In today’s operations allocating production asset surveillance across well sites may seem infeasible, where just a handful of engineers are tasked with watching over and responding to alarms for hundreds of locations.
But technologies like artificial intelligence (AI) and machine learning are changing what is possible in upstream production operations management, allowing cost-effective scaling and deployment of previously inaccessible intelligence.
This shift to smarter autonomous operations can help reduce downtime risk and dramatically improve operational efficiency.
Overcoming upstream challenges
Many Australian well sites are remote and sprawling, and practical constraints can limit the level of instrumentation, control and intervention. Evolving conditions over time require adaptive methods that can make automated processes either too costly or too difficult to implement using current strategies.
Often engineers are overwhelmed by multiple production-well alarms because the alarms operate within tighter parameter bounds to track whether an asset is operating in an optimal region, hence requiring regular threshold adjustments as conditions evolve. This can result in staff missing important events that lead to asset and production downtime.
They also typically only learn about events after they happen. Figure 1 shows a challenging electrical submersible pump (ESP) well with multiple gas interference events. Within the course of seven months, there was cumulative downtime of about 100 days, almost 100 stop-start cycles (Hz=0), and a total of four days of the system being in stressful low flow conditions.
Tracking and prioritising events is still largely a manual process.
Figure 2 shows a timeline of an actual incident on a high-value ESP well, where a real-time AI-based detection engine was tested. The engine was fed real-time signals from the ESP system, such as pump discharge and intake pressures, motor speed, current and temperature, and wellhead pressure. It was engineered for robustness, with the ability to accommodate different combinations of available measurements, and account for data quality issues such as missing, frozen or faulty sensor data.
In this incident, the system was able to raise an issue during restart, balancing sensitivities that can lead to false alarms and gathering sufficient evidence before raising the flag. While the solution demonstrated significant value by providing an early alert of an existing critical event, there was still 23 minutes from the point-of-detection to shut-down, due to the use of largely manual processes. Could the system be intelligent enough to diagnose the situation, recover on its own and prevent a costly shutdown?
During operation, ESPs will be subject to multiple stressful events and the normal wear and tear associated with a running mechanical device – contributing to the eventual failure of the pump.
This is shown in Figure 3 on the top timeline as a red operating zone. The quicker operators can detect and resolve the event, avoiding the red zone, the less stress the pump will experience, increasing the pump’s lifespan, improving production time and reducing intervention costs.
Meanwhile, another challenge facing the oil and gas industry is mounting retirements. As skilled, seasoned employees leave the workforce, they take decades of critical knowledge about production assets and processes.
Deploying more intelligent production capabilities addresses these challenges by capturing crucial process knowledge and enabling higher levels of automation within the control system at the edge.
Intelligent Automation at the Edge
The industry has a rich history of modelling and simulation tools and operations know-how. What determines the success of more decentralised intelligence is effectively packaging, deploying and maintaining these elements at scale.
Ideally, solutions should integrate with a production asset’s IoT-enabled control panel rack and remote terminal unit managed centrally from the cloud. By deploying this intelligence at the edge, operators can get the required necessary response times for closed-loop automation and optimisation. Advanced automation can be done in a reliable manner, without being susceptible to factors like wireless communications disruptions, bandwidth limitations and cost.
Back to the previous ESP example, an AI-based solution deployed in the control system can recognise high-risk situations by constantly evaluating the probability and severity of issues and act immediately. Because ESPs are located downhole, they require adequate flow for cooling the motor and pump. In a low-flow situation, a significant amount of energy can potentially be released locally around the ESP, requiring immediate resolution. The solution can adjust equipment operations based on the specific type of low-flow event detected and continuously monitor the impact.
Since the system can proactively adjust controls in early moments before the conditions escalate, it avoids unnecessary shutdowns, protects production assets, and extends their useful operating life.
Operators with limited resources must prioritise which wells to dedicate attention to, based on metrics like production rates, leaving lower-tier wells to trip and risk prolonged shutdowns. In an era of efficiency, AI-based solutions that scale avoid making drastic trade-offs.
Higher Level of Knowledge
When operationalised at scale, AI provides a higher level of decision support to experts, improving the management of production assets. This allows centralised management to immediately begin capturing, prioritising, resolving and classifying events.
Over time, a treasure trove of knowledge can then be accumulated and used to continuously improve. In the ESP case, as more events are validated and properly catalogued, supervised learning techniques can be deployed to retrain engines to improve performance metrics.
Instead of losing valuable knowledge as workers retire, this captured knowledge can be shared across the workforce, to drive decision making. As more intelligence is made accessible, the gap between production engineers and operations will narrow, enabling greater collaboration and help harvest previously untapped efficiencies.
Additionally, there are elements of the AI-based solution that can continuously learn about each well and its events. With this mechanism, the solution will adapt to improve its ability to solve problems based on each well’s unique history.
Reimagining well sites
Real-time, plug-and-play AI-based solutions are already being tested to help drive better decision-making.
They are designed for maximum impact and minimal disruption, having the ability to scale to many assets with minimal set up and maintenance over time. These solutions are already being used in the cloud to detect and prioritise events, and at the edge to autonomously resolve critical events and improve by learning over time.
Soon, intelligent solutions like these will be a competitive necessity for producers that want to not only improve their performance and profitability, but also retain critical operations knowledge before it walks out the door.
By Jonathan Chong, Advanced Technology R&D Manager, Sensia, a Joint-Venture between Rockwell Automation and Schlumberger.