Abstract

Methane (CH4) microseepage from petroleum basins is a significant contributor to the atmospheric CH4 budget. However, research about CH4 migration and release mechanism is still very limited. This work seeks to theorize and verify the migration and release mechanism of CH4 microseepage via field measurement and physical simulation, which, to the best of our knowledge, has not been reported in literature. Fluxes of CH4 microseepage from Dawanqi oilfield were measured, and three manifestations of release were observed, namely, continuous, flat, and episodic. Based on field observations, bench-scale physical simulation of CH4 migration through geological features of the oilfield was further conducted for 290 days. The results show that CH4 migration is mainly driven by buoyancy and diffusion. In continuous release, CH4 migration is mainly driven by buoyancy. In flat release, CH4 migration is dominated by diffusion. At low pressure, CH4 migrates upward slowly. As buoyancy increases, CH4 eventually break through the capillary pressure of the pore throat, causing spikes in CH4 concentrations in the layers above and reproducing episodic release observed during field measurement. Via field observation and verification by physical simulation, this work theorizes the migration mechanism of CH4 microseepage and its correlation with release types observed and confirms that counterbalance of buoyancy force and capillary pressure plays a critical role in episodic release of CH4 from oilfield. The findings of this study shed light on the migration mechanism and release manifestations of CH4 microseepage under different geological conditions and improve accuracy of estimating the flux of CH4 microseepage into atmosphere.

1. Introduction

Microseepage, the slow and diffuse migration of gaseous hydrocarbons from underground petroleum reservoirs, is an important process that influences gas-oil exploration and the atmospheric methane (CH4) budget [1, 2].

Since the 1930s, geologists and geochemists have extensively exploited the presence of CH4 and light alkanes in soil for oil and gas exploration [35]. For example, breakthroughs were made in oil and gas exploration in China in the 1950s, with the observation of oil or gas seepage above reservoirs, such as the Karamay oilfield [6]. Since the 1980s, microseepage has been measured and modeled in several petroleum basins of North America and Europe [79]. Klusman et al. suggested that knowing the gas flux was also valuable for petroleum exploration and applied the closed chamber method for gas flux measurements in petroleum geology [10]. Several studies found that drylands are not necessarily a net sink for atmospheric CH4. A substantial portion of drylands occur over sedimentary basins that host natural gas and oil reservoirs, where gas migration to the surface takes place, producing positive fluxes of CH4 into the atmosphere [11, 12]. Accordingly, research includes microseepage, together with other geological CH4 exhalation processes (mud volcanoes, oil-gas seeps, and submarine seepage), in the atmospheric budget of natural CH4 sources [1315].

Numerous studies have confirmed that microseepage of CH4 on the surface (soil, subsoil, and shallow aquifers) originates from deep gas-oil reservoirs [1618]. There has been a preliminary understanding of CH4 emissions from shale gas extraction [1922], production/abandoned oil and gas wells, and their effects on the environment in recent years [2325]. However, little research focused on the migration and release mechanisms of hydrocarbons [26, 27]. Most researchers have agreed that hydrocarbons that originate from oil reservoirs migrate upward to the surface by buoyancy, advection, and water dissolution [2629]. Compared with other researchers, we observed three different release manifestations, namely, continuous, flat, and episodic release, on the surface in field work. And on the basis of field results, we discussed the migration mechanism of CH4 microseepage corresponding to the different release manifestations and potential impact of various geological conditions.

Here, we present the steps of the work (Figure 1). First, Dawanqi oilfield was selected, which has a shallow reservoir. Then, field flux data was measured in Dawanqi oilfield [18]. Next, we built an analog experimental system to study migration mechanism of CH4 microseepage according to concept model. Finally, CH4 microseepage release manifestations and migration mechanism were discussed by analyzing the results of field work and physical simulations.

2. Geologic Setting

Dawanqi oilfield, located in the western part of the Kuqa-Baicheng depression (Tarim Basin), formed during the terminal stage of the Himalayan movement. The top-down strata of the Dawanqi oilfield (Figure 2(c)) are Quaternary (Q), Neogene Kangcun Formation (N1-2k), Jidike Formation (N1j), Paleogene Suweiyi Formation, Jurassic, and Triassic coal bearing [6] (see Supplementary Material (available here) for all abbreviations). Hydrocarbon reservoirs are relatively shallow (170–700 m) in the Quaternary to Neogene sandstones; oil and gas were generated in the Triassic and Jurassic coal-bearing formations [30, 31]. The thickness of oil bearing strata reaches 528.9 m, and the depth of groundwater table is 4.2 m [18, 32]. The area has highly faulted and fractured conditions. Three groups of fractures in the northeast (NE), northwest (NW), and eastwest (EW) directions cut the strata into blocks. Geochemical anomalies can be detected when oil and gas that is enriched in block reservoirs migrate to the surface along faults. Gas reservoirs have at least 89 vol% of thermogenic CH4 (δ13C: −18‰ to −38‰ Vienna Pee Dee Belemnite (VPDB)), with C2+ alkanes (8 vol%), N2 (2 vol%), and CO2 (0.5 vol%) [33]. During the fieldwork of this study, groundwater can be encountered 4.2 meters below surface.

