Effect of biogas generation on radon emissions from landfills receiving radium-bearing waste from shale gas development

Abstract

Dramatic increases in the development of oil and natural gas from shale formations will result in large quantities of drill cuttings, flowback water, and produced water. These organic-rich shale gas formations often contain elevated concentrations of naturally occurring radioactive materials (NORM), such as uranium, thorium, and radium. Production of oil and gas from these formations will also lead to the development of technologically enhanced NORM (TENORM) in production equipment. Disposal of these potentially radium-bearing materials in municipal solid waste (MSW) landfills could release radon to the atmosphere. Risk analyses of disposal of radium-bearing TENORM in MSW landfills sponsored by the Department of Energy did not consider the effect of landfill gas (LFG) generation or LFG control systems on radon emissions. Simulation of radon emissions from landfills with LFG generation indicates that LFG generation can significantly increase radon emissions relative to emissions without LFG generation, where the radon emissions are largely controlled by vapor-phase diffusion. Although the operation of LFG control systems at landfills with radon source materials can result in point-source atmospheric radon plumes, the LFG control systems tend to reduce overall radon emissions by reducing advective gas flow through the landfill surface, and increasing the radon residence time in the subsurface, thus allowing more time for radon to decay. In some of the disposal scenarios considered, the radon flux from the landfill and off-site atmospheric activities exceed levels that would be allowed for radon emissions from uranium mill tailings.

Implications:

Increased development of hydrocarbons from organic-rich shale formations has raised public concern that wastes from these activities containing naturally occurring radioactive materials, particularly radium, may be disposed in municipal solid waste landfills and endanger public health by releasing radon to the atmosphere. This paper analyses the processes by which radon may be emitted from a landfill to the atmosphere. The analyses indicate that landfill gas generation can significantly increase radon emissions, but that the actual level of radon emissions depend on the place of the waste, construction of the landfill cover, and nature of the landfill gas control system.

Introduction

The dramatic increase in development of shale formations containing natural gas (shale gas) and other hydrocarbons through hydraulic fracturing has the potential to generate large quantities of drill cuttings, flowback water, and produced water. These organic-rich shale gas formations often contain elevated concentrations of naturally occurring radioactive materials (NORM), such as uranium, thorium, and radium (Kargbo et al., 2010; Ortiz and Anthony, 1993; Rowan et al., 2011). Drill cuttings may be disposed in on-site pits or sent to solid waste landfills. Flowback and produced water may be treated on-site, used for dust control on roads, or sent to publically owned waste water treatment facilities. NORM in water sent to waste water treatment plants may accumulate in treatment plant sludge, which is then land-farmed or sent to solid waste landfills (Walter, 2007). Longer-term, radium and other NORM may accumulate in pipes and other production equipment and must be periodically removed. The radioactive material from these fixtures thus becomes technologically enhanced NORM (TENORM).

In 1999, the U.S. Department of Energy (DOE) sponsored a study of potential human exposures to radionuclides associated with the disposal of oil and gas extraction and production (E+P) wastes containing TENORM in municipal solid waste (MSW) landfills (Smith et al., 1999, 2003). This DOE-sponsored study considered a range of radionuclides potentially present in the sludge and scale waste, and multiple exposure pathways. Included in the radionuclides considered were radium-226 (²²⁶Ra) and radium-224 (²²⁴Ra), and their volatile progeny radon-222 (radon or ²²²Rn; half-life 3.8 days) and radon-220 (thoron or ²²⁰Rn; half-life 55 sec). The exposure assessment was performed using the multimedia model RESRAD Version 5.782 (Yu et al., 1993). The only significant exposure identified by this study was from indoor radon due to vapor-phase diffusion into a building constructed on the landfill after its closure. Radon emissions during the operating phase of the landfill were not considered. The study concluded that states should consider regulations allowing the disposal of limited quantities of TENORM wastes containing up to 50 pCi/g of ²²⁶Ra in MSW landfills and allow disposal of wastes containing higher concentrations on a case-by-case basis. With respect to water treatment residues, the radioactive materials in water treatment residues are primarily radionuclides in the uranium-238 (²³⁸U) and thorium-232 (²³²Th) decay series (Table 1). Except for ²²²Rn and ²²⁰Rn, the radionuclides in these decay series will be present primarily in the solid and liquid states in the residues. ²²²Rn and ²²⁰Rn, produced directly by the decay of ²²⁶Ra and ²²⁴Ra, respectively, will be released primarily in the vapor phase. Interagency Steering Committee on Radiation Standards (ISCORS) found little risk to human health from the nongaseous radioactive materials potentially present in the water treatment plant residues (ISCORS, 2003). Despite these studies, disposal of drilling wastes from the Marcellus Shale in the northeastern United States in nonhazardous waste landfills has triggered public concern (Finger, 2010). Potential exists for similar concerns to develop in other areas with significant shale gas development.

