Abstract

The quality of storm-water discharge is a growing concern in the United States owing to a steady increase of harmful pollutants entering freshwater. Many congressional mandates that require local governments to reduce the impact of storm-water discharge on the environment have greatly increased the need for economically and environmentally viable solutions to pollution reduction. One such solution is that of constructed wetlands. Wetlands naturally function as a water-purifying system that removes contaminants such as organic material, suspended solids, pathogens, nutrients, and heavy metals and that lowers biological oxygen demand (BOD) levels (Pastor et al. 2003). Storm water and wastewater have conventionally been treated in off-site locations by different physical, chemical, and biological systems; however, for larger industrial units, it has become more cost-effective and environmentally beneficial to treat storm water on-site with constructed wetlands (Environmental Protection Agency [EPA] 2009). The use of constructed wetlands for large industrial facilities allows for more sustainable means of removing harmful pollutants on-site before they enter the local water source rather than sending the water to the regional water treatment facility (Campbell and Ogden 1999). The growth in consumption of forest products for biofuels has led to increased waste by-products generated from the bio-oil manufacturing processes. This waste is often found in the storm-water runoff from these facilities and can have harmful effects on the environment if not treated properly. A rapid development of widely applicable, low-cost methods for treatment of this type of wastewater is a high priority because several bio-oil refineries will start operating throughout the United States in the next few years (KiOR 2013). An efficient way to deal with the resulting wastewater produced needs to be examined.

Since the passing of the Clean Water Act in 1972, the country has made great efforts in reducing the amount of pollution released into surface waters. For this study, the contaminants that were treated in the constructed wetlands originated from the thermochemical processing of biomass. Hydrothermal processing occurs when wet biomass undergoes thermal treatment at high pressures to produce liquid hydrocarbons (Mohan et al. 2006). The source of the wastewater for this study comes from hydrothermal processing of biomass.

Industrial biomass conversion sites contain a number of contaminants that are found in the storm-water runoff. These include aromatic hydrocarbons, such as benzene, toluene, ethylbenzene, and xylene, as well as other residual chemicals derived from petroleum. The other pollutants found in the contaminated wastewater derived from the bio-oil process include aldehydes, ketones, carboxylic acids, phenols, and furans (Klass 1998, Basu 2010, Caldeira-Pires et al. 2013).

There are many ways of dealing with contaminants that are often found in water, such as filtration, absorption through plant uptake, volatilization, and the microbial activity within a constructed wetland (Baker and Herson 1994). Some of the dominating organisms that are often found in a constructed wetland system are those that can break down hydrocarbons, aldehydes, ketones, carbolic acids, and phenols (Vymazal 2011, Weaver et al. 2012, Zhao et al. 2012). These include but are not limited to Rhodococcus aetherivorans, Archaeoglobus, Borrelia burgdorferi, and Achromobacter (Nelson and Wolverton 2011, Ansola et al. 2014).

Objectives

This study proposed to amend the waste-contaminated storm-water runoff from biomass conversion to a bio-oil facility through a simulated constructed wetland. The hypothesis of this study was that the contaminated rainwater can be remediated by constructed wetlands and safely released back into the native waterways. The results of this study help determine if wetlands can rapidly reduce the levels of bio-oil–contaminated storm-water runoff. This study also examined what levels of bacterial microbial communities occur during contamination and how they fluctuate during the breakdown of toxins. Previous studies at the Department of Sustainable Bioproducts at Mississippi State University have demonstrated the feasibility of kenaf fiber and wood shavings to remove toxins and crude oil from the bio-oil wastewater (Moghbeli et al. 2014). This study provides a foundation to future research on bio-oil wastewater–contaminated storm water and its remediation by constructed wetlands.

Methods

The experimental constructed wetlands consisted of sediments of pea gravel at the bottom, sand, and wetland clay soil on top, layered in equal volumes in six 87-liter plastic bins. The bins were connected to each other in a closed system, with the first bin attached also to a small-scale aquatic pump used for initial delivery of contaminated water into the system. Native wetland plants were randomly planted in five bins, with the first bin containing only gravel and sand, being left without any plants or soil, thus serving as a filtration system. The plant species included bulrush (Scirpus californicus), softrush (Juncus effusus), water hyacinth (Eichhornia crassipes), duckweed (Lemna minor), pickerelweed (Pontederia), duckpotato (Sagittaria lancifolia), native canna (Canna spp.), buttonbush (Cephalanthus occidentalis), cutgrass (Leersia oryzoides), and native lotus (Nelumbo lutea).

