The new reality of oil and gas economics has had a major effect on the upstream industry. For most organizations the main goal has been to reduce production costs, with many turning to advanced technology. A key initiative showing initial success has been the digitalization of the oil asset. The vision is that increasing the amount of data collected and the ability to make decisions based on data will improve proactive management and reduce opex.
This digitalization is exciting, but it is only an enabling step. It involves instrumenting and collecting broader sets of information to achieve process insights.
Future of oil and gas economics
Long-term energy scenarios show continued demand for hydrocarbons but at a flattened growth rate, suggesting prices will likely remain near current levels. To sustain this, new production from existing and new assets must be brought online. This drives the need for future upstream investments sustainably and at close to today’s breakeven costs, meaning structural industry changes are needed. These include
- A move toward standardized versus one-of-a-kind designs enabled through the capture of best-practice unit and module designs via data and models, which reduce overall capex;
- An increased collaboration between industry players focusing on better use of people and resources across the execution chain; and
- A breakthrough in using data to achieve higher summits of reliability with the combined use of techniques such as machine learning, deep-insight process models and statistical models to turn Big Data into production-predictive and prescriptive knowledge. This has the promise of significantly driving opex reductions.
On the capex side, operators are driving engineers toward adoption of lower risk standardized and modularized asset designs. On the opex side, putting buyer pressure on contractors has resulted in short-term results but not a sustainable longer-term tactic. Instead, fundamental improvements in efficiency, capex designs and reliability are essential.
Promise of automation productivity
The upstream segment is characterized by its technical complexity, remote production environments, the challenge of developing the experienced technical experts who can guide these dynamic and ever-changing production environments and the expense of putting technical teams in place at the asset.
The upstream industry has lagged overall in terms of the types of automation common in other manufacturing sectors. The first area that has evolved rapidly is the automation of the drilling domain, specifically advanced directional drilling. Improving asset performance and reliability through technology is the current frontier. Today’s upstream world is ripe to benefit from a productivity growth opportunity.
Many organizations look at data analytics for uptime solely in the context of equipment maintenance, but there is a wealth of information types that provide valuable insights to enable machine learning tools to understand patterns leading to process problems and failure:
- Equipment data, including embedded sensor data and asset history data;
- Process data, including historian-based capture of all instrumentation across a process, unit or site;
- Maintenance data, including maintenance history, frequency, severity and equipment lifetime;
- Process safety data, including safety and asset integrity incidents associated with equipment and processes;
- Condition data, including results of measurements and inspections related to corrosion, equipment degradation and metal fatigue; and
- Enterprise resource planning, which provides a variety of insights into asset performance and yields.
The combination of these data types—understanding interactions of the hydrocarbon flows, the process and the equipment—allows prescriptive strategies to modify operating strategies, maximize uptime and minimize maintenance costs.
What to do with production data
More data are becoming available from equipment like turbines and new compressors as well as large pumps and subsea modules. However, equipment that can play a spoiler role in production levels and uptime have limitations on information collection and instrumentation costs. It is not simply about using data to understand equipment. Equipment interacts with the flowing hydrocarbons and the processes; those complexities and dynamics must be unraveled to optimize assets.
To make future unconventional assets feasible to produce, approaches such as machine learning combined with advanced optimization modeling will be key to achieving the economics that are required.
There are converged technology components that turn data into predictive knowledge. Advanced Big Data process historian applications can capitalize on the production accounting and allocation, online key equipment monitoring and safety systems data streams coming from the largest and most complex fields. Operations advisory and key performance indicator visualization systems in the form of operator-friendly dashboard tools take the data power of any process historian and turn it into an asset optimization and reliability weapon by removing people from remote and expensive-to-staff offshore environments.
Design, operations and maintenance have long persisted as isolated worlds of automation. The current upstream environment demands a different approach.
An upstream enterprise achieves operational excellence with a strong life-cycle view of the process and the asset and by continuously improving its assets to improve operability, maintainability and uptime. To take advantage of that, a feedback loop captures the digital image of that well-operated and optimized asset to use in the next similar asset development project.
Further innovations will provide even more insights and capabilities into optimizing designs for operability and maintainability. A few of the key breakthroughs include module-costing models within capital cost estimation tools. These provide a powerful environment to compare modular construction with “stick-built” construction. Volumetric model-based conceptual estimates in the capital cost estimation tools capture completed projects as cost models that can be reused in a low-risk way through powerful relocation and resizing models. Integrated economics enable the rapid translation of process models into total installed costs.
Knowledge of the process
Data alone are not able to provide the predictive intelligence operators and decision-makers require in such dynamic environments. Models encapsulating the interactions of the reservoir, production processes, equipment and economics provide the opportunity to fully leverage available data.
The key breakthroughs enabling models to be deployed as analytic engines in upstream include advanced solvers, equation-oriented solution methods for process models, rigorous models capturing operating realities versus design specifications and user-friendly dynamic modeling. This is crucial to understand the performance of gas- and oil-gathering networks and their interaction with production systems. Recent breakthroughs include packaging the advanced modeler experience into easy-to-use templates that show the models’ value in speeding up the startup and shutdown of offshore operations from days to hours. Each such contraction results in incremental revenue opportunities. Innovation in encapsulating flow-assurance thermodynamics in areas such as hydrate formation and rigorous gathering network hydrodynamics into the general process-modeling environment has provided access to the general process engineer, resulting in economic and safety benefits for upstream operations.
There are many opportunities to achieve significant economic benefits both in the capex and opex domains as well as in incremental production. The challenge is to make sense of which technologies match the business priorities best and assemble them into a business solution.