3. Field Work and Physical Simulations

3.1. Flux Measurements

Field work took place in summer 2014 and 2019 and winter 2015 in the Dawanqi oilfield. The region is characterized by rare vegetarian, low land productivity, and severe climate. The annual temperature ranged from -20°C in winter to 30°C in summer [18]. Figure 2(d) shows that there are 124 flux measurement points across three MT1, MT2, and MT3 [16, 18].

Microseepage CH4 fluxes were performed with a portable laser-based gas analyzer (UGGA, LGR915-0011, USA; detection limit of 5 ppbv CH4 and 1σ precision of 0.6 ppbv) combined with a closed accumulation chamber (net volume of and effective height of 7 cm). Each flux measurement was based on accumulation times of ca. 20 min. The sampling interval along each transect varied from 50 to 300 m, depending on suitable ground conditions for installing the closed chamber. The control site was in an area located outside the petroleum field, 50 km from the field boundary.

3.2. Analog Experimental System

To study migration mechanism of CH4 microseepage, physical simulations were conducted in an intermedia-scale apparatus, which had internal dimensions of . Since migration of CH4 microseepage corresponds to many factors, such as dynamic systems, channel conditions, and geochemical shielding, it was not possible to consider all geological aspects within one apparatus. The main controlling factors (gas source, caprock, and migration channel) were chosen to establish a concept model (Figure 2(a)). In terms of migration channels, microfissures in strata were considered, and complex factors such as faults, unconformities, and strata tilt were excluded.

The cell was fitted with sampling ports at 20 cm intervals along the height of the cell. Layers 1–5 were packed with cement and quartz sand. In Figure 3(b), layer 1 and layers 2–5 represent the simulated caprock and overlying strata, respectively. The porosity, permeability, and breakthrough pressure of simulated caprock were 16.48%, 0.095 mD, and 1.56 MPa, respectively, equivalent to caprock V [34]. The height of saturation zone was about to 15 cm at layer 1. This layer provided a wet-sealing system which avoids gas emitted. In the top, 25 cm soil (layers 6 and 7) was supported by a stainless-steel frame.

A point source injection was used to replicate buoyancy-driven vertical migration, placed 6.5 cm from the bottom boundary (Figure 4). Natural gas was supplied from a cylinder, with batch composition of CH4 (89%), C2+ (4.4%), C3+ (2.3%), C4+ (1.7%), C4+ (0.2%), and N2 (2.4%), in accordance with a wet gas reservoir of the Kuqa oil-gas system. Gas pressure was measured at the injection point of cylinder and apparatus using a gas pressure meter connected to the gas injection tube, providing a constant gas flow during injection.

Based on the pressure program, the entire injection process was divided into three phases, namely, phase 1 (days 1-78), phase 2 (days 79-119), and phase 3 (days 120-290). The total gas pressure of Phase 1 and Phase 2 is 0.02MPa and 0.1MPa, respectively. Finally, phase 3 was initiated on day 120 with a total gas pressure of 0.2 MPa.

The sampling tube fitted on one end with sampling probe and the other end of the tube extending outside. 50 μL gas samples were extracted from the sampling tube for composition analyses. The CH4 concentrations in the gas samples were analyzed with an Agilent 6890 N gas chromatograph equipped with a flame ionization detector. The concentrations of CH4 were determined with a precision of 0.01 ppm.

4. Results

4.1. Release Types of CH4 Microseepage

During field measurements, release of CH4 microseepage can be categorized as three types, namely, continuous (Figure 5), flat (Figure 6), and episodic release (Figure 7). As shown in Table 1, continuous release mainly occurs near faults, with CH4 fluxes substantially higher than other release types. In continuous release, CH4 concentration elevates quickly and stably, at the rate of 0.007 ppm/second at 545-12 and by 0.285 ppm/second at 555-16 (Figure 5). In contrast, flat release refers to the case where CH4 concentration changes at nugatory rates (<10-4 ppm/second; see Figure 6(a)).

