[Table of Contents] [Next Section]

Section 4 - Treatibility Study Results and Data Evaluation

4.1 - Sampling Results

The purpose of this section is to summarize the sampling results and provide an evaluation of the data collected during the benchscale and pilot-scale electrokinetic system at SWMU O-1. Soil samples were collected during system installation to characterize the pre-study conditions and during process operations to determine contaminant transport performance within the SWMU O-1 area.

4.1.1   Initial Baseline Soil Sample Data

Soil samples were collected in July 1997 near test excavation No. 3, see Figure 2.12, as part of the initial pilot test installation phase. The results of the 1997 sample analyses are presented in Table 4.1.

The analytical results show a significant amount of chromium contaminant concentrations in the general vicinity of the field test. However, cadmium and PCE concentrations collected from the area were relatively low. The highest concentrations of metal contaminants were detected in anode well A2 (0 to 1 feet bgs) during the initial electrokinetic pilot test installation. Anode well A1 (0 to 1 feet bgs) showed the largest amount of PCE concentration within the identified electrokinetic pilot test installation area. The average concentrations of chromium, cadmium, and PCE are 544 mg/kg, 1.32 mg/kg, and 2.88 mg/kg, respectively. The total mass of each contaminant represented by the average concentrations within SWMU O-1 test area is approximately 16 pounds of chromium and less than 1 pound each of cadmium and PCE.

The soils collected for conducting a laboratory benchscale test contained approximately 783 mg/kg chromium and 2.5 mg/kg cadmium.

4.1.2   System Performance Soil Sample Data

The locations of the system performance soil sampling profiles are shown in Figure 3.1. The samples were collected from depths ranging between 0 and 4 feet bgs. Results of metal and VOC analysis generated from both Lynntech and Parsons ES sampling events are discussed below.

Metal Contamination

Results of system performance sampling data indicated that metal contamination at the SWMU O-1 site was very heterogeneous, both laterally and vertically. Lynntech metal analysis results show that the starting chromium concentration between the electrodes was greater with soil depth, particularly at 3 feet and below as compared to the surface contamination. It seems that the chromium level concentration at these depths was not much affected by the electrokinetic process, because the deeper soils showed no significant acidification. The acid produced electrochemically and transported electrokinetically through the soil was not enough to acidify the soil at a depth of 3 to 4 feet.

Figure 12 of Lynntech's technical report (Lynntech, 1997a) consisting of 10 graphs, summarizes the results of chromium analysis performed on soil core samples taken during the process operation. The data at the end of the process were obtained by sampling the trenches excavated on January 6, 1998, as shown in Figure 4.1 of this report. The data are grouped in the graph for each particular depth, e.g., 0 to 6", 6 to 12", etc., and changes in chromium with respect to time are shown for a particular depth. Noted problems from the performance sampling events conducted by Lynntech include inadequate sample volumes collected at necessary locations and the inability to collect soil samples at depth. Both of these noted problems were due to large limestone rocks which inhibited the ability to collect necessary data.

Volatile Organic Compound Contamination

Results of analyses for PCE from the initial test startup indicate that there was very little PCE contamination within the test location (3.64 mg./kg). Reviewing the most recent soil gas distribution (Figure 2.11), PCE soil gas concentrations are highger outside of the excavated areas. Results of analyses for PCE from the final sampling event indicate a significant amount (48.9 mg/kg) of PCE was available within the test area soils. The data indicates potential movement of PCE into the treatment from areas outside of the field scale treatment zone.

Based on the screening analytical results of the soil sampling performed during the initial pilot test startup (September 1997), the November 1997 performance sampling activities, (results of analysis shown in Table 4.2); and the pilot test final sampling event, (results of analysis shown in Table 4.3); the electrokinetic process for volatile organic compounds indicates that the process will transport VOCs through electroosmosis to the cathode well.

Geotechnical Data

Soil samples collected from fill material for the laboratory benchscale test also included a geotechnical analysis, results of which are presented in Section IV.A.1 Table 2 of Lynntech's technical report (Lynntech, 1997a) provided in Appendix A.

