Lead-based paint testing technologies: summary of an EPA/HUD field study.
Author: Schmehl, R. L.; Cox, D. C. Dewalt, F. G. Source: American Industrial Hygiene Association Journal v. 60 no4 (July/Aug. 1999) p. 444-51 ISSN: 0002-8894 Number: BAST99049841 Copyright: The magazine publisher is the copyright holder of this article and it is reproduced with permission. Further reproduction of this article in violation of the copyright is prohibited.
Keywords: lead-based paint, chemical test kits, X-ray fluorescence.
The U.S. Centers for Disease Control and Prevention (CDC) has recognized lead poisoning as one of the most common and devastating environmental diseases of young children.(FN1) Lead-based paint (LBP) in older housing, especially LBP in poor condition, is often implicated either directly or indirectly in elevated blood lead levels of children younger than 6 years of age. Exposure to lead in paint can come from paint chips, from dust caused by abrasion of paint on friction or impact surfaces, and from the deterioration of interior or exterior painted surfaces. Reflecting an increasing awareness of the public health threat posed by LBP, the federal government first enacted the Lead-Based Paint Poisoning Prevention Act of 1971,(FN2) which was later amended by enactment of the Housing and Community Development Act of 1987. In October 1992 the Residential Lead-Based Paint Hazard Reduction Act(FN3) became law and became the most comprehensive federal legislation regarding lead-based paint and lead-based paint hazards.
To meet the requirements of Section 1017 of the Residential Lead-Based Paint Hazard Reduction Act, the U.S. Department of Housing and Urban Development (HUD), with substantial input from U.S. Environmental Protection Agency (EPA), prepared and issued its guidelines(FN4) for the conduct of federally supported work involving LBP in housing. To support the development of the LBP inspection chapter of the guidelines and update the lead testing fact sheets available from the National Lead Information Center Clearinghouse, EPA and HUD sponsored a field evaluation study of the available lead-testing technologies.(FN5,6) Chemical test kits and X-ray fluorescence instruments, which were the portable lead-testing technologies commonly in use at the time of the study, were of particular interest because of the advantages of speed, cost, and minimal destructiveness that they offered relative to laboratory analysis. This article describes the design of the study and summarizes its major findings.
PAINT TESTING TECHNOLOGIES
It was acknowledged that the most accurate method for measuring the amount of lead in paint was to remove a sample of paint from the surface and subject it to laboratory analysis. In the laboratory the sample is pulverized and dissolved in an acid solution before its lead content is assayed using any of a number of complex instruments. This process requires considerable time for sample collection, transport, and laboratory turnaround before a lead measurement is obtained. Laboratory measurement is also expensive: Paint sample collection requires skill, which increases associated labor costs, and the per-sample cost of laboratory analysis alone is substantial ($5 to $25 in 1997 U.S. dollars). The laboratory cost varies with the laboratory, the analysis turnaround time, and the number of samples submitted, with fewer samples commanding a higher unit price. Damage from sampling for laboratory analysis is an unavoidable consequence, making the repair of the damaged surfaces an additional cost.
To overcome the disadvantages of laboratory measurement, and especially to reduce the cost of performing LBP testing, the lead detection industry developed several portable (nonlaboratory) technologies to test for lead in paint. At the start of the EPA/HUD field study in 1993 the two most commonly used portable technologies for LBP testing were chemical test kits and portable X-ray fluorescence (XRF) instruments.
Chemical test kits indicate the presence of lead in paint with a color change that occurs when chemicals in the kit react with lead, which is referred to as a positive classification to lead. A measurement obtained with a test kit is, therefore, not quantitative, but classified either positive or negative, depending on whether the prescribed color change was observed. At the time of the study several brands of chemical test kits were marketed to homeowners and other nonprofessional lead testers, and at least one chemical test kit was developed for, and only made available to, certified professionals. The cost per test of the commercial test kits ranged from approximately $1 to $5 (1997 U.S. dollars). In addition, no formal technical training was required of users of the commercial test kits, there was less destruction of painted surfaces than from paint sampling for laboratory measurements, and test results were quickly obtained.
The other portable technology is a form of X-ray spectrometry--a method for identifying elements by detecting characteristic energy emission. The first reported use of X-ray spectrometry was in 1912, when an electron beam system was used to excite atoms that caused the characteristic energy to be emitted.(FN7) In 1925 the methodology was refined by bombarding specimens with X-ray photons instead of electrons, which began the emergence of XRF spectrometry.(FN8) Laboratory XRF instruments became commercially available in the early 1950s(FN9) and the portable XRF instruments, designed for measuring lead in paint, came into use much more recently. In addition to measuring lead in paint, field portable XRF instruments have been used for identifying elements in a variety of substances such as soil, water, and steel.
