Alfajri, Reza (2015-12). Probabilistic Reserves and Resources Estimation of the West Virginia Marcellus Shale Play Using the MCMC PDCA Method. Master's Thesis. Thesis uri icon

abstract

  • The Marcellus shale is currently the most productive shale play in the United States. In 2015, the Marcellus shale play led in natural gas production per rig and had the highest shale gas production in the United States. Several reports and articles have been published on Marcellus shale play reserves/resources estimates. Some of these estimates were deterministic, while some were probabilistic. These published estimates are all now outdated. Updated probabilistic reserves and resources estimates for the Marcellus shale play are needed. The Marcellus shale play covers six states with the two most productive states being Pennsylvania (PA) and West Virginia (WV). Between these two states, only WV has monthly production data in its production reports; PA production is reported semi-annually. The Markov Chain Monte Carlo (MCMC) method has been successfully used to quantify uncertainty in production forecasts and estimated ultimate recovery (EUR) for the Barnett shale and Eagle Ford shale. There are 20 shale plays that have been discovered in United States. Confirmation of the reliability of the MCMC method using other shale play data is still needed. The objectives of this work are to generate probabilistic reserves and resources estimates for the WV Marcellus shale play and to confirm the reliability of the MCMC method in quantifying uncertainty in production forecasts using production data from the WV Marcellus shale play. Based on geology and initial gas-liquid-ratio (GLR) analysis, the WV Marcellus shale play was divided into liquid-rich and dry-gas regions. A hindcast study was performed to confirm the reliability of the MCMC method in forecasting production and estimating reserves in the WV Marcellus shale play. Type probabilistic decline curves were then generated to forecast Technically Recoverable Resources (TRR) at 20 years (TRR20) for the wells in both the liquid-rich and dry-gas regions. Reserves and resources for the WV Marcellus shale play were estimated by performing Monte Carlo simulation. Based on the WV Marcellus shale play analysis, reserves and resources estimates for PA, Ohio (OH), and entire Marcellus shale play are then extrapolated. Hindcast study results show that the MCMC Probabilistic Decline Curve Analysis (PDCA) method is able to reliably quantify uncertainty in production forecasts and reserves estimates in the WV Marcellus shale play. The total WV NGL reserves and resources range from a P10 of 0.12 billion barrels NGL (BBNGL) to a P90 of 0.58 BBNGL, with a P50 of 0.23 BBNGL. The total WV gas reserves and resources range from a P10 of 81.54 trillion cubic feet (TCF) to a P90 of 283.81 TCF, with a P50 of 145.091 TCF. These estimates are generally much higher than most of the previously published estimates.
  • The Marcellus shale is currently the most productive shale play in the United States. In 2015, the Marcellus shale play led in natural gas production per rig and had the highest shale gas production in the United States. Several reports and articles have been published on Marcellus shale play reserves/resources estimates. Some of these estimates were deterministic, while some were probabilistic. These published estimates are all now outdated. Updated probabilistic reserves and resources estimates for the Marcellus shale play are needed. The Marcellus shale play covers six states with the two most productive states being Pennsylvania (PA) and West Virginia (WV). Between these two states, only WV has monthly production data in its production reports; PA production is reported semi-annually.

    The Markov Chain Monte Carlo (MCMC) method has been successfully used to quantify uncertainty in production forecasts and estimated ultimate recovery (EUR) for the Barnett shale and Eagle Ford shale. There are 20 shale plays that have been discovered in United States. Confirmation of the reliability of the MCMC method using other shale play data is still needed. The objectives of this work are to generate probabilistic reserves and resources estimates for the WV Marcellus shale play and to confirm the reliability of the MCMC method in quantifying uncertainty in production forecasts using production data from the WV Marcellus shale play.

    Based on geology and initial gas-liquid-ratio (GLR) analysis, the WV Marcellus shale play was divided into liquid-rich and dry-gas regions. A hindcast study was performed to confirm the reliability of the MCMC method in forecasting production and estimating reserves in the WV Marcellus shale play. Type probabilistic decline curves were then generated to forecast Technically Recoverable Resources (TRR) at 20 years (TRR20) for the wells in both the liquid-rich and dry-gas regions. Reserves and resources for the WV Marcellus shale play were estimated by performing Monte Carlo simulation. Based on the WV Marcellus shale play analysis, reserves and resources estimates for PA, Ohio (OH), and entire Marcellus shale play are then extrapolated.

    Hindcast study results show that the MCMC Probabilistic Decline Curve Analysis (PDCA) method is able to reliably quantify uncertainty in production forecasts and reserves estimates in the WV Marcellus shale play. The total WV NGL reserves and resources range from a P10 of 0.12 billion barrels NGL (BBNGL) to a P90 of 0.58 BBNGL, with a P50 of 0.23 BBNGL. The total WV gas reserves and resources range from a P10 of 81.54 trillion cubic feet (TCF) to a P90 of 283.81 TCF, with a P50 of 145.091 TCF. These estimates are generally much higher than most of the previously published estimates.

ETD Chair

publication date

  • December 2015