Land Treatment and the Toxicity Response of Soil Contaminated with 7Jvood Preserving Wste

Scott G. Huling Daniel F. Pope *John E. Matthews Judith L. Sims Ronald C. Sims Darwin L. Sorensen

Scott G. HulJng is an

(P.E> d t b tbe USEPA Robert S Kerr Emrironmental Researcb Laboratory (RS-lU.) in Ada, Ohlabonra. He bas a BS is Bndmnental Sciencefron, tbe Uduersity of Texas at San Antonio, an M.S. in Environmental Engine&ngJrom tbe UniversUy of Texas at Austi% and is a doctoral candidute at the Department of Cbemistry a d Environnoefftal l?ngine&ng at tbe University of Arizona in Tucson, Arizona Dr. Daniel F. Pope bas reseambed &lopatby, Motransformation of woodpreserving and petroleum wastes, and designed SOU and water Moremediation systems. Currently be assists EPA- RSKERL witb site cbaractetization and remediatioa Jobn E. Mattbews, environmental scienlist, received a RS. and an M.S. in aquatic biologyfrom Utab State Uniuersity (USU). f f e worked at tbe EPA- RSKERL for 26 years in

environmentacengineer

Soik contaminated m'th woodpresm'ng wastes, includingpentachlo- rophenol (PCP) and creosote, are treated at field-scale in an engineered prepared-bed system consisting of two one-acre land Ireatment units (LTUs). Ihe concentration of selected indicator compounds of treatment w o r m a n c e included PCP, pyene, and total carcinogenic plycyclic aromatic hydrocarbons (TCPAHs) was monitored in the soil by taking both composited soilsamples at multiple points in lime, and discrete soil samples at twopoinfi in time. i%e mean concentration of the indicator compounds and the 95-percent confidence interval (CI> of the composite and discrete samples agreed relatively well, and fkt-order degradation rate kinetics satisfactorily represented the mean chemical concentration loss of indica- tor compounds in the LTU. Toxicity of the soil, as measured by MicrotoxTM assay of the soil extracts, indicated that toxicity reduction corresponded with indicator compound disappearance. No toxicity effects were observed with time in treated layen ofsoil (rifts) buried beneath highly contaminated lifts of newly applied soil. nis indicated that vertical migration of soluble contaminants from such lifts had little eflect on the microbialactiuity in the underlying treated soil.

The U.S. EPA established the Bioremediation Field Initiative in 1990 to help develop bioremediation as an effective remediation technology. One objective of the initiative was to obtain and disseminate field-based data and information from field experiences concerning the implementation and performance of bioremediation techniques. As part of the initiative, evaluation of surface soil treatment in a prepared-bed system was conducted at the Champion International Superhnd Site (Libby Site), a former wood preserving facility in Libby, Montana. The prepared-bed system consisted of two one-acre lined land treatment units (LTUs). The field performance evaluation was designed and conducted by Utah State University (USU) with support from the U.S. EPA, Robert S. Kerr Environ- mental Research Laboratory. Champion International mainrains and oper- ates the LTUs at the Libby Site.

The objective of the LTU performance cvaluation was to conducr a comprehensive field study to assess performance of the L'I'Us in terms of treatment effectiveness, rates of disappearance of contaminants of interest,

CCC 1051-5658/95/050241-15 (0 1995 John Wiley & Sons, Inc.

41

S c m HULING DANIEL POPE JOHN M A ~ S JUDITH Sws RONALD SIMS Dmwm SORENSEN

tbe areas of subsurfme invest#gations, land tre- a d SOU Mo+enredkrtlorr, and is tbe preddent ofJMCQ 1% E d v W CmsuWng. Juditb L Shns is a reseatrb assistant professor at tbe Utab Water Researcb Laboratory at USU. A s a soil scientist, ber researcb interests include tbe development of gutdunce concerning tbe application of remedial tecbnologies to contaminated soil systems. Dr. Ronald C Sims is a professor and bead of tbe LUvision of Envitvnmental Engineering, Department of Civil and Environmental Engineering at USU. His researcb interests include assessment and development of treatment tecbnologies for soils contaminated witb bazardm wastes. Dr. Darwin L Smensen, an environmental microbiologist, is a researcb associate professor of tbe Utab Water Researcb Laboratory at USU. He received a Pb.n from Colorado State University in 1982 and bis researcb interests include intrinsic bimenaediation of subsurfae systems.

and detoxification of the contaminated soil. Results of soil detoxification measured using the MicrotoxTM assay, as a function of treatment of the contaminated soil in the LTUs, are presented in this article.

SITE DESCRIPTION Contamination of soils at the Libby Site resulted from wood preserving

operations from 1946 to 1969. Contaminants primarily consisted of residuals from creosote and penrachlorophenol (PCP) wood preservatives. Contami- nated soilsfrom three primarysourceareas (tank farm, bun dip area, andwdste pit) were excavated, screened to remove rocks larger than one inch in diameter, and placed in an excavated waste pit area. Contaminants of concern at the site include: (1) polycyclic aromatic hydrocarbons (PAHs), (2) PCP, and (3) dioxins (impurities in technical grade PCP). Target remediation levels for the contaminated soils are: 88 mgkg TCPAHs, 8.0 m@g naphthalene, 8.0 mgkg phenanthrene, 7.3 mdkg pyrene, 37 mgkg PCP, and I 0.001 mgkg dioxin equivalency. Total carcinogenic polycyclic aromatic hydrocarbons (TCPAHs) consist of four-ring compounds (bcnz(a)anthracene, chrysene); five-ring compounds (benzo(b)flouranthene, benzo(k)flouranthene, benzo(a)pyrene, dibenz(a,h)anthracene); and six-ring coinpounds (benzo(g,h,i)perylene, indeno(1,2,5cd)pyrene).

