United States Office of Office of Solid EPA/540/4-91/003 Environmental Protection Research and Waste March 1991 Agency Development and Emergency Response Ground-Water Issue CHARACTERIZING SOILS FOR HAZARDOUS WASTE SITE ASSESSMENTS R. P. Breckenridgel, J. R. Williams2, and J. F. Keckl INTRODUCTION tion provides data types required for decision The Regional Superfund Ground Water making in three distinct RI/FS tasks: Forum is a group of ground-water scentists representing EPA’s Regional Offices, orga- 1.. Determination of the nature and extent of nized to exchange up-to-date information re- soil contamination. lated to ground-water remediation at hazard- ous waste sites. Soil characterization at 2. Risk assessment, and determination of hazardous waste sites is an issue identified by risk-baaed soil dean-up levels. the forum as a concern of CERCLA decision- makers. 3. Determination of the potential effective- ness of soil remediation alternatives. To address this issue, this paper was pre- pared through support from EMSL-LV and Identification of data types required for the first RSKERL, under the direction of R. P. task, determination of the nature and extent of Breckenridge, with the support of the contamination, is relatively straight forward. Superfund Technical Support Project For The nature of contamination is related to the further information contact Ken Brown, EMSL- types of operations conducted at the site. LV Center Director, at FTS 545-2270 or R. P. Existing records, if available, and interviews Breckenridge at FTS 583-0757. with personnel familiar with the site history are good sources of information to help determine Site investigation and remediation under the the types of contaminants potentially present. Superfund program is performed using the This information may be used to shorten the CERCLA remedial investigation/feasibility Iist of target analytes from the several hundred study (RI/FS) process. The goal of the RI/FS contaminants of concern in the 40 CFR Part process is to reach a Record of Decision 284 list (Date 7-l-89). Numerous guidance (ROD) in a timely manner. Soil characteriza- documents are available for planning all 1 ldaho National Engineering Laboratory, Environmental Science and Technology Group, Idaho Falls, ID 83415. 2 Soil Scientist, U.S. EPA/R.S. Kerr Environmental Research Laboratory, Ada, OK 74820 SuperfundTechnology Support Center for Monitoring and Site Characterization, Environmental Monitoring Systems Laboratory Las Vegas, NV Superfund Technoloqy Support Center for Ground-Water Fate and Transport, Robert S. Kerr Environmental Research Laboratory Ada, OK aspects of the subsequent sampling effort (US EPA 1987a, Concerns 1988a, 1988b, and Jenkins et at., 1988). This paper addresees two ooncerns related to soil characteriza- The extent of contamination is also related to the types of tion for CERCLA remedial response. The first concern is the operations conducted at the site. Existing records, if available, applicability of traditional soil declassification methods to CERCLA and interviews with personnel familiar with the site history are soil characterization. The second is the identification of soil aIso good sources of information to help determine the extent of characterization data types required for CERCLA risk assess- contamination potentially present. The extent of contamination ment and analysis of remedial alternatives. These concerns are is dependent on the nature of the contaminant source(s) and the related, in that the Data Quality Objective (DQO) process extent of contaminant migration from the source(s). Migration addresses both. The DQO process was developed, in part, to routes may include air, via volatilization and fugitive dust emis- assist CERCLA decision-makers in identifying the data types, sions; overland flow; direct discharge; Ieachate migration to data quality,and data quantity required to support decisions that ground water and surface runoff and erosion. Preparation of a must be made during the RI/FS process. Data Quality Objec- preliminary site conceptual model is therefore an important step tives for Remedial Response Activities: DevelopmentProcess in planning and directing the sampling effort The conceptual (US EPA 1987b) is a guidebook on developing DQOs. This model should identify the most likely locations of contaminants process as it relates to CERCLA soil characterization is dis- in soil and the pathways through which they move. cussed in the Data Quality Objective section of this paper. The data type requirements for tasks 2 and 3 are frequently less Data types required for soil oharacterization must be determined well understood. Tasks 2 and 3 require knowledge of both the early in the RI/FS process, using the DQO process. Often, the nature