Episodic release is defined as spikes of CH4 concentration measurement (Figure 7), which occurs along faults or oil-gas area. Episodic release can be further divided into two types, i.e., the spike and the flat type. The spike type, shown in Figures 7(a) and 7(c), features spikes of CH4 concentration followed by rapid decline. The flat type, on the other hand, presents high CH4 concentration plateaued for more than 100 seconds, due to consistently high pressure (Figures 7(b) and 7(d)). Moreover, Figure 7 indicates that the interval for episodic release is ca. 2000 seconds, demonstrating that the episodic release is a stochastic event primarily due to the counterbalance between buoyancy and capillary pressure (see Section 5.1).

4.2. Physical Simulations of CH4 Microseepage

The initial pressure was set to 0.02 MPa for phase 1 of the physical simulation conducted (days 1–78), as the breakthrough pressure for microseepage was relatively low. The pressure was then increased to 0.1 MPa for phase 2 (days 79–119) and 0.2 MPa for phase 3 (days 120-190).

4.2.1. CH4 Concentration in Simulated Caprock and Strata

As shown in Table 2 and Figure 8(a), CH4 concentration gradually decreases from layer 1 to layer 3. As shown in Figures 8(b)–8(d), CH4 concentrations at all these layers remain relatively stable during phases 1 and 2 but rapidly increase from days 140 to 150 during phase 3, resulting high standard deviation (SD) variations. At layer 1, the SD substantially increases to 5827.44, accompanied by apparent episodic release. Note that the episodic release (days 142–145) was delayed from pressure increase at the start of phase 3 (day 120). This delay indicates that pressure increase does not directly lead to episodic release. Only after the buoyancy increases and breaks through the capillary pressure can episodic release occur.

4.2.2. CH4 Concentration in Soil

Figures 8(e) and 8(f) show that CH4 concentrations (8.55–1919.26 ppm) at layer 7 are higher than those at layer 6 (6.45–800 ppm). Several spikes can be seen on both layers, which are similar to the episodic release observed in the field measurements at 543-28-5 (see Figure 7(c)). The presence of microfissure or fault in simulated caprock or strata, as confirmed by scanning electron microscopic (SEM) imaging (see Figure 9), suggests that priority paths may exist and thus cause the spikes observed on layers 6 and 7. This mechanism differs from that for layers 1-3 at days 142-150 (namely, buoyancy; see Section 4.2.1). These findings also imply that failure to observe and record episodic release events may substantially underestimate CH4 fluxes.

5. Discussions

5.1. Release Mechanism of CH4 Microseepage in Field Work

Continuous release (Figure 5) is mainly distributed along faults, which form preferential pathway for the upward migration of CH4 microbubbles. CH4 migration along these channels is evidenced by several studies [3537]. Preferential pathway, which has the minimum resistance and maximum buoyancy, determines the direction of CH4 migration [38]. In addition, fault sealability can influence the migration and release mechanism of CH4 microseepage. Unoxidized CH4 can be released into the atmosphere and manifest as continuous release when faults extend to the Earth’s surface [39].

Robertson et al. advocate that diffusion or buoyancy may cause continuous or discontinuous gas migration [40]. Flat release shown in Figures 6(a) and 6(b) implies substantially lower CH4 migration rate than continuous release driven by buoyancy. Therefore, flat release is likely to be driven by diffusion, which is caused by concentration gradient, regardless of dynamic systems, porosity, permeability, or capillary pressure. In addition, diffusion plays essential role in oil and gas loss of shallow reservoirs. It is estimated that Dawanqi oilfield, which is geographically characterized with shallow hydrocarbon reservoirs, highly distributed fault systems, and poor cover conditions, lost 5.5%, 25.6%, and 77.9% of the reserves in the Quaternary (2.0 Ma), Pliocene (5.2 Ma), and Miocene (23.2 Ma), respectively, mainly due to diffusion [32]. However, some other researchers suggest that diffusion is not the main reason for gas migration [1, 41]. After all, reservoirs would not have formed or preserved if oil and gas could diffuse through the overlying strata and easily escape to the surface. Nonetheless, diffusion should not be ignored especially when CH4 migrates to a well-sealed area (such as mudstone). If there were no faults or microfissures in the overlying strata, CH4 can diffuse upward slowly and exhibit flat release on the surface. The release rate is higher than that in the control site (Figure 7(c)) but lower than in episodic and continuous release area.