The LTUs are one-acre each and consist of a liner system (60 mL high density polyethylene (HDPE) liner, compacted clay, geotextile filter fabric, and geonet); a leachate collection system (perforated HDPE piping, gravel drains, and sump); rainfall runoffAeachate storage (two 50,000-gallon tanks); and a passive moisture control system (perforated HDPE piping). Design criteria were total containment of contaminated soils, surface runoff, and leachate, with treatment and ultimate disposal of all contami- nated soils within the LTUs.

Contaminated soils stockpiled in the waste pit area are pretreated by tilling, adding nutrients, and controlling moisture. Soil from the waste pit area was placed in the LTlJs in approximately 6-inch layers (lifts). The LTUs were irrigated to maintain soil moisture content at 40 to 70 percent of field capacity in the treatment zone, and for dust control. Water application was managed to minimize the volume of ledchdte collccted in the L l z l sump. Nutrients (usually ammonium sulfate and ammonium phosphate) were added to the LTUs either as solids or dissolved in irrigation water. Amounts added were based on a desired 12-3O:l soil carbon-to-nitrogen ratio and a 10: 1 nitrogen-to-phosphorus ratio. Treatment-zone soil was monitored for total organic carbon, total Kjeldahl nitrogen, and total phosphorus. Nutrients were added as needed. The LTUs were tilled weekly, when weather conditions permitted, using a tractor-mounted rototiller or similar equipment. LTU operations were discontinued during winter, and new lifts were not applied until the contaminant cleanup levels were achieved.

EXPEWMENTAL APPROACH The toxicity of water extracts of soil in the L'l'lls was evaluated using

the MicrotoxTM assay. The concentration of indicator compounds in the LTUs was also measured to evaluate treatment performance and to relate

42 RXMEDIATION/SPRING 1995

LAND TREATMENT AND THE Toxrcm RESPONSE OP SOIL CONTAMINATJID WITH WOOD PRESERVING W m

Exhibit 1. LTUs 1 and 2: Quadrant Numbers and Location of Sampling Core Numbers for Obtaining Discrete Soil Samples.

to the toxicity measurements. Intensive soil sampling in the LTUs was conducted over a two-year period. Discrete soil samples were collected and analyzed for PCP, pyrene, TCPAHs, phenanthrene, and naphthalene. PCP, pyrene, and TCPAHs were used as indicator parameters and correlated to toxicity response. Concurrently, Champion International conducted routine operational monitoring of the LTUs using composited soil samples. Discrete samples (individual samples taken from one location) were used by Utah State University (USU) instead of composite samples (samples taken from more than one location and mixed before analysis) in order to more closely delineate the range of contaminant concentrations found in the LTUs. Results of analyses of composite samples tend toward the mean concentration of the sampled area and do not show the full range of concentrations actually present. Data from both the discrete and composited samples were used to measure treatment and were correlated to the toxicity response with treatment.

MATERLALS AND METHODS

Discrete Soil Sample Collection, Handling, and Analysis Discrete soil sampling locations were determined by overlaying a

rectangular grid on the LTUs. The location of the origin of the grid was randomly selected, and each point of intersection of the grid within the LTU bed was designated as a sampling point.' The locations of the sampling points within LTUs 1 and 2 are presented in Exhibit 1. There were 32

REMEDIATION/SPRING 1995 43

S u x r HULING DANIEL POPE JOHN hksmimvs JUDITH SPAS RONALD SPAS DARWIN SORENSEN

Exhibit 2. Lift Loading and Sampling Dates for Discrete Samples.

L N 1 Lift No. Lift Application Date Date Sampled

1 6/29/89 5/6/91, 9/18/91, 9/1/92 2 8/8/89 same as lift 1 3 7/11/90 same as lift 1 4 5/9/91 9/18/91, 9/1/92 5 7/23/91 7/27/91 6 5/5/92 -

LTU 2 1 7/25/91 9/18/91, 9/1/92

sampling points per LTU; at ieast 30 samples were collected during each sampling event; a subset, usually 18-20 samples, was analyzed. A Giddings tube sampler was driven into the soil to the desired depth, and the soil core was removed from the tube and placed in an I-CHEM jar. Soils were extracted (U.S. EPA Method 3550) using sonication which ensures intimate contact of sample matrix with extraction solvent (1 :I methylene chloride- to-acetone ratio). PCP was analyzed using gas chromatography (GC) (U.S. EPA Method 8040), and PAH compounds using gas chromatography/mass spectrometry (GC/MS) (U.S. EPA Method 8270).

Soil sampling in LTU 1 began after the first three lifts had been applied, and began in LTU 2 following application of the first lift. The accurate thickness of lifts afier initial placement of approximately six inches of contaminated soil is unknown due to natural and managed soil compac- tion. Therefore, sampling depths are approximately correlated with the lifts placed in the units. The first lift is well-correlated with depth since the different textures OF the underlying sand and contaminated soil were easily distinguishable, However, lifts 2 and 3 were difficult to distinguish, and samples collected from these lifts were a composite of the two lifts. Lift 4 was compacted prior to application of lift 5, so it was readily distinguished due to density differences in the soil core sample. A summary of the soil application dates and discrete soil sanipling dates are provided in Exhibit 2.

