US20140124265A1 - Systems and methods for expert systems for underbalanced drilling operations using bayesian decision networks - Google Patents

Systems and methods for expert systems for underbalanced drilling operations using bayesian decision networks Download PDF

Info

Publication number
US20140124265A1
US20140124265A1 US13/827,746 US201313827746A US2014124265A1 US 20140124265 A1 US20140124265 A1 US 20140124265A1 US 201313827746 A US201313827746 A US 201313827746A US 2014124265 A1 US2014124265 A1 US 2014124265A1
Authority
US
United States
Prior art keywords
recommendations
inputs
drilling
node
receive
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Abandoned
Application number
US13/827,746
Inventor
Abdullah Saleh Hussain Al-Yami
Jerome Schubert
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Saudi Arabian Oil Co
Texas A&M University System
Original Assignee
Saudi Arabian Oil Co
Texas A&M University System
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority to US201261722027P priority Critical
Application filed by Saudi Arabian Oil Co, Texas A&M University System filed Critical Saudi Arabian Oil Co
Priority to US13/827,746 priority patent/US20140124265A1/en
Assigned to THE TEXAS A&M UNIVERSITY SYSTEM reassignment THE TEXAS A&M UNIVERSITY SYSTEM ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: SCHUBERT, JEROME, DR
Publication of US20140124265A1 publication Critical patent/US20140124265A1/en
Assigned to SAUDI ARABIAN OIL COMPANY reassignment SAUDI ARABIAN OIL COMPANY ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: HUSSAIN AL-YAMI, ABDULLAH SALEH
Abandoned legal-status Critical Current

Links

Images

Classifications

    • EFIXED CONSTRUCTIONS
    • E21EARTH DRILLING; MINING
    • E21BEARTH DRILLING, e.g. DEEP DRILLING; OBTAINING OIL, GAS, WATER, SOLUBLE OR MELTABLE MATERIALS OR A SLURRY OF MINERALS FROM WELLS
    • E21B44/00Automatic control systems specially adapted for drilling operations, i.e. self-operating systems which function to carry out or modify a drilling operation without intervention of a human operator, e.g. computer-controlled drilling systems; Systems specially adapted for monitoring a plurality of drilling variables or conditions
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N7/00Computer systems based on specific mathematical models
    • G06N7/005Probabilistic networks

Abstract

Systems and methods are provided for an underbalanced drilling (UBD) expert system that provides underbalanced drilling recommendations, such as best practices. The UBD expert system may include one or more Bayesian decision network (BDN) model that receive inputs and output recommendations based on Bayesian probability determinations. The BDN models may include: a general UBD BDN model, a flow UBD BDN model, a gaseated (i.e., aerated) UBD BDN model, a foam UBD BDN model, a gas (e.g., air or other gases) UBD BDN model, a mud cap UBD BDN model, an underbalanced liner drilling (UBLD) BDN model, an underbalanced coil tube (UBCT) BDN model, and a snubbing and stripping BDN model.

Description

    PRIORITY CLAIM
  • This application claims priority to U.S. Provisional Patent Application No. 61/722,027 filed on Nov. 2, 2012, entitled “Systems and Methods for Expert Systems for Underbalanced Drilling Operations Using Bayesian Decision Networks,” the disclosure of which is hereby incorporated by reference in its entirety.
  • BACKGROUND OF THE INVENTION
  • 1. Field of the Invention
  • This invention relates generally to the drilling and extraction of oil, natural gas, and other resources, and more particularly to evaluation and selection of underbalanced drilling systems.
  • 2. Description of the Related Art
  • Oil, gas, and other natural resources are used for numerous energy and material purposes. The search for extraction of oil, natural gas, and other subterranean resources from the earth may cost significant amounts of time and money. Once a resource is located, drilling systems may be used to access the resources, such as by drilling into various geological formations to access deposits of such resources. The drilling systems rely on numerous components and operational techniques to reduce cost and time and maximize effectiveness. For example, drill strings, drill bits, drilling fluids, and other components may be selected to achieve maximum effectiveness for a formation and other parameters that affect the drilling system. Typically, many years of field experience and laboratory work are used to develop and select the appropriate components and operational practices for a drilling system. However, these techniques may be time-consuming and expensive. Moreover, such techniques may produce inconsistent results and may not incorporate recent changes in practices and opinions regarding the drilling systems.
  • SUMMARY OF THE INVENTION
  • Various embodiments of methods and systems for expert systems for determining underbalanced drilling operations using Bayesian decision networks are provided herein. In some embodiments, a system is provided that includes one or more processors and a non-transitory tangible computer-readable memory. The non-transitory tangible computer-readable memory includes an underbalanced drilling expert system executable by the one or more processors and configured to provide one or more underbalanced drilling recommendations based on one or more inputs. The underbalanced drilling expert system includes an underbalanced drilling Bayesian decision network (BDN) model. The underbalanced drilling BDN model includes a first section having a formation indicators uncertainty node configured to receive one or more formation indicators from the one or more inputs, a formation considerations decision node configured to receive one or more formation considerations from the one or more inputs, and a first consequences node dependent on the formation indicators uncertainty node and the formation considerations decision node and configured to output the one or more underbalanced drilling recommendations based on one or more Bayesian probabilities calculated from one or more formation indicators and the one or more formation considerations. The underbalanced BDN model includes a second section having a planning phases uncertainty node configured to receive one or more planning phases from the one or more inputs, a planning phases recommendations decision node configured to receive one or more planning phases recommendations from the one or more inputs, and a second consequences node dependent on the planning phases uncertainty node and the planning phases recommendations decision node and configured to output the one or more underbalanced drilling recommendations based on one or more Bayesian probabilities calculated from one or more planning phases and the one or more planning phases recommendations. Finally, the underbalanced drilling BDN model also includes a third section having a an equipment requirements uncertainty node configured to receive one or more equipment requirements from the one or more inputs, an equipment recommendations decision node configured to receive one or more equipment recommendations from the one or more inputs, and a third consequences node dependent on the equipment requirements uncertainty node and the equipment recommendations decision node and configured to output the one or more underbalanced drilling recommendations based on one or more Bayesian probabilities calculated from the one or more equipment requirements and the one or more equipment recommendations.
