by Jack Caldwell
Can you characterize acid rock drainage? The problem is so common in mining: spatial variability and data quality. You know the author is onto something when you see a description of kriging and other statistical techniques . I wonder that anybody has the gall to present their data without a statistical evaluation of its accuracy in the presence of the brilliance that has been brought to bear on the issue.
And yet they do, as is amply demonstrated by the other papers on this topic. See the specifics of predictive attempts from mines in the Canadian Cordillera , Leadville, Colorado [507 and 738], Great Falls, Montana , Thompson, Manitoba , and Cerro de Pasco, Peru .
You may be interested in the characterization of formation and process as detailed in papers on antimony leaching from stibnite , sphalerite decomposition , pyrite oxidation , iron sulfides , goethite , and the emanation of gas from piles . I give no further details as you obviously know your ARD if you seek out these papers.
My favorite paper  on the topic is from Noranda and Quebec where the solution is feasible: keep the acid generating materials below the water table and hence free of oxygen and the pernicious tendency to generate waters of the wrong pH.
Currently the most intensive and rigorous prediction program to quantify the long-term development of acid drainage from waste rock dumps is that being undertaken for the Diavik Diamond Mine in the far north of Canada. A significant paper  describes construction and study of two large-scale waste rock piles to provide information about the performance of sulfide bearing rock in artic conditions.
The objectives of the study are to quantify the release of acidity, metals, and other solutes from sulfide-bearing mine waste piles in the Arctic environment, to develop conceptual models of water flow in unsaturated piles, understand the different physical mechanism leading to cooling of large stockpiles in cold climates, including advective flow, quantify and assess the value of small-scale measurement in predicting larges-scale, long-term processes, and undertake rigorous testing of models of physio-chemical processes in such dumps. It is hoped the work will lead to development of protocols for design, management, closure, and regulation of such facilities.
The Australians report  advances in the concurrent use of acid base accounting (ABA) methods, including consideration of net acid production potential (NAPP) and net acid generation (NAG) to predict acid generation and augment conventional testing. Other papers in this session report on the use of paste to predict acid generation potential , comparison of predicted and measured neutralization potential of intrusive rocks , simulation of the neutralization capacity of silicate rocks , geochemical characterization of Cluff Lake, Canada mine waste rock for use in prediction of acid generation , and scale-up of mass loading rates in arid artic environments .
Modern computers and modeling techniques enable us to model any process we choose. The only issues are time, money, realistic and representative data, and a case history for validation. Witness the modeling of the world climate and the onset of global warming. I have read of the many world-climate models that sometimes agree and often disagree. Sadly most of my readers will not be around in a hundred years to see which model is correct. A similar situation prevails in modeling acid drainage: it will take hundreds of years to decide if a specific model is truly correct. This does not deter us, however, and here is an overview of papers on modeling acid mine drainage.
Modeling of the gold slimes dams in the South African Witwatersrand using PYROX, PHREEQC, and spreadsheets establish that buffering will continue for thousands of years and that it may take between 1,000 and 10,000 years for the piles to generate acid drainage. This fascinating paper  says that the results can be used to evaluate mitigation measures, but unfortunately does not say what these mitigation measures are. I cannot myself conceive of implementing 10,000 years of mitigation in those areas where I grew up and saw the growth and performance of gold slimes dams as we called them and I still do. I know what it cost (about one billion dollars) to deal with uranium mill tailings at twenty-four piles in the United States; could this be replicated on the Witwatersrand-does it need to be replicated?
HYDROGEOCHEM was used to model uranium mine waste rock dumps at Pocos de Caldas, Brazil. Modeling showed that acid drainage will continue for at least 600 years. These results will be used to decide on closure works . Maybe they will have to spend comparable sums to those spent in the US on uranium mine closure and stabilization. But as always it is easier and cheaper to model than to act.
At the Svartliden Gold Mine in Sweden, the fourteen percent of potentially acid generating waste rock will be encapsulated into benign rock. Modeling with SULFIDOX showed that encapsulation will limit acid seepage. I do not attempt here to detail the processes incorporated into the model and described in the paper; enough that is seems to be a pretty comprehensive way to model waste rock dumps .
POSSUM is an object-oriented program that models the evolution of mine water quality, accounting for contaminant sources and sinks, water level change (due to seasonal fluctuations), reaction kinetics, variable flow regimes, infiltration through porous media, and inflow from secondary roadways. The case history is an underground coal mine in the UK. POSSUM solves for concentrations in both solid and aqueous phases using sequential finite difference methods . The paper does not tell us how long the process will continue or what is to be done to control them.
HYDRUS and POLYMIN were used to model hypothetical waste rock dumps and their moisture conditions and acid generation . In Quebec this leads to steady state conditions and acid generation that may then be addressed by unspecified measures.
Geochemical modeling of the Antamina waste dumps included collection and collation of field and laboratory data, and water quality modeling to simulate short and long-term trends. Data collection will continue and model results refined and calibrated . Similarly at the Tyrone Mine in New Mexico, data collection and use of Acid Base Accounting (ABA), Acid Producing Potential (APP), and Acid Neutralizing Potential (ANP) were used to model waste rock piles and predict their future performance . Specifically, SWMS-2D flow modeling showed that 15% of rainfall infiltrates the pile, 95% of that flows through the matrix, and less than 5% through the macro-pore channels. The long residence time, 47 years, in the matrix results in different seepage chemistry than from the short residence time, less than 2.5 days, in the macro-pores.
| Prev | Next | Back To Top |