Composite Soil Sample Collection, Handling, and Analysis Composited soil samples collected by Champion International were

analyzed in the on-site laboratory and were primarily used to guide day- to-day operation of the remedial actions. A composite soil sample was collected from each of the four quadrants (Exhibit 1). An equal aiiioiint of soil from the top six inches was collected from two to four lwations within each quadrant, composited in a stainless steel bowl, and mixed, for a total of four samples at each sampling event. PAHs and PCP were analyzed using a modified U.S. EPA Method 8100 and 8040, respectively.'

44 &,MBDIATION/SPRING 1995

The procedures are relatively simple and inexpensive

Toxicity The Microtoxm assay is a general toxicity assay that measures the

microbial response, as light output, to an aqueous sample. This assay was used in this study to measure the effect of a complex chemical e m a from contaminated soil on a single test species under specifk test conditions. This assay is not intended to provide information on toxicity from a human health or safety perspective.

The Microtoxm assay can be used as a screening test to measure the relative toxicity of chemicals in soils. This information can be correlated with the degradation of the chemical compounds and provide a general indicator of the toxicity reduction. The procedure has been used success- fully in previous studies to screen and predict the treatability potential of waste in soils'; to estimate an appropriate range of waste loading rates and detoxification potential'; to determine appropriate application rates for specific wastes5$ to assess the detoxification of a complex petroleum waste in soils'; and to determine the relative toxicity of various chemical types and isomers5

The MicrotoxTM assay is a simple, standardized acute toxicity tesc8 that utilizes a suspension of marine luminescent bacteria (Photobacterium phosphoreum) as the bioassay organism and a "challenging" solution. P. phosphoreum are bioluminescent microorganisms that produce light output which represents a complex chain of biochemical reactions involving the luciferin-luciferase system. The Microtoxm assay involves subjecting a suspension of P. phosphowum (i.e., the "test reagent") to several dilutions of a soil sample extract (i.e., the "challenging solution") and measuring the light output to evaluate the relative toxicity. The light output of the organism is monitored under defined conditions of exposure time (t) and test temperature (TI. A dose-response curve is generated and used to determine the effective concentration (EC50) of the test sample that is required to cause a 50-percent reduction in light output after five minutes. The procedures used to perform this test and to calculate the EC50 are outlined in the MicmtoxrM Manual." The reduction in light output of the microorganisms represents physiological inhibition, not just mortality, indicating the presence of toxic constituents in the test sample, and therefore a measure of the toxicity.

The advantages of MicrotoxTM are that the procedures are relatively simple and inexpensive; the results can be obtained within a relatively short time-frame (less than one hour for the assay); a small volume of sample (as little as 10 mL) is r e q ~ i r e d ~ . ~ ; and a statistically significant number of test organisms are used.lo

The soil sample extracts utilized in the testing procedure were obtained by extracting approximately 25 g of each soil sample with 100 mL of distilled, deionized water (DUW) in glass jars with screw-cap lids. The samples were extracted for approximately 24 hours in a rotating box at 30 rpm. Approximately 50 mL of supernatant from each sample were transferred to separate Nalgene centrifuge tubes. Suspended solids were then removed from the extracts by centrifugation for 20 minules at 5000 rpm. The osmotic pressure of the extracts was adjusted by adding NaCl 10

