Chlorinated solvents remain the most common class of contaminants at hazardous waste sites in the United States in general, as well as for the Department of Defense specifically. Bioremediation has emerged as a promising technology for addressing chlorinated solvents with relatively low capital costs, minimal (or no) secondary waste streams, minimal hazard to workers and the environment, in situ contaminant destruction, low maintenance, and minimal site disturbance. However, not all contaminated sites have significant populations of the most important bacteria required for efficient biodegradation of these contaminants, namely, Dehalococcoides spp (Dhc). In those cases, bioaugmentation (adding a concentrated culture of the desired bacteria to a site) is becoming widely used to address potential biological limitations to degradation. While this has been demonstrated to be effective on a small scale, no rigorous full-scale demonstrations have been performed to evaluate different strategies for achieving successful growth and distribution of Dhc bacteria to achieve site cleanup goals.
The overall objective of this work was to compare the cost and performance of full-scale bioaugmentation of chlorinated solvent contaminated groundwater using passive and active bacterial distribution approaches. The technical objectives for this demonstration were as follows:
The relative pros and cons of active recirculation and passive inject-and-drift strategies for large-scale bioaugmentation of chlorinated solvents in groundwater were evaluated in a side-by-side comparison at the Seal Beach Naval Weapons Station (NAVWPNSTA) Seal Beach Site 70 in the City of Seal Beach, California. Three phases of activities were completed for each of the treatment cells, as follows:
Bench-scale testing showed that complete dechlorination of trichloroethene (TCE) to ethene could be achieved even in the presence of high concentrations of sulfate, as long as sulfate-reducing conditions prevailed. While DNA analysis revealed low concentrations of native Dhc at the site in a few locations, it was determined that not all of the known functional genes for dechlorination were present. Specifically, the vcrA gene was absent in site groundwater. As this functional gene is present in commercially available dechlorination cultures, it was tentatively selected as an appropriate biomarker for the bioaugmented culture pending results of DNA analysis of groundwater samples following the pre-conditioning phase.
Baseline groundwater sampling confirmed that initial conditions were mildly reducing, with high concentrations of TCE (maximum concentrations of 140,000 micrograms per liter [µg/L] for the active cell, and 60,000 µg/L in the passive cell), with very little conversion to cis-1,2-dichloroethene (cis-DCE). During pre-conditioning, electron donor was distributed throughout most of the passive cell, and throughout the upgradient portion of the active cell. Where electron donor was distributed, sulfate-reducing conditions were generally achieved, and in some locations, TCE transformation to cis-DCE was observed. However, almost no vinyl chloride (VC) was detected, and Dhc detections were few and at very low concentrations. Most importantly for the DNA analysis of groundwater samples, no detections of the vcrA functional gene were observed, confirming its utility as a biomarker of the bioaugmentation culture.
Bioaugmentation of both treatment cells occurred in January 2009. Following bioaugmentation and during injection of one percent sodium lactate, considerable increases in numbers of Dhc bacteria (ranging from > 106 gene copies/L to > 109 gene copies/L) and all three functional genes (tceA, bvcA, and vcrA) were observed in all wells in the upper portion of the active cell. Overall, conversion of TCE to ethene was proceeding effectively in the upgradient third to half of the active treatment cell, but was not observed at the monitoring well two-thirds of the way down the treatment cell axis.
In the passive treatment cell, the electron donor distribution appeared to improve over time using the original monthly injection frequency. During the post-bioaugmentation phase, TCE and DCE were mostly removed, with VC and ethene observed for the first time at injection wells PIW-2 and -3 within 2 weeks after inoculation in January 2009. As of October 2009, total chlorinated volatile organic compounds (CVOCs) continue to remain low at all three injection wells. However, little to no dechlorination was observed in the upper portion of the passive cell during the post-bioaugmentation phase, possibly due to inhibition of dechlorination due to the presence of other contaminants such as chloroform. In contrast, complete reductive dechlorination of TCE to ethene was observed in the central and lower portion of the passive cell.
The growth of DHC was measured in each cell using DNA analysis of groundwater samples based on the total number of cells at the end of the study compared to the number injected, as well as by tracking increases over time at monitoring wells. Growth was very similar in both cells, with about a two order magnitude increase in cell numbers estimated in each. It was also observed that concentrations at injection wells were sustained above about 106 gene copies/L throughout the test, and concentrations at monitoring wells increased to concentrations approximately equal to the injection wells by the end of the test. As with the first measure of growth, the two bioaugmentation strategies appeared equally effective based on this analysis.
Comparing and contrasting the distribution of Dhc by the two bioaugmentation strategies was the key objective of this demonstration. Based on previous studies of bacterial transport in general, and bioaugmentation specifically, groundwater velocity appeared to be one of only a few parameters than can be easily manipulated during bioremediation that might have a significant impact on transport of Dhc. Relative distribution efficiency of passive vs. active transport was assessed by comparing travel time of injected Dhc to travel time of the conservative tracer (iodide) used in Phase 2 of the demonstration. The groundwater velocity in the active cell was 1 to 1.8 feet per day (ft/d), and for the passive cell it was 0.22 to 0.44 ft/d, a difference of approximately a factor of 5. The tracer and Dhc data indicated that bacterial transport was not significantly retarded compared to groundwater flow in either the active or passive cells. In fact, first arrival of Dhc was faster than that of the conservative tracer in the majority of the passive cell monitoring wells. In the active cell, Dhc transport velocity appeared to be approximately equal to that of the conservative tracer. These results demonstrate that Dhc was transported more rapidly relative to groundwater flow under passive conditions than active recirculation. This is consistent with previous indications that retardation of Dhc transport relative to a conservative tracer increases with groundwater velocity. The net result was that the passive distribution strategy provided effective distribution of Dhc (along with complete dechlorination to ethene) over a larger portion of the treatment cell than was achieved with active recirculation.
Projected implementation costs for a "typical" application (not including the intensive monitoring required for a rigorous demonstration) of bioaugmentation at a 0.5-acre site using the active and passive approach were estimated based on the demonstration costs. Most of the costs are similar (e.g., start-up, general construction, monitoring, and performance assessment) because they are common to both active and passive approaches. However, the construction and operations and maintenance (O&M) costs for the active approach are approximately three times as high as for the passive approach. The result is an estimated cost for the active approach of $2.5M, compared to $1.5M for the passive approach. The primary drivers for this cost increase are the significantly higher amount of lactate required, and the higher costs for construction and maintenance of recirculation systems. For a site like Seal Beach, the benefits of implementing an active recirculation approach do not appear to be justified by the increased costs.
It should be noted, however, that some sites have conditions that would lead to more significant benefits for recirculation systems. For sites with very high groundwater flow velocities, recirculation might be needed to manage residence time within the treatment zone to avoid potential offsite migration of partially chlorinated byproducts such as cis-DCE and VC. Such a site would also allow electron donor to be distributed over a much larger distance prior to being degraded than was possible at Seal Beach, which would also increase the benefit. On the other hand, sites with very low groundwater velocities might make a passive system impractical because very little distribution can be achieved without enhancing the hydraulic gradient. What this demonstration indicates is that for sites that are closer to the "average" in terms of groundwater velocity, passive bioaugmentation systems are likely to be more cost-effective than active systems.