地下水 铬污染修复

A lthough not as common as solvent or fuel products, chromium contamination of groundwater is relatively widespread. The U.S. EPA has estimated that as many as 1300 sites in the United States may have groundwater contaminated with chromium—some of which date to World War II. As with most groundwater contaminants, chromi- um contamination is traditionally remediated with an extraction and treatment system—“pump and treat”. Groundwater is extracted from the aquifer through a well network and conveyed to an aboveground treat- ment plant, where the chromium is removed using anion exchange or precipitated with various treatments. In many ways, chromium contamination has been an ideal candidate for pump and treat, because the most common chemical form, hexavalent chromium [Cr(VI)], is highly soluble and not readily adsorbed onto sediment surfaces. However, because of well- known limitations inherent to methods that require groundwater extraction, such as exponentially de- creasing response to treatment and diffusion-limited 464 A ■ ENVIRONMENTAL SCIENCE & TECHNOLOGY / DECEMBER 1, 2002 © 2002 American Chemical Society DECEMBER 1, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY ■ 465 A rates of extraction, pump-and-treat remediation of chromium-contaminated groundwater is not always satisfactory (1–3). Thus, alternative treatment meth- ods have been developed. Most are in situ methods that treat the impacted groundwater in the aquifer and eliminate the extraction step. This article pro- vides an overview of this technology. Chromium primer Chromium has three relatively stable valence states: Cr(0) (the metallic state), Cr(III), and Cr(VI). Other va- lence states are possible, but are not very stable. Cr(III) is relatively insoluble in water under common envi- ronmental conditions (pH 6–9), forming hydroxides and oxyhydroxides alone, and a solid solution with iron (4, 5). However, Cr(VI) is quite soluble and mo- bile in the environment. Chromium has various industrial uses, including chrome plating, steel making, corrosion inhibition, wood preservation, well drilling (as a fluid additive), biocides, and paint and primer pigments. Smelters that produce chromium can also be a source of con- In Situ Treatment of CHROMIUM- CONTAMINATED GROUNDWATER New technologies show promise for removing chromium(VI) pollution at lower cost. J O N A T H A N F R U C H T E R U. S. D EP AR TM EN T OF E N ER GY IN SE TS : P AC IF IC N OR TH W ES T N AT IO N AL L AB OR AT OR Y tamination, because the recovery of chromium from chromite ores requires oxidation to Cr(VI) before con- version to other forms. Cr(VI) is usually shipped commercially in the dichromate form (Cr2O7 2–), because the chromium is more concentrated. However, in dilute, near-neutral pH aqueous solution, Cr(VI) commonly forms hy- drochromate (HCrO4 –) and chromate (CrO4 2–) anions. These anions are generally poorly adsorbed on soils and sediments because they already have nu- merous uncharged or negatively charged surface sites at near-neutral pHs. Because they are sparingly sol- uble, barium chromate precipitates may form under some circumstances. However, the hydrochromate and chromate ions generally are not retarded but flow unimpeded in groundwater aquifers (6). Cr(VI) concentrations in groundwater in the United States are federally regulated under the Safe Drinking Water Act, with a maximum contaminant level of 0.100 mg/L as total chromium, and under the Clean Water Act with an ambient water quality crite- ria of 0.011 mg/L as chromium. Some U.S. states and other countries have implemented more stringent standards than the U.S. government for chromium in groundwater. There is continued uncertainty as to the appropriate drinking water standard for both hexavalent and total chromium. The in situ approach In situ remediation of chromium-contaminated groundwater involves chemically or biologically re- ducing Cr(VI) to Cr(III), which is less toxic, less solu- ble, and less mobile than Cr(VI). In addition, Cr(III) can then precipitate as a hydroxide, usually as a solid solution with ferric iron hydroxide, and will be effec- tively immobilized (7, 8). This reduction is usually a permanent solution, because Cr(III) is not easily re- oxidized to Cr(VI) under conditions that occur in most natural groundwater environments. Cr(III) oxides and hydroxides occur naturally in small concentrations in the sediment and soil in many groundwater aquifers (crustal average concentration = 102 ppm) (9). Although manganese oxides could re- oxidize Cr(III), Cr(VI) is rarely detected in these aquifers. In many instances, the chromium oxidation by manganese oxides in soils and sediments appears to be limited by surface alteration effects (10). Various approaches to reducing Cr(VI) in situ have been developed and tested. These methods usually in- volve adding some already reduced compound to act as a source of electrons. System selection and design