  • In some embodiments, a computer-implemented method for an underbalanced drilling expert system having an underbalanced drilling Bayesian decision network (BDN) model is provided. The method includes receiving one or more inputs and providing the one or more nodes of a first section of the underbalanced drilling BDN model. The one or more nodes include a formation indicators uncertainty node and a formation considerations decision node. Additionally, the method further includes determining one or more underbalanced drilling recommendations at a consequences node of the first section of the underbalanced drilling BDN model, the determination comprising a calculation of one or more Bayesian probabilities based on the one or more inputs and providing the one or more underbalanced drilling recommendations to a user.
  • Additionally, in some embodiments, a system having one or more processors and a non-transitory tangible computer-readable memory is provided. The memory the memory includes an underbalanced drilling expert system executable by the one or more processors and configured to provide one or more underbalanced drilling recommendations based on one or more inputs. The underbalanced drilling expert system includes a flow underbalanced drilling Bayesian decision network (BDN) model. The flow underbalanced drilling BDN model includes a first section having a tripping types uncertainty node configured to receive one or more tripping types from the one or more inputs, a permeability level uncertainty node configured to receive one or more permeability levels from the one or more inputs, a tripping options decision node configured to receive one or more tripping options from the one or more inputs, and a first consequences node dependent on the tripping uncertainty node, the permeability level uncertainty node, and the tripping options decision node and configured to output the one or more flow underbalanced drilling recommendations based on one or more Bayesian probabilities calculated from the one or more tripping types, the one or more permeability levels, and the one or more tripping options. The flow underbalanced drilling BDN model also includes a second section having a connection types uncertainty node configured to receive one or more connection types from the one or more inputs, a connection options decision node configured to receive one or more connection options from the one or more inputs, and a second consequences node dependent on the connection uncertainty node and the connection options decision node and configured to output the one or more flow underbalanced drilling recommendations based on one or more Bayesian probabilities calculated from the one or more connection types and the one or more connection options. Finally, the foam underbalanced drilling BDN model includes a third section having a flow drilling types uncertainty node configured to receive one or more flow drilling types from the one or more inputs, a flow drilling options decision node configured to receive one or more flow drilling options from the one or more inputs, and a third consequences node dependent on the flow drilling uncertainty node and the flow drilling options decision node and configured to output the one or more flow underbalanced drilling recommendations based on one or more Bayesian probabilities calculated from the one or more flow drilling types and the one or more flow drilling options.
  • Further, in some embodiments a computer-implemented method for an underbalanced drilling expert system having a flow underbalanced drilling (UBD) Bayesian decision network (BDN) model is provided. The method includes receiving one or more inputs and providing the one or more inputs to one or more nodes of a first section of the flow underbalanced drilling BDN model. The one or more nodes include a tripping uncertainty node configured to receive one or more tripping types, a permeability level uncertainty node configured to receive one or more permeability levels, and a tripping options decision node a tripping options decision node configured to receive one or more tripping options. The method further includes determining one or more underbalanced drilling recommendations at a consequences node of the first section of the flow underbalanced drilling BDN model by calculating one or more Bayesian probabilities based on the one or more inputs and providing the one or more underbalanced drilling recommendations to a user.
  • Additionally, in some embodiments a system is provided having one or more processors and a non-transitory tangible computer-readable memory accessible by the one or more processors. The memory includes an underbalanced drilling expert system executable by the one or more processors and configured to provide one or more underbalanced drilling recommendations based on one or more inputs. The underbalanced drilling expert system includes a gaseated underbalanced drilling Bayesian decision network (BDN) model. The gaseated underbalanced drilling BDN model includes a first section having a gas injection process uncertainty node configured to receive one or more gas injection process types from the one or more inputs, a gas injection processes characteristics decision node configured to receive one or more gas injection process characteristics from the one or more inputs, and a first consequences node dependent on the gas injection process uncertainty node and the gas infection processes considerations decision node and configured to output the one or more gaseated underbalanced drilling recommendations based on one or more Bayesian probabilities calculated from the one or more gas injection process types and the one or more gas injection process characteristics. The gaseated underbalanced drilling BDN model also includes a second section having a fluid volume limits uncertainty node configured to receive one or more fluid volume limits from the one or more inputs, a fluid volume limits requirements decision node configured to receive one or more fluid volume limits requirements from the one or more inputs, and a second consequences node dependent on the fluid volume limits uncertainty node and the fluid volume limits requirements decision node and configured to output the one or more gaseated underbalanced drilling recommendations based on one or more Bayesian probabilities calculated from the one or more fluid volume limits and the one or more fluid volume limits requirements. Additionally, the gaseated underbalanced drilling BDN model includes a third section having a kick type uncertainty node configured to receive one or more kick types from the one or more inputs, a kicks recommendations decision node configured to receive one or more kicks recommendations from the one or more inputs, and a third consequences node dependent on the kick type uncertainty node and the kicks recommendations decision node and configured to output the one or more gaseated underbalanced drilling recommendations based on one or more Bayesian probabilities calculated from the one or more kick types and the one or more kicks recommendations. Finally, the gaseated underbalanced drilling BDN model includes a fourth section having an operational considerations uncertainty node configured to receive one or more operational considerations from the one or more inputs, an operational recommendations decision node configured to receive one or more operational recommendations from the one or more inputs, and a fourth consequences node dependent on the operational considerations uncertainty node and the operational recommendations decision node and configured to output the one or more gaseated underbalanced drilling recommendations based on one or more Bayesian probabilities calculated from the one or more operational recommendations and the one or more operational recommendations.