REMEDIATION/SPRING 1995 45

SCOTT HULING DANIEL POPE JOHN &ITHEWS JUDITH SIMS RONALD SIMS D ~ W T N SORENSEN

I I I I

~~~ ~~-

Exhibit 3. PCP Concentration in Soil (First-Order Degradation Model).

Cornposited Soil Samples LTU2,Li f t l

Ln Wpl (mglkp)

I

5

4.5 1 "1 3 Ln[PCP]

bring the extract concentration to 2 percent NaCl, as required for the Microtoxm assay. Dilutions of 50,25,12.5, and 6.25 percent were prepared for each aqueous extract. The P. phosphoreum suspension was then challenged at each sample concentration (dilution), and the light output was recorded at t = 0 and t = 5 minutes. A positive reagent control (blank) was tested concurrently with each test sample. This control blank consisted of DDW adjusted to 2 percent NaCl and l0p.L of the MicrotoxTM reagent (=lo6 P.phosphot.eum). Light loss for the control blank and the test samples was compared with the light loss caused by a known toxic standard (sodium pentachlorophenate at 10 mg/L). One toxic control standard was run for every ten samples.

RESULTS

PCP, Pyrene, and TCPAH Soil Concentration

Composited Soil Samples, Multiple Sampling Events Soil concentration data for PCP, pyrenc, and TCPAHs have been

plotted to evaluate degradation kinetics (first-order model), Exhibits 5 5 , respectively. The soil samples were collected from LTU 2, lift 1 starting 7/25/91 (day one) through 10/21/91 (day 881, representing 13 sampling events. The data scatter in these figures illustrates the Variability of soil contaminant concentration, which is reasonable and expected. Linear regression was used to calculate degradation rates (k), correlation coeffi- cients (s), and half-lives (tJ which are summarized in Exhibit 6 . The kinetic data indicate that PCP, pyrene, and TCPAHs each degrade as a function of time in the LTUs. However, the low correlation coefficients (9)

46 R.EMEDIATION/SPIUNG 1995

TBEATMENT AND THB T o x m RESPONSE OF Son ~ONTAMINATZD WITH WOOD PRXSERVING W m

Exhibit 4. Pyrene Concentration in Soil (First-Order Degradation Model).

Cornposited Soil Samples L T u 2 , L i f t l

Ln [Pymnel (ma/kg) 5 r

L

8

3.5 -

~ni~yrene] - -0.01 %(t) + 4.53 2.5 - f '2 = 0.55 = I .

8

2 1 I I I I

0 20 40 60 a0 100 Tlme (days)

-%it 5. TCPAH Concentration in Soil (First-Order Degradation Model).

Cornposited Soil Samples L T u 2 , L m l

Ln VCPAHs] (mglkg)

6

5.5

5

4.5

4

3.5

B

I I I I

0 20 40 60 80 100 Time (days)

from regression analysis indicate that time of treatment explains 55-68 percent of the change in concentration with time. It is reasonable to expect that the low correlation is due to the inherent variability of contaminant concentrations in the LTU.

Discrete Soil Samples, Two Sampling Events The mean and !%-percent confidence interval (Cl) for the mean for

PCP, and pyrene and TCPAHs soil concentrdtions are presented in

~

~EDIATION/SPRING 1995 47

Sccyrr HULING DANIEL POPE JOHN MA~HEWS JUDITH SIMS RONALD SIMS D a m SORENSEN

Exhibit 6. Summary of Chemical and Kinetic Data for Composite and Discrete Samples LTlJ 2, Lift 1.

Composite Samples Discrete Samples

Correlation Deet.ad. Deet;ld coeffldent ( P 2 ) r n i t i a I M a n Cone. Rate, k W-Li fe Xnitial Mean Conc. Rate, k --Life

ContPmLnvlt Quadrants 1,2,3,4 (rngllcg) (11-41 Wday) (days) (melkS, (n-20) (Uday) (days)

PCP .97, 90, .70, .76 115.1 -.0192 36 101.4 -.0163 43 Pyrene .88, .90, .a, .48 90.5 - . o m 45 84.9 -.0125 55 TCPAHS .88, .m, .70, .58 279.5 -.0183 38 204.0 -.0127 55

~

Exhibit 7. PCP Concentration in Discrete Soil Samples, L T U 2, Lift 1, Day 1 and llay 53'

Quadrant 1 Quadrant 2 Quadrant 3 Quadrant 4 LTU

Day 1 (7/27/91) 141.3 (0-305) 66.6 (41.3-91.8) 71.3 (54.7-87.9) 126.4 (51.5-201.3) 101.4 (66.1-136.7) Day 53 (9/18/91) 23.9 (0-52.4) 31.5" (13.7-43.4) 73.6 (21.2-1 26.0) 36.0 (2.7-69.3) 42.7- (26.4- 59.0)

* Mean concentration in mg/kg (95% CI for the mean), day 1 (n-20). day 53 (11-19) " Excluding one value in quadnnt 2 (1 168 mdkg)

Exhibit 8. Pyrene and TCPAH Concentration in Discrete Soil Samplcs, LTIJ 2, Lift 1, Day 1 and Day 54'.

Day 1,7/27/91 wene 108.0 (34.0-181) 89.4 02.6-146.0) 59.5 (46.3-72.6) 83.3 (36.5-130.0) 84.9 (65.0-1050)

Day 1, 7/27/91 TCPAHS 26.3 (80.9-444) 62.6-240) I j j (12j-184) 248 (91.9-404) 204 (153.3-255) .* 43.2 (0-94.0) Day 54,9/19/31 Pyrene ** .. 0.

Day 54,9/19/91 TCPAHs ** .. It .. 103 (1.2-204)

Mean concentration in mdkg (95% C1 for the mean), day 1 (n-20). day 54 (n-6) " Insufficient number of samples collected to repod mean and 95% Q:

Exhibits 7 and 8, respectively. The soil samples wcre collectcd from LTU 2, lift 1 on 7/27/91 and 9/18/91. These data reprcscnt discrete soil samples collected in the four quadrants of LTU 2.

Similar to the cornposited sample data set, contaminant concentration variability is also observed in the discrete sample data set. For example, the standard error is 35 percent of the mean for PCP (day one) and in Exhibit 7, quadrant 3, the mean concentration was greater at day 53 than at day one, while the overall mean in LTU 2 showed a statistically significant decrease with time. Likewise, one value (1 168 mg/kg) for PCP in quadrant 2 was one to two orders of magnitude grcatcr than the othcr 19 values in this data set. The mean pyrene and TCPAH concentrations (Exhibit 8)

