The effect of grainsize of sediments in the CO₂/CH₄ replacement process within a hydrate lattice: An experimental report

Highlights

• The effect of sediment grainsize on the CO₂/CH₄ replacement process was tested.

• Two lab-scale twin reactors were used for experiments.

• A lower size of grains led to a faster and more abundant formation of CH₄ hydrates.

• Conversely, the larger size of grains favoured the replacement process.

• Better results in terms of CH₄ recovery and CO₂ capture were achieved.

Abstract

The effects of natural parameters and their related properties, on the formation and dissociation of gas hydrates, have been widely investigated and satisfying information can be found elsewhere in literature. Conversely, there is a substantial lack of similar studies related to CO₂/CH₄ replacement processes. This research aims to contribute in this sense and shows replacement tests in presence of two pure silica porous media, whose only difference stays in the size of grains. The results confirmed what asserted in literature for the simple formation of hydrates and provided new insights about the replacement process. It was found that the lower size of grains favoured the formation of hydrates and their stability. Precisely for this reason, the fine sediment hindered the replacement, because the higher stability of methane hydrates reduced the capability of CO₂ molecules to penetrate within the sediment and take the place of methane molecules. This study proved once again that, natural elements capable to act as promoter for the formation of hydrates, often have an inhibiting effect on the replacement processes.

Introduction

Natural gas hydrates consist of ice-like crystalline structures, where gaseous molecules, referred as “guest” are physically trapped within solid cages made with water molecules (also called “hosts”) [1]. These structures were accidentally discovered in 1778 and their existence was officially proved in 1810 [2]. However, only in the mid-1960 it was confirmed the existence of enormous naturally-occurring deposits of gas hydrates diffused worldwide [3]. To date, natural gas hydrates are known as the most abundant available energy source [4]. These compounds were mainly discovered in continental margins and slope sediments (up to 97%) [5,6], where their distribution is strongly dependent on the specific marine geological features [7,8], and in permafrost regions (approximately 3%) [9,10]. The most abundant offshore hydrate reservoirs were discovered in the South China Sea, Gulf of Mexico, Sea of Japan, South Korean Sea, Bearing Strait and Indian Ocean, while the most important onshore deposits were found in the permafrost regions of Alaska, Mackenzie Delta, Siberia and Qinghai-Tibet Plateau [11]. The storage capacity of gas hydrate ranges from 160 to 180 Nm³gas/m³hyd, thus ensuring high energy density in the presence of methane as guest. According to the current estimations, up to 3*10¹⁵ m³ of methane are present worldwide under form of hydrate [3].

However, the current technologies, available for the recovery of methane, are still far from making the process competitive and completely sustainable [12]. The most relevant problems can be found in the location and morphology of deposits, difficulties associated to the heat and mass transfer through the layers of the reservoir, environmental concerning and costs required for their exploitation. These critical issues motivate the need of further scientific efforts to better understand the formation and dissociation mechanisms of gas hydrates and the influence of external elements, such as water composition, properties of the sediment, etc., on the process. The formation of hydrates comprehends two distinguished phases: the initial nucleation and the catastrophic growth phase [1]. The first phase consists of the formation of the first clusters of hydrates, which than continue their growth via collisions with the surrounding clusters and consequent formation of bigger agglomerates. As soon as the so-called “critical size” is reached, the cluster becomes capable to grow spontaneously, without the need of further collisions. The following growth of hydrates represents the catastrophic growth phase and is function of the presence of host and guest molecules and the presence of suitable thermodynamic conditions. The critical size is obtained when the system reaches the maximum of Gibbs free energy; it represents a key element for the process because separates the first part, where the formation is stochastic and cannot be accurately predicted, from the second part, whose trend can be defined as a function of the process conditions [13]. The stochastic nature of the nucleation phase, mainly depends on the probability that labile clusters have to form and the probability of their growth, which is associated to the collision with similar entities, diffused in the bulk phase.

As a consequence of it, the formation of the first nuclei depends diffusion of guest molecules within the sediment, the occurrence of enough diffused and extended gas–liquid interfaces, the presence of elements capable to promote the formation of clusters (also called “nucleation sites”). All these parameters are strongly related to the properties of the sediment hosting the formation process [14,15]. The intimate contact between gas and water molecules is particularly affected by the permeability and the porosity of the sediment and the promotion of hydrate nucleation depends on the morphology and the chemical composition of grains and rocks. The strong connection between gas hydrates formation and the properties of sediments hosting the process, allowed to the production of numberless theoretical and experimental studies on the formation and dissociation of gas hydrates in presence of different typologies of sediments [16], [17], [18].

Hydrates form and deposit in sediments and may assume several different morphologies, mostly as a function of the type of gas source [19].

