Basic XRF Field Applications
Today, most site investigations and remedial actions make use of mobile laboratories because of the need for fast and accurate results. The EPA has been promoting a flexible approach to site cleanup that recognizes site-specific decisions and data needs. This approach, known as TRIAD, has been designed to reduce costs, improve decision certainty, and expedite site closeout.
Some of the most successful TRIAD case histories involve the use of X-ray Fluorescence (XRF) screening for metals. Prior to the advent of field-worthy XRF systems, site investigation and remediation for metal contaminants proceeded in a more classical manner of waiting for fixed laboratory results or paying a premium for rush results.
In some cases, fixed laboratory methods were performed in a mobile laboratory setting with good results. This practice involved the use of concentrated acids and time consuming digestion steps and was useful for only a limited number of elements. This is no longer a problem with the availability of field portable and transportable Energy Dispersive X-ray Fluorescence Spectrometers (EDXRF).
Reporting limits are element, instrument, and sample matrix dependent. Each dependent variable contributes to the final achievable reporting limit. Once the suite of elements and instrument conditions are modeled, the only variable left is the sample matrix which can vary from site to site, or even within a particular site.
XRF techniques are not very useful for light elements (elements between atomic numbers 11-18). The remaining elements, those between atomic numbers 19 and 92, are candidates for this technique.
Reporting limits are also instrument dependent and can vary between manufacturers for a variety of reasons.
For radioisotope sources, the following factors impact the reporting limit: isotope used (Fe-55, Cd-109, Am-241, Cm-244); half life of the isotope; and the amount of isotope in millicuries. For X-ray tube sources, the energy level of tube and the current level of tube both impact the reporting limit. Both radioisotope and X-ray tube sources are affected by count time during analysis; detector type (gas filled, or Si(Li); grain size of sample; and moisture of sample.
Applications for EDXRF are varied and include lead investigations at battery recycling businesses, remedial investigations of military and public shooting ranges, munitions demolition areas, weapons manufacturing facilities, copper ore assay from mine tailing piles, and many industrial and household remedial actions for leaded paint removal.
As always, the benefit of a field analytical method is in providing near real time data so that time-critical decisions can be made in the field. This concept works best when the target analytes are limited to the main chemicals of concern for a particular site. The following application examples are from sites that had a limited list of target analytes so that the XRF method could be customized to maximize sample throughput and sensitivity.
Example 1. Lead Remediation at a Landfill Burn Site
This project had several time critical aspects: the rainy season was approaching, and there was a narrow window of time to complete all of the site work before a protected species of bird returned to the area. The project involved removal of 20,000 tons of lead-contaminated soil and ash from a burn pit operation at an abandoned land fill. The lead had contaminated the entire area surrounding the landfill, which bordered on an elementary school and a public lake. The required reporting limit for lead was 15 mg/kg, and the sample throughput was to be 40 samples per day. Confirmation samples were taken initially at a 10 percent frequency, and dropped to a five percent frequency once the correlation was determined.
A transportable XRF unit was set up on location in a mobile laboratory, complete with fume hood, generator power, and water. All samples were prepared by sieving, drying, and grinding prior to analysis. The calibration model used was Fundamental Parameters and results were provided at a rate of four samples per hour. The data correlated at R=0.988 (see North Chollas for data spanning nearly three orders of magnitude). The project was completed within the allotted time frame and budget.
The XRF approach was approved by city and state agencies, with the stipulation that a correlation be determined at the start of the project and confirmation samples be split throughout the project, initially at a 10 percent frequency, then decreasing to no less than five percent.
Example 2. Lead Remediation at an Active Battery Recycling Shop
This project was a State Superfund cleanup of lead-contaminated soil from a battery recycling shop located in a residential area. The initial site study called for removal of 1,315 cubic yards of soil for transport and disposal at a Class I landfill at an approximate cost of $620,000. By using the EDXRF system, the volume of soil actually sent was reduced to 666 cubic yards, which saved $306,000. Other critical aspects of the project involved minimization of exposure risks to nearby residents and site workers. By using the rapid turnaround results from the EDXRF system, the site was cleaned up to an action level of 220 mg/kg with correlation to EPA method 6010 of R=0.989.
The State of California, under the State Superfund program, sanctioned this project. The XRF approach was initially approved with the stipulation that site-specific calibration be performed. The samples were quantified by using Fundamental Parameters software, and since the correlation was >0.900, the site-specific calibration was not needed.
Example 3. Multiple Elements at Navy Dump Site
This project was a technology demonstration of both the EDXRF and a soil washing technique at a known contaminated site, Hunter’s Point. This site was adjacent to a densely populated area. It had been used by the Navy and private industry for a variety of activities, including transformer and battery storage and vehicle maintenance activities. The site was moderately contaminated with lead concentrations above non-residential clean up levels. Other elements of concern were mercury, antimony, copper, zinc, and chromium. During the two-week project, the EDXRF system was used to help optimize the soil washing system prior to the actual demonstration. The demonstration yielded 80 washed samples of which 23 were sent for confirmation.
The correlation for antimony was rather poor and was a function of the solubility of antimony in the acid digestion used in EPA Method 6010. Aside from antimony, all correlation coefficients were >0.990. Mercury was not detected in any of the samples, nor by the confirmation laboratory at a level >10 mg/kg. This demonstration helped confirm the findings from an earlier study published by the California Military Environmental Coordination Committee that a correlation coefficient of 0.9 or greater indicates that the field XRF data may be considered definitive (i.e. equivalent in data quality to CLP methods).
XRF techniques are rapidly gaining acceptance as a viable field solution to rapid metal analysis in a field setting. XRF is no longer considered an emerging technique and is in widespread use by state and federal agencies, as well as by environmental professionals throughout the country. Whether portable or transportable, the EDXRF technique is proving its worth as a rapid, accurate, and cost-effective tool for site investigation and remedial activities.
- USEPA Field Analytic Technologies Encyclopedia (FATE), Online resource: http://fate.clu-in.org. Last updated, January, 2003.
USEPA TRIAD Report, Online resource: http://www.epa.gov/tio/triad. Last updated November, 2005.
- TN Spectrace, Spectrace Instruments EDXRF Users School Manual, Spectrace Instruments, Fort Collins, Colorado.
Merck & Co., Inc., The Merck Index, Eleventh Edition, Table of Radioactive Isotopes, Merck & Co., Inc., Rahway, New Jersey, 1989.
Bertin, E.P., Introduction to X-Ray Spectrometric Analysis, Plenum Publishing, New York, 1978.
Panuscka, Barber, Smith, Mobile EDXRF Screening at Lead Contaminated Soil Removal Project, Conference Proceedings, Hazwaste World Superfund XVII, October, 1996, Washington D.C.
CMECC, Field Analytical Measurement Technologies, Applications, and Selection, April 1996.
Barber, M, Application Analysis Report, Onsite Environmental Laboratories, Bay Area Conversion Action Team (BADCAT).