REMEIXAT~ON/SPIUNG 1995

LAND T X E A ~ AND THE Toxrcrr~ RESPONSE OF Son, CONTUINATED wrrx WOOD P ~ E ~ E ~ ~ I N G WASTE

Exhibit 9. Mean Concentration of PCP, Pyrene, and TCPAHs of the Composited and Discrete Soil Samples in LTU 2, Lift 1..

Day 1 Composite (7/26/91) 115.1 (100.1-130.1) 90.5 (71.0-110.0) 279.5 (223336) Day 1 Discrete (7/27/91) 101.4 (66.1-136.7) 84.9 (65.0-105.0) 204.0 (153.3-255) Day 60 Composite @/23/91) 45.9 (5.4-86.4) 43.4 (17.1-69.7) 92.5 (34.6-150.4) Day 54 Discrete (9/18-19/91) 42.7" (26.4-59.0) 43.2 (0-94.0) 103 (1.2-204)

Mean concenuation in mdkg 05% Q for the mean), PO clay 1, n-2O; clay 54. n-19, Pyrene and TCPAl-I day 1, n-20 clay 54. n-6 " Excluding one value in quadrant 2 (1168 mg/kgl

decreased, but the decrease was not statistically significant (95-percent CI) for the LTU or any of the quadrants. This variability is believed to be attributed to: (1) hot spots from the heterogeneous placement of highly contaminated soil in the LTU from the waste pit area (stockpile), and (2) to the presence of high-concentration soil-waste oil conglomerates.

Comparfson of Composited versus Discrete Soil Samples Indicator compound half-life values from the composited samples

were calculated directly from the first-order degradation rate constant (k) in the regression equation; the half-life values from the discrete samples were calculated assuming first-order degradation rate kinetics (i.e., C(t)/Co = e-k'). The data from the discrete samples, collccted at two points in time (days one and 53) for PCP, pyrene, and TCPAHs, were used to calculate k; t,, was calculated directly from k and summarized in Exhibit 6. PCP, pyrene, and TCPAH half-lives for the composited samples were 36.1,44.7, and 37.9 days, respectively; and 42.5, 55.4, and 54.8 days, for the discrete samples, respectively. Although the data are variable, half-lives based on the mean values for PCP, pyrene, and TCPAH using data sets derived from different sampling approaches only differ by 15 percent, 19 percent, and 31 percent, respectively. Additionally, while the mean and 95-percent CI for the PCP, pyrene, and TCPAH concentrations for discrete and composited soil samples in LTU 2, lift 1 (Exhibit 9) indicite variability, these values also agree relatively well.

Pentachlorophenol The PCP soil concentration data for composited soil samples (Exhibit

3) were linearized using a first-order model and regressed by quadrant (Exhibit 6). The correlation coefficients in quadrants 1 and 2 indicate that the PCP soil concentration conformed relatively well to first-order degra- dation rate kinetics. The lower correlation coefficient in quadrants 3 and 4 indicate that other factors, in addition to time of treatment, affected data Variability.

PCP decreased by 58 percent during 53 days (7/27/91 to 9/18/91)

REMEDUTION/~PRING 1995 49

Scan HI"G DANIEL POPE JOHN M A ~ ~ E W S J U D ~ SIMS RONALD SIMS DARWPI SOaeJm

Regression analysis of pyrene curd TCPAH concentration data from cornposited soil samples indicates that the soil pyrene concentration conformed to first- order degradation kinetics in quadrants 1 and2.

based on the mean PCP concentration in discrete samples collected in LTU 2. PCP in composite samples decreased by 60percent during the 6Oday period (7/26/91 to 9/23/91) based on the mean concentration. The half- lives using the discrete and composite data were within 15 percent, that is, 42.5 and 36.1 days, respectively. These half-life values are greater than those reported by Dasappa and Loehn" and fall within the range reported for a variety of soils by McGinnis.l2

Results of a laboratory study using spiked radiolabeled PCP and soil from LTU 1 and 2 indicated that PCP was metabolized to CO, and water by indigenous soil microorganisms at temperatures and moisture contents representative of the field conditions in LTUs 1 and 2.13 Incorporation of radiolabeled carbon into the soil solid phase was also a significant treatment process. Volatilization of PCP at field-scale was estimated to be minor due to the insignificant losses observed in laboratory "flow-through" soil microcosms. Therefore, the loss of PCP at field-scale, as indicated by both the discrete and composite data, can be attributed to both biotic and abiotic processes.

Pyrene and TCPAHs The pyrene concentrations decreased by 49 percent during 54 days

(7/27/91 to 9/19/91) based on the mean pyrene concentration in discrete samples and decreased 52 percent during the 60-day period (7/26/91 to 9/23/91) in the composited samples. The half-lives using the discrete and composite data were within 19 percent, that is, 55.4 and 44.7 days, respectively. These values are consistent with other values reported for

