A New Method for Measuring Solvent Diffusivity in Heavy Oil by Dynamic Pendant Drop Shape Analysis (DPDSA)

Summary This paper presents a new experimental method and its computational scheme for measuring solvent diffusivity in heavy oil under practical reservoir conditions by DPDSA. In the experiment, a see-through windowed high-pressure cell is filled with a test solvent at desired pressure and temperature. Then, a heavy-oil sample is introduced through a syringe delivery system to form a pendant oil drop inside the pressure cell. The subsequent diffusion of the solvent into the pendant oil drop causes its shape and volume to change until an equilibrium state is reached. The sequential digital images of the dynamic pendant oil drop are acquired and digitized by applying computer-aided digital image-acquisition and-processing techniques. Physically, variations of the shape and volume of the dynamic pendant oil drop are attributed to the interfacial tension reduction and the well-known oil-swelling effect as the solvent gradually dissolves into heavy oil. Theoretically, the interfacial profile of the dynamic pendant oil drop is governed by the Laplace equation of capillarity, and the molecular diffusion process of the solvent into the pendant oil drop is described by the diffusion equation. An objective function is constructed to express the discrepancy between the numerically predicted and experimentally observed interfacial profiles of the dynamic pendant oil drop. The solvent diffusivity in heavy oil and the mass-transfer Biot number are used as adjustable parameters and thus are determined once the minimum objective function is achieved. This novel experimental technique is tested to measure diffusivities of carbon dioxide in a brine sample and a heavy-oil sample, respectively. It should be noted that, with the present technique, a single diffusivity measurement can be completed within an hour and only a small amount of oil sample is required. The interface mass-transfer coefficient at the solvent/heavy-oil interface can also be determined. In particular, this new technique allows the measurement of solvent diffusivity in an oil sample at constant prespecified high pressure and temperature. Therefore, it is especially suitable for studying the mass-transfer process of injected solvent into heavy oil during solvent-based post-cold heavy-oil production (post-CHOP).

This paper presents a new experimental method and its computational scheme for measuring gas diffusivity in heavy oil at high pressures and elevated temperatures by the dynamic pendant drop volume analysis (DPDVA). In the experiment, a see-through windowed high-pressure cell is first filled with a test gas at a prespecified pressure and temperature. Then, a heavy oil sample is introduced by using a syringe pump to form a pendant drop inside the pressure cell. Due to the oil swelling effect, the subsequent dissolution of the gas into the pendant oil drop causes its volume to increase until the saturation state is reached. The sequential digital images of the dynamic pendant oil drop are acquired and analyzed by applying computer-aided image acquisition and processing techniques to measure the oil drop volumes at different times. A mass-transfer model is developed theoretically to describe the diffusion process of the gas into the pendant heavy oil drop. This model is numerically solved by applying the semidiscrete Galerkin finite element method. The volume of the dynamic pendant oil drop is calculated from the numerically predicted transient gas concentration distribution inside the pendant oil drop. The gas diffusivity in heavy oil and the swelling factor of gas-saturated heavy oil are, thus, determined by finding the best fit of the theoretically calculated volumes of the dynamic pendant oil drop to the experimentally measured data. This novel experimental technique is applied to measure CO2 diffusivities in a heavy oil sample and the swelling factors of a CO2-saturated heavy oil at P = 2, 3, 4, 5, and 6 MPa and T = 23.9 °C.

