Huang, Hai (Xi'an Shiyou University and Shaanxi Key Laboratory of Advanced Stimulation Technology for Oil & Gas Reservoirs) | Babadagli, Tayfun (University of Alberta) | Andy Li, Huazhou (University of Alberta) | Develi, Kayhan (Istanbul Technical University)
The fracture-surface characteristics (such as roughness and fractal dimensions) may greatly affect the proppant transport during hydraulic fracturing operation. Few researches have focused on investigating the proppant transport in vertical fracture with actual surface characteristics. As a continuation of our previous study (
This paper reports the results of laboratory scale screening of different chemicals for their microeumlsion generation capabilities to be eventually recommended for non-thermal heavy oil recovery (chemical flooding). The study was performed through visualization of microemulsions generated using vials and microscopic images. The impact of salinities of brine on the emulsification was studied thoroughly in order to identify the synergy between the selected chemicals and the heavy oil. An alcohol propoxy sulfate surfactant from the Alfoterra series, Alfoterra S23-7S-90, a nonionic surfactant HORA-W10, gave good emulsion formation results at low salinity conditions (2.5 wt. %, 3.8 wt. %). Polysorbate-type nonionic surfactant Tween 20 gave good emulsion formation at high salinity conditions (6.35 wt. %, 7.6 wt.%). Their emulsion formation performance with a crude heavy oil of viscosity 4,812cP and 11.74 °API helped create an initial correlation of performance with the composition of crude oil and synthetic brine samples of various salinities. Attempts were also made to stabilize oil-in-water emulsions formed with Alfoterra S23-7S-90, HORA-W10, Tween 20 using nanofluids (metal oxides), sodium carbonate, and an anionic polyacrylamide-based polymer (PolyFlood MAX-165). Emulsions were visualized under the Axiostar plus transmitted-light microscope and their stability was studied in order to screen the most optimal chemical (or chemical combinations) available for low cost heavy oil recovery.
Primary recovery of heavy-oil is remarkably low due to high viscosity and low energy by solution gas exsolution to drive the oil. Gas injection to improve foamy flow and also to dilute the oil in such reservoirs has been proposed as a secondary recovery method. However, because of the high costs of injected gases, efforts are needed to optimize the process by selection of proper gas type (or gas combinations) and suitable injection scheme. To achieve this goal, an experimental procedure was followed with rigorous analyses of the output. A 1.5 m long and 5 cm diameter sand-pack was first saturated with brine, which was replaced with dead oil. Then, gas solvents were injected to dead-oil containing core-holder until nearly reaching 500 psi followed by a two-day soaking period. Pressures all along the sand-pack were recorded with eight pressure transducers. Different combinations of various gas solvents (methane, CO2, and air) aiming to select the most competitive and economic formula were tested with a certain set of pressure depletion rates.
The physics of the foamy oil flow for different solvent mixtures and depletion conditions were analyzed using pressure profiles acquired, recorded oil/gas data with time, and gas chromatography and SARA analyses of the produced gas and oil. Three huff-n-puff cycles were applied. Compared with other light hydrocarbon solvents and carbon dioxide, air has its high advantage in terms of accessibility and lowered cost. Hence, attention was given to air that was mainly used to pressurize the system and increase oil viscosity due to oxidation process with an expectation of better foam quality when injected with other gases such as CO2 and methane. Methane (CH4) yielded the quickest response in terms of gas drive but, in the long run, CO2 was observed to be more effective technically. Air was observed to be effective if mixed with CO2 or methane from an economics point of view. To sum up the results, air Huff-n-Puff (HnP) followed by 2-cycles of CH4 HnP yielded 36.21% recovery, while air HnP followed by 2-cycles of CO2 HnP delivered 30.36% oil. When the gases are co-injected, air 50%-CO2 50% and air 50%-CH4 50% recovered 29.85% and 23.74% of total oil-in-place, respectively.