  • In some embodiments, a computer-implemented method for an underbalanced drilling expert system having a gaseated underbalanced drilling Bayesian decision network (BDN) model is provided. The method includes receiving one or more inputs and providing the one or more inputs to one or more nodes of a first section of the gaseated underbalanced drilling (UBD) BDN model. The one or more nodes include a gas injection process uncertainty node configured to receive one or more gas injection process types, and a gas injection processes characteristics decision node configured to receive one or more gas injection process characteristics. Additionally, the method includes determining one or more underbalanced drilling recommendations at a consequences node of the first section of the gaseated UBD BDN model by calculating one or more Bayesian probabilities based on the one or more inputs and providing the one or more underbalanced drilling recommendations to a user.
  • Moreover, in some embodiments a system is provided having one or more processors and a non-transitory tangible computer-readable memory accessible by the one or more processors. The memory includes an underbalanced drilling expert system executable by the one or more processors and configured to provide one or more underbalanced drilling recommendations based on one or more inputs. The underbalanced drilling expert system includes a foam underbalanced drilling (UBD) Bayesian decision network (BDN) model. The foam underbalanced drilling BDN model includes a first section having a foam systems considerations uncertainty node configured to receive one or more foam systems considerations from the one or more inputs, a foam systems recommendations decision node configured to receive one or more foam systems recommendations from the one or more inputs and a first consequences node dependent on the foam systems considerations uncertainty node and the foam systems recommendations decision node and configured to output the one or more underbalanced drilling recommendations based on one or more Bayesian probabilities calculated from the one or more foam systems considerations and the one or more foam systems recommendations. The foam underbalanced drilling BDN model also includes a second section having a foam systems designs uncertainty node configured to receive one or more foam system designs from the one or more inputs, a foam system designs recommendations decision node configured to receive one or more foam system designs recommendations from the one or more inputs, and a second consequences node dependent on the foam systems designs uncertainty node and the foam system designs recommendations decision node and configured to output the one or more underbalanced drilling recommendations based on one or more Bayesian probabilities calculated from the one or more foam system designs and the one or more foam system designs recommendations.
  • In some embodiments, a computer-implemented method for an underbalanced drilling expert system having a foam underbalanced drilling (UBD) Bayesian decision network (BDN) model is provided. The method includes receiving one or more inputs and providing the one or more inputs to one or more nodes of a first section of the foam UBD BDN model. The one or more nodes include a foam systems considerations uncertainty node configured to receive one or more foam systems considerations and a foam systems recommendations decision node configured to receive one or more foam systems recommendations. Additionally, the method includes determining one or more underbalanced drilling recommendations at a consequences node of the first section of the foam UBD BDN model, by calculating one or more Bayesian probabilities based on the one or more inputs and providing the one or more underbalanced drilling recommendations to a user.
  • Additionally, in some embodiments a system is provided having one or more processors and a non-transitory tangible computer-readable memory accessible by the one or more processors. The memory includes an underbalanced drilling expert system executable by the one or more processors and configured to provide one or more underbalanced drilling recommendations based on one or more inputs. The underbalanced drilling expert system includes a gas underbalanced drilling (UBD) Bayesian decision network (BDN) model. The gas underbalanced drilling BDN model includes a first section having a rotary and hammer drilling uncertainty node configured to receive one or more rotary and hammer drilling types from the one or more inputs, a rotary and hammer drilling recommendations decision node configured to receive one or more rotary and hammer drilling recommendations from the one or more inputs, and a first consequences node dependent on the rotary and hammer drilling uncertainty node and the rotary and hammer drilling recommendations decision node and configured to output the one or more air and gas underbalanced drilling recommendations based on one or more Bayesian probabilities calculated from the one or more rotary and hammer drilling types and the one or more rotary and hammer drilling recommendations. The gas underbalanced drilling BDN model includes a second section a gas drilling considerations uncertainty node configured to receive one or more gas drilling considerations from the one or more inputs, a gas drilling considerations recommendations decision node configured to receive gas drilling considerations recommendations from the one or more inputs, and a second consequences node dependent on the gas drilling considerations uncertainty node and the gas drilling considerations recommendations decision node and configured to output the one or more air and gas underbalanced drilling recommendations based on one or more Bayesian probabilities calculated from the one or more gas drilling considerations and the one or more gas drilling considerations recommendations. Additionally, the gas underbalanced drilling BDN model includes a third section having a gas drilling operations uncertainty node configured to receive one or more gas drilling operations from the one or more inputs, a gas drilling operations recommendations decision node configured to receive gas drilling operations recommendations from the one or more inputs, and a third consequences node dependent on the gas drilling operations uncertainty node and the gas drilling operations recommendations decision node and configured to output the one or more air and gas underbalanced drilling recommendations based on one or more Bayesian probabilities calculated from the one or more gas drilling operations and the one or more gas drilling operations recommendations. Finally, the gas underbalanced drilling BDN model includes a fourth section having a gas drilling rig equipment uncertainty node configured to receive one or more gas drilling rig equipment from the one or more inputs, a gas drilling rig equipment recommendations decision node configured to receive gas drilling rig equipment recommendations from the one or more inputs, and a fourth consequences node dependent on the gas drilling rig equipment uncertainty node and the gas drilling rig equipment recommendations decision node and configured to output the one or more air and gas underbalanced drilling recommendations based on one or more Bayesian probabilities calculated from the one or more gas drilling rig equipment and the one or more gas drilling rig equipment recommendations.