In this paper, a new experimental technique is presented to study the solvent mass transfer in heavy oil and the resultant oil-swelling effect by applying dynamic pendant drop volume analysis (DPDVA). In the experiment, a pendant drop of heavy oil is formed inside a visual high-pressure cell, which is initially filled with a solvent (e.g. propane). As the solvent gradually dissolves into heavy oil, the volume of the pendant oil drop keeps increasing due to the oil-swelling effect. The sequential digital images of the dynamic pendant oil drop are acquired and analyzed to determine its volumes at different times. Theoretically, a previously formulated mathematical model is applied to describe the solvent mass transfer in heavy oil and the oil-swelling effect. The diffusion coefficient of the solvent in heavy oil and the oil-swelling factor are determined by finding the best fit of the theoretically predicted volumes of the dynamic pendant oil drop to the experimentally measured data. Experimental tests are conducted for the heavy oil-propane system at constant pressures of P = 0.4 − 0.9 MPa and a constant temperature of T = 23.9 °C. It is found that both the diffusion coefficient and the oil-swelling factor of the heavy oil-propane system increase with pressure. The major advantage of this new technique is that simultaneous measurements of the solvent diffusivity in heavy oil and the oil-swelling factor can be completed within two hours at the pre-specified constant pressure and temperature. Introduction In the vapour extraction (VAPEX) process, a solvent (e.g. methane, ethane, propane, butane or carbon dioxide) is injected into a heavy oil reservoir at a pressure close to its vapour pressure(1). Previous studies have already shown that molecular diffusion of the injected solvent in heavy oil plays a vital role in the VAPEX process(2–4). Thus the diffusion coefficient of the solvent in heavy oil under the actual reservoir conditions becomes an important parameter in the reservoir simulation and field design of the VAPEX process. In the literature, there are several experimental methods for measuring solvent diffusivity in heavy oil. These experimental methods can be roughly categorized into conventional and non-conventional methods. Conventional methods involve compositional analysis of liquid samples taken from the heavy oil-solvent mixture at different times and locations during a diffusion test(5). These methods are expensive, intrusive and time-consuming, especially if the diffusion test is conducted at a high pressure. In addition, compositional analysis of the heavy oil-solvent mixture is prone to large experimental error. Non-conventional methods measure the change of a property of the heavy oil-solvent system during the molecular diffusion process. This property can be the gas volume(6), oil-gas interface position inside a capillary(7), laser light intensity of the heavy oil-gas mixture(8), gas pressure(9) and shape of a pendant heavy oil drop surrounded by a gaseous solvent(10). In particular, the decaying gas pressure is measured while the molecular diffusion of the solvent into heavy oil proceeds within a closed high-pressure diffusion cell. This is referred to as the pressure decay method, which has been applied to measure the diffusivities of methane, ethane, propane, nitrogen and carbon dioxide in crude oils(11–14).

Enhanced solvent dissolution into in-situ upgraded heavy oil under different pressures

Assessing heavy oil composition of a green field, with an accepted level of uncertainty so it can be used for refinery capacity planning is a critical challenge. Knowledge of heavy crude oil properties is vital to understand potential adverse impact on process performance and total costs of the whole value chain, downstream (corrosion, catalyst deactivation, fouling and pumping) and upstream (corrosion, fouling, obstruction, producibility, lifting, pumping and transportation). The usefulness of heavy oil properties is highly dependent upon how samples are representing the bulk of hydrocarbon resources that will be developed and the reliability of the sampling procedure which becomes even more challenging when performed in a geologically complex, multilayered, supergiant green field with wide variations of fluid properties vertically and aerially. In this paper we present a field case with the methodology used to design, plan and execute a heavy oil sampling and essay for refinery capacity planning and the lessons learned, in an area of the field representing the first phase of development, in a supergiant heavy oil green field located in north of Kuwait. This heavy oil sampling and essay was planned and executed under high levels of uncertainty, using previous crude assays, PVT data from appraisal and thermal pilot wells and reservoir static and dynamic model. Extensive statistical analysis was applied to understand and map uncertainties related to sampling, completeness and representativeness of laboratory tests vs. accepted international standards. The methodology was applied by a multidisciplinary project team involving specialists from upstream and downstream who performed the work using project management tools to design and implement the execution of sampling in the field. The project took near one year to complete and results are now used for refinery capacity planning at short and midterm. Lessons learned were documented so the future sampling and assays can be improved as data from more wells is going to be available during execution of development phase.

