Supervising Scientist Report 121
Klessa DA, Koehnken L and Johnston JF
Supervising Scientist, 1997
ISBN 0 642 24321 2
- SSR121 - Estimation of water quality over time within the Queen and King Rivers. Mount Lyell Remediation Research and Demonstration Program (PDF - 4.5 MB)
- Preliminary pages (PDF - 611 KB)
- Chapters 1-4 (PDF - 942 KB)
- Chapters 5-7 (PDF - 1.7 MB)
- References and Appendices (PDF - 504 KB)
A combined chemical-speciation and hydrological model has been developed to predict down stream changes in water chemistry in the Queen and King Rivers in response to possible remediative measures developed as part of the Mount Lyell Remediation Research and Demonstration Program (MLRRDP). A century of mining at the Mount Lyell copper mine in Queenstown, Tasmania, has resulted in massive volumes of tailings, smelter slag and acid drainage entering the Queen River, a tributary of the King River, and ultimately Macquarie Harbour. Acid drainage continues to enter the river system from the pumping of water from the active underground mine workings, and the draining of water from inactive mine workings and waste rock dumps. The aim of the MLRRDP, a cooperative program involving the Tasmanian and Commonwealth governments, is to develop remediation plans for the lease site, rivers and harbour which, if implemented, will begin to reverse the environmental impacts of historic mining practices, and promote the recovery of aquatic ecosystems.
The combined chemical-speciation and hydrological model was developed in order to fulfil the following objectives: predict changes in Queen and King River water chemistry in response to further development of the Mount Lyell mine or remediative measures to counter pollution from the Mount Lyell mine lease site; provide information which will assist in the development of a strategy for the biological recovery of the King and Queen Rivers based upon remediative measures at the mine lease site; supply boundary values and concentrations for pH and metals in the King River to permit the development of scenarios for the hydrodynamic modelling of Macquarie Harbour; and, to compare biological response under test conditions with speciation predictions to help elucidate the principal factors determining ecotoxicological reaction.
The model consists of two parts: the first deals with dilution of mine effluent by unimpacted river water and the second with speciation predictions based upon the composition of the mix and its chemical equilibria using MINTEQ2A (ver 3.11). The mixing model was based on a single source of acid drainage and two unimpacted diluents, namely the Queen and King River. Unimpacted waters in both these rivers typically have a very low suspended solids content and alkalinity. The composition of acid drainage was taken as the median concentration of constituents from a combined Conveyor Tunnel and North Lyell Tunnel source representing about 78% of the Cu load from the Mount Lyell lease. The diluents were also based on median concentrations, albeit from a very limited data set. Mixing of acid drainage with river water was modelled using either a geometric dilution series for the Queen River and King River with the power station off, or in the case of the power station on, Lake Burbury providing 80 cumecs and median flow from other sources.
Because results from MLRRDP investigations indicate that the major effort in countering the impact of acid drainage will have to be liming so long as mine effluent has direct passage to local creeks and rivers, the scenarios which were modelled were equivalent to the proportion of total acid drainage which might be neutralised under different remediation strategies. The four most likely remediation scenarios which were modelled were: 8% neutralisation of acid drainage (current situation); 65% neutralisation of acid drainage corresponding to the neutralisation of Conveyor Tunnel discharges; 80% neutralisation of acid drainage corresponding to the treatment of the North Lyell Tunnel discharges in addition to the Conveyor Tunnel; and 99% neutralisation of acid drainage sources which also includes the discharges from the West Lyell waste rock dumps. Neutralisation endpoints of pH 6.5 and pH 5.5 were modelled for each of the above described scenarios, as was the inclusion/exclusion of a copper-recovering Solvent Extraction/Electrowinning (SX/EW) plant prior to neutralisation.
In the model, the neutralised acid drainage was re-mixed with the remaining acid drainage sources and 'released' downstream. It was assumed that the precipitation products formed from the neutralisation reaction were not discharged into the Queen River. A conservative approach was taken for the downstream speciation modelling in that the sorption of metals by either naturally occurring suspended matter or precipitates, which were predicted to form from components of the acid drainage, was not invoked. The concentration of natural suspended solids in Lake Burbury water, the principal source of the King River, is exceedingly low (typically 1-2 ntu), and are therefore unlikely to influence solution phase chemistry to any marked degree.
The model results have extremely important implications for the development of remediation strategies and the recovery of the Queen and King Rivers. One of the most important findings is that there is little scope for recovery of the King River based on the predicted bioavailability and toxicity of copper unless almost all (>99%) of the acid drainage currently entering Haulage Creek is neutralised to pH 6.5. Additionally, the use of SX/EW technology in conjunction with neutralisation of acid drainage to pH 5.5 confers less advantage in terms of potential copper toxicity to aquatic life compared to neutralisation to pH 6.5. Hence, if SX/EW is to proceed, no advantage will be realised by the downstream ecosystem unless all raffinate is neutralised to pH 6.5. The model results also indicate that even if this very high target of 99% treatment of acid drainage can be achieved, the ecological recovery of the Queen River may continue to be hampered due to toxicological impacts associated with copper.
Aluminium, another metal of concern for the recovery of the river system, was more difficult to evaluate due to a high degree of uncertainty about the relative toxicities of Al species to aquatic organisms, and because of the paucity of available data pertaining to background concentrations in the King River catchment. However, model results suggest that aluminium concentrations may be reduced to tolerable concentrations if somewhere between 80% and 99% of the acid drainage is neutralised to a pH of 5.5, conditions under which copper would continue to remain the limiting factor.
The toxicological impacts of fluoride, manganese and iron were also examined with the conclusion that fluoride is present in sufficiently high concentration in acid drainage to pose constraints to biological recovery and liming to pH 6.5 is required to promote the precipitation of fluorite during neutralisation. Manganese and iron were found to probably be far less of a constraint to biological recovery in the Queen and King Rivers than either copper or aluminium.
The speciation predictions from the model could be refined through the collection of additional geochemical input data, especially with respect to background concentrations of metals in the King River catchment, and DOC concentrations.
Validation of model predictions was outside the scope of this project. Clearly, however, there is a need to test the predictions provided under the modelled scenarios by laboratory-based mixing experiments and, if necessary, to adjust the model to take account of kinetics and chemical processes such sorption and precipitation. In particular, the importance of a suspended Fe oxy-hydroxide phase in river water derived from the mixing of acid and neutralised drainage with river water requires to be elucidated with respect to its role to adsorb, and possibly coprecipitate Cu.