A Review of Bioremediation andNatural Attenuation of MTBELawrence C. Davis and Larry E. EricksonDepartments of Biochemistry and Chemical Engineering, Kansas State University, Manhattan, KS 66506; [email protected] (forcorrespondence)Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/ep.10028Methyl tert-butyl ether (MTBE) has been the focus ofmuch attention because it is used in large amountsand was reportedly relatively recalcitrant to bioreme-diation or natural attenuation. Beginning with a fewpapers a decade ago, evidence has been presented that,in fact, under suitable conditions it is amenable tobioremediation. Many species from widely disparatemicrobial genera are able to consume it either as a solecarbon source or as a cometabolite. Optimal conditionsdiffer from site to site. Both aerobic and anaerobicconditions may permit MTBE degradation, with arange of electron acceptors, from oxygen throughFe(III), Mn(IV), sulfate, nitrate, and methanogenesis.MTBE metabolism in the vadose zone can be highlyactive. The published literature suggests that naturalpopulations are adapting to MTBE, and reported ratesof biodegradation appear to be larger in the morerecent literature. Plants may serve as efficient conduitsto withdraw MTBE from the wet subsurface, releasing itto the atmosphere or the vadose zone, where it may bemetabolized or diffuse into the atmosphere where it isquickly photodegraded. The major remaining issuesare the time required to attain specified criteria ofcleanup or whether augmentation is necessary for ef-fective remediation. © 2004 American Institute of ChemicalEngineers Environ Prog, 23: 243–252, 2004Keywords: MTBE, bioremediation, phytoremedia-tionINTRODUCTIONMTBE, more formally methyl tert-butyl ether, hasbeen extensively used as an octane booster and oxy-genate additive in reformulated gasoline in the UnitedStates. With MTBE being distributed nationwide, thepotential for contamination of soil and groundwater issubstantial. Efforts to develop strategies for cleanuphave produced a large number of publications, a selec-tion of which are discussed below [1– 60]. MTBE hascertain advantages over ethanol in the behavior ofgasoline blends to which it is added, and it is efficientlyproduced from iso-butylene. Current production (2003)of MTBE is over 200,000 barrels per day (⬃8 ⫻ 106L)in the United States and U.S. usage is about 1/3 greater[12].Physical and organoleptic properties of MTBE makeit a challenge to deal with effectively. Unlike mostgasoline constituents, MTBE is highly soluble in water.At room temperature its solubility is about 50 g/L, 20times greater than that of BTEX, the most soluble gas-oline constituents (benzene, toluene, xylenes, andethyl benzene). As discussed by Kinner [30] it alsopartitions strongly from air to water. The dimensionlessHenry’s Law constant is in the range of 0.01 to 0.04,depending on temperature [10], whereas that of ben-zene is about 0.2. Thus MTBE is more likely to dissolvein water and less likely to volatilize from gasoline orwater to air than is the case for other gasoline constit-uents.Some of the important physical and chemical prop-erties of MTBE and BTEX compounds are shown inTables 1 and 2. Because of its small Henry’s constant,MTBE is expensive to remove from water by aeration.Removal using adsorption is expensive because thequantity that adsorbs to activated carbon is relativelysmall based on the small values of the octanol/waterpartition coefficient and sorption coefficient. Keller etal. [27] evaluated four physicochemical treatment tech-nologies and reported estimated costs associated witheach technology. When no air treatment is required, airstripping is the lowest-cost technology for high flowrates, whereas hollow-fiber membranes are cost effec-tive for low flow rates. Granular activated carbon ismost cost effective if air treatment costs are included.Advanced oxidation processes are the most expensiveof the four options.Although health effects of low-level MTBE contam-© 2004 American Institute of Chemical EngineersEnvironmental Progress (Vol.23, No.3) October 2004 243ination are uncertain and disputed (see [13]), it has apotent taste impact in water at levels of 10 –30 ␮g/L(ppb). Thus even small spills of MTBE are detectable,and the USEPA [49] issued a health advisory recom-mending that drinking water levels be kept below20–40 ␮g/L. This would provide a very large safetymargin compared to known biological effects. Manystates have implemented regulations on MTBE contam-ination of water and a number have banned its use ingasoline. Through the year 2000, 38 states had actionlevels, cleanup levels, or drinking water standards forMTBE [37].The drinking water levels varied but were all below250 ␮g/L. As of March 2003, restrictions or outrightbans were pending in 16 states. Five states proposingbans or severe restriction depend on MTBE for oxy-genates and account for 45% of MTBE use nationwide[12]. Thus if implemented, these restrictions couldmarkedly reduce new incidents of MTBE contamina-tion of groundwater.One liter of MTBE/ha could contaminate all theyearly rainfall on that area to a level of about 100 ppb(assuming 1 m/yr rain). Avoiding such contaminationin areas where there is a lot of reformulated gasolineuse has proven challenging. When MTBE constitutes10–15% of the gasoline, containment systems have tomaintain nearly perfect efficacy to avoid some contam-ination of groundwater. This has been a particular issuein California, which consumes a large fraction of allU.S. MTBE, although the expected magnitude of theproblem is debated [26, 33].Given the widespread nature of MTBE contamina-tion and the intense efforts to remediate it, a largenumber of studies have been published that detail anextensive amount of research. Prince [40] provided anexcellent critical summary of what was known aboutmicrobial degradation of MTBE. More recently, Fayolleet al. [14], Fiorenza et al. [16], and Seagren and Becker[45] provided reviews that describe some of the addi-tional progress that has occurred. Fayolle et al. [14]examined the broader issues of remediating other ox-ygenates such as tert-amyl methyl ether and t-butylalcohol. Fiorenza et al. [16] considered the distributionof contamination problems, the nature of plumes, andexamples of natural attenuation. Seagren and Beckerreviewed the natural attenuation of BTEX and MTBE[45]. Here we focus on a limited selection of