Regression analysis of pyrene and TCPAH concentration data from composited soil samples indicates that the soil pyrene concentration conformed to first-order degradation kinetics in quadrants 1 and 2 (Exhibit 6). Similar to PCP, lower correlation coefficient in quadrants 3 and 4 indicate that other factors, in addition to time of trcatment, affected data variability.

TCPAH concentrations decreased 50 percent during 54 days (7/27/91 to 9/19/91) based on the mean TCPAH concentration in discrete samples in LTU 2. However, the mean TCPAH concentration in composite samples was estimated to decrease 67 percent during a 60-day period (7/26/91 to 9/23/91). The half-lives using the discrete and composite data were 54.8 and 37.9 days, respectively, a variation of 31 percent.

md.7.12.14

Soil Toxicity Response MicrotoxTM analyses were performed on discrete soil samples collected

from LTU 2, lift 1, on 7/27/91 (day one), two days after lift 1 was applied. The ten samples represented replicates from five locations on the sample grid in Exhibit 1. An additional ten discrete soil samples were collected from LTU 2, lift 1, on 9/18/91 (day 531, and eight of these samples represented replicates from four grid locations. MicrotoxTM results from day one and day 53 are presented in Exhibits 10 and 11, respectively. These data indicate that toxicity, a s measured by Microtoxl", decreased signifi-

50 REMEDIATION/SPRING 1995

LAND TREATMENT AND THE Toxrcm RESPONSE OF Son. CONTAMINATED WITH WOOD PRESERWNG WASTE

Exhibit 10. Soil Microtox Toxicity (EC50) Values Discrete Soil Samples Collected 7/27/91 LTU 2, Lift 1, Day 1.

Exhibit 11. Soil Microtox Toxicity (EC50) Values Discrete Soil Samples Collected 9/18/91 LTU 2, Lift 1, Day 53.

-No. EC50 ]lo+ High' Corr.Coef. GridNo. ECSO Low. High. Corr.Coef.

34 34 43 43 45 45 54 54 57 57

3.73 2.54 3.03 1.89 5.09 3.51 7.21 4.42 8.44 6.15 7.1 6.62 7.84 5.7 8.89 8.08 6.74 3.87 7.88 4.82

Mean 6.60

5.5 4.86 7.37 11.75 11.59 7.63 10.77 9.77 11.74 12.89

0.9950 0.9939 0.9937 0.9840 0.9920 0.9997 0.9926 0.9992 0.9810 0.9822

34 34 43 45 52 52 54 54 58 58

NDR" NDR NDR NDR 47.5 24.1 93.7 0.9724 NDR NDR NDR 8.86 6.68 11.76 0.9933 9.01 7.85 10.33 0.9984

'35% CI for the EC50 95% C1 for the ECjO

** NDR = No Dose Responsc

cantly from day one (mean EC50 6.6) to day 53 where seven of the ten soil samples had no dose response (NDR), as did background samples. The three values in Exhibit 11, where an EC50 value was measured, reflect the variability of contaminant concentrations in LTU 2 that was observed earlier. For example, while the overall mean PCP concentration decreased (Exhibit 7) from day one (101.4 mdkg, n=20, 95-percent CI 66.1-136.7) to day 53 (42.7 mg/kg, 11-19, 95-percent CI 26.4-59.01, there were random high concentrations as indicated in Exhibit 12, which may have resulted in measurable EC50 Values.

The reduction in MicrotoxTM EC50 from 7/27/71 to 9/18/91 (Exhibits 10 and 11) corresponded with reductions in PCP, pyrene, and TCPAHs concentrations noted in Exhibits 7 and 8. A similar correspondence between toxicity reduction, as measured by MicrotoxTM, and PCP soil c~ncentration,~*" and PAH soil c~ncentration'.'~.'~ has been observed in laboratory microcosms.

MicrotoxTM analyses were performed on discrete soil samples collected from lift 4, which was loaded on 5/9/91 in LTU 1. Monitoring data indicated that lift 4 had reached cleanup goals by 7/13/91 ; correspondingly, lift 5 was loaded 7/23/91. Seven discrete soil samples were collected for MicrotoxTM analyses from lift 4 on two different dates, 9/18/91 and 9/1/72, representing 133 and 481 days from lift application. This lift was sampled on the same dates for analysis of PCP, pyrene, and TCPAH soil concentration, and results are presented in Exhibit 3. All MicrotoxTM values for the 14 discrete samples collected in lift 4 were nontoxic. These results are consistent with those observed in LTU 2. For example, in LTU 2 on day 53, seven out of ten MicrotoxTM values indicated NDR when the mean PCP, pyrene, and TCPAH concentrations were 42.7 mdkg, 43.2 mdkp,, and 103 mdkg,

REMEDIATION/SPB.ING 1995 51

SColT HULING DANIEL POPE JOHN M A ~ S JUDITH SIMS RONALD SMS DARWIN SOENSEN

Exhibit 12. PCP Concentration in Discrete Soil Samples, LIV 2, Lift 1, Day 53.

PCPQulndrant LTU Quadrant Core No. PCP (mg/kg) M e a n (mg/kS, 95Oh CI for M e a n -

1 36 45 47 48 39

2 38 40 42 44 54 55 58 59 Go 49 51 52 63 64

LTU Mean

49.1 23.9 4.5-52.4 7.9 15.8 23.0 37.6

14.9 1167.8 37.6 21.7 104

118.6 86.4 37.2 15.2 83.0 24.8 30.1 26.9 42.7'

34.1 31 .Y 13.7-43.4'

73.6 21.2- 126.0

36.0 2.7-69.3

26.4-59.0'

Excluding one value in quadcint 2. core no. 42

respectively (Exhibits 7 and 8). On 9/18/91 and 9/1/92 in LTU 1, lift 4, the mean PCP, pyrene, and TCPAHs concentrations were clearly below these values (Exhibit 13, and MicrotoxTM ECSO values indicate no toxicity.