Episodic release (Figure 7) occurs both along faults and oil-gas areas. Heterogeneous environments can cause lateral or pulsed gas transport in porous media [4244]. As CH4 trapped under caprock gradually accumulates, its buoyant force increases. Eventually, buoyancy overcomes the capillary pressure, so that CH4 breaks through the pore throat in caprock [4547], migrates upward, and causes microseepage [48, 49]. After that, the buoyancy decreases below the threshold, and hydrocarbons start to accumulate again. Another example of this process is the Old Faithful geyser in Yellowstone National Park in Wyoming, USA. In addition, episodic release is also caused by discontinuous gas flow. As shown in the scheme of flux measurement (Figure 10), CH4 floats to the top of the closed chamber (blue area) owing to low density. If the gas in the entire chamber had not thoroughly mixed, the top part rich in CH4 would be sampled, hence a concentration spike in the measurement. Subsequently, due to diffusion driven by concentration gradient, CH4 concentration measurement started to decrease and eventually returned to average concentration of CH4 accumulated in the chamber, which was still higher than the initial value measured at the beginning of sampling process. Barometric pressure and wind intensity may also influence CH4 release types on the surface [5052]. In the Tarim Basin where this study was conducted, the variation of the barometric pressure was relatively small; however, the velocity (0.3-5 m/s) and direction of the wind changed rapidly based on field measurements, whose impact on CH4 release manifestations observed remains to be clarified.

5.2. Migration Mechanism of CH4 Microseepage in Physical Simulations

In our previous study, Wang et al. have demonstrated that vertical migration of CH4 microseepage is driven by buoyancy at layers 1–3 and by diffusion at layers 4 and 5 in physical simulation [33]. Combined with field measurement, we therefore advocate that either buoyancy or diffusion will be the dominant driving force under different circumstances. Specifically, CH4 microseepage is mainly driven by buoyancy in intense tectonic deformation, highly fracture/microfracture zone, and poorly sealed channels; when CH4 migrates to highly sealed caprock or vadose zone, diffusion becomes the leading migration mechanism.

Admittedly, the bench-scale apparatus used in this work is much lower in pressure and smaller in scale than the actual oil reservoir, in addition to many other geological features that cannot be reproduced perfectly. Nonetheless, the physical simulation results of this work theorize the influence of buoyancy force, capillary pressure, and preferential pathway on release types of CH4 microseepage.

6. Summary and Conclusions

The main conclusion remarks of this study can be summarized as follows: (i)Three CH4 microseepage release manifestations were observed in field measurement: (1) continuous, (2) flat, and (3) episodic(ii)The results of physical simulations illustrate that these CH4 release manifestations can be attributed to various factors of CH4 migration, such as preferential pathway, buoyancy force, and capillary pressure(iii)It is theorized that in each CH4 release manifestation, its migration is subjected to different geological conditions and dominated by one specific mechanism of buoyancy or diffusion and thus differs from one another(iv)In continuous release, CH4 migrates along preferential pathway (microfissure or fault) at high speed and concentration, mainly driven by buoyancy. In flat release, CH4 migration is dominated by diffusion, causing its concentration to increase at a rate lower than continuous release but higher than control. In episodic release, CH4 travels along microfissures driven by buoyancy at highly fluctuating concentration(v)CH4 fluxes might be underestimated if it fails to record episodic release events

Surveys on geological CH4 fluxes over the past 20 years have confirmed their significant contribution to atmospheric CH4, but more thorough investigation on CH4 microseepage mechanisms, emission monitoring, and prediction models will be essential to improving the accuracy of quantifying greenhouse gas emissions in the future. The CH4 microseepage release manifestations and corresponding CH4 migration mechanism discussed in this paper are based on field monitoring data and physical simulations, which may differ from the actual geological environment. After all, the migration mechanism changes with the change of the geological conditions.

Data Availability

The data used to support the findings of this study are included within the article.

Disclosure

Part of the research was carried out at the Wuxi Research Institute of Petroleum Geology, Research Institute of Petroleum Exploration and Production, SINOPEC, Wuxi, China.

Conflicts of Interest

The authors declare that there are no conflicts of interest regarding the publication of this paper.

Acknowledgments

This work was supported by the Joint Funds of the National Natural Science Foundation of China (grants U2003101 and 41872126) and the Graduate Scientific Research Foundation of Hangzhou Dianzi University (CXJJ2021030 and 2022R407B059).

Supplementary Materials

All abbreviations within the article and their corresponding full names are listed in the table. (Supplementary Materials)