Cutting Downtime in Half with Deep Neural Networks

Cutting Downtime in Half with Deep Neural Networks: Advanced Predictive Maintenance Two Weeks Ahead CASE STUDY

Introduction Fueling efficiency: Battling ESP failures in remote In the fossil fuel industry, operational efficiency and component reliability are pivotal for seamless business processes. The down-hole mining sector often encounters energy sources deep under the earth’s surface and in remote locations that need innovative energy extraction solutions. However, challenges such as unexpected failures of Electrical Submersible Pumps (ESP) can lead to extended downtime that costs companies millions of dollars.

Challenge Revolutionizing downtime reduction for a global energy giant Our client, one of the largest integrated energy companies in the world, needed a way to shorten downtimes by proactively identifying breakdowns before they happened. The oil wells’ remote locations and the pumps’ depths compounded the challenge of addressing failures quickly. Traditional maintenance practices typically involve periodic inspections or scheduled maintenance at fixed intervals, which can be time-consuming or not viable when a well is continuously running. Unplanned failures are common and can lead to extended downtime, further impacting productivity and profitability Addressing the unforeseen Unplanned failures of ESPs were a critical problem for our client. These unexpected breakdowns cause significant downtime, impacting production schedules and operational profitability. Minimizing downtime is incredibly important to avoid multi-million-dollar losses, primarily due to the weeks it takes to schedule, travel to, and repair non-producing wells.

Smart ESP management: reducing downtime, boosting ROI

Periodic replacement and fixed-interval maintenance are not profitable when doing so requires stopping the production of a well for potentially no reason. If an ESP was found to be running fine and had plenty of life left, the well experienced significant downtime (and lost revenue) for little to no return. Our client needed a more intelligent and proactive approach. The goal was to allocate maintenance resources just in time to schedule replacement before an ESP failure occurs, optimizing maintenance costs and ESP lifetime value and reducing unplanned downtime.

Solution Mastering ESP failure prevention with deep learning

Fractal needed first to understand the types of ESP failures to address the challenge. We analyzed the challenge and root cause analysis data to do this. Armed with this knowledge, we harnessed the power of Deep Neural Network (DNN) models, leveraging historical sensor data, to effectively differentiate between standard and aberrant equipment conditions. Drawing upon our extensive proficiency in predictive maintenance solutions and a comprehensive grasp of our client’s data and operations, we engineered a model capable of forecasting equipment failures well in advance, enabling the scheduling of repairs with minimal disruption to operations.

The base CatBoost model employed an autoencoder mod- el as a feature input, and to refine our approach, we lever- aged domain expertise from subject matter experts (SMEs) and conducted Physics-driven failure mode analysis. We further customized the model by incorporating informa- tion about each specific machine’s operating conditions and insights gathered from other machines to improve the model’s accuracy. Fractal deployed this solution on Azure, using Azure Data Lake for storage, Databricks for data analysis, and Azure Machine Learning for model training. Our offering: Engineered ESP success through predictive maintenance precision

TOOL

AIM

RESULT

To differentiate between normal and deviated equipment states

Improved accuracy in predicting the condition of equipment Enhanced capacity to predict equipment conditions An accurate representation of the intricate conditions that equipment may face Improved accuracy of the model in addressing ESP system conditions A significant boost in the precision and timeliness of equipment condition predictions

Deep Neural Network

An input to the primary CatBoost model

Autoencoder Model

To gain insights from SMEs and physics-driven failure mode analysis

Domain Expertise & Analysis

To learn from other similar machines to enhance model accuracy

Customized Machine Conditions

To increase precision and timeliness in predictions

Additional Model Enhancements

Outcome Maximizing Uptime and Profitability

Instantaneous Impact Implementing the predictive maintenance model resulted in a substantial reduction in unplanned downtime. By identifying equipment issues at least two weeks before they escalated, the company could schedule maintenance in time to get a team out to the remote site before the ESP failed, leading to additional revenue generated and optimized maintenance operations. Sustainable Gains Over time, the benefits of this solution will continue to grow. In particular, the model’s ROI will continue to improve as additional failure data is generated. This will reduce or eliminate downtime before a team can repair a well. It will also decrease false positives for pumps with remaining useful life. This would help see a substantial revenue increase from the additional uptime of pumps in the field.

Profitability Enhancement

Cost Savings & Revenue Growth

Downtime Reduction

The model significantly reduces the downtime between a failure and a response, resulting in additional revenue.

Knowing what has failed and where leads to reduced turnaround time and increased profitability.

Substantial cost savings and revenue growth are realized through intelligent and proactive maintenance.

Enable better decisions with Fractal Fractal is one of the most prominent providers of Artificial Intelligence to Fortune 500® companies. Fractal’s vision is to power every human decision in the enterprise, and bring AI, engineering, and design to help the world’s most admired companies. Fractal’s businesses include Crux Intelligence (AI driven business intelligence), Eugenie.ai (AI for sustainability), Asper.ai (AI for revenue growth management) and Senseforth.ai (conversational AI for sales and customer service). Fractal incubated Qure.ai, a leading player in healthcare AI for detecting Tuberculosis and Lung cancer. Fractal currently has 4000+ employees across 16 global locations, including the United States, UK, Ukraine, India, Singapore, and Australia. Fractal has been recognized as ‘Great Workplace’ and ‘India’s Best Workplaces for Women’ in the top 100 (large) category by The Great Place to Work® Institute; featured as a leader in Customer Analytics Service Providers Wave™ 2021, Computer Vision Consultancies Wave™ 2020 & Specialized Insights Service Providers Wave™ 2020 by Forrester Research Inc., a leader in Analytics & AI Services Specialists Peak Matrix 2022 by Everest Group and recognized as an ‘Honorable Vendor’ in 2022 Magic Quadrant™ for data & analytics by Gartner Inc. For more information, visit fractal.ai

Corporate Headquarters Suite 76J, One World Trade Center, New York, NY 10007

Get in touch

Page 1 Page 2 Page 3 Page 4 Page 5 Page 6 Page 7

Made with FlippingBook - PDF hosting