Across the globe, oil and gas fields operate with tens to thousands of wellsites, a number set to grow in the years to come to meet increasing energy demands.
As these industrial assets grow in size and complexity, there's a pressing need for more efficient and sustainable ways to maximise production while reducing costs and emissions, especially in high-volume assets. The limitations of traditional control methods are pushing the industry towards autonomous operations, enabled by more intelligent control systems.
Addressing the Current Control Challenge
Control systems are key to running oil and gas fields effectively. They operate at each wellsite, and their efficiency and adaptability significantly impact the performance across the entire field and asset. Both the capability of these systems and the ease with which engineers and operators can manage them contribute to optimising performance at all levels.
A notable hurdle in optimising an oil and gas field using existing control systems is the "traditional control bottleneck".
Despite advancements in surveillance systems for detecting optimisation opportunities across well fields, the static and single-variable nature of traditional control systems remains a significant challenge for production engineers, field operators and control room operators to manage each day.
The need for routine monitoring and adjustments in these systems escalates as production systems grow in scale and complexity. Some wellsites may not even have a controller option currently available due to limited surface or subsurface equipment. Current, individual wellsite controllers entails considerable time and effort from production engineers and control room operators, hampering real-time responsiveness and broader asset network optimisation.
The use of single-asset, static-configured, single-variable controllers within multi-asset, multivariable, time-varying systems reveals two primary issues:
The time required in managing each controller.
The performance of the controller itself.
The dynamic operational environment of oil and gas fields means that even minor monitoring lapses or delayed responses, day or night, can trigger a host of issues, such as production loss and elevated maintenance and intervention costs. Where existing feedback controllers do result in instability, equipment failures are costly. Preventing even a handful of these failures each year can be worth hundreds of thousands of dollars.
Addressing these issues is crucial for advancing towards autonomous operations, which target unified connectivity and enhanced capabilities for wellsites operating within a network.
Elevating Control Systems: A Smart Control Paradigm
Enhancing an asset's autonomous function begins by upgrading the autonomy of its control systems, which lays the groundwork for building more autonomous business processes.
Introducing smart control technologies is a key step forward. Utilising machine learning and data analytics, smart control facilitates real-time decision-making, addressing the constraints of traditional controllers like proportional-integral-derivative (PID) control.
Smart control aligns with modern control theory, moving beyond basic linear control methods to include adaptive and predictive control techniques. In context of the automation pyramid and the Purdue Model of Control Hierarchy, smart control acts as a supervisory software layer (e.g. level 2.5) to form a Modified Purdue Model, that integrates data from the process control level with the operations and business layers.
Unlike traditional hardware-centric control systems, smart control providing a way to “codify” engineering knowledge and principles into the way a control system is optimised and operates.
Smart control facilitates a richer integration and modeling of data to unearth optimisation opportunities. It not only introduces business-level strategies into the controllers' operation but also provides more integrated, data-driven insights on asset performance up to asset managers, fostering a harmonised, data-driven decision-making environment across an asset-driven organisation.
How Does Smart Control Differ from Existing Control Systems?
Smart control advances traditional control systems by adapting and optimising autonomously, unlike conventional systems that often require manual adjustments. It embodies a "set it right, let it run" approach, allowing organisations to achieve more with less. This is because smart control necessitates less detailed configuration for every minor change in the dynamic subsurface or conduits, streamlining operations and saving valuable time.
A key difference is a smart controllers ability to utilise a wide range of data to learn from. It taps into both real-time and historical data, embodying a form of "memory-based learning" where past data informs current control strategies. This stands in contrast to traditional systems that may only react to immediate deviations from a single signal, lacking the insight from analysing trends over time. Without signal availability (e.g. a failed downhole sensor), existing controllers may not even be able to function.
Additionally, smart control being software-based can come with user-friendly interfaces, making system interaction simpler and more intuitive for engineers. Smart control offers a more intelligent, responsive, and engineer-centric approach to optimising industrial control systems.
How Does Smart Control Work?
Smart control automates industrial operations by continuously checking multivariable systems and making corrections to maintain optimal performance (e.g. in dynamic environments). It connects different system parts or even multiple systems for coordinated control. It uses smart algorithms (machine learning modelling) to analyse data, adapt to changes, and optimise performance. It learns from past data to improve control strategies over time, makes decisions based on real-time or historical data analysis, and reduces human intervention by automating routine tasks. It also can offer diagnostic information and predictive maintenance insights.
