Nikolaou, Michael2018-03-122018-03-12December 22017-12December 2http://hdl.handle.net/10657/2928Hydraulic fracturing is essential for oil and gas production from unconventional low-permeability reservoirs. Countless number of variables contribute towards a successfully hydraulic fractured well. Understandings of these variables and its relationships with well production can generate valuable insight which can be used to build better wells for the future. Estimation of well production performance is extremely important to maintain a positive cash flow for any company. Current methodology to relate these variables with production is very tedious and might require expensive and complex simulation models. With the recent advancement in data capturing technologies and emergence of Internet of Things there exist an inordinate opportunity to use the existing data using sophisticated data driven models. Therefore, a framework for building a data driven solution is provided. The model is capable of taking existing wellbore data, completions data, fracture operation data, historical production data and first month production to predict the decline curves of producing wells. This will help in minimizing risks, promote better investment decisions and provide statically backed decline curves in a very short time.application/pdfengThe author of this work is the copyright owner. UH Libraries and the Texas Digital Library have their permission to store and provide access to this work. Further transmission, reproduction, or presentation of this work is prohibited except with permission of the author(s).Decline curve analysisProduction predictionCompletion parametersPartial least squaresPLSHydraulic fracturesUnconventional wellsMachine learningData scienceData analytics2018-03-12Thesisborn digital