Site-Specific Equilibrium Partitioning Sediment Benchmarks for PAHs in Sediments at Manufactured Gas Plants

Susan Kane Driscoll, Menzie-Cura & Assoc., Inc., 8 Winchester Place, Suite 202, Winchester, MA, 01890, Tel: 781-782-6123, Fax: 781-756-1610, Email: driscols@menziecura.com
Ben Amos, Menzie-Cura & Assoc., Inc., 8 Winchester Place, Suite 202, Winchester, MA, 01890, Tel: 781-782-6128, Fax: 781-756-1610, Email: bamos@menziecura.com
Meg McArdle, Menzie-Cura & Assoc., Inc., 8 Winchester Place, Suite 202, Winchester, MA, 01890, Tel: 781-782-6121, Fax: 781-756-1610, Email: mcardle@menziecura.com
Charles A. Menzie, Menzie-Cura & Assoc., Inc., 8 Winchester Place, Suite 202, Winchester, MA, 01890, Tel: 781-782-6150, Fax: 781-756-1610, Email: camenzie@menziecura.com
Andrew Coleman, EPRI, 3412 Hillview Avenue, Palo Alto, CA, 94304, Tel: 650-855-2249, Fax: 650-855-8588, Email: AColeman@epri.com

From the 1800s, manufactured gas plants (MGPs) produced byproducts, such as coal tars, that contain high concentrations of polycyclic aromatic hydrocarbons (PAHs). The presence of PAHs in phases such as coal tar or soot that do not partition freely may trigger expensive clean-up actions that are not based on site-specific risks.  Sediment quality benchmarks, such as the Equilibrium Partitioning Sediment Benchmarks (ESBs), may be overprotective if the characteristics of the sediment reduce the bioavailability and toxicity. The draft USEPA Bioavailability Procedure uses measured or estimated concentrations of PAHs in porewater to estimate the fraction of total PAHs that are bioavailable. The objective of this study is to examine whether the Bioavailability Procedure can be used to develop site-specific ESBs for PAHs that are conservative predictors of sediment toxicity at MGP sites. Sediments were analyzed for 34 PAHs, total organic carbon, soot carbon, and sediment toxicity. Porewater was analyzed for PAHs and organic carbon. The sum of the ESB toxic units (Sum-TUs) were calculated from: 1) concentrations of PAHs in bulk sediment, 2) concentrations of PAHs in porewater, or 3) concentrations of PAHs and soot carbon in sediment. Results indicate that: 1) Sum-TUs based on concentrations of PAHs in bulk sediment correctly predicted lack of toxicity in sediments with concentrations of total PAHs less than 100 mg/kg, over predicted toxicity at concentrations between 200 to 300 mg/kg, and correctly predicted toxicity at concentrations greater than 300 mg/kg. 2) Sum-TUs based on measured concentrations of PAHs in porewater were somewhat variable, but overestimated toxicity in fewer samples than estimates based on bulk sediment. 3) Sum-TUs estimated from measurements of PAHs in bulk sediment and soot carbon were the most accurate predictors of toxicity.  This research demonstrates that the Bioavailability Procedure is useful in assessing impacts of PAHs in sediments at MGP sites.

Trace Metal Concentrations in the Sediment Cores of Pulicat Lake, East Coast of India

N. Jayaraju, Department of Geology, S.V.University, Tirupati-517 502, India, Email: naddimi_raju@yahoo.com

Threat to the fragile lake ecosystem is alarming across the world.  The present study area, Pulicat lake has no exemption in the Indian context.  To realise this objective, five sediment core samples were collected from the water depths varying between 1.0 to 4.5 m in Pulicat lake.  The study examines the concentration and probable source of the trace metals (Co, Cr, Mn, Pb, Ni and Zn).  Investigations reveal that Ni and Cr are rich in Kalangi estuarine sediments compared to other areas of the lake.  This may be due to the mixing of Industrial outfalls and bioturbation. The cores collected from Northern part of the lake is least affected by the metal mobility.  The study shows that Southern and channel parts of the lake are highly succeptable and an endemic threat awaits to the lake environment.  In addition, it appears that the lake Pulicat, the second largest in India is gradually turning into a garbage bin along East Coast of India.

Monitoring of PCDD/Fs Distribution in Marine Sediments according to Different Particle Sizes

Se-Jin Lee, School of Environmental Science & Engineering, Pohang University of Science and Technology, Pohang, 790-784, Korea, Tel: +82-54-279-8323, Fax: +82-54-279-8299 Email: kiwii4u@postech.ac.kr
Ji-Hoon Kim, School of Environmental Science & Engineering, Pohang University of Science and Technology, Pohang, 790-784, Korea, Tel: +82-54-279-8323, Fax: +82-54-279-8299 Email: residual@postech.ac.kr
Yoon-Seok Chang, School of Environmental Science & Engineering, Pohang University of Science and Technology, Pohang, 790-784, Korea, Tel: +82-54-279-8323, Fax: +82-54-279-8299, Email: yschang@postech.ac.kr
Myeong-Hee Moon, Department of Chemistry, Yonsei University, Seoul, 120-749, Korea, Tel: +82-2-2123-5634, Fax: +82-2-364-7050, Email: mhmoon@yonsei.ac.kr

