Quantification of Vapor Intrusion Pathways into a Slab-on-Ground

Jan 6, 2009 - Vapor discharge from the uncovered open ground soil adjacent to the building and subsequent advection into the building was unlikely due...
0 downloads 0 Views 1MB Size
Environ. Sci. Technol. 2009, 43, 650–656

Quantification of Vapor Intrusion Pathways into a Slab-on-Ground Building under Varying Environmental Conditions BRADLEY M. PATTERSON* AND GREG B. DAVIS CSIRO Land and Water, Private Bag No. 5, Wembley WA 6913, Australia

Received May 15, 2008. Revised manuscript received November 3, 2008. Accepted November 21, 2008.

Potential hydrocarbon-vapor intrusion pathways into a building through a concrete slab-on-ground were investigated and quantified under a variety of environmental conditions to elucidate the potential mechanisms for indoor air contamination. Vapor discharge from the uncovered open ground soil adjacent to the building and subsequent advection into the building was unlikely due to the low soil-gas concentrations at the edge of the building as a result of aerobic biodegradation of hydrocarbon vapors. When the building’s interior was under ambient pressure, a flux of vapors into the building due to molecular diffusion of vapors through the building’s concrete slab (cyclohexane 11 and methylcyclohexane 31 mg m-2 concrete slab day-1) and short-term (up to 8 h) cyclical pressure-driven advection of vapors through an artificial crack (cyclohexane 4.2 × 103 and methylcyclohexane 1.2 × 104 mg m-2 cracks day-1) was observed. The average subslab vapor concentration under the center of the building was 25,000 µg L-1.Basedonthemeasuredbuilding’sinteriorvaporconcentrations and the building’s air exchange rate of 0.66 h-1, diffusion of vapors through the concrete slab was the dominant vapor intrusion pathway and cyclical pressure exchanges resulted in a near zero advective flux. When the building’s interior was under a reduced pressure (-12 Pa), advective transport through cracks or gaps in the concrete slab (cyclohexane 340 and methylcyclohexane 1100 mg m-2 cracks day-1) was the dominant vapor intrusion pathway.

Introduction Hydrocarbon-vapor intrusion into buildings can be a significant driver of health risk and the level of site remediation required (1). Also, the understanding of the various vapor intrusion pathways can guide field assessment procedures to better quantify vapor intrusion into buildings. While a number of studies have investigated vapors in open ground conditions (e.g., (2)), there are few (1, 3) detailed field studies on vapor intrusion into built structures. For a building with a concrete slab-on-ground, the pathways for vapor intrusion include (i) discharge of vapors from the uncovered open ground soil adjacent to the building and subsequent advection into the building (3), (ii) molecular diffusion of vapors through the building’s concrete slab (3, 4), and (iii) pressure* Corresponding author phone: +61-8-93336276; fax: +61-8-9333 6211; e-mail: [email protected]. 650 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 3, 2009

