Abstract: The aim of this study was to operate an upflow anaerobic packed bed reactor (UAPB) containing sulfate reducing bacteria (SRB) under acidic conditions similar to those found in acid mine drainage (AMD). The UAPB was filled with sand and operated under continuous flow at progressively lower pH and was shown to be capable of supporting sulfate reduction at pH values of 6.0, 5.0, 4.5, 4.0 and 3.5 in a synthetic medium containing 53.5 mmol l(-1) lactate. Sulfate reduction rates of 553-1052 mmol m(-3) d(-1) were obtained when the influent solution pH was progressively lowered from pH 6.0 to 4.0, under an optimal flow rate of 2.61 ml min(-1). When the influent pH was further lowered to pH 3.5, sulfate reduction was substantially reduced with only about 1% sulfate removed at a rate of 3.35 mmol m(-3) d(-1) after 20 days of operation. However, viable SRB were recovered from the column, indicating that the SRB population was capable of surviving and metabolizing at low levels even at pH 3.5 conditions for at least 20 days. The changes in conductivity in the SRB column did not always occur with changes in pH and redox potential, suggesting that conductivity measurements may be more sensitive to SRB activity and could be used as an additional tool for monitoring SRB activity. The bioreactor containing SRB was able to reduce sulfate and generate alkalinity even when challenged with influent as low as pH 3.5, indicating that such treatment systems have potential for bioremediating highly acidic, sulfate contaminated waste waters. (c) 2006 Elsevier Ltd. All rights reserved.

Abstract: Membrane replacement can be a major operating cost of a membrane plant. During the development of a novel desalination technique (the SPARRO process) for treating calcium sulphate scaling mine waters the expected life of tubular cellulose acetate membranes operating in the seeded-slurry mode was investigated.During four operating phases of the plant over a five-year period more than 9 000 h of operating data were obtained. Performance data showed that each operating phase was dominated by either membrane fouling or membrane hydrolysis. Membrane fouling was observed to begin near the front-end of the membrane stack and proceed towards the back. Hydrolysis, on the other hand, occurred first in the tail end of the stack and moved backwards towards the Front end modules. Although two detailed membrane autopsies were carried out no definitive statement can be made in respect of the causes of either membrane hydrolysis or membrane fouling. However, suggestions are presented to explain the observed fouling phenomenon in relation to the turbidity of the pretreated feed water and the presence of chlorine. It is proposed that the presence of radioactive isotopes in the mine water which become concentrated in the process contributes to the observed membrane hydrolysis. A membrane life of up to two years is projected for an improved pretreatment arrangement.