Table 1. Uranium and thorium decay series leading to production of radon and thoron

^238U Series	Half-Life	^232Th Series	Half-Life
^238U	4.5 × 10^9 yr	^232Th	1.4 × 10^10 yr
^234Th	24 days	^228Ra	5.8 yr
^234Pa	1.2 min	^228Ac	6.1 hr
^234U	2.4 × 10^5 yr	^228Th	1.9 yr
^230Th	7.7 × 10^4 yr	^224Ra	3.7 days
^226Ra	1600 yr	^220Rn	56 sec
^222Rn	3.8 days	^216Po	0.15 sec
^218Po	3.1 min	^212Pb	11 hr
^214Pb	27 min	^212Bi	61 min
^214Bi	20 min	^212Po and ^208Tl	3 × 10^-7 sec, 3.1 min
^214Po	1.6 × 10^−4 sec	^208Pb	Stable
^210Pb	22 yr
^210Bi	5 days
^210Po	140 days
^206Pb	Stable
Notes: Half lives from Argonne National Laboratory, EVS, Human Health Fact Sheet, August 2005. http://www.ead.anl.gov/pub/doc/natural-decay-series.pdf.

Purpose and Scope

Although previous studies of radon emissions from landfills (ISCORS, 2003; Smith et al., 1999, 2003) considered atmospheric radon emissions due to diffusion through the landfill cover, they did not consider the effect on radon emissions of landfill gas (LFG) generation due to biological decay of organic matter in the land disposal site, which could increase radon emissions by creating a pressure gradient between the refuse and the atmosphere. They also did not consider point-source emissions from LFG systems that are required at most MSW landfills. This paper provides estimates of the radon emissions and atmospheric radon activities that could result from disposal of radium-bearing drilling and water treatment wastes in an MSW landfill with LFG generation, including the effect of LFG control systems. Walter et al. (2005) considered the effect of gas generation from anaerobic hydrocarbon biodegradation in a land disposal site dedicated to E+P waste (a monofill).

Radon Emissions from Landfills

This paper focuses on vapor-phase emissions of ²²²Rn or radon, rather than ²²⁰Rn or thoron, due to radon's significantly longer half-life (Table 1). Radon emissions from radium-bearing wastes disposed in MSW landfills and dedicated land burial sites will be controlled by advective and diffusive transport processes, and partitioning between the gas, liquid, and solid phases in the refuse (Figure 1). Anaerobic biodegradation of organic matter in the solid waste and sludge produces gas that can increase radon emissions by creating a pressure gradient between the refuse and the atmosphere, thus adding a potentially significant advective flow component to the slower diffusive flow of radon. Internal gas generation will cause the radon to enter the accessible environment through the surface of the landfill, vents in the landfill cover, and the exhaust system of landfill gas control systems (such as flares, gas turbines, or internal combustion engines). The movement of radon in the gas phase is slowed somewhat by partitioning of the radon between the gas, water, and organic matter in the disposal site, thus reducing emissions to the environment by allowing more time for its radioactive decay in the subsurface.

Figure 1. Illustration of radon transport processes in a landfill and landfill construction used in numerical simulations (thicknesses not to scale).

Diffusion

Transport in the vapor phase has generally been considered the dominant transport process controlling the movement of radon from soil to the atmosphere, except in the case of vapor migration into buildings and subsurface structures where barometric pressure changes can induce adective flow. The tracer diffusion coefficient for radon in air is reported to be 1.2 × 10⁻⁵ m²/sec (Nazaroff, 1992). The diffusion coefficient in a landfill environment may be slightly higher than that normally assumed for soil because the temperature in refuse undergoing biodegradation is higher than the average soil temperature at a given location. The diffusion coefficient of radon in a landfill environment could also be affected, to a small extent, by the composition of the gas. In any event, the uncertainty associated with the effective diffusion coefficient due to tortuous diffusion pathways (tortuosity effects) in the subsurface environment is probably greater than those associated with temperature and gas composition. The effective, vapor-phase diffusion coefficient is reduced in the surface by the porosity, water content, and tortuosity of the soil and refuse. This research used the Millington model (Millington, 1959) to account for the influence of these factors on the effective diffusion coefficient.