In general, XRF instruments direct high-energy gamma and X-rays into a surface. These high-energy rays strike atoms in the surface, causing electrons to be ejected from their orbits, or shells. Characteristic X-ray energy is emitted when another electron fills the void in the shell. The emitted energy is detected by the XRF instrument and converted to a quantitative measure. For the lead atom, characteristic frequencies are emitted from the K- and L-shells, its two innermost electron orbits. Energy emitted from these shells (energy bands) will be referred to as K X-rays and L X-rays, respectively.
The portable XRF instruments that were evaluated in the EPA/HUD field study were generically referred to as either direct readers or spectrum analyzers. These designations are related to what can be displayed by the instrument. Direct readers display only the lead content values, whereas the spectrum analyzers are capable of displaying an energy spectrum and lead content values. Another distinguishing characteristic of these instruments is the energy band from which a quantitative measure is made: the K X-rays or L X-rays. K X-rays, being of higher energy than L X-rays, are attenuated less by paint than L X-ray energy. For this reason, K X-ray measurements are generally more useful for detecting lead in deeper layers of paint. The XRF instruments in the field study made K X-ray measurements, L X-ray measurements, or both.
The detection of lead in paint is complicated by the material underlying the paint, the substrate, and how it affects the high-energy rays directed into the paint by the XRF instrument. Some of the gamma and high-energy X-rays from the radiation source that are directed into the paint will hit the underlying substrate and "bounce" back to the instrument's detector. This phenomenon, referred to as backscatter (or Compton effect) interferes with the detection of the energy emitted from the lead atom. Thus, the XRF instrument must account for this backscatter to determine accurately the amount of lead in the paint. The substrate effect on the backscatter--some substrates are more dense while others are more absorbing--will be referred to as substrate bias. The portable XRF instruments in the field study used different methods to account for substrate bias with differing degrees of success, since substrate materials such as brick, concrete, drywall, metal, plaster, and wood can have widely varying properties. A testing method to further account for substrate bias is called substrate correction and is discussed below.
Unlike chemical test kits, XRF measurements are quantitative and are expressed in milligrams of lead per square centimeter (mg/cm2). Due to the potential radiation hazard and instrument cost, varying from $7000 to $45,000 (1997 U.S. dollars), XRF instruments are intended for use by professional testers only. However, the XRF instruments evaluated in the study usually returned a measurement within a minute and without damage to the painted surface. This speed of measurement allows extensive testing to be performed on a given day and makes XRF testing an affordable alternative to laboratory analysis for residential lead testing.
DESIGN OF THE FIELD STUDY
The purpose of the EPA/HUD field study was to evaluate portable lead-testing technologies under real-world conditions. Six chemical test kit brands and six XRF instrument models were chosen for evaluation. To perform a controlled evaluation, all 12 products were tested on a common set of test locations chosen from painted building components. Test locations were selected to reflect the range of conditions under which LBP was found in residential structures, and in sufficient quantity for accurate estimates to be derived. A pilot study, conducted in March and April 1993, was based on 100 test locations chosen from multifamily public housing units in Louisville, Ken. The "full" study, conducted from July to October 1993, was based on 750 test locations chosen from single-family public housing units in Denver, Colo., and 440 test locations chosen from multifamily public housing units in Philadelphia, Pa.
Five of the six chemical test kits evaluated in the EPA/HUD field study were commercially available, representing each different type of test kit known to be marketed in the spring of 1993. The other, referred to as the State sodium sulfide kit, was available only to professional testers certified by the State of Massachusetts. Table I identifies the chemical test kits that were evaluated in the field study.
The evaluated XRF instrument models are identified in Table II. Two of these, the Pb Analyzer (TN Technologies, Inc., Round Rock, Tex.) and the XL (Niton Corp., Bedford, Mass.), were prototypes that were not available to professional testers at the time the study began. Three of the four commercially available instruments were widely used in testing for LBP at the time of the study, the exception being the X-MET 880 (Outokumpu Electronics, Bend, Ore.), which was more typically used for lead testing in soil. The MAP-3 (Scitec Corp., Kennewick, Wash.) and Pb Analyzer provided both K X-ray and L X-ray measurements. The Microlead I (Warrington, Inc., Austin, Tex,) and XK-3 (Princeton Gamma-Tech, Inc., Princeton, N.J.) provided only K X-ray measurements, and the XL and the X-MET 880 provided only L X-ray measurements. The MAP-3, Pb Analyzer, XL, and X-MET 880 are spectrum analyzers. The Microlead I and XK-3 are direct readers.
Development of a design for the EPA/HUD field study was complicated by two major factors:.