The experimental study of effect of microwave heating time on the heavy oil properties: Prospects for heavy oil upgrading

A representative clean heavy oil sample (<1% BS&W) is crucial to obtain accurate fluid property data, which are required for heavy oil production, processing, and transportation designs. However, it is always a challenge to remove water from a heavy oil sample (<20° API) due to its near-water density (10° API) and high viscosity (slow separation). The objectives of this study include qualifying three heavy oil dewatering methodologies, determining their impact on heavy oil properties, and recommending suitable method(s) for preparing heavy oil samples for laboratory fluid phase behavior and property measurements. The three demulsification methods – centrifugation, the addition of demulsifiers, and distillation – were systematically evaluated and compared. Three dead heavy oil samples ranging from 20 to 11 API gravity degrees and with viscosities ranging from 100 to 5,000 mPa·s at 310.9 K (38°C) were selected. A stable emulsion with a 30 vol% synthetic brine were prepared using each heavy oil sample, which was used to evaluate each of the demulsification methods. Viscosity, density, and composition of the dewatered samples were measured and compared with the corresponding original oil data. This study demonstrated that centrifugation had the minimum impact on sample properties, although it was difficult to reduce the residual water content below the dewatering target of less than 1 wt%, especially for sample densities nearest the water density. The addition of chemical demulsifiers followed by mild centrifugation resulted in samples that met the dewatering target. The change in fluid properties caused by the demulsifier method was similar to the minor changes observed by the centrifugation method. Distillation could remove water from all three emulsions to below the dewatering target, if appropriate experimental procedures were followed. However, distillation had the strongest impact on dewatered oil properties due to the loss of light/intermediate components, despite precautions to minimize the mass loss. In current open literature, the effect of the heavy oil sample preparation procedure on the measured heavy oil properties and PVT data is rarely reported in detail, which makes it a challenge to properly perform the data interpretation and use them in relative modelling and design work. This simple, systematic study provides insight into the effect of three common dewatering procedures and guidance for developing reliable laboratory sample preparation protocols for heavy oils.

Experimental investigation of comparing electromagnetic and conventional heating effects on the unconventional oil (heavy oil) properties: Based on heating time and upgrading

Characterization of a heavy oil–propane system in the presence or absence of asphaltene precipitation

In this paper, a new experimental technique is presented to study the solvent mass transfer in heavy oil and the resultant oil-swelling effect by applying dynamic pendant drop volume analysis (DPDVA). In the experiment, a pendant drop of heavy oil is formed inside a visual high-pressure cell, which is initially filled with a solvent (e.g. propane). As the solvent gradually dissolves into heavy oil, the volume of the pendant oil drop keeps increasing due to the oil-swelling effect. The sequential digital images of the dynamic pendant oil drop are acquired and analyzed to determine its volumes at different times. Theoretically, a previously formulated mathematical model is applied to describe the solvent mass transfer in heavy oil and the oil-swelling effect. The diffusion coefficient of the solvent in heavy oil and the oil-swelling factor are determined by finding the best fit of the theoretically predicted volumes of the dynamic pendant oil drop to the experimentally measured data. Experimental tests are conducted for the heavy oil-propane system at constant pressures of P = 0.4 - 0.9 MPa and a constant temperature of T= 23.9°C. It is found that both the diffusion coefficient and the oil-swelling factor of the heavy oil-propane system increase with pressure. The major advantage of this new technique is that simultaneous measurements of the solvent diffusivity in heavy oil and the oil-swelling factor can be completed within two hours at the prespecified constant pressure and temperature.

Heavy crude oil upgrading using nanoparticles by applying electromagnetic technique