  • Further, in some embodiments a computer-implemented method is provided for an underbalanced drilling expert system having a gas underbalanced drilling (UBD) Bayesian decision network (BDN) model. The method includes receiving one or more inputs and providing the one or more inputs to one or more nodes of a first section of the gas underbalanced drilling BDN model. The one or more nodes include a rotary and hammer drilling uncertainty node and a rotary and hammer recommendations decision node. Additionally, the method includes determining one or more underbalanced drilling recommendations at a consequences node of the first section of the gas underbalanced drilling BDN model, by calculating one or more Bayesian probabilities based on the one or more inputs and providing the one or more underbalanced drilling recommendations to a user.
  • In some embodiments, a system is provided having one or more processors and a non-transitory tangible computer-readable memory accessible by the one or more processors. The memory includes an underbalanced drilling expert system executable by the one or more processors and configured to provide one or more underbalanced drilling recommendations based on one or more inputs. The underbalanced drilling expert system includes a mud cap underbalanced drilling (UBD) Bayesian decision network (BDN) model. The mud cap underbalanced drilling BDN model includes a first section having a mud cap drilling types uncertainty node configured to receive one or more mud cap drilling types from the one or more inputs, a mud cap drilling types recommendations decision node configured to receive one or more mud cap drilling types recommendations from the one or more inputs, and a first consequences node dependent on the mud cap drilling types uncertainty node and the mud cap drilling types recommendations decision node and configured to output the one or more underbalanced drilling recommendations based on one or more Bayesian probabilities calculated from the one or more mud cap drilling types and the one or more mud cap drilling types recommendations.
  • In some embodiments a computer-implemented method is provided for an underbalanced drilling expert system having a mud cap underbalanced drilling (UBD) Bayesian decision network (BDN) model. The method includes receiving one or more inputs and providing the one or more inputs to one or more nodes of a first section of the mud cap UBD BDN model. The one or more nodes include mud cap drilling types uncertainty node configured to receive one or more mud cap drilling types a mud cap drilling types recommendations decision node configured to receive one or more mud cap drilling types recommendations. The method further includes determining one or more underbalanced drilling recommendations at a consequences node of the first section of the mud cap UBD BDN model by calculating one or more Bayesian probabilities based on the one or more inputs and providing the one or more underbalanced drilling recommendations to a user.
  • Moreover, in some embodiments a system is provided having one or more processors and a non-transitory tangible computer-readable memory accessible by the one or more processors. The memory includes an underbalanced drilling expert system executable by the one or more processors and configured to provide one or more underbalanced drilling recommendations based on one or more inputs. The underbalanced drilling expert system includes an underbalanced liner drilling (UBLD) Bayesian decision network (BDN) model. The UBLD BDN model includes a first section having a UBLD plans uncertainty node configured to receive one or more UBLD plans from the one or more inputs, a UBLD plans recommendations decision node configured to receive one or more UBLD plans recommendations from the one or more inputs, and a first consequences node dependent on the UBLD planning uncertainty node and the UBLD planning recommendations decision node and configured to output the one or more underbalanced liner drilling recommendations based on one or more Bayesian probabilities calculated from the one or more UBLD plans and the one or more UBLD plans recommendations. The UBLD BDN model also includes a second section having a UBLD solvable problems uncertainty node configured to receive one or more UBLD solvable problems from the one or more inputs, a UBLD advantages decision node configured to receive one or more UBLD advantages from the one or more inputs, and a second consequences node dependent on the UBLD problems uncertainty node and the UBLD advantages decision node and configured to output the one or more underbalanced drilling recommendations based on one or more Bayesian probabilities calculated from the one or more UBLD solvable problems and the one or more UBLD advantages. Additionally, the UBLD BDN model includes a third section having a UBLD considerations uncertainty node configured to receive one or more UBLD considerations from the one or more inputs, a UBLD considerations recommendations decision node configured to receive one or more UBLD considerations recommendations from the one or more inputs, and a third consequences node dependent on the UBLD considerations uncertainty node and the UBLD recommendations decision node and configured to output the one or more underbalanced liner drilling recommendations based on one or more Bayesian probabilities calculated from the one or more UBLD considerations and the one or more UBLD considerations recommendations.
  • In some embodiments a computer-implemented method is provided for an underbalanced drilling expert system having an underbalanced drilling liner (UBLD) Bayesian decision network (BDN) model. The method includes receiving one or more inputs and providing the one or more inputs to one or more nodes of a first section of the UBLD BDN model. The one or more nodes include a UBLD plans uncertainty node configured to receive one or more UBLD plans and a UBLD plans recommendations decision node configured to receive one or more UBLD plans recommendations. The method also includes determining one or more underbalanced drilling recommendations at a consequences node of the first section of the UBLD BDN model by calculating one or more Bayesian probabilities based on the one or more inputs and providing the one or more underbalanced drilling recommendations to a user.
