Complete report (PDF-file in German, 1.6 MB)

1. Introduction and objectives

The cooperative project of the Technical University of Berlin with the Environmental Research Center in Leipzig/Halle (UFZ), called "Heavy metal mobilization and migration in complex mining and smelting heaps by the example of the smelting location Helbra", developed an extensive database which should be qualitatively analyzed regarding the leaching behaviour of different waste dump components from the production of the Kupferschiefer and its smelting. The results should contribute to an overall view of material flows in these waste dump bodies.

For this purpose five lysimeter boxes with the components Kupferschiefer, Theisen-sludge, low-temperature carbonized Theisen-sludge, slag and host rocks (e.g. limestones and minor sandstones) were set up in a field test. This study conducted a three-dimensional process analysis. In a first step variations of element concentrations within these groups were examined. Thus the natural range of variation and inhomogeneity of the element concentrations in the raw material within the lysimeter boxes should be determined. In the second step the element concentrations in relation to a depth profile (top-center-bottom) in the lysimeter boxes have been examined. The results of the leaching experiment provided in the last step the possibility to regard the dynamics of the mobilization of the heavy metals over a time span. This work tried to determine a dynamic reaction profile in the end and to create thereby the hazard potential of pollutant discharges from the waste dumps. Since the quantity of the leachate was not measured during the test series, unfortunately, and climatic data is only incompletely recorded, a complete quantitative balance on heavy metal contents could not be executed.

2. Discussion of the results, summary and outlook

Discussion

In this study a detailed investigation of heavy metal mobility behaviour from waste dump material of typical smelting products of the Kupferschiefer has been taken place. From these investigations conclusions for long-term behaviour of components, an estimation of the endangerment and also the possible application as building materials (in particular the slag in road construction, and hydraulic engineering) had to be discussed.

Firstly, it was stated that variations in the leaching boxes represent generally mobilization processes. Some elements, such as sulphur in the Kupferschiefer; calcium, potassium, cobalt and manganese in the host rocks; manganese in the slag; cadmium, cobalt and lead in the low-temperature carbonized Theisen-sludge and subordinated cadmium, copper, lead and zinc in the Theisen-sludge, too, could represent partly variances by layering or an inhomogeneity by loading the boxes, since their variance is close to the natural variability of the samples. These were evaluated as not critical for the discussion of mobilization types.

In the second step the distribution of element concentrations in the lysimeter boxes has been regarded in comparison before and after the leaching experiments. Generalized mobilization types could be determined. The depletion type is the dominating distribution for almost all parameters in the Theisen-sludge and low-temperature carbonized Theisen-sludge. This could have been caused by the high reactivity of the samples due to their small grain-size and larger mobility of the heavy metals under oxidizing and sour conditions (sulphide oxidation). A substantial vertical displacement of heavy metals does not seem to have taken place in the Kupferschiefer and in the slag. Generally, thus the indifferent mixing type of the mobilization types for these components outweighs. In the case of the slag this could be due to its small reactivity (sulphide droplets are included in a glass matrix) and in the case of the Kupferschiefer due to its higher carbonate content, which buffers the release of acids from sulphide oxidation and provides partial adjustment of heavy metals. The host rocks, which possess a still higher carbonate content, show highest element concentrations at the bottom of the lysimeter box (enrichment-type). This could be explained partially by the buffering effect of the higher carbonate content. Blow-outs of Theisen-sludge particles could have been effective for a up to 7-fold enrichment in the case of arsenic.

In the third step hydro-geochemical investigations on the leachates have been executed. It applies to these that no water or heavy metal balance could have been executed, since the quantity of the leachates had not been documented. Thus, only relative investigations on the leachates have been executed by the author.

Firstly, a closer look at the temporal development of pH value and electrical conductivity has been taken. Because of the high sulphide concentrations, especially in the Theisen-sludge and low-temperature carbonized Theisen-sludge, it came - due to sulphide oxidations of the formerly under rather reducing conditions stored material - to the formation of a very low pH value (3-4), which constantly rose in the Theisen-sludge in the following leachates up to a value around 6, whereas the low-temperature carbonized Theisen-sludge remained hardly changed at pH values around 3. This shows particularly the very high reactivity of Theisen-sludge samples and low-temperature carbonized Theisen-sludge. Little by little the lumpy components are encased by an oxide coating and transformation of sulphide to sulphate is almost stopped.