The potential for vertical migration of the contaminants being treated or their transformation products is an important concern for LTU perfor- mance. Typically, vertical migration is evaluated using chemical analysis of soil pore liquid or soil core samples. Alternatively, an indirect measure- ment of this process is to evaluate the MicrotoxTM response in buried lifts. At wood preserving sites, PCP and its decomposition products (i,e., teua-, tri-, di-, and chlorophenols) are very toxic to P. phosphoreurn.l2 PCP and tetrachlorophenol salts are routinely used as a control in the calibration of MicrotoxTM. In this performance evaluation, samples from buried lifts were assayed by MicrotoxTM to determine whether vertical migration of contami- nants occurred from upper lifts.

Exhibit 14 presents thk concentrations of PCI', pyrene, and TCPAH in LTU 1, lift 5, when it was frst applied onto lift 4 , and on the two later dates (9/18/91,9/1/92) when lift 4 was sampled for MicrotoxTM analyses. Contami- nant concentrations in LTU 1, lift 5 on 7/27/91, were greater than values in LTU 2, lift 1, day one (Exhibits 7 and 81, which yielded a mean EC50 of 6.6. It was assumed that the E G O of soil emcts from LTU 1, lift 5 on 7/27/91 would have also been high. The MicrotoxTM EC50 values for the 14 discrete samples

52 REMEDIATION/SPRING 1995

LAND T-TMENT AND THE Toxrcnr RESPONSE OF Son. CONTAMINATED WITH WOOD PRESERVING WASTE

Exhibit 13. PCP, Pyrene, TCPAH Soil Concenua- tions Discrete Soil Samples, LTU 1, Lift 4'.

Exhibit 14. PCP, Pyrene, TCPAH Soil Concentra- tions Discrete Soil Samples, LTU 1, Lift 5'.

Date PB Pyrrnc T C P B

9/18/91 20.7 4.6 33.0

11-21 n920 n=20 (8.7-32.6) (2.8-6.6) (18.5-47.5)

9/1/92 10.5 3.9 41.0 (7.0-13.9) (1.3-6.1) (13.3-64.4) 11-20 11-18 n-18

Date PCP Pyrenc TCP- -~ ~~

7/27/91 119.4 135 254

11-20 n520 n920 (93.9-145.0) (102-167) (179-328)

9/18/91 40.5 35.3 103

11-20 n-20 1-1-20 (28.9-52.1) (14.5-56.0) (52.5-153)

9/1/92 16.9 4.3 37.1

n-20 n=19 n=19 (13.3-20.5) (1.4-7.3) (23.1-51.2)

Mean concentration in mdkg (95% C1 for the mean) Man concentration in n d k g (95% C1 for the man)

collected from the buried lift 4, LTU 1 on 9/18/91 and 9/1/92 (56 and 292 days after application of lifl 5, respectively) were nontoxic. Lift 6 was applied on 5/5/92, and an additional set of discrete soil samples were collected from LTU 1, lifi 1: ten on 5/8/91, eight on 9/18/91 and nine on 9/1/92 (27 total); and from lifts 2 and 3: nine on 5/7/91, ten on 9/18/91, and nine on 9/1/92 (28 total). Microtoxm results for the discrete soil samples indicated that all 55 samples from lifts 1, 2, and 3 were nontoxic.

These data indicate that loading contaminated lifts onto lifts that had previously reached the cleanup goals had no measurable effect on the MicrotoxTM response in lower lifts in LTU 1. The Microtoxm assay is generally more sensitive to toxicants (PCP, PAHs) than other indicators of soil microbial activity in the presence of wood preservative and petroleum- related chemi~als .~ .~ .~ . '~ Therefore, soil microbial activity in LTU 1 would not be expected to be negatively impacted by the wdter soluble contaminants.

An evaluation of the decrease in mutagenic potential using the Ames assay in LTU 1 was undertaken in a separate study.17 Initial mutagenic potential of soil applied to LTU 1 was considered to be approximately 330 revertants per gram of soil (weighted activity). Results of mutagenicity testing for lift 1 sampled September 1989 (three months after application of lift 1) indicated detoxification to soil background levels (less than 150 revertants per gram of soil). Soil samples taken 1.5 months after application of lift 2 (November 1989) showed only thrcc of nine samples with mutagenic activities near or greater than 150 revertants per gram of soil (indicative of contaminated soil). Results obtained in another study indicated that treatment of contaminated soil in LTU 1 reduced the, mutagenic potential of the solvent extracts of soil (measured as weighted activity or revertants per gram of soil) to background levels after approxi- mately three months of treatment.'*

REMEDLATION/SPRING 1995 53

Sco'rr HULING DANIEL POPE JOHN M A ~ S JUDITH SIMS RONALD SIMS D ~ W I N SORENSEN

CONCLUSIONS A statistically significant decrease in PCP, pyrene, and TCPAH concen-

trations occurred at field-scale as determined by composite or discrete soil sampling and analysis. Based on mean concentrations, first-order degra- dation rate kinetics satisfactorily represented the chemical loss for these compounds. Additionally, good agreement was observed between composited and discrete soil samples for the mean concentrations and half- Iives of PCP, pyrene, and TCPAHs.

Detoxification occurred in PCP and creosote Contaminated soil being treated at field-scale in prepared-bed )and treatment units. Detoxification occurred over the same time-frame as the degradation of indicator compounds, PCP, pyrene, and TCPAHs. No increase in toxicity in lower lifts of soil was observed when highly contaminated soil was applied to lifts that had previously undergone remediation. This indicated that vertical migration of water-soluble Contaminant extracts from such lifts has little effect on soil microbial activity in the underlying treated soil. Overall, soil treatment included degradation of target PCP and PAH compounds to target remediation levels and detoxification of thc soil.