  • Moreover, in some embodiments a system is provided having one or more processors and a non-transitory tangible computer-readable memory accessible by the one or more processors. The memory includes an underbalanced drilling expert system executable by the one or more processors and configured to provide one or more underbalanced drilling recommendations based on one or more inputs. The underbalanced drilling expert system includes an underbalanced coil tube (UBCT) Bayesian decision network (BDN) model. The UBCT BDN model includes a first section having a UBCT preplanning uncertainty node configured to receive one or more UBCT preplans from the one or more inputs, a UBCT preplanning requirements decision node configured to receive one or more UBCT preplan requirements from the one or more inputs, a first consequences node dependent on the UBCT preplanning uncertainty node and the UBCT preplanning recommendations decision node and configured to output the one or more UBCT drilling requirements based on one or more Bayesian probabilities calculated from the one or more UBCT preplans and the one or more UBCT preplan requirements. The UBCT BDN model also includes a second section having a UBCT considerations uncertainty node configured to receive one or more UBCT considerations from the one or more inputs, a UBCT recommendations decision node configured to receive one or more UBCT recommendations from the one or more inputs, and a second consequences node dependent on the UBCT considerations uncertainty node and the UBCT recommendations decision node and configured to output the one or more underbalanced drilling recommendations based on one or more Bayesian probabilities calculated from the one or more UBCT considerations and the one or more UBCT recommendations.
  • In some embodiments a computer-implemented method is provided for an underbalanced drilling expert system having an underbalanced coil tube (UBCT) Bayesian decision network (BDN) model. The method includes receiving one or more inputs and providing the one or more inputs to one or more nodes of a first section of the UBCT BDN model. The one or more nodes include a UBCT preplanning uncertainty node configured to receive one or more UBCT preplans and a UBCT preplanning requirements decision node configured to receive one or more UBCT preplan requirements. The method also includes determining one or more underbalanced drilling recommendations at a consequences node of the first section of the UBCT BDN model, the determination comprising a calculation of one or more Bayesian probabilities based on the one or more inputs and providing the one or more underbalanced drilling recommendations to a user.
  • Further, in some embodiments another system is provided having one or more processors and a non-transitory tangible computer-readable memory accessible by the one or more processors. The memory includes an underbalanced drilling expert system executable by the one or more processors and configured to provide one or more underbalanced drilling recommendations based on one or more inputs. The underbalanced drilling expert system includes a snubbing and stripping Bayesian decision network (BDN) model. The snubbing and stripping BDN model includes a first section having a snubbing types uncertainty node configured to receive one or more snubbing types from the one or more inputs and a snubbing types recommendations decision node configured to receive one or more snubbing types recommendations from the one or more inputs, and a first consequences node dependent on the snubbing types uncertainty node and the snubbing types recommendations decision node and configured to output the one or more underbalanced recommendations based on one or more Bayesian probabilities calculated from the one or more snubbing types and the one or more snubbing types recommendations. The snubbing and stripping BDN model also includes a second section having a snubbing units uncertainty node configured to receive one or more snubbing units from the one or more inputs, a snubbing units recommendations decision node configured to receive one or more snubbing units recommendations from the one or more inputs, and a second consequences node dependent on the snubbing units uncertainty node and the snubbing units recommendations decision node and configured to output the one or more stripping and snubbing recommendations based on one or more Bayesian probabilities calculated from the one or more snubbing units types and the one or more snubbing units recommendations. Additionally, the snubbing and stripping BDN model includes a third section having a snubbing operations uncertainty node configured to receive one or more snubbing operations from the one or more inputs, a snubbing operations recommendations decision node configured to receive one or more snubbing operations recommendations from the one or more inputs, and a third consequences node dependent on the snubbing operations uncertainty node and the snubbing operations recommendations decision node and configured to output the one or more stripping and snubbing recommendations based on one or more Bayesian probabilities calculated from the one or more snubbing operations and the one or more snubbing operations recommendations. Finally, the snubbing and stripping BDN model also includes a fourth section having a stripping procedures uncertainty node configured to receive one or more stripping procedures from the one or more inputs, a stripping procedures recommendations decision node configured to receive one or more stripping procedures recommendations from the one or more inputs, and a fourth consequences node dependent on the stripping procedures uncertainty node and the stripping procedures recommendations decision node and configured to output the one or more stripping and snubbing recommendations based on one or more Bayesian probabilities calculated from the one or more stripping procedures and the one or more stripping procedures recommendations.
  • Finally, in some embodiments another computer-implemented method is provided for an underbalanced drilling expert system having a snubbing and stripping Bayesian decision network (BDN) model. The method includes receiving one or more inputs and providing the one or more inputs to one or more nodes of a first section of the snubbing and stripping BDN model. The one or more nodes include snubbing types uncertainty node configured to receive one or more snubbing types and a snubbing types recommendations decision node configured to receive one or more snubbing types recommendations. The method also includes determining one or more underbalanced drilling recommendations at a consequences node of the first section of the snubbing and stripping BDN model, the determination comprising a calculation of one or more Bayesian probabilities based on the one or more inputs and providing the one or more underbalanced drilling recommendations to a user.