This is more obvious in the samples Kupferschiefer, host rocks and slag. A pH value between 6 and 8 has been adjusted and rose only slightly - however constantly - during the investigations. This could be due to the low reactivity in the case of the slag, the Kupferschiefer and the host rocks could have buffered the acid release due to their higher carbonate content. The low pH value in the Theisen-sludge and low-temperature carbonized Theisen-sludge in interaction with oxidizing conditions, from which one can proceed because of the constant contact of the leachate with atmospheric oxygen, provide large mobilities for heavy metals.

Therefore a reversed picture results concerning the conductivity, too. Very high electrical conductivities showed all five samples in the first four leachates, however, particularly the Theisen-sludge and the low-temperature carbonized Theisen-sludge. Afterwards, a prominent decrease of the electrical conductivity in all leachates could be observed and thereupon a continuous slow decrease. Since the sulphate anion concentration is the dominating species in all samples, it has to be assumed that sulphide oxidation controls the electrical conductivity, at a smaller scale chlorides and fluorides in the Theisen-sludge, too.

This is acknowledged by good negative correlation of pH values with electrical conductivities especially in the low-temperature carbonized Theisen-sludge and in the Theisen-sludge itself. The Kupferschiefer and the host rocks follow with lower correlation coefficients. The slag shows only a low negative correlation and clarifies thereby the indifferent and almost inert behaviour. This could happen because of the fact that the sulphide droplets, which could set sulphate ions free, are inside a glass matrix and can only react with atmospheric agents at breaking edges.

Further investigations on correlative relations in between pH values, temperature, the amount of precipitation and reaction times of atmospheric agents resulted in best positive correlations of conductivity and temperature in the leachate out of the low-temperature carbonized Theisen-sludge. This could be explained by the fact that the lysimeter boxes are very small in size and changes of temperature can influence reaction flows from outside. I.e. that exotherm and endotherm reactions can be influenced by changes of temperature. This might rather be effective in the larger dumps, since their own microclimate might prevail in their core due to their size and compactness. This must be considered during the transfer of results from lysimeter experiments to dump scales.

The low-temperature carbonized Theisen-sludge has shown the best negative correlation of the amount of precipitation to the electrical conductivity, too. The time span between the samplings is generally not effective on the other parameters, since low correlation coefficients are pointed out. An exception represents the slag sample. This could be because of the low reactivity of the slag, whose effect could have been removed by the longer time span spent by the precipitate and the lower infiltration pressure by following percolation water.

The chapter about threshold values has dealt with the question, which temporal development the leaching process will take and how could the results be evaluated regarding threshold value lists (in this study the threshold value lists from Netherlands and Berlin). At the same time species conditions for each component in the leachate has been discussed. This was important for further investigations such as ion balances. Almost all components exceed the category C of the Netherlands or Berlin threshold value list for SO42- and NO3-. This is especially true for zinc and cadmium in the Theisen-sludge and low-temperature carbonized Theisen-sludge and for copper in the low-temperature carbonized Theisen-sludge even 100- to 10.000-fold, respectively.

These high concentrations are due to the high reactivity of the fine-grained samples, the low pH value resulting from it and the oxidizing conditions, which particularly provide a large mobility of the heavy metals from the Theisen-sludge and the low-temperature carbonized Theisen-sludge. The concentrations decrease constantly in almost every sample and have a local minimum in the fourth leachate. Only Ca2+ and SO42- in the host rocks and in the slag possess a local maximum there. The existence of a minimum can be explained with little precipitation, in addition, small temperatures (frost) in this time, which caused a strong decrease of the availability of the solvent (rain water). Generally, many reactions might run slower at lower temperatures, too.

Precipitations and incrustations could have occurred (see secondary minerals such as gypsum and anglesite), which provide a fixation of heavy metals. Their mobilization could hardly be renewed or there is a delay due to kinetic inhibitions. The formation of the local maximum for Ca2+ and SO42- in the fourth leachate could be explained by possible formation from the dissolution of gypsum crusts in host rocks and slag (Matheis and Jahn, 1996). Gypsum is stable only up to a pH value of 6.2. The rising pH value in the leachates could indicate a more effective dissolution of these gypsum crusts. Since the origin of the nitrate in this investigation cannot be derived from anthropogenetic loads (fertilizer, faeces), the source of the nitrate does not remain clarified.

