at 21% oxygen and a pressure
of 0.825atm, and one may use the ideal gas law to obtain the mass of oxygen in the mine


where Mr is the molecular mass of oxygen ( 32gmol'l),xgg is the mol fraction of Oxygen in
air (0.21 for 21%), P is the atmospheric pressure (84 000Pa), R is the universal gas constant
(8.3 l4JK"mol") and T is the temperature (298K for room temperature). This gives 228ton of

If, during the flooding of the mine, all the oxygen in the mine air is used to oxidise pyrite, an
iron concentration of at most 0.1gl.'land pH level of 3 is expected to result according to
Equation 3. This compares to a value of Ojgii-lwhich was measured in West Rand AMD,
and implies that most of the oxidation leading to AMD happens before mine closure
(assuming that ventilation ceases when operation stops). '

Once mine voids are flooded, the oxygen supply used by bacterial action (Equation 1) is
limited to the dissolved oxygen in the ground water penetrating the abandoned mine void.
The maximum (worst case scenario) concentration of dissolved oxygen in water at 25�C is the
value at saturation, 6mgffl (Wikipedia, 2010-11-05), which translates into only 6ton of
oxygen in the mine void ooded with saturated water compared to 228ton of oxygen of the
air~l1ed void.

Page 160

The Southern A ican Institute of Mining and Metallurgy
Water in the Southern Aican Minerals Industry 201 I

R Huberts

The oxygen availability has therefore been severely restricted on the ooding of the mine
void, and now an iron concentration of only 3mg�4and a pH level of 4 can be achieved
according to Equation 3. This compares with a maximum standard of 0.3mg�'1 and pH 5.5-
7.5 set for the purification of waste water or efuent (Government Gazette, 1984), so
treatment would still be required.

Over a period of time, the AMD in the mine void will therefore be replaced by a water
solution that can only contain a maximum of 3mg�4 of iron. If one considers a short period
of time dt, an iron mass balance (in mg) can be set up as follows:


V dC
_ -1 _ a ___
C - 3mg! lfm QdT  (S)

where C is the iron concentration in the AMD (mglfl), Q is the decantation flow-rate (E64), V
is the volume of the mine void (m3) and t is time (measured in days). Note that the iron
concentration in the AMD and the mine void are assumed to be equal 4 this would therefore
give a conservative (high) value of C. Assuming an initial AMD iron concentration of
500mgB4, integration of Equation 5 (and leaving out the units) gives

C = 3 +4912 �''*l$7'_ (6)

It can be seen that the concentration of iron in the AMD is a function of Q/V, the decant rate.
Using the values for Q and V reported in the introduction, the decant rate for the West Rand
Basin, in Edlms, varies between 0.3 and 1, while that for the Central Rand Basin is expected
to be between 0.1 and 0.2. The expected variation in iron concentration in the AMD with
time is given in Figure 1 for different decant rates.


500 p. . .   Rf
p12, Decent rate/
400 h fd*'m
E 30o -, """"'�"
\ 1 ' -.--.o.2
E 1 - ,
200 "+i\'"*:" ?-;% r "term "-0 3
1  __1
100 i
o so


Ti metyears

Figure 1. AMD iron concentration variation from a fully ooded mine

For a high decant rate, most of the iron is ushed out of the mine cavity in the rst ten years.
For lower decant rates, the process is dragged out over a lifetime period.

Page 161

The Southern Aican Institute of lllining and Metallurgy
Water in the Southern Aican lllinerats Industry 20] I
R Huberts

The iron and acid in the AMD that resulted from pyrite oxidation would need to be treated,
using lime for example. If it is assumed that the iron precipitates out as ferric hydroxide, any
iron left in the ferrous state would need to be oxidised in a pro-treatment step and the total
equation for neutralisation with lime is as follows:

2CHO+FBSQ+3 . 750241. 5H2O->FB(OI'I)3(S)+ZCEISO4(S) 

Note that in the left hand side of the equation the oxidation of pyrite using oxygen happens
rst, and the products are then neutralised by the CaO. The concentration of iron in the
AMD, determined by the decant rate, will indicate how much pyrite was oxidised, and the
rate of lime consumption can then be calculated using this and the ow-rate of the AMD in
conjunction with Equation 7.

