Rahul Jha serves as US Upstream Materials Operations Manager at ExxonMobil, where he leads large-scale materials and logistics operations supporting upstream production across a network of warehouses and field assets.
In this role, he led the operationalisation of ExxonMobil’s SAP-based materials management system across Permian ExxonMobil facilities, establishing a standardised, system-driven approach to inventory visibility, material flow, and execution readiness for critical field operations.
Rahul’s work centers on engineering and operations management in unconventional oil and gas, with a focus on improving hydrocarbon recovery from complex oil sands systems. At Imperial Oil’s Kearl oil sands operation, a majority-owned affiliate of ExxonMobil, he developed and implemented production optimisation strategies that contributed to record output, reaching approximately 270 000 bpd in 2023 and increasing further to 281 000 bpd in 2024 through coordinated planning, constraint management, and real-time operational alignment.
A central element of this work was the development and application of Dynamic Limit Diagrams, a novel methodology within an ExxonMobil-affiliated operation that enabled teams to identify system constraints, coordinate across mining and processing units, and make daily production decisions based on integrated capacity visibility. The approach delivered measurable production gains and generated interest for broader deployment across global assets.
His contributions also include the application of advanced process control and machine learning to froth treatment system performance – the stage in oil sands processing where bitumen froth is separated from water and solids using solvent – where targeted improvements in recovery, solvent efficiency, and plant stability have delivered measurable operational and economic value. These efforts include structured optimisation programs addressing solvent losses, heat integration, and interface control in oil sands processing systems.
Rahul is a co-inventor on patented technologies related to oil sands processing and advanced measurement systems, and a published author in peer-reviewed journals including Energy & Fuels and Colloids & Surfaces A. His work reflects a consistent focus on translating engineering fundamentals and analytical methods into practical improvements in resource recovery and operational reliability.
This interview examines how engineering discipline, AI-enabled tools, and system-level thinking are being applied to improve recovery, reliability, and production performance in unconventional oil and gas operations.
Ellen Warren (EW): Rahul, you have led engineering and operational initiatives across extraction, froth treatment, and production planning in unconventional oil sands systems. How has your approach to improving hydrocarbon recovery evolved across these roles, and what principles now guide your work?
Rahul Jha (RJ): Bitumen recovery improvements increase production at virtually no additional cost, while also reducing GHG intensity. I approach these challenges using a data-over-opinion principle, applying advanced statistical analysis and data analytics to identify the primary factors that influence recovery. Once the key drivers are established, I determine which of them are controllable and quantify their impact using Pareto Principle. In most cases, the majority of improvement can be achieved by optimising just three to four controllable factors.
EW: At Imperial Oil’s Kearl oil sands operation, you contributed to record production levels through coordinated planning and optimisation. What were the key operational changes that enabled that step-change in performance?
RJ: Record production was primarily enabled by designing and implementing the processes, people systems, and operating routines required for a 24/7 team to focus continuously on the primary production constraint. This team worked in close coordination with multiple operations groups to make timely interventions in both plant and mine operations, ensuring the facility remained in its optimal operating state. During equipment reliability events, the team provided management with clear quantification of production losses, allowing leadership to rapidly prioritise resources and minimise downtime.
A key innovation supporting this performance was the development of first-of-their-kind Dynamic Limit Diagrams, updated in real time to identify the active constraint within a highly complex, integrated production system. Kearl consists of numerous interconnected subsystems, and determining the true limiting factor at any moment required substantial expert knowledge of system interactions. By embedding this expert knowledge into automatically generated diagrams, the team created a practical decision-support capability that made system constraints visible across the operation. This allowed technical and operating teams to focus on the highest-value interventions, accelerate constraint removal, and improve overall production performance. The innovation contributed directly to stronger operational execution and helped enable record production outcomes.
EW: Before you developed and implemented Dynamic Limit Diagrams, what operational limitations or decision gaps were preventing teams from maximising production at Kearl, and how did this methodology change the way constraints were identified and managed in real time?
RJ: As I mentioned, the Dynamic Limit Diagram was built using numerous inputs informed by years of operating experience, process knowledge, and system understanding. This expert knowledge was translated into a simple, practical constraint diagram that made it easier for operations teams, as well as new engineers responsible for plant optimisation, to quickly identify the true production constraints within the system. The diagram also highlighted the operating levers available to relieve those constraints, thereby enabling faster decisions and improving efforts to maximise production and oil sands recovery.