Based on the data from the geotechnical analysis, none of these physical parameters appear to limit the potential effectiveness of electrokinetic remediation in the SWMU O-1 area. In fact, the values from the geotechnical testing indicate that the soil type and its hydraulic permeability are suitable for the application of electorkinetic process at the O-1 site.

4.1.4   Quality Assurance Summary

All laboratory metals data collected as aprt of the electrokinetic treatability study by Parsons ES were validated using the Air Force Center for Environmental Excellence (AFCEE) Quality Assurance Project Plan (QAPP) version 1.1 to ensure generation of defensible data. An ITIR is provided as Appendix B to this report. Deviations from the QAPP or the analytical method and a discussion of the overall usability of the data are presented in the ITIR. Because of laboratory inaccuracies in methods used for quantifying VOC present in samples analyzed by SW-8260, data generated during the previous investigations are considered screening data only. These data are not validated and, thus, are not considered to determine quantity and extent of VOC contamination or to provide defensible data.

The analytical techniques and QA/QC procedures established and used by Lynntech for analysis of soil and well water are documented in Section III.4 "Methods and QA/QC Procedures." The data generated by Lynntech are also considered screening data.

In summary, all metal samples were prepared and analyzed within the specified holding times. Only EPA-approved analytical procedures were used. Noted laboratory problems and deviations from the QAPP are summarized below.

1. Hexavalent chromium samples were analyzed at 48 hours, which exceeded the 24-hour holding time. The hexavalent chromium result in samples gathered during the investigations are considered to be usable; however, they are flagged "R".

2. All ICP serial dilution criteria were met, except for barium, copper, nickel, and silver during the liner investigation. The barium, copper, and nickel results in the associated samples were considered to be estimated and flagged "J". The silver result already had a more severe qualifier of "F". The zinc result in the affected samples was flagged "B" to indicate that the blank analyte was found in the affected sample and associated blank. The selenium results in samples OX-1-97-SEW-1, OX-1-97-SEW-2, and OX-1-97-SEW-3 was considered to be unusable and flagged "R" because of continuing calibration criteria not being met.

4.2 - Laboratory Benchscale Treatability Results

Electrokinetic experiments were performed in the laboratory by Lynntech to provide data to support the field pilot test. The results of the laboratory test are detailed in Section IV.A of Lynntech's technical report provided in Appendix A.

4.2.1   Determination of Optimum Extractant for SWMU O-1 Soil

The results of the batch tests performed on the O-1 soils showed the highest chromium removal from the soil was obtained when oxalic acid and citric acid were components of the extractant solutions. The results demonstrate that by lowering the soil pH, higher efficiency of chromium removal can be achieved. In most cases when leachant solutions included citric acid, solubilization of chromium in the soil was enhanced due to the possible formation of stable chromium complexes with citrates. The citric acid was chosen because it is environmentally benign and additionally provides anions which form soluble salts and/or complexes in soil with chromium and other heavy metals at the CSSA SWMU O-1 site.

The initial pH of SWMU O-1 soil for the benchscale test was approximately 10, and Lynntech reported the soil exhibited a high buffering capacity. Tests were performed to determine the most efficient acid for acidifying the soil and extracting chromium from soil. It is important to note that the benchscale test found no difficulties in solubilizing cadmium within the SWMU O-1 soils collected. However, extraction experiments concentrated on chromium extraction because cadmium concentration in the soil matrix was low (2-3 mg/kg) and therefore provided little data to support the anticipated field scale test. A detailed discussion and summary of results of cadmium extraction from SWMU O-1 soils is provided in Appendix A, Section IV.A.4 of Lynntech's Technical Report.