Gas (advective) movement

The movement of biogenic gases, changes in barometric pressure, and gas movement induced by the operation of landfill gas control/extraction systems can also transport radon from landfills to the atmosphere. Such advective processes will cause the radon to enter the accessible environment through the surface of the disposal site, through vents in a landfill cover, and the exhaust system of landfill gas extraction systems (such as flares, turbines, or internal combustion engines).

Distribution phenomena

The extent to which radon released from radium in the landfill reaches the atmosphere is strongly dependent on its residence time in the subsurface given the relatively short half-lives of ²²⁰Rn and ²²²Rn. Retardation of the diffusive and advective movement of radon due to partitioning between the solid, liquid, and gas phases (Figure 1) within a MSW landfill can significantly increase its subsurface residence time.

Although radon is classed as an inert gas, it is known to sorb on solids such as activated carbon (National Academy of Sciences, 1999). The solid phase or phases in a landfill consist of the organic and inorganic materials in the refuse and the soil used for daily, interim, and final cover. The refuse deposited in MSW landfills in the United States consists largely of paper and paperboard (29% by weight), yard wastes (13%), food waste (14%), wood (6%), and miscellaneous other materials such as rubber, leather, plastics, metals, and glass (38%) (U.S. Environmental Protection Agency [EPA], 2011). Thus, given the high organic content of MSW and radon's affinity for activated carbon, sorption on the organic material in a landfill would potentially retard its movement and limit emissions to the environment.

Turtialinen et al. (2000) described the sorption of radon on activated carbon as primarily a physical process in the micropores of the carbon as opposed to being due to a chemical affinity of radon for the carbon. However, radon has a relatively high affinity for organic liquids, as evidenced by its octanol-water partitioning coefficient (K _ow) of approximately 32 and for soil organic carbon, with a water–organic carbon distribution coefficient (K _oc) averaging 23 mL/g (Wong et al., 1992). Schery and Whittlestone (1989) reported data indicating a distribution coefficient of 5–10 mL/g for peat moss at moisture contents of 5–10%. They report distribution coefficients between 0.01 and 0.1 mL/g for moist soil. Based on the nature of their tests, the distribution coefficients they report include the effect of the organic carbon content on the distribution coefficient, although they did not report its value.

Although the extent to which radon can be expected to sorb onto organic matter in a land disposal environment is uncertain, sorption onto organic matter should be regarded as potentially a significant process if the radon-generating material is mixed with organic waste. Assuming that radon sorption is proportional to the organic carbon content of the landfill solids, the solid-water distribution coefficient (K _d) can be approximated by

(1)

where K _oc is the organic carbon distribution coefficient (mL/g) and f _oc is the fraction organic carbon (dimensionless).

The relationship between the water activity and the solid activity is given by

(2)

where C _s is the activity in solid (picocuries per gram; pCi/g) and C _w is the activity in water (pCi/mL). A picocurie is represents 3.7 × 10⁻² radioactive decays per second or 3.7 × 10⁻² becquerels (Bq).

Water is the primary liquid phase in most MSW landfills, although small quantities of organic liquids, such as waste oil, paint, and household cleaners and solvents, will also be present in most landfills. Under equilibrium conditions, the partitioning of radon between gas and water can be described by Henry's Law:

(3)

where H _D is the dimensionless Henry's coefficient. The dimensionless Henry's coefficient H _D for radon is reported to be approximately 3.3 (National Research Council, 1998). However, based on data by Boyle (1911), H _D is temperature dependent and could be greater than 5 within a landfill where temperatures can reach 40–50 °C. A higher value of H _D would increase the gas-phase activity of radon and increase the radon surface flux.

