Spe-192002-ms case Study Applied Machine Learning to Optimise pcp completion Design in a cbm field


Data Requirements / SCADA / Completion Data



Download 1.89 Mb.
View original pdf
Page3/8
Date22.08.2023
Size1.89 Mb.
#61902
1   2   3   4   5   6   7   8
spe-192002-ms
Data Requirements / SCADA / Completion Data
The data that exists in atypical production company will require processing in order for it to be used with this framework. Pump and completion information will need to be combined with time series production information so that each production period or pump installation can be represented as a row in the final data set. Care must betaken to ensure that missing data is either not used or imputed with approximate values as this will add more uncertainty to the final output.
Stationary representations of a CBM Well
Describing the dynamic behavior of a well over the course of its producing life is challenging. Practically.
a well over a production period can be described or represented by it average gas and water production or a vector containing the time series production. Ideally, key design parameters such water rate, differential pressure, gas volume fractions and solids volume fractions can be used in selecting an ideal design. In practice however, this data is rarely available. The first section of this framework will be to convert dynamic production data into a surrogate well model that can be used to forecast both water and gas rates. Both the traditional literature and new approach will be discussed.
Material Balance Modeling
The Langmuir volume is the maximum amount of gas that can be adsorbed to coal or shale at infinite pressure. The Langmuir pressure, or critical desorption pressure, is the pressure at which one half of the
Langmuir volume can be adsorbed.
(1)
The Langmuir pressure and volume constants are usually obtained from lab tests. They can vary significantly between wells.
The King material balance method is the most thorough model describing the gas stored both in the fractures as well as the matrix of the coal seam. This model is also analogous to the P/Z method for conventional methods.
(2)
The Jensen and Smith model makes simplifications to equation by assuming that the gas stored in the cleat space (i) is small and therefore negligible. The simplification allows for the solution to be solved more readily.
(3)
The next step is to uncover the change in permeability of the fracture system due to the change in subsurface stresses. A simple method of quantifying the change in the permeability of the system is to
Downloaded from http://onepetro.org/SPEAPOG/proceedings-pdf/18APOG/2-18APOG/D021S016R002/1220497/spe-192002-ms.pdf/1 by Vedanta Limited - Cairn Oil & Gas user on 28 June 2023


4
SPE-192002-MS
observe the change in the inflow performance. Although diffusion is the mechanism through which gas is desorbed from the coal matrix, its direct application is not achievable without the use complex methods such as numerical simulation. A simple inflow relationship (Equation 4
) is used that is derived from pseudo pressure and Darcy's law.
(4)
Though applying the fitted Jensen and Smith tank model to the daily gas rate it is possible to estimate the daily reservoir pressure. Applying the daily reservoir pressure and the daily bottom hole pressure into equation allows the inflow performance to be estimated.
There are two major effects that control the fracture permeability of coal formations. The first being compaction as a result of dewatering and lack of support within the coal formation, this causes a reduction in fracture permeability. The second being matrix shrinkage where the coal shrinks as gas is removed through desorption, this causes an increase in fracture permeability.
The Shi-Durucan model is an accepted geomechanical permeability model for coal formations, it accounts for both major effects of compaction and matrix shrinkage.
(5)
Equation 5
describes how the initial permeability is affected by changes ineffective stress. It is this change ineffective stress that is described by the Shi-Durucan model in equation The Shi-Durucan model is semi empirical as it requires the use of two fitting constants sand b. The model also employs physical material characteristics such as the Young's Modulus and Poisson's ratio. It relates the change in pressure to the change ineffective stress.

Download 1.89 Mb.

Share with your friends:
1   2   3   4   5   6   7   8




The database is protected by copyright ©ininet.org 2024
send message

    Main page