As stated by the classical Thomson equation, the pore scale thermodynamics of solvent is different from bulk conditions being critically controlled by capillary characteristics. This equation shows that the boiling points decrease remarkably as the pore size and interfacial tension become smaller. This paper investigates this phenomenon for hydrocarbon solvents experimentally and compares the results with the values obtained from the Thomson equation to test its applicability in modelling heavy-oil recovery by solvents under non-isothermal conditions. As an initial step, the boiling temperature of different single component solvents (heptane and decane) was measured by saturating Hele-Shaw type cells with variable apertures (ranging from 0.04 mm to 5 mm) and monitoring the boiling process. One experiment was run with a thickness of 12 mm to represent the bulk case. As the aperture (pore size) became smaller, the boiling point temperature decreased. For example, the measured boiling temperatures of heptane and decane were approximately 57.7°C and 107.4°C for the aperture values less than 0.15 mm, which were considerably lower than the "bulk" values (around 40%). In the next step, the same experiments were repeated using micromodels representing porous media. The micromodel (grain diameter of 0.15 mm and a pore throat of 0.075 mm) was designed with uniform properties (constant grain diameter and pore throat). By using the Thomson equation, the boiling points of the selected liquids were mathematically computed and compared with the experimental results from Hele-Shaw experiments.
The roughness of fractures may play an important role in affecting the migration and placement of proppants during hydraulic fracturing operations. Previous studies focused on investigating the proppant transport in smooth vertical fractures, which did not consider the effect of the fracture-surface roughness. We examine the migration of proppants in rough and vertical fractures and then quantitatively reveal the effect of roughness on the instantaneous proppant transport and final proppant placement. Two types of rock samples (marble and granite) are fractured with the Brazilian test and molded to manufacture 20 × 20 × 5 cm transparent replicas. The surface roughness of these rock samples was first characterized by fractal dimensions. Then, the dyed fracturing fluid with a given proppant loading was injected into the rough vertical fracture. In each test, the inlet pressures were continuously monitored in order to obtain the differential pressure across the fracture model while the proppants were being transported in the fracture. The process was videotaped to real-time track the proppant distribution in the rough fracture.
The proppant-transport behavior in the rough and vertical fracture was observed to be totally different from that in the smooth fracture. The major experimental findings include the following: 1) The proppant in a rough vertical fracture does not progress as a regular sand bank that commonly occurs in the smooth fracture, but rather an irregular-shape sand clusters with fractal characteristics; 2) In the rough and vertical fracture, the phenomenon of proppant bridging is visually observed, and such phenomenon is more likely to occur in the location with a larger roughness height. This implies rough fracture could promote a wider spreading of the proppant in the fracture compared to smooth fractures, and; 3) The existence of roughness enhances the vertical displacement of fluid containing proppants. These effects are also favorable for obtaining a better filling of the proppants in the fracture. Our experimental study reveals the mechanisms of proppant transport and distribution in real vertical fractures under the influence of roughness effect.
Late cycles of cyclic steam stimulation (CSS) are characterized by a decreasing heavy-oil recovery and an increasing water cut. Nickel nanoparticles can be used to promote aquathermolysis reactions between water and heavy oil in steam-injection processes, thereby increasing the recovery factor (RF). In this paper, detailed investigations were performed to determine the optimal operational parameters and answers to the following questions:
CSS experiments were conducted between temperatures of 150 and 220°C. Steam generated under these temperatures was injected into sandpacks saturated with Mexican heavy oil. Powder-form nickel nanoparticle was introduced into this process to boost the oil production. In an attempt to obtain the optimal concentration, different concentrations were tested. Next, oil sands without any nanoparticle additives were first added into the cylinder. Then, only one-third of the sandpack was mixed with nickel nanoparticles near the injection port. Experiments were executed to study the effects of temperature, nickel concentrations, and nanoparticle-penetration depth on the ultimate oil recovery and produced oil/water ratios after each cycle. Produced-oil quality and emulsion formation were evaluated with gas-chromatography (GC) analysis, viscosity measurements, saturates/asphaltenes/resins/aromatics (SARA) tests, and microscopic analysis of the effluents.
