Many of the explosives that occur as groundwater contaminants at Department of Defense (DoD) sites are nitroaromatic compounds (NACs) such as TNT or nitramines such as RDX. Under favorable conditions, NACs react rapidly with zero-valent iron (Fe ), which suggests that permeable reactive barriers containing zero-valent iron (FePRBs) might be useful in the remediation of groundwater contaminated with explosives. Unfortunately, reduction of NACs by iron metal produces aromatic amines as the primary products, and these products are still substances of regulatory concern. As a result, full-scale implementation of FePRBs to treat explosives-contaminated groundwater has been delayed until an effective treatment for the amines is developed and tested. The objective of this SERDP Exploratory Development (SEED) project was to develop an oxidative treatment step to treat the products of nitro reduction, so that the combination would form a sequential reactive treatment zone (SRTZ) that could be used to reach treatment goals for TNT-contaminated groundwater under field conditions.

The primary engineering goal of the project was to develop a sequence of reactive treatment zones consisting of two permeable reactive barriers—one to reduce nitro groups to amines, followed by another PRB or open trench to oxidize the amines to immobilized products. The first treatment zone was planned to be a conventional FePRB, and the second zone would have three key elements: a buffer zone, an oxidizing zone, and a method of oxidant delivery. The quartz sand buffer zone was intended to prevent excessive precipitation of iron oxides at the downgradient end of the FePBR. The oxidizing zone would be either an open trench or gravel-packed PBR, and the oxidant delivery method would range from sparging and direct injection to passive infiltration. Bench-scale columns first were prepared with 100% construction-grade, granular iron metal to generate simulated effluent from an FePRB.

Contrary to previous studies with FePRBs, many of the columns exhibited complete TNT degradation with no reaction products detected in the effluent. A range of follow-up studies was initiated to explain the differences between results of batch and column tests and to reassess the robustness of simple FePRBs for treatment of TNT-contaminated groundwater. It was found that break-through of TNT and its reduction products only occurs at unrealistically high flow velocities and/or low loadings of iron. Little evidence for TNT or its transformation products was found by eluting the columns or extracting the iron. Results of this study suggest that full-scale FePRBs (alone) may be an effective remediation tool for TNT and possibly other nitrated compounds.

This SEED project transitioned to a core SERDP project (ER-1232) to determine (1) whether the apparently complete removal of TNT and its reduction products really is complete and irreversible, (2) whether any field conditions will significantly alter the applicability or long-term performance of this technology, and (3) to what extent these results apply to RDX.

FePRBs are already a proven technology for in situ and passive remediation of several contaminants, including chlorinated solvents and chromate. If the follow-on project proves successful, it could expand the scope of application of FePRBs to groundwater contaminated with a variety of explosives, including TNT and possibly even RDX.