Assuming equilibrium partitioning between the gas, liquid, and solid phases in the waste, the relationship between the radon gas-phase activity and its activity per bulk volume of landfilled material can be described by

(4)

where is the gas-volume partitioning factor (dimensionless); θ_T is the total porosity (dimensionless); S _w is the water saturation (dimensionless); and ρ_b is the dry bulk density (g/cm³). The calculations reported in this paper use a K _oc of 23 mL/g to calculate sorption of radon on organic matter with f _oc of 0.20 for the refuse and 0.005 for the vegetative layer soil. The results also assume H _D equals 3.3.

Radon emanation factor

In addition to these transport processes, the amount of radon that enters into the landfill pore space depends on the rate at which it can escape the solid, radium-bearing substrate. The rate is generally characterized by a radon emanation factor. Radon emanation factors of 0.22 for sludge and 0.05 for scale have been estimated (Rogers and Associates, 1997). The emanation factor determines the fraction of radon produced by ²²⁶Ra decay that actually escapes from the solid TENORM matrix into the void space of the subsurface medium.

Gas generation and emission processes

The rate and duration of gas production in a landfill depends on the organic content of the waste, the age of the waste, the volume of the landfill, and the availability of water to support the microbial community and the fermentation process. The production of gas increases the gas pressure until total gas emissions to the atmosphere and soil surrounding the landfill equal the rate of gas production. The over-pressure in the fill with respect to atmospheric pressure depends on the gas permeability of the cover, surrounding soil, and liner, and on the nature of the gas collection system, if any.

In the case of modern MSW landfills constructed in the United States that meet the requirements of 40 CFR 258 Subtitle D, the landfill or individual cells will have liners that restrict the movement of landfill gas into the surrounding soil. For such landfills, most of the gas will leave the landfill either as surface emissions through the landfill cover or through a LFG collection system. Active gas collection systems are required for landfills estimated to emit more than 500 Mg/yr of nonmethane organic compounds (NMOCs). Such systems typically involve a network of LFG collection wells penetrating the refuse or a perimeter LFG collection system, and a flare or other combustion system to burn the methane and NMOCs. In some cases, the LFG may be used as fuel for internal combustion engines or gas turbines to generate electricity, or scrubbed and conveyed to an off-site user via a gas pipeline. Even at landfills that do not exceed the 500 Mg/yr threshold, active or passive LFG collection may be required to prevent off-site migration of methane or to control odors.

Landfill gas generation rates vary from about 1 × 10⁻³ to 8 × 10⁻³ cubic meters per kilogram refuse per year (m³/kg-yr) (Emcon Associates, 1992). In the absence of site-specific measurements, the Landgem model (Pelt et al., 1998) is often used to estimate emissions based on the age and volume of the landfill. The Landgem model is based on an exponential decay equation of the form:

(5)

where q is LFG generation rate (m³/kg-day); L _R is the refuse methane generation potential (m³/kg); α is the refuse degradation rate (day⁻¹); t is the time after disposal (day); and t ₀ is the age of refuse (day). The factor of 2 in eq 5 comes from the assumption that methane and carbon dioxide are generated in equal volumes. Values of α range from 8 × 10⁻⁶ to 6 × 10⁻⁴ day⁻¹. Values of L ₀ range from 0.006 to 0.27 m³/kg and depend on the organic composition of the refuse, its moisture content, and other site-specific factors.

Simulation of Potential Radon Emissions

Radon emissions from land disposal of TENORM will be highly site specific. For the purposes of this study, a likely worst-case (conservative) disposal condition was assumed in which the TENORM was deposited in a discrete layer similar to that assumed by Smith et al. (1999, 2003). The reports by Smith et al. are unclear as to the depth of burial assumed during active landfill operations. The disposal scenario used here assumes that the TENORM waste is deposited on the surface of the refuse before the cell is closed (Figure 1). Other disposal scenarios are possible. For example, the TENORM waste could be delivered to the landfill periodically and become incorporated as discrete layers sandwiched between daily cover or the TENORM could be mixed (commingled) with conventional refuse. Walter et al. (2005) found that mixing the TENORM with refuse significantly reduced radon emissions.