  • BRIEF DESCRIPTION OF THE DRAWINGS
  • FIG. 1 is a block diagram that illustrates a system in accordance with an embodiment of the present invention;
  • FIG. 2 is a schematic diagram of a computer and an underbalanced drilling expert system in accordance with an embodiment of the present invention;
  • FIGS. 3A-3I are block diagrams of processes of an underbalanced drilling expert system in accordance with an embodiment of the present invention;
  • FIG. 4 is a schematic diagram of an example of a Bayesian decision network model for the selection of a swelling packer in accordance with an embodiment of the present invention;
  • FIGS. 5-8 are tables of the probability states associated with the nodes of the Bayesian decision network model of FIG. 4;
  • FIG. 9 is a table of input utility values assigned to a consequences node of the Bayesian decision network model of FIG. 4;
  • FIG. 10 is a table of total probability calculations for drilling fluid types of the Bayesian decision network model of FIG. 4;
  • FIG. 11 is a table of Bayesian probability determinations for the Bayesian decision network model of FIG. 4;
  • FIG. 12 is a table of consequences based on the Bayesian probability determinations depicted in FIG. 11;
  • FIG. 13 is a table of expected utilities based on the consequences depicted in FIG. 12;
  • FIG. 14 is a table of consequences based on the probability states depicted in FIG. 8;
  • FIG. 15 is a table of expected utilities based on the consequences depicted in FIG. 14;
  • FIGS. 16A-16I are schematic diagrams that depict a general UBD BDN model and inputs to the general UBD BDN model in accordance with an embodiment of the present invention;
  • FIG. 17 is a schematic diagram that depicts a selected input to the general UBD BDN model of FIG. 16A;
  • FIG. 18 is a table that depicts the output from the general UBD BDN model of FIG. 16A;
  • FIGS. 19A-19H are schematic diagrams that depict a flow UBD BDN model and inputs to the flow UBD BDN model in accordance with an embodiment of the present invention;
  • FIGS. 20A and 20B are schematic diagrams that depict selected inputs to the flow UBD BDN model of FIG. 19A;
  • FIG. 21. Is a table that depicts the output from the flow UBD BDN model of FIG. 19A;
  • FIGS. 22A-22I are schematic diagrams that depict a gaseated UBD BDN model and inputs to the gaseated UBD BDN model in accordance with an embodiment of the present invention;
  • FIGS. 23A and 23B are schematic diagrams that depict a selected input to and an output from the gaseated UBD BDN model of FIG. 22A;
  • FIGS. 24A-24E are schematic diagrams that depict a foam UBD BDN model and inputs to the foam UBD BDN model in accordance with an embodiment of the present invention;
  • FIGS. 25A and 25B are schematic diagrams that depict a selected input to and an output from the foam UBD BDN model of FIG. 24A;
  • FIGS. 26A-26I are schematic diagrams that depict an air and gas UBD BDN model and inputs to the air and gas UBD BDN model in accordance with an embodiment of the present invention;
  • FIGS. 27A and 27B are schematic diagrams that depict a selected input to and an output from the air and gas UBD BDN model of FIG. 26A;
  • FIGS. 28A and 28B are schematic diagrams that depict another selected input to and an output from the air and gas UBD BDN model of FIG. 26A;
  • FIGS. 29A-29G are schematic diagrams that depict a mud cap UBD BDN model and inputs to the mud cap UBD BDN model in accordance with an embodiment of the present invention;
  • FIGS. 30A and 30B are schematic diagrams that depict a selected input to and an output from the mud cap UBD BDN model of FIG. 29A;
  • FIGS. 31A and 31B are schematic diagrams that depict another selected input to and an output from the mud cap UBD BDN model of FIG. 29A;
  • FIGS. 32A-32G are schematic diagrams that depict a UBLD BDN model and inputs to the UBLD BDN model in accordance with an embodiment of the present invention;
  • FIGS. 33A and 33B are schematic diagrams that depict a selected input to and an output from the UBLD BDN model of FIG. 32A;
  • FIGS. 34A and 34B are schematic diagrams that depict another selected input to and an output from the UBLD BDN model of FIG. 32A;
  • FIGS. 35A-35E are schematic diagrams that depict a UBCTD BDN model and inputs to the UBCTD BDN model in accordance with an embodiment of the present invention;
  • FIGS. 36A and 36B are schematic diagrams that depict a selected input to and an output from the UBCTD BDN model of FIG. 35A;
  • FIGS. 37A-37I are schematic diagrams that depict a snubbing and stripping BDN model and inputs to the snubbing and stripping BDN model in accordance with an embodiment of the present invention;
  • FIGS. 38A and 38B are schematic diagrams that depict a selected input to and an output from the snubbing and stripping BDN model of FIG. 37A;
  • FIGS. 39A and 39B are schematic diagrams that depict a selected input to and an output from the snubbing and stripping BDN model of FIG. 37A;
  • FIG. 40 is a block diagram that depicts a process for constructing a BDN model in accordance with an embodiment of the present invention; and
  • FIG. 41 is a block diagram of a computer in accordance with an embodiment of the present invention.
  • While the invention is susceptible to various modifications and alternative forms, specific embodiments thereof are shown by way of example in the drawings and will herein be described in detail. The drawings may not be to scale. It should be understood, however, that the drawings and detailed description thereto are not intended to limit the invention to the particular form disclosed, but to the contrary, the intention is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the present invention as defined by the appended claims.
