Reliability improvement for predicting acid-forming potential of rock samples using static tests.

Abstract

In predicting the acid-forming potential of rock samples, a combination of acid-base accounting (ABA) and net acid generation (NAG) tests has been commonly used. While simple and economical, this method sometimes shows low reliability such as categorizing certain samples as uncertain (UC). ABA and NAG tests were modified to selectively recover valid minerals in nature and substituted for the original tests. ABA test overestimated acid-producing capacity (in the case of weathered samples) and acid-neutralizing capacity (in the case of plagioclase-including samples) compared to the modified ABA test. NAG test yielded lower NAG pH compared to modified NAG test for samples with high total C content and low total S content. By comparing the correlation coefficients between acid generation amounts by the two evaluation methods, it was confirmed that modified evaluation method (MEM) has a much higher reliability (R 2=0.9582) than existing evaluation method (EEM) (R 2=0.5873). It was also concluded that exploiting advantages of both EEM and MEM is recommended where EEM is initially applied for general classification and a supplemented static test of MEM is executed for the purpose of correcting the error of UC categorized samples.