An ion balance has been executed examining the quality of the chemical analysis. The Kupferschiefer component showed a very good cation/anion equilibrium, or the analysis was good, since it measured all parameters correctly. For the slag and the host rocks an anion deficit or a cation surplus has been determined, which has been probably caused by the alkalinity (HCO3-), which has not been measured. However, it can be assumed that this species is already led into the system by rain water. The Theisen-sludge and the low-temperature carbonized Theisen-sludge indicate a cation deficit. The reason could be that more anions were possibly needed for complexing reactions.

Since high ion concentrations are available and these influence each other mutually, ion activities were calculated, with which the following hierarchical cluster analyses were executed.

A simple hierarchical cluster analysis according to the weighted average value method (Davis, 1986) was executed in order to distinguish statistically homogeneous groups due to their relative similarity. Besides the activities of the species, temperature, the amount of precipitation, pH values and electrical conductivities were parameters, which were considered in the analysis. Firstly, it should be mentioned that the examination of the executed grouping by formation of cophenetic correlation coefficients and their comparison with the original correlation generally led to good results. Subsequently, a closer look at the heavy metals has been taken. In the Kupferschiefer the hierarchical cluster analysis resulted in grouping zinc, calcium and pH value together. This clarifies the influence, which the carbonate content in the Kupferschiefer sample has on the pH value and thus on the release of zinc. In the slag sample no heavy metals were leached and there is no genuine grouping recognizable, too. This reflects again the indifferent behaviour of the slag during these leaching experiments, which became already clear with the mobilization trends. The exceptional position of the parameter pH value has particularly to be mentioned in the cluster analysis of parameters in the slag leachates. The parameters in the Theisen-sludge sample show a high correlation among themselves. Environmentally relevant metals like cadmium, zinc, and nickel form a group with sulphate as an anion. This shows preferential occurrence with sulphate from sulphide oxidation in this sample. Copper forms a subgroup with manganese and lead is independent of the remaining heavy metals. This might be because of the fact that lead and at a smaller scale also copper are less soluble than zinc under these conditions. This clarifies that lead and zinc, which behave similarly in the primary dispersion as sulphides, can differ with respect to same Eh-/pH-conditions in the secondary dispersion. In contrast to the Kupferschiefer leachate, calcium, and thus basic properties of gypsum, does not have an influence. This is shown in the very small relative similarity to heavy metal clusters and to pH values, too. Something similar results from the leachate out of the low-temperature carbonized Theisen-sludge, in which calcium and lead form a cluster. Copper, however, forms the cadmium-zinc-nickel cluster. This higher mobility is probably due to significantly lower pH values in the low-temperature carbonized Theisen-sludge. The leachate from the host rocks does not indicate heavy metal contents. The calculated calcium-magnesium-sulphate cluster points out the dolomitic component in the host rocks, which essentially are dolomitic limestones as country rocks of the Kupferschiefer, and thus clarifies the effectiveness of this grouping method.

Holmstrom (1999) determined during investigations on sulphide bearing waste dump components of different types (with and without carbonate) that the carbonate-free component achieved a very low pH value with time and released high heavy metal concentrations. Contrasting to it the pH value remained high in the carbonate bearing component and prevented releases of heavy metals. The components in this investigation behaved similarly. However, the carbonate-free samples Theisen-sludge and the low-temperature carbonized Theisen-sludge achieved very low initial pH values due to their high reactivity and therefore a very high heavy metal release, which - as in the mentioned investigation - was reduced with time.

Summary and outlook

In order to execute a final estimation of endangerment, the parameters which influence the mobility of pollutants are summarized again. An attempt to discuss the hazard potential of the individual groups has been made. The factor permeability dominates the physical parameters, and has a crucial influence on the solution and transportation of pollutants. Compaction, swelling capacity of the deposited material and the precipitation rate control the availability of the solvent, water. This is of special importance for the extremely fine-grained samples of the Theisen-sludge and low-temperature carbonized Theisen-sludge.

This permits good wetting with water and a high chemical reactivity due to large surfaces. The formation of jointing in the slag dumps enlarges permeability, too. The slag, however, did not react with the atmospheric agents for the sulphide droplets are included in the glass. However, the distribution of Theisen-sludge suspensions in the dumps (pond IX and pond X) results in an endangerment with possible long-term effects of heavy metal release. The Kupferschiefer shows a good cleavage parallel to layering, achieved thereby a gradually enlarging surface and thereby could be an endangering potential due to the physical parameters. The host rocks do not represent physically caused endangerment due to their lumping. Although numerous reactions between solid and solution phase run relatively slowly for they are kinetically restrained, the parameter time was not effective besides in the slag sample. All other components showed up very reactive, particularly the Theisen-sludge and the low-temperature carbonized Theisen-sludge. The temperature could have had an influence due to the small size of the leaching boxes, like good correlation results in the low-temperature carbonized Theisen-sludge show. The remaining samples behave indifferently to the parameter temperature.

