Predicting the occurrence of acid rock drainage is important and challenging for the mining industry. However, there are weaknesses in using the existing predictive methods since the mineralogy and other affecting factors are site-specific. Acid-base accounting tests such as the standard Sobek method and modified Lawrence method are widely used to predict acid rock drainage because they are simple and cost-effective.

Various actual and composite samples were tested using the Sobek and Lawrence techniques in order to study the influence of sample properties (bulk elemental composition, and mineralogy) on their acid neutralisation potential, Samples include pyritic shale samples, rock samples, various mineral samples, pyrite-mineral blends, and pyrite-shale blends. The effect of acid strength and volume on the acid neutralisation potential was also considered.

It was found that the Neutralisation Potential (NP) was influenced by the moles of acid per gram sample, the test method and the sample properties. The samples tested exhibited a range of acid neutralisation potentials, where all the minerals tested except pyrite consume acid.

Among the pyrite-mineral blends, it was found that the NP of blends of pyrite with minerals serpentine and chlorite are higher in the Sobek test than in the Lawrence test. This may be due to the severe experimental conditions employed in the Sobek test, resulting in more dissolution of the acid neutralising minerals. The NP of pyrite-shale blends increased with increasing mass ratio of shale. The shales contribute to the neutralisation of the acid generated from the oxidation of pyrite in the blends.