The goal of any in situ aquifer treatment method is to deliver an appropriate reagent to the contamina- tion in the aquifer. Therefore, one way to classify the different treatment options is by reagent and deliv- ery system. The correct option is then a matter of choosing the right combination for the site condi- tions. To chemically reduce Cr(VI), reagents usually consist of reduced forms of three elements: carbon for most biological remediation systems and iron or sul- fur for abiotic approaches and certain specialized bi- ological systems. Delivery systems include trenches, infiltration galleries, groundwater wells, and direct- push injection. Choosing the correct approach begins with char- acterizing the site by geochemical, hydrological, and geological means, along with describing the nature and distribution of the contaminants. These charac- teristics are tied together with a site conceptual model. Frequently, a numerical model is used to make the site description more quantitative. Treatability studies in the lab and sometimes at the field scale also help facilitate the selection process. The following are examples of successful remediation systems. In situ treatment methods These methods involve abiotic approaches, usually involving reduced iron or sulfur compounds as the electron donor, and sometimes both. Permeable reactive barriers. Permeable reactive barriers, or treatment walls, treat groundwater as it flows away from the source and through the aquifer. The permeable barrier cuts off only the flow of con- taminants, but not the groundwater. In the trench- and-fill barrier configuration, a trench is excavated and filled with a chemically reactive medium. Care must be taken to ensure that the hydraulic conduc- tivity is equal to or higher than that of the surround- ing aquifer to allow groundwater to flow through the treatment zone. As the water flows through the zone, the chromium is reduced by the reactive medium and subsequently precipitates as a Cr(III) hydroxide. Because permeable reactive barriers are passive, their operation and maintenance costs are low. Another benefit is that the contaminants that require treatment and disposal are not brought to the sur- face. Barriers can be used even if the contaminant source has not been identified or well characterized, and the natural groundwater flow pattern is unaltered. However, this approach cannot treat contaminant sources and is not suitable for all geologic or hydro- logic regimes. For example, hydraulic conductivities 466 A ■ ENVIRONMENTAL SCIENCE & TECHNOLOGY / DECEMBER 1, 2002 FIGURE 1 Cross-sectional schematic of a trench- and-fill permeable reactive barrier The barrier, which uses metallic iron to reduce contaminants, has found some use in treating chromium. Fill Reactive cell Treated groundwater Aquitard Contaminant plume DECEMBER 1, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY ■ 467 A that are low may not permit sufficient groundwater flow to treat significant volumes of contaminated water. Groundwaters that are very high in dissolved constituents, such as calcium, may form precipitates that clog the barrier’s pores. Another limitation is that the barrier must eventually be replaced. If this hap- pens too often, the cost may be higher than alterna- tive approaches. Iron particle barriers. These barriers are the most common form of permeable systems. They rely on metallic iron [Fe(0)] for chemically reducing contam- inants (Figure 1). Although originally developed to treat chlorinated organic solvents, they have found some use in treating chromium and other metallic contaminants (11, 12). Particulate iron has the ad- vantage that each atom can donate up to three elec- trons, giving the barrier significant redox (reduction/ oxidation) capacity. When iron is present, Cr(III) can precipitate as a mixed iron–chromium hydroxide, which has a lower solubility than pure chromium hydroxide. This type of barrier is most frequently emplaced using trenching techniques, although other methods have been used. Often sheet piles are used to facilitate the installation of the barrier. Barriers emplaced by trenching have most commonly been restricted to depths of 10 m or less below the surface (13). Various methods for in- stallation at deeper depths have been investigated, in- cluding vibrating beams and jet grouting, but these become increasingly difficult at greater depths. In ad- dition, the high pHs that form in these barriers may lead to precipitation of various minerals, with subse- quent plugging, so Fe(0) barriers rarely use their full redox capacity (14). Nevertheless, iron particle barri- ers for Cr(VI) reduction appear to be operating suc- cessfully in the United States (15). Some difficulties with passivation of the iron particles in a Cr(VI) bar- rier have been reported by Danish scientists (16). Other permeable reactive barriers. Other forms of permeable reactive barriers have been used to treat