100    of. _.__---- "KW.
% 8O
E '-.
0 1
0 20 40 60 80

Figure 2. Lime consumption rate for 0.2�d"'m'3 decant rate and 70M�d" AMD ow-rate

The consumption of lime will never cease as it is assumed that the ground water seeping into
the mine void is saturated with oxygen, and that all oxygen is used to leach pyrite. It can be
seen that the consumption of lime can potentially drop to very low levels (bottoming out at
about 0.5td"), even be it after a couple of decades.

One way of decreasing the consumption rate of lime faster to 0.5tdl is to neutralize more
timeously and closer to the source of AMD, right in the mine void. CaO as a neutralising
agent would probably be unsuitable, as this could cause the water to become basic in mine
void pockets where the ooding water is not that acidic. Another option would be ground
CaCOg (marble) which is cheaper and not very soluble in water (Wikepedia, 2010); therefore
proposed to be less likely to have a negative effect on the environment:

2C3CO3+F�S3+3.75O1+I.5HgO->F�(OH)3(5)+2CEISO4(S)+ZCOZ(E) ' 

According to Equation 8, one 25kg CaCO3 bag every 4.5m would suffice to neutralise
SOOmgEJAMD in a 2m diameter tunnel. Access to place the bags would be problematical
once mining activity has ceased, so it may be suggested that the bags be put down as soon as
(or even before) a section of the mine has been worked-out and is closed off, as part of the
normal mine operations. The neutralization of AMD in this way needs to be investigated

Page 162

The Southern A ican Institute of /ifining and Metallurgy
Water in the Southern African Minerals Industry 201 I

R Huberts

properly before taking the matter further, but if this is possible it would be an effective way of
paying dues to the environment.

3. Conclusions

The production of AMD that impacts on the environment seems to start before the sealing of
a mine. After the ooding of a mine, bacteria involved in AMD production are denied access
to atmospheric oxygen, leading to dramatically decreased oxidation of sulphide minerals
surrounding a mine void. As the only source of oxygen is contained in the water seeping into
a mine void, the concentration of iron in the AMD will decrease from an initial value (say
SOOmgEJ) to a maximum of about 3mg4 over a period of time, leading to an ever decreasing
AMD treatment operating cost.

On the other hand, once a disused mine is ooded, then draining the water again will re-
establish the atmospheric supply of oxygen and re-kindle the production of AMD, reversing
the trend.

A model was set up assuming that the effluent has the same composition as the solution in the
mine void. This could be tested and model parameters could be established using actual data
obtained from AMD. The nal treatment requirement will depend on the degree of oxygen
saturation of the incress.

Current measures for AMD treatment are largely reactive. A prevention strategy, for example
by allowing for the (future) in-situ neutralisation of AMD as part of current working mine
operations, could be investigated further.

4 Acknowledgement

The author acknowledges the receipt of AMD composition analysis obtained from Mr J.
Luvuno, from the University of Johannesburg, South Africa.

5 References

CARTE BLANCl-IEMNET. South Africa. 1 August 2010.

MININGWEEKLYCOM. Aurora says Grootvlei pumping subsidy increased,
for-aurorasgrootvlei-mine-ZO1 0-08-04

VERBAAN, B, AND HUBERTS, R. An electrochemical study of the bacterial leaching of
synthetic Ni3S2lnt. J. Of Mn. Proa, 24,1988.pp 185-202.

HUBERTS, R. Modelling of ferrous sulphate oxidation by iron oxidizing bacteria  a
chemiosmotic and electrochemical approach. University of the Witwatersrand
PhD dissertation, 1994

oxidation by dissolved oxygen and by ferric iron, Geochimica et
CosmochimicaActa, June 1987,Vol. SI, Issue 6. ppl561-l571.

httpzf/, accessed 2010-11-05

httpz/, accessed 2010-1 1-O3

Requirements for the purication of waste water or efuent, Government gazette, 18 May
1984, no. 9225.

Page 163

Open in Browser | Download 12 KB