The Strategic Advantage of MemComputing in Oil and Gas: Midstream
“The real competition is between supply chains, not companies” – Martin Christopher
The oil and gas industry is one of the largest international supply chains that fuels our global economies and daily lives. Given the immense size and scale of operations, companies in this industry face tremendous logistical challenges and compete to maintain efficient supply chains.
It is the job of the quantitative analysts, operations researchers, and data scientists within oil and gas companies to constantly improve the efficiency and reliability of their supply chains.
They strive day in and day out to achieve their goals of maximizing production, reducing capital and operating expenses, and improving their carbon footprint. They use tools such as mathematical optimization solvers to address many of these challenges, which are often embedded in their sophisticated supply chain planning software.
However, today’s best-in-class technologies cannot handle the computations at scale to provide truly optimal solutions.
When considered in their entirety, these logistical problems are too large and complex to solve at one time. Therefore, they must be broken down and solved in pieces or by ignoring some inputs and constraints.
This technique provides solutions, but they are sub-optimal. Thus, there is still room for significant improvement as these companies struggle to streamline their ever-growing data sets into actionable insights for better decision-making.
As a result, companies are turning to new advanced computing tools, like the Virtual MemComputing Machine, to tackle their most difficult supply chain challenges in the midstream sector.
The midstream sector plays a critical role in the energy chain. It handles the transportation and storage of oil and gas from the upstream sector (exploration and production) to the downstream sector (marketing and distribution).
As you can imagine, there are complex domestic and international scheduling, routing, and forecasting challenges that are difficult to manage. Organizations plan these operations months or even years in advance, whether it’s delivering workers and materials to remote/offshore locations or transporting oil and gas from rigs to refineries via massive tanker ships, trains, or pipelines.
In many cases, the logistical coordination required to execute these efforts becomes too large and complex to solve using traditional methods. This complexity is due in part to the challenges associated with:
- Integrating different supply chains
- Various modes of transportation (pipeline, rail, truck, ship, helicopter)
- The human resources required to load, unload, conduct, drive and pilot these vehicles
- Sensitive time windows
- Economic uncertainties
- And an unending variety of other variables and constraints
Therefore, these companies must rely on heuristics or manual approaches to solve these enormous logistical problems. Depending on the size and scale of the application, this can still be a very time-consuming process and result in approximate or sub-optimal solutions, which leads to unavoidable costs.
And as with any supply chain, disruptions are inevitable. Changes in demand, bad weather, and unexpected downtime challenge companies to quickly adjust their supply chains to minimize cost and time to recovery.
Indeed, in order to remain competitive, companies are actively seeking new technologies to gain more visibility into their operations to improve their supply chains’ efficiency, reliability, and resilience.
The MemComputing Advantage
MemComputing is working with a number of the world’s “supermajor” oil and gas companies on optimizing complex applications in the midstream sector. These are the types of problems MemComputing is currently focusing on:
- Vessel purchasing/leasing/chartering decisions
- Cargo planning
- Transport scheduling
- Vehicle routing
- Refinery scheduling
These problems come in all shapes and sizes. However, the Virtual MemComputing Machine (VMM) excels at providing truly optimal solutions for the largest problems at scale (involving thousands of variables and operational/economic constraints) that are otherwise intractable for today’s technology. VMM capabilities include:
- Increased computational capacity
The VMM, at its core, is a mathematical optimization solver, but it isn’t constrained by classical computing limitations the way current methods are. Instead, the VMM leverages computational memory to overcome classical computing bottlenecks and solve optimization problems orders of magnitude faster and more accurately than today’s state-of-the-art methods.
This proprietary technology handles the largest, most complex optimization problems at the immense scale required by oil and gas operations and many other industries.
2. Rapidly react to supply chain disruptions
Should a disruption occur, such as unexpected weather conditions, companies use the Virtual MemComputing Machine to re-optimize their operations in minutes. The mathematical model representing the supply chain application can easily include constraints to mirror current conditions, allowing our clients to quickly re-optimize their schedule, routes, and plans while minimizing the time to recovery and mitigating losses to resume operations.
3. Scenario planning
Clients leveraging the Virtual MemComputing Machine can also run various what-if scenarios that consider changes in supply and demand as well as fluctuations in the inventory of materials, resources, and capacity levels. This flexibility allows companies to prepare for future problems or weigh their options before deciding on a logistical endeavor in the future.
The VMM represents a strategic competitive advantage for companies operating in the midstream, as proven in the following real-world use cases:
Helicopter Fleet Schedule Optimization
A large oil and gas company employed the Virtual MemComputing Machine to schedule a fleet of helicopters for offshore deliveries to oil rigs and support vessels.
This particular problem is often intractable due to the number of helicopters, their different capacities, the number of offshore locations, cargo, personnel, and time expectations of the deliveries.
The Virtual MemComputing Machine provided a near-optimal schedule that was 24% more efficient than best-in-class optimization solvers and solved the problem orders of magnitude faster. Not only was the solution found in seconds, but the resulting savings in time, fuel, costs, and reduced maintenance was proven to deliver $10s to $100s of millions in annual savings depending on the fleet size.
Maritime Crude Oil Delivery Scheduling
Another large integrated oil and gas company wanted to determine the optimal monthly shipping schedule of Very Large Crude Carriers (VLCCs) to ports while also considering:
- Many dozens of VLCCs, along with the variety of grades of crude oil that they carried
- The size and depth of ports
- Refineries and their inventory of different grades of crude oil and blends
The problem is intractable in that it would take years to find an optimal solution using current methods. Instead, the Virtual MemComputing Machine converged quickly to satisfy all constraints and provide far better solutions than today’s best approaches in about 1 hour. The solution we found demonstrates the potential to save $10s to $100s of millions annually for problems of this magnitude in the O&G industry.
The oil and gas industry is experiencing transformational times, as reflected by the intensifying trade tensions, turbulent global economics, and the disruptive impact of the Covid-19 pandemic.
Now more than ever, companies are pursuing new innovative technologies to improve the efficiency and resilience of their supply chains to maintain an edge over their competitors.
With MemComputing, companies in oil and gas are optimizing logistics across the entire energy chain to realize new operational efficiencies that dramatically reduce capital and operating expenses, increase profitability, and improve the environmental outlook of their operations.