Experimental results show that the best concentration of nickel nanoparticles, which gives the highest ultimate oil RF, is 0.20 wt% of initial oil in place (IOIP) under 220°C, whereas the nickel concentration of 0.05 wt% provides the highest RFs at the early stages. A lower temperature of 150°C provides a much-lower RF than 220°C, which is mainly because of a lower level of aquathermolysis reactions at 150°C. By analyzing the compositions of produced oil and gas samples with GC and SARA tests, we confirm that the major reaction mechanism during the aquathermolysis reaction is the breakage of the carbon/sulfur (C/S) bond; the nickel nanoparticles can act as catalyst for the aquathermolysis reaction; and the catalytic effect becomes less remarkable from cycle to cycle. One run of the experiment to test the effect of particle-penetration depth revealed that the nickel nanoparticles distributed near the injection port greatly contributed to the ultimate RF.
Capillary imbibition tests are commonly applied to measure wettability-alteration potential of chemicals. However, these tests are exhaustive, time-consuming, and expensive, and the underlying physics of the alteration process from a surface-chemistry point of view is often limited and/or unexplained. Contact-angle measurement is a quicker and more-feasible screening tool to assess the emerging wettability modifiers. It also provides visual data on the mechanics of the wettability-alteration process. This paper focuses on contact-angle measurements as a means to evaluate the wettability alteration on mineral plates and porous-rock samples. Imidazolium ionic liquids were tested at different concentrations. To study the effect of pH on the wettability, sodium chloride and sodium borate were used at different concentrations. The composition of divalent ions was varied because of their possible use with low-high-salinity water as wettability-alteration agents. Unmodified and surface-modified silica, zirconium, and alumina nanoparticles were also tested.
Contact-angle measurements were performed initially on mica, marble, and calcite plates. Experiments were repeated on polished surfaces of Berea sandstone, Indiana limestone, and cleaned Grosmont carbonate cores. Oils (pure and solvent-mixed crude oils) with different viscosities and densities were used to test the effect of oil type on the process. The images were obtained by an single-lens reflex (SLR) camera at different temperatures ranging from 25 to 80°C. By testing with different concentrations, the optimum chemicals were found for different mineral-plate/porous-rock systems. Then, the results were cross checked with the imbibition tests performed on the same samples to validate the contact-angle-measurement observations.
Thermal-stability tests were also performed in case of their use during or after a thermal method. For the thermal-stability tests, contact-angle experiments were conducted in a high-pressure andhigh-temperature (up to 200°C) cell. It was shown that certain ionic liquids and nanofluids are stable at high temperatures and can be efficiently used at low concentrations.
Our recent experimental studies on superheated solvent injection for heavy-oil recovery showed that when a solvent is injected into the reservoir, the process is highly sensitive to pressure and temperature. The effects of these parameters on the recovery factor (RF) are accentuated when the operating conditions are closer to the saturation curve of the solvent injected. This paper investigates this process and formulates the optimal field-scale application conditions that yield the maximum profit, as a continuation of previous work. To achieve this, a hypothetical field-scale numerical model was constructed, and the key parameters identified through the aforementioned sensitivity analysis were incorporated. Then, the injection process was simulated for a two-horizontal injection/production pattern. An optimization study was performed to identify the relative contributions of the effective parameters (pressure, temperature, and injection rate) and to propose an optimal application scheme with a genetic algorithm (GA). The critical pressure and temperature yielding maximum production and highest profit considering solvent retrieval were defined for different injection rates and application scenarios. Our results indicate that, at the end of the hot solvent-injection process, an important volume of solvent is left in the reservoir, and its volume depends on the injection–production scheme selected. Nevertheless, if the project is performed under appropriately selected operational parameters (obtained through the optimization processes) and followed by the proper process to retrieve the solvent from the reservoir (low-temperature steam or hot-water applications), it can make the hot solvent-injection process profitable.
