U.S. Geological Survey

Radon Potential of the Upper Midwest


(Note: numbers in parentheses refer to numbered references at the end of this paper.)

INTRODUCTION

Assessing the geologic radon potential of the United States has been the subject of a two-year project by the U.S. Geological Survey and the Environmental Protection Agency, in cooperation with the Association of American State Geologists. Indoor radon data from the State/EPA Residential Radon Survey, individual state radon surveys, and other sources were compared with bedrock and surficial geology, aerial radiometric data, soil properties, and soil and water radon studies. A numeric Radon Index and Confidence Index have been developed as part of this project to quantitatively describe geologic radon potential on a regional scale in a consistent manner across the country (1). Ten regional reports describing the geologic radon potential of the United States give detailed evaluations of the geologic radon potential of each state.

This paper focuses on the geologic radon potential of the upper Midwestern states that are covered mainly by Pleistocene-age continental glacial deposits. Glacial deposits constitute a challenging geologic setting for which to determine radon potential because the glacial deposits underlying any given area may reflect local source rocks, i.e., the underlying and nearby bedrock, or source rocks from many miles away, for example, cobbles of Canadian Shield crystalline rocks found in northern Kansas. Because of the crushing and grinding action of the glaciers, glacial deposits commonly have higher permeability and sometimes have higher radon emanation coefficients than their bedrock sources. This is not to say, however, that all glacial deposits have high permeability, or that it is the relatively high permeability of glacial deposits that causes many homes built on them to have elevated indoor radon concentrations. In fact, a large proportion of homes (40-70 percent) with elevated (> 4 pCi/L) indoor radon levels, and a significant proportion of homes (5-10 percent) with high (> 20 pCi/L) indoor radon levels occur on clay-rich tills with low permeability in parts of North and South Dakota, Minnesota, and Iowa. Conversely, relatively few homes in Michigan (about 14 percent of those sampled) had indoor radon levels exceeding 4 pCi/L, although most of Michigan's soils fall in the moderate to high permeability category. A brief discussion of the main factors involved and their use in the USGS's geologic radon potential assessments follows.

METHODS AND DATA USED

The assessment of geologic radon potential was made using five main types of data: 1) geologic (lithologic), 2) aerial radiometric, 3) soil characteristics, including soil moisture and permeability, 4) indoor radon data, and 5) building architecture (specifically, whether homes in each area are predominantly slab-on-grade or crawl space construction, as opposed to homes with basements). These elements were examined and integrated to produce estimates of radon potential.

The geologic factor is based on the type and distribution of lithologic units and other geologic features in each area being assessed. Rock types with naturally high uranium concentrations (considered to be greater than 2.5 parts per million (ppm) for the purpose of this assessment) that are most likely to cause indoor radon problems include carbonaceous black shales, glauconite-bearing sandstones, some fluvial sandstones, phosphorites and phosphatic sediments, chalk, some carbonates, some glacial deposits, bauxite, lignite, some coals, uranium-bearing granites and pegmatites, metamorphic rocks of granitic composition, felsic and alkalic volcanoclastic and pyroclastic volcanic rocks, syenites and carbonatites, and many sheared or faulted rocks. Rock types least likely to cause radon problems include marine quartz sands, non-carbonaceous shales and siltstones, some clays and fluvial sediments, metamorphic and igneous rocks of mafic composition, and mafic volcanic rocks (2). Exceptions exist within these general lithologic groups because of the occurrence of localized uranium deposits such as roll-front deposits.

Although many cases of extreme indoor radon levels can be traced to high radium and (or) uranium concentrations in bedrock and sediments, some structural features, most notably faults and shear zones, have been identified as sites of localized uranium concentrations and have been associated with some of the highest known indoor radon levels. Two of the highest known indoor radon occurrences in the United States are associated with sheared fault zones near Boyertown, Pennsylvania (3, 4), and in Clinton, New Jersey (5, 6).

Aerial radiometric data provide an estimate of the surficial concentrations of radon parent materials (uranium, radium) in rocks and soils. Equivalent uranium (eU) is calculated from the counts received by a gamma-ray detector in the energy window corresponding to emissions from bismuth-214, with the assumption that uranium and its decay products are in secular equilibrium. It is expressed in units of parts per million (ppm) of equivalent uranium. Gamma radioactivity may also be expressed in terms of a radium concentration; 3 ppm eU corresponds to approximately 1 picocurie per gram (pCi/g) of radium-226. The aerial radiometric data used for assessing radon potential in this study were collected as part of the National Uranium Resource Evaluation (NURE) program of the 1970s and early 1980s (7). The NURE aerial radiometric data were collected by aircraft in which a gamma-ray spectrometer was mounted, flying approximately 122 m (400 ft) above the ground surface. Smoothing, filtering, recalibrating, and matching of adjacent quadrangle data sets were performed to compensate for background, altitude, calibration, and other types of errors and inconsistencies in the original data set. The corrected data were then used to produce a contour map of eU values for the conterminous United States (8).

Soil surveys prepared by the U.S. Soil Conservation Service (SCS) and by state Cooperative Agricultural Extension Services provide data on soil characteristics. The reports are commonly available in county formats and state summaries, and usually contain both generalized and relatively detailed maps of soils in the area. For state or regional-scale radon characterizations, soil maps are compared to geologic maps of the area, and the textural descriptions, shrink-swell potential, depth to seasonal high water table, permeability, and other relevant characteristics of each soil group noted.

Three major sources of indoor radon data were used. All data reflect 2-7 day screening measurements collected with charcoal canister detectors. The first and largest source of data is from the State/EPA Residential Radon Surveys (9, 10). During the period 1986-1992, 42 states completed EPA-sponsored radon surveys, with a total of more than 59,000 participants. The State/EPA Residential Radon Surveys were designed to be comprehensive and statistically significant at the state level, and were subjected to high levels of quality assurance and control. The surveys collected screening measurements in the lowest livable areas of the home. The target population for the surveys included owner-occupied, single family, detached housing units (11). Participants were selected randomly from telephone-directory listings.

The second source of indoor radon data comes from independent state surveys or utility company surveys. Examples include indoor radon surveys conducted by the states of Delaware, Florida, Oregon, New Hampshire, New Jersey, New York, and Utah, and a survey of radon in homes of the Pacific Northwest by the Bonneville Power Administration. The third source of indoor radon data comes from The Radon Project, formerly of the University of Pittsburgh (12, 13). This ongoing effort has been collecting indoor radon data since 1985, and to date has accumulated approximately 175,000 indoor radon measurements for the entire country. The data are collected from homeowners that purchase radon detectors from The Radon Project on a voluntary basis. The data are therefore non-random, although a limited bias reduction technique has been applied to the data set by eliminating certain values based on responses to a standard questionnaire (13).

[Back][Home] [Next]


URL: http://energy.cr.usgs.gov/radon/midwest2.html
Server information: webman@energy.cr.usgs.gov
13 October 1995