Article pubs.acs.org/est
Evaluating the Development of Biocatalytic Technology for the Targeted Removal of Perchlorate from Drinking Water Justin M. Hutchison, Jeremy S. Guest, and Julie L. Zilles* Department of Civil and Environmental Engineering, University of Illinois at Urbana-Champaign, Urbana, Illinois 61801, United States S Supporting Information *
ABSTRACT: Removing micropollutants is challenging in part because of their toxicity at low concentrations. A biocatalytic approach could harness the high affinity of enzymes for their substrates to address this challenge. The potential of biocatalysis relative to mature (nonselective ion exchange, selective ion exchange, and whole-cell biological reduction) and emerging (catalysis) perchlorate-removal technologies was evaluated through a quantitative sustainable design framework, and research objectives were prioritized to advance economic and environmental sustainability. In its current undeveloped state, the biocatalytic technology was approximately 1 order of magnitude higher in cost and environmental impact than nonselective ion exchange. Biocatalyst production was highly correlated with cost and impact. Realistic improvement scenarios targeting biocatalyst yield, biocatalyst immobilization for reuse, and elimination of an electron shuttle could reduce total costs to $0.034 m−3 and global warming potential (GWP) to 0.051 kg CO2 eq m−3: roughly 6.5% of cost and 7.3% of GWP of the background from drinking water treatment and competitive with the best performing technology, selective ion exchange. With less stringent perchlorate regulatory limits, ion exchange technologies had increased cost and impact, in contrast to biocatalytic and catalytic technologies. Targeted advances in biocatalysis could provide affordable and sustainable treatment options to protect the public from micropollutants.
■
Massachusetts (6 and 2 μg L−1, respectively).9−11 The best available technologies for perchlorate currently include wholecell biological reduction and ion exchange.11,12 However, these technologies are hampered by the low concentration of perchlorate and the low ratios of perchlorate to cocontaminating compounds (often 1:1000 to 1:10 000), as well as the costs and environmental impacts associated with resin regeneration and/or disposal. Reverse osmosis was excluded from this analysis due to its high costs and impacts.13,14 Considering new technologies, palladium−rhenium catalysts have been proposed for perchlorate removal. 15 Biocatalytic treatment using perchlorate reductase and chlorite dismutase has also recently been demonstrated to selectively reduce perchlorate in groundwater samples.5,6 A fundamental understanding of the parameters governing sustainability trade-offs in biocatalytic processes is critical to prioritize research and development and to protect both public health and our environment. To this end, we integrated performance models with techno-economic analysis (TEA) and life cycle assessment (LCA) in a Monte Carlo framework to
INTRODUCTION A growing number of micropollutants are threatening drinking water supplies, with the problem exacerbated by the increasing water reuse required by growing populations and larger fluctuations in water availability caused by climate change.1,2 Biocatalytic technologies could harness the speed and selectivity of biological enzymes to remove micropollutants under environmental conditions typical of drinking water. Compared to its current uses, such as pharmaceutical production,3,4 the use of biocatalysts in drinking water or wastewater treatment represents a more challenging application: high volume, low value products, and, for micropollutants, low concentrations compared to cocontaminants (often >3 orders-of-magnitude difference). To our knowledge, only two biocatalytic applications have been investigated for water or wastewater treatment: perchlorate reduction for drinking water treatment5,6 and removal of phenolic compounds in wastewater treatment.7,8 The economics and environmental impacts of biocatalytic water treatment have not previously been evaluated; we do so here for the endocrine-disruptor perchlorate. Perchlorate causes irreversible effects on mental development at low concentrations, with the US Environmental Protection Agency (EPA) proposing drinking water standards of 15 μg L−1 and with stricter existing water regulations in California and © XXXX American Chemical Society
Received: Revised: Accepted: Published: A
February 14, 2017 May 3, 2017 May 12, 2017 May 12, 2017 DOI: 10.1021/acs.est.7b00831 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology Table 1. Performance Models for Perchlorate Removal Technologies performance equationa
