BOUZAHZAH et al. 2012. 9 th International Conference on Acid Rock Drainage (ICARD 2012) May 20 - 26, 2012, Ottawa, Ontario Page 1 of 12 Modification and automation of the humidity cell test protocol to favor tailings reactivity Hassan Bouzahzah 1 , Mostafa Benzaazoua 1,2 and Bruno Bussière 1 1 Université du Québec en Abitibi-Témiscamingue, Rouyn-Noranda, QC, CANANDA, Hassan.Bouzahzah@uqat.ca 2 INSA Lyon LGCIE, Villeurbanne, France, Abstract Previous laboratory work showed that the standard ASTM humidity cell (HC) kinetic test can be influenced by several factors that have significant impact on sulfide oxidation. This could lead to different leaching rates and, in some cases, to different acid mine drainage interpretations and predictions. In this study, modifications to the standard HC protocol are discussed. Mine tailings with a high acid-generating potential was submitted to both standard and modified ASTM HC kinetic test for a period of 20 weeks. Keeping samples permanently at a degree of saturation between 40% and 60 % was the main protocol modification, a range where tailings are the most reactive, as demonstrated in previous work. Results showed that under the modified ASTM protocol, sulfide oxidation rate was different by several orders of magnitude compared to the standard protocol. But, maintaining tailings degree of saturation in a range of 40-60% during the entire test requires daily operator intervention which is expensive and time consuming. This regulation can be automated with the use of a controller, a moisture sensor and a computer. The setup of the automated saturation control system and its operating mode is presented in this paper. The system can simultaneously control several humidity cells. Key Words: Acid mine drainage, prediction, ASTM humidity cell, Introduction Most of the active and closed mine sites are surrounded by solid wastes stored as waste rock piles and tailings impoundments which can contain significant proportions of sulfide minerals. When exposed to atmospheric conditions, these sulfidic materials may undergo oxidation producing acid mine drainage (AMD) which represents a major source of acidity and metals for the surrounding environment (Kleinmann et al., 1981). Static tests have often been used to assess, as a screening step, the acid- generating potential (AGP), but sometimes, they are unable to assess the long-term AGP (due to a large uncertainty zone, Ferguson and Morin, 1991). For these uncertain materials, kinetic tests are commonly used to provide a direct measurement of acid generation, neutralization rates, and long-term AGP. Column and humidity cell tests are the most commonly used kinetic test types to predict AGP and the lag time to AMD onset. They are based on sample alteration under laboratory controlled conditions simulating relatively intense weathering conditions (Bowell et al.,2006; Frostad et al. ,2002; Lapakko and White ,2000; Lawrence ,1990; Price and Kwong ,1997; Sapsford et al., 2009). Previous laboratory works had shown opposite results on AGP prediction for the same tailings depending on the type of kinetic test used. Benzaazoua et al., (2008) noticed a certain reactivity inhibition when tailings sample from Doyon mine (Québec) were submitted to the ASTM standard normalized humidity cells. These authors reported that the pH of leachate remained neutral over 364 days corresponding to 52 cycles (Figure 1) whereas Demers et al., (2008) noticed that the same sample submitted to column kinetic test became acidic after 380 days, which corresponds to only 10 cycles, as showed in Figure 1. Moreover, Sapsford et al., (2009) reported a difference in the calculated sulfate release rate (mg/kg/cycle) from a crushed waste rock submitted in duplicate to ASTM humidity cell tests. They concluded that the aeration system malfunctioned; as a result, one cell became much drier than the other during the test, depriving the waste rock of interstitial water required for pyrite oxidation. Bowell et al., (2006) had observed the same phenomenon when analyzing weathering of Volcanogenic