The White Whale team posing for a socially distanced photo in Crescent Heights

Article

White Whale’s Production Forecasting Solution with AI

Written for BOE Report

The heavily nuanced and complex landscape of the energy industry, especially in the case of unconventional reservoirs, has led to a significant breakthrough in the converging worlds of energy and data analytics. Following years of iterative development in collaboration with some of the top producers in the region, Calgary-based White Whale Analytics has released its latest oil and gas solution.

DeepSea is a versatile end-to-end AI platform with applications in a wide variety of industries and contexts. As a software platform for oil and gas analytics, it has its roots in simulating and optimizing production scenarios to create the best outcomes for White Whale’s clients. Most recently, its world-class machine learning algorithms have been applied to forecasting EUR for thousands of wells in the Montney region with significant results.

The latest technology stack learns from historical data to generate a more accurate and unbiased look at reserves estimates as far as 10 years into the future. In addition to the obvious benefits of creating accurate forecasts with such longevity, DeepSea’s approach eliminates bias inherent to conventional techniques, and does so in a processing time of mere minutes.

Hand-fitting type curves to production data is difficult to do with accuracy and near impossible to do within this time frame manually. Production data can be complicated and noisy regardless of the conditions in the field, but the early stages of a well’s life cycle can be especially unreliable for detecting trends in the data. Furthermore, unconventional reservoirs are especially complex and can be a problematic environment in which to establish and understand the relationships that may exist between wells.

DeepSea mitigates these challenges with cutting-edge clustering algorithms to create as accurate a forecast as possible even when given limited or noisy data. It also pre-processes data, eliminating bad data points resulting from the early stages of a well’s life cycle, or from simply anomalistic behaviour. However, while these anomalistic data points are “trimmed” from the EUR forecasts, they are not forgotten. DeepSea’s built-in anomaly detection technology checks daily data against forecasted values autonomously to identify wells which are behaving abnormally, acting as a guide directing engineers’ attention to wells in need and saving dollars in downtime costs.

White Whale's Anomaly Detection
A demo snapshot of White Whale’s well anomaly detection technology output.

From pre-processing bad data to creating forecasts with pinpoint accuracy, every step of the way is automated within DeepSea, streamlining workflows for engineers and simplifying the tedious work in upstream operations, leaving room for the experts there to reach for bigger ideas. In addition to time saved on behalf of the engineers who would, typically, create these types of forecasts manually, DeepSea’s production forecasting technology offers a host of benefits to oil and gas producers across production methods. In the SAG-D space, for example, DeepSea has achieved the following for White Whale clients:

  • Increases in overall production by margins of 2-10%
  • Reductions in steam-to-oil ratio and, therefore, water usage, by margins of 10%
  • Improved environmental performance by reducing energy and air emissions intensity throughout the plant

As a whole, the world of data analytics, machine learning and AI has formed a symbiotic relationship with the oil and gas industry as the latter answers the call to take advantage of their inherent wealth of data on a widespread scale. In the market for data analytics and AI software solutions for oil and gas, DeepSea differentiates itself by answering complex problems with advanced methods, and within a simplistic interface. The complexity of building and working with a data analytics dashboard is minimized with DeepSea’s fluid, thoughtful design, and users of all types and backgrounds are empowered to realize value in their data and processes from DeepSea.

Since early in its life, White Whale has aimed to build empowering solutions like this to help its clients make data-driven decisions using less time, space, and energy across the board. In countless problems and scenarios, scientifically-sound models for organizations of all sizes have been delivered by White Whale with unparelled deployment times. Within just three weeks from the initial consultation, White Whale secures access to relevant data and builds a working prototype of the solution, allowing the user to begin their validation and testing.

In this latest production forecasting solution, however, the timeline is much shorter than that. Working with public data updated on a monthly basis, White Whale has built a standardized prototype for any producer in the Montney region, offering insights into both their own and their industry peers’ data. This showcases DeepSea as a robust short- and long-term forecasting tool for all reservoir variables, within a familiar context for energy producers in Alberta and B.C.

White Whale's Montney Solution
A demo graphical user interface of White Whale’s Montney solution.

Across the industries it delivers solutions to, White Whale has demonstrated a skillfulness in decomposing the individual problems it addresses into core concepts that are commonly felt to create scalable solutions. As a result of this process and a skilled data science team behind the scenes, DeepSea is the energy software platform that knows the industry profoundly.

Read More