Chemical parameters control solution and precipitation reactions between the fixed phase and the solvent water. Numerous equilibrium reactions are affected by pH values and redox potentials, whereby these two parameters dominate leachates. They control chemical environments and thus the mobility as well as fixation of heavy metal cation- and anion-complexes. However, redox potentials were not measured, unfortunately. Probably it is possible to assume oxidizing conditions, since the leachates were constantly in contact with atmospheric oxygen. Adsorptive bonding or complex bonding can be very effective, additionally. Organic compounds can cause complexing reactions as well and thus lead to additional mobilization. The first chemical reactions form an initial Eh-pH-environment by the interaction of dissolution, precipitation and transportation. This could lead to a reduction of the mobilization of heavy metals, or an increase (Scheffer & Schachtschabel, 1982). The influence of the pH value is evident for the samples Theisen-sludge and low-temperature carbonized Theisen-sludge due to the sulphide oxidation. On the other hand, this investigation revealed that the carbonate content ensured especially in the host rocks and at a smaller scale in the Kupferschiefer higher pH values and an immobilization of heavy metals occurred (buffering effect). The slag seems to have behaved chemically inert except for the edges. Complexing reactions seem to be indicated in the organic-rich Theisen-sludge, whereby fluorides could have been important, too.

Biological parameters could not be considered in this study, but might have been effective, additionally, for most redox equilibria are kinetically restrained reactions (Scheffer & Schachtschabel, 1982) and would run in extremely long periods without catalysis.

Particularly favourable geochemical prerequisites (environmental conditions) for pollutant mobility, availability and thus the danger of a groundwater contamination with heavy metals within areas with acidic tendencies show - results presented in this study - the samples Theisen-sludge and low temperature carbonized Theisen-sludge. Their reclamation is highly necessary. For this reason the old Theisen-sludge dump location in the pond IX was mitigated. However, the Theisen-sludge dispersedly distributed in the jointed slag dump represents still an endangerment, which is pointed out by high heavy metal concentrations in the percolation water ("Stadtborn seepage"). At present these are cleaned from zinc in a neutralization plant. However, the produced neutralization mud has to be deposited separately.

Theisen-sludge from the old pond IX has been relocated in a waste dump of host rocks and low-grade Kupferschiefer-ore under a geo textile cover (pond X). In the case of a break-through of solutions, the release of higher heavy metal concentrations would be rather prevented by high carbonate contents of this dump. The Kupferschiefer follows the Theisen-sludge and the low-temperature carbonized Theisen-sludge concerning the endangerment priority, since heavy metal release is clearly lower and, on the other hand, it occurs mixed with low-grade Kupferschiefer-ore with dumps of host rocks, which do not represent a hazard potential. Slag, which represents volumetrically the largest dump component in Helbra, behaves chemically almost inertly and therefore it could be used as secondary raw material (road construction (paving-stones, crushed stone material)), as long as it is not interspersed by Theisen-sludge out of the old pond IX. Schreck (1997) mentioned a study which does not recommend the use of slag for dwellings due to its radio toxicity.

Further investigations are necessary, which should particularly deal with the questions for water balance, temperature, Eh and the seasonal climatic influence. Therefore, experiments should be continued during the whole year considering all available climatic data, in order to be able to meet more exact predictions to mobilization processes and hazard potential. Besides this, pollutant inputs by precipitation (wet deposition) and atmosphere (dry deposition) as well as the pH value of the precipitation should be included into a balance, too. Generally, the transmissibility of the leaching experiments on the dumps, its heterogeneities or filling of the lysimeter boxes with the sample material should be checked.

Thermodynamic considerations and saturation calculations could be employed on the basis of expected reactions and reactions indicated by secondary minerals. Moreover, pH, Eh and temperature dependence of the stability constants should be considered. It should be noticed particularly for the results of the presented correlation analysis that correlations rely on 7 values (7 leachates). A more reliable statistical result would be given by more values (and thus more leachates). This should be considered in a possible subsequent project, if a multivariate-statistical analysis of the leachates should be intended.