(1) the large number of measurements to be made at each test location, representing all chemical test kits and XRF instruments, plus the collection of at least one paint sample for laboratory analysis; and.
(2) the destructiveness of paint sample collections and chemical test kit applications to the painted surfaces.
Because multiple destructive testing could not be conducted at exactly the same place, subareas of each test location were allocated to test kits, XRF instruments, and paint sampling in a consistent manner to ensure comparability of results. This was done by using a standardized format or template that was demarked on each test location. The "standard" templates used in the pilot and full studies are shown at the top of Figure 1, with several variations shown below them. A template was applied horizontally if possible, or else rotated, as necessary, to fit on a painted building component. In a few cases, the template needed to be altered, as shown, to fit. A paint sample was collected from one of the areas marked with a "P" that was randomly selected. Randomization was also used to allocate testing areas marked with a "T" to the six chemical test kits evaluated in the study. The area marked with an "X" was allocated to XRF testing.
TESTING PROCEDURES
TEST LOCATION IDENTIFICATION
Prior to the start of field testing, test locations were selected from each of the housing units used in the study. For each test location a template was drawn on the painted surface and assigned a unique number.
CHEMICAL TEST KIT MEASUREMENT
After all test locations were identified and marked with templates in a housing unit, the chemical test kit operators began to collect measurements. With one exception, the test kit operators did not have specific scientific background or prior training, but were chosen to represent typical homeowners who might purchase a test kit for personal use. The exception was the operator of the State sodium sulfide kit, who was a Massachusetts state-certified inspector. The nonprofessional testers were screened to ensure that they were able to detect the color changes that signified a positive indication of lead. The nonprofessional chemical test kits were rotated among the nonprofessional testers during the course of the study to make it difficult for a tester to become more proficient with any one test kit. After each tester had completed testing at a test location, the used area of the template was covered to prevent subsequent testers from observing the result.
PAINT SAMPLE COLLECTION AND LABORATORY ANALYSIS
For each test location, paint samples were collected for laboratory analysis. The paint samples were analyzed by inductively coupled plasma atomic emission spectroscopy. After the paint samples were collected, the unused P areas of the templates were also scraped to expose the substrate. This created a bare area in each template that had the same dimensions as the X area. This bare area became one of the XRF instrument testing areas.
XRF INSTRUMENT TESTING
XRF instrument testing began after paint sampling was completed in a housing unit. The XRF instrument operators were trained and licensed employees of lead testing companies. XRF instrument testing was conducted at the designated X area and at the bare area of each template. Measurements on the bare areas were made both directly on the bare substrate and on National Institute of Standards and Technology (NIST) standard reference material (SRM) films placed over the bare areas. The NIST SRMs used for this purpose had a lead level of 1.02 mg/cm2. The measurements taken on a NIST SRM film (with a 1.02 mg/cm2 lead level) placed over bared areas were obtained to examine the performance of substrate correction, a procedure used with some XRF instruments to reduce the bias associated with the substrate underlying the paint.
XRF testing was based on a nominal 15-second reading time for a full-strength radioactive source. Actual reading times were adjusted either automatically by the instrument or manually by the operator to account for the decay of the radioactive source in the instrument due to its age. Accordingly, the actual reading time would necessarily increase as the source decays.
EVALUATION OF TESTING RESULTS
For chemical test kits evaluated in the EPA/HUD field study, classification accuracy was the primary criterion used to measure performance. XRF instruments were evaluated by both classification accuracy and bias and precision estimates. Classification accuracy was measured relative to the federal standards of 1.0 mg/cm and 0.5% lead by weight for defining LBP, as designated by Title X. Many state and local governments also have adopted these standards.
Classification accuracy was reported in terms of misclassified results called false negative and false positive results. Relative to a given standard, a lead test result was classified as a false negative if the laboratory measurement was equal to or greater than the standard (positive) and, for chemical test kits, a color change was not observed (negative). Similarly, for XRF instruments a lead measurement was classified as a false negative if the laboratory measurement was equal to or greater than the standard (positive) and the XRF measurement was less than the standard (negative). In other words, a false negative result is one in which the laboratory measurement was classified positive but tested negative according to a chemical test kit or XRF instrument result for the same test sample. A lead test result was classified as false positive if the laboratory measurement was less than the standard (negative) and, for chemical test kits, a color change was observed (positive). For XRF instruments a lead test result was classified as a false positive if the laboratory measurement was less than the standard (negative) and the XRF measurement was equal to or greater than the standard (positive). In other words, a false positive result is one in which the laboratory measurement was classified negative but tested positive according to a chemical test kit or XRF instrument result for the same sample. None of the products evaluated in the study achieved a perfect performance, i.e., no false negatives or false positives with respect to either federal standard.