Segregated natural gas sources can provide abundant quantities of gas for hydrates production and lead to the production of hydrates in massive, nodular, vein and bedded morphologies [5,20,21]. Differently, disseminated and limited quantities of natural gas lead to gas hydrates in grain-enveloping, pore-spacing, cementing and frame-supporting morphologies [22], [23], [24]. The production of gas hydrates on lab-scale facilities, is carried out by following four possible strategies: (i) the mixed sample preparation; (ii) the excess gas; (iii) the excess water and (iv) the dissolved gas method [5]. The mixed sample preparation method consists of preform methane hydrates in a high-pressure vessel and then mix them with sediment to form the hydrate-bearing sediment [25]. The excess gas technique is base on injecting gas on a preformed mixture of water/ice and sediments [26]; if the pore gas is then replaced with pore water, the excess water method is obtained [27,28]. Finally, the dissolved gas strategy consists of injecting the guest compound in the pore space of a fully saturated sediment [29,30]. Cementing and grain-enveloping morphologies are often obtained with the excess gas and excess water methods [31], while pore-spacing and frame-supporting morphologies are obtained with the mixed sample preparation and the dissolved gas methods [32]. This latter morphology always requires high saturation levels [33]. Natural gas hydrates reservoirs were found to exist in a wide range of different sediments, like medium, fine and very fine sand, coarse silt, coarse sand and gravel [34]. It has been proved that the hydrate saturation level is higher in coarser sediments than in finer sediments [35]. Oshima and colleagues studied the deposits of hydrates in the Eastern Nankai Trough [35]. They found that the saturation level was higher in the sandy layers than in the muddy layers. Similarly, the studies carried out on the Alaskan North Slope reservoirs, revealed that the saturation level is higher in the sand and coarse silty layers than in the fine silt and clayey layers [36]. Natural gas saturation is a key parameter for gas hydrates formation. It is function of several parameters, such as the rate of methanogenesis, porosity and permeability of the sediments and advection rate of gas [37]. The saturation level strongly depends on the grain size and surface area of sediments [36]. The saturation degree is a crucial parameter because up to 90% of natural hydrates reservoirs have been discovered in fine-grained sediments, which are currently considered the most promising typology of sediment for the production of gas from hydrates and the efficiency of the process is proportional to the saturation degree [36].

Since 1980, a new promising solution, to improve the efficiency of methane recovery from hydrate-bearing sediments, was proposed: it consists of the replacement of methane molecules with carbon dioxide, into pre-existing water cages [38]. This strategy allows to reach three key targets at the same time: the improved recovery of methane, the storage of carbon dioxide in permanent form and the preservation of sediments, thank to the possibility of avoiding the dissociation of water cages [39]. The feasibility of the CO₂/CH₄ replacement process depends on two main factors: both the types of guests form the same typology of structure, or the Cubic sI Structure [1]; moreover, at the same thermodynamic conditions, the enclathration of carbon dioxide is favoured over methane [40]. The reason stays in the lower enthalpy of formation, which is equal to-57.98 kJ/mol for carbon dioxide hydrates, while is equal to-54.49 kJ/mol for methane hydrates. These advantageous conditions lead to the existence of a narrow thermodynamic region in which the replacement process is thermodynamically favoured and spontaneously occurs [41]. Theoretically, the exchange ratio between the two types of molecules is equal to one; however, this value cannot be spontaneously reached. The sI structure contains two different cavities, the relatively small 5¹² cage and the relatively large 5¹²6² cage. The molecule of methane is capable to fit both of them, while the molecule of carbon dioxide, due to its larger size, encounters more difficulties in fitting the small cavities. For that reason, the maximum efficiency achievable is equal to 75% [42]. Moreover, in field applications, the real efficiency is often lower, due to the action of external variables (such as the presence of salts and ions dissolved in water, capable to reduce the activity of the host molecules, low permeability of the reservoir, ice formation, methane hydrates reformation, and so on) on the process. In this sense, the physical and chemical properties of the sediment can strongly affect the efficiency of the replacement process.

However, while the properties of sediments and their effect on the methane hydrates formation and dissociation processes, have been widely explored [43], [44], [45], [46], similar studies have been only occasionally carried out for the CO₂/CH₄ replacement process and there is a substantial lack of information in literature.

This work aims to provide useful information to understand how the geometrical properties of the sediment can intervene on the CO₂/CH₄ replacement. This process is mainly based on modifying the local conditions of the reservoir, in order to generate instability for methane hydrates, while keeping the condition suitable for the formation and stability of carbon dioxide hydrates. However, due to the relatively little difference between the phase boundary equilibrium conditions of the two respective species and due to practical difficulties in intervening with high accuracy in the natural reservoirs, the replacement process is often carried out at local conditions close to the ones required for methane hydrate stability.

As a consequence of it, if the sediment is capable to provide high stability for methane hydrates, the efficiency of the replacement process could be drastically lower than that commonly highlighted with lab-scale experiments. Previous researches already proved that the properties of sediment may alter the formation of dissociation conditions of gas hydrates and, in some cases, the entity of such variation is function of the gaseous species involved in the process [47,48].