Marine sediment is an important sink of hydrophobic organic compounds (HOCs) entering from terrestrial and atmosphere. We obtained marine sediments from southeastern coast in Korea and separated into particle sizes. The combination of Pinched-SPLITT (split-flow thin fractionation) technique and high resolution gas chromatograpy/high resolution mass spectrometry (HRGC/HRMS) was used. Separated sediments were processed for HRGC/HRMS analysis using a multiresidue method based on US EPA method 1613. From these analyses, we tried to see the PCDD/Fs distribution in different particle sizes, especially fine particles (< 63 mm); 20-63, 10-20, 5-10, 2-5, < 2 µm. Levels of PCDD/Fs in bulk sediments were similar to previous studies and each sediment sample showed some different homologue profiles due to different environment condition. PCDD/F levels in separated sediments tended to increase as particle sizes decreased. These trends in different particle sizes were related to amount of organic carbon contained in each separated particle. It is considered that small particles have higher surface areas than larger particles and organic carbon can adsorb and capture PCDD/Fs. Fine particles showed higher contamination and this represents the effects of organohalogenated contaminants can be different in particle sizes. Furthermore, the smaller particles have the more potential health risk to marine environments such as deposit-feeder.

Preliminary Sediment Sampling at the Big Lost River Sinks

Christopher J. Martin, S. M. Stoller Corporation, 1780 First Street, Idaho Falls, ID  83401, Tel: 208-525-9358, Fax: 208-525-3364, Email: cmartin@stoller.com
Douglas Halford, S. M. Stoller Corporation, 1780 First Street, Idaho Falls, ID  83401, Tel: 208-525-9358, Fax: 208-525-3364, Email: dhalford@stoller.com
Dr. Richard Marty, S. M. Stoller Corporation, 1780 First Street, Idaho Falls, ID  83401, Tel: 208-525-9358, Fax: 208-525-3364, Email:rmarty@stoller.com

One aspect of the effects of nuclear and chemical waste on plants and animals not addressed by the Comprehensive Remedial Investigation/Feasibility Study Ecological Risk Assessment for the Idaho National Engineering and Environmental Laboratory (INEEL) was the magnitude of contaminant transport down the Big Lost River into environmentally sensitive areas such as the Big Lost River Sinks.  Transport in river systems is a known route for the movement of non-volatile radioactive and conventional contaminants.  Studies of fallout plutonium in rivers have shown that the most important transport pathway is through binding of contaminants to the fine-grained sediment followed by subsequent downstream movement of these particles.  The objective of this project was an initial assessment of contaminant transport in the Big Lost River system to the area of the Big Lost River Sinks, concentrating in the areas of primary sedimentation within the Sinks.  The objective was met through the collection and analysis of sediment samples from a cross-section of depositional environments within the area of the Sinks for metals listed in part 264.24 of the Resource Conservation and Recovery Act and gross radionuclides.  The first sampling was done to a depth of 30 cm.  Additional sediments were collected to one meter (1 m) or refusal for future analysis.  Statistical analyses of the various contaminants measured were carried out using nonparametric methods, specifically the Kruskal-Wallace ANOVA for comparisons of multiple sample groups and the Mann-Whitney U test for paired comparisons.  Statistical analysis showed that the concentrations of the radionuclides and metals measured in this initial assessment were statistically the same or lower than the background values used, with the exception of aluminum, barium, and chromium.  This presentation will discuss the sample results and statistical analysis, highlighting data limitations and recommendations.  

Effects of Data Sampling in Geostatistical Modeling of Sediment Contaminant Concentrations

Kandiah Ramanitharan, Department of Civil & Environmental Engineering, Tulane University, New Orleans, LA 70118, Tel: 504-865-7313, Fax: 504-862-8941

Sampling strategies and consequent modeling play important roles in characterization of contaminated sediments. This paper analyses how the sample size and density influence the variogram fitting and the kriging in the aquatic sediment contaminant concentration modeling. Ordinary kriging and ordinary cokriging models are considered for the study. Concentration data of three heavy metals (Cr, Ni and Cd) measured at Duwamish River bottom sediment and the percentages of different sediment soil particles (Clay, silt, and sand) are used in this case study. Three subsets are created from each full data set by randomly picking 25%, 50% and 75% of the full dataset. Each subset is created in a manner to be the subset of any larger set. Whenever a data set is not of normal distribution, either log-normal transformation or Box-Cox transformation is found suitable to pretreat the data before variogram fitting. Micro-spatial correlation is modeled with spherical, Gaussian and exponential variograms with nugget effects. After the transformation, the trend component is fitted and removed from the data, and the residual is used to fit the variogram. The fitted variogram is used in the kriging and the kriged values are cross validated with the data measurements in the subset. Minimization of Root Mean Square (RMS) errors of cross validation result is used as the best model parameter-selecting criterion together with an additional condition of having same spread in measured data and cross validated values. In addition, the model is further validated by using the data that are removed from a full set to make a subset as the testing set for the particular subset. The results show that the model predictions rapidly improve with number of data until a relatively ‘steady’ in the cross validation RMS error is achieved. Cokriging with other heavy metals considerably improve the predictions. While cokriging with the soil constituent percents provides better predictions than those can be achieved in the univariate kriging, these cokriged predictions are much inferior to those obtained with the cokriging with other heavy metals. Results are discussed, and the future research goals are identified.

Top