Type of Document Dissertation Author Ansari, Esmail Author's Email Address email@example.com URN etd-03232016-135113 Title Mathematical Scaling and Statistical Modeling of Geopressured Geothermal Reservoirs Degree Doctor of Philosophy (Ph.D.) Department Petroleum Engineering Advisory Committee
Advisor Name Title Hughes, Richard G Committee Chair Dutrow, Barbara Committee Member Sears, Stephen O Committee Member Tyagi, Mayank Committee Member Bourdin, Blaise Dean's Representative Keywords
- predictive model
- inspectional analysis
- dimensionless numbers
- scaling groups
- geothermal reservoir
- experimental design
- dimensional analysis
- response surface
- Screening model
- statistical model
Date of Defense 2016-02-17 Availability unrestricted AbstractThe interest for developing geopressured-geothermal reservoirs along the US Gulf Coast is increasing for securing energy needs and reducing global warming. Identifying the most attractive candidate reservoirs for geothermal energy production requires quick and simple models. Analytical models are not always available and simulating each case individually is expensive. The use of scaling and statistical modeling is one approach to translate the output of a simulator into quick models with general applicability at all scales. The developed models can quickly estimate temperature and thermal energy recovery from the geopressured-geothermal reservoirs. These models can screen large databases of reservoirs to select the most attractive ones for geothermal energy production.
This study presents two different designs for extracting energy from geopressured-geothermal reservoirs: Regular line drive and Zero Mass Withdrawal (ZMW). First, the governing partial differential equations describing each design are derived from the fundamental equations. Inspectional analysis on the partial differential equations of each design provides the most succinct and meaningful form of the dimensionless numbers for scaling the designs. The dimensionless numbers are tested and verified by selecting models with identical dimensionless numbers but different dimensional parameters.
For creating the response models, statistics is used to find the important dimensionless numbers for predicting the response systematically. A procedure is used to compare all possible models and select the best one. These simplified final models are then presented and the performance of the simplified models is assessed using testing runs. Applications of these models are presented.
To test the response models, two field cases from southern Louisiana are evaluated: the Gueydan Dome reservoir and the Sweet Lake reservoir. The Gueydan Dome reservoir (Vermilion parish, LA) is investigated using an optimization algorithm and it is concluded that the temperature map should be used for pre-development heat extraction assessments. The Sweet Lake reservoir (Cameron Parish, LA) is studied using this conclusion.
Filename Size Approximate Download Time (Hours:Minutes:Seconds)
28.8 Modem 56K Modem ISDN (64 Kb) ISDN (128 Kb) Higher-speed Access Ansari_diss.pdf 7.34 Mb 00:33:59 00:17:28 00:15:17 00:07:38 00:00:39
If you have questions or technical problems, please Contact LSU-ETD Support.