Simulations of radon emissions were performed using a numerical model of gas flow and radon transport. The program simulates emissions of radiogenic gases from a land disposal site containing an internal source of gas. Flow and transport processes represented in the model include the following:

•	Internal generation of a carrier gas due to a first-order decay process
•	Advective flow of an ideal carrier gas governed by Darcy's Law
•	Internal generation of a contaminant gas due to first-order decay of a parent constituent
•	Vapor-phase molecular diffusion of the contaminant gas
•	Advective-dispersive transport of the contaminant gas
•	Retardation of the contaminant gas due to linear partitioning between subsurface solid, water, and gas phases
•	First-order decay of the contaminant gas

The subsurface flow and transport model is linked to a semianalytical Gaussian atmospheric dispersion model to allow computation of atmospheric activities of the contaminant gas from surface and point sources.

Radon emissions were simulated using a three-dimensional model it represent a landfill cell 200 m wide by 200 m long by 25 m deep. The simulations were performed starting with the initial pressure in the subsurface produced by LFG from the refuse below the TENORM layer. The initial age of the refuse was specified as that resulting in a short-term LFG generation rate for 11-yr-old refuse, approximately equal to the time-weighted average rate over a 30-yr operational period, with a landfill gas generation potential of 0.1 m³/kg and refuse degradation rate (α) of 1.4 × 10⁻⁴ day⁻¹.

The ²²⁶Ra activity in the cells containing TENORM waste was specified as 5 × 10⁴ pCi/kg (50 pCi/g), as assumed by Smith et al. (1999, 2003). For the purposes of simulating surface emissions, the TENORM was assumed to be spread uniformly across the landfill cell as a discrete layer 2.5 m thick. This TENORM distribution represents an upper bounding case for the landfill cell unless higher radium activity waste is allowed to be placed in the landfill.

The assumption of an initial radon activity of zero implies that radon in the TENORM is not in secular equilibrium with radium when the waste is deposited in the landfill. Although this assumption may not be correct with respect to the solid TENORM matrix, it is probably correct for the pore space of the TENORM waste because any gas-phase radon within the waste would be lost when it is initially deposited in the landfill. In any event, the simulation period (typically 50 days) was long enough for the radon activities in landfill to reach a nearly constant value with respect to the rate of landfill gas production.

Two landfill cover conditions were simulated. The first assumed a landfill with a conventional “RCRA” cover as specified 40 CFR 258 Subtitle D, illustrated in Figure 1, consisting of (from the surface downward) 0.5 m of vegetated soil cover, 0.5 m of gravel capillary barrier and drainage layer, and 0.5 m of compacted clay. Material properties assumed in the simulations are listed in Table 1. The second condition considered a cover with the compacted clay replaced with a high-density polyethylene geomembrane liner with a saturated hydraulic conductivity (water permeability) of 2 × 10⁻¹³ cm/sec (Schroeder et al., 1994). The hydraulic properties assigned to the materials comprising the model with the RCRA cover are listed in Table 2 and the transport properties are listed in Table 3. These properties were selected to be representative of in-place materials used in cover construction. Flow and transport across the surface layer is controlled by the difference between a prescribed atmospheric pressure, and the pressure and radon activity in the uppermost model cell.

Table 2. Hydraulic properties assigned to material types used in landfill model

Material	Saturated Hydraulic Conductivity (K h) (m/sec)	Porosity	Water Saturation	Vertical Anisotropy (K z/K h)
Vegetative layer	2 × 10^−5	0.3	0.2	0.1
Drainage layer	1 × 10^−3	0.3	0.01	1
Clay barrier	2 × 10^−7	0.4	0.3	0.1
Geomembrane	2 × 10^−14	0.01	NA	NA
TENORM waste	2 × 10^−5	0.4	0.3	0.1
Refuse	1 × 10^−4	0.4	0.3	0.1

Table 3. Transport properties assigned to material types used in landfill model

Material	Bulk Density (kg/m^3)	^226Ra (pCi/kg)	Rn Emanation Factor	Dispersivity (m)
Vegetative layer	1500	0	0.1	1
Drainage layer	1500	0	0.1	1
Clay barrier	1500	0	0.1	1
TENORM waste	1500	50,000	0.2	1
Refuse	1300	0	0.1	1

Simulation scenarios

The following six scenarios were simulated:

1.	Conventional cover with no LFG production
2.	Conventional cover with no LFG collection system
3.	Conventional cover with a passive LFG collection system
4.	Conventional cover with an active LFG collection system
5.	Geomembrane cover with passive LFG collection system
6.	Geomembrane cover with active LFG collection system

The scenario of a geomembrane cover with no LFG collection was not considered reasonable because a geomembrance cover would generally require a LFG collection system to prevent excessive pressure build-up in the landfill cell.

