Mapping completion design trends in a compartmentalized tight oil reservoir for rapid evaluation using artificial neural networks

Y. Bansal, T. Ertekin, Z. T. Karpyn

Research output: Contribution to conferencePaperpeer-review

Abstract

Lower oil prices pose a threat to developing tight oil reservoirs economically. Typically, completion cost varies between 50-70% of the total capital costs of drilling unconventional wells (EIA, 2016). Major challenges in justification of drilling these new wells lie with the uncertainty in production, and thus, with the efficiency of drilling and completion design of these wells. Identifying a suitable completion design can be a complex problem as tight oil systems are discontinuous hydrocarbon sources. These resources often present challenges in building good geological models. However, abundant data for production, well logs, seismic, drilling and completions are available through recent developments in these formations. This information can be utilized to characterize these reservoirs and identify case-specific and cost-effective completion strategies. This paper discusses the utilization of the data to map completion trends in tight oil systems to optimally complete these wells. The available data is used to generate trends for completion parameters with the application of artificial neural networks. It is shown that the complex relationship among geological properties, well logs, production history, and completions methodology can be established with this methodology. The suggested completion design may enable to lower the effective cost of production from these wells. In addition, the proposed ANN based proxy takes only a few seconds to suggest completion design for a new well. It is also shown that the proxy model developed with this application can easily be utilized with available optimization algorithms to optimize the completion parameters in a time efficient manner.

Original languageEnglish (US)
DOIs
StatePublished - 2017
EventSPE Abu Dhabi International Petroleum Exhibition and Conference 2017 - Abu Dhabi, United Arab Emirates
Duration: Nov 13 2017Nov 16 2017

Other

OtherSPE Abu Dhabi International Petroleum Exhibition and Conference 2017
CountryUnited Arab Emirates
CityAbu Dhabi
Period11/13/1711/16/17

All Science Journal Classification (ASJC) codes

  • Geochemistry and Petrology

Fingerprint Dive into the research topics of 'Mapping completion design trends in a compartmentalized tight oil reservoir for rapid evaluation using artificial neural networks'. Together they form a unique fingerprint.

Cite this