ResX software provides a streamlined reservoir modeling and data conditioning framework.

Many Silicon Valley fueled buzzwords—digitalization (or digitization), big data, cloud-based solutions, artificial intelligence (AI), machine learning—have been ringing in our ears for some time now.  While these words quickly entered the corporate vernacular, the application of these technology visions, let alone their adoption, is much slower—especially in the upstream oil and gas industry.  

In reality, beyond the buzz, the industry has been working with advances in hardware, software and infrastructure technology for years. Take decline curve analysis or Kriging. We may not instantly think of these as applications of machine learning, but they have been part of software code for decades.  The limited use of this approach to “learn” from data is not because of a lack of applications areas, but more the lack of readily accessible data.  In fact, the toolbox of sophisticated algorithms to assist in modeling the subsurface has been waiting for the expansion of digital technology and data lakes. This is now driving innovations in many areas of our industry from assisting real-time operations to strategic field development planning.

At Resoptima, we are using the latest software technology to enable better reservoir management.  Our focus is to provide a software solution that helps asset teams address several of the key subsurface modeling cross-domain questions including:

  • How can we condition static and dynamic data consistently? 
  • What is the best way to propagate uncertainty systematically across models? 
  • How can we dramatically increase the reliability in our model predictions? 
  • Is there a more efficient way to keep models updated as new data arrive?

ResX is Resoptima’s signature software and it has been developed to bridge the gap that has existed between the geoscientists and reservoir engineers’ modeling tools. The technology behind ResX takes off from the ensemble Kalman family of algorithms. The idea behind the algorithms is to use a series of measurements over time (measurements that may contain noise or errors) to estimate or predict an unknown variable at a future state.  This methodology has applications in navigation systems, weather forecasting, economics, and, more recently, reservoir modeling.  In our case, Resoptima has developed fit-for-purpose machine learning algorithms in ResX to enable a robust quantification of uncertainty across static and dynamic data conditioning and modeling.  The result is a solution that improves the efforts and results of reservoir modeling and management.

The digital era is clearly fueling the growth of data science solutions to domain-specific problems and enabling new expressions of traditional subsurface modeling.  ResX is part of this wave and is something truly new in the digital space of the subsurface modeling world. To see ResX in action contact Resoptima.