driven advection of vapors through cracks or gaps in the building’s substructure (5, 6). These cracks or gaps can result from utility conduits entering a building or substructure macropores that could provide preferential vapor migration pathways. Also the building’s construction technique can results in cracks or gaps between the walls and the concrete slab. Subslab to in-building pressure differentials may be caused by (i) short-term barometric pressure changes as a result of regular oscillations in atmospheric winds and pressure fields often referred to as atmospheric tides (7) or tidal-induced water table fluctuations in coastal areas (8), (ii) longer-term meteorological induced barometric pressure changes (9, 10), (iii) rainfall events (8), (iv) thermal differences between indoors and outdoors (6), (v) wind loading on the buildings superstructure (6, 9), and (vi) imbalanced building ventilation (6). Modeling of vapor intrusion processes (e.g., (11)) can illustrate the kinds of patterns in subsurface vapor and oxygen concentrations and indoor air concentrations that can arise under different environmental site conditions. However, detailed experimental quantification and comparison of the different vapor intrusion flux pathways into a slab-on-ground building has not been undertaken. The current study provides otherwise unavailable data and constraints for validation of vapor intrusion models. In this paper, we present the results of laboratory and field research to assess and quantify the roles of different vapor intrusion pathways from a hydrocarbon-contaminated sandy vadose zone to the base of and into a slab-on-ground building, to identify the most dominant vapor intrusion pathways under different building and meteorological conditions. Site Description. The field research was conducted within and around an abandoned building located 50 km south of Perth, Western Australia. The building was of double brick construction with dimensions 18 m × 13.5 m, with a 19-cm thick concrete slab-on-ground, and a 10-m wide concrete apron on three sides. The fourth side of the building (front of the building) was adjacent to an uncovered open ground. Plumbing and electrical utilities were generally located within the double brick cavity around the perimeter of the building. Visual inspection showed no obvious cracks or gaps in the concrete floor slab or between the walls and floor. This building overlaid a sandy vadose zone that was contaminated with LNAPL (light nonaqueous phase liquid) kerosene in the zone of water table fluctuation at ∼3 m below ground. Further details of the site and soil properties are given in the Supporting Information.

Material and Methods Monitoring Installations. To monitor soil gas composition, the site was instrumented with online volatile organic compound (VOC) probes (12), online oxygen probes (13) and mini piezometers (2-mm i.d. and 3-mm o.d. nylon tubes, with the ends covered with nylon mesh to avoid clogging by soil media) for manual collection of soil gas. The VOC probes operate on the diffusion of hydrocarbon vapors through a semipermeable membrane into a chamber that is intermittently purged with gas to transport the hydrocarbon vapors to a hydrocarbon specific gas sensor above ground. The VOC probes were precalibrated using cyclohexane vapor at four different concentrations to provide a multipoint calibration curve. Additionally, the site was also instrumented with temperature probes and a second set of mini piezometers for use with online pressure transducers. Online instrumen10.1021/es801334x CCC: $40.75

Published 2009 by the American Chemical Society

Published on Web 01/06/2009

q ) Deff

∂C ∂z

(1)

where Deff is an effective hydrocarbon diffusion coefficient through concrete. These calculations assume hydrocarbon consumption does not occur within the concrete. Haghighat et al. (4), reviewed diffusion coefficients of various VOCs through solid concrete and quoted Deff values for methane (4.6 × 10-7 m2 s-1), ethyl acetate (5.1 to 39 × 10-8 m2 s-1), n-octane (1.0 to 3.0 × 10-7 m2 s-1), and n-dodecane (2.4 × 10-7 m2 s-1). To provide a site-specific Deff, a laboratory diffusion experiment, similar to the one-flow method described by Meininghaus and Uhde (14), was conducted on a concrete core sample collected from the center of the building. Details of the laboratory diffusion experiment are given in the Supporting Information. For longer time periods, eq 2 can be used to calculate Deff and R (the capacity factor or accessible porosity) from the slope and intercept respectively, of the line relating the cumulative mass per unit area to time (15-17), where M is the cumulative mass, Co is the delivery chamber constant concentration, A is the cross sectional area of the concrete core, d is the thickness of the concrete core, and t is time. t Rd M ) CoA Deff d 6