Oxalic acid exhibited the highest chromium removal, approximately 72%, recorded during the long term (16 hours) titration experiments. Citric acid exhibited the highest removal in the short term (2 hour) titration experiments. Results indicate that oxalic acid appeared to be the most efficient acidifier for SWMU O-1 soil, and citric acid can be efficiently used as a complexing agent and solubilizer for chromium removal. However, it was observed that oxalic acid can form a hard calcium oxalate precipitate which can increase soil resistance; therefore, oxalic acid was not considered to be compatible with the electrokinetic soil processing at the O-1 site. Thus, citric acid was used as a principal component of the leachant solution in further tests.

Citrate is one of the most powerful complexing agents for metals. It is a tri-protic acid that can form a number of bonds with cations and can react with some cations sufficiently to convert them to uncharged, or even negatively charged species. These citrate complexes have different ionic mobilization in an electric field which could affect the movement direction in which the movement of metals occurs within the treatment area. Because of the limitations for this study effort, no data was collected regarding the complexation of metals with citric or hydrochloric acid.

During the batch leachability tests, it was found that when citric acid was added to the O-1 soil, soil swelling occurred. Concerns were noted during the benchscale testing, that this soil swelling would have adverse effects relative to the maintenance of the electrokinetic field process unit. Tests were performed to find ways of controlling the soil swelling during leachant addition. The test performed with regard to control of soil swelling included a mixing of hydrochloric, and/or formic acid with citric acids at varying concentrations to provide data on the ideal leachant solution. The swelling was determined by measuring the change in volume of the soil occupying a testing vial. Results of the laboratory effort to control soil swelling were duplicated in the field at the SWMU O-1 site using plastic 5-gallon containers. The soil swelling test results determined that hydrochloric acid used in a combination with citric acid provided solubilization of heavy metals, as well as calcium salts, and was environmentally acceptable.

4.2.2   Determination of the Valence State of Chromium

One of the concerns in the electrokinetic soil process is that the chromium ion contaminant could exist or be converted to its hexavalent state (Cr[VI]), which is more mobile than Cr(III) and is a highly toxic form of chromium. The benchscale laboratory test included efforts to identify the valence state of the chromium ions after extraction. The O-1 soil was titrated using citric acid and chromium ion species analyzed in the filtrate after extraction. It was found that the majority of chromium contaminant in the solution after extraction was in lower valence than the Cr(VI).

Additional tests were performed to identify the possible conversion of lower valence state chromium into the hexavalent form. Soil samples were spiked with the Cr(VI) ions and extraction was performed using varying concentrations of benign organic acid (citric and acetic acid) solutions. Test results indicate that use of citric acid was capable of efficiently reducing the spiked Cr(VI) to its lower valence form. Acetic acid showed no reduction of hexavalent chromium. It was concluded that citric acid, when added to contaminated O-1 soil, prevents the conversion of chromium from its lower valence state(s), and helpskeep chromium contaminant in the lower valence state ionic forms. Further discussions on the valence state of chromium are found in Section IV.A.3 of Lynntech's technical report provided in Appendix A.

4.2.3   Benchscale Electrokinetic Removal of Chromium from O-1 Soil

Benchscale laboratory electrokinetic experiments were performed on SWMU O-1 soil as described in Section 3 of this report. Results of the benchscale tests performed are discussed in detail within Lynntech's technical report provided in Appendix A. A summary of these results is disccused below.

The laboratory electrokinetic tests were conducted over a period of 36 days. During the first 19 days a two percent solution of citric acid was added to the cathode well to control the pH were a basic solution is formed due to the electrochemical hydrogen evolution reaction. This addition also provides soluble anions and serves as a complexant for metal contaminants in the soil. The electroosmosis flow from the anode toward the cathode helped in the distribution of citric acid in soil. During the first 19 days of operation, the pH of the soil showed no significant change. Since previous batch tests confirmed that chromium removal was directly related to lowering the soil pH, a more concentrated citric acid (40%) was added to both the anode and cathode wells. In addition, to speed up the soil acidification, citric acid was injected in the soil between the wells. A soil pH of 2-3 was achieved after five weeks of continuous treatment.

