Methane microseepage in drylands: soil is not always a CH₄ sink

Abstract

Drylands are considered a net sink for atmospheric methane and a main component of global inventories for greenhouse gas budgets. However, a significant portion of drylands occur over sedimentary basins hosting natural gas and oil reservoirs, with gas migration to the surface (named “microseepage”) producing positive atmospheric CH₄ fluxes. In this overview, we summarize the outcomes of microseepage surveys performed in different petroleum basins, describe how the microseepage area is estimated and what are the emission factors that can be used for a preliminary global emission estimate. Microseepage frequently overcomes methanotrophic consumption occurring in dry soil throughout large areas, and it is enhanced by faults and fractures in the rocks. Fluxes of a few tens to hundreds of mg m⁻² d⁻¹ are frequent for oil–gas fields, globally estimated at ∼4 million km². However, microseepage may potentially exist over a wider area (∼8 million km², i.e. 15% of global drylands), including the total petroleum system. Based on a relatively large and geographically dispersed data-set of emission factors from different hydrocarbon-prone basins in the USA and Europe, upscaling suggests that global microseepage emission exceeds 10 Tg y⁻¹; it cannot be ignored in the atmospheric methane budget and in assessments for the sink potential of dry soil.

Keywords:

• methane
• seepage
• geological sources
• soil uptake

Introduction

Dryland soil is considered a net biotic sink of atmospheric methane, with global uptake on the order of 30 ± 15 Tg y⁻¹ (Intergovernmental Panel on Climate Change 2001) or 20 ± 3 Tg y⁻¹ (Potter et al. 1996). However, some projections fall within the range of 5–58 Tg y⁻¹ (Dorr et al. 1993) indicating that there is still substantial uncertainty regarding the magnitude of this global sink. Negative gas flux, generally on the order of −5 to −1 mg m⁻² d⁻¹ (Dong et al. 1998), is due to methanotrophic oxidation by CH₄ ⁻ consuming bacteria in the soil. Methanotrophic oxidation occurs in grasslands, temperate and boreal forest soils, desert soils, fertilized soils, humisol, moss-derived peat soils, tundra soils, and unflooded paddy soils (Minami and Takata 1997). Soil is considered a source of methane only in wet conditions, in the presence of methanogenic bacteria (in all wetlands, including rice paddies, bogs and flooded soils; Batjes and Bridges 1994).

In the 1980s and 1990s, some anomalies with respect to dryland behavior (i.e. positive fluxes instead of negative fluxes) were found in South America. Unexpected emissions of methane to the atmosphere (>1 mg m⁻² d⁻¹) were found in two areas of dry grassland or savanna soils within the Orinoco Valley and in the Guyana Shield of northeastern Venezuela (Hao et al. 1988; Scharffe et al., 1990). The measurements were criticized and considered erroneous by other researchers, as the authors had no explanation for the positive methane flux (P. Crutzen, personal communication). However, Hao et al. (1988) suggested the possibility of gas release through upward diffusion from underground natural gas reservoirs near Chaguarama, in the region investigated. At a later date, Hao et al. (1988) were deemed correct since their “biological” survey was actually conducted over one the largest petroleum systems in the world (the Orinoco Petroleum Belt; Erlich and Barrett 1992). The area investigated by Scharffe et al., (1990), near the Guri dam and south of the Orinoco Belt, is located in an area that corresponds with important SW-NE deep fault systems containing highly fractured and permeable mylonites, characterizing the regional brittle tectonics of the Guyana shield.

Later, under the framework of studies for hydrocarbon seepage from sedimentary basins, positive fluxes of methane in dry lands were reported for several sites in the USA (Klusman et al. 1998, 2000a). Indeed, the occurrence of methane and light alkane anomalies in dry soils has been extensively used by geologists and geochemists as a tool for oil and gas exploration since the 1930s (Laubmeyer 1933), and then in recent times (Jones and Drozd 1983; Klusman 1993; Tedesco 1995; Schumacher and LeSchack 2002; Abrams 2005). Microseepage is in fact recognized as a slow but pervasive, diffuse exhalation from soils of light gaseous hydrocarbons, mainly methane, resulting from natural gas migrations stored in underground hydrocarbon accumulations (Figure 1; Brown 2000; Etiope and Klusman 2002, 2010). Methanotrophic oxidation can partially attenuate microseepage which, over a long time frame, can produce calcium carbonate veins and dispersed calcium carbonates of distinct isotopic composition. Indirect methods, such as microbial prospecting (e.g. Tucker and Hitzman 1996; Wagner et al. 2002), remote sensing (e.g. Van der Meer et al. 2002), and magnetic measurements (e.g. Liu et al. 2004) have then confirmed the existence of microseepage throughout large areas in oil–gas fields on various continents. Nevertheless, these studies have focused exclusively on the soil detection of anomalous concentrations of methane and light alkanes. Soil-atmospheric flux measurements, being time consuming, and unnecessary for oil–gas explorations, were not made since understanding atmospheric impacts was not an objective. As a consequence, the available data-set for microseepage fluxes is rather poor. Only recently, since 2000, have a large number of flux data sets been acquired throughout dry soil areas in hydrocarbon-prone sedimentary basins in Europe and Asia, specifically in Italy, Romania, Greece, Azerbaijan (Etiope et al. 2002, 2004a, 2004b, 2006), and China (Tang et al. 2008). Together with measurements made in the United States by the Colorado School of Mines (e.g. Klusman et al. 2000a), the surveys listed form the only systematic estimates for microseepage fluxes released to the atmosphere.