  • DETAILED DESCRIPTION
  • As discussed in more detail below, provided in some embodiments are systems, methods, and computer-readable media for an underbalanced drilling (UBD) expert system based on Bayesian decision network (BDN) models. In some embodiments, the UBD expert system includes a user interface and incorporates probability data based on expert opinions. The UBD expert system may include multiple BDN models, such as a general UBD model, a flow UBD drilling model, a gaseated (i.e., aerated) UBD drilling model, a foam UBD model, a gas (e.g., air or other gases) UBD model, a mud cap UBD model, an underbalanced liner drilling (UBLD) model, an underbalanced coil tube (UBCT) drilling model, and a snubbing and stripping model. Each model may include multiple sections and may receive inputs and provide outputs, such as recommendations, based on the inputs. The inputs to an uncertainty node of a BDN model may include probabilities associated with each input, or a user may select a specific input for the uncertainty node. Based on these inputs, and the inputs to a decision node, a model may put recommendations from a consequences node.
  • FIG. 1 is a block diagram that illustrates a system 100 in accordance with an embodiment of the present invention. The system 100 includes a formation 102, a well 104, and an underbalanced drilling (UBD) system 106. The system 100 also includes an underbalanced drilling expert system 108 for use with the underbalanced drilling system 106. As described further below, the underbalanced drilling expert system 108 may be implemented on a computer and may include one or more Bayesian decision networks to evaluate inputs and output recommended UBD operations for use with the underbalanced drilling system 106. As will be appreciated, the well 104 may be formed on the formation 102 to provide for extraction of various resources, such as hydrocarbons (e.g., oil and/or natural gas), from the formation 102. In some embodiments, the well 104 is land-based (e.g., a surface system) or subsea (e.g., a subsea system).
  • The underbalanced drilling system 106 may develop the well 104 by drilling a hole into the formation 102 using a drill bit, e.g., a roller cone bits, drag bits, etc. The underbalanced drilling system 106 may generally include, for example, a wellhead, pipes, bodies, valves, seals and so on that enable drilling of the well 104, provide for regulating pressure in the well 16, and provide for the injection of chemicals into the well 104. As used herein, the term underbalanced drilling refers to a drilling operation in which the wellbore pressure is purposely maintained at a lower pressure than the fluid pressure in the formation 102. Accordingly, the UBD drilling system 106 may include, for example, dry air systems, mist systems, aerated mud systems, gaseated systems, foam systems (e.g., stable foam systems) and other suitable systems. During operation, various UBD-specific scenarios may occur that require adjustments to different parameters of the UDB drilling system 106, such as different equipment, different operations, different tripping, different flow, different connections, different gas injections, different gas and fluid volumes, well kicks, different foams, different air and gas systems, different mud caps, different underbalanced liners, different underbalanced coil tubes, and snubbing and stripping. In some embodiments, the well 104, underbalanced drilling system 106 and other components may include sensors, such as temperature sensors, pressure sensors, and the like, to monitor the drilling process and enable a user to gather information about well conditions.
  • The underbalanced drilling system 106, well 104, and formation 102 may provide a basis for various inputs 112 to the underbalanced drilling expert system 108. For example, as described below, temperature ranges, the formation 102, and potential hole problems may be provided as inputs 112 to the underbalanced drilling expert system 108. The underbalanced drilling expert system 108 may access an expert data repository 114 that includes expert data, such as probability data used by the underbalanced drilling expert system 108. The expert data may be derived from best practices, expert opinions, research papers, and the like. As described further below, based on the inputs 112, the underbalanced drilling expert system 108 may output recommendations for the underbalanced drilling system 106. For example, the underbalanced drilling expert system 108 may provide the optimal equipment, UBD operations, tripping, connections, flow drilling operations, gas injection processes, air and gas operations, and so on as described further below. Based on these recommendations, different practices may be selected and used in the UBD drilling system 106
  • FIG. 2 depicts a computer 200 implementing an underbalanced drilling expert system 202 in accordance with an embodiment of the present invention. As shown in FIG. 2, a user 204 may interact with the computer 200 and the underbalanced drilling expert system 202. In some embodiments, as shown in FIG. 2, the underbalanced drilling expert system 202 may be implemented in a single computer 200. However, in other embodiments, the underbalanced drilling expert system 202 may be implemented on multiple computers in communication with each other over a network. Such embodiments may include, for example, a client/server arrangement of computer, a peer-to-peer arrangement of computers, or any other suitable arrangement that enables execution of the underbalanced drilling expert system 202. In some embodiments, the underbalanced drilling expert system 202 may implemented as a computer program stored on a memory of the computer 200 and executed by a process of the computer 200.
  • In some embodiments, the underbalanced drilling expert system 202 may include a user interface 206 and an expert data repository 208. The user interface 206 may be implemented using any suitable elements, such as windows, menus, buttons, web pages, and so on. As described in detail below, the underbalanced drilling expert system 202 may include one or more Bayesian decision network (BDN) models 210 that implemented Bayesian probability logic 212. The BDN models 210 may evaluate selections of inputs and associated probabilities 214 and output a decision 216 from the BDN model. In the embodiments described herein, the BDN model 210 may include nine different BDN models related to UDB drilling: a general approach to UBD model, a flow UBD drilling model, a gaseated (i.e., aerated) UBD drilling model, a foam UBD model, a gas (e.g., air or other gases) UBD model, a mud cap UBD model, an underbalanced liner drilling (UBLD) model, an underbalanced coil tube (UBCT) drilling model, and a snubbing and stripping model. Each model may include multiple sections and is described in further detail below. The UBD expert system 202 may include any one or combination of the models mentioned above. The BDN models 210 may then calculate Bayesian probabilities for the consequences resulting from the selected inputs, and then output recommended operations. For each BDN model, the output may include a table of probabilities for various recommendations, a single recommendation based on the highest Bayesian probability, or expected utility values for each BDN model to enable to user to evaluate and select the operation having the optimal expected utility for the selected inputs.