Two landfill gas control systems were considered: a system of passive perimeter wells and an active landfill gas extraction system. The passive vent system consisted of 16 vent wells located along the perimeter of the cell, as illustrated in Figure 2A. Each well was assumed to fully penetrate the refuse, but to be sealed-off through the cover and TENORM layer. Each well was maintained at atmospheric pressure so that the landfill gas emission rate at each well was proportional to the internal landfill pressure and the extracted gas was released at each well location.

Figure 2. Configuration of LFG collection wells. (A) Passive vent wells. (B) Active collection system with central vent well.

The active landfill gas extraction system consisted of five wells at the locations illustrated in Figure 2B. Landfill gas was assumed to be extracted at a constant rate from each well and conveyed to a central stack near the center of the cell. The extraction rates were specified so the total extraction rate was equal to approximately 95% of the total landfill gas generation rate based on the assumption that the landfill gas extraction system is not 100% effective. Specifically, the total landfill gas generation rate was 0.2 standard m³/sec and the extraction rate at each well was 0.04 m³/sec.

The atmospheric radon activities were computed using an analytical, Gaussian dispersion model without building wake effects (Zannetti, 1990). Emissions from landfill gas passive vent wells are represented as point sources emitted from a stack height of 2 m. Emissions from the active extraction system were released from a 6-m-high stack located at the central extraction well. The Gaussian dispersion calculations were based on a wind speed of 1 m/sec from left to right across the landfill, as depicted in Figure 2A and B. Surface emissions directly through the landfill cover were simulated by representing each model cell as a point source and using superposition to compute the total atmospheric activity.

Simulation Results

The results of the simulations are shown in Figure 3 in terms of the radon emission rate due to diffusion and advection through the surface of the landfill, and releases from the LFG control systems, if present. The components of the radon fluxes are listed in Table 4. The simulation results for the clay cover with LFG generation and no LFG control indicate that LFG generation increases the radon emission rate by a factor of approximately 6 relative to emissions by diffusion only with no LFG production. Although the simulations for the clay cover with passive and active LFG control decrease total emissions relative to the no control simulation, the total emissions are still 2 to 3 times higher than the case with no LFG generation. For the clay cover, the LFG control systems significantly reduce the advective flow through the landfill surface by decreasing the gas pressure in the landfill and causing a portion of the radon to flow through the refuse, where it decays, without significantly reducing the diffusive flow. The cover with the geomembrane essentially eliminated any surface releases of radon and limited the releases from the LFG control wells by allowing radon to decay within the refuse. Figure 4 shows the rate of decay of radon within the subsurface of the landfill versus the total emission rate to the atmosphere. More radon decays in the subsurface for the landfills with LFG controls because the control wells cause radon to flow through the refuse, thus increasing the subsurface residence time and decay, before being emitted to the atmosphere. Total radon emissions for the geomembrane scenarios were significantly less than those for the scenario with no LFG production due to a longer subsurface residence time for the radon.

Table 4. Summary of Radon emission rates from the landfill model simulations

	Emitted through Surface (pCi/sec)
Scenario	Advective	Diffusive	Emitted Through Wells (pCi/sec)	Total Atmospheric Emission (pCi/sec)	Decayed in Subsurface (pCi/sec)	Ratio Decayed to Emitted	Effective Surface Flux (pCi/m^−2-sec)
No gas control	0	3.9 × 10^5	0	3.9 × 10^5	5.1 × 10^6	13	9.8
Clay cover—No control	1.2 × 10^6	1.0 × 10^6	0	2.2 × 10^6	6.6 × 10^6	1.2	55
Clay cover—Passive	2.7 × 10^5	1.1 × 10^6	2.4 × 10^4	1.4 × 10^6	3.7 × 10^6	2.6	35
Clay cover—Active	6.6 × 10^4	7.6 × 10^5	6.6 × 10^4	8.9 × 10^5	4.3 × 10^6	4.8	22
Geomembrane—Passive	Negligible	Negligible	6.3 × 10^4	6.3 × 10^4	5.2 × 10^6	83	1.6
Geomembrane—Active	Negligible	Negligible	7.3 × 10^4	7.3 × 10^4	5.3 × 10^6	72	1.8

Figure 3. Summary of radon emission rates for the various landfill cover and LFG collection scenarios.

