Use of SF 6 to label drilling air in unsaturated, fractured rock studies: risk of over-purging Gregg R. Davidson* Department of Geology and Geological Engineering, University of Mississippi, University, MS 38677, USA Received 10 December 2000; accepted 12 November 2001 Editorial handling by R.L. Bassett Abstract Sulfur hexafluoride was injected as a tracer gas into the air stream during air-drilling of a borehole in the unsaturated zone of a partially-welded, fractured tuff at Apache Leap, Arizona. One-meter intervals were later isolated at multiple depths and pumped to purge drilling air from each interval. The volume of air purged (at 1 atm, 20  C), ranged from a low of 0.3 m 3 in an unfractured interval, to a high of 252 m 3 in a highly fractured interval. The concentration of SF 6 remained high throughout the purge in all intervals and often increased over time. Measurements of d 13 C, 14 C and CO 2 concentration indicated that atmospheric air was eventually drawn into several of the intervals in spite of the fact that SF 6 concentrations remained high. Possible explanations include mixing of atmospheric air drawn through fractures with partially-purged matrix air, and delayed removal of SF 6 relative to atmospheric gases due to adsorption of SF 6 within the tuff matrix, dissolution into pore water, or diffusion from dead-end pores with restricted openings. In this system, following a long delay between drilling and purging, it was found that the risk of contamination from surface air by over-purging was substantially greater than the risk of contamination from residual drilling air by under-pur- ging. # 2002 Elsevier Science Ltd. All rights reserved. 1. Introduction Subsurface investigations that require drilling of sampling wells or boreholes inevitably introduce foreign gases or solutions into the system. These fluids must be purged from the system in order to obtain air or water samples that represent in situ conditions. Introduction of a tracer into the drilling fluid during placement of a borehole provides one mechanism for quantitatively monitoring a subsequent purge. Ideally, water or air is purged from the system until the tracer drops below detection. Continued pumping should then draw uncontaminated formation fluids. In this study, a bore- hole was completed in a fractured tuff using SF 6 -label- led atmospheric air as the drilling fluid. During the subsequent purge of air from isolated intervals within the borehole, evidence was obtained that indicates that reliance on a reputedly conservative tracer can still result in over-purging. Sulfur hexafluoride has become a fairly common tra- cer gas in subsurface studies, with applications for tracking flow in both the saturated and unsaturated zone. The growing use of SF 6 as a tracer is derived from its non-toxic nature (Lester and Greenberg, 1950), its low detection limits using standard gas chromatography (e.g. Wanninkhof et al., 1991), and the perception that it does not significantly adsorb to solid substrates (e.g. Watson et al., 1991; Wilson and Mackay, 1993, 1996; Olschewski et al., 1995). It is worth noting that the toxicity study by Lester and Greenberg (1950), invari- ably cited in SF 6 -related papers, was very limited in scope. Rats were exposed to an atmosphere of up to 80% SF 6 for up to 1 day, with no deleterious effects observed during an undisclosed period of observation. Applications of SF 6 in saturated environments typi- cally have not been done to monitor purging, but have 0883-2927/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved. PII: S0883-2927(02)00044-6 Applied Geochemistry 17 (2002) 1361–1370 www.elsevier.com/locate/apgeochem * Tel.: +1-601-232-5824; fax: +1-601-232-5998. E-mail address: davidson@olemiss.edu (G.R. Davidson).