chromium contamination. One particularly inexpen- sive alternative uses sawdust, compost, and limestone (17 ). This barrier is actually a cross between chemi- cal and biological methods, because the compost acts as a carbon source for microbial populations, which are active in the reduction of the chromium. Naturally occurring zeolite coated with cationic surfactants has also been investigated as an adsorbant for chromium in permeable reactive barriers (18). In situ redox manipulation (ISRM). ISRM tech- nology creates a permeable subsurface treatment zone to reduce mobile chromium in groundwater to an insoluble form. The permeable treatment zone is created by reducing Fe(III), which is present as sur- face oxides, to Fe(II) within the aquifer sediments (19). Some of the Fe(III) in 2:1 smectite clays is also reduced by injecting sodium dithionite (Na2S2O4) into the aquifer (Figure 2) (20). Sodium dithionite is a FIGURE 2 In situ redox manipulation (ISRM) process This technology creates a permeable subsurface treatment zone in aquifer sediments, where mobile chromium in groundwater is reduced to a less soluble and mobile form. RM-X are monitoring wells. Mobile field lab Injection solution RM-2 RM-5 RM-7 RM-6 RM-8 RM-1a RM-1b RM-4 RM-3 RM-9 Injection well Static viewer level Vadose zone Office/storage/trailer Low- permeability unit Permeable treatment zone Groundwater flow Contaminant plume from upgradient source High- permeability unit Low- permeability unit strong reducing agent that has several desirable char- acteristics for this type of application, including in- stability in the natural environment and reaction and degradation products that ultimately oxidize to sul- fate. This instability is beneficial because it means that the reaction period is rapid, and that after a period of several days, no dithionite remains in the aquifer. Potassium carbonate/bicarbonate is added to the injection solution as a pH buffer to enhance the stability of dithionite during the reduction of avail- able iron (by buffering H+ generated during iron reduction). As with permeable reactive barriers, an ISRM treat- ment zone is placed perpendicular to the groundwa- ter flow to intercept the contaminant plume. This geometry is created by a series of overlapping injec- tion/withdrawal wells. Advantages include the use of conventional groundwater wells, which leads to eas- ier installation at greater depths, and the ability of a single injection of dithionite to create a treatment zone that will last for many years. The technology is limited to clastic (silt, sand, and gravel) aquifers that have sufficient hydraulic conductivity to allow the reagent injection (>10–2 cm/s). The aquifer materials must also contain at least small amounts of reactive iron compounds (0.01–0.1%), although this is not usu- ally a concern. Most aquifers contain considerably more total iron than this, but only a fraction of it is reactive. The longevity of the barrier depends on sev- eral factors, including the amount of reactive iron, the concentration of oxygen and contaminant—both of which will reoxidize the barrier—in the ground- water, and the groundwater flow velocity. Chemically enhanced pump and treat. Also called geochemical fixation, chemically enhanced pump and treat adds a chemical reducing agent to the treat- ed groundwater before it is reinjected into the aquifer. In this way, the residual Cr(VI) that is not actually re- moved during the groundwater extraction phase can be treated in situ, alleviating some of the problems of conventional pump and treat (21). The reagent of choice, usually sodium metabisul- fite, Na2S2O5 , or calcium polysulfide, CaSx, is a func- tion of site geochemistry. Ferrous sulfate and sodium bisulfide have also been used. However, ferrous sul- fate injection can lower pH as a result of reactions that are similar to acid mine drainage, and sulfides can cause precipitation, which may clog the aquifer. Chemically enhanced pump and treat can be used to treat source areas. Another advantage is that treat- ed water does not need to be discharged to the sur- face. However, the approach needs a sufficiently permeable aquifer, removal of reduction reaction products if they become too concentrated, and reg- ulatory approval to reinject treated water. Electrochemical methods Electrochemical or electokinetic remediation places a series of electrodes into the contaminated zone, to which a low-voltage (50–150 V), direct current charge is applied (22). Contaminant ions in the water will migrate toward the electrode of opposite charge, which is called electromigration. Because hydrogen ions will migrate, the pH will decrease at the anode and increase at the cathode. For groundwater remediation, the electrodes can simply be placed in slotted nonmetallic wells, such as those made of polyvinyl chloride. The drift veloc- ities of the contaminant ions are relatively slow, around 1 cm per day, so that the