Using Multivariate Analysis to Optimize Development in Western Canadian Unconventional Resource Plays

Tyler Schlosser | McDaniel & Associates

 November 20, 2019

Doors open 11:30am | Announcements begin at 11:45 am
Hyatt Hotel, Imperial Ballroom 5/7/9 | 700 Centre Street SE, Calgary AB T2G 5P6

 This event is SOLD OUT 

Tickets for this luncheon will be emailed out 1 week prior to the event
The cut-off for ticket sales is 4:00pm, five business days before the event. November 13, 2019

Predicting and optimizing production performance, recovery and economics in large resource plays like the Montney and Duvernay requires integrating different types of information from a variety of data sources. One of the most important sources of information to accomplishing this is a consistent and, hopefully, predictive geological model that provides coverage across the entire play. A multivariate analysis workflow is outlined which incorporates geospatial, geological, fluid composition, completion, pad design and parent-child effect data together in one model. Identifying and deriving data features that offer the most predictive and interpretive capability to is important in arriving at a model. Advantages and difficulties related to this workflow when compared to more traditional reserves or development planning workflows are explored.

The challenges of model interpretation, data dimensionality and correlation, integrating domain knowledge and causal constraints, unexplained variance, model transparency and interpretability, and model application are discussed. Progressing through the workflow from training to prediction to optimization is iterative and usually involves many people. Applied workflow examples and some findings from real projects done in the Montney and Duvernay are highlighted, including some field development scenarios and economic optimization results.


As Director of Business Intelligence at McDaniel & Associates, Tyler Schlosser applies his 16 years of machine learning experience to the oil and gas industry for completion and field development optimization, performance prediction and a deeper understanding of the drivers of asset value and productivity. Tyler is also experienced in the technical and economic evaluation of oil and gas reserves and resources, probabilistic economic and commodity price forecasting, and risk modeling.