Reservoirs containing very heavy oil or extremely heterogeneous/fractured geology are not convenient for steamflooding and even cyclic steam injection. Then, steam can be used to heat the reservoir and accelerate the recovery by gravity drainage. Two well-known applications of this method are steam assisted gravity drainage (SAGD) and thermally assisted gas oil gravity drainage. Although the latter is not commercially applied, the former is a proven technology with remarkable production in Canada and Venezuela. Due to the risks caused by the cost and solvent retention, no large scale applications of solvent injection with steam have been implemented. An alternative is to use chemicals as suggested a few decades ago to alter the interfacial forces and improve microscopic displacement. This paper presents experimental results on testing -new generation- chemicals for their capability in recovery improvement.
Sandpack experiments were conducted to evaluate the incremental in oil recovery by chemical additives compared to sole steam injection. Steam and chemicals were heated and introduced to the system from separate channels at the entrance of the vertically situated sandpack (30 cm long, 5 cm in diameter). To generate a purely gravity dominated system (pressure differential of 10–25 psi) a back pressure regulator was used. The chemicals used include thermally stable surface agents, such as surfactants (AAS J1111, O352, LTS-18), Tween 20, biodiesel), ionic liquid (BMMIM BF4), high pH solution (NaBO2), solvent (heptane), and nanoparticles (SiO2). The oil selected was 20,000 cp crude.
Incremental recoveries were monitored and related to the thermal stability of the chemicals. A comparative analysis was provided as to their contribution to the reduction of the cost (less steam and lower temperature) and chemicals were classified based on their recovery improvement performance and thermal stability. Through this experimental schematic, the highest increment in oil recovery was achieved by LTS-18 but also combined a high duration of the experiment with a high water consumption. This reduces the result in economical favorable conditions of the LTS-18. Biodiesel had the best performances in steam-to-oil ratio (SOR) and its effects needs to be further investigated. Tertiary injection of hot water with a surfactant was inefficient. Ionic liquid increased the oil recovery in the tertiary stage after the core was flooded with a low quality steam by 20%.
Over a 6-month time frame in 2016, I was able to attend three SPE conferences on heavy oil in different countries spanning three continents (Canada, Peru, and Kuwait). Despite regional differences in the applications, potentials, problems, and technological needs, the common theme in all conferences was “low cost.” Cost optimization in heavy-oil production was discussed from technical and economic perspectives, not only in the technical sessions but also in numerous panel discussions.
Such optimization can be achieved through numerical modeling to suggest general optimal strategies and development plans or by using proper real-time data acquisition (production optimization) for prompt decisions while operations are ongoing. This requires continuous monitoring of the processes as seen in many steam-assisted-gravity-drainage operations or other types of steam-injection applications. I selected two papers about advanced monitoring techniques as suggested reading in this issue.
Moreover, chemical and nanomaterial additives to water and steam have received a great deal of attention. Low-interfacial-tension (microemulsion) and low-salinity injection in heavy oils in sands and carbonates and wettability alteration in carbonates were common topics at conferences held over the past year. I selected one review paper for additional reading and one experimental work as a summary paper on this subject. Apparently, modeling efforts on advanced (but unconventional) technologies such as electromagnetic heating have continued. You will find a detailed mathematical analysis of the process in one of the papers summarized.
Despite the recent economic downturn, we were able to hear the outcome of current field practices at pilot or demonstration scale. Papers detailing small-scale cyclic-steam-injection applications in Kuwait and Oman were worth reading, and one article on this specific subject is included here. Considering these activities in the Middle East, effective transfer of technologies from North America to that part of the world will become highly critical in the near future.
Recommended additional reading at OnePetro: www.onepetro.org.
SPE 181160 State-of-the-Art Review of the Steam Foam Process by Eric Delamaide, IFP Technologies Canada, et al.
SPE 180732 An Integrated Probabilistic Work Flow for Primary and Thermal Performance Prediction of a Large Extraheavy-Oil Field by Raushan Kumar, Chevron, et al.
SPE 181431 Horizontal Steam-Injection Flow Profiling Using Fiber Optics by Mahdy Shirdel, Chevron Energy Technology Company, et al.SPE 180726 SAGD Production Observations Using Fiber-Optic Distributed Acoustic and Temperature Sensing: SAGD DAS—Listening to Wells To Improve Understanding of Inflow by Warren MacPhail, Devon, et al.