Figure 4. Radon mass decayed versus radon emissions from the LFG system.

The simulated spatial distribution of atmospheric radon activities for the for the case on no LFG generation is shown in Figure 5A and for the passive LFG vent scenarios with LFG generation in Figure 5B and C. Comparison of Figure 5A with Figure 5Bfurther illustrates the significant increase in radon emissions resulting from LFG generation in the landfill.

Figure 5. Simulated atmospheric radon activity at 2-m elevation. (A) Landfill with no LFG production, emissions by diffusion only. (B) Landfill with clay cover and passive LFG vent system. (C) Landfill with geomembrane cover and passive LFG vent system. Note that atmospheric activities are computed analytically only at locations in the computational grid. The plumes would extend outside the computational grid (color figure available online).

Figure 6 shows the simulated atmospheric plumes for the landfill with clay cover and no LFG control (Figure 6A), and the two active LFG control scenarios. The atmospheric plumes for the active extraction systems appear to be disconnected from the source at the central well because the stack height was set at a height of 6 m to represent a flare, whereas the radon activities are computed at a height of 2 m to represent the breathing zone. Under the simulated wind conditions (1 m/sec, constant direction), the simulations for the landfills with the clay cover, with or without LFG collection systems, result in radon plumes exceeding 0.5 pCi/L (approximately the average radon background in the United States) (United Nations Scientific Committee on the Effects of Atomic Radiation, 2000), extending well off the disposal cell. The simulations for the scenarios with the geomembrane cover result in significantly smaller radon plumes because the geomembrane eliminates essentially all of the emissions through the surface of the landfill.

Figure 6. Simulated atmospheric radon activities at 2-m elevation. (A) Landfill with clay cover, LFG generation, and no LFG control system. (B) Landfill with clay cover and active LFG control system. (C) Landfill with geomembrane cover and active LFG control system. Note that atmospheric activities are computed analytically only at locations in the computational. The plumes would extend outside the computational grid.

The atmospheric plumes shown in Figures 5 and 6 represent conservative activities because the simple atmospheric dispersion model does not account for variations in wind speed, direction, or atmospheric stability conditions. These factors are site specific and incorporating them into the analysis would decrease average radon activities, but increase the plume spread and areal coverage.

Summary and Conclusions

The simulation results clearly indicate that LFG generation in MSW landfills can significantly increase atmospheric release of radon from TENORM disposed in the landfill. The emissions and resulting atmospheric activities from a specific landfill will depend on the manner in which the TENORM is placed in the landfill, the LFG-generating properties of the refuse in the landfill, the nature of the LFG control system, and, to a great extent, the construction of the landfill cover. In particular, mixing the TENORM with the refuse may reduce atmospheric emissions.

Estimating the human health effects from exposure to radon activities reported in this study was beyond the scope of this investigation. In addition, the findings of this study indicate that actual radon exposures would be highly site specific depending on the manner and volume of TENORM disposal and the environmental setting of the land disposal unit. For these reasons, the potential atmospheric radon activities presented above were simply compared to reported natural radon activities in air and relevant guidance and standards for radon exposure.

According to United Nations Scientific Committee on Effects of Atomic Radiation (2000), the average global radon activity is 0.3 pCi/L. The average indoor air radon activity is reported to be 1.2 pCi/L in the United States.²⁵ The EPA (2009) has recommended an indoor air radon activity of 4 pCi/L as a recommended action level for mitigation but has a long-term goal of reducing indoor activities to ambient atmospheric levels of approximately 0.4 pCi/L. The International Commission on Radiological Protection (ICRP) (1993) previously recommended action levels for dwellings and work places of 200 to 600 Bq/m³ (5.4–16 pCi/L) and 500–1500 Bq/m³ (13.5–40.5 pCi/L) for an equilibrium factor of 0.4, which is typical for indoor or underground exposure. Several equivalent definitions for the equilibrium factor have been used for situations in which radon decay progeny are not in equilibrium with radon (i.e., have lower activity concentrations than radon). The equilibrium factor is the air concentration ratio of potential alpha energy for the actual mixture of radon decay products to the potential alpha energy if all decay products were in equilibrium with radon. More recently, the ICRP (2009) recommends applying occupational radiological protection requirements to existing exposure situations with radon activities at 1000 Bq/m³ (27 pCi/L) or more. The Health Physics Society (2009) has published an update on perspectives and recommendations on indoor radon. The best estimate for effective dose due to indoor radon exposure is 3–7 milliSieverts per year (mSv/yr) at 4 pCi/L (300–700 mrem/yr).