NOTES

1. Mason, B.J. 1983. Prepnrution of Soil Sumpling Pro[ocoC i'bchniqiies r4nd .YtrW.?gies. EPA-600/4-83-020. Environmental Monitoring Systems Lalwratory, 1J.S. Environmentnl Protection Agency, h s Vegas, Nevada.

2. Woodward Clyde Consultants. May, 1790. No Migrrrtioir Petition Report: rand T i z a t m t chlfts, Libby Montana. Prepared for Champion International Corporation, Stamford, Connecticut.

3. Matthews, J.E. and A.A. Uulich. "AToxicity Reduction Tcsl Systcin to Assist in Predicting Land Treatability of Hazardous Organic Wastes." In: J.K. I'ctros, Jr. et al., Ed., Hazardous andIndustrialSolid Wustc Testing: Fourth Symposium. I'hil:itlclphin, ASI"M/SI'P 8x6, 1984.

4. Matthews, J.E. and L. Hastings. "Evaluation o f Toxicity 'I'cht Procedure for Screening Treatability Potential of Waste in Soil." T i i c f p Assewnrrzt 2: 26 j-281, 1987.

5. Loehr, Ray C., TreatahiIity Potentiul forEPA ListedHazcwdotis Wnste in Soil, EPA/6002- 89/011, US. EPA Robert S. K e n Environmcntal Rcscxch I;ibomtory, A h , Okl:ihonia, March, 1989.

6. Sims, R.C., D.L. Sorensen, WJ. Dourette, and I..I.. 1I;istings. 19861~ WastdSofl Treatability Studies: Metbodologies mad Resiilts. Volume 2. Wmte Loading Impacts on Soil Degradation, Transformation, und Ivmohilimtioiz. 1.:13A/600/6-s6/003h, Rohcr t S . Kerr Environmental Research Laboratory, U S . Environriicntal I'rotcction Agency, Ada, Okln- homa . 7. Symons, U.D., and R.C. Sims. 1988. "As.ses?;ing 1)ctoxifir';ition of :i Complex 1 I:iz:irclous Waste, Using the MicrotoxTM Assay." Arch. o/liiit~imii. Coizt. mid Taz. 1 7 497-505.

8. Microbics Corp., 1992. MIcrotm" iMunrrrr1, Microhics Corp., Carlsharl, Cilifornia.

9. Sims, R.C., J.L. Sims, D.L. Sorensen, and L.L. Iiastings. 19th. WastdSoil Trcf&&li!v Studies: Methodologies and Results. Volrimc 1. IPA/600/6-86/003a, Rolxrt S. Kcrr Environ- mental Research Laboratory, U.S. Environmental Protection Agency, Ada, Oklahonia.

10. Dutka, U.J. and G. Uitton. Toxicity Testing lJsiizg Mionoqoiiisms, CRC I'rcss, Inc., I3oca Raton, Florida, 1986.

11. Dasappa, S.M. and R.C. Lwhr. 'Toxicity Hcditction in Contaniinalcd Soil I~ioremcdiation

54 REMEDLATION/SPIUNG 1995

LAND TREATMENT AND THE Tox~cm RESPONSE OF SOIL CONTAMINATED w r r ~ WOOD PRESERVING WASTE

Processes." Waf. Re$. Vol. 25, No. 9, pp. 1121-1130, 1391.

12. McGinnis, G.D., H. Borazjani, D.F. Pope, D.A. Strobel, and L.K. McFarland. On-Slte Treatment of C m o t e and Pentacblorophenol SIUdges and Contaminated Soil. EPA/600/ 2-91/019. May, 1991, USEPA-RSKERL.

13.Sims, R.C., J.L. S i m , D.L Sorensen, and J.E. McLean. Champion Internattonal S u m n d Slte, Ltbb, Montana: Biomnedtation FfeM Perfomnce Evaluation of the prepared-BedLand TEahnent Sysrem. Robert S. Ken Environmental Research Laboratory, U.S. EPA, Ada, Oklahoma, DRAFT, July, 1334.

14. Bulman,T.L.,S. Lesage, P.J.A. Fowlie,andM.D. Webber. ~ e P ~ ~ t ~ c e o f P ~ s l n S o t 1 . PACE Report No. 85-2, Petroleum Association for Conservation of the Canadian Environ- ment. 1202-275 Slater St. Ottawa, Ontario, Canada. November, 1985.

15. Abbott, C. and R.C. Sims. "Use of Uioassays to Monitor PAH Containination in Soil." pnxeedings of the 10th Natlonal Supejlrnd Conference, Washington D.C., November 27- 29, 1989.

16. Wang X., X. Yu, and R. Uartha. 1990. "Effect of Bioremediation on Polycyclic Hydrocarbon Residues in Soil." Enuiron. Sci. and TechnoL 24, 1086-1089.

17. April\, W., R.C. Sins, J.L. Siins, and J.E. Matthews. 1990. "Assessing Detoxification and Degradation of Wood Preserving and I'etroleum Wisics in Contaminated Soil." Waste Mgmt. and Res. 8: 45-65.

18. Donnelly, K.C., C.S. Anderson, J.C. Thomas, K.W. Brown, D.J. Manek, and S.H. Safe. 1992. Bacterial Mutagenicity of Soil Extracts from a 13iorcincdiation Facility Treating Wood- preserving Waste. lourn. Huz. Mat. 30: 71-81.

REFERENCE

Sims, R.C. 1990. Soil Remediation Techniques at Unconuolled Hazardous Waste Sites: A Critical Review. lourn. Air G Waste Mgmt. A s . 40(5): 703-732.

ACKNOWLEDGMENTS

The authors wish to acknowledge the significant support of the Champion International staff at Libby, Montana, specifically: Ralph Heinert, Dave Cosgriff, Gerald Cosgriff, Jim Carraway, and Jim Davidson. Without the cooperation and support of these individuals, this project could not have been accomplished. Also, we would like to acknowledge the support from Bert E. Dledsoe and Dr. Mary Randolph of the EPA-RSKERL, and Jon Ginn, a doctoral candidate in the Department of Environmental Engineering at Utah State University. Logan, Utah. Aldiough the research descrilxd in this article has been funded by the U.S. EPA, jt has not k e n subject to internal Agency peer review and therefore does not necessarily reflect the views of the Agency and n o official endorsement should be inferred.