In this study, we use a custom-designed visual cell to investigate nonequilibrium interactions between liquid propane (C3(1)) and a heavy oil sample (7.2°API) at 55°C. The heavy oil sample is taken from Clearwater Formation located in Western Canadian Sedimentary Basin (WCSB). We inject C3(1) into the visual cell containing the heavy oil sample (pressure buildup process) and allow the injected C3(1) to interact with the oil sample (soaking process). After the pressure buildup process, we observe three phases in the visual cell: 1) heavy oil (0.67 mol), 2) liquid C3 (C3(1),0.60 mol), and gaseous C3 (C3(g), 0.20 mol). We measure visual-cell pressure and observe the C3-heavy oil interactions during the pressure buildup and soaking processes. Nonequilibrium interactions occurring at the interfaces of C3(1)-heavy oil and C3(1)-C3(g) are recorded over time. The results show that complete mixing of heavy oil with C3(1) occurs in two stages. First, upward extracting flow of hydrocarbon components from bulk heavy oil phase toward C3(1) phase form a distinguished layer (L1) during the soaking process. The extracted oil components become denser over time and move down (draining flows) towards the C3(1)-heavy oil interface due to gravity. The gradual color change of L1 from colorless to black suggests the mixing of hydrocarbon components from heavy oil. After L1 becomes homogenous, a second layer (L2) is formed at the upper part of the bulk C3(1) phase (above L1). Extracting and draining flows becomes active once again, leading to mixing of oil components from L1 into L2. At equilibrium condition, heavy oil and C3(1) are completely mixed and form a single homogenous phase.

During a solvent-based heavy oil recovery process, significant viscosity reduction is achieved through sufficient solvent dissolution and possible asphaltene precipitation. In the second viscosity reduction mechanism, the heavy oil is in-situ upgraded as the asphaltene content in the produced heavy oil is greatly lowered. In this paper, the respective and synergistic effects of asphaltene content and solvent concentration on heavy oil viscosity are studied by measuring viscosities of heavy oil samples with different asphaltene contents and/or solvent concentrations. More specifically, a deasphalted heavy oil sample is obtained by using the standard ASTM method to extract asphaltenes from a crude heavy oil. Then reconstituted heavy oil sample is prepared by adding the extracted asphaltenes into the deasphalted heavy oil at a different asphaltene content each time and its corresponding viscosity is measured at the atmospheric pressure. It is found that the viscosity of such prepared heavy oil is sensitive to the asphaltene content. In particular, the viscosity of the deasphalted heavy oil is reduced by 13.7 times in comparison with that of the original heavy oil at T=23.9°C. This indicates that high asphaltene content results in high viscosity of the heavy oil. On the other hand, the viscosities of propane-saturated heavy oil samples with different asphaltene contents are measured at high pressures. As the equilibrium propane concentration reaches certain value, the viscosity of the crude heavy oil-propane system is reduced by almost two orders. This result shows that solvent dissolution plays a dominant role in heavy oil viscosity reduction. When the equilibrium propane concentration is high enough, all the three reconstituted heavy oil-propane systems have extremely low viscosities. In this case, effect of asphaltene content on heavy oil viscosity becomes negligible.

Experimental measurement and parametric study of CO2 solubility and molecular diffusivity in heavy crude oil systems

In this study, we use a custom-designed visual cell to investigate nonequilibrium interactions between liquid propane (C3(l)) and a heavy oil sample (7.2 deg API) at varying experimental conditions. We inject C3(l) into the visual cell containing the heavy oil sample (pressure-buildup process) and allow the injected C3(l) to interact with the oil sample (soaking process). We measure visual-cell pressure and visualize the C3/heavy oil interactions during the pressure-buildup and soaking processes. Nonequilibrium interactions occurring at the interfaces of C3(l)/heavy oil and C3(l)/C3(g) are recorded with respect to time. The results show that the complete mixing of heavy oil with C3(l) occurs in two stages. First, upward extracting flows of oil components from bulk heavy oil phase toward C3(l) phase form a distinguished layer (L1) during the soaking process. The extracted oil components become denser over time and move downward (draining flows) toward the C3(l)/heavy oil interface due to gravity. The gradual color change of L1 from colorless (color of pure C3(l)) to black suggests the mixing of oil components with C3(l). After L1 appears to be uniform, a second layer (L2) is formed above L1 in the bulk C3(l) phase. Extracting and draining flows become active once again, leading to the mixing of oil components from L1 into L2. At final conditions, heavy oil and C3(l) appear to be mixed and form a single uniform phase.