The contaminated storm water consisted of rainwater subsequently polluted at predetermined levels with bio-oil wastewater acquired from a biomass conversion facility. Four levels of contamination, representative of storm-water runoff that would occur on-site, were obtained by mixing 230 mL, 630 mL, 830 mL, and 2.5 liters of wastewater with 240 liters of rainwater (1,000×, 500×, 300×, and 100× dilution, respectively). The first three dilutions were circulated through the bin system once, and the lowest dilution was applied three times, thus providing the opportunity to measure the effects of increased contamination levels through Phases 1 to 3 and the steady, highest, repeated levels of contamination in Phases 4 to 6. After Phase 3 of the sampling, the new wetland plants were reestablished for Phases 4 to 6. Three repetitions of each sample were taken for all the tests. Sampling of water and soil was performed on the first and tenth day of each phase. In addition, control samples—unpolluted rainwater and contaminated storm water—were analyzed along with the samples from each bin (i.e., microcosm). The samples were analyzed for (1) BOD, (2) microtoxicity, (3) pH, (4) microbial counts, and (5) microbial bacterial community of both soil and water. All results were processed within 1 or 2 days of the sampling event. The BOD analysis were performed according to the standardized test by Enviro-Lab, Inc., in Starkville, Mississippi (EPA 2009). Microtoxicity was tested according to the Microtox Model 500 (M500) 100 percent toxicity procedure. Microbial counts were performed by streaking 10⁴ dilutions of water samples on agar plates after 5 days, and the bacterial community was analyzed through terminal restriction fragment length polymorphism (T-RFLP) according to the method of Kirker et al. (2012).

The results were pairwise compared using SAS, and nonmetric multidimensional scaling (NMDS) of T-RFLP data was performed in PC-ORD software. For NMDS analysis, the data were analyzed for difference based on the Jaccard distance.

Results

pH of water samples

According to the EPA (2009) and the Clean Water Act (1972), permissible discharge limits for storm-water pH levels should remain within a range of 6.5 to 8.5. A majority of the 24 samples of the pH results measured in each microcosm were found to meet permissible discharge levels after 10 days within the wetlands. During Phases 1 to 3, the pH levels increased from acidic values to neutral levels within the first 10 days in the wetlands. The same was observed in Phases 4 to 6 despite a significantly higher contamination level. Data in Table 1 represent average values of all six microcosms.

Table 1 Water quality results measured over time in six phases of the study.^a

^a Values represent an average of all three repetitions for each of the six microcosms at each sampling event for each phase. Within each row, values with different letters denote significant differences at the 0.05 level.

View Fullscreen

Bacterial fragment data of soil and water samples

The T-RFLP analysis identified 6,429 individual bacterial fragments in the soil and the water samples that were representative of 209 distinct taxonomic units (phylotypes). A total of 3,171 individual fragments were detected in the water, and 3,258 fragments were identified in the soil bacteria DNA samples. The T-RFLP data were exported into PC-ORD version 5.0 to determine species richness and diversity.

The microbial species richness in the soil was similar on Days 0 and 10 (Table 2). However, there was a shift in bacterial phylotypes among the phases. Phase 1 was similar in richness to Phases 4 and 5 even though the levels of contamination were substantially different. Phases 2 and 3 were similar in richness and were the only phases to have a higher level of richness on Day 0 than on Day 10. No significant differences were observed in bacterial community richness among the six microcosms.

Table 2 Species richness measured in water and soil samples by terminal restriction fragment length polymorphism.^a

^a The results are compared over time (Days 0 to 10) and by levels of contamination (Phases 1 to 6). Within each row, values with different letters denote significant differences at the 0.05 level.

View Fullscreen

The results from the water microbial species richness test (Table 2) displayed a trend similar to what was seen in the microbial colonies test. Microbial richness was high at the initial extraction and then declined after 10 days in Phases 1 and 3. The opposite was again true for phases of high contamination levels. During these phases, the richness increased with time.

Species diversity of soil is shown in Figure 1. Examining patterns among the phases, NMDS analysis showed species communities distributed in very distinct trends. Phases 1 to 3 clustered together, indicating a similar microbial community. Phases 4 and 6 clustered away from each other, with Phase 5 transitioning between the two clusters, indicating a gradual shift in species with concentration.

View Figure

• Download as Powerpoint
• Download Figure

NMDS analysis of water species diversity (Fig. 2) showed a somewhat similar trend to soil species diversity. Phases 1 to 3 had a similar composition with some overlap with Phases 5 and 6. However, Phase 4 showed a transitional cluster between Phases 5 and 6 owing in part to the replanting that occurred after Phase 3. This indicates a dissimilarity of Phase 4 that could be because of the new plant species added to the soils at that phase.

View Figure

• Download as Powerpoint
• Download Figure

Discussion

Conclusions

• • 

  The constructed wetlands were successful at lowering BOD and toxicity levels and achieving permissible pH levels when the concentration of contaminated storm water was less than or equal to 300× dilution. Much of the BOD reduction was a result of volatilization of the contaminated wastewater.

• • 

  When the concentration of contaminated water exceeded 300× dilutions, the constructed wetlands were successful only at achieving permissible pH discharge levels. Better results may be achievable with the longer residence time in the wetlands.

• • 

  The microbial colonies from Phases 1 to 3 thrived during initial contamination and waned after contamination levels were depleted. During Phases 4 to 6, the microbial colonies thrived in growth after adjusting to the high levels of contamination for 10 days. This is due in part to particular bacteria communities that flourish on the consumption of particular contaminants and other bacteria communities that cannot thrive under those specific conditions. There is potential research for determining which populations thrive under higher bio-oil wastewater contamination and which prosper under lower levels.

• • 

  The majority of bacterial communities compared were similar in composition for Phases 1 to 3 and different for Phases 4 to 6. Additional studies will be required to identify the individual species linked to bio-oil wastewater degradation.

It is acknowledged that the results for Phases 4 to 6 may improve in reduced levels of contamination should the storm water remain in the constructed wetlands longer. However, the amount of contamination lost through volatilization may increase as well.