Identify rod pump anomalies using AI



Companies can now measure almost every single aspect of business activity due to various analytic tools and management software. Often, they have advanced datasets to explore their performance because there are so many metrics to measure.

While navigating various datasets, anything outside the norm would be considered an anomaly. Different anomalies can have an enormous impact on business decisions, operational results, and management’s ability to respond rapidly and accurately.

To help oil and gas operators with Rod Pump management, we've developed Emerson's Zedi Autonomous Rod Pump Management. The application is an Artificial Intelligence (AI) engine that proactively identifies common operational anomalies and makes the appropriate adjustments to the rod pump stroke and speed, all without operator intervention.


How we identify issues

Rod pumping equipment can encounter various problems over its lifetime. hardware failures such as parted rods and worn tubing can be partially mitigated with proper design and maintenance practices. Wellbore-related issues such as wax build-up, gas interference, or gas locking are more unpredictable.

Using Emerson’s Zedi Cloud SCADA platform along with the Zedi Go Mobile app, operators can remotely enter, monitor, and manage production data including adjusting pumping parameters without traveling to the site. Then include Zedi Autonomous Rod Pump application, to take full advantage of AI algorithms to help you identify anomalies in near real time.Autonomous rod pump management

As pump cards and other data are pulled into the Zedi cloud platform, the AI engine analyzes all parameters and indicates an assessment on the Zedi Access dashboard. When an issue is identified, the analytics engine can autonomously issue commands to the rod pump’s pump off controller to adjust the variable frequency drive’s (VFD) operation to correct the situation.

Corrective actions consist of sequences of learned commands to the VFD to adjust the stroke rate as needed based on the type and severity of the anomaly, just as an operator would. Instructions can be autonomously delivered to the pump off controller to make the appropriate corrective action to the pump, or raised as an alarm to the operator, identifying the issue and recommending action.

The AI engine can identify and correct the most common modes of failure in rod pump wells and recognize a problematic situation that falls outside of the current set of identified anomalies.

Our solution allows you to prioritize daily tasks, so you can spend your time troubleshooting wells and minimize the need to make unnecessary trips to the field.

A growing list of anomalies

Our experts continue to advance our AI engine, so its variety of detectable anomalies expands. Over time, our technology will be more valuable as the range of predictive and autonomous capabilities increases. Identified anomalies currently include:

  • Fluid Pounding
  • Gas Interference
  • Gas Lock
  • Parted Rods
  • Sticking Pump Barrel
  • Waxing
  • Sanding
  • Stuck Piston
  • Traveling Valve Stuck Open
  • Traveling Valve Stuck Closed

We want to ensure our autonomous solution can provide your operations team with real-time knowledge, allowing them to make intelligent decisions and schedule the work necessary before going to the field. Optimize your rod pumps with our AI solution so you can increase production revenue due to reduced downtime, fewer maintenance costs, and improved safety onsite.LEARN MORE >


Topics: Artificial Lift, Zedi Go, Zedi Access, Oil and Gas, Analytics, Production Data, Autonomous, Rod Pump, AI, SaaS SCADA, Artificial Intelligence

Subscribe to get new blog posts >

Contact us >

Recent Posts