Paper No. 9
Presentation Time: 8:00 AM-12:00 PM
MINERALOGICAL CHARACTERIZATION OF LIMESTONE REACTED WITH ACID MINE DRAINAGE IN A PULSED LIMESTONE BED TREATMENT SYSTEM
Armoring of limestone is a common cause of failure in limestone-based systems for treatment of acid-mine drainage (AMD). Limestone is the least expensive material available for acid neutralization, but is not typically recommended for highly acidic, Fe-rich waters due to the armoring problem. To minimize armor formation and enhance limestone dissolution to create alkalinity, a new AMD treatment technology that uses CO2 in a pulsed limestone bed (PLB) reactor was developed by the USGS at the Leetown Science Center, Kearneysville, WV. The effects of the new technology are examined by characterization of limestone before and after treatment, with and without the PLB process, for AMD from an inactive coal mine at the Friendship Hill National Historical Site, PA. Previously, a constructed wetland at the site failed due to the elevated Fe (150 mg/L) and acidity (1,000 mg/L CaCO3) of the AMD. In experiments without the PLB process, limestone was almost completely armored with reddish-colored ochre within 48 hours of contact with AMD in a static, fluidized bed reactor. Effluent pH increased briefly from the inflow pH of 2.9 to over 7 but then decreased to <4 during 48 hours of contact. Limestone grains developed a rind of gypsum encapsulated by a 10- to 30-micrometer thick Fe-Al hydroxysulfate armor. Isolation of calcite from further reaction with AMD by gypsum and ochre armor prevented the limestone from generating any additional alkalinity in the system. With the PLB process, armoring was suppressed and most limestone grains completely dissolved resulting in an effluent pH of >6 after 3 months of operation. Limestone removed from the PLB pilot plant was a mixture of unarmored, rounded and etched limestone grains (59 %) and partially armored limestone (8 %) and refractory mineral grains (33 %). Residual reactor grains had thicker, more Al-rich armors and lacked the gypsum rind that developed without pulsing. Replenishment of limestone during reactor operation resulted in a mixture of partially dissolved new calcite grains and armored grains. Al-rich zones developed in the interior parts of armor rims in both the PLB and static bed samples in response to pH changes at the solid/solution interface. The PLB process appears to effectively circumvent the armor problem and allows greater utilization of limestone for AMD treatment.