and extent of contamination, the environmental fate and first soil data types related to risk assessment and remedial transport of the contaminants, and an appreciation of the need alternative selection available during a CERCLA site investiga- for quality data to select a viable remedial treatment technique. tion are soil textural descriptions from the borehole logs pre- pared by a geologist during investigations of the nature and Contaminant fate and transport estimation is usually performed extent of contamination. These boreholes might include instal- by computer modeling. Site-specific information about the soils Iation of ground-water monitoring wells, or soil boreholes. Typi- in which contamination occurs, migrates, and interacts with, is cally, borehole logs contain soil Iithology and textural descrip- required as input to a model. The accyof the model output tions, based on visual analysis of drill cuttings. is no better than the accuracy of the input information. Preliminary site data are potentially valuable, and can provide The purpose of this paper is to provide guidance to Remedial modelers and engineers with data to begin prepare Project Managers (RPM) and On-Scene Coordinators (OSC) conceptual model and perform scoping calculations. Soil tex- ooncerning soil characterization data types required for ture affects movement of air and water in soil, infiltration rate, decision-making in the CERCLA RI/FS process related to risk porosity, water holding capacity, and other parameters. assessment and remedial alternative evaluation for contami- Changes in lithology identify heterogeneities in the subsurface nated soils. Many of the problems that arise are due to a lack of (i.e., Iow permeability layers, etc.). Soil textural declassification is understandng the data types required for tasks 2 and 3 above. therefore important to contaminant fate and transport modeling, This paper describes the soil chactererization data types re- and to screening and analysis of remedial alternatives. How- quired to conduct model based risk assessment for task 2 and ever, unless collected properly, soil textural descriptions are of the selection of remedial design for task 3. The information limited value for the following reasons: presented in this paper is a compilation of current information from the literature and from experience combined to meet the 1. There are several different systems for classification of soil purpose of this paper. particles with respect to size. To address this problem it is important to identify which system has been or will be used EMSL-Las Vegas and RSKERL-Ada convened a technical to classify a soil so that data can be properly compared. committee of experts to examine the issue and provide technical Figure 1 can be used to compare the different systems (Gee guidance based on current scientific information. Members of and Bauder, 1986). Keys to Soil Taxonomy (Soil Survey the cornmittee were Joe R. Williams, RSKERL-Ada; Robert G. Staff, 1990) provides details to one of the more useful Baca, Robert P. Breckenridge, Alan B. Crookett and John F. systems that should be consulted prior to classifying a site’s Keck from the Idaho National Engineering Laboratory, Idaho soils. Falls, ID; Gretchen L Rupp, PE, University of Nevada-Las Vegas; and Ken Brown, EMSL-LV. 2. The accuracy of the field classification is dependent on the skill of the observer .To overcome this concern RPMs and This document was compiled by the authors and edited by the OSCs should collect soil textural data that are quantitative members of the committee and a group of peer reviewers. rather than qualitative. Soil texture can be determined from a soil sample by sieve analysis or hydrometer. These data Characterization of a hazardous waste site should be done types are superior to qualitative description based on visual using an integrated investigative approach to deterrnine quickly analysis and are more Iikely to meet DQOs. and cost effectively the potential health effects and approach response measures at a site. An integrated approach involves 3. Even if the field person accurately classifies a soil (e.g., as consideration of the different types and sources of contami- a silty sand or a sandy loam), textural descriptions do not nants, their fate as they are transported through and are parti- afford accurate estimations of actual physical properties tioned, and their impact on different parts of the environment. required for modeling end remedial alternative evaluation, 2 such as hydraulic conductivity. For exarnple, the hydraulic The types of decisions vary throughout the RI/FS process, but conductivity of silty-sand can range from 10-5 to 10-1 cm/sec in general they become increasingly quantitative as the pro- (four orders