Figure 1. Sketch of gas microseepage from natural hydrocarbon reservoirs.

Today, it is known that the positive flux of methane, or microseepage, can reach levels of tens, hundreds, and thousands of mg m⁻² d⁻¹ throughout large areas, especially surrounding macro-seeps such as those associated with mud volcanoes (Etiope and Milkov 2004; Etiope et al. 2004a, 2004b). Microseepage is quite common and pervasive within petroliferous and sedimentary basins at lower rates.

Methane microseepage can also occur over coal-beds but generally it is not considered as a natural source, being almost always produced by dewatering of coal strata induced by mining activity or preparation of rocks for coal-bed methane (CBM) production. Natural seepage of thermogenic gas, unrelated to mining, was however reported in the Ruhr basin in Germany (Thielemann et al. 2000). So, the existence of significant microseepage related to coal-beds cannot be excluded; its actual role as a methane source must be assessed by field measurements in coal basins not perturbed by mining. Positive fluxes of methane from the soil can, then, occur in geothermal areas (Etiope 1999; Klusman et al. 2000b; Etiope et al. 2007) where methane is produced by high temperature inorganic reactions or the thermal breakdown of organic matter. In all cases, microseepage can vary seasonally with methanotrophic oxidation being lower in colder and dry seasons.

However, microseepage is only one component of a wider class of geological CH₄ sources, including macro-seeps (e.g. mud volcanoes), geothermal, and submarine emissions (Etiope 2004; Kvenvolden and Rogers 2005) which emit, in total, ∼53 ± 11 Tg CH₄ y⁻¹ (Etiope et al. 2008) representing the second most important natural source of methane after wetlands. Geo-methane emissions are now considered as a new source category, both in the updated (2009) European EMEP/EEA air pollutant emission inventory guidebook (EMEP/EEA 2009) and in a new U.S. EPA report on natural sources for methane (U.S. EPA 2010).

Global area potentially affected by microseepage

More than 75% of the world's petroliferous basins contain surface macro-seeps (Clarke and Cleverly 1991) around which microseepage occurs, typically in wide halos. Independent of macro-seeps, Klusman et al. (1998; 2000a) assumed that microseeping areas included all of the sedimentary basins in a dry climate, with petroleum and gas generation processes at depth, in an area that has been estimated to be ∼43.4 × 10⁶ km². The flux data available today suggests that microseepage corresponds closely to the spatial distribution of hydrocarbon reservoirs and portions of sedimentary basins that are, or that have been in the past, at temperatures >70°C (thermogenesis). Accordingly, Etiope and Klusman (2010) assumed that microseepage may occur within the so-called total petroleum system (TPS), a term used in petroleum geology (Magoon and Schmoker 2000) to describe the whole hydrocarbon-fluid system in the lithosphere including the essential elements and processes needed for oil and gas accumulations, migrations, and seeps. In the TPS, gas migrations to the surface can occur through advective mechanisms (Brown 2000; Etiope and Martinelli 2002) driven by two main factors: (1) excessive pressure gradients in the rocks and (2) permeable pathways (fractures, faults, and permeable sedimentary horizons). Wherever both factors exist, microseepage to the surface can easily result.

In the world, 42 countries produce 98% of the petroleum, 70 countries produce 2%, and 70 countries produce 0%. So a TPS, and as a consequence the potential for microseepage, occurs in 112 (42 + 70) countries. The first consideration suggests that microseepage is potentially a very common phenomenon and widespread on all continents. The global area of potential microseepage was preliminarily assessed by an analysis for the distribution of oil–gas fields within all of the 937 petroleum provinces or basins, as reported in the GIS data-set of the U.S. Geological Survey World Petroleum Assessment 2000 and related maps (Etiope and Klusman 2010). For each province, a polygon was drawn that enclosed gas–oil field points in interactive maps, and the area was estimated by graphic software. Significant gas–oil field zones were found to occur in at least 120 provinces, with the total area of gas–oil field zones estimated to be between 3.5 and 4.2 million km² (Etiope and Klusman 2010), ∼7% of global dryland area. Global TPS area has been estimated to be ∼8 million km² (i.e. 15% of global dryland and twice global wetland area; Matthews and Fung 1987). It must be emphasized that such estimates refer to potential areas of microseepage. Actual microseepage areas are unknown. Microseepage from coal deposits outside TPS is not considered at this stage, because available flux measurements are quite scarce and in most cases, as previously mentioned, the seepage is not natural but induced by the mining activity or dewatering associated with CBM production. Thus, areas with coal deposits can be partially considered as natural microseepage sources, but the lack of a statistically significant amount of data prevents any estimation of the global area and related emission factors.