  • As described below, a user 204 may use the user interface 206 to enter selections 210 of inputs for the BDN model 210. The associated probabilities for the inputs may be obtained from the expert data repository 208. Based on the inputs 210, a user 204 may receive the outputs 212 from the BDN model 210, such as recommended UBD operations and expected utility values. The output 212 may be provided for viewing in the user interface 206. Further, as explained below, a user may return to the underbalanced drilling expert system 202 to add or change the inputs 214. The BDN model 210 may recalculate the outputs 216 based on the added or changed inputs 214 and the Bayesian probability logic 212. The recalculated outputs 216 may then provide additional or changed recommended underbalanced drilling practices and expected utility values. Here again, the outputs 216 may be provided to the user in the user interface 206. The user 204 may use a single BDN model of the UBD expert system 202, or may use multiple models of the UBD expert system 202, such as two, three, four, five, six, seven, eight, or nine models of the UDB expert system 202.
  • FIGS. 3A-3I each depict a process corresponding to a BDN model that may be implemented in a UBD expert system in accordance with an embodiment of the present invention. As explained below, a UBD expert system may include any one or combination of the BDN models described below, and thus may executed any one or combination of the processes described in FIGS. 3A-3I. FIG. 3A depicts a process 300 of the operation of a general UBD BDN model of a UBD expert system in accordance with an embodiment of the present invention. Initially, a user interface for an underbalanced drilling expert system may be provided to a user (block 302). From the user interface, various selections of inputs may be received. For example, selections of formation indicators may be received (block 304) by the underbalanced drilling expert system. As explained below, a user may enter a selection of one or more possible formation indicators into the underbalanced drilling expert system. Additionally, selections of UBD planning phases may be received (block 306) by the underbalanced drilling expert system. As explained below, inputs may be provided at any node of a BDN model of the underbalanced drilling expert system. Additionally, in some embodiments, equipment requirements may also be selected and received by the underbalanced drilling expert system (block 308). Finally planned operations a UBD system may be selected and received by the model (block 310). As mentioned above, any one of or combination of these selections may be received. As described below, the BDN model enables a user to enter inputs at any node of the BDN model.
  • Next, the received selections may be provided as inputs to uncertainty nodes of a general UBD BDN model of the UBD expert system (block 310), and the selected inputs may include associated probability states, as determined from expert data 312. Next, the data from the uncertainly nodes may be combined (i.e., propagated to) a consequence node of the general UBD BDN model based on the expert systems data (block 312). The propagation and determination of consequences is based on the Bayesian logic described below in FIGS. 4-15 and implemented in the UBD BDN model described below and illustrated in FIGS. 16A-161. Next, general recommendations and expected utility values may be calculated by the general UBD BDN model (block 316). Finally, the recommendations and expected utility values may be output in a user interface of the UBD expert system (block 318).
  • FIG. 3B depicts a process 312 of the operation of flow UBD BDN model of an UBD expert system in accordance with an embodiment of the present invention. The process 312 illustrates inputs and flow recommendations of the underbalanced drilling expert system, as illustrated further below. Initially, a user interface for an underbalanced drilling expert system may be provided to a user (block 313). From the user interface, various selections of inputs may be received. For example, selections of tripping types may be received (block 314) by the underbalanced drilling expert system. As explained below, a user may enter a selection of one or more possible tripping types into the underbalanced drilling expert system. Additionally, selections of connection types may be received (block 315) by the underbalanced drilling expert system. Finally, in some embodiments, flow drilling types may also be selected by a user and received by the underbalanced drilling expert system (block 316). As explained above, any one of or combination of the selections described above may be input by a user and received by the UBD expert system.
  • Next, the received selections may be provided as inputs to uncertainty nodes of a flow UBD BDN model of the UBD expert system (block 317), and the selected inputs may include associated probability states, as determined from expert data 318. Next, the data from the uncertainly nodes may be combined (i.e., propagated to) a consequence node of the flow UBD BDN model based on the expert systems data (block 319), as based on the Bayesian logic described below in FIGS. 4-15 and implemented in the flow UBD BDN model described below and illustrated in FIGS. 19A-19H. Next, recommendations and expected utility values may be calculated by the BDN model (block 320). Finally, the recommendations and expected utility values may be output in a user interface of the UBD expert system (block 321).
  • FIG. 3C depicts a process 324 of the operation of another model of an underbalanced drilling expert system in accordance with an embodiment of the present invention. The process 324 illustrates a gaseated (i.e., aerated) UBD BDN model of the underbalanced drilling expert system, as illustrated further below. Again, a user interface for an underbalanced drilling expert system may be provided to a user (block 325). From the user interface, various selections of inputs may be received. For example, selections of a gas injection process may be received (block 326) by the underbalanced drilling expert system. As explained below, a user may enter a selection of one or more gas injection processes into the underbalanced drilling expert system. Additionally, selections of gas and fluid volume limits may be received (block 327) by the underbalanced drilling expert system. In some embodiments, kick types of a UBD system may also be selected by a user and received by the underbalanced drilling expert system (block 328). Finally, a user selection of operational considerations of a gaseated UBD system may also be received by the underbalanced drilling expert system (block 329). Any one of or combination of these selections may be received, as the gaseated UBD BDN model enables a user to enter inputs at any node of the BDN model.
  • Next, the received selections may be provided as inputs to uncertainty nodes of a gaseated UBD BDN model of the underbalanced drilling expert system (block 330), and the selected inputs may i