Energy & Sustainability

Machine learning has a powerful impact on energy and sustainability. A broad range of industries such as oil and gas, mining, geothermal, wind, and solar can solve their challenges with SpeedWise® Machine Learning (SML). SML algorithms can help companies save significant costs by forecasting energy usage, demand, and anomalies.

QRI BRINGS SML TO THE AWS ENERGY COMPETENCY PARTNER PROGRAM

The AWS Energy Competency Program is designed to identify, validate, and promote AWS Partners with demonstrated deep expertise and technical proficiency within this unique industry, including proven customer success developing solutions across the value chain, from production operations and optimization, to commodities trading, new energy solutions, and more. AWS Energy Competency Partners are vetted and validated against a high bar to attain the AWS Competency designation.

To receive the designation, AWS Partners undergo a rigorous technical validation process, including a customer reference audit. The AWS Energy Competency provides energy customers the ability to more easily select skilled Partners to help accelerate their digital transformations with confidence.

Use Cases

Pore Pressure Prediction

Machine Learning can be used to predict the Real Time Pore Pressure and Fracture Gradient. Pore pressure and fracture gradients are amongst the most critical items in mud weight design while drilling wells onshore and offshore. The actual pore pressure will dictate where to set the casing, how much the actual mud weight should be, what potential problems are, and more. Optimizing mud weight design can prevent potential problems such as unexpected mud losses and lost circulation, potentially much higher cost of drilling and the time delays in the drilling schedule and drilling rig operations, and certainly, the lack of safety of the crew. Check out our YouTube Video to watch a use case video about how to use Machine Learning to predict the Pore Pressure prediction.

Predict the Rate of Penetration of a Well

Machine Learning can be used to predict the Rate of Penetration at which a well can be drilled. The rate of penetration is the speed at which a drill bit breaks the rock under it to deepen the borehole. A common scenario is of a drilling engineer that would like to determine the best drilling operating parameters such as the RPM, drilling flow rate, drilling fluid characteristics, and more by using historical information. Check out our YouTube Video to watch a use case video about how to use Machine Learning to predict the rate of penetration of a well.

Initial Production

Machine Learning can be used to predict the Initial Production (IP) for a recompletion well. There are many features that can be used to help predict the IP such as the location of the well, the perforated net pay, the hydrocarbon pore thickness, cumulative water cut and GOR, and more. Check out our YouTube Video to watch a use case video about how to use Machine Learning to predict IP.

Completion Design

Machine Learning can be used to select the best completion design for a new well in an unconventional oil and gas reservoir. Hydraulic fracturing is a standard well stimulation technique used in tight oil & gas reservoirs to improve well productivity. The process typically involves high-pressure injection of "fracking fluid" into the reservoir to create cracks or fractures. The fracking fluid primarily contains water and proppant, which is a thickening agent. One of the biggest decisions that completion engineers need to make is how much fracking fluid and proppant they need to inject into the well to maximize the rate of return on investment. Check out our YouTube Video to watch a use case video about how to use Machine Learning to predict the best completion design.

Predict Anomalies for Energy Usage

Machine learning models can analyze energy consumption anomalies by identifying patterns of energy usage. For example, if a warehouse manager is informed that there is an unusually high usage of energy, they can further investigate their data to understand which machines may be utilizing the high energy. This allows the warehouse manager to catch any defects in expensive machinery, resulting in savings from large financial losses.

Predict Energy Demand

Machine learning algorithms can detect patterns in energy consumption data to provide valuable insights to building managers, energy companies, and utility companies. Insights can inform managers on daily or monthly energy usage and can help managers optimize strategies that result in energy-saving policies.

Predict Energy Price

Energy companies can utilize machine learning algorithms to make pricing recommendations. Large amounts of data can be used with SML to derive correlations between energy supply and demand. Based on these insights, pricing managers are able to make better informed pricing strategies and recommendations.

Wind Energy Production Forecast

A common challenge for power-grid operators is to predict the wind production as precisely as possible. SML’s algorithms can be used to reliably predict wind production based on historical data and variables such as wind speed, wind direction, date and time, and more. Predicting wind production output ahead of time can help significantly control and reduce costs.

Solar Energy

Machine learning algorithms can detect patterns in temperature, solar irradiation, and historic weather data to forecast weather conditions. Accurate weather conditions can help grid operators predict grid supply more accurately by reducing the uncertainty of the variable energy output.