technology
impacts considered
biocatalyst
rreactor*Δt = K m*(ln[So] − ln[S1]) + ([S0] − [S1])
substrate (eq S.8), temperature (eq S.9), inhibitors (S.2)
chemical catalyst
ln[S1] − ln[S0] = − rREACTOR_CAT × Δt
substrate (eq S.20), inhibitorsb
whole-cell
TED = (FEDoxygen )/(fe,oxygen ) + (FEDnitrate )/(fe,nitrate ) + (FEDperchlorate)/(fe ,perchlorate ) −
[RClO4 ][Cl−] [RCl][ClO−4 ]
non-selective IX
ClO4 − = K Cl
selective IX
[S0] − [S1] = mass of resin loading capacity
substratec, inhibitors (S.24)
substrate (eq S.25), inhibitors (eqs S.26−28) substrate (S.6)
Variable definitions: rreactor = maximum activity of the reactor, Δt = hydraulic retention time, Km = half saturation constant, S0 = perchlorate influent concentration, S1 = perchlorate regulatory limit, rreactor_cat = rate constant for chemical catalysts, TED = total electron donor, FED = fraction of electron donor, fe = fraction of electrons, K = equilibrium constant, RClO4 = perchlorate sorbed to resin, RCl = chloride sorbed to resin, ClO4− = bulk perchlorate concentration, Cl− = bulk chloride concentration. bThe impact of oxygen and nitrate were accounted for in the electron donor demand without modeling the mechanism for removal. cThe amount of electron donor was accounted for in the reduction of perchlorate but the kinetic effects of perchlorate concentration were not. a
parameters were from soluble protein fractions from the perchlorate-reducing bacterium Azospira oryzae strain PS (ATCC number BAA-33) and incorporated the effects of real-world groundwater components.6 Specifically, temperature effects were modeled using the Arrhenius equation based on previous experiments,6 and the effects of competition from nitrate were modeled as noncompetitive inhibition.5,6 Other groundwater components have negligible effects.6 Substrate inhibition and product inhibition are also negligible, as detailed in the SI, Section S.2.6,18 For chemical catalysts, the performance model used first order reaction rate kinetics to estimate catalyst dosing for the reactor.19 Oxygen and nitrate poison the chemical catalysts and must be removed.20 This was accounted for by increasing the amount of electron donor, without modeling the technology required to remove the oxygen and nitrate. The required pH adjustment before and after treatment was included in the costs and impacts. The catalyst was assumed to have a five year life, for consistency with a previous LCA21 and to account for expected improvements resulting from technology development, although to our knowledge this life span has not yet been demonstrated experimentally. The whole-cell bacterial reduction model was based on the stoichiometry for facultative perchlorate/nitrate reducing bacteria,5 with acetic acid as the electron donor. The electron donor demand included the reduction of preferred electron acceptors oxygen and nitrate,22 in addition to perchlorate, with corresponding cell synthesis requirements for all three electron acceptors. As with temperature, the effects of perchlorate concentration on kinetics were neglected due to insufficient data in the literature. Performance of the nonselective ion exchange systems was modeled according to Guter23,24 with published selectivity coefficients for perchlorate, nitrate, bicarbonate and sulfate to chloride (Table S.1). The model assumed plug flow and instantaneous equilibrium of sorbed and aqueous phases25 and calculated the number of treatable bed volumes of drinking water. Because nitrate desorption from the resin could result in effluent nitrate concentrations greater than influent, the model terminated treatment and initiated regeneration as needed to prevent violation of regulatory limits (44 mg NO3− L−1 or perchlorate).26 The computational intensity of the nonselective
elucidate the economic and environmental implications of biocatalysis for drinking water treatment, using perchlorate as a model contaminant. Costs and impacts were compared to four alternative technologies at different stages of development and implementation, and results were leveraged to identify technology development pathways for biocatalytic processes.
■
METHODS Water Characteristics Distribution. Data for anion concentrations were obtained from the United States Geological Survey (USGS) based on well samplings from ten states.16 Empirical distributions were created based on the five anions: perchlorate, nitrate, sulfate, bicarbonate, and chloride from wells with detectable perchlorate. Wells with more than 44 mg L−1 nitrate (