EW: Oil sands operations involve tightly coupled systems across mining, hydrotransport, extraction, and froth treatment. How do you ensure alignment across these subsystems when optimising for overall production rather than individual unit performance?
RJ: We ensured alignment by using the Dynamic Limit Diagram to identify the active constraints across all major subsystems, rather than focusing on any single unit in isolation. In addition, we developed operating guidelines, operating envelopes, and optimisation levers that helped teams make decisions supporting global optimisation of the entire operation, rather than unit-level optimisation.
EW: Your work in froth treatment optimisation highlights the importance of variables such as interface control, solvent balance, and mixing energy. How do these factors influence recovery, and how do you translate that understanding into actionable operating strategies?
RJ: Froth treatment optimisation began with advanced statistical analysis to identify the key variables driving recovery, including interface control, solvent balance, and mixing energy. Based on this understanding, we implemented an advanced multivariable process control strategy using Dynamic Matrix Control (DMC) to optimise these variables simultaneously. This allowed the plant to operate closer to its true optimum, increasing recovery while reducing solvent loss.
EW: Several of your initiatives focus on reducing solvent losses and improving efficiency. How do you approach optimisation problems where recovery, cost, and environmental performance must be managed simultaneously?
RJ: I approach these problems through integrated process optimisation, recognising that recovery, cost, and environmental performance are closely linked. When recovery is improved, the operation produces more valuable output from the same feed, energy, and solvent inputs, which reduces unit costs and improves environmental performance simultaneously. My focus is therefore on identifying the critical operating variables, understanding their system-wide impact, and implementing strategies that move the plant closer to its optimal operating window. This allows one optimisation effort to deliver benefits across production efficiency, cost reduction, and environmental performance.
EW: You have led the application of machine learning and advanced control technologies in operating environments. Where do these tools deliver the most value, and what conditions need to be in place for successful deployment?
RJ: I have found that machine learning and advanced control deliver the most value in areas where product quality or recovery is influenced by multiple interacting variables that are difficult to optimise manually. In my case, the work began with data analysis to identify the factors affecting product quality, followed by the application of reinforcement learning, a novel machine learning technique to improve performance.
For successful deployment, several conditions must be in place: reliable instrumentation and sensors, strong process understanding, an operations team with a continuous improvement mindset, and management willing to accept measured operational risk. Any plant trial carries potential downside, but with proper engineering safeguards those risks can be mitigated, even though some residual risk always remains.
EW: Your role in operationalising ExxonMobil’s SAP-based materials management system reflects a different dimension of operations leadership. How do materials visibility and logistics discipline influence production reliability in upstream operations?
RJ: One of the leadership skills I have developed is the ability to stay calm, evaluate trade-offs, and make sound decisions even when information is incomplete. That skill was especially valuable in helping implement SAP effectively across the Permian assets. The system has given us much stronger visibility into materials across warehouses and yards, with real-time updates that help ensure the right materials reach the business at the right time. It has also enabled us to manage key KPIs more effectively, identify the root causes of material delays, and solve them in a more systematic way.
EW: You are a co-inventor on patented technologies and have published research on oil sands processability and solids behaviour. How have these technical contributions shaped the way you diagnose and address recovery and reliability challenges in operating facilities?
RJ: My patents involve practical solutions to reliability and recovery challenges. One uses IR technology for continuous pipe monitoring, which can help detect pipe thinning or failure early, improve repair planning, and reduce process safety risk in highly flammable solvent service.
The other focuses on recovering warm water from a very difficult stream containing asphaltenes, solids, and water. Recovering this water can improve energy efficiency in energy-intensive oil sands processing.
EW: As unconventional oil and gas operations continue to evolve, what capabilities and operating models will be required to sustain improvements in recovery, efficiency, and system reliability?
RJ: I believe unconventional oil and gas operations will require greater instrumentation, broader use of advanced machine learning, and stronger advanced control systems to move toward more automated operations. These capabilities can improve recovery, efficiency, and reliability while also helping operators work more safely.
I also see significant potential for robotics in surveillance, equipment condition monitoring, and remote inspection and repair. Over time, this kind of operating model can make facilities safer, more reliable, and more efficient.
The views expressed in this interview are solely those of Rahul Jha and do not necessarily reflect the views of ExxonMobil.