Spatial variability of tight oil well productivity and the impact of technology
PhD candidate MIT
Research Assistant for the MIT Energy Initiative.
Topic: Shale Production
When: Friday April 20, 2018
Time: 11:00AM - 12:00PM EST
Where: Auditorium Schlumberger-Doll Research Center, Cambridge MA
1 Hampshire Street
Cambridge, MA 02139-1578 USA
RSVP is required- click here to RSVP
New well productivity levels have increased steadily across the major shale gas and tight oil basins of North America since large-scale development began a decade ago. These gains have come about through a combination of improved well and hydraulic fracturing design, and a greater concentration of drilling activity in higher quality acreage, the so called “sweets spots.” Accurate assessment of the future potential of shale and tight resources depends on properly disentangling the influence of technology from that of well location and the associated geology, but this remains a challenge. This presentation describes how regression analysis of the impact of design choices on well productivity can yield highly erroneous estimates if spatial dependence is not controlled for at a sufficiently high resolution. Two regression approaches, the spatial error model and regression-kriging, are advanced as appropriate methods and compared to simpler but widely used regression models with limited spatial fidelity. A case study in which these methods are applied to a large contemporary well dataset from the Williston Basin in North Dakota reveals that only about half of the improvement in well productivity is associated with technology changes, but the simpler regression models substantially overestimate the impact of technology by attributing location-driven improvement to design changes. Overestimating technology’s role in well productivity has important implications for future resource availability and economics, and the development choices of individual operators.
Justin Montgomery is a PhD candidate at MIT in the Computational Science & Engineering (CSE) program and a Research Assistant for the MIT Energy Initiative. His research is focused on using machine learning and statistical modeling approaches to improve the forecasting of unconventional oil and gas well production rates and facilitate more optimal development strategies. Justin has worked in the petroleum industry on projects involving analytics, learning, and optimization in unconventional onshore drilling and completions (hydraulic fracturing) as well as in reservoir engineering and development planning for offshore discoveries. He has been interviewed on Bloomberg Television and been invited to speak about his research at the U.S. Department of Energy and the Center for Strategic & International Studies. Justin holds a B.S. in Mechanical Engineering from Texas A&M University and an S.M. in Technology and Policy from MIT. In his spare time, he enjoys running on the Charles River and singing in a rock band.
For more details please check out Justin's page http://web.mit.edu/jbm/www/
Only SPE members can RSVP at this web site. If you are a non-SPE member please send a confirming email to Jean Elkhoury at firstname.lastname@example.org with SPE-NYNE April 20 in the subject line. The email needs to contain your name, company name and citizenship.
For those who would like to join online, please follow the link below
→Join Skype Meeting
Trouble Joining? Try Skype Web App
Join by phone
+1 844 713 3285 (US) (SLB) English (United States)
Conference ID: 47053549
According to Society of Petroleum Engineers policy, embargoed country nationals will not be admitted to the meeting.