Chromium contaminant removal accomplished after 36 days of electrokinetic treatment in teh benchscale tests was in the range between 60% (in the center of the bed) and 94% (near the electrode wells). Very little movement of chromium was observed during the first 19 days of treatment due to the soil pH being close to neutral. Acidification of the soil facilitated chromium solubilization, transport, and removal from the O-1 soil.

After 36 days of continuous laboratory treatment, the soil was sectioned in three portions - soil near the anode, middle of the cell, and near the cathode. Each soil sample was then homogenized and analyzed to determine total amount of chromium remaining in the soil. The sample analysis of the soil portion nearest the cathode showed a 99.8% removal efficiency. The sample analysis of the soil portion between the electrodes showed a removal of 84%. A 64% removal of chromium in soil nearest the anode indicated that the chromium was transported and accumulated toward the anode. The movement and accumulation of chromium toward the anode well indicates the chromium in the soil is in the anionic form and is reportedly controlled by electromigration and dielectrophoresis.

4.3 - Field Pilot Scale Treatability Results

The objective of the field pilot scale treatability study was to compile data for an evaluation of the effectiveness for heavy metal and PCE contaminant removal from O-1 soil under conditions encountered at the actual SWMU O-1 field site. A detailed discussion on the result of the field study is presented in Lynntech's technical report provided in Appendix A of this report. The following is a summary of the results observed by Lynntech.

The field demonstration of Lynntech's electrokinetic process, in operation for 89 days, clearly demonstrates an almost linear dependence of chromium removal was 34%, which occurred near the anodes were soil was acidified to a pH of 2 to 4. Only 13% of the chromium was removed near the cathode where less efficient soil acidification occurred. No measurable amount of chromium removal was noticed in the middle regions between the electrodes. Similar to the benchscale results, migration of chromium to the anode indicated that transport of the negatively charged chromium ions was exclusively by electromigration.

The following sections discuss the results of process operation and monitoring, chromium contaminant transport and removal observations and the correlation between chromium removal and soil acidification. In addition, process cost estimates are also discussed.

4.3.1   Process Operation and Monitoring

Total voltage applied to the field study system was 14.0 volts and a current per anode was in the range between 2.5 to 3.0 amps, and at the cathodes 8 to 10 amps. This yielded total current to the field of 18 to 20 amps. The applied pulsing regime was: on/off 20 seconds/5 seconds. The process unit was turned on August 15, 1997, and was shut down on December 23, 1997. The total energy consumption during the processing period was estimated at 538.2 kilowatts.

Based on the findings of the benchscale study. soil acidification was the most critical step in the treatment of SWMU O-1 soils. To accelerate this process, the concentration of citric acid/hydrochloric acid used to neutralize the electrochemically produced base in the cathode wells was increased to 5% citric acid and 10% hydrochloric acid. To further expedite the soil acidification, addition injection wells were mounted between the electrode wells. Both horizontal and vertical injection wells were installed. The acid addition was controlled through a solenoid valve with the flow adjusted to the soil hydraulic permeability.

4.3.2   Chromium Contaminant Transport and Removal Observations

Lynntech, Inc. reported the results of chromium contaminant transport observed from monitoring the process control zones (profiles) as described in Section 3.3.2 of this report. The two control zones were :

1. Cathode well 2 - mid-point - anode well 6; and

2. Cathode well 1 mid-point - anode well 1.

The methodology also included data gathered in the vicinity of each anode and cathode well. Samples were taken approximately every three weeks and analyzed for chromium by Lynntech. Sample locations are shown in Figure 3.1.

Results of chromium analyses from the process monitoring sampling effort are presented as Figure 12 in Lynntech's technical report. The generated monitoring data, from the field pilot scale study. was plotted on three dimensional graphs. The data included chromium levels or pH, with respect to location and time. Time is plotted on the x-axis; sample location between the cathode and anode wells are plotted on the y-axis; and the z-axis plots the corresponding chromium concentration of pH. The use of the three dimensional graphs provides a determination of temporal and spatial changes in contaminant concentration or pH of the O-1 soil. The three dimensional graphs for the monitored control zones are provided as Figure 13, Figure 14, Figure 15, Figure 16, and Figure 17 in Lynntech's technical report.