electrokinetic method is not applicable to fast-moving groundwa- ters. The slow drift velocity also requires relatively close well spacing, another potential limitation. Attempts to increase the drift velocity using higher voltages can lead to problems with soil heating. However, it can be useful for treating unsaturated soils and slow-moving groundwater in tight aquifers in which the permeability is too low to permit other types of in situ remediation. Highly mobile anions such as chromium are good candidates for electroki- netic remediation, because they drift through the aquifer with little or no adsorption. A successful pilot- scale demonstration of the electrokinetics technolo- gy for vadose-zone Cr(VI) contamination has been conducted, but there have been no reported deploy- ments for groundwater chromium contamination to date (23). Biological in situ methods Reduction of Cr(VI) by living organisms either can occur inside the cell or can be mediated in solution by extracellular enzymes. It can involve direct reduc- tion of Cr(VI) or biological reduction of another metal species, such as iron, followed by abiotic reduction of chromium by the reduced metal. Microbial reduction. Microbial reduction of Cr(VI) has been known for over two decades, with early stud- ies showing that facultatively anaerobic Pseudomonas species are capable of catalyzing direct metabolic re- ductions of Cr(VI) to Cr(III) (24). Since that time, nu- merous investigators have shown that bacterial reduction of Cr(VI) is a widespread trait across sev- eral chemotrophic and phototrophic bacterial genera. In situ reduction of chromium by bacteria can be achieved by the introduction of nutrients (electron donors), microbes (bioaugmentation), or both. Nutrients may be sugars (e.g., molasses) or organic acids such as acetate, which can be used by many microorganisms (25), or lactate, which is metabolized by a restricted number of organisms. Injection of nu- 468 A ■ ENVIRONMENTAL SCIENCE & TECHNOLOGY / DECEMBER 1, 2002 PA CI FI C N OR TH W ES T N AT IO N AL L AB OR AT OR Y DECEMBER 1, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY ■ 469 A trients alone assumes that a suitable population of in- digenous metal-reducing or metal-accumulating or- ganisms exists at the site. Recent research has shown that such organisms occur widely, so bioaugmenta- tion should not be needed (26). Several species of bacteria, yeast, and algae cul- tured in the laboratory are capable of sequestering metals or changing the redox status of the aquifer so that the metals precipitate or are more easily adsorbed (27). The unicellular yeast Saccaromyces cerevisiae demonstrates the most favorable results for metal accumulators. Reduction of Cr(VI) to Cr(III) by microorganisms can be direct or indirect. Direct enzymatic reduction can be achieved by two types of bacteria: dissimila- tory metal-reducing bacteria that can use metals as electron acceptors for growth, or the fermentative and other anerobic metabolic groups that reduce met- als, especially relatively easy-to-reduce metals like Cr(VI), as a byproduct of their primary metabolic ac- tivity. An example of a dissimilatory metal-reducing bacterium is Shewanella oneidensis, strain MR-1 (28). A fermentative bacterium that has been shown to re- duce Cr(VI) is Enterobactor cloacae, strain HO1 (29). Cr(VI) reduction by mixed cultures enriched from soil samples has also been demonstrated in the labora- tory (26). Numerous other bacterial genera have also been shown to reduce Cr(VI). The indirect approach uses iron-reducing bacte- ria, such as Shewanella alga, strain BrY, to reduce iron oxides and iron-containing clay minerals in aquifer materials to the ferrous state (30). In this way, a re- ducing barrier of ferrous iron, similar to that described in the ISRM barrier section, can be created. Microbial remediation offers relatively low costs and only uses environmentally benign, carbon-based reducing agents rather than sulfur or iron. The ulti- mate result of carbon metabolism is usually CO2 as opposed to sulfates or ferric iron salts, which may cause secondary problems. However, there are four concerns: Nutrients must be injected periodically over the entire remediation period, which may be years; sufficient formation permeability is needed to allow injection of nutrients; achieving microbial growth where it is needed can be difficult; and unwanted for- mation plugging can occur as a result of excessive growth near injection points. Phytoremediation. Phytoremediation uses plants to remediate contaminated soil and groundwater through uptake, accumulation/sequestration, or bio- chemical degradation. All vascular plants take up met- als through their root systems, and some can accumulate and store large amounts. Laboratory stud- ies and small-scale field studies of the uptake/ sequestration of several metal contaminants in plants, including chromium, have been conducted. Phyto- remediation