136879 research outputs found
Sort by
Inoculant Rate and Nodulation on Peanut and suggestions for proper handling and application of Rhizobium Inoculants
Get PDF
Ionic Surfactant Systems for EOR Applications in High-Temperature Shale Reservoirs
Get PDFPrevious experimental studies and limited field results have shown that the application of surfactants can be effective in improving recovery from unconventional liquid-rich reservoirs. Most of these studies and field applications have been conducted at temperatures below 200���. The application of surfactants for EOR has not been evaluated at higher temperature conditions. However, many of the potential EOR candidates in the lower 48 are reservoirs that have temperatures greater than 200���. For instance, temperatures greater than 320��� in the Eagle Ford liquid-rich window have been reported. Therefore, this study aims to design surfactant systems that are thermally stable and effective for high-temperature EOR applications in shale reservoirs.
This study provides a complete workflow for surfactant selection for high-temperature EOR applications and presents guidelines for designing surfactant systems for these high-temperature conditions. Single ionic surfactant systems, as well as novel blends of ionic surfactants, were evaluated. The behavior of the different ionic surfactant systems was investigated through contact angle (CA) and interfacial tension (IFT) measurements for a wide range of temperatures up to 350��� and pressures of up to 5000 psia. These measurements were performed with crude oil and oil-saturated rock samples obtained from a high-temperature Eagle Ford reservoir. Surfactant systems that were effective in altering wettability favorably at high-temperature conditions were further evaluated by Surfactant Assisted Spontaneous Imbibition in order to quantify their impact on improving oil recovery. Surfactant aqueous stability, time stability, and solubility were also investigated.
CA and IFT measurements reveal that temperature has a significant impact on the effectiveness of surfactant systems. At low-temperature conditions, some ionic surfactants were effective in altering wettability favorably. However, at increased temperatures, ionic surfactants showed limitations in altering the wettability favorably above a certain temperature limit tagged the ionic surfactant temperature limit (ITL). To ensure the applicability of surfactants at high-temperature, the ITL has to be above the reservoir temperature. Means of improving the ITL, as well as designing surfactants with higher ITL, became the paramount objective. Furthermore, it was revealed that the use of higher surfactant concentration and higher salinity systems improved the ITL, thereby making surfactant systems more effective for high-temperature applications. Optimal surfactant molecular structures to improve the ITL of the surfactant system are described. Novel co-surfactants systems, which are blends of cationic and zwitterionic surfactant systems, showed the most promising results for high-temperature applications with recovery factors up to 22.6% OOIP.
This study establishes a foundation in understanding the behavior of surfactant molecules at high-temperature conditions and unlocks the potential for improving oil recovery significantly in high-temperature unconventional reservoirs
Path Planners and Controllers for Autonomous Navigation of Robots in Unstructured Domains
Get PDFAutonomous navigation of robots involves two components (i) Perception and (ii) Navigation. Navigation involves performing path planning, motion planning and dynamic control in a sequence utilizing the information from perception. Path planning involves generating a path that the robot should traverse along to accomplish the mission. Motion planning utilizes these waypoints to come up with motion primitives for the robot to navigate. Dynamic controls then utilizes the motion primitives to select appropriate actuator commands to actuate the robot. This dissertation proposes solutions addressing path planning, MCAR (Meta Reinforcement-Learning based Contextual Adaptive-Control using Representations) and motion planning, RLP (Reshaping Local Path Planner).
Control of non-linear dynamic systems in the presence of unknowns from a nominally narrow distribution may be solved using classical control paradigms or learning based paradigms. Classical control paradigms employ feedback control strategies. Learning based paradigms employ data-driven techniques and strategies for controlling the systems. However, when the unknowns are drawn from a nominally broad distribution, the above techniques may not provide acceptable control performance or acceptable adaptation across the space of unknowns. Classical techniques require an extensive engineering effort to model the behavior of the unknowns with the dynamic system. Learning based techniques require a large amount of data and sophisticated algorithms to capture the data representation. While dealing with unknowns from a broad distribution, classical controls typically employ adaptive control techniques and learning based controls employ the scheme of ���learning to learn��� or more broadly meta-learning techniques to adapt between the unknown tasks. Adaptive control schemes utilize adaptation laws that leverage the observation of the variation of states, inputs and observations to pick a suitable adaptation law. The adaptation law is then utilized to modify the classical control policy. Meta Learning on the other hand utilizes two schemes i.e model-agnostic meta learning (MAML) paradigms and task-inference paradigms that both leverage a bi-level optimization scheme (for policy and adaptation). The proposed control approach in this thesis, MCAR utilizes the in-built knowledge of the dynamic system, along with adaptation law estimates and a structure of the system unknowns obtained using a history of states and inputs. We demonstrate the working of this technique on a class of dynamic systems that represent an ackerman drive car along different surfaces of varying traction (from asphalt - a surface of high traction to ice - a surface of low traction) for the task of motion-planning. We deploy the algorithm on a series of simulation scenarios to demonstrate the working of MCAR. With the initial part of the thesis focusing on motion planning of robotic vehicles, the next part focuses on robust path planning techniques to generate a modified local path given a desired global path that maybe temporally stale, may lie in the presence of obstacles and may not respect the constraints of minimum turning radius of the robotic vehicle.
Techniques to solve this problem of path planning include classical paradigms like potential field methods (PFMs) which rely on rolling-out a local path minimizing a ���virtual potential function��� that lie along the path, active exploration methods (AEMs) that rely on searching for the best possible actuation scheme to navigate the RV and learning based paradigms which utilize a learned understanding of the path and the obstacle configuration to generate a local path. PFMs suffer from stagnation issues where the RV can get struck in convex obstacle configurations. AEMs suffer from a large actuation search space. Learning based techniques suffer from a lack of repeatability and a lack of explainability. Two fundamental concepts enable the RLP algorithm (a) a path-following synthetic vehicle whose steering actions are non-myopically optimized to result in a smooth traversible path that meets path curvature constraints, and (b) a path-aware turning moment-field that enables obstacle avoidance while eluding the typical local-minimum-induced stagnation associated with PFMs. The use of the combination of the two concepts results in a reduced action space over which optimization needs to be performed towards minimizing the path deviation subject to obstacle avoidance, and thus results in an efficient algorithm that can be implemented online. RLP is demonstrated in simulations and field experiments, performing real time local path planning and obstacle avoidance on two different vehicle platforms (ackerman steered 2-axled vehicle, and a differential-steering 4-axled vehicle) in an unstructured off-road terrain