The Two Pillars of Smart Control: Connectivity and Capability
Industry 4.0 technologies are pioneering the connectivity between Information Technology (IT) and Operational Technology (OT) for unified control and production management systems. This marks a new phase where on-site controllers (i.e. OT systems) can be dedicated to process safety, and more advanced optimisation modelling can be done remotely (i.e. IT systems) where there is a requirement for interconnectivity across assets and data-intensive modelling, to deliver more intelligent, autonomous operations.
This transition also presents a chance to refine the automation architecture for remote, industrial assets, fostering a more streamlined and effective operational framework across front-line teams.
The sophisticated algorithms encapsulated within smart control serves as the intelligence hub of the operation, merging domain knowledge and engineering principles with machine learning to provide setpoint recommendations.
Utilising the abundant data gathered from daily operations, smart control employs data-driven modeling to foresee how to enhance multiple strategic objectives, such as boosting production performance while reducing costs and emissions, in line with broader sustainability goals.
Smart control makes the concept of autonomous operations achievable, enabling assets to function more efficiently with less engineering intervention, and aligning with wider sustainability and operational excellence objectives.
Smart control harnesses the data from your operations, converting it into valuable insights and automatic adjustments, learning each day. It's about optimising your assets within operational limits and constraints through advanced analytical modelling.
Here are the core advantages that smart control brings to address the primary challenges associated with traditional control systems:
1. Enhanced Asset Performance
By introducing more sophisticated algorithms, smart control helps maximise production and minimise both cost and emissions. It's designed to get the most out of your assets while keeping them within the safe and efficient operational boundaries you define.
2. Time Efficiency
Less time is needed to maintain the controller. This efficiency lets engineers focus more on optimising the asset itself rather than getting bogged down with configuring and maintaining individual control loops. It translates to more time for higher-value tasks and projects.
3. Simplified Setup in Mature Assets
For mature assets, smart control alleviates the need for installing additional downhole sensors to maintain a high-performance control system.
4. Improved Connectivity
Bridging the communication among different disciplines, teams, and assets within organisations, smart control fosters enhanced data visibility for strategic decisions. It nurtures a collaborative environment, making information flow seamlessly across the board.
For organisations operating assets with well-established control and surveillance systems, envisaging a 1-5% improvement across various value drivers like production, reliability, cost, engineering time, and emissions provides a realistic metric to evaluate the potential of smart control technologies. It’s prudent to measure the net benefits against the costs associated with implementing these technologies, while also accounting for technical risks and potential upsides. This assessment forms the groundwork for an informed decision, aligning the technological investments with the overarching goals of enhancing operational efficiency and sustainability through autonomous operations.
Example: Traditional vs. Smart Control
For example, consider a field with 200 wellsites. In this example, half the sites are controlled on pressure and the other half on flow rate, requiring four parameters per well, totalling 800 setpoints to manage.
Alternatively, with smart control, there’s one smart controller for the entire field. Global operating limits are pre-programmed into an intelligent control algorithm with 2 parameters to function per field, reducing the total setpoints to manage for optimisation purposes to just 2.
This reduces the configuration effort by over 99%, empowering engineers and operators with scalable control levers. This is the engineering workforce multiplier that data-driven modelling can offer.
Example: SmartLift™ Software Driving Autonomous Operations
At SIG ML, we offer smart control for network wellsite optimisation of artificial lift wellsites, now proven to operate successfully, with closed-loop automation, at scale.
SmartLift™ provides coordinated control of artificial lift systems, bringing multiple wellsites previously operated independently into a common control software, allowing for efficient handling of low-level, repetitive decisions across hundreds or thousands of assets simultaneously.
SmartLift™, carefully and securely integrated into your daily monitoring, analysis, and control systems, can learn from the data you collect and be guided by engineering knowledge and principles. For connectivity, SIG ML has partnered with Kelvin.ai, to form a commercialised smart control solution.
Learn more here .
At SIG ML, we offer smart control and decision software applications to optimise industrial assets in less time, with less engineering effort.
The future of engineering is data-driven.
Contact us today to for a collaborative discussion on trialling this technology.
Author: Sam Bost, Founder & CEO