A comparison of data generated from monitoring control zone 2 (cathode well 1 - anode well 1) for layer 0 to 6 inches (Figure 13) and layer 6 to 12 inches (Figure 14) show corresponding trends in chromium removal from 0 to 1 soil. The data shows elevated chromium contamination approximately midway between the cathode well 1 and anode well 1 at the start of the electrokinetic process. Through time, the chromium migrated toward both the anode and cathode wells, resulting in a significant decrease in chromium contamination in the middle treatment area. The movement of chromium was accomplished by electromigration toward the anode well and by dielectrophoresis and electroosmosis toward the cathode well. Chromium contamination of soil at depths greater than 1 foot showed no significant movement or acidification. This is most likely due to the limited time available for acidifying the soil.

As observed in the benchscale study there is an almost linear dependence of chromium removal on soil pH. Figure 15 of Lynntech's technical report shows pH values corresponding to chromium data obtained for core samples taken at 0 to 6 inches in depth (Figure 13). Comparing the pH changes and chromium concentration, a clear correlation between pH and chromium removal was obtained. The three dimensional graphs illustrate that the soil was most successfully acidified near the anode (pH 2-3) where an electrochemically produced acid entered the soil in addition to the acid injected through an injection well system. Efficient contaminant movement at these locations was observed.

Comparison of Figure 16 and Figure 17 in Lynntech's technical report demonstrate the same type of correlation between the pH and chromium removal for the first control zone (profile 1) between anode well 6 and cathode well 2. The changes in pH near the anode in time closely follows the changes in chromium removal int he vicinity of the specific electrode. The pH between anode well 6 and cathode well 2, toward cathode well 2 remained mostly unchanged. In this region no significant chromium removal was observed. Figure 18, presented in Lynntech's technical report, clearly depicts the correlation between the soil acidity and the chromium concentration in the vicinity of the anode well. Table 4.4 shows the average percentages of removal of chromium for the different regions within the control zones of SWMU O-1. The average removal was calculated by comparing the initial value of chromium (25 September 1997 data) with the values obtained for samples taken at the end of the process (6 January 1998).

Table 4.4 - Average Percentage Removal of Chromium from O-1 Soil

The concentration of chromium found in the electrode wells and effluent was in the range of 10 to 1000 parts per million (ppm). This indicated that most of the chromium was concentrated on the packing material of the electrode wells. Packing material surrounding each well was sampled, and results of analyses indicate that chromium concentrated during the process within the areas as expected.

4.3.3   Process Cost Estimates

A cost estimate for the operation of the electrokinetic soil processing system was prepared by Lynntech and presented in Section IV.C. of Lynntech's technical report found in Appendix A. The estimated costs were based on the energy consumption during the three months of process operation and cost for chemicals used in the process. The total operation cost per ton of soil was reported to be $490. The majority of the cost was due to the large amount fo chemicals used to acidify the soil.

4.4 - Investigation-Derived Waste Management Summary

IDW generated from the electrokinetic system removal effort included excavated packing material and surrounding soil material, personal protective equipment (PPE), well effluent, and miscellaneous debris. The PPE and miscellaneous debris were placed into plastic bags and disposed of along with general plant trash. Excavated soils generated were placed into labeled 55-gallon containers. Approximately six containers of excavated soil material were generated. One container of the cathode well effluent was generated. Accordingly, the containers which were labeled as containing IDW were sampled 9 January 1998, for proper waste characterization as specified in 30 TAC 335 Subchapter R. Results of TCLP analyses for the IDW soils identified PCE present at 0.081 milligrams per liter (mg/L), and chromium at 1.31 mg/L/. The IDW soils meet nonhazardous Class i criteria levels. The containerized soils were properly disposed of an at off-site landfill. Results of TCLP analysis for IDW effluent water identified no VOCs, and chromium present at 38.0 mg/L. The IDW effluent water is classified as a hazardous waste (EPA code D007) and was properly disposed of an an off-site landfill.

[Next Section]