Straightforward application of these relationships for indoor ²²²Rn to radon emanation, dispersion, and outdoor exposure is not advised due to large expected differences in the equilibrium factor. Radiological dose from radon is dominated by the radioactive decay products of radon. As specified in 10 CFR Part 20, Appendix B, Tables 1 and 2, derived air concentrations for occupational inhalation (and effluent concentrations in air for ²²²Rn with decay products removed relative to ²²²Rn with decay products are 2 orders of magnitude different. The derived air concentration for a given radionuclide is the air concentration that corresponds to the annual occupational dose limit for a 2000-hr exposure during light work with a breathing rate of 1.2 m³/hr. For air effluents, the ²²²Rn activity limits are 10 pCi/L without decay products and 0.1 pCi/L with decay products. When emitted to the atmosphere from the landfill surface or LFG control system, the radon plume is expected to be depleted in decay products. Although radioactive in-growth will occur during atmospheric transport of the plume, the transport time near the landfill will be short relative to the time for radon decay products to reach equilibrium levels. However, as the plume moves further downwind, decay products will build up even though the plume becomes more dispersed. These uncertainties in the outdoor exposure scenario to radon from landfills make estimating doses beyond the scope of this study.

No clear regulatory standards were identified that would apply to atmospheric radon from TENORM placed in a MSW landfill. In its guidance on protective cleanup levels for radioactive contamination at Comprehensive Environmental Response, Compensation, and Liability Act of 1980 (CERCLA) sites, the EPA (1997) recommends a maximum radiological dose limit of 15 mrem/yr for those sites at which a dose assessment is conducted. The EPA considers this limit to be inclusive of all radioactive contaminants of concern at a site including radon. The Nuclear Regulatory Commission uses 25 mrem/yr as a standard for unrestricted land use when terminating a licensed site (10 CFR Part 20). The regulatory standard most closely corresponding to emissions from land disposal of radon-generating TENORM would be that for inactive uranium mill tailings sites in 40 CFR Part 192.02. This standard limits the average surface release of ²²²Rn to 20 pCi/m²-sec for the site and the annual average ²²²Rn activity in air “at or above any location outside the disposal site” to 0.5 pCi/L. As indicated by Figure 7, the simulated radon fluxes for the scenarios analyzed exceed the uranium mill tailings regulatory flux limits for the cases without a geomembrane cover, and exceed the off-site radon activity limit in all cases (Figures 5 and 6).

Figure 7. Equivalent ²²²Ra surface flux computed for the various landfill scenarios. Equivalent surface flux is the total emission rate from the landfill surface and LFG control wells divided by the landfill cell area.

Nomenclature

f oc	=	= fraction organic carbon, dimensionless
	=	= gas generation rate per unit mass, L3/m-t [m3/kg-day]
	=	= time after disposal, t [day]
	=	= age of refuse, t [day]
	=	= gas activity, m/L3 [pCi/L]
	=	= gas activity per bulk volume, m/L3 [pCi/L]
	=	= activity in water, m/L3 [pCi/L]
	=	= dimensionless Henry's coefficient
	=	= radon flux, m/L2-t [pCi/m2-sec]
	=	= distribution coefficient, L3/m [mL/g]
K h	=	= horizontal hydraulic conductivity [m/sec]
	=	= organic carbon distribution coefficient, L3/m [mL/g]
	=	= oil-water distribution coefficient, dimensionless
	=	= vertical hydraulic conductivity [m/sec]
	=	= refuse methane generation potential, L3/m [m3/kg]
	=	= gas-volume partitioning factor, dimensionless
	=	= water saturation, dimensionless
	=	= refuse degradation rate, t−1 [day]
	=	= total porosity, dimensionless
	=	= dry bulk density, m/L3 [kg/m3]

Acknowledgments

This work was funded, in part, by Southwest Research Institute Internal Research and Development Project No. 20-8096. We appreciate the helpful comments from the anonymous reviewers.