English

Abstract

Manganese removal by limestone beds, as pioneered by Vail and colleagues (Pyrolusite Systems) can be effective for passively removing Mn from acid mine drainage. This paper reports rates of removal and sizing concepts based on data from eight Mn removal beds. A representative chemical reaction for Mn removal is Mn 2+ + 0.5O2 + H2O = MnO2 + 2H + (1) Laboratory studies indicate that Mn removal is catalyzed by bacteria and by Mn oxide surfaces, and that at constant pH and oxygen saturation, the rate of Mn removal from solution can be expressed as d(Mn)/dt = k (Mn)(MnO2) ( 2 ) where (Mn) is the concentration of Mn, (MnO2) is the surface area of Mn oxide (including the effect of bacteria), k is a rate constant and t is time. Integration of the resulting simplified equation gives the relation Log ((Mn)/(Mn0)) = -k1St/2.3 (3) where Mn0 is the influent Mn concentration (time 0), S is the surface area of Mn precipitation and k1 is a rate constant including effects of bacteria and surfaces. Based on equation (1), the rate may also be proportional to (O2) 0.5 and (H + ) -2 . The Mn vs. retention time data for six of the field sites closely fit a rate constant (k1) of 10 -3.35 hr -1 (ms 2 /mv 3 ) -1 where S (ms 2 /mv 3 ) is the surface area of limestone per cubic meter of bed. Retention time is estimated assuming 50% porosity. Two sites show faster Mn removal. An effect of pH may be present but is not large since the field data show that the limestone buffers pH to between 6.5 and 7.5. A dissolved O2 effect is undoubtedly present but cannot be resolved with the available data. The influent water should be well aerated and open to the atmosphere. Also, Fe and Al must be essentially lacking in the influent or the bed will plug with Fe or Al precipitate. The following equation can be used to size beds for manganese removal: A(m 2 ) = -0.276Q Log((Mn)/(Mno))/(k1SD) (4) where A is the bed area, Q is the flow rate (L/min) and D is the depth (thickness, m) of the bed.