Global estimate of methane emission by microseepage

Preliminary models have suggested that hydrocarbon-prone sedimentary basins in dry climates may produce a mean microseepage flux in the subsoil of 4.42 mg CH₄ m⁻² d⁻¹ (Klusman et al. 1998; 2000a). Assuming 90% methanotrophic consumption for this microseepage rate in dry soils suggests a global emission of methane to the atmosphere of at least 7 Tg y⁻¹.

Present surveys confirm microseepage in the range of some tens of mg m⁻² d⁻¹, and more rarely hundreds or thousands of mg m⁻² d⁻¹ in smaller areas, corresponding with faults in hydrocarbon reservoirs or around macro-seeps. Close to macro-seeps (mud volcanoes or other gas seeps) the flux can reach 10⁵ mg m⁻² d⁻¹ (Etiope et al. 2004a, 2004b). Accordingly, it is not possible to consider an average value for a given TPS, and homogeneously identifiable areas with different microseepage levels should be taken into account.

A wide geographically dispersed data-set of microseepage flux data from the USA and Europe (563 data points, mostly averages of 2 or 3 measurements from the same site made with different closed-chamber systems and for different seasons) suggests that at least three main levels of microseepage should be considered (Etiope and Klusman 2010):

	Level 1: High microseepage (>50 mg m−2 d−1)
	Level 2: Medium microseepage (5–50 mg m−2 d−1)
	Level 3: Low microseepage (0–5 mg m−2 d−1).

Levels 1 and 2 occur mainly in sectors hosting macro-seepage sites, and in sedimentary basins in general, during winter. Of 563 measurements, 276 were positive fluxes (49%) and 3% were in the level 1 range (mean of 210 mg m⁻² d⁻¹). Level 2 represented ∼12% of the surveyed areas (a mean of 14.5 mg m⁻² d⁻¹). Level 3 was common in winter, far from macro-seepage zones, and accounted for ∼34% of the sedimentary zones surveyed (a mean of 1.4 mg m⁻² d⁻¹). Scaling up to a global microseepage area could be based on considering the average flux from each of the three microseepage levels, and from assuming that the percentage of the occurrence of different levels (3%, 12%, and 34%) is valid at the global scale. The assumption did not give large errors since the result was not particularly sensitive when changing the percentages for several units. Accordingly, scaling up to all gas–oil field areas gave a total microseepage on the order of 11–13 Tg y⁻¹. Extrapolating to the global potential microseepage area (TPS: ∼8 million km²) provided an emission on the order of 25 Tg y⁻¹. These estimates are in agreement with the lower limit of 7 Tg y⁻¹ as initially suggested by Klusman et al. (1998) and Etiope and Klusman (2002). However, more measurements in various areas and for different seasons are needed to refine the three-level classification and the actual area of seepage.

Conclusions

A significant portion of drylands occurs over sedimentary basins hosting natural gas and oil reservoirs where gas migration to the surface (microseepage) takes place, producing positive fluxes of methane to the atmosphere, that range from a few to hundreds of mg m ⁻² d⁻¹. These fluxes have been determined at an increasing numbers of locations. Based on a relatively large and geographically dispersed data-set from different hydrocarbon-prone sedimentary basins, it has been possible to improve global estimates of microseepage extent and emission rates to the atmosphere. Global emissions should be in the range of 10–25 Tg y⁻¹, which is comparable to some estimates for the global methane sink in drylands (Potter et al. 1996). Nevertheless, microseepage was not considered among the natural methane sources (e.g. Wuebbles and Hayhoe 2002) and it has been just mentioned in the latest IPCC 2007 report (Denman et al. 2007). If global microseepage is actually >10 Tg y⁻¹ it cannot be neglected in the atmospheric source/sink methane budget.

Indeed, microseepage is only one component of a wider range of geological CH₄ sources, which include onshore macro-seeps (mud volcanoes plus other seeps), submarine macro-seeps, and geothermal emissions. All of these geological sources would produce from 42 to 64 Tg CH₄ y⁻¹ (Etiope et al. 2008), representing the second largest natural source of methane, after wetlands. Also, this source would represent ∼10% of the total emission of methane to the atmosphere. However, microseepage has a specific role in the global methane budget. Intrinsically, this suggests that not all drylands are a methane sink and basically that drylands, that are a part of petroleum basins and sedimentary basins having undergone thermogenesis, may not show methane uptake but a positive exhalation to the atmosphere that varies seasonally. If this is the case, global estimates of soil uptake should be re-assessed by removing ∼7 to 15% of global drylands areas. The consensus value, 30 Tg y⁻¹ (e.g. IPCC 2001), could be overestimated; the large uncertainties in global uptake (5 to 58 Tg y⁻¹) suggest that further efforts should be made to improve estimates, and the present article outlined that such efforts should necessarily include microseepage surveys.