(

FIGURE 1. Contour plots showing vadose zone soil gas concentrations of VOC and oxygen concentrations. The plots show a cross section view through the center line of the building from the center of the building to the uncovered open space adjacent to the building. tation was monitored every 30 min for the oxygen probes, 60 min for the temperature and pressure probes, and 8 h for the VOC probes. Monitoring instrumentation was emplaced at different depth intervals at 3 main locations: (i) the center of the building, (ii) midway between the center and edge of the building, and (iii) at the outer edge of the building. Monitoring instrumentation under the concrete slab for each depth interval was installed by first drilling a 60-cm diameter hole through the concrete slab, then hand augering to the required depth interval. Probes (VOC, oxygen, temperature) and mini piezometers were then installed in the hole and the hole was backfilled with drill spoils. The top of the hole was then sealed using a concrete slurry mixture to form a 19-cm thick concrete plug. Wires and mini piezometers protruding through the hole were vibrated during sealing to ensure good contact to the concrete and no visible gaps. Locations of the monitoring installations are shown in Figure 1. The pressure difference above and below the concrete slab was also determined using a hand-held manometer (TSI DP-Calc, model 8705) connected to the open gas sampling tube located near the center of the building, and extending 0.25 m below the concrete slab. The pressure difference inside and outside of the building was also determined using the hand-held manometer and a gas sampling tube extending to the outside of the building through a small gap under an external door. Details of the manual gas sampling from the minipiezometers and analysis are given in the Supporting Information. Vapor Flux Through a Noncracked Concrete Slab. Hydrocarbon vapor flux (q) through a concrete slab can be estimated using the standard Fickian law (eq 1):

)

(2)

The concentration gradient was based on the concrete slab thickness and field determined steady-state hydrocarbon concentration above and below the concrete slab. VOC concentrations above the concrete slab were determined from gas samples collected (using a glass syringe) within the building, while VOC vapor concentrations immediately below the concrete slab were determined from gas samples collected from the 0.15 m mini piezometer at the central location. Vapor Flux Through a Cracked Concrete Slab. To assess vapor flux through a cracked concrete slab, a 1.23-cm diameter hole was drilled through the concrete slab near the center of the building as an artificial crack. Over this hole a 7.95 L PVC flux vented hood was cemented in place to monitor vapor emissions from the artificial crack and enable pressure equilibration (18, 19). Further details of the flux hood are given in the Supporting Information. Vapor flux through the artificial crack was calculated using eq 3, where CFinitial is the initial vapor concentration within the flux hood prior to vapor intrusion, CFfinal is the final vapor concentration within the flux hood prior to vapor intrusion, VF is the volume of gas within the flux hood, tVI is the time of vapor intrusion into the flux hood, and Acrack is the surface area of the drilled hole. qcrack )

(CFfinal - CFinitial) ⁄ VF Acrack . tVI

(3)

Building’s Air Ventilation Rate. A number of sulfur hexafluoride (SF6) tracer tests were conducted to determine the building’s air ventilation rate, similar to the method describe by Said (20).Two initial tracer tests (fully sealed and partly sealed) were conducted under the ambient building’s pressure conditions, and a third tracer test was conducted under a reduced pressure. Details of the air ventilation tracer tests are given in the Supporting Information. The building’s gas exchange rate was then calculated (eq 4), where Aex is the building’s air ventilation rate (exchanges per time), C is the tracer gas concentration at time t, and Co is the tracer gas concentration at time to. Aex )

Ln(C ⁄ Co) -(t - to)

(4)

Predicted Building VOC Concentration. A predicted building VOC concentration, assuming diffusion through the VOL. 43, NO. 3, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

651

concrete slab was the only vapor intrusion pathway, could be determined (eq 5) based on the mass balance calculation of VOC flux into the building from the vadose zone through the concrete slab, and VOC flux out of the building through ventilation. For eq 5, CB is the VOC concentration within the building. Aslab is the effective building concrete slab floor area in contact with VOC contaminated vadose zone; this was assumed to be the 110 m2 back half of the building (accounting for internal walls) with the 10 m wide concrete apron at the back and sides of the building. qslab is the flux of VOC through the concrete slab, and VB is the volume of air within the building (760 m3). CB )

Aslab . qslab Aex . VB

(5)

Results and Discussion Vadose Zone VOC and Oxygen Distribution. VOC concentrations were up to 47,000 µg L-1 in the zone of NAPL contamination with the BTEX